jdenny-ornl wrote: > > > Updated to pass `-O2` instead, this was the actual behavior that passed, > > > so it's mostly not a functional change. > > > > > > So you just changed the tests not to use -O1? Doesn't this patch then > > represent a regression in what's supported? > > I wouldn't consider something "supported" if it only worked because we > overruled the user's `-O1` flag and replaced it with `-O2`.
Cases that used to compile successfully (even if more optimized than expected) will no longer. Sounds like a regression. https://github.com/llvm/llvm-project/pull/102972 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
