cerc-io/lighthouse
16
0
Fork 0
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity

Use SmallVec for TreeHash packed encoding (#3581)

Browse Source 
## Issue Addressed

NA

## Proposed Changes

I've noticed that our block hashing times increase significantly after the merge. I did some flamegraph-ing and noticed that we're allocating a `Vec` for each byte of each execution payload transaction. This seems like unnecessary work and a bit of a fragmentation risk.

This PR switches to `SmallVec<[u8; 32]>` for the packed encoding of `TreeHash`. I believe this is a nice simple optimisation with no downside.

### Benchmarking

These numbers were computed using #3580 on my desktop (i7 hex-core). You can see a bit of noise in the numbers, that's probably just my computer doing other things. Generally I found this change takes the time from 10-11ms to 8-9ms. I can also see all the allocations disappear from flamegraph.

This is the block being benchmarked: https://beaconcha.in/slot/4704236

#### Before

```
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 980: 10.553003ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 981: 10.563737ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 982: 10.646352ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 983: 10.628532ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 984: 10.552112ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 985: 10.587778ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 986: 10.640526ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 987: 10.587243ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 988: 10.554748ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 989: 10.551111ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 990: 11.559031ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 991: 11.944827ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 992: 10.554308ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 993: 11.043397ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 994: 11.043315ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 995: 11.207711ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 996: 11.056246ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 997: 11.049706ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 998: 11.432449ms
[2022-09-15T21:44:19Z INFO  lcli::block_root] Run 999: 11.149617ms
```

#### After

```
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 980: 14.011653ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 981: 8.925314ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 982: 8.849563ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 983: 8.893689ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 984: 8.902964ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 985: 8.942067ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 986: 8.907088ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 987: 9.346101ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 988: 8.96142ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 989: 9.366437ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 990: 9.809334ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 991: 9.541561ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 992: 11.143518ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 993: 10.821181ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 994: 9.855973ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 995: 10.941006ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 996: 9.596155ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 997: 9.121739ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 998: 9.090019ms
[2022-09-15T21:41:49Z INFO  lcli::block_root] Run 999: 9.071885ms
```

## Additional Info

Please provide any additional information. For example, future considerations
or information useful for reviewers.
This commit is contained in:
Paul Hauner 2022-09-16 08:54:06 +00:00
parent bde3c168e2
commit b0b606dabe
15 changed files with 44 additions and 41 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
consensus/ssz_types/src/bitfield.rs
View File

@ -620,7 +620,7 @@ impl<N: Unsigned + Clone> tree_hash::TreeHash for Bitfield<Variable<N>> {
tree_hash::TreeHashType::List
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("List should never be packed.")
}
@ -641,7 +641,7 @@ impl<N: Unsigned + Clone> tree_hash::TreeHash for Bitfield<Fixed<N>> {
tree_hash::TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Vector should never be packed.")
}

2
consensus/ssz_types/src/fixed_vector.rs
View File

@ -167,7 +167,7 @@ where
tree_hash::TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Vector should never be packed.")
}

2
consensus/ssz_types/src/variable_list.rs
View File

@ -184,7 +184,7 @@ where
tree_hash::TreeHashType::List
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("List should never be packed.")
}

