On Mon, May 02, 2016 at 10:03:51AM -0600, Chris Murphy wrote:
> On Mon, May 2, 2016 at 8:58 AM, Dominik 'Rathann' Mierzejewski
> <dominik(a)greysector.net> wrote:
> > On Monday, 02 May 2016 at 15:24, Stephen Gallagher wrote:
> > [...]
> >> All of the major stakeholders that usually trigger a mass rebuild (GCC,
glibc,
> >> etc.) have been notified directly and are on board with this. This
announcement
> >> was to ensure that no one was left surprised by this in case we missed
telling
> >> anyone directly.
> >
> > I'm not so sure about GCC. It's been updated to 6.1 in both F25 and
F24
> > and it's causing bugs already. For example, anything trying to link
> > against boost built with gcc-6.0.0 will fail to link at the moment, see
> > bug
https://bugzilla.redhat.com/show_bug.cgi?id=1331983 .
>
> It's built in koji for F25 and F24, but it's not been bodhi pushed to
> updates-testing for either. I think this is the usual and expected way
> things go?
There are no known ABI incompatibilities between gcc-6.1.1 (i.e. snapshot
a few hours after the GCC 6.1 release) and recent 6.0.0 snapshots
(development snapshots months to weeks before the 6.1 release that happened
last week). So, are you sure it is really in between
gcc-6.0.0-0.20.fc{24,25} and gcc-6.1.1-1.fc{24,25}, and not something you
just attribute to newer gcc because you see newer version on it?
failed.
Rebuilding boost makes it succeed again. In the first case,
6.0.0-0.20.fc25 was in the buildroot. 6.1.1-1.fc25 was in the latter.
Boost was the same in both (1.60.0-5.fc25).
Regards,
Dominik
--
Fedora
"Faith manages."
-- Delenn to Lennier in Babylon 5:"Confessions and Lamentations"