Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix: stmgr: make the tipset and height agree when estimating gas #10216

Merged
merged 3 commits into from
Feb 10, 2023

Conversation

Stebalien
Copy link
Member

@Stebalien Stebalien commented Feb 8, 2023

Related Issues

Should fix filecoin-project/ref-fvm#1648.

Proposed Changes

Specifically re-execute all messages in the current tipset, tacking the new message onto the end. That way, the epoch is the epoch of the current tipset.

We could try to "make" a fake block and use that, but that's unlikely to work well.

Checklist

Before you mark the PR ready for review, please make sure that:

  • Commits have a clear commit message.
  • PR title is in the form of of <PR type>: <area>: <change being made>
    • example: fix: mempool: Introduce a cache for valid signatures
    • PR type: fix, feat, build, chore, ci, docs, perf, refactor, revert, style, test
    • area, e.g. api, chain, state, market, mempool, multisig, networking, paych, proving, sealing, wallet, deps
  • Tests exist for new functionality or change in behavior
  • CI is green

@maciejwitowski
Copy link
Contributor

maciejwitowski commented Feb 9, 2023

Tested #10216, it works now.

@snissn
Copy link
Contributor

snissn commented Feb 9, 2023

This test also confirms this PR works #10206

@@ -142,14 +136,14 @@ func (sm *StateManager) callInternal(ctx context.Context, msg *types.Message, pr
buffStore := blockstore.NewTieredBstore(sm.cs.StateBlockstore(), blockstore.NewMemorySync())
vmopt := &vm.VMOpts{
StateBase: stateCid,
Epoch: vmHeight,
Epoch: ts.Height(),
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we're changing the value of Epoch here from vmHeight (ts.Height()+1) to ts.Height() is that off by 1 change ok?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, that's the point.

Comment on lines 110 to 121
// Unless executing on a specific state cid, apply all the messages from the current tipset
// first. Unfortunately, we can't just execute the tipset, because that will run cron. We
// don't want to apply miner messages after cron runs in a given epoch.
if stateCid == cid.Undef {
stateCid = ts.ParentState()
tsMsgs, err := sm.cs.MessagesForTipset(ctx, ts)
if err != nil {
return nil, xerrors.Errorf("failed to lookup messages for parent tipset: %w", err)
}
priorMsgs = append(tsMsgs, priorMsgs...)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Are we sure this isn't breaking StateCall / lotus state compute-state?

Exchanges rely on this to track transactions.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That uses

func ComputeState(ctx context.Context, sm *StateManager, height abi.ChainEpoch, msgs []*types.Message, ts *types.TipSet) (cid.Cid, []*api.InvocResult, error) {

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Which is also incorrect? Well, it's correct if you don't pass any messages.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hm. It does affect StateCall. I'm going to rework this a bit.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done. I now only do this for CallWithGas.

Specifically re-execute all messages in the current tipset, tacking the new
message onto the end. That way, the epoch is the epoch of the current tipset.

We could try to "make" a fake block and use that, but that's unlikely to
work well.
@Stebalien Stebalien enabled auto-merge (squash) February 10, 2023 20:47
@Stebalien Stebalien merged commit 01e616f into release/v1.20.0 Feb 10, 2023
@Stebalien Stebalien deleted the steb/fix-tipset-selection branch February 10, 2023 20:49
dumikau pushed a commit to protofire/lotus that referenced this pull request Jun 8, 2023
…in-project#71)

* Test and fix eth_FeeHistory

* Add test for eth_getStorageAt

* Add test for eth_call

* Change EthBlock.Extradata type to EthBytes

* fix: EthAPI: Drop hack in GetStorageAt

* Eth JSON-RPC: implement web3_clientVersion

* fix lint

* add genesis for ipc gateway

* Remove debug logging

* chore: cli

Standarize cli/code functions similar to: filecoin-project#9317

- cctx.NArg() instead of cctx.Args().xxx
- Add check for args and print help on functions that did not have it

