v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
#![cfg(not(debug_assertions))]
|
|
|
|
|
2021-07-09 06:15:32 +00:00
|
|
|
use std::fmt;
|
|
|
|
use std::sync::Mutex;
|
2021-12-13 20:43:22 +00:00
|
|
|
use std::time::Duration;
|
2021-07-09 06:15:32 +00:00
|
|
|
|
|
|
|
use beacon_chain::test_utils::{
|
2023-02-14 12:33:38 +00:00
|
|
|
AttestationStrategy, BeaconChainHarness, BlockStrategy, EphemeralHarnessType,
|
2021-07-09 06:15:32 +00:00
|
|
|
};
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
use beacon_chain::{
|
2020-10-01 01:41:58 +00:00
|
|
|
BeaconChain, BeaconChainError, BeaconForkChoiceStore, ChainConfig, ForkChoiceError,
|
Use the forwards iterator more often (#2376)
## Issue Addressed
NA
## Primary Change
When investigating memory usage, I noticed that retrieving a block from an early slot (e.g., slot 900) would cause a sharp increase in the memory footprint (from 400mb to 800mb+) which seemed to be ever-lasting.
After some investigation, I found that the reverse iteration from the head back to that slot was the likely culprit. To counter this, I've switched the `BeaconChain::block_root_at_slot` to use the forwards iterator, instead of the reverse one.
I also noticed that the networking stack is using `BeaconChain::root_at_slot` to check if a peer is relevant (`check_peer_relevance`). Perhaps the steep, seemingly-random-but-consistent increases in memory usage are caused by the use of this function.
Using the forwards iterator with the HTTP API alleviated the sharp increases in memory usage. It also made the response much faster (before it felt like to took 1-2s, now it feels instant).
## Additional Changes
In the process I also noticed that we have two functions for getting block roots:
- `BeaconChain::block_root_at_slot`: returns `None` for a skip slot.
- `BeaconChain::root_at_slot`: returns the previous root for a skip slot.
I unified these two functions into `block_root_at_slot` and added the `WhenSlotSkipped` enum. Now, the caller must be explicit about the skip-slot behaviour when requesting a root.
Additionally, I replaced `vec![]` with `Vec::with_capacity` in `store::chunked_vector::range_query`. I stumbled across this whilst debugging and made this modification to see what effect it would have (not much). It seems like a decent change to keep around, but I'm not concerned either way.
Also, `BeaconChain::get_ancestor_block_root` is unused, so I got rid of it :wastebasket:.
## Additional Info
I haven't also done the same for state roots here. Whilst it's possible and a good idea, it's more work since the fwds iterators are presently block-roots-specific.
Whilst there's a few places a reverse iteration of state roots could be triggered (e.g., attestation production, HTTP API), they're no where near as common as the `check_peer_relevance` call. As such, I think we should get this PR merged first, then come back for the state root iters. I made an issue here https://github.com/sigp/lighthouse/issues/2377.
2021-05-31 04:18:20 +00:00
|
|
|
StateSkipConfig, WhenSlotSkipped,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
};
|
2021-10-07 11:24:57 +00:00
|
|
|
use fork_choice::{
|
2022-07-25 23:53:26 +00:00
|
|
|
CountUnrealized, ForkChoiceStore, InvalidAttestation, InvalidBlock, PayloadVerificationStatus,
|
|
|
|
QueuedAttestation,
|
2021-10-07 11:24:57 +00:00
|
|
|
};
|
2021-10-14 02:58:10 +00:00
|
|
|
use store::MemoryStore;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
use types::{
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
test_utils::generate_deterministic_keypair, BeaconBlockRef, BeaconState, ChainSpec, Checkpoint,
|
|
|
|
Epoch, EthSpec, Hash256, IndexedAttestation, MainnetEthSpec, SignedBeaconBlock, Slot, SubnetId,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
pub type E = MainnetEthSpec;
|
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
pub const VALIDATOR_COUNT: usize = 32;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
/// Defines some delay between when an attestation is created and when it is mutated.
|
|
|
|
pub enum MutationDelay {
|
|
|
|
/// No delay between creation and mutation.
|
|
|
|
NoDelay,
|
|
|
|
/// Create `n` blocks before mutating the attestation.
|
|
|
|
Blocks(usize),
|
|
|
|
}
|
|
|
|
|
|
|
|
/// A helper struct to make testing fork choice more ergonomic and less repetitive.
|
|
|
|
struct ForkChoiceTest {
|
2023-02-14 12:33:38 +00:00
|
|
|
harness: BeaconChainHarness<EphemeralHarnessType<E>>,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
2020-10-01 01:41:58 +00:00
|
|
|
/// Allows us to use `unwrap` in some cases.
|
|
|
|
impl fmt::Debug for ForkChoiceTest {
|
|
|
|
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
|
|
|
|
f.debug_struct("ForkChoiceTest").finish()
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
impl ForkChoiceTest {
|
|
|
|
/// Creates a new tester.
|
|
|
|
pub fn new() -> Self {
|
2021-10-14 02:58:10 +00:00
|
|
|
let harness = BeaconChainHarness::builder(MainnetEthSpec)
|
|
|
|
.default_spec()
|
|
|
|
.deterministic_keypairs(VALIDATOR_COUNT)
|
|
|
|
.fresh_ephemeral_store()
|
|
|
|
.build();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
Self { harness }
|
|
|
|
}
|
|
|
|
|
2020-10-01 01:41:58 +00:00
|
|
|
/// Creates a new tester with a custom chain config.
|
|
|
|
pub fn new_with_chain_config(chain_config: ChainConfig) -> Self {
|
2021-10-14 02:58:10 +00:00
|
|
|
let harness = BeaconChainHarness::builder(MainnetEthSpec)
|
|
|
|
.default_spec()
|
|
|
|
.chain_config(chain_config)
|
|
|
|
.deterministic_keypairs(VALIDATOR_COUNT)
|
|
|
|
.fresh_ephemeral_store()
|
|
|
|
.build();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
Self { harness }
|
|
|
|
}
|
|
|
|
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
/// Get a value from the `ForkChoice` instantiation.
|
|
|
|
fn get<T, U>(&self, func: T) -> U
|
|
|
|
where
|
2020-06-17 02:50:32 +00:00
|
|
|
T: Fn(&BeaconForkChoiceStore<E, MemoryStore<E>, MemoryStore<E>>) -> U,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
{
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
func(
|
|
|
|
&self
|
|
|
|
.harness
|
|
|
|
.chain
|
|
|
|
.canonical_head
|
|
|
|
.fork_choice_read_lock()
|
|
|
|
.fc_store(),
|
|
|
|
)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert the epochs match.
|
|
|
|
pub fn assert_finalized_epoch(self, epoch: u64) -> Self {
|
|
|
|
assert_eq!(
|
|
|
|
self.get(|fc_store| fc_store.finalized_checkpoint().epoch),
|
|
|
|
Epoch::new(epoch),
|
|
|
|
"finalized_epoch"
|
|
|
|
);
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert the epochs match.
|
|
|
|
pub fn assert_justified_epoch(self, epoch: u64) -> Self {
|
|
|
|
assert_eq!(
|
|
|
|
self.get(|fc_store| fc_store.justified_checkpoint().epoch),
|
|
|
|
Epoch::new(epoch),
|
|
|
|
"justified_epoch"
|
|
|
|
);
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert the epochs match.
