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(baseapp): Utilizing voting power from VEs in ValidateVoteExtensions #17518

Merged
merged 8 commits into from
Aug 25, 2023

Conversation

davidterpay
Copy link
Contributor

@davidterpay davidterpay commented Aug 23, 2023

Description

Closes: #17517

This PR updates the voting power calculations that are used to verify Vote Extensions in ValidateVoteExtensions. Instead of utilizing the current bonded tokens of a validator, we utilize the voting power provided by the vote extension itself. This ensures the liveness of networks using vote extensions since validator set updates take a few blocks to propagate to comet.


Author Checklist

All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.

I have...

  • included the correct type prefix in the PR title
  • [] added ! to the type prefix if API or client breaking change
  • targeted the correct branch (see PR Targeting)
  • provided a link to the relevant issue or specification
  • followed the guidelines for building modules
  • included the necessary unit and integration tests
  • added a changelog entry to CHANGELOG.md
  • included comments for documenting Go code
  • updated the relevant documentation or specification
  • reviewed "Files changed" and left comments if necessary
  • run make lint and make test
  • confirmed all CI checks have passed

Reviewers Checklist

All items are required. Please add a note if the item is not applicable and please add
your handle next to the items reviewed if you only reviewed selected items.

I have...

  • confirmed the correct type prefix in the PR title
  • confirmed ! in the type prefix if API or client breaking change
  • confirmed all author checklist items have been addressed
  • reviewed state machine logic
  • reviewed API design and naming
  • reviewed documentation is accurate
  • reviewed tests and test coverage
  • manually tested (if applicable)

@davidterpay davidterpay marked this pull request as ready for review August 23, 2023 21:58
@davidterpay davidterpay requested a review from a team as a code owner August 23, 2023 21:58
@github-prbot github-prbot requested review from a team, alexanderbez and julienrbrt and removed request for a team August 23, 2023 21:58
percentSubmitted := math.LegacyNewDecFromInt(sumVP).Quo(math.LegacyNewDecFromInt(totalVP))
if percentSubmitted.LT(VoteExtensionThreshold) {
return fmt.Errorf("insufficient cumulative voting power received to verify vote extensions; got: %s, expected: >=%s", percentSubmitted, VoteExtensionThreshold)
if totalVP > 0 {
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Not sure if this should ever be possible, but adding this as a sanity check.

Copy link
Contributor

Choose a reason for hiding this comment

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

its totalVP a signed integer? lmfao 🤣

I'd flip the logic around, personally. <= 0

Copy link
Contributor

Choose a reason for hiding this comment

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

lol better safe than sorry...

Copy link
Contributor Author

Choose a reason for hiding this comment

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

kek

@davidterpay davidterpay changed the title [Bug Fixes]: Utilizing voting power from VEs in ValidateVoteExtensions fix(baseapp): Utilizing voting power from VEs in ValidateVoteExtensions Aug 23, 2023
CHANGELOG.md Outdated
@@ -52,6 +52,7 @@ Ref: https://keepachangelog.com/en/1.0.0/
### Bug Fixes

* (types) [#16583](https://github.com/cosmos/cosmos-sdk/pull/16583), [#17372](https://github.com/cosmos/cosmos-sdk/pull/17372) Add `MigrationModuleManager` to handle migration of upgrade module before other modules, ensuring access to the updated context with consensus parameters within the same block that executes the migration.
* (baseapp) [#17518](https://github.com/cosmos/cosmos-sdk/pull/17518) Utilizing voting power from vote extensions (comet) instead of the current bonded tokens (baseapp) to determine if a set of vote extensions are valid.
Copy link
Contributor

Choose a reason for hiding this comment

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

Suggested change
* (baseapp) [#17518](https://github.com/cosmos/cosmos-sdk/pull/17518) Utilizing voting power from vote extensions (comet) instead of the current bonded tokens (baseapp) to determine if a set of vote extensions are valid.
* (baseapp) [#17518](https://github.com/cosmos/cosmos-sdk/pull/17518) Utilizing voting power from vote extensions (CometBFT) instead of the current bonded tokens (x/staking) to determine if a set of vote extensions are valid.

@alexanderbez alexanderbez added the backport/v0.50.x PR scheduled for inclusion in the v0.50's next stable release label Aug 24, 2023
@davidterpay
Copy link
Contributor Author

davidterpay commented Aug 24, 2023

@alexanderbez two of the CI jobs are failing but im not sure how to correct them.

  • Detect leaking resources and bad code patterns
  • Warn about consensus code changes

Any guidance there?

@@ -112,19 +119,14 @@ func ValidateVoteExtensions(
return fmt.Errorf("failed to verify validator %X vote extension signature", valConsAddr)
}

sumVP = sumVP.Add(bondedTokens)
Copy link
Contributor

Choose a reason for hiding this comment

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

this is such a footgun. bondedTokens is actually useless in the ValidatorStore interface.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

yea wasn't sure if updating the staking keeper's BondedTokensAndPubKeyByConsAddr would break anything elsewhere. alternatively looks like we could just use ValidatorByConsAddr(context.Context, sdk.ConsAddress) (ValidatorI, error)

Copy link
Contributor Author

Choose a reason for hiding this comment

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

ValidatorI has the pubkey we want

Copy link
Contributor

@itsdevbear itsdevbear left a comment

Choose a reason for hiding this comment

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

nits.

