Re: [patch] Fix for bz14333 -- race between atexit() and exit()

[Paul Pluzhnikov <ppluzhnikov at google dot com>]

On Mon, Sep 25, 2017 at 3:53 PM, H.J. Lu <hjl.tools@gmail.com> wrote:
>
> On 9/20/17, Paul Pluzhnikov <ppluzhnikov@google.com> wrote:
> > On Wed, Sep 20, 2017 at 9:57 AM, Carlos O'Donell <carlos@redhat.com> wrote:
> >> On 09/20/2017 10:50 AM, Joseph Myers wrote:
> >>> I note the commit message just said "Fix BZ 14333".  Can people please
> >>> include the more detailed descriptions of changes made, as in their
> >>> mailing list postings, in the commit message?
> >
> > Sorry about that.
> >
> > (I've been using the ChangeLog entry as the commit message, but I
> > noticed that few others do.)
> >
>
> New tests fail at random on i686:
>
> https://sourceware.org/bugzilla/show_bug.cgi?id=22207


With default "ulimit -s" of 8192, the test can try to create 1024
threads with a total of 8GiB of RAM usage, which is a bit too much for
a 32-bit system.

I'll send a patch.

Thanks,
-- 
Paul Pluzhnikov