optimization/7130: miscompiled code for gcc-3.1 in powerpc linux with -funroll-all-loops

[Alan Modra]

The following reply was made to PR optimization/7130; it has been noted by GNATS.

From: Alan Modra <amodra@bigpond.net.au>
To: Andreas Jaeger <aj@suse.de>
Cc: yozo@cs.berkeley.edu, gcc-gnats@gcc.gnu.org, gcc-patches@gcc.gnu.org
Subject: Re: optimization/7130: miscompiled code for gcc-3.1 in powerpc linux with -funroll-all-loops
Date: Fri, 12 Jul 2002 09:12:20 +0930

 On Thu, Jul 11, 2002 at 02:06:55PM +0200, Andreas Jaeger wrote:
 > Alan Modra <amodra@bigpond.net.au> writes:
 > 
 > > (And this time, I waited for bootstrap to get through building the stage2
 > >  compiler before firing off emails.)
 > 
 > 
 > What about the testsuite?  Any new regressions or passes?
 
 New regressions unfortunately.  Patch withdrawn until I can figure out
 what is going wrong.
 
 > I guess this is on Powerpc-linux?
 
 Yes.
 
 -- 
 Alan Modra
 IBM OzLabs - Linux Technology Centre