[PATCH] D25485: [DAG] optimize negation of bool

[Sanjay Patel via llvm-commits]

spatel added a comment.

Thanks, Simon. I'll make the suggested changes and get this in.

Note that the MIPS tests for sdiv/urem/srem with i1 values *are* universal folds (can't divide by zero, so these disappear?), but I'm wondering how those patterns would appear in the backend. We do have IR-level folds for those in InstSimplify.


https://reviews.llvm.org/D25485