projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 7cc8377) | patch
[CostModel][X86] Adjust pre-SSE41 fp scalar select costs to account for vector ops
authorSimon Pilgrim <llvm-dev@redking.me.uk>
Fri, 6 May 2022 10:41:45 +0000 (11:41 +0100)
committerSimon Pilgrim <llvm-dev@redking.me.uk>
Fri, 6 May 2022 10:41:55 +0000 (11:41 +0100)
commitd21bf514940fd3b6368796e3ad22e1910c8598c6
treeb3a1443ae80e64c6e64e353e9b823ffc4a398268tree | snapshot
parent7cc8377f2c572a919ecb2d22b2039acf3e6b745acommit | diff
[CostModel][X86] Adjust pre-SSE41 fp scalar select costs to account for vector ops

Based off the script from D103695, we now mainly use BLENDV or OR(AND,ANDN) to select scalar float/double ops
llvm/lib/Target/X86/X86TargetTransformInfo.cpp diff | blob | history
llvm/test/Analysis/CostModel/X86/vselect-cost.ll diff | blob | history
llvm/test/Transforms/SLPVectorizer/X86/crash_cmpop.ll diff | blob | history
Domain: System / Toolchain;