projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: a0e8642) | patch
[X86] Disable commuting for the first source operand of zero masked scalar fma intrin...
authorCraig Topper <craig.topper@intel.com>
Wed, 4 Mar 2020 22:33:26 +0000 (14:33 -0800)
committerCraig Topper <craig.topper@intel.com>
Wed, 4 Mar 2020 22:35:53 +0000 (14:35 -0800)
commit6ca96765c7e6f63b45e6c311918a648ef684ea20
treec5b52ad8e67861b76691c2e4787d1bae61c3d2d7tree | snapshot
parenta0e86420ae72e4dc38c3a341943d4f14139e1e4bcommit | diff
[X86] Disable commuting for the first source operand of zero masked scalar fma intrinsic instructions.

I believe this is the correct fix for D75506 rather than disabling all commuting. We can still commute the remaining two sources.

Differential Revision:m https://reviews.llvm.org/D75526
llvm/lib/Target/X86/X86InstrInfo.cpp diff | blob | history
llvm/test/CodeGen/X86/avx512-intrinsics.ll diff | blob | history
Domain: System / Toolchain;