* fix: rpcenc: Don't hang when source dies

* make docsgen-cli

make docsgen-cli

* fix: EthAPI: Handle EthCall return correctly

* feat: vm: Assert empty object CID when dumping state

* Update codeql to v2

* Enable code ql for release branches

* less strict ArgsCheck

Less strict ArgsCheck

* feat: wdpost: Emit more detailed errors

* Fix panic in EthGetCode

* Update codeql to v2

* Enable code ql for release branches

* Add actors to circ-supply

- Consider funds to EAM as locked
- Add evm, placeholder and ethaccount to circ supply

* devnets starts with nv17

* Use new kit.DisableEthRPC method in test

* make gen

* init git submodules in build-docker CI

* Remove old Snapcraft and AppDir folders

* Always include checked in files in docker context

The .dockerignore file is symlinked to the .gitignore file, and checked
in files should not be removed from our docker context otherwise they
result in dirty git state when we build our images.

* Check git state in Dockerfile to catch .dockerignore bugs

* make jen

- Add builtin.EthereumAddressManagerActorAddr to builtin.go.template and make jen
- Rename to EthereumAddressManagerActorAddr to match pattern of other actors (CronActorAddr/etc)

* chore: all: bump go-libipfs to replace go-block-format

Includes changes from:
- ipfs/go-block-format#37
- ipfs/boxo#58

* init git submodules in build-docker CI

* Remove old Snapcraft and AppDir folders

* Always include checked in files in docker context

The .dockerignore file is symlinked to the .gitignore file, and checked
in files should not be removed from our docker context otherwise they
result in dirty git state when we build our images.

* Check git state in Dockerfile to catch .dockerignore bugs

* Fix lint errors

* Skip some rpc conformance tests

* make gen

* Changed skip message

* fix: don't replace the method in EthSendRawTransaction (filecoin-project#10129)

This will just cause signature validation to fail.

* Remove stm: #integration comment

* Fix comment

* fix: devnets: default to starting from nv17

* Refactor to ensure conformance test can run in circleci

* go fmt

* Standardize path variable

* Update to go-state-types v0.10.0-alpha-11

* chore: node: migrate go-bitswap to go-libipfs/bitswap

This was migrated in ipfs/boxo#63.

* simplify transaction hash db queries, prevent fd leaks

* Refactor: Unify EthTx to FilecoinMessage methods v2 (filecoin-project#10095)

* Refactor: Unify EthTx to FilecoinMessage methods

* Filecoin messages can again be converted to Eth Txs

* All BLS messages should calculated tx hash with unsigned message

* Refactor newEthTxReceipt

* fill in from and to for non-eth transactions

* Hoist nil check out of newEthTxFromMessageLookup

---------

Co-authored-by: Aayush <[email protected]>
Co-authored-by: Raúl Kripalani <[email protected]>

* Populate transaction hash database if the database has not been set up before

* rpc: Switch eth_subscribe to reverse calls

* gateway: eth_subscribe support

* itests: Fix TestEthSubscribeLogs

* ethtypes: Serialize EthFilterID/EthSubscriptionID correctly

* eth rpc: Params are optional in eth_subscribe

* fix: cli: add ArgsUsage

fix: cli: add ArgsUsage

* make libp2p user agent overridable

* move UserAgentOption upwards

* Eth JSON-RPC: from in eth_getTransactionByHash is not correctly populated filecoin-project#1614

* Move assigning `from` address to the API level

* Improve the error message

* Add test for EthGetTransactionByHash

* fix: should not serve non v0 api in v1

* Fix: typo

Fix: typo

* Fix: typos

Fix: typos

* Fix: typo

Fix: typo

* Fix: typo

Fix: typo

* Fix: typo

Fix: typo

* Fix: typos

Fix: typos

* Fix: typos

Fix: typos

* test: fevm: add in tests for deploying, destroying contracts, recursive calls, sending value (filecoin-project#10082)

adds the following tests to itests/fevm_test.go:
 - recursive tests
 - delegate call tests
 - delegate call recursive tests
 - revert tests
 - destruct tests
 - contract deploy address tests
 - send value to contracts
 - gas limit on value transfer tests
 - sending value to destroyed contracts
adds the test to itests/fevm_address_test.go:
 - deploy contract and confirm address is different second deploy

* merge

* tests that use create2 and destroy to validate evm state

* wip: ipc_gateway in genesis

* deps: Update go-jsonrpc to v0.2.1

* Test: assert all fields returned from EthGetTransactionByHash have expected values

Related to  filecoin-project#10151 (comment)

* Fix merge conflicts

* Allow f4 address to send to all address types if ID address exists on chain

* fix: extend LOTUS_CHAIN_BADGERSTORE_DISABLE_FSYNC to the markset

Without doing this walking a badger markset on a non-nvme knocks the
node hopelessly out of sync during a compaction.

* review fixes

* gateway: Support all EthModule methods

* node builder: Use gateway eth module in lite mode

* feat: ethrpc: Support filtering by address in subscribe

* itests: Fix TestEthFilterAPIDisabledViaConfig

* Check decoding params for new methods

* Review fixes

* feat: evm: align events implementation with FIP-0049 and FIP-0054. (filecoin-project#10152)

- Event keys are now t1, t2, t3, t4 for topics; and d for data.
- ref-fvm no longer stores events in the blockstore for us. It just
  returns events to the client, who is now responsible for handling
  them as it wishes / according to its configuration.
- Add a flag to VMOpts to have the events AMT be written in the blockstore.
- Add a flag to the ChainStore to advertise to the rest of the system
  if the ChainStore is storing events.
- Enable that flag if the EthRPC is enabled (can also add an explicit
  configuration flag if wanted).

* cli: fix extend cmd to get the right sector number

* todo: rebase master to include release/v1.20.0

* feat: compute a better gas limit for recursive external contract calls

* make gen

* fix: ethtypes: Correct 'no uncles' hash in NewEthBlock

* LOTUS_FEVM_ENABLEETHRPC: Fix env variable name in error

* fix: ethtypes: Correct 'no transactions' hash in NewEthBlock

* fix: null rounds: pass correct timestamp to the FVM.

* itest: fix FEVM tests for upstream changes

* itest: fix: test comment

* itest: fix remaining fevm failures

* improve evm error handling in itests (filecoin-project#10161)

* use WithValue language for test

* clean up test for recursive delegate call count. improved readability (filecoin-project#10195)

* chain: explicitly check that gasLimit is above zero

* add bundle git tag from pack.sh into builtin_actors_gen

* fix: worker: add all task type flag

Add all flag for the `lotus-worker tasks enable/disable` cmd

* Update cmd/lotus-worker/tasks.go

Co-authored-by: Łukasz Magiera <[email protected]>

* Update cmd/lotus-worker/tasks.go

Co-authored-by: Łukasz Magiera <[email protected]>

* Update cmd/lotus-worker/tasks.go

Co-authored-by: Łukasz Magiera <[email protected]>

* make docsgen-cli

make docsgen-cli

* retry make docsgen-cli

retry make docsgen-cli with the new usage

* remove bundle-gen from make gen - not actually useful today anyway

* fvm: chore: update FVM

This:

1. Updates the builtin actors bundle (for actors v10).
2. Updates the event entry type to include the codec.
3. Removes the cbor encoding and zero trimming from event data.

I've chose to:

1. _Not_ add codec handling to the event filtering system for now.
2. _Skip_ events with unexpected codecs.

We don't actually _allow_ these events in the FVM right now, and it
simplifies the implementation.

However, I _am_ recording the codecs in the database so we don't have to
migrate it later.

* Option to set-seal-delay seconds

Add the option to set-seal-delay in seconds

* make docsgen-cli

make docsgen-cli

* Error if backup file exists

Error out if backup file already exists

* fix: ethrpc: Don't send sub notifs in array

* Update eth_subscribe itests to work with correct responses

* test: build: ignore git tags in embedded metadata

* fix: build: use actual v9 bundles for butterfly

* fix: ethrpc: emit correct bloom filters (filecoin-project#10194)

* skip problematic test vectors.

