[PATCH] D100072: [AMDGPU] Allow -amdgpu-unsafe-fp-atomics to ignore denorm mode
Tony Tye via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed Apr 7 16:57:05 PDT 2021
t-tye added inline comments.
================
Comment at: clang/docs/ClangCommandLineReference.rst:3002
+
+Enable unsafe floating point atomic instructions. Bypasses denormal mode checks and allows to use atomics on a potentially unsupported memory types. (AMDGPU only)
+
----------------
Enable generation of unsafe floating point atomic instructions. May generate more efficient code, but may not respect rounding and denormal modes, and may give incorrect results for certain memory destinations. (AMDGPU)
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D100072/new/
https://reviews.llvm.org/D100072
More information about the llvm-commits
mailing list