projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 2ee4ec6) | patch
AMDGPU: Fix backwards s_cselect_* operands
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Tue, 19 May 2020 14:16:45 +0000 (10:16 -0400)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Wed, 27 May 2020 13:26:09 +0000 (09:26 -0400)
commit833996cef1381115b0077ab5694e189463f5d02e
tree4b371180ff0783129ec9955f525646f7bf997e46tree | snapshot
parent2ee4ec6b6f6d0571288db69b824f6773717d2cf7commit | diff
AMDGPU: Fix backwards s_cselect_* operands

The vector equivalent has backwards operands, but the scalar version
does not. The passes that use these hooks aren't enabled by default,
so this doesn't really change anything.
llvm/lib/Target/AMDGPU/SIInstrInfo.cpp diff | blob | history
llvm/test/CodeGen/AMDGPU/early-if-convert.ll diff | blob | history
Domain: System / Toolchain;