[CostModel][X86] Improve sum reduction costs.
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Sat, 12 Oct 2019 13:21:50 +0000 (13:21 +0000)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Sat, 12 Oct 2019 13:21:50 +0000 (13:21 +0000)
commit1b59a16c0b8bd89445b97eee0fde2a152001dbaf
tree4a8db0f7b553bc9e38ed955915a44b71eb0df974
parentba229557dd208a39086c5e2bd8db619dca7d9b45
[CostModel][X86] Improve sum reduction costs.

I can't see any notable differences in costs between SSE2 and SSE42 arches for FADD/ADD reduction, so I've lowered the target to just SSE2.

I've also added vXi8 sum reduction costs in line with the PSADBW codegen and discussions on PR42674.

llvm-svn: 374655
llvm/lib/Target/X86/X86TargetTransformInfo.cpp
llvm/test/Analysis/CostModel/X86/reduce-add.ll
llvm/test/Analysis/CostModel/X86/reduction.ll
llvm/test/Transforms/SLPVectorizer/X86/remark_horcost.ll