* feat: actors: Add bytecode getter

* itests: Test EVM bytecode getters, check bytecode hash

* ethrpc: Add missing fields to EthTx

* Update cli/backup.go

Co-authored-by: Łukasz Magiera <[email protected]>

* feat: eth cli: Strip out empty spaces around contract bytes

* fix: chain: make sure the head is empty, not the code

* bump test vectors.

* remove test vectors bleeding edge job.

We no longer need it because specs-actors is deprecated.
v7 vectors have been merged to master.

* fix: Don't call WalletExport in msg signing flows

* updates butterflynet reset artifacts

* bootstrap node multiaddr
* new genesis file

these files are necessary for others to build lotus and join the new butterflynet, reset on February 08, 2023

* fix: gas: update ffi & correct the message inclusion cost in nv18 (filecoin-project#10228)

Co-authored-by: Raúl Kripalani <[email protected]>

* refactor: use EthHash for event topics

This ensures they're always 32 bytes and padded, as required.

* fix: eth: strict event parsing

We now enforce the following rules:

1. No duplicate topics or data.
2. Topics must have 32 byte keys.
3. Topics may not be skipped. (e.g., no t1 & t3 without a t2).
4. Raw codecs.

We _don't_ require that topics/data be emitted in any specific order.

We _skip_ events with unknown keys. We _drop_ events that violate the
above rules.

* Apply suggestions from code review

Co-authored-by: raulk <[email protected]>

* fix: eth: log on unexpected events

We can remove these later as we add more event types, but this will aid
in debugging.

* fix: make gen

* fix config.yml

* implement itest and handle optional params

* fix typo

* fix: stmgr: make the tipset and height agree when estimating gas (filecoin-project#10216)

* fix: stmgr: make the tipset and height agree when estimating gas

Specifically re-execute all messages in the current tipset, tacking the new
message onto the end. That way, the epoch is the epoch of the current tipset.

We could try to "make" a fake block and use that, but that's unlikely to
work well.

* fix: stmgr: only apply tipset messages for CallWithGas

* fix: itest: window post dispute

* eth_feeHistory: parse block param correctly.

* add todo

* Update to FFI v1.20.0-rc1

* fix bad test.

* improve TODO.

* Update to go-state-types v0.10.0-rc2

* disable adding git tags to bundle metadata

* Update actors to v10.0.0-rc.1

* eth: FIP-0055: implement final version of transitory delegated signature. (filecoin-project#10239)

* make lint happy and re-generate devgen.car

* setup ipc gateway in genesis and assign initial balance

* override with ipc-actors type to load ipc bundle automatically

* set balance of reward actor according to network name and minor fix

* linter fixes

* make linter happy :)

* deploy ipc subnet actor

* fix config.yml

---------

Co-authored-by: Ian Davis <[email protected]>
Co-authored-by: Aayush <[email protected]>
Co-authored-by: ychiao <[email protected]>
Co-authored-by: cryptoAtwill <[email protected]>
Co-authored-by: Phi <[email protected]>
Co-authored-by: Łukasz Magiera <[email protected]>
Co-authored-by: Geoff Stuart <[email protected]>
Co-authored-by: Jiaying Wang <[email protected]>
Co-authored-by: Geoff Stuart <[email protected]>
Co-authored-by: Łukasz Magiera <[email protected]>
Co-authored-by: Jennifer Wang <[email protected]>
Co-authored-by: Ian Davis <[email protected]>
Co-authored-by: Jorropo <[email protected]>
Co-authored-by: Steven Allen <[email protected]>
Co-authored-by: Raúl Kripalani <[email protected]>
Co-authored-by: Richard Guan <[email protected]>
Co-authored-by: Maciej Witowski <[email protected]>
Co-authored-by: omahs <[email protected]>
Co-authored-by: snissn <[email protected]>
Co-authored-by: Peter Rabbitson <[email protected]>
Co-authored-by: 0x5459 <[email protected]>
Co-authored-by: Travis Person <[email protected]>
Co-authored-by: Anton Evangelatov <[email protected]>
Co-authored-by: ognots <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Illegal argument error when calling block.difficulty
5 participants