Repository navigation
chore(quest): drop ts-scalar-ice, fixed within #130 - #162
Conversation
Quest-Claim: 77de3a41c99e0263
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: 47a2d9e
No actionable bugs found. Deleting this stale quest and its index entry is appropriate: a31650f added both the ok_or_error guard and the enum-error fixture cases, and those remain in the reviewed tree. This is simpler than maintaining a second task for an already-covered crash; the deletion changes no runtime code or public API.
Verification: static inspection of the deletion, original fix, current scalar guard, retained splits fixture and driver flags. I did not run the reported reproducer, differential tests or quest checks.
|
Landing: the maintainer decided to delete the quest. Merged main without conflicts; Codex reviewed head 47a2d9e with a thumbs-up and no findings; CI passed. (Written by Claude Opus 5.5) |
Proposes deleting quest/m2/ts-scalar-ice.md. The crash it plans to reproduce was fixed inside #130 itself, before the ICE report was found.
Findings
ok_or_errorand the report. The report's line numbers match that commit's parent, wherescalar_callpassedok_or's error toscalar_operandwithout checking it.ts/fixtures/lang/src/splits.rs(second, andsplit_first().ok_or(err)with an enum error). On currentmain, removing theok_or_errorguard makeslangpanic with the report's backtrace:take_apart(theShape::Enumarm) ←scalar_source←scalar_operand←scalar_call. With the guard, thelangsnapshot and differential tests pass.checked by find_scalarsassumption inscalar.rsagainstdef_ok,producer,ok_or_error, andwalk. Each one is guarded by the same rule. The one mismatch is thatok_or_erroraccepts anOperand::RuntimeCheckserror thatscalar_operandtreats as unreachable. It can't be reached: the driver passes-Zub-checks=no, which foldsub_checks()toconst falsein the translated crate (checked in MIR, both inline and through a local), and std never passes it took_or. Following the rule to reproduce bugs before fixing them, I left it alone.Decisions
main, and the existing fixture proves it. The unreachableRuntimeChecksmismatch stays as is, since there is no crash to reproduce.Public API: none.
Checks:
quest checkok;just ts test --test fixtures langpasses onmain, and fails with the report's backtrace when the guard is removed.🤖 Generated with Claude Code
(Written by Claude Opus 5.5)