|
|
|
|
pub fn assert_best_justified_epoch(self, epoch: u64) -> Self {
|
|
|
|
assert_eq!(
|
|
|
|
self.get(|fc_store| fc_store.best_justified_checkpoint().epoch),
|
|
|
|
Epoch::new(epoch),
|
|
|
|
"best_justified_epoch"
|
|
|
|
);
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
2020-10-01 01:41:58 +00:00
|
|
|
/// Assert the given slot is greater than the head slot.
|
|
|
|
pub fn assert_finalized_epoch_is_less_than(self, epoch: Epoch) -> Self {
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
assert!(self.harness.finalized_checkpoint().epoch < epoch);
|
2020-10-01 01:41:58 +00:00
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert there was a shutdown signal sent by the beacon chain.
|
2022-02-28 22:07:48 +00:00
|
|
|
pub fn shutdown_signal_sent(&self) -> bool {
|
|
|
|
let mutex = self.harness.shutdown_receiver.clone();
|
|
|
|
let mut shutdown_receiver = mutex.lock();
|
|
|
|
|
|
|
|
shutdown_receiver.close();
|
|
|
|
let msg = shutdown_receiver.try_next().unwrap();
|
|
|
|
msg.is_some()
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert there was a shutdown signal sent by the beacon chain.
|
|
|
|
pub fn assert_shutdown_signal_sent(self) -> Self {
|
|
|
|
assert!(self.shutdown_signal_sent());
|
2020-10-01 01:41:58 +00:00
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Assert no shutdown was signal sent by the beacon chain.
|
2022-02-28 22:07:48 +00:00
|
|
|
pub fn assert_shutdown_signal_not_sent(self) -> Self {
|
|
|
|
assert!(!self.shutdown_signal_sent());
|
2020-10-01 01:41:58 +00:00
|
|
|
self
|
|
|
|
}
|
|
|
|
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
/// Inspect the queued attestations in fork choice.
|
|
|
|
pub fn inspect_queued_attestations<F>(self, mut func: F) -> Self
|
|
|
|
where
|
|
|
|
F: FnMut(&[QueuedAttestation]),
|
|
|
|
{
|
|
|
|
self.harness
|
|
|
|
.chain
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.canonical_head
|
|
|
|
.fork_choice_write_lock()
|
2022-07-25 23:53:26 +00:00
|
|
|
.update_time(self.harness.chain.slot().unwrap(), &self.harness.spec)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
func(
|
|
|
|
self.harness
|
|
|
|
.chain
|
|
|
|
.canonical_head
|
|
|
|
.fork_choice_read_lock()
|
|
|
|
.queued_attestations(),
|
|
|
|
);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Skip a slot, without producing a block.
|
|
|
|
pub fn skip_slot(self) -> Self {
|
|
|
|
self.harness.advance_slot();
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
2020-07-01 02:45:34 +00:00
|
|
|
/// Skips `count` slots, without producing a block.
|
|
|
|
pub fn skip_slots(self, count: usize) -> Self {
|
|
|
|
for _ in 0..count {
|
|
|
|
self.harness.advance_slot();
|
|
|
|
}
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
2020-10-01 01:41:58 +00:00
|
|
|
/// Build the chain whilst `predicate` returns `true` and `process_block_result` does not error.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
pub async fn apply_blocks_while<F>(self, mut predicate: F) -> Result<Self, Self>
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
where
|
2021-07-09 06:15:32 +00:00
|
|
|
F: FnMut(BeaconBlockRef<'_, E>, &BeaconState<E>) -> bool,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
{
|
|
|
|
self.harness.advance_slot();
|
2020-08-26 09:24:55 +00:00
|
|
|
let mut state = self.harness.get_current_state();
|
|
|
|
let validators = self.harness.get_all_validators();
|
|
|
|
loop {
|
|
|
|
let slot = self.harness.get_current_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let (block, state_) = self.harness.make_block(state, slot).await;
|
2020-08-26 09:24:55 +00:00
|
|
|
state = state_;
|
2021-07-09 06:15:32 +00:00
|
|
|
if !predicate(block.message(), &state) {
|
2020-08-26 09:24:55 +00:00
|
|
|
break;
|
|
|
|
}
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
if let Ok(block_hash) = self.harness.process_block_result(block.clone()).await {
|
2021-03-17 05:09:57 +00:00
|
|
|
self.harness.attest_block(
|
|
|
|
&state,
|
|
|
|
block.state_root(),
|
|
|
|
block_hash,
|
|
|
|
&block,
|
|
|
|
&validators,
|
|
|
|
);
|
2020-10-01 01:41:58 +00:00
|
|
|
self.harness.advance_slot();
|
|
|
|
} else {
|
|
|
|
return Err(self);
|
|
|
|
}
|
2020-08-26 09:24:55 +00:00
|
|
|
}
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
2020-10-01 01:41:58 +00:00
|
|
|
Ok(self)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Apply `count` blocks to the chain (with attestations).
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
pub async fn apply_blocks(self, count: usize) -> Self {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self.harness.advance_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
self.harness
|
|
|
|
.extend_chain(
|
|
|
|
count,
|
|
|
|
BlockStrategy::OnCanonicalHead,
|
|
|
|
AttestationStrategy::AllValidators,
|
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Apply `count` blocks to the chain (without attestations).
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
pub async fn apply_blocks_without_new_attestations(self, count: usize) -> Self {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self.harness.advance_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
self.harness
|
|
|
|
.extend_chain(
|
|
|
|
count,
|
|
|
|
BlockStrategy::OnCanonicalHead,
|
|
|
|
AttestationStrategy::SomeValidators(vec![]),
|
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Moves to the next slot that is *outside* the `SAFE_SLOTS_TO_UPDATE_JUSTIFIED` range.
|
|
|
|
///
|
|
|
|
/// If the chain is presently in an unsafe period, transition through it and the following safe
|
|
|
|
/// period.
|
|
|
|
pub fn move_to_next_unsafe_period(self) -> Self {
|
|
|
|
self.move_inside_safe_to_update()
|
|
|
|
.move_outside_safe_to_update()
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Moves to the next slot that is *outside* the `SAFE_SLOTS_TO_UPDATE_JUSTIFIED` range.
|
|
|
|
pub fn move_outside_safe_to_update(self) -> Self {
|
2021-11-08 07:29:04 +00:00
|
|
|
while is_safe_to_update(self.harness.chain.slot().unwrap(), &self.harness.chain.spec) {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self.harness.advance_slot()
|
|
|
|
}
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Moves to the next slot that is *inside* the `SAFE_SLOTS_TO_UPDATE_JUSTIFIED` range.
|
|
|
|
pub fn move_inside_safe_to_update(self) -> Self {
|
2021-11-08 07:29:04 +00:00
|
|
|
while !is_safe_to_update(self.harness.chain.slot().unwrap(), &self.harness.chain.spec) {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self.harness.advance_slot()
|
|
|
|
}
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Applies a block directly to fork choice, bypassing the beacon chain.