30
consensus/tree_hash/src/impls.rs
View File

@ -14,8 +14,8 @@ macro_rules! impl_for_bitsize {
TreeHashType::Basic
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
self.to_le_bytes().to_vec()
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
PackedEncoding::from_slice(&self.to_le_bytes())
}
fn tree_hash_packing_factor() -> usize {
@ -41,7 +41,7 @@ impl TreeHash for bool {
TreeHashType::Basic
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
(*self as u8).tree_hash_packed_encoding()
}
@ -62,7 +62,7 @@ macro_rules! impl_for_lt_32byte_u8_array {
TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
unreachable!("bytesN should never be packed.")
}
@ -87,10 +87,10 @@ impl TreeHash for U128 {
TreeHashType::Basic
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
let mut result = vec![0; 16];
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
let mut result = [0; 16];
self.to_little_endian(&mut result);
result
PackedEncoding::from_slice(&result)
}
fn tree_hash_packing_factor() -> usize {
@ -109,10 +109,10 @@ impl TreeHash for U256 {
TreeHashType::Basic
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
let mut result = vec![0; 32];
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
let mut result = [0; 32];
self.to_little_endian(&mut result);
result
PackedEncoding::from_slice(&result)
}
fn tree_hash_packing_factor() -> usize {
@ -131,10 +131,10 @@ impl TreeHash for H160 {
TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
let mut result = vec![0; 32];
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
let mut result = [0; 32];
result[0..20].copy_from_slice(self.as_bytes());
result
PackedEncoding::from_slice(&result)
}
fn tree_hash_packing_factor() -> usize {
@ -153,8 +153,8 @@ impl TreeHash for H256 {
TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
self.as_bytes().to_vec()
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
PackedEncoding::from_slice(self.as_bytes())
}
fn tree_hash_packing_factor() -> usize {

11
consensus/tree_hash/src/lib.rs
View File

@ -8,13 +8,16 @@ pub use merkleize_padded::merkleize_padded;
pub use merkleize_standard::merkleize_standard;
use eth2_hashing::{hash_fixed, ZERO_HASHES, ZERO_HASHES_MAX_INDEX};
use smallvec::SmallVec;
pub const BYTES_PER_CHUNK: usize = 32;
pub const HASHSIZE: usize = 32;
pub const MERKLE_HASH_CHUNK: usize = 2 * BYTES_PER_CHUNK;
pub const MAX_UNION_SELECTOR: u8 = 127;
pub const SMALLVEC_SIZE: usize = 32;
pub type Hash256 = ethereum_types::H256;
pub type PackedEncoding = SmallVec<[u8; SMALLVEC_SIZE]>;
/// Convenience method for `MerkleHasher` which also provides some fast-paths for small trees.
///
@ -109,7 +112,7 @@ pub enum TreeHashType {
pub trait TreeHash {
fn tree_hash_type() -> TreeHashType;
fn tree_hash_packed_encoding(&self) -> Vec<u8>;
fn tree_hash_packed_encoding(&self) -> PackedEncoding;
fn tree_hash_packing_factor() -> usize;
@ -125,7 +128,7 @@ where
T::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
T::tree_hash_packed_encoding(*self)
}
@ -146,7 +149,7 @@ macro_rules! tree_hash_ssz_encoding_as_vector {
tree_hash::TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
unreachable!("Vector should never be packed.")
}
@ -169,7 +172,7 @@ macro_rules! tree_hash_ssz_encoding_as_list {
tree_hash::TreeHashType::List
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
unreachable!("List should never be packed.")
}

4
consensus/tree_hash/tests/tests.rs
View File

@ -1,5 +1,5 @@
use ssz_derive::Encode;
use tree_hash::{Hash256, MerkleHasher, TreeHash, BYTES_PER_CHUNK};
use tree_hash::{Hash256, MerkleHasher, PackedEncoding, TreeHash, BYTES_PER_CHUNK};
use tree_hash_derive::TreeHash;
#[derive(Encode)]
@ -18,7 +18,7 @@ impl tree_hash::TreeHash for HashVec {
tree_hash::TreeHashType::List
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
unreachable!("List should never be packed.")
}

6
consensus/tree_hash_derive/src/lib.rs
View File

@ -150,7 +150,7 @@ fn tree_hash_derive_struct(item: &DeriveInput, struct_data: &DataStruct) -> Toke
tree_hash::TreeHashType::Container
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Struct should never be packed.")
}
@ -231,7 +231,7 @@ fn tree_hash_derive_enum_transparent(
tree_hash::TreeHashType::Container
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Enum should never be packed")
}
@ -288,7 +288,7 @@ fn tree_hash_derive_enum_union(derive_input: &DeriveInput, enum_data: &DataEnum)
tree_hash::TreeHashType::Container
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Enum should never be packed")
}

