On Wed, Apr 22, 2026 at 06:55:14PM +0200, Jakub Jelinek wrote:
> On Wed, Apr 22, 2026 at 05:41:22PM +0100, Alice Carlotti wrote:
> > On Wed, Apr 22, 2026 at 04:55:49PM +0200, Jakub Jelinek via Gcc wrote:
> > > We have branched for the GCC 16 release.  All changes on the 
> > > releases/gcc-16
> > > branch require release manager approval now.
> > 
> > I pushed some bug fixes to master just after the branch.  There are 7 
> > commits
> > in total; two bug fixes, one testsuite fix, and four miscellaneous
> > improvements.  Are they ok for the branch now, or do I need to wait until 
> > after
> > the release?
> > 
> > Misc.:    e65f4568e9c (upstream/trunk, upstream/master) aarch64 testsuite: 
> > Merge exts_sve2 into exts
> > Test fix: 478ec2ec373 aarch64 testsuite: Fix gating of sme-lutv2 asm tests
> > Misc.:    5a9ddd6860e Output mode switching transitions to dump file
> > Misc.:    378c191c283 aarch64: Extend comment about saving/restoring zt0
> > Misc.:    fabd6595c67 Fix some mode switching doc/comment typos
> > Bug fix:  960c8597962 aarch64: Fix ZA state transition [PR119210]
> > Bug fix:  3d13dace31d aarch64: Add ZTO_REGNUM and SME_STATE_REGNUM use to 
> > movt
> 
> At this point we ideally don't want to change the branch at all, or change
> it only when gcc doesn't build at all on important target, ABI problems are
> found or some P1 appears.
> I think none of the above are of that kind, so they can wait for 16.2

Ok, agreed on that.

> (if
> they should be backported at all, not judging any of that, just mentioning
> that miscellaneous improvement category maybe doesn't match the branch
> criteria (regression bugfixes or documentation fixes)).

Three of them are comment/doc fixes that I think were explicitly approved for
backport in an earlier review.  I thought Pinski had approved backporting the
dump change (to my surprise - I wasn't asking for it), but I actually had just
misread the email.

I'll leave all of the backporting (including earlier versions) until after the
release.

Alice

> 
>       Jakub
> 

Reply via email to