This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Don't use IFUNC resolver for longjmp or system in libpthread (bug 21041)


Hi Andreas, Bartłomiej,

Le 06/11/2017 à 14:30, Bartłomiej Piotrowski a écrit :
> On 2017-11-06 09:48, Andreas Schwab wrote:
>> On Nov 01 2017, Romain Naour <romain.naour@gmail.com> wrote:
>>
>>> Hi All,
>>>
>>> Le 08/08/2017 à 16:50, Andreas Schwab a écrit :
>>>> Unlike the vfork forwarder and like the fork forwarder as in bug 19861,
>>>> there won't be a problem when the compiler does not turn this into a tail
>>>> call.
>>>>
>>>> Andreas.
>>>>
>>>> 	* nptl/pt-longjmp.c (longjmp, siglongjmp): Don't use IFUNC resolver.
>>>> 	* nptl/pt-system.c (system): Likewise.
>>>>
>>>
>>> What's the status of this issue ?
>>> This patch has been merged in master (upcoming 2.27) but the ticket is still
>>> "NEW" [1].
>>
>> The bug is about moving a symbol between libraries, as such still
>> unresolved.
>>
>>> It's safe to backport this patch to glibc 2.25 and 2.26 stable branches ?
>>
>> You can try.
>>
>> Andreas.
>>
> 
> I backported this to 2.26 in Arch some time ago. Looks fine on my end.

Thank you both for the feedback.

I did a runtime test with ffmpeg [1] and the "Relink" message is gone.
No regression so far.

[1] http://lists.busybox.net/pipermail/buildroot/2017-October/205865.html

Best regards,
Romain

> 
> Bartłomiej
> 


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]