Files
clang-p2996/llvm/test/CodeGen/AMDGPU/llvm.maxnum.f16.ll
Harald van Dijk ceb744eb2f [AMDGPU] Fix canonicalization of truncated values. (#83054)
We were relying on roundings to implicitly canonicalize, which is
generally safe, except with roundings that may be optimized away.

Fixes #82937.
2024-03-13 12:08:39 +00:00

40 KiB