Re: [hurd,commited 0/3] glibc 2.44 hurd fixes

Samuel Thibault <[email protected]> Fri, 24 Jul 2026 23:26:41 +0200
Newsgroups gmane.os.hurd.cvs,gmane.comp.lib.glibc.alpha
Organization I am not organized
Message-ID <amPYkbBbsg1nH4Xn@end>
Hello,

Joseph Myers, le ven. 24 juil. 2026 20:14:37 +0000, a ecrit:
> On Tue, 21 Jul 2026, Samuel Thibault wrote:
> 
> > This fixes build/run against the latest 2.44 commits for GNU/Hurd
> > 
> > Samuel Thibault (3):
> >   hurd: Add __fstatat64 to dl-sysdep.c
> >   hurd: Add __readlink and readlink to dl-sysdep.c
> >   hurd: demangle pointers on fork's longjmp
> > 
> >  sysdeps/mach/hurd/Versions             |  4 +-
> >  sysdeps/mach/hurd/dl-sysdep.c          | 53 ++++++++++++++++++++++++++
> >  sysdeps/mach/hurd/i386/ld.abilist      |  1 +
> >  sysdeps/mach/hurd/i386/localplt.data   |  2 +
> >  sysdeps/mach/hurd/i386/longjmp-ts.c    |  5 +++
> >  sysdeps/mach/hurd/x86_64/ld.abilist    |  1 +
> >  sysdeps/mach/hurd/x86_64/localplt.data |  2 +
> >  sysdeps/mach/hurd/x86_64/longjmp-ts.c  |  5 +++
> >  8 files changed, 72 insertions(+), 1 deletion(-)
> 
> My bot is showing localplt test failures because the expected references 
> to readlink and __fstatat64 don't show up.
> 
> https://sourceware.org/pipermail/libc-testresults/2026q3/014839.html

Oh, I didn't think it could depend on the compiler here. I have made it
optional.

Thanks,
Samuel