diff options
author | Ana Mihajlovic <Ana.Mihajlovic@amd.com> | 2025-03-28 11:20:17 +0100 |
---|---|---|
committer | GitHub <noreply@github.com> | 2025-03-28 11:20:17 +0100 |
commit | f7a034d400860501a26e3429e1c6a9f310f07f76 (patch) | |
tree | 35564571954c2ecd64a4e8b01e7ab74aea6d1cd0 /clang/lib/Frontend/TestModuleFileExtension.cpp | |
parent | 0d64f5adbaea173380668bc4280bb816bdb3a0de (diff) | |
download | llvm-f7a034d400860501a26e3429e1c6a9f310f07f76.zip llvm-f7a034d400860501a26e3429e1c6a9f310f07f76.tar.gz llvm-f7a034d400860501a26e3429e1c6a9f310f07f76.tar.bz2 |
[AMDGPU] (x or y) xor -1 -> x nor y (#130264)
Added pattern so s_nor is selected for ((i1 x or i1 y) xor -1) instead
of s_or and s_xor . This patch is for i1 divergent. The ballot in the
test is added for the retrieval of lanemask. The control flow is needed
because the combiner can't pass through phi instructions.
Diffstat (limited to 'clang/lib/Frontend/TestModuleFileExtension.cpp')
0 files changed, 0 insertions, 0 deletions