jfb requested changes to this revision.
jfb added a comment.
This revision now requires changes to proceed.
Given http://wg21.link/p0418 I think this requires an update. I don't think the
old behavior is worth supporting in older `-std=` versions, unless we're
worried that it would make code les
jfb added a comment.
Warnings look correct. Could they be made more useful by suggesting that
`acq_rel` be `acquire` instead, and `release` be `relaxed` instead?
https://reviews.llvm.org/D22711
___
cfe-commits mailing list
cfe-commits@lists.llvm.or
EricWF added a comment.
ping. Clang correctly lowers the invalid memory orders, so do we want these
diagnostics?
https://reviews.llvm.org/D22711
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/
EricWF created this revision.
EricWF added reviewers: majnemer, jfb, rsmith.
EricWF added a subscriber: cfe-commits.
Clang currently issues diagnostics if the success memory order argument is
invalid, but it doesn't check the failure ordering argument for cmpxchg atomic
operations.
This patch ad
