[PATCH] D44644: [X86] Use silvermont cost model overrides for goldmont as well.

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon added a comment.

In https://reviews.llvm.org/D44644#1047756, @craig.topper wrote:

> Turns out almost all of the slow things in the SLM table have been improved in GLM. So this isn't the right table to use. The only thing that didn't change much was floating point division.


I'm hoping to start work on PR36550 reasonably soon - a better approach might be to (a) ensure that the SLM model matches what the TTI says it should and (b) decide how best to provide a GLM model. What do you think?


https://reviews.llvm.org/D44644