[PATCH] D98243: [LV] Account for the cost of predication of scalarized load/store
Sjoerd Meijer via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue Mar 16 06:57:14 PDT 2021
SjoerdMeijer accepted this revision.
SjoerdMeijer added a comment.
This revision is now accepted and ready to land.
This looks very reasonable to me.
PS Nothing wrong with a descriptive function name: `useEmulatedMaskMemRefHack`:-)
================
Comment at: llvm/lib/Transforms/Vectorize/LoopVectorize.cpp:6781
+ Cost += TTI.getScalarizationOverhead(
+ Vec_i1Ty, APInt::getAllOnesValue(VF.getKnownMinValue()), false, true);
+ Cost += TTI.getCFInstrCost(Instruction::Br, TTI::TCK_RecipThroughput);
----------------
Nit: I see there's precedent of not doing it in this file, but
.., false /* Insert */, true /* Extract */);
would have helped readability for me.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D98243/new/
https://reviews.llvm.org/D98243
More information about the llvm-commits
mailing list