fix bech32 prefix in evidence #8461

yihuang · 2021-01-28T05:26:46Z

Description

When decoding the consensus address, it should use the dynamic configured prefix, rather than the constant.

Problem:

when cosmos based chain customized prefix, double sign behaviour don't get jailed.

Before we can merge this PR, please make sure that all the following items have been
checked off. If any of the checklist items are not applicable, please leave them but
write a little note why.

Targeted PR against correct branch (see CONTRIBUTING.md)
Linked to Github issue with discussion and accepted design OR link to spec that describes this work.
Code follows the module structure standards.
Wrote unit and integration tests
Updated relevant documentation (docs/) or specification (x/<module>/spec/)
Added relevant godoc comments.
Added a relevant changelog entry to the Unreleased section in CHANGELOG.md
Re-reviewed Files changed in the Github PR explorer
Review Codecov Report in the comment section below once CI passes

tomtau · 2021-01-28T05:41:27Z

technically, this may be a consensus-breaking change for non-gaia live networks on 0.41 but probably ok to go with that?

alessio · 2021-01-28T06:39:42Z

Hi @yihuang, and thanks for taking the time the submit this patch and help make Cosmos SDK better.

If the bug is confirmed and it is reproducible in master, we generally accept PRs in master first. Once a PR is fixed in master, we cherry-pick the respective commit and open PRs against the supported stable branches.

Thus, would you mind re-targeting this PR to master please?

yihuang · 2021-01-28T07:19:51Z

Hi @yihuang, and thanks for taking the time the submit this patch and help make Cosmos SDK better.

If the bug is confirmed and it is reproducible in master, we generally accept PRs in master first. Once a PR is fixed in master, we cherry-pick the respective commit and open PRs against the supported stable branches.

Thus, would you mind re-targeting this PR to master please?

Done.

I didn't confirmed in master, because master needs some extra changes for our tests to run. But the patch should be good in master too.

codecov · 2021-01-28T07:28:14Z

Codecov Report

Merging #8461 (4f55c93) into master (784a9a6) will increase coverage by 0.01%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #8461      +/-   ##
==========================================
+ Coverage   61.47%   61.49%   +0.01%     
==========================================
  Files         630      630              
  Lines       36398    36399       +1     
==========================================
+ Hits        22377    22384       +7     
+ Misses      11671    11664       -7     
- Partials     2350     2351       +1

Impacted Files	Coverage Δ
x/evidence/types/evidence.go	`87.50% <100.00%> (+19.75%)`	⬆️

robert-zaremba

I confirm, this indeed causes problem, and the changeset is a proper fix.

robert-zaremba · 2021-01-28T09:24:01Z

@yihuang - thanks for catching this and contributing back to the SDK.

@alessio , indeed , the problem is in the following lines:

evidence := types.FromABCIEvidence(tmEvidence)
k.HandleEquivocationEvidence(ctx, evidence.(*types.Equivocation))

Above, the evidence doesn't have a prefix from config, however, when decoding, in Equivocation.GetConsensusAddress() sdk.ConsAddress, we use config to get the right prefix.

Another fix would be to use []byte instead of string for Equivocation.ConsensusAddress. This way we would avoid bech32 roundtrip at all. Probably we would need to add one more type for "presentation layer".

I approved this PR, because it solves the issue, we can think if we would prefer to alternative solution I mentioned above (I slightly prefer a solution with two types: one for model / operations, and one for presentation) - it's more bulletproof, we avoid bugs with wrong cons/val etc... prefixes.

Finally we have to add a test - ideally in this PR, but I'm OK to do it in a new PR. The test should be more or less following:

evidence := types.FromABCIEvidence(tmEvidence)
consAddr := evidence.GetConsensusAddress()
/// now we do a check:
assert.EqualValues(e.Validator.Address, consAddr)

alessio · 2021-01-28T10:55:56Z

Finally we have to add a test - ideally in this PR, but I'm OK to do it in a new PR

@yihuang feel like adding a test by chance?

yihuang · 2021-01-28T11:46:01Z

unit test added, manually confirmed that without the change, the test fails.

alessio · 2021-01-28T11:49:46Z

@tomtau dixit:

technically, this may be a consensus-breaking change for non-gaia live networks on 0.41 but probably ok to go with that?

This would go right away into v0.42 I suppose

alessio · 2021-01-28T11:52:25Z

@yihuang last bit: can you add a changelog entry please?

