[PATCH] D114779: [LV][X86] Sink `LoopVectorizationCostModel::useEmulatedMaskMemRefHack()` further into TTI, disable for X86/AVX2+

[Roman Lebedev via Phabricator via llvm-commits]

lebedev.ri added a comment.

As per @fhahn's "someone more familiar with X86 cost modeling should review the claim the cost model is fixed :)"

@RKSimon given all the costmodel fixes, i claim that nowadays the relevant parts of
the X86 costmodel (at least as of AVX2+) are correct, and the hack is no longer needed.
(D111460 <https://reviews.llvm.org/D111460> already landed under the same pretense.)

Do you agree with my assessment?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D114779/new/

https://reviews.llvm.org/D114779