projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: cf44372) | patch
[X86] Don't turn avx masked.load with constant mask into masked.load+vselect when...
authorCraig Topper <craig.topper@intel.com>
Thu, 6 Jun 2019 05:41:27 +0000 (05:41 +0000)
committerCraig Topper <craig.topper@intel.com>
Thu, 6 Jun 2019 05:41:27 +0000 (05:41 +0000)
commit9226ba6b376ea2a221e97c2f674841a496869f4d
tree4dbad74ee134cbc4f7fc9e18a54f86f44ae2e833tree | snapshot
parentcf44372137f245b9ce940e12bbaedfe2c23a3bc4commit | diff
[X86] Don't turn avx masked.load with constant mask into masked.load+vselect when passthru value is all zeroes.

This is intended to enable the use of an immediate blend or
more optimal instruction. But if the passthru is zero we don't
need any additional instructions.

llvm-svn: 362675
llvm/lib/Target/X86/X86ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/X86/masked_load.ll diff | blob | history
Domain: System / Toolchain;