|
|
|
|
///
|
|
|
|
/// Asserts the block was applied successfully.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
pub async fn apply_block_directly_to_fork_choice<F>(self, mut func: F) -> Self
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
where
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
F: FnMut(&mut SignedBeaconBlock<E>, &mut BeaconState<E>),
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
{
|
2020-08-26 09:24:55 +00:00
|
|
|
let state = self
|
|
|
|
.harness
|
|
|
|
.chain
|
|
|
|
.state_at_slot(
|
|
|
|
self.harness.get_current_slot() - 1,
|
|
|
|
StateSkipConfig::WithStateRoots,
|
|
|
|
)
|
|
|
|
.unwrap();
|
|
|
|
let slot = self.harness.get_current_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let (mut signed_block, mut state) = self.harness.make_block(state, slot).await;
|
|
|
|
func(&mut signed_block, &mut state);
|
2020-08-26 09:24:55 +00:00
|
|
|
let current_slot = self.harness.get_current_slot();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
self.harness
|
|
|
|
.chain
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.canonical_head
|
|
|
|
.fork_choice_write_lock()
|
2021-10-07 11:24:57 +00:00
|
|
|
.on_block(
|
|
|
|
current_slot,
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
signed_block.message(),
|
|
|
|
signed_block.canonical_root(),
|
2021-12-13 20:43:22 +00:00
|
|
|
Duration::from_secs(0),
|
2021-10-07 11:24:57 +00:00
|
|
|
&state,
|
|
|
|
PayloadVerificationStatus::Verified,
|
2021-11-15 00:26:42 +00:00
|
|
|
&self.harness.chain.spec,
|
2022-07-25 23:53:26 +00:00
|
|
|
CountUnrealized::True,
|
2021-10-07 11:24:57 +00:00
|
|
|
)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap();
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Applies a block directly to fork choice, bypassing the beacon chain.
|
|
|
|
///
|
|
|
|
/// Asserts that an error occurred and allows inspecting it via `comparison_func`.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
pub async fn apply_invalid_block_directly_to_fork_choice<F, G>(
|
2020-10-19 05:58:39 +00:00
|
|
|
self,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
mut mutation_func: F,
|
|
|
|
mut comparison_func: G,
|
|
|
|
) -> Self
|
|
|
|
where
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
F: FnMut(&mut SignedBeaconBlock<E>, &mut BeaconState<E>),
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
G: FnMut(ForkChoiceError),
|
|
|
|
{
|
2020-08-26 09:24:55 +00:00
|
|
|
let state = self
|
|
|
|
.harness
|
|
|
|
.chain
|
|
|
|
.state_at_slot(
|
|
|
|
self.harness.get_current_slot() - 1,
|
|
|
|
StateSkipConfig::WithStateRoots,
|
|
|
|
)
|
|
|
|
.unwrap();
|
|
|
|
let slot = self.harness.get_current_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let (mut signed_block, mut state) = self.harness.make_block(state, slot).await;
|
|
|
|
mutation_func(&mut signed_block, &mut state);
|
2020-08-26 09:24:55 +00:00
|
|
|
let current_slot = self.harness.get_current_slot();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let err = self
|
|
|
|
.harness
|
|
|
|
.chain
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.canonical_head
|
|
|
|
.fork_choice_write_lock()
|
2021-10-07 11:24:57 +00:00
|
|
|
.on_block(
|
|
|
|
current_slot,
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
signed_block.message(),
|
|
|
|
signed_block.canonical_root(),
|
2021-12-13 20:43:22 +00:00
|
|
|
Duration::from_secs(0),
|
2021-10-07 11:24:57 +00:00
|
|
|
&state,
|
|
|
|
PayloadVerificationStatus::Verified,
|
2021-11-15 00:26:42 +00:00
|
|
|
&self.harness.chain.spec,
|
2022-07-25 23:53:26 +00:00
|
|
|
CountUnrealized::True,
|
2021-10-07 11:24:57 +00:00
|
|
|
)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.err()
|
|
|
|
.expect("on_block did not return an error");
|
|
|
|
comparison_func(err);
|
|
|
|
self
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Compares the justified balances in the `ForkChoiceStore` verses a direct lookup from the
|
|
|
|
/// database.
|
|
|
|
fn check_justified_balances(&self) {
|
|
|
|
let harness = &self.harness;
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let fc = self.harness.chain.canonical_head.fork_choice_read_lock();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
let state_root = harness
|
|
|
|
.chain
|
|
|
|
.store
|
Separate execution payloads in the DB (#3157)
## Proposed Changes
Reduce post-merge disk usage by not storing finalized execution payloads in Lighthouse's database.
:warning: **This is achieved in a backwards-incompatible way for networks that have already merged** :warning:. Kiln users and shadow fork enjoyers will be unable to downgrade after running the code from this PR. The upgrade migration may take several minutes to run, and can't be aborted after it begins.
The main changes are:
- New column in the database called `ExecPayload`, keyed by beacon block root.
- The `BeaconBlock` column now stores blinded blocks only.
- Lots of places that previously used full blocks now use blinded blocks, e.g. analytics APIs, block replay in the DB, etc.
- On finalization:
- `prune_abanonded_forks` deletes non-canonical payloads whilst deleting non-canonical blocks.
- `migrate_db` deletes finalized canonical payloads whilst deleting finalized states.
- Conversions between blinded and full blocks are implemented in a compositional way, duplicating some work from Sean's PR #3134.
- The execution layer has a new `get_payload_by_block_hash` method that reconstructs a payload using the EE's `eth_getBlockByHash` call.
- I've tested manually that it works on Kiln, using Geth and Nethermind.
- This isn't necessarily the most efficient method, and new engine APIs are being discussed to improve this: https://github.com/ethereum/execution-apis/pull/146.
- We're depending on the `ethers` master branch, due to lots of recent changes. We're also using a workaround for https://github.com/gakonst/ethers-rs/issues/1134.
- Payload reconstruction is used in the HTTP API via `BeaconChain::get_block`, which is now `async`. Due to the `async` fn, the `blocking_json` wrapper has been removed.
- Payload reconstruction is used in network RPC to serve blocks-by-{root,range} responses. Here the `async` adjustment is messier, although I think I've managed to come up with a reasonable compromise: the handlers take the `SendOnDrop` by value so that they can drop it on _task completion_ (after the `fn` returns). Still, this is introducing disk reads onto core executor threads, which may have a negative performance impact (thoughts appreciated).
## Additional Info
- [x] For performance it would be great to remove the cloning of full blocks when converting them to blinded blocks to write to disk. I'm going to experiment with a `put_block` API that takes the block by value, breaks it into a blinded block and a payload, stores the blinded block, and then re-assembles the full block for the caller.
- [x] We should measure the latency of blocks-by-root and blocks-by-range responses.
- [x] We should add integration tests that stress the payload reconstruction (basic tests done, issue for more extensive tests: https://github.com/sigp/lighthouse/issues/3159)
- [x] We should (manually) test the schema v9 migration from several prior versions, particularly as blocks have changed on disk and some migrations rely on being able to load blocks.
Co-authored-by: Paul Hauner <paul@paulhauner.com>
2022-05-12 00:42:17 +00:00
|
|
|
.get_blinded_block(&fc.fc_store().justified_checkpoint().root)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap()
|
|
|
|
.unwrap()
|
2021-07-09 06:15:32 +00:00
|
|
|
.message()
|
|
|
|
.state_root();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let state = harness
|
|
|
|
.chain
|
|
|
|
.store
|
|
|
|
.get_state(&state_root, None)
|
|
|
|
.unwrap()
|
|
|
|
.unwrap();
|
|
|
|
let balances = state
|
2021-07-09 06:15:32 +00:00
|
|
|
.validators()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.into_iter()
|
|
|
|
.map(|v| {
|
|
|
|
if v.is_active_at(state.current_epoch()) {
|
|
|
|
v.effective_balance
|
|
|
|
} else {
|
|
|
|
0
|
|
|
|
}
|
|
|
|
})
|
|
|
|
.collect::<Vec<_>>();
|
|
|
|
|
|
|
|
assert_eq!(
|
|
|
|
&balances[..],
|
2022-12-13 09:57:26 +00:00
|
|
|
&fc.fc_store().justified_balances().effective_balances,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
"balances should match"
|
2022-12-13 09:57:26 +00:00
|
|
|
);
|
|
|
|
assert_eq!(
|
|
|
|
balances.iter().sum::<u64>(),
|
|
|
|
fc.fc_store().justified_balances().total_effective_balance
|
|
|
|
);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Returns an attestation that is valid for some slot in the given `chain`.
