From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 15127 invoked by alias); 24 Oct 2002 05:36:01 -0000 Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-prs-owner@gcc.gnu.org Received: (qmail 15112 invoked by uid 71); 24 Oct 2002 05:36:01 -0000 Date: Wed, 23 Oct 2002 22:36:00 -0000 Message-ID: <20021024053601.15111.qmail@sources.redhat.com> To: nobody@gcc.gnu.org Cc: gcc-prs@gcc.gnu.org, From: Alan Modra Subject: Re: optimization/8328: powerpc64 ICE in reload Reply-To: Alan Modra X-SW-Source: 2002-10/txt/msg00925.txt.bz2 List-Id: The following reply was made to PR optimization/8328; it has been noted by GNATS. From: Alan Modra To: gcc-gnats@gcc.gnu.org Cc: Subject: Re: optimization/8328: powerpc64 ICE in reload Date: Thu, 24 Oct 2002 15:05:03 +0930 On Thu, Oct 24, 2002 at 11:18:38AM +0930, Alan Modra wrote: > holding a common subexpression, "(void *) &loc". The first one being > that gcse incorrectly inserts code around a label. We replace No, I was wrong about this. The inserted code is actually correct, if not ideal. > Note how insn 383 has moved before "exit". Also, I believe that gcse > should be using register DI 131 here as that is the one used in a > previous replacement of "(void *) &loc". And this one is related to this FIXME in gcse.c, I think: /* FIXME: If it turns out that the insn isn't deletable, then we may have unnecessarily extended register lifetimes and made things worse. */ So we're down to just the reload problem. -- Alan Modra IBM OzLabs - Linux Technology Centre