[glibc] s390x: Regenerated ULPs

[Paul Zimmermann]

> From: Florian Weimer <fweimer@redhat.com>
> Cc: libc-alpha@sourceware.org
> Date: Thu, 09 Jan 2025 12:11:35 +0100
> 
> * Stefan Liebler via Glibc-cvs:
> 
> > commit 82688ca3107c93d9eba6440981e473d1877b6281
> > Author: Stefan Liebler <stli@linux.ibm.com>
> > Date:   Thu Jan 9 10:02:21 2025 +0100
> >
> >     s390x: Regenerated ULPs
> >     
> >     ULPs update needed if build with gcc 11.5 or 10.5.
> >
> > Diff:
> > ---
> >  sysdeps/s390/fpu/libm-test-ulps | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/sysdeps/s390/fpu/libm-test-ulps b/sysdeps/s390/fpu/libm-test-ulps
> > index 76a1f3c7e5..92e2cc7f77 100644
> > --- a/sysdeps/s390/fpu/libm-test-ulps
> > +++ b/sysdeps/s390/fpu/libm-test-ulps
> > @@ -7,10 +7,12 @@ ldouble: 1
> >  
> >  Function: "acos_downward":
> >  double: 1
> > +float: 1
> >  ldouble: 1
> >  
> >  Function: "acos_towardzero":
> >  double: 1
> > +float: 1
> >  ldouble: 1
> >  
> >  Function: "acos_upward":
> 
> The acosf function is expected to be correctly rounded for all rounding
> modes, so this is an actual bug.

this might not be the case with gcc < 12 due to
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=57245

Stefan, please can you send the output of the failing tests?

Paul