|
|
|
|
///
|
|
|
|
/// Also returns some info about who created it.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
async fn apply_attestation_to_chain<F, G>(
|
2020-10-19 05:58:39 +00:00
|
|
|
self,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
delay: MutationDelay,
|
|
|
|
mut mutation_func: F,
|
|
|
|
mut comparison_func: G,
|
|
|
|
) -> Self
|
|
|
|
where
|
2023-02-14 12:33:38 +00:00
|
|
|
F: FnMut(&mut IndexedAttestation<E>, &BeaconChain<EphemeralHarnessType<E>>),
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
G: FnMut(Result<(), BeaconChainError>),
|
|
|
|
{
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let head = self.harness.chain.head_snapshot();
|
2020-08-26 09:24:55 +00:00
|
|
|
let current_slot = self.harness.chain.slot().expect("should get slot");
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
let mut attestation = self
|
|
|
|
.harness
|
|
|
|
.chain
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.produce_unaggregated_attestation(current_slot, 0)
|
|
|
|
.expect("should not error while producing attestation");
|
|
|
|
|
|
|
|
let validator_committee_index = 0;
|
|
|
|
let validator_index = *head
|
|
|
|
.beacon_state
|
|
|
|
.get_beacon_committee(current_slot, attestation.data.index)
|
|
|
|
.expect("should get committees")
|
|
|
|
.committee
|
|
|
|
.get(validator_committee_index)
|
|
|
|
.expect("there should be an attesting validator");
|
|
|
|
|
2020-06-18 09:11:03 +00:00
|
|
|
let committee_count = head
|
|
|
|
.beacon_state
|
|
|
|
.get_committee_count_at_slot(current_slot)
|
|
|
|
.expect("should not error while getting committee count");
|
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
let subnet_id = SubnetId::compute_subnet::<E>(
|
|
|
|
current_slot,
|
|
|
|
0,
|
|
|
|
committee_count,
|
|
|
|
&self.harness.chain.spec,
|
|
|
|
)
|
|
|
|
.expect("should compute subnet id");
|
2020-06-18 09:11:03 +00:00
|
|
|
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let validator_sk = generate_deterministic_keypair(validator_index).sk;
|
|
|
|
|
|
|
|
attestation
|
|
|
|
.sign(
|
|
|
|
&validator_sk,
|
|
|
|
validator_committee_index,
|
2021-07-09 06:15:32 +00:00
|
|
|
&head.beacon_state.fork(),
|
2020-08-26 09:24:55 +00:00
|
|
|
self.harness.chain.genesis_validators_root,
|
|
|
|
&self.harness.chain.spec,
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
)
|
|
|
|
.expect("should sign attestation");
|
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
let mut verified_attestation = self
|
|
|
|
.harness
|
|
|
|
.chain
|
Batch BLS verification for attestations (#2399)
## Issue Addressed
NA
## Proposed Changes
Adds the ability to verify batches of aggregated/unaggregated attestations from the network.
When the `BeaconProcessor` finds there are messages in the aggregated or unaggregated attestation queues, it will first check the length of the queue:
- `== 1` verify the attestation individually.
- `>= 2` take up to 64 of those attestations and verify them in a batch.
Notably, we only perform batch verification if the queue has a backlog. We don't apply any artificial delays to attestations to try and force them into batches.
### Batching Details
To assist with implementing batches we modify `beacon_chain::attestation_verification` to have two distinct categories for attestations:
- *Indexed* attestations: those which have passed initial validation and were valid enough for us to derive an `IndexedAttestation`.
- *Verified* attestations: those attestations which were indexed *and also* passed signature verification. These are well-formed, interesting messages which were signed by validators.
The batching functions accept `n` attestations and then return `n` attestation verification `Result`s, where those `Result`s can be any combination of `Ok` or `Err`. In other words, we attempt to verify as many attestations as possible and return specific per-attestation results so peer scores can be updated, if required.
When we batch verify attestations, we first try to map all those attestations to *indexed* attestations. If any of those attestations were able to be indexed, we then perform batch BLS verification on those indexed attestations. If the batch verification succeeds, we convert them into *verified* attestations, disabling individual signature checking. If the batch fails, we convert to verified attestations with individual signature checking enabled.
Ultimately, we optimistically try to do a batch verification of attestation signatures and fall-back to individual verification if it fails. This opens an attach vector for "poisoning" the attestations and causing us to waste a batch verification. I argue that peer scoring should do a good-enough job of defending against this and the typical-case gains massively outweigh the worst-case losses.
## Additional Info
Before this PR, attestation verification took the attestations by value (instead of by reference). It turns out that this was unnecessary and, in my opinion, resulted in some undesirable ergonomics (e.g., we had to pass the attestation back in the `Err` variant to avoid clones). In this PR I've modified attestation verification so that it now takes a reference.
I refactored the `beacon_chain/tests/attestation_verification.rs` tests so they use a builder-esque "tester" struct instead of a weird macro. It made it easier for me to test individual/batch with the same set of tests and I think it was a nice tidy-up. Notably, I did this last to try and make sure my new refactors to *actual* production code would pass under the existing test suite.
2021-09-22 08:49:41 +00:00
|
|
|
.verify_unaggregated_attestation_for_gossip(&attestation, Some(subnet_id))
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.expect("precondition: should gossip verify attestation");
|
|
|
|
|
|
|
|
if let MutationDelay::Blocks(slots) = delay {
|
|
|
|
self.harness.advance_slot();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
self.harness
|
|
|
|
.extend_chain(
|
|
|
|
slots,
|
|
|
|
BlockStrategy::OnCanonicalHead,
|
|
|
|
AttestationStrategy::SomeValidators(vec![]),
|
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
mutation_func(
|
|
|
|
verified_attestation.__indexed_attestation_mut(),
|
|
|
|
&self.harness.chain,
|
|
|
|
);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
2020-08-26 09:24:55 +00:00
|
|
|
let result = self
|
|
|
|
.harness
|
|
|
|
.chain
|
|
|
|
.apply_attestation_to_fork_choice(&verified_attestation);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
comparison_func(result);
|
|
|
|
|
|
|
|
self
|
|
|
|
}
|
2020-09-18 05:14:31 +00:00
|
|
|
|
|
|
|
/// Check to ensure that we can read the finalized block. This is a regression test.
|
|
|
|
pub fn check_finalized_block_is_accessible(self) -> Self {
|
|
|
|
self.harness
|
|
|
|
.chain
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.canonical_head
|
|
|
|
.fork_choice_read_lock()
|
|
|
|
.get_block(&self.harness.finalized_checkpoint().root)
|
2020-09-18 05:14:31 +00:00
|
|
|
.unwrap();
|
|
|
|
|
|
|
|
self
|
|
|
|
}
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
2021-11-08 07:29:04 +00:00
|
|
|
fn is_safe_to_update(slot: Slot, spec: &ChainSpec) -> bool {
|
|
|
|
slot % E::slots_per_epoch() < spec.safe_slots_to_update_justified
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
2022-02-28 22:07:48 +00:00
|
|
|
#[test]
|
|
|
|
fn justified_and_finalized_blocks() {
|
|
|
|
let tester = ForkChoiceTest::new();
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let fork_choice = tester.harness.chain.canonical_head.fork_choice_read_lock();
|
2022-02-28 22:07:48 +00:00
|
|
|
|
|
|
|
let justified_checkpoint = fork_choice.justified_checkpoint();
|
|
|
|
assert_eq!(justified_checkpoint.epoch, 0);
|
|
|
|
assert!(justified_checkpoint.root != Hash256::zero());
|
|
|
|
assert!(fork_choice.get_justified_block().is_ok());
|
|
|
|
|
|
|
|
let finalized_checkpoint = fork_choice.finalized_checkpoint();
|
|
|
|
assert_eq!(finalized_checkpoint.epoch, 0);
|
|
|
|
assert!(finalized_checkpoint.root != Hash256::zero());
|
|
|
|
assert!(fork_choice.get_finalized_block().is_ok());
|
|
|
|
}
|
|
|
|
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
/// - The new justified checkpoint descends from the current.
|
|
|
|
/// - Current slot is within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_with_descendent_inside_safe_slots() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_inside_safe_to_update()
|
|
|
|
.assert_justified_epoch(0)
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(2);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// - The new justified checkpoint descends from the current.
|
|
|
|
/// - Current slot is **not** within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`
|
|
|
|
/// - This is **not** the first justification since genesis
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_with_descendent_outside_safe_slots() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch <= 2)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_outside_safe_to_update()
|
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.assert_best_justified_epoch(2)
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(3);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// - The new justified checkpoint descends from the current.
|
|
|
|
/// - Current slot is **not** within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`
|
|
|
|
/// - This is the first justification since genesis
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_first_justification_outside_safe_to_update() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_to_next_unsafe_period()
|
|
|
|
.assert_justified_epoch(0)
|
|
|
|
.assert_best_justified_epoch(0)
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.assert_best_justified_epoch(2);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// - The new justified checkpoint **does not** descend from the current.
|
|
|
|
/// - Current slot is within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`
|
|
|
|
/// - Finalized epoch has **not** increased.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_with_non_descendent_inside_safe_slots_without_finality() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_inside_safe_to_update()
|
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.apply_block_directly_to_fork_choice(|_, state| {
|
|
|
|
// The finalized checkpoint should not change.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.finalized_checkpoint().epoch = Epoch::new(0);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
// The justified checkpoint has changed.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().epoch = Epoch::new(3);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
// The new block should **not** include the current justified block as an ancestor.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().root = *state
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.get_block_root(Epoch::new(1).start_slot(E::slots_per_epoch()))
|
|
|
|
.unwrap();
|
|
|
|
})
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(3)
|
|
|
|
.assert_best_justified_epoch(3);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// - The new justified checkpoint **does not** descend from the current.
|
|
|
|
/// - Current slot is **not** within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`.
|
|
|
|
/// - Finalized epoch has **not** increased.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_with_non_descendent_outside_safe_slots_without_finality() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_to_next_unsafe_period()
|
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.apply_block_directly_to_fork_choice(|_, state| {
|
|
|
|
// The finalized checkpoint should not change.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.finalized_checkpoint().epoch = Epoch::new(0);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
// The justified checkpoint has changed.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().epoch = Epoch::new(3);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
// The new block should **not** include the current justified block as an ancestor.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().root = *state
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.get_block_root(Epoch::new(1).start_slot(E::slots_per_epoch()))
|
|
|
|
.unwrap();
|
|
|
|
})
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.assert_best_justified_epoch(3);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// - The new justified checkpoint **does not** descend from the current.
|
|
|
|
/// - Current slot is **not** within `SAFE_SLOTS_TO_UPDATE_JUSTIFIED`
|
|
|
|
/// - Finalized epoch has increased.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_checkpoint_updates_with_non_descendent_outside_safe_slots_with_finality() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.move_to_next_unsafe_period()
|
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.apply_block_directly_to_fork_choice(|_, state| {
|
|
|
|
// The finalized checkpoint should change.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.finalized_checkpoint_mut().epoch = Epoch::new(1);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
|
|
|
|
// The justified checkpoint has changed.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().epoch = Epoch::new(3);
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
// The new block should **not** include the current justified block as an ancestor.
|
2021-07-09 06:15:32 +00:00
|
|
|
state.current_justified_checkpoint_mut().root = *state
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.get_block_root(Epoch::new(1).start_slot(E::slots_per_epoch()))
|
|
|
|
.unwrap();
|
|
|
|
})
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(3)
|
|
|
|
.assert_best_justified_epoch(3);
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Check that the balances are obtained correctly.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn justified_balances() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.current_justified_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_justified_epoch(2)
|
|
|
|
.check_justified_balances()
|
|
|
|
}
|
|
|
|
|
|
|
|
macro_rules! assert_invalid_block {
|
2022-02-25 00:10:17 +00:00
|
|
|
($err: tt, $($error: pat_param) |+ $( if $guard: expr )?) => {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
assert!(
|
|
|
|
matches!(
|
|
|
|
$err,
|
|
|
|
$( ForkChoiceError::InvalidBlock($error) ) |+ $( if $guard )?
|
|
|
|
),
|
2021-10-22 04:49:51 +00:00
|
|
|
)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
};
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1
|
|
|
|
///
|
|
|
|
/// assert block.parent_root in store.block_states
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_block_unknown_parent() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let junk = Hash256::from_low_u64_be(42);
|
|
|
|
|
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks(2)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_invalid_block_directly_to_fork_choice(
|
|
|
|
|block, _| {
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
*block.message_mut().parent_root_mut() = junk;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
},
|
|
|
|
|err| {
|
|
|
|
assert_invalid_block!(
|
|
|
|
err,
|
|
|
|
InvalidBlock::UnknownParent(parent)
|
|
|
|
if parent == junk
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1
|
|
|
|
///
|
|
|
|
/// assert get_current_slot(store) >= block.slot
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_block_future_slot() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks(2)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_invalid_block_directly_to_fork_choice(
|
|
|
|
|block, _| {
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
*block.message_mut().slot_mut() += 1;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
},
|
2021-02-10 23:29:49 +00:00
|
|
|
|err| assert_invalid_block!(err, InvalidBlock::FutureSlot { .. }),
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1
|
|
|
|
///
|
|
|
|
/// assert block.slot > finalized_slot
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_block_finalized_slot() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_invalid_block_directly_to_fork_choice(
|
|
|
|
|block, _| {
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
*block.message_mut().slot_mut() =
|
|
|
|
Epoch::new(2).start_slot(E::slots_per_epoch()) - 1;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
},
|
|
|
|
|err| {
|
|
|
|
assert_invalid_block!(
|
|
|
|
err,
|
|
|
|
InvalidBlock::FinalizedSlot { finalized_slot, .. }
|
|
|
|
if finalized_slot == Epoch::new(2).start_slot(E::slots_per_epoch())
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1
|
|
|
|
///
|
|
|
|
/// assert get_ancestor(store, hash_tree_root(block), finalized_slot) ==
|
2021-07-09 06:15:32 +00:00
|
|
|
/// store.finalized_checkpoint().root
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
///
|
|
|
|
/// Note: we technically don't do this exact check, but an equivalent check. Reference:
|
|
|
|
///
|
|
|
|
/// https://github.com/ethereum/eth2.0-specs/pull/1884
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_block_finalized_descendant() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let invalid_ancestor = Mutex::new(Hash256::zero());
|
|
|
|
|
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.assert_finalized_epoch(2)
|
|
|
|
.apply_invalid_block_directly_to_fork_choice(
|
|
|
|
|block, state| {
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
*block.message_mut().parent_root_mut() = *state
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.get_block_root(Epoch::new(1).start_slot(E::slots_per_epoch()))
|
|
|
|
.unwrap();
|
2021-07-09 06:15:32 +00:00
|
|
|
*invalid_ancestor.lock().unwrap() = block.parent_root();
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
},
|
|
|
|
|err| {
|
|
|
|
assert_invalid_block!(
|
|
|
|
err,
|
|
|
|
InvalidBlock::NotFinalizedDescendant { block_ancestor, .. }
|
|
|
|
if block_ancestor == Some(*invalid_ancestor.lock().unwrap())
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
macro_rules! assert_invalid_attestation {
|
2022-02-25 00:10:17 +00:00
|
|
|
($err: tt, $($error: pat_param) |+ $( if $guard: expr )?) => {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
assert!(
|
|
|
|
matches!(
|
|
|
|
$err,
|
|
|
|
$( Err(BeaconChainError::ForkChoiceError(ForkChoiceError::InvalidAttestation($error))) ) |+ $( if $guard )?
|
|
|
|
),
|
|
|
|
"{:?}",
|
|
|
|
$err
|
2021-10-22 04:49:51 +00:00
|
|
|
)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
};
|
|
|
|
}
|
|
|
|
|
|
|
|
/// Ensure we can process a valid attestation.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn valid_attestation() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|_, _| {},
|
|
|
|
|result| assert_eq!(result.unwrap(), ()),
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// This test is not in the specification, however we reject an attestation with an empty
|
|
|
|
/// aggregation bitfield since it has no purpose beyond wasting our time.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_empty_bitfield() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.attesting_indices = vec![].into();
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(result, InvalidAttestation::EmptyAggregationBitfield)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert target.epoch in [expected_current_epoch, previous_epoch]
|
|
|
|
///
|
|
|
|
/// (tests epoch after current epoch)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_future_epoch() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.data.target.epoch = Epoch::new(2);
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::FutureEpoch { attestation_epoch, current_epoch }
|
|
|
|
if attestation_epoch == Epoch::new(2) && current_epoch == Epoch::new(0)
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert target.epoch in [expected_current_epoch, previous_epoch]
|
|
|
|
///
|
|
|
|
/// (tests epoch prior to previous epoch)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_past_epoch() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(E::slots_per_epoch() as usize * 3 + 1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.data.target.epoch = Epoch::new(0);
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::PastEpoch { attestation_epoch, current_epoch }
|
|
|
|
if attestation_epoch == Epoch::new(0) && current_epoch == Epoch::new(3)
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert target.epoch == compute_epoch_at_slot(attestation.data.slot)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_target_epoch() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(E::slots_per_epoch() as usize + 1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.data.slot = Slot::new(1);
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::BadTargetEpoch { target, slot }
|
|
|
|
if target == Epoch::new(1) && slot == Slot::new(1)
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert target.root in store.blocks
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_unknown_target_root() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let junk = Hash256::from_low_u64_be(42);
|
|
|
|
|
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.data.target.root = junk;
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::UnknownTargetRoot(root)
|
|
|
|
if root == junk
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert attestation.data.beacon_block_root in store.blocks
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_unknown_beacon_block_root() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let junk = Hash256::from_low_u64_be(42);
|
|
|
|
|
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _| {
|
|
|
|
attestation.data.beacon_block_root = junk;
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::UnknownHeadBlock { beacon_block_root }
|
|
|
|
if beacon_block_root == junk
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert store.blocks[attestation.data.beacon_block_root].slot <= attestation.data.slot
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_future_block() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::Blocks(1),
|
|
|
|
|attestation, chain| {
|
|
|
|
attestation.data.beacon_block_root = chain
|
Use the forwards iterator more often (#2376)
## Issue Addressed
NA
## Primary Change
When investigating memory usage, I noticed that retrieving a block from an early slot (e.g., slot 900) would cause a sharp increase in the memory footprint (from 400mb to 800mb+) which seemed to be ever-lasting.
After some investigation, I found that the reverse iteration from the head back to that slot was the likely culprit. To counter this, I've switched the `BeaconChain::block_root_at_slot` to use the forwards iterator, instead of the reverse one.
I also noticed that the networking stack is using `BeaconChain::root_at_slot` to check if a peer is relevant (`check_peer_relevance`). Perhaps the steep, seemingly-random-but-consistent increases in memory usage are caused by the use of this function.
Using the forwards iterator with the HTTP API alleviated the sharp increases in memory usage. It also made the response much faster (before it felt like to took 1-2s, now it feels instant).
## Additional Changes
In the process I also noticed that we have two functions for getting block roots:
- `BeaconChain::block_root_at_slot`: returns `None` for a skip slot.
- `BeaconChain::root_at_slot`: returns the previous root for a skip slot.
I unified these two functions into `block_root_at_slot` and added the `WhenSlotSkipped` enum. Now, the caller must be explicit about the skip-slot behaviour when requesting a root.
Additionally, I replaced `vec![]` with `Vec::with_capacity` in `store::chunked_vector::range_query`. I stumbled across this whilst debugging and made this modification to see what effect it would have (not much). It seems like a decent change to keep around, but I'm not concerned either way.
Also, `BeaconChain::get_ancestor_block_root` is unused, so I got rid of it :wastebasket:.
## Additional Info
I haven't also done the same for state roots here. Whilst it's possible and a good idea, it's more work since the fwds iterators are presently block-roots-specific.
Whilst there's a few places a reverse iteration of state roots could be triggered (e.g., attestation production, HTTP API), they're no where near as common as the `check_peer_relevance` call. As such, I think we should get this PR merged first, then come back for the state root iters. I made an issue here https://github.com/sigp/lighthouse/issues/2377.
2021-05-31 04:18:20 +00:00
|
|
|
.block_at_slot(chain.slot().unwrap(), WhenSlotSkipped::Prev)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap()
|
|
|
|
.unwrap()
|
|
|
|
.canonical_root();
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::AttestsToFutureBlock { block, attestation }
|
|
|
|
if block == 2 && attestation == 1
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert target.root == get_ancestor(store, attestation.data.beacon_block_root, target_slot)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_inconsistent_ffg_vote() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
let local_opt = Mutex::new(None);
|
|
|
|
let attestation_opt = Mutex::new(None);
|
|
|
|
|
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, chain| {
|
|
|
|
attestation.data.target.root = chain
|
Use the forwards iterator more often (#2376)
## Issue Addressed
NA
## Primary Change
When investigating memory usage, I noticed that retrieving a block from an early slot (e.g., slot 900) would cause a sharp increase in the memory footprint (from 400mb to 800mb+) which seemed to be ever-lasting.
After some investigation, I found that the reverse iteration from the head back to that slot was the likely culprit. To counter this, I've switched the `BeaconChain::block_root_at_slot` to use the forwards iterator, instead of the reverse one.
I also noticed that the networking stack is using `BeaconChain::root_at_slot` to check if a peer is relevant (`check_peer_relevance`). Perhaps the steep, seemingly-random-but-consistent increases in memory usage are caused by the use of this function.
Using the forwards iterator with the HTTP API alleviated the sharp increases in memory usage. It also made the response much faster (before it felt like to took 1-2s, now it feels instant).
## Additional Changes
In the process I also noticed that we have two functions for getting block roots:
- `BeaconChain::block_root_at_slot`: returns `None` for a skip slot.
- `BeaconChain::root_at_slot`: returns the previous root for a skip slot.
I unified these two functions into `block_root_at_slot` and added the `WhenSlotSkipped` enum. Now, the caller must be explicit about the skip-slot behaviour when requesting a root.
