This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Near 1000 arm-eabi neon FAIL in C testsuite (PR38697 and PR39361)
- From: Laurent GUERBY <laurent at guerby dot net>
- To: Richard Guenther <richard dot guenther at gmail dot com>
- Cc: Paul Brook <paul at codesourcery dot com>, Richard Earnshaw <rearnsha at arm dot com>, gcc <gcc at gcc dot gnu dot org>, John David Anglin <dave at hiauly1 dot hia dot nrc dot ca>
- Date: Wed, 11 Mar 2009 18:22:08 +0100
- Subject: Re: Near 1000 arm-eabi neon FAIL in C testsuite (PR38697 and PR39361)
- References: <1236547150.11347.1924.camel@localhost> <84fc9c000903081445x5ab2e60do9ba2a9a9fe4fa661@mail.gmail.com>
On Sun, 2009-03-08 at 22:45 +0100, Richard Guenther wrote:
> On Sun, Mar 8, 2009 at 10:19 PM, Laurent GUERBY <laurent@guerby.net> wrote:
> > Hi,
> >
> > Trunk on armv5tel-linux-gnueabi (compile farm gcc50) currently fails
> > about 1000 C tests:
> >
> > http://gcc.gnu.org/ml/gcc-testresults/2009-03/msg00810.html
> >
> > 99% of those fail are neon FAIL:
> >
> > ...
> > FAIL: gcc.target/arm/neon/vst4_laneu32.c scan-assembler vst4.32[ \\t]+\\
> > \\{(([dD][0-9]+\\\\[[0-9]+\\\\]-[dD][0-9]+\\\\[[0-9]+\\\\])|([dD][0-9]+\
> > \\\[[0-9]+\\\\], [dD][0-9]+\\\\[[0-9]+\\\\], [dD][0-9]+\\\\[[0-9]+\\\\],
> > [dD][0-9]+\\\\[[0-9]+\\\\]))\\\\}, \\\\[[rR][0-9]+\\\\]!?([ \
> > \t]+@[a-zA-Z0-9 ]+)?\\n
> > FAIL: gcc.target/arm/neon/vst4_laneu8.c (test for excess errors)
> > FAIL: gcc.target/arm/neon/vst4_laneu8.c scan-assembler vst4.8[ \\t]+\\\\{(([dD][0-9]+\\\\[[0-9]+\\\\]-[dD][0-9]+\\\\[[0-9]+\\\\])|([dD][0-9]+\\\\[[0-9]+\\\\], [dD][0-9]+\\\\[[0-9]+\\\\], [dD][0-9]+\\\\[[0-9]+\\\\], [dD][0-9]+\\\\[[0-9]+\\\\]))\\\\}, \\\\[[rR][0-9]+\\\\]!?([ \\t]+@[a-zA-Z0-9 ]+)?\\n
> >
> > A while ago I opened (but there were far less neon FAIL):
> >
> > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38697
> >
> > There is also:
> >
> > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39361
> >
> > arm-eabi is listed as primary platform here:
> >
> > http://gcc.gnu.org/gcc-4.4/criteria.html
> >
> > Do you know why all these tests are failing?
>
> See http://gcc.gnu.org/ml/gcc-patches/2009-03/msg00442.html
r144713, just after the above patch commit, indeed has
far less neon failures:
http://gcc.gnu.org/ml/gcc-testresults/2009-03/msg01066.html
I still have received no answer or hint about why
some neon tests are still failing:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=38697
Sincerely,
Laurent