AaronBallman wrote: > It is an ugly situation that a point release fixes one problem but breaks > another.
100% agreed, it's an unfortunate situation to be in. > Preferably get intended behavior in both cases but we don't live in a perfect > world and do live with compromises. Agreed, though there's not much room for compromise here because we either give the correct value or we don't. Is this the only builtin you're having issues with, or is this impacting a bunch of them? If it's just one or two builtins, would using a compiler version check help NVCC avoid the disruption? https://github.com/llvm/llvm-project/pull/111660 _______________________________________________ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
