From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gdb-return-13360-listarch-gdb=sources.redhat.com@sources.redhat.com>
Received: (qmail 12692 invoked by alias); 10 Apr 2003 06:02:55 -0000
Mailing-List: contact gdb-help@sources.redhat.com; run by ezmlm
Precedence: bulk
List-Subscribe: <mailto:gdb-subscribe@sources.redhat.com>
List-Archive: <http://sources.redhat.com/ml/gdb/>
List-Post: <mailto:gdb@sources.redhat.com>
List-Help: <mailto:gdb-help@sources.redhat.com>, <http://sources.redhat.com/ml/#faqs>
Sender: gdb-owner@sources.redhat.com
Received: (qmail 12684 invoked from network); 10 Apr 2003 06:02:54 -0000
Received: from unknown (HELO hrtades9.atea.be) (194.78.143.106)
  by sources.redhat.com with SMTP; 10 Apr 2003 06:02:54 -0000
Received: from hrtades10.atea.be (siemens.atea.be [139.10.143.141]) by hrtades9.atea.be with SMTP (Microsoft Exchange Internet Mail Service Version 5.5.2653.13)
	id 1RFVN75Z; Thu, 10 Apr 2003 08:02:53 +0200
Received: by siemens.atea.be with Internet Mail Service (5.5.2653.19)
	id <H6JKS72G>; Thu, 10 Apr 2003 08:02:53 +0200
Message-ID: <57FD2C3A246F76438CA6FDAD8FE9F19597E9D6@hrtades7.atea.be>
From: Vermeulen Jan <Jan.Vermeulen@siemens.com>
To: 'Peter Barada' <pbarada@mail.wm.sps.mot.com>
Cc: peter@baradas.org, gdb@sources.redhat.com, Peter.Barada@motorola.com
Subject: RE: How to build application to run under the ppc simulator?
Date: Thu, 10 Apr 2003 06:02:00 -0000
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
X-SW-Source: 2003-04/txt/msg00092.txt.bz2

Peter,

> >> and it bombs with:
> >> 
> >> >Program terminated with signal SIGSEGV, Segmentation fault.
> >> >The program no longer exists.
> >> >[Switching to process 0]
> >> >(gdb) 
> >
> >This is quite normal, because your program is executed correctly and
> >executes an 'blr' instruction at the end of the main 
> routine. This makes the
> >simulated CPU jump to what's currently in the link register, 
> which can be
> >anything. There it will do crazy things untill it finally 
> segfaults :)
> >
> >If in doubt, just 'stepi' the program, watching the assembly 
> be executed.
> 
> I can't stepi from the start, I have to run it.  How(actually,
> *where*) can I set a breakpoint at the *first* instruction so I can
> *then* stepi? 

If your only function is 'foo', why not try "break foo" at the gdb command?

I just tried it out (same example as you) and it works. I can stepi in the
program and see that it jumps to 0x0 after the 'blr' instruction at the end
of 'foo' where it segfaults.

HTH,
Jan