This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
Re: 20 GCC HEAD regressions, 4 new, with your patch on 2004-06-16T20:00:03Z.
- From: Roger Sayle <roger at eyesopen dot com>
- To: David Edelsohn <dje at watson dot ibm dot com>
- Cc: Andrew Pinski <pinskia at physics dot uc dot edu>, <gcc-regression at gcc dot gnu dot org>
- Date: Thu, 17 Jun 2004 11:12:20 -0600 (MDT)
- Subject: Re: 20 GCC HEAD regressions, 4 new, with your patch on 2004-06-16T20:00:03Z.
On Thu, 17 Jun 2004, David Edelsohn wrote:
> > native gcc.sum gcc.c-torture/execute/divconst-2.c
>
> This failure appeared yesterday on the powerpc-ibm-aix testresults.
> The log output is:
>
> /gcc/dje/src/gcc/testsuite/gcc.c-torture/execute/divconst-2.c: In function `r':
> /gcc/dje/src/gcc/testsuite/gcc.c-torture/execute/divconst-2.c:10: internal compiler error: in expand_mult_const, at expmed.c:2614
>
> I think this may be related to Roger's recent expmed.c changes.
Yep, this looks like mine. The assertion checking on line 2613
needs to mask both val and val_so_far by the current machine mode.
I'm preparing a patch that I'll test on powerpc-ibm-aix5.2.0.0.
Sorry for the inconvenience,
Roger
--