Re: Bug#882158: stretch-pu: package glibc/2.24-11+deb9u2
- To: Aurelien Jarno <email@example.com>, firstname.lastname@example.org, Julien Cristau <email@example.com>
- Cc: Cyril Brulebois <firstname.lastname@example.org>, email@example.com
- Subject: Re: Bug#882158: stretch-pu: package glibc/2.24-11+deb9u2
- From: "Adam D. Barratt" <firstname.lastname@example.org>
- Date: Wed, 14 Feb 2018 21:11:49 +0000
- Message-id: <[🔎] email@example.com>
- In-reply-to: <20180114105231.GA23743@aurel32.net>
- References: <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org> <email@example.com> <20171209132245.GA4899@aurel32.net> <firstname.lastname@example.org> <email@example.com> <20180114105231.GA23743@aurel32.net>
Control: tags -1 + pending
On Sun, 2018-01-14 at 11:52 +0100, Aurelien Jarno wrote:
> On 2018-01-13 17:26, Julien Cristau wrote:
> > Control: tag -1 confirmed
> > On Sat, Dec 9, 2017 at 14:22:45 +0100, Aurelien Jarno wrote:
> > > Unfortunately it didn't make in 9.3 due to the regression
> > > introduced wrt
> > > /etc/ld.so.nohwcap (see bug#883394). The issue is due to the
> > > conversion
> > > of libc6-i686 into a transitional package between jessie and
> > > stretch, and
> > > dropping the postinst and postrm script handling the removal of
> > > /etc/ld.so.nohwcap after the upgrade. The problem always existed
> > > in
> > > stretch, but the probability for it to happen has been greatly
> > > increased
> > > by the fix for #882272. The issue doesn't affect buster/sid as
> > > the
> > > transitional package has been removed.
> > >
> > > I have fixed the issue in version 2.24-11+deb9u3 by reintroducing
> > > the
> > > postinst and postrm scripts in the transitional package. You will
> > > find
> > > below the corresponding patch.
> > >
> > > Thanks for considering it for 9.4.
> > >
> > Assuming that's been tested in all the various scenarios, please go
> > ahead.
> Thanks, I have just uploaded it. Note that the issue with the nvidia
> drivers has been found (providing the default version of the
> without TLS) and has already been fixed in sid.
Flagged for acceptance; sorry for the delay.