Additionally, I replaced `vec![]` with `Vec::with_capacity` in `store::chunked_vector::range_query`. I stumbled across this whilst debugging and made this modification to see what effect it would have (not much). It seems like a decent change to keep around, but I'm not concerned either way.
Also, `BeaconChain::get_ancestor_block_root` is unused, so I got rid of it :wastebasket:.
## Additional Info
I haven't also done the same for state roots here. Whilst it's possible and a good idea, it's more work since the fwds iterators are presently block-roots-specific.
Whilst there's a few places a reverse iteration of state roots could be triggered (e.g., attestation production, HTTP API), they're no where near as common as the `check_peer_relevance` call. As such, I think we should get this PR merged first, then come back for the state root iters. I made an issue here https://github.com/sigp/lighthouse/issues/2377.
2021-05-31 04:18:20 +00:00
|
|
|
.block_at_slot(Slot::new(1), WhenSlotSkipped::Prev)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap()
|
|
|
|
.unwrap()
|
|
|
|
.canonical_root();
|
|
|
|
|
|
|
|
*attestation_opt.lock().unwrap() = Some(attestation.data.target.root);
|
|
|
|
*local_opt.lock().unwrap() = Some(
|
|
|
|
chain
|
Use the forwards iterator more often (#2376)
## Issue Addressed
NA
## Primary Change
When investigating memory usage, I noticed that retrieving a block from an early slot (e.g., slot 900) would cause a sharp increase in the memory footprint (from 400mb to 800mb+) which seemed to be ever-lasting.
After some investigation, I found that the reverse iteration from the head back to that slot was the likely culprit. To counter this, I've switched the `BeaconChain::block_root_at_slot` to use the forwards iterator, instead of the reverse one.
I also noticed that the networking stack is using `BeaconChain::root_at_slot` to check if a peer is relevant (`check_peer_relevance`). Perhaps the steep, seemingly-random-but-consistent increases in memory usage are caused by the use of this function.
Using the forwards iterator with the HTTP API alleviated the sharp increases in memory usage. It also made the response much faster (before it felt like to took 1-2s, now it feels instant).
## Additional Changes
In the process I also noticed that we have two functions for getting block roots:
- `BeaconChain::block_root_at_slot`: returns `None` for a skip slot.
- `BeaconChain::root_at_slot`: returns the previous root for a skip slot.
I unified these two functions into `block_root_at_slot` and added the `WhenSlotSkipped` enum. Now, the caller must be explicit about the skip-slot behaviour when requesting a root.
Additionally, I replaced `vec![]` with `Vec::with_capacity` in `store::chunked_vector::range_query`. I stumbled across this whilst debugging and made this modification to see what effect it would have (not much). It seems like a decent change to keep around, but I'm not concerned either way.
Also, `BeaconChain::get_ancestor_block_root` is unused, so I got rid of it :wastebasket:.
## Additional Info
I haven't also done the same for state roots here. Whilst it's possible and a good idea, it's more work since the fwds iterators are presently block-roots-specific.
Whilst there's a few places a reverse iteration of state roots could be triggered (e.g., attestation production, HTTP API), they're no where near as common as the `check_peer_relevance` call. As such, I think we should get this PR merged first, then come back for the state root iters. I made an issue here https://github.com/sigp/lighthouse/issues/2377.
2021-05-31 04:18:20 +00:00
|
|
|
.block_at_slot(Slot::new(0), WhenSlotSkipped::Prev)
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.unwrap()
|
|
|
|
.unwrap()
|
|
|
|
.canonical_root(),
|
|
|
|
);
|
|
|
|
},
|
|
|
|
|result| {
|
|
|
|
assert_invalid_attestation!(
|
|
|
|
result,
|
|
|
|
InvalidAttestation::InvalidTarget { attestation, local }
|
|
|
|
if attestation == attestation_opt.lock().unwrap().unwrap()
|
|
|
|
&& local == local_opt.lock().unwrap().unwrap()
|
|
|
|
)
|
|
|
|
},
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/// Specification v0.12.1:
|
|
|
|
///
|
|
|
|
/// assert get_current_slot(store) >= attestation.data.slot + 1
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn invalid_attestation_delayed_slot() {
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks_without_new_attestations(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.inspect_queued_attestations(|queue| assert_eq!(queue.len(), 0))
|
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|_, _| {},
|
|
|
|
|result| assert_eq!(result.unwrap(), ()),
|
|
|
|
)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
v0.12 fork choice update (#1229)
* Incomplete scraps
* Add progress on new fork choice impl
* Further progress
* First complete compiling version
* Remove chain reference
* Add new lmd_ghost crate
* Start integrating into beacon chain
* Update `milagro_bls` to new release (#1183)
* Update milagro_bls to new release
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Tidy up fake cryptos
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* move SecretHash to bls and put plaintext back
Signed-off-by: Kirk Baird <baird.k@outlook.com>
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Rough beacon chain impl working
* Remove fork_choice_2
* Remove checkpoint manager
* Half finished ssz impl
* Add missed file
* Add persistence
* Tidy, fix some compile errors
* Remove RwLock from ProtoArrayForkChoice
* Fix store-based compile errors
* Add comments, tidy
* Move function out of ForkChoice struct
* Start testing
* More testing
* Fix compile error
* Tidy beacon_chain::fork_choice
* Queue attestations from the current slot
* Allow fork choice to handle prior-to-genesis start
* Improve error granularity
* Test attestation dequeuing
* Process attestations during block
* Store target root in fork choice
* Move fork choice verification into new crate
* Update tests
* Consensus updates for v0.12 (#1228)
* Update state processing for v0.12
* Fix EF test runners for v0.12
* Fix some tests
* Fix broken attestation verification test
* More test fixes
* Fix typo found in review
* Add `Block` struct to ProtoArray
* Start fixing get_ancestor
* Add rough progress on testing
* Get fork choice tests working
* Progress with testing
* Fix partialeq impl
* Move slot clock from fc_store
* Improve testing
* Add testing for best justified
* Add clone back to SystemTimeSlotClock
* Add balances test
* Start adding balances cache again
* Wire-in balances cache
* Improve tests
* Remove commented-out tests
* Remove beacon_chain::ForkChoice
* Rename crates
* Update wider codebase to new fork_choice layout
* Move advance_slot in test harness
* Tidy ForkChoice::update_time
* Fix verification tests
* Fix compile error with iter::once
* Fix fork choice tests
* Ensure block attestations are processed
* Fix failing beacon_chain tests
* Add first invalid block check
* Add finalized block check
* Progress with testing, new store builder
* Add fixes to get_ancestor
* Fix old genesis justification test
* Fix remaining fork choice tests
* Change root iteration method
* Move on_verified_block
* Remove unused method
* Start adding attestation verification tests
* Add invalid ffg target test
* Add target epoch test
* Add queued attestation test
* Remove old fork choice verification tests
* Tidy, add test
* Move fork choice lock drop
* Rename BeaconForkChoiceStore
* Add comments, tidy BeaconForkChoiceStore
* Update metrics, rename fork_choice_store.rs
* Remove genesis_block_root from ForkChoice
* Tidy
* Update fork_choice comments
* Tidy, add comments
* Tidy, simplify ForkChoice, fix compile issue
* Tidy, removed dead file
* Increase http request timeout
* Fix failing rest_api test
* Set HTTP timeout back to 5s
* Apply fix to get_ancestor
* Address Michael's comments
* Fix typo
* Revert "Fix broken attestation verification test"
This reverts commit 722cdc903b12611de27916a57eeecfa3224f2279.
