[PATCH] D34174: [x86] replace div/rem with shift/mask for shuffle combining

[Sanjay Patel via Phabricator via llvm-commits]

spatel added a comment.

In https://reviews.llvm.org/D34174#779996, @chandlerc wrote:

> This makes lots of sense to me. Honestly, fixing the multiplies might be worth it -- if you can benchmark >1% improvement, I might do it. I'm thinking about the architectures where integer multiplies are *very* slow (atom at least, also some ARM chips)


Thanks for the review! Let me mark the multiplies with a TODO in this commit, and then I'll check that as a follow-up. I suspect that we can probably do better still by avoiding this entirely in some cases, but I need to run some more experiments.


https://reviews.llvm.org/D34174