We need to always use the state-tree from the tipset _after_ the message
executed. If we use any other state-tree, we might not find the address
we're trying to resolve.
This change also has some implication for pending messages: there's no
guarantee we'll be able to generate a 0x-style address for a pending
native message. So, instead of trying, I've removed support for pending
native messages from the Eth API. Messages from EthAccounts will still
work, and native messages will still show up in blocks/traces, they just
won't show up as "pending". Which should affect exactly nobody.
I'm also taking this opportunity to cleanup some edge-cases:
1. Pass contexts where appropriate.
2. Remove all state access from `ethTxHashFromSignedMessage`.
Part of #11355
* upgrade calibnet by removing move_partitions from miner actor in actor v12
* cids for buggy bundles
* revert changes to v12 tar
* upgrade system actor state
* update based on manifest
* nit: clean up some comments
* chore: rename param to oldBuggyMinerCID
* refactor, ensure both buggy bundles are loaded
* update to actors v12.0.0-rc.3
* fix: load second buggy bundle for UpgradeWatermelonFixHeight
* add calibration fix2 upgrade epcoh
* update mainnet upgrade epoch
---------
Co-authored-by: Aayush <arajasek94@gmail.com>
Co-authored-by: jennijuju <jiayingw703@gmail.com>
* refactor: make GetEmbeddedBuiltinActorsBundle take network bundle name
* update calibnet actor bundle to v12.0.0-rc.2, but include v12.0.0-rc.1 as calibrationnet-buggy.car
* wip: calibnet unbork migration
* calibnet: add buggy miner actor CID to actorMeta
* fix incorrect buggy calibnet manifest
* make UpgradeWatermelonFixHeight a build param
* calibnet patch: check whether network is calibration from init actor state
* add sanity checks to the v12 calibnet patch upgrade
* address review
After changing in prev commit to use to ethereum addresses the
comparison does not make sense against builtin actors. This
fixes that by storing also the filecoin addresses in each trace
Also renamed filecoin related fields to Filecoin prefix.
Also remove requirement call to InvokeContract needed to come
from a evm actor
The sequence number used for replay detection was being updated before message validation confirmed that the message originated from the correct host. This would allow one host A to create a message with the ID of another host B that could then update the cached sequence number for B. While the message from A would fail validation and be ignored, the cached sequence number for B would get updated. This would lead to a temporary DoS for host B as its messages were incorrectly rejected as replays.
This fixes the issue by setting the cached sequence number after message validation.
