projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 89315a2) | patch
AVX-512: Fixed a bug in shuffle lowering 32-bit mode
authorElena Demikhovsky <elena.demikhovsky@intel.com>
Thu, 15 Oct 2015 11:35:33 +0000 (11:35 +0000)
committerElena Demikhovsky <elena.demikhovsky@intel.com>
Thu, 15 Oct 2015 11:35:33 +0000 (11:35 +0000)
commit9d0865a1a6166f993e52662bfb42a1b0e35bb9f5
treeef7dea1a2d729dbb7eca030af80e25f9068876ddtree | snapshot
parent89315a2bfd431ce9737f3b48ff53b0ba085c32bfcommit | diff
AVX-512: Fixed a bug in shuffle lowering 32-bit mode

AVX-512 bit shuffle fails on 32 bit since we create a vector of 64-bit constants.
I split 8x64-bit const vector to 16x32 on 32-bit mode.

Differential Revision: http://reviews.llvm.org/D13644

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@250390 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp diff | blob | history
test/CodeGen/X86/vector-shuffle-512-v8.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors