On Fri, Mar 9, 2018 at 6:55 PM Andre McCurdy <[email protected]> wrote:
> On Fri, Mar 9, 2018 at 6:25 PM, Bystricky, Juro > <[email protected]> wrote: > > This was morty upgrade, which did not have those patches to begin with, > I did not remove them. The patches were not in rocko either, in the end I > got them from master (gcc6.4 was eventually removed completely), > > Backporting of ARM specific patches was not the objective. ( I was not > aware that 6.4 had an ARM specific regression which broke the pi3 builds). > > Long story short, all patches should be there now, please review them, > see the commit message for more details. > > I'll make sure all other gcc6 upgrades contain the ARM patches as well. > > Comparing my morty branch (with update to gcc 6.4 by backporting > incremental gcc updates from master) and today's morty-next (which > updates gcc 6.2 -> 6.4 with a single commit) the only differences are: > > - My branch lacks all the new x86 specific patches (OK, since they > are appearing here first) > > - Today's morty-next is lacking one patch, which was added to oe-core > master as a backport (from gcc master) to gcc 6.3: > > > http://git.openembedded.org/openembedded-core/commit/?id=d2631f45a057c53797b7ba657662f35f66a2b04e > > I don't have any strong opinions on the patch, but presumably it fixed > some issue Khem was seeing, it still applies cleanly to gcc 6.4 and we > have it in rocko so on balance I'd say it should probably be in morty > too. > Yes this patch fixes an arm regression > > Overall this looks great though and I'm very happy to see gcc updates > landing in the release branches. Thanks Juro! >
-- _______________________________________________ Openembedded-core mailing list [email protected] http://lists.openembedded.org/mailman/listinfo/openembedded-core
