[PATCH] D61638: [LV] Move getScalarizationOverhead and vector call cost computations to CM. (NFC)

[Florian Hahn via Phabricator via llvm-commits]

fhahn created this revision.
fhahn added reviewers: Ayal, hsaito, dcaballe, rengolin.
Herald added subscribers: rogfer01, rkruppe, hiraditya.
Herald added a project: LLVM.

This reduces the number of parameters we need to pass in and they seem a
natural fit in LoopVectorizationCostModel. Also simplifies things for
D59995 <https://reviews.llvm.org/D59995>.

As a follow up refactoring, we could only expose a expose a
shouldUseVectorIntrinsic() helper in LoopVectorizationCostModel, instead
of calling getVectorCallCost/getVectorIntrinsicCost in
InnerLoopVectorizer/VPRecipeBuilder.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D61638

Files:
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/lib/Transforms/Vectorize/VPRecipeBuilder.h

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D61638.198428.patch
Type: text/x-patch
Size: 9741 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190507/3eee06b9/attachment.bin>