[PATCH] D64285: [InstCombine] Fold select (icmp sgt x, -1), lshr (X, Y), ashr (X, Y) to ashr (X, Y))

[Sanjay Patel via Phabricator via llvm-commits]

spatel added a comment.

In D64285#1580944 <https://reviews.llvm.org/D64285#1580944>, @nikic wrote:

> So to answer your question, it's about both code duplication and redundant matching. And while the death by a thousand cuts may be unavoidable, we should still try to not hasten it along unduly...


I agree. Let's not add extra code for the 'exact' case if we don't need to. Ie, if we can do all of these transforms in instcombine alone by dropping the 'exact' flag, do that. Put another way: unless there's evidence that dropping the exact flag is going to cause a missed optimization, go with the 'less code is better' solution. We have precedence in instcombine/instsimplify for that trade-off.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D64285/new/

https://reviews.llvm.org/D64285