This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Fix va_start on x86_64 after any named argument passed on the stack that needed padding (PR target/44942)
This is OK for mainline and branches (after a couple of days,
perhaps). Even if it is not a regression, the change is well isolated
and without impact on other compiler functionality, so I guess that
wrong-code fix outweights relatively small risk.
Agreed, we always tried to backport API breakages back to branches as
soon as possible.