From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 21095 invoked by alias); 26 Jan 2003 16:21:25 -0000 Mailing-List: contact gdb-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sources.redhat.com Received: (qmail 21080 invoked from network); 26 Jan 2003 16:21:25 -0000 Received: from unknown (HELO crack.them.org) (65.125.64.184) by 172.16.49.205 with SMTP; 26 Jan 2003 16:21:25 -0000 Received: from nevyn.them.org ([66.93.61.169] ident=mail) by crack.them.org with asmtp (Exim 3.12 #1 (Debian)) id 18crQ5-0001PP-00; Sun, 26 Jan 2003 12:22:13 -0600 Received: from drow by nevyn.them.org with local (Exim 3.36 #1 (Debian)) id 18cpXY-0000to-00; Sun, 26 Jan 2003 11:21:48 -0500 Date: Sun, 26 Jan 2003 16:21:00 -0000 From: Daniel Jacobowitz To: Jan Hoogerbrugge Cc: gdb@sources.redhat.com Subject: Re: two gdb porting questions Message-ID: <20030126162147.GA3439@nevyn.them.org> Mail-Followup-To: Jan Hoogerbrugge , gdb@sources.redhat.com References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.1i X-SW-Source: 2003-01/txt/msg00419.txt.bz2 On Sun, Jan 26, 2003 at 11:50:55AM +0100, Jan Hoogerbrugge wrote: > >From: Daniel Jacobowitz > > >> * I want to use stabs to pass frame information from my compiler to > >> gdb. Which stab number(s) should I use for this? How do I get this > >> information in my xxx-tdep.c given the start address of the > >> function? (I use ELF in the case that matters) > > > >Is it absolutely necessary that you use stabs? There are clean ways to > >express this information in DWARF-2, and if you're using ELF, there's > >really no excuse. > > My compiler (which is not gcc) happens to generate stabs. So I have to do > it with stabs. Currently the compiler misuses a N_ROSYM to pass frame info > to its debugger. However, gdb does not like this. > > Are there other targets where a stab is used for frame information? > > The instruction format of the debug target is so complex that decoding the > prologue is not a real option. We did something similar for an unpublished port; as a local hack it wasn't too bad and we did it just about the same way you're describing. You just stuff the value in a new field in the symbol struct. Eventually we arranged to generate DWARF-2 instead, though, and it went away. -- Daniel Jacobowitz MontaVista Software Debian GNU/Linux Developer