Skip to content

Drop fuzz_coverage target from CI#39249

Closed
krinkinmu wants to merge 1 commit intoenvoyproxy:mainfrom
krinkinmu:disable-fuzz-coverage-ci
Closed

Drop fuzz_coverage target from CI#39249
krinkinmu wants to merge 1 commit intoenvoyproxy:mainfrom
krinkinmu:disable-fuzz-coverage-ci

Conversation

@krinkinmu
Copy link
Copy Markdown
Contributor

Commit Message:

I'm trying to switch back to static linking for coverage tests in Envoy CI to work around a bug in Clang/LLVM (see
llvm/llvm-project#32849).

However the change that switches to static linking fails fuzz_coverage with what seems like some kind of resource constraint (e.g., linker gets killed, I suspect, as a result of OOM, since static linking requires more resources).

@yanavalsov suggested that fuzz_coverage target might not be the best way to measure fuzzing coverage (see
#39030 (comment)) and given that I think we should at least consider disabling it until we can make it better. That's what this PR does.

NOTE: @phlax also mention that he did some work to migrate coverage to EngFlow builds in the past and there might be PR that we could finish that will migrate fuzz_coverage to EngFlow. There is a chance that it will address the issue - I'm mentioning it here to give a complete overview of the possible alternatives here.

Additional Description:

You can find some related discussion in #39030 (specifically #39030 (comment), #39030 (comment), #39030 (comment), #39030 (comment) and #39030 (comment)).

@phlax suggested in slack that @adisuissa might be the person more familiar with the coverage targets in Envoy CI and should be involved.

Risk Level: low (in the sense that disabling this target on CI should not break anything, but let's discuss if there are some strategic risks with disabling this test target).
Testing: N/A
Docs Changes: N/A
Release Notes: N/A
Platform Specific Features: N/A

+cc @phlax @yanavlasov @adisuissa

I'm trying to switch back to static linking for coverage tests in Envoy
CI to work around a bug in Clang/LLVM (see
llvm/llvm-project#32849).

However the change that switches to static linking fails `fuzz_coverage`
with what seems like some kind of resource constraint (e.g., linker gets
killed, I suspect, as a result of OOM, since static linking requires
more resources).

@yanavalsov suggested that fuzz_coverage target might not be the best
way to measure fuzzing coverage (see
envoyproxy#39030 (comment))
and given that I think we should at least consider disabling it until we
can make it better. That's what this PR does.

NOTE: @phlax also mention that he did some work to migrate coverage to
EngFlow builds in the past and there might be PR that we could finish
that will migrate fuzz_coverage to EngFlow. There is a chance that it
will address the issue - I'm mentioning it here to give a complete
overview of the possible alternatives here.

Signed-off-by: Mikhail Krinkin <mkrinkin@microsoft.com>
@repokitteh-read-only
Copy link
Copy Markdown

As a reminder, PRs marked as draft will not be automatically assigned reviewers,
or be handled by maintainer-oncall triage.

Please mark your PR as ready when you want it to be reviewed!

🐱

Caused by: #39249 was opened by krinkinmu.

see: more, trace.

@krinkinmu krinkinmu marked this pull request as ready for review April 28, 2025 09:43
@krinkinmu
Copy link
Copy Markdown
Contributor Author

@phlax mentioned that I should really test this PR is staging first, so I'm closing it for now (I cannot move it to back to draft) and will re-open after I can do more testing.

@krinkinmu krinkinmu closed this Apr 28, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant