Re: [PR19826] fix non-LE TLS in static programs

[Florian Weimer <fweimer at redhat dot com>]

On 12/02/2016 05:45 PM, Carlos O'Donell wrote:
> On 12/02/2016 11:32 AM, Florian Weimer wrote:
> > On 12/02/2016 05:25 PM, Carlos O'Donell wrote:
> >
> > > > This patch broke non-optimized global-dynamic TLS on aarch64
> > > > (for shared builds), as shown by my new tst-tls-manydynamic test
> > > > case.
> > >
> > > Does reverting Alex's patch fix tst-tls-manydynamic?
> >
> > It's sufficient to revert the patch to nptl/allocatestack.c.  I don't
> > know if it is really needed, I and I have no idea how to make
> > initialization at this point safe because the threads themselves
> > could be using the DTV concurrently, so we cannot enlarge it.
> >
> > I filed <https://sourceware.org/bugzilla/show_bug.cgi?id=20915> so
> > that we keep track of this regression and fix it before the release.
>
> You don't quite answer my question.

Well, obviously, because as I wrote, the patch broke the 
tst-tls-manydynamic test case.

> If bug 20915 is a release blocker because of the new regression, and
> I think it should be, then we should immediately add it to the release
> blockers list for 2.25:
>
> https://sourceware.org/glibc/wiki/Release/2.25

Thanks, I added it.

Florian