[CostModel][X86] Add CostKinds handling for fmul ops
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Fri, 2 Sep 2022 15:55:12 +0000 (16:55 +0100)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Fri, 2 Sep 2022 15:57:23 +0000 (16:57 +0100)
commit11765b77be84d793ebedc5b5436c463490746131
tree6b8b79c9b084c5767098eeb64b3271c24a4bd8a8
parenta2ec18ee043f7c8caa0e9e329721182b9f1a5dcd
[CostModel][X86] Add CostKinds handling for fmul ops

This was achieved with an updated version of the 'cost-tables vs llvm-mca' script D103695

As we're using 'typical' worst case values, not all cost entries come from a single CPU - e.g. the latency/throughput from haswell but the size-latency(uops) from zen1/alderlake-e due to 'double pumping'
llvm/lib/Target/X86/X86TargetTransformInfo.cpp
llvm/test/Analysis/CostModel/X86/arith-fp-latency.ll
llvm/test/Analysis/CostModel/X86/arith-fp-sizelatency.ll
llvm/test/Analysis/CostModel/X86/intrinsic-cost-kinds.ll