glibc 2.2.5pre1

[Martin Schwidefsky]

> > I tried glibc-2.2.5pre1 on S/390 and found a few things.
We found two more things. The math test suite found reported failures
for j0(4.0) and j0(-4.0), in both cases 1 ulp. I regenerated the
libm-test-ulps file for s/390.
The second thing is a test case failure in linuxthreads found by
Uli (Weigand) while testing with gcc 3.1. The option -fno-inline
does not prevent the automatic inlining of functions and gcc 3.1
inlines a function that it may not inline. -fno-inline-functions
takes case of it.

> Thanks, I've checked this in.  I assume that I'll get a separate patch
> if anything is needed on the trunk.
You'll get a separate patch on libc-hacker shortly.

2002-01-14  Martin Schwidefsky <schwidefsky@de.ibm.com>

     * linuxthreads/Makefile (CFLAGS-tst-cancel.c): Add
     -fno-inline-functions to prevent automatic inline.
     * sysdeps/s390/fpu/libm-test-ulps: Update.

(See attached file: glibc-2.2.5pre1-s390-2.diff)

blue skies,
   Martin

Linux/390 Design & Development, IBM Deutschland Entwicklung GmbH
Schönaicherstr. 220, D-71032 Böblingen, Telefon: 49 - (0)7031 - 16-2247
E-Mail: schwidefsky@de.ibm.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: glibc-2.2.5pre1-s390-2.diff
Type: application/octet-stream
Size: 6192 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20020114/31f901f1/attachment.obj>