This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: Testsuite failure ?
- To: gcc-bugs at gcc dot gnu dot org
- Subject: Re: Testsuite failure ?
- From: David Ronis <ronis at onsager dot chem dot mcgill dot ca>
- Date: Fri, 25 Aug 2000 08:56:18 -0400 (EDT)
- Cc: manfredh at redhat dot com, aoliva at redhat dot com
- Reply-To: David Ronis <ronis at onsager dot chem dot mcgill dot ca>
You were right, it's the extra alignment flags that are causing the problems.
A bootstrap with
BOOT_CFLAGS=-O3 -march=i686 -funroll-loops -ffast-math -fomit-frame-pointer
eliminates all the libio and libstc++ testsuite failures, leaves the gcc and g77
failures unchanged, and now gives:
=== g++ Summary ===
# of expected passes 5974
# of unexpected failures 12
# of expected failures 96
# of untested testcases 8
(down from 805 unexpected failures). I'm not sure if this is simply due to the
fact that libio and libstc++ now test our correctly (they had 100% failure rates
before).
I was using BOOT_CFLAGS=-O3 -march=i686 -funroll-loops -fomit-frame-pointer
-fforce-mem -fforce-addr -ffast-math -malign-double -malign-jumps=2
-malign-functions=2 -malign-loops=2, and I'll start adding back these flags
(starting with the -malign ones) in order to see which one starts the problems.
David