This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch] Strength reduction for autoincrement generation
- To: Bernd Schmidt <bernds at redhat dot com>
- Subject: Re: [patch] Strength reduction for autoincrement generation
- From: Michael Hayes <m dot hayes at elec dot canterbury dot ac dot nz>
- Date: Sun, 11 Feb 2001 12:48:30 +1300 (NZDT)
- Cc: Michael Hayes <m dot hayes at elec dot canterbury dot ac dot nz>, gcc-patches at gcc dot gnu dot org
- References: <14956.44398.355351.244881@ongaonga.elec.canterbury.ac.nz> <Pine.LNX.4.30.0101231323350.1128-100000@host140.cambridge.redhat.com>
I sent the following a while ago:
> > Without this patch autoincrement addressing are only generated with
> > extremely good luck. A discussion (well a monologue) of the problem
> > can be found on gcc.gnu.org/ml/gcc/2001-01/msg00548.html.
> >
> > OK to commit?
>
> > 2001-01-23 Michael Hayes <m.hayes@elec.canterbury.ac.nz>
> >
> > * loop.c (loop_giv_reduce_benefit): Tweak autoincrement generation
> > heuristic.
and after a short discussion the thread died. This patch provides
a large improvement for autoincrement generation with small loops:
to be blunt, without it the generated code for the c4x is appalling.
Can I commit it?
Michael.