projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: e2b87de) | patch
[X86][AVX2] Vectorized i16 shift operators
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Mon, 25 May 2015 17:49:13 +0000 (17:49 +0000)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Mon, 25 May 2015 17:49:13 +0000 (17:49 +0000)
commit4da23583b63eb5014454fded95cab1fe651e800e
treebcbc5397cd8166e19fb6b2d816de920f191c4c17tree | snapshot
parente2b87dee1ee19859ea83415eace353fa023bddb0commit | diff
[X86][AVX2] Vectorized i16 shift operators

Part of D9474, this patch extends AVX2 v16i16 types to 2 x 8i32 vectors and uses i32 shift variable shifts before packing back to i16.

Adds AVX2 tests for v8i16 and v16i16

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@238149 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86ISelLowering.cpp diff | blob | history
lib/Target/X86/X86TargetTransformInfo.cpp diff | blob | history
test/CodeGen/X86/avx2-vector-shifts.ll diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors