diff options
| author | Matthias Krüger <matthias.krueger@famsik.de> | 2023-10-14 19:22:17 +0200 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-10-14 19:22:17 +0200 |
| commit | a8cda30f320189b5746b798d357930c4295199f6 (patch) | |
| tree | b4c29d6323238d507ea3d1766e6e0113005aa5e0 /tests/codegen/array-codegen.rs | |
| parent | 4dd4d9b489d71b8020eedea1001ad2b0904601be (diff) | |
| parent | 58d62fc271457a5d1cfdfc947c5039df6326a45e (diff) | |
| download | rust-a8cda30f320189b5746b798d357930c4295199f6.tar.gz rust-a8cda30f320189b5746b798d357930c4295199f6.zip | |
Rollup merge of #116591 - Zalathar:flaky-hash, r=Mark-Simulacrum
Don't accidentally detect the commit hash as an `fadd` instruction I've seen some reports of `tests/codegen/target-feature-inline-closure.rs` spuriously failing because it thinks the hash in the rustc version number contains an `fadd` instruction. https://github.com/rust-lang/rust/pull/116085#issuecomment-1751174916 https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/Is.20.60tests.2Fcodegen.2Ftarget-feature-inline-closure.2Ers.60.20flakey https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/Strange.20.5Cn.20in.20output.20of.20assert.20.23108341/near/395811335 This PR tries to make that not happen by adding a `CHECK-LABEL` directive that will match the line with the rustc version string, preventing the previous `CHECK-NOT` from seeing it.
Diffstat (limited to 'tests/codegen/array-codegen.rs')
0 files changed, 0 insertions, 0 deletions