Landing: d338c42d76fa
Project / Subsystem
gcc / amdgcn
Date
2026-07-09
Author
Andrew Stubbs
Commit
d338c42d76fa7ddec361fc344bf0c6add63c306b
Source
github
Perf win
No
Breaking
No
All attributes
- project
- gcc
- subsystem
- amdgcn
- patch_id
- —
- commit_hash
- d338c42d76fa7ddec361fc344bf0c6add63c306b
- source_type
- github
- headline
- AMDGCN testsuite fixes for conditional floating-point operations
- tldr
- Typographical errors in AMDGCN test cases for `cond_fmaxnm` and `cond_fminnm` are corrected.
- author
- Andrew Stubbs
- outcome
- committed
- performance_win
- false
- breaking_change
- false
- series_id
- —
- series_parts
- []
- tags
-
- • amdgcn
- • gpu
- • testsuite
- • bugfix
- discussion_id_link
- —
- bugzilla_pr
- —
- date
- 2026-07-09T00:00:00.000Z
GCC’s AMDGPU (AMDGCN) testsuite has been fixed to correct typos in test cases related to conditional floating-point minimum and maximum operations. The tests cond_fmaxnm_3.c, cond_fmaxnm_4.c, cond_fminnm_3.c, and cond_fminnm_4.c now correctly scan for SFmode and DFmode operations, resolving issues that previously caused incorrect assembler scan patterns.