This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: fix execute/20011008-3.c (fwd)
- From: Joern Rennecke <joern dot rennecke at st dot com>
- To: rth at redhat dot com
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Wed, 17 Apr 2002 16:27:31 +0100
- Subject: Re: fix execute/20011008-3.c (fwd)
- Organization: SuperH UK Ltd.
- References: <200203170011.AAA04829@meolyon.local>
- Reply-to: joern dot rennecke at st dot com
amylaar@onetel.net.uk wrote:
> > On Sat, Feb 09, 2002 at 11:58:22AM +0000, Joern Rennecke wrote:
> > However, if the loop is not completely unrolled, the biv will be computed
> > both inside the loop and afterwards, yet the set inside the loop can
> > no eliminated since it is not dead on the looping path.
Richard Henderson wrote:
> Any suggetions on dealing with this then?
How about setting some flag in the loop struct to indicate that a final value
is known, but no set emitted; the loop unroller can then test that flag and
emit the set when needed.
--
--------------------------
SuperH
2430 Aztec West / Almondsbury / BRISTOL / BS32 4AQ
T:+44 1454 462330