[PATCH] D152166: [InstCombine] Expand contractable fmuladd into fmul + fadd

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added a comment.

In D152166#4455742 <https://reviews.llvm.org/D152166#4455742>, @fhahn wrote:

> Is there a test that shows the actual benefit of this, e..g due to further combines? One potential issue is that (parts of) the cost model won't consider `fadd/fmul` pairs as single instruction, so. there may be negative fallout in things like SLPVectorizer.

This is a question for whether we want the original combine here or to remove it. If we're going to do this it at all it should use the minimal required flags


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D152166/new/

https://reviews.llvm.org/D152166