This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: libm-test-ulps
- From: Andreas Jaeger <aj at suse dot de>
- To: Greg Schafer <gschafer at zip dot com dot au>
- Cc: libc-alpha at sources dot redhat dot com
- Date: Tue, 09 Jul 2002 08:37:38 +0200
- Subject: Re: libm-test-ulps
- References: <20020709145432.A446@tigers-lfs.nsw.bigpond.net.au>
Greg Schafer <gschafer@zip.com.au> writes:
> Hi
>
> I'm currently testing:-
>
> gcc-3.1 (release)
> binutils-2.12.1 (release)
> glibc-2.2.5 (release)
>
> for i686-pc-linux-gnu
>
> on an Intel Celeron 533 with everything compiled as -march=i686
> by default.
>
> "make check" for glibc fails a couple of the math tests.
>
> I followed the instructions in the source to construct a new
> "libc/sysdeps/i386/fpu/libm-test-ulps" file.
>
> I've attached it to this message just in case it is useful
> to you.
Those errors might be an indication of non-stable algorithms, rounding
erros in the CPU, and problems in the test framework. Since the math
tests have been rewritten for glibc 2.3, it doesn't make sense to
commit this to mainline - and we're not going to apply such patches to
glibc 2.2.x.
>
> Thanks for working on glibc.
>
> Greg
>
> ps - any idea if a glibc-2.2.6 release is on the horizon?
Ulrich has to answer this. I would like to see a release soon,
Andreas
--
Andreas Jaeger
SuSE Labs aj@suse.de
private aj@arthur.inka.de
http://www.suse.de/~aj