[PATCH] D29404: [ValueTracking] emit a warning when we detect a conflicting assumption (PR31809)

[Eli Friedman via Phabricator via llvm-commits]

efriedma added a comment.

It seems fine to emit an optimization remark here.  I don't recall the details about the different levels of optimization remarks off the top of my head, though.


https://reviews.llvm.org/D29404