2
consensus/types/src/execution_block_hash.rs
View File

@ -67,7 +67,7 @@ impl tree_hash::TreeHash for ExecutionBlockHash {
Hash256::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
self.0.tree_hash_packed_encoding()
}

4
consensus/types/src/graffiti.rs
View File

@ -7,7 +7,7 @@ use serde::{de::Error, Deserialize, Deserializer, Serialize, Serializer};
use ssz::{Decode, DecodeError, Encode};
use std::fmt;
use std::str::FromStr;
use tree_hash::TreeHash;
use tree_hash::{PackedEncoding, TreeHash};
pub const GRAFFITI_BYTES_LEN: usize = 32;
@ -159,7 +159,7 @@ impl TreeHash for Graffiti {
<[u8; GRAFFITI_BYTES_LEN]>::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
self.0.tree_hash_packed_encoding()
}

4
consensus/types/src/participation_flags.rs
View File

@ -3,7 +3,7 @@ use safe_arith::{ArithError, SafeArith};
use serde_derive::{Deserialize, Serialize};
use ssz::{Decode, DecodeError, Encode};
use test_random_derive::TestRandom;
use tree_hash::{TreeHash, TreeHashType};
use tree_hash::{PackedEncoding, TreeHash, TreeHashType};
#[derive(Debug, Default, Clone, Copy, PartialEq, Deserialize, Serialize, TestRandom)]
#[serde(transparent)]
@ -78,7 +78,7 @@ impl TreeHash for ParticipationFlags {
u8::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
self.bits.tree_hash_packed_encoding()
}

6
consensus/types/src/payload.rs
View File

@ -7,7 +7,7 @@ use std::convert::TryFrom;
use std::fmt::Debug;
use std::hash::Hash;
use test_random_derive::TestRandom;
use tree_hash::TreeHash;
use tree_hash::{PackedEncoding, TreeHash};
#[derive(Debug)]
pub enum BlockType {
@ -175,7 +175,7 @@ impl<T: EthSpec> TreeHash for BlindedPayload<T> {
<ExecutionPayloadHeader<T>>::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> PackedEncoding {
self.execution_payload_header.tree_hash_packed_encoding()
}
@ -244,7 +244,7 @@ impl<T: EthSpec> TreeHash for FullPayload<T> {
<ExecutionPayload<T>>::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
self.execution_payload.tree_hash_packed_encoding()
}

2
consensus/types/src/slot_epoch.rs
View File

@ -16,7 +16,7 @@ use crate::{ChainSpec, SignedRoot};
use rand::RngCore;
use safe_arith::{ArithError, SafeArith};
use serde_derive::{Deserialize, Serialize};
use ssz::{ssz_encode, Decode, DecodeError, Encode};
use ssz::{Decode, DecodeError, Encode};
use std::fmt;
use std::hash::Hash;
use std::iter::Iterator;

4
consensus/types/src/slot_epoch_macros.rs
View File

@ -290,8 +290,8 @@ macro_rules! impl_ssz {
tree_hash::TreeHashType::Basic
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
ssz_encode(self)
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
self.0.tree_hash_packed_encoding()
}
fn tree_hash_packing_factor() -> usize {

2
crypto/bls/src/macros.rs
View File

@ -7,7 +7,7 @@ macro_rules! impl_tree_hash {
tree_hash::TreeHashType::Vector
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
unreachable!("Vector should never be packed.")
}

2
testing/ef_tests/src/cases/common.rs
View File

@ -43,7 +43,7 @@ macro_rules! uint_wrapper {
<$wrapped_type>::tree_hash_type()
}
fn tree_hash_packed_encoding(&self) -> Vec<u8> {
fn tree_hash_packed_encoding(&self) -> tree_hash::PackedEncoding {
self.x.tree_hash_packed_encoding()
}