From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 22889 invoked by alias); 14 Sep 2011 19:36:01 -0000 Received: (qmail 22839 invoked by uid 22791); 14 Sep 2011 19:36:00 -0000 X-SWARE-Spam-Status: No, hits=-2.2 required=5.0 tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,RCVD_IN_DNSWL_NONE,RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from elasmtp-galgo.atl.sa.earthlink.net (HELO elasmtp-galgo.atl.sa.earthlink.net) (209.86.89.61) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Wed, 14 Sep 2011 19:35:44 +0000 Received: from [70.170.59.51] (helo=macbook2.local) by elasmtp-galgo.atl.sa.earthlink.net with esmtpa (Exim 4.67) (envelope-from ) id 1R3vFI-0006WB-5I for gdb@sourceware.org; Wed, 14 Sep 2011 15:35:44 -0400 Message-ID: <4E710209.4030600@earthlink.net> Date: Wed, 14 Sep 2011 19:36:00 -0000 From: Stan Shebs User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.7; rv:6.0.2) Gecko/20110902 Thunderbird/6.0.2 MIME-Version: 1.0 To: gdb@sourceware.org Subject: Re: How to use gdbserver for X86 host with arm-linux target ? References: <4E6DEBD1.4080007@parrot.com> <201109121448.18875.pedro@codesourcery.com> In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-ELNK-Trace: ae6f8838ff913eba0cc1426638a40ef67e972de0d01da94048a426f539f4db8f59560b8633aa3cdf350badd9bab72f9c350badd9bab72f9c350badd9bab72f9c X-IsSubscribed: yes Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org X-SW-Source: 2011-09/txt/msg00049.txt.bz2 On 9/13/11 10:28 AM, Doug Evans wrote: > [...] > Seems like this another example where gdb suffers from having two code > bases (gdb, gdbserver) to implement the same functionality. > Ironically, any discussion of sharing vs duplication would have been happening a few dozen steps away from where you and I were sitting at Cygnus back in 1993. :-) I don't remember if there was much talk about it back then, but if there was, it probably seemed like an impossibility, especially considering the shoestring that Cygnus operated on at the time. The whole idea of a target vector interface was still pretty new, and control flow was in reality much more intertwined and config-macro-dependent than the target vector pretended. I think all new debugger designs are separating the low-level inferior control sufficiently to build debugger and server using the same code if desired. Stan stan@codesourcery.com