@tomtau I was thinking again about your comment, and yes, this is potentially breaking for v0.41 networks, except the Cosmos Hub.

robert-zaremba · 2021-01-28T12:44:27Z

I was thinking again about your comment, and yes, this is potentially breaking for v0.41 networks, except the Cosmos Hub.

Breaking change term is a ambiguous. This change is backward compatible, but creates valid state changes which were not valid before. So it's a soft fork.

alessio · 2021-01-28T12:49:53Z

Yep, the change is potentially state breaking

yihuang · 2021-01-28T14:02:28Z

I was thinking again about your comment, and yes, this is potentially breaking for v0.41 networks, except the Cosmos Hub.

Breaking change term is a ambiguous. This change is backward compatible, but creates valid state changes which were not valid before. So it's a soft fork.

Isn't it a hard fork? old version won't jail the faulting validator, new version will, they won't agree on the state ever after?

yihuang · 2021-01-28T17:08:17Z

@yihuang last bit: can you add a changelog entry please?

Done

Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com>

tac0turtle · 2021-02-02T09:14:53Z

@alessio & @robert-zaremba was this backported?

robert-zaremba · 2021-02-03T00:39:57Z

@marbar3778 - this is scheduled f the upcoming release.

fix bech32 prefix in evidence (cosmos#8461) Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

robert-zaremba · 2021-02-04T17:06:05Z

For the record:

Based on the current release schedule, this won't go into 0.41.x line because it's a state machine breaking change
This doesn't affect Gaia - so we don't need this patch for Gaia update
However, 0.41.x is buggy for anyone else and we need to communicate clearly that nobody except Gaia should use 0.41.x (nor 0.40.x) and should wait for 0.42.x.

robert-zaremba · 2021-02-04T17:10:04Z

cc: @clevinson , @zmanian - check above.

fix bech32 prefix in evidence (cosmos#8461) Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

* add disclaimer for v0.41 series for non-hub SDK chains due to #8461 * update issue link * update issue link, pt2 * fix bad issue link

* add disclaimer for v0.41 series for non-hub SDK chains due to #8461 * update issue link * update issue link, pt2 * fix bad issue link (cherry picked from commit b19ac2c) # Conflicts: # CHANGELOG.md

… (#8708) * Add warning notice in changelog for v0.41.x bug (ref: #8461) (#8707) * add disclaimer for v0.41 series for non-hub SDK chains due to #8461 (cherry picked from commit b19ac2c) Co-authored-by: Cory <cjlevinson@gmail.com> Co-authored-by: Alessio Treglia <alessio@tendermint.com>

Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

fix bech32 prefix in evidence (cosmos#8461) Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

amaury1093 · 2021-03-04T16:15:33Z

@Mergifyio backport release/v0.42.x

Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> (cherry picked from commit d6e4a2e) # Conflicts: # CHANGELOG.md

mergify · 2021-03-04T16:16:55Z

Command backport release/v0.42.x: success

Backports have been created

#8787 fix bech32 prefix in evidence (bp #8461) has been created for branch release/v0.42.x

yihuang requested review from aaronc, alessio and alexanderbez as code owners January 28, 2021 05:26

yihuang force-pushed the fix_bech32_prefix_in_evidence branch from d99db72 to cecc19e Compare January 28, 2021 07:17

yihuang changed the base branch from release/v0.41.x to master January 28, 2021 07:17

yihuang force-pushed the fix_bech32_prefix_in_evidence branch 2 times, most recently from 41e9023 to 03521ef Compare January 28, 2021 08:46

robert-zaremba approved these changes Jan 28, 2021

View reviewed changes

yihuang force-pushed the fix_bech32_prefix_in_evidence branch from 03521ef to 556ed19 Compare January 28, 2021 11:45

yihuang force-pushed the fix_bech32_prefix_in_evidence branch from 556ed19 to d0d778a Compare January 28, 2021 11:47

alessio approved these changes Jan 28, 2021

View reviewed changes

yihuang force-pushed the fix_bech32_prefix_in_evidence branch from d0d778a to 4f55c93 Compare January 28, 2021 17:07

fix bech32 prefix in evidence

c722083

Co-authored-by: Tomas Tauber <2410580+tomtau@users.noreply.github.com>

yihuang force-pushed the fix_bech32_prefix_in_evidence branch from 4f55c93 to c722083 Compare January 28, 2021 17:08

fedekunze added C:x/evidence A:automerge Automatically merge PR once all prerequisites pass. labels Jan 29, 2021

