EricWF marked 2 inline comments as done.
================
Comment at: include/atomic:581
@@ +580,3 @@
+ || __f == memory_order_acq_rel, ""))) \
+ __attribute__ ((__unavailable__("memory order argument to atomic operation
is invalid")))
+#endif
----------------
jfb wrote:
> This isn't checking for the following requirement from the standard:
>
> > The failure argument shall be no stronger than the success argument.
>
> I think that's OK because I intend to remove that requirement from C++ :)
>
> Should we nonetheless enforce the requirement until the standard drops it? If
> so, "stronger" isn't well defined by the standard, details here:
> https://github.com/jfbastien/papers/blob/master/source/D0418r1.bs
Because "stronger" lacks a solid definition I choose not to implement checks
for that requirement. I believe Clang does as well.
> I think that's OK because I intend to remove that requirement from C++ :)
Yet another reason to not implement it.
https://reviews.llvm.org/D22557
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
