[PATCH] D102934: [unroll] Use value domain for symbolic execution based cost model

[Nikita Popov via Phabricator via llvm-commits]

nikic accepted this revision.
nikic added a comment.
This revision is now accepted and ready to land.

In D102934#2778255 <https://reviews.llvm.org/D102934#2778255>, @reames wrote:

> If you're happy with the direction, do you mind if I land them without separate review?  i.e. Can I interpret this as an LGTM for the split patches?
>
> I'd mildly prefer not splitting the patches, but I'll do whatever will minimize the review back-and-forth.  :)

Okay, I don't feel particularly strongly about this, so let's say this LGTM modulo the note about `ConstantExpr::getCompare()`.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D102934/new/

https://reviews.llvm.org/D102934