krzysz00 wrote: v. In the context of this PR, would you be okay with me dropping the whole non-integral section and for now pretending that AMDGPU pointers are "unstable" to keep the current optimization restrictions in place
Yeah, that seems reasonable, just stick a comment somewhere pointing out that the instability is a lie. https://github.com/llvm/llvm-project/pull/105735 _______________________________________________ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
