Bug#1099624: libc6 2.41 pthread cause pari to FTBFS

[Bill Allombert <ballombe@debian.org>]

On Thu, Mar 06, 2025 at 02:17:46PM +0100, Bill Allombert wrote:
> On Thu, Mar 06, 2025 at 09:50:35AM +0100, Bill Allombert wrote:
> > On Thu, Mar 06, 2025 at 07:02:25AM +0100, Aurelien Jarno wrote:
> > > Hi Bill,
> > > 
> > > On 2025-03-05 21:30, Bill Allombert wrote:
> > > > Package: libc6
> > > > Version: 2.41-7
> > > > Severity: normal
> > > > 
> > > > Hello Aurelien,
> > > > 
> > > > During a test build pari FTBFS due to a test-suite failure in a test involving pthread.
> > > > this failure is only reproducible with libc6 2.41.
> > 
> > Thanks for looking into this!
> 
> Apparently this happens after a thread does a longjmp to itself.

Maybe I was mistaken. It seems the issue is with pthread_exit, but
maybe this is implemented by longjmp ?

The main failure mode is a double free, but if I remove the free(),
I get instead 

Thread 430 "gp-sta" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffda20c46c0 (LWP 2255037)]
__longjmp_cancel () at ../sysdeps/x86_64/__longjmp.S:147
warning: 147    ../sysdeps/x86_64/__longjmp.S: Aucun fichier ou dossier de ce nom

Cheers,
-- 
Bill. <ballombe@debian.org>

Imagine a large red swirl here.