From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-94985-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 9691 invoked by alias); 27 Apr 2004 17:55:27 -0000
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
Received: (qmail 9672 invoked from network); 27 Apr 2004 17:55:27 -0000
Received: from unknown (HELO NUTMEG.CAM.ARTIMI.COM) (217.40.111.177)
  by sources.redhat.com with SMTP; 27 Apr 2004 17:55:27 -0000
Received: from mace ([192.168.1.25]) by NUTMEG.CAM.ARTIMI.COM with Microsoft SMTPSVC(6.0.3790.0);
	 Tue, 27 Apr 2004 18:53:58 +0100
From: "Dave Korn" <dk@artimi.com>
To: <gcc@gcc.gnu.org>
Subject: RE: Error while using LOG_LINKS information.
Date: Tue, 27 Apr 2004 18:48:00 -0000
MIME-Version: 1.0
Content-Type: text/plain;
	charset="US-ASCII"
Content-Transfer-Encoding: 7bit
In-Reply-To: <20040427163515.40040.qmail@web40307.mail.yahoo.com>
Message-ID: <NUTMEGUzk0GyUJPEUeC00000054@NUTMEG.CAM.ARTIMI.COM>
X-OriginalArrivalTime: 27 Apr 2004 17:53:58.0781 (UTC) FILETIME=[9DC86AD0:01C42C80]
X-SW-Source: 2004-04/txt/msg01275.txt.bz2

> -----Original Message-----
> From: gcc-owner On Behalf Of kernel_learner
> Sent: 27 April 2004 17:35

> Hi all,
> 
>  IN a machine-dependent-reorg pass that I am writing I
> wrote a simple piece of code which traverses the
> LOG_LINKS:
> 
> for(;insns;insns=NEXT_INSN(insns))
> {
>    for(link = LOG_LINKS (insns); link; link =
> XEXP(link, 1));
> }
> This gives me a problem while building gcc:
> 
> make[2]: *** [libgcc/./_muldi3.o] Error 1
> make[2]: Leaving directory
> `/home/SOK_STUFF/or1k/b-gcc/gcc'
> make[1]: *** [libgcc.a] Error 2
> make[1]: Leaving directory
> `/home/SOK_STUFF/or1k/b-gcc/gcc'
> make: *** [all-gcc] Error 2
> 
> I am sure that the problem is cause by this code since
> if I take it away it builds PERFECT!
> 
> Is the LOG_LINKS information corrupted or invalid at
> the machine dependent reorganization part. Or am I
> traversing it wrong?

  From my reading of the docs, LOG_LINKS (insn) is only valid if the insn is
of type insn, call_insn or jump_insn.  You should make sure in your outer
loop that insn is of one of those types before the inner loop is allowed to
run, otherwise you might be applying LOG_LINKS to a code_label, barrier or
note.

  Or it could be some kind of corruption, if that doesn't fix it.


    cheers, 
      DaveK
-- 
Can't think of a witty .sigline today....