Co-authored-by: Kirk Baird <baird.k@outlook.com>
Co-authored-by: Michael Sproul <michael@sigmaprime.io>
2020-06-17 01:10:22 +00:00
|
|
|
.inspect_queued_attestations(|queue| assert_eq!(queue.len(), 1))
|
|
|
|
.skip_slot()
|
|
|
|
.inspect_queued_attestations(|queue| assert_eq!(queue.len(), 0));
|
|
|
|
}
|
2020-07-01 02:45:34 +00:00
|
|
|
|
|
|
|
/// Tests that the correct target root is used when the attested-to block is in a prior epoch to
|
|
|
|
/// the attestation.
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn valid_attestation_skip_across_epoch() {
|
2020-07-01 02:45:34 +00:00
|
|
|
ForkChoiceTest::new()
|
|
|
|
.apply_blocks(E::slots_per_epoch() as usize - 1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-07-01 02:45:34 +00:00
|
|
|
.skip_slots(2)
|
|
|
|
.apply_attestation_to_chain(
|
|
|
|
MutationDelay::NoDelay,
|
|
|
|
|attestation, _chain| {
|
|
|
|
assert_eq!(
|
|
|
|
attestation.data.target.root,
|
|
|
|
attestation.data.beacon_block_root
|
|
|
|
)
|
|
|
|
},
|
|
|
|
|result| result.unwrap(),
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
)
|
|
|
|
.await;
|
2020-07-01 02:45:34 +00:00
|
|
|
}
|
2020-09-18 05:14:31 +00:00
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn can_read_finalized_block() {
|
2020-09-18 05:14:31 +00:00
|
|
|
ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
2020-09-18 05:14:31 +00:00
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-09-18 05:14:31 +00:00
|
|
|
.check_finalized_block_is_accessible();
|
|
|
|
}
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
#[test]
|
|
|
|
#[should_panic]
|
|
|
|
fn weak_subjectivity_fail_on_startup() {
|
|
|
|
let epoch = Epoch::new(0);
|
|
|
|
let root = Hash256::from_low_u64_le(1);
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(Checkpoint { epoch, root }),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config);
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_pass_on_startup() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let epoch = Epoch::new(0);
|
|
|
|
let root = Hash256::zero();
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(Checkpoint { epoch, root }),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config)
|
|
|
|
.apply_blocks(E::slots_per_epoch() as usize)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_shutdown_signal_not_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_passes() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(2);
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let checkpoint = setup_harness.harness.finalized_checkpoint();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(2)
|
|
|
|
.assert_shutdown_signal_not_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_fails_early_epoch() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(2);
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let mut checkpoint = setup_harness.harness.finalized_checkpoint();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
checkpoint.epoch = checkpoint.epoch - 1;
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 3)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap_err()
|
|
|
|
.assert_finalized_epoch_is_less_than(checkpoint.epoch)
|
|
|
|
.assert_shutdown_signal_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_fails_late_epoch() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(2);
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let mut checkpoint = setup_harness.harness.finalized_checkpoint();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
checkpoint.epoch = checkpoint.epoch + 1;
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 4)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap_err()
|
|
|
|
.assert_finalized_epoch_is_less_than(checkpoint.epoch)
|
|
|
|
.assert_shutdown_signal_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_fails_incorrect_root() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(2);
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let mut checkpoint = setup_harness.harness.finalized_checkpoint();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
checkpoint.root = Hash256::zero();
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 3)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap_err()
|
|
|
|
.assert_finalized_epoch_is_less_than(checkpoint.epoch)
|
|
|
|
.assert_shutdown_signal_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_epoch_boundary_is_skip_slot() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
|
|
|
// first two epochs
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap();
|
|
|
|
|
|
|
|
// get the head, it will become the finalized root of epoch 4
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let checkpoint_root = setup_harness.harness.head_block_root();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
setup_harness
|
|
|
|
// epoch 3 will be entirely skip slots
|
|
|
|
.skip_slots(E::slots_per_epoch() as usize)
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 5)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(5);
|
|
|
|
|
|
|
|
// the checkpoint at epoch 4 should become the root of last block of epoch 2
|
|
|
|
let checkpoint = Checkpoint {
|
|
|
|
epoch: Epoch::new(4),
|
|
|
|
root: checkpoint_root,
|
|
|
|
};
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
// recreate the chain exactly
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.skip_slots(E::slots_per_epoch() as usize)
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 5)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(5)
|
|
|
|
.assert_shutdown_signal_not_sent();
|
|
|
|
}
|
|
|
|
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
#[tokio::test]
|
|
|
|
async fn weak_subjectivity_check_epoch_boundary_is_skip_slot_failure() {
|
2020-10-01 01:41:58 +00:00
|
|
|
let setup_harness = ForkChoiceTest::new()
|
|
|
|
// first two epochs
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap();
|
|
|
|
|
|
|
|
// get the head, it will become the finalized root of epoch 4
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
let checkpoint_root = setup_harness.harness.head_block_root();
|
2020-10-01 01:41:58 +00:00
|
|
|
|
|
|
|
setup_harness
|
|
|
|
// epoch 3 will be entirely skip slots
|
|
|
|
.skip_slots(E::slots_per_epoch() as usize)
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 5)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.apply_blocks(1)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.assert_finalized_epoch(5);
|
|
|
|
|
|
|
|
// Invalid checkpoint (epoch too early)
|
|
|
|
let checkpoint = Checkpoint {
|
|
|
|
epoch: Epoch::new(1),
|
|
|
|
root: checkpoint_root,
|
|
|
|
};
|
|
|
|
|
|
|
|
let chain_config = ChainConfig {
|
|
|
|
weak_subjectivity_checkpoint: Some(checkpoint),
|
2021-09-22 00:37:28 +00:00
|
|
|
..ChainConfig::default()
|
2020-10-01 01:41:58 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
// recreate the chain exactly
|
|
|
|
ForkChoiceTest::new_with_chain_config(chain_config.clone())
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch == 0)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap()
|
|
|
|
.skip_slots(E::slots_per_epoch() as usize)
|
2021-07-09 06:15:32 +00:00
|
|
|
.apply_blocks_while(|_, state| state.finalized_checkpoint().epoch < 6)
|
Use async code when interacting with EL (#3244)
## Overview
This rather extensive PR achieves two primary goals:
1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.
Additionally, it achieves:
- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
- I had to do this to deal with sending blocks into spawned tasks.
- Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
- We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
- Avoids cloning *all the blocks* in *every chain segment* during sync.
- It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough :sweat_smile:)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.
For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273
## Changes to `canonical_head` and `fork_choice`
Previously, the `BeaconChain` had two separate fields:
```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```
Now, we have grouped these values under a single struct:
```
canonical_head: CanonicalHead {
cached_head: RwLock<Arc<Snapshot>>,
fork_choice: RwLock<BeaconForkChoice>
}
```
Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.
## Breaking Changes
### The `state` (root) field in the `finalized_checkpoint` SSE event
Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:
1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.
Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java#L171-L182) it uses [`getStateRootFromBlockRoot`](https://github.com/ConsenSys/teku/blob/de2b2801c89ef5abf983d6bf37867c37fc47121f/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java#L336-L341) which uses (1).
I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.
## Notes for Reviewers
I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.
I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".
I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.
I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.
Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.
You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.
I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.
Co-authored-by: Mac L <mjladson@pm.me>
2022-07-03 05:36:50 +00:00
|
|
|
.await
|
2020-10-01 01:41:58 +00:00
|
|
|
.unwrap_err()
|
|
|
|
.assert_finalized_epoch_is_less_than(checkpoint.epoch)
|
|
|
|
.assert_shutdown_signal_sent();
|
|
|
|
}
|