[flang-commits] [flang] [flang] Skip optimized bufferization on volatile refs (PR #140781)
Asher Mancinelli via flang-commits
flang-commits at lists.llvm.org
Wed May 21 06:49:43 PDT 2025
ashermancinelli wrote:
> @jeanPerier
>
> To clarify, I am not asking you to update the patch to ignore VOLATILE effects here (we would still need a bit more thinking to make sure things are OK for the cases more complex than x = x +y from my example). I just do not want future us to believe it would be incorrect to do so if we need do.
Thank you for the thoughtful review! I wonder if there's somewhere I should collect those ideas along with the design of volatile so we can iterate on it (or at least keep it documented somewhere). Maybe something in `docs/`?
https://github.com/llvm/llvm-project/pull/140781
More information about the flang-commits
mailing list