This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
Re: 60 GCC regressions, 24 new, with your patch on 2002-07-30T23:53:00Z.
- From: David Edelsohn <dje at watson dot ibm dot com>
- To: gcc-regression at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Cc: Franz dot Sirl-kernel at lauterbach dot com, amodra at bigpond dot net dot au, bernds at redhat dot com, rth at redhat dot com, toon at moene dot indiv dot nluug dot nl, Geoff Keating <geoffk at redhat dot com>, zack at codesourcery dot com
- Date: Wed, 31 Jul 2002 01:24:00 -0400
- Subject: Re: 60 GCC regressions, 24 new, with your patch on 2002-07-30T23:53:00Z.
The new g77 failures are ICEs in validate_condition_mode. I am
guessing that the g77 failures are due to Toon's patch and the following
fragment from validate_condition_mode:
/* These should never be generated except for
flag_unsafe_math_optimizations. */
if (mode == CCFPmode
&& ! flag_unsafe_math_optimizations
&& (code == LE || code == GE
|| code == UNEQ || code == LTGT
|| code == UNGT || code == UNLT))
abort ();
I suspect that the test needs to add flag_finite_math_only, but I won't be
investigate this further until tomorrow. If someone wants to analyze this
and verify the proposed solution, be my guest.
Thanks, David