fitzgen opened PR #13381 from fitzgen:kotlin-benchmark-bug to bytecodealliance:main:
See each commit for details.
fitzgen requested cfallin for a review on PR #13381.
fitzgen requested wasmtime-compiler-reviewers for a review on PR #13381.
fitzgen requested wasmtime-core-reviewers for a review on PR #13381.
fitzgen updated PR #13381.
:thumbs_up: alexcrichton submitted PR review.
:speech_balloon: alexcrichton created PR review comment:
This local is never set in this test? (the branching condition is a bit confusing to me, not sure what the intention was )
:speech_balloon: alexcrichton created PR review comment:
ditto with lack of set
:speech_balloon: alexcrichton created PR review comment:
Similar to above, this local is never set?
:memo: fitzgen submitted PR review.
:speech_balloon: fitzgen created PR review comment:
It is there to make sure that the loop is an actual loop, but that the second use+assert is not on the back edge, which was required to trigger this bug. I'll tweak things to make that more clear, and we don't need a local that is implicitly zero to do that.
fitzgen updated PR #13381.
fitzgen has enabled auto merge for PR #13381.
fitzgen added PR #13381 Fix the copying collector bug(s) triggered by the Kotlin benchmarks to the merge queue.
:check: fitzgen merged PR #13381.
fitzgen removed PR #13381 Fix the copying collector bug(s) triggered by the Kotlin benchmarks from the merge queue.
Last updated: Jun 01 2026 at 09:49 UTC