projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: f3e9047) | patch
[AMDGPU] Fix lowering of AMDGPU::G_CTTZ_ZERO_UNDEF to AMDGPU::G_AMDGPU_FFBL_B32
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Tue, 17 Aug 2021 16:57:01 +0000 (17:57 +0100)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Tue, 17 Aug 2021 17:09:57 +0000 (18:09 +0100)
commitfb81271e8b44cafff3abf4c05d1485d4b4832e6e
treecbfa127e4e7cd41e833221090cadd441523f765ftree | snapshot
parentf3e9047249d05ff2fb79076dbfbbdad4a35fbc63commit | diff
[AMDGPU] Fix lowering of AMDGPU::G_CTTZ_ZERO_UNDEF to AMDGPU::G_AMDGPU_FFBL_B32

As mentioned on D107474, there was a copy+paste typo repeating G_CTLZ_ZERO_UNDEF that coverity reported as dead code.

Differential Revision: https://reviews.llvm.org/D108210
llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp diff | blob | history
llvm/test/CodeGen/AMDGPU/GlobalISel/regbankselect-cttz-zero-undef.mir diff | blob | history
Domain: System / Toolchain;