@julienrbrt
Copy link
Member

@alexanderbez two of the CI jobs are failing but im not sure how to correct them.

  • Detect leaking resources and bad code patterns
  • Warn about consensus code changes

Any guidance there?

Those have been broken since the upgrade to go 1.21. Feel free to ignore. cc @elias-orijtech

@@ -86,7 +93,7 @@ func ValidateVoteExtensions(
}

valConsAddr := sdk.ConsAddress(vote.Validator.Address)
bondedTokens, cmtPubKeyProto, err := valStore.BondedTokensAndPubKeyByConsAddr(ctx, valConsAddr)
_, cmtPubKeyProto, err := valStore.BondedTokensAndPubKeyByConsAddr(ctx, valConsAddr)
Copy link
Member

@julienrbrt julienrbrt Aug 24, 2023

Choose a reason for hiding this comment

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

If we actually need to ignore the value, we should delete this method from x/staking and update the ValidatorStore interface.
It was added recently and has never been released (see #17164).

Copy link
Contributor

Choose a reason for hiding this comment

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

yea agreed, though I think we should think in more detail around the footgun of CometBFT power vs whats int he staking module being different and either document the weirdness or define access patterns.

Copy link
Contributor

Choose a reason for hiding this comment

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

Ditto. Let's address this @davidterpay and we can merge.

Copy link
Contributor Author

@davidterpay davidterpay Aug 24, 2023

Choose a reason for hiding this comment

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

I'll add some docs on why we are utilizing the voting power on CometBFT instead of most recent bonded tokens on base app. But at a high level, liveness of the network is dependent on the consensus power each validator has as derived by base app. But this voting power has a delay in terms of relaying to CometBFT. What this means is that for vote extensions to be in sync with the underlying consensus engine, they should default to voting power from the VEs themselves - not from baseapp at the current block. Otherwise you have liveness issues.

Copying this over from the open issue related to this.

Say there are two validators in the network, val1 and val2:

Block H - 1:

  • Val1 has a voting power of 80 (from the perspective of the consensus engine CometBFT)
  • Val2 has a voting power of 20 (from the perspective of the consensus engine CometBFT)

Only val1 signs off on block H - 1 which constitutes a valid block (> 2/3+). In block H-1, there is a redelegation transaction that moves 30 voting power from val1 to val2.

Block H:

  • is currently being constructed and contains the signed vote extensions from Block H - 1.

Within ValidateVoteExtensions,

  • Val1 has voting power of 50 (from the perspective of the application)
  • Val2 has a voting power of 50 (from the perspective of the application)

When validate vote extensions is called in prepare/process proposal, it will see that Val1 only has a voting power of 50 and will reject the set of vote extensions (even though it should still be 80 since this is what comet is utilizing).

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Do you think this should be added to the ADR or within the code itself?

Copy link
Member

@julienrbrt julienrbrt Aug 24, 2023

Choose a reason for hiding this comment

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

I'd say both 😬 or link the ADR in the code for explanation.

@davidterpay
Copy link
Contributor Author

Updated BondedTokensAndPubKeyByConsAddr to GetPubKeyByConsAddr

@alexanderbez alexanderbez enabled auto-merge August 25, 2023 17:40
@alexanderbez alexanderbez added this pull request to the merge queue Aug 25, 2023
Merged via the queue into cosmos:main with commit bb0c866 Aug 25, 2023
mergify bot pushed a commit that referenced this pull request Aug 25, 2023
…ns (#17518)

Co-authored-by: Aleksandr Bezobchuk <alexanderbez@users.noreply.github.com>
(cherry picked from commit bb0c866)

# Conflicts:
#	CHANGELOG.md
#	docs/architecture/adr-064-abci-2.0.md
@elias-orijtech
Copy link
Contributor

@alexanderbez two of the CI jobs are failing but im not sure how to correct them.

  • Detect leaking resources and bad code patterns
  • Warn about consensus code changes

Any guidance there?

Those have been broken since the upgrade to go 1.21. Feel free to ignore. cc @elias-orijtech

These should be upgraded to 1.21 now. Sorry for the trouble.

tac0turtle pushed a commit that referenced this pull request Aug 28, 2023
…ns (backport #17518) (#17552)

Co-authored-by: David Terpay <35130517+davidterpay@users.noreply.github.com>
Co-authored-by: Julien Robert <julien@rbrt.fr>
@lasarojc lasarojc mentioned this pull request Nov 9, 2023
20 tasks
@faddat faddat mentioned this pull request Nov 8, 2024
12 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
backport/v0.50.x PR scheduled for inclusion in the v0.50's next stable release C:x/staking Type: ADR
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[Bug]: ValidateVoteExtensions Voting Power Mismatch
6 participants