This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: decrement and branch optimization broken?
- To: Jan Hubicka <jh at suse dot cz>
- Subject: Re: decrement and branch optimization broken?
- From: Roman Zippel <zippel at linux-m68k dot org>
- Date: Fri, 27 Jul 2001 19:41:50 +0200
- CC: gcc at gcc dot gnu dot org, john at feith dot com
- References: <20010727170559.F30003@atrey.karlin.mff.cuni.cz>
Hi,
Jan Hubicka wrote:
> I still think combine should be able to produce the instruction in the
> sequence you point out.
I've found the problem, recog_for_combine removes the old notes prior to
calling recog, but this pattern looks for REG_NONNEG.
This change was introduced with this patch:
Fri Aug 4 11:43:49 2000 John Wehle (john@feith.com)
* combine.c (recog_for_combine): Remove the old notes
prior to attempting to recognize the new pattern.
(distribute_notes): Ignore REG_NONNEG notes.
http://gcc.gnu.org/ml/gcc-patches/2000-06/msg00564.html
bye, Roman