Merge branch 'master' into fix_bech32_prefix_in_evidence

877203a

amaury1093 mentioned this pull request Feb 8, 2021

docs: update master docs #8353

Merged

10 tasks

clevinson mentioned this pull request Feb 10, 2021

Update master CHANGELOG.md with v0.41, enable building docs for v0.41 #8562

Merged

9 tasks

clevinson mentioned this pull request Feb 23, 2021

v0.42 Tracking Issue #8668

Closed

7 tasks

clevinson mentioned this pull request Feb 25, 2021

Add warning notice in changelog for v0.41.x bug (ref: #8461) #8707

Merged

9 tasks

clevinson added a commit that referenced this pull request Feb 25, 2021

add disclaimer for v0.41 series for non-hub SDK chains due to #8461

130d95c

clevinson added a commit that referenced this pull request Feb 26, 2021

Add warning notice in changelog for v0.41.x bug (ref: #8461) (#8707)

b19ac2c

* add disclaimer for v0.41 series for non-hub SDK chains due to #8461 * update issue link * update issue link, pt2 * fix bad issue link

hydrogen18 added a commit to akash-network/cosmos-sdk that referenced this pull request Mar 2, 2021

backport fix cosmos#8461 into this version

4a6dc6e

mergify bot mentioned this pull request Mar 4, 2021

fix bech32 prefix in evidence (bp #8461) #8787

Closed

ethanfrey mentioned this pull request Mar 5, 2021

Provide availability guarantees for IBC Packet data cosmos/ibc-go#74

Closed

4 tasks

This was referenced Mar 11, 2021

Bump github.com/cosmos/cosmos-sdk from 0.40.0-rc5 to 0.42.1 celestiaorg/celestia-app#45

Closed

build(deps): bump github.com/cosmos/cosmos-sdk from 0.41.4 to 0.42.1 maticnetwork/heimdall#600

Closed

tomtau mentioned this pull request Sep 17, 2021

feat: static checking for common bug patterns #10190

Closed

4 tasks

tomtau mentioned this pull request Nov 16, 2021

Static Analyzer for the SDK cosmos/gosec#1

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix bech32 prefix in evidence #8461

fix bech32 prefix in evidence #8461

yihuang commented Jan 28, 2021 •

edited

Loading

tomtau commented Jan 28, 2021

alessio commented Jan 28, 2021 •

edited

Loading

yihuang commented Jan 28, 2021 •

edited

Loading

codecov bot commented Jan 28, 2021 •

edited

Loading

robert-zaremba left a comment

robert-zaremba commented Jan 28, 2021

alessio commented Jan 28, 2021

yihuang commented Jan 28, 2021

alessio commented Jan 28, 2021

alessio commented Jan 28, 2021

robert-zaremba commented Jan 28, 2021

alessio commented Jan 28, 2021 •

edited

Loading

yihuang commented Jan 28, 2021

yihuang commented Jan 28, 2021

tac0turtle commented Feb 2, 2021

robert-zaremba commented Feb 3, 2021

robert-zaremba commented Feb 4, 2021 •

edited

Loading

robert-zaremba commented Feb 4, 2021

amaury1093 commented Mar 4, 2021

mergify bot commented Mar 4, 2021

fix bech32 prefix in evidence #8461

fix bech32 prefix in evidence #8461

Conversation

yihuang commented Jan 28, 2021 • edited Loading

Description

tomtau commented Jan 28, 2021

alessio commented Jan 28, 2021 • edited Loading

yihuang commented Jan 28, 2021 • edited Loading

codecov bot commented Jan 28, 2021 • edited Loading

Codecov Report

robert-zaremba left a comment

Choose a reason for hiding this comment

robert-zaremba commented Jan 28, 2021

alessio commented Jan 28, 2021

yihuang commented Jan 28, 2021

alessio commented Jan 28, 2021

alessio commented Jan 28, 2021

robert-zaremba commented Jan 28, 2021

alessio commented Jan 28, 2021 • edited Loading

yihuang commented Jan 28, 2021

yihuang commented Jan 28, 2021

tac0turtle commented Feb 2, 2021

robert-zaremba commented Feb 3, 2021

robert-zaremba commented Feb 4, 2021 • edited Loading

robert-zaremba commented Feb 4, 2021

amaury1093 commented Mar 4, 2021

mergify bot commented Mar 4, 2021

yihuang commented Jan 28, 2021 •

edited

Loading

alessio commented Jan 28, 2021 •

edited

Loading

yihuang commented Jan 28, 2021 •

edited

Loading

codecov bot commented Jan 28, 2021 •

edited

Loading

alessio commented Jan 28, 2021 •

edited

Loading

robert-zaremba commented Feb 4, 2021 •

edited

Loading