From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 5699 invoked by alias); 23 Jan 2006 17:37:00 -0000 Received: (qmail 5690 invoked by uid 22791); 23 Jan 2006 17:37:00 -0000 X-Spam-Check-By: sourceware.org Received: from nevyn.them.org (HELO nevyn.them.org) (66.93.172.17) by sourceware.org (qpsmtpd/0.31.1) with ESMTP; Mon, 23 Jan 2006 17:36:59 +0000 Received: from drow by nevyn.them.org with local (Exim 4.54) id 1F15cf-0005Ox-5z; Mon, 23 Jan 2006 12:36:57 -0500 Date: Mon, 23 Jan 2006 17:47:00 -0000 From: Daniel Jacobowitz To: David Lamy-Charrier Cc: GDB List Subject: Re: breakpoint commands no more working due to MI front-end Message-ID: <20060123173657.GA20650@nevyn.them.org> Mail-Followup-To: David Lamy-Charrier , GDB List References: <6541ed4c0601230241p2646be6bm87251b26a935bb7e@mail.gmail.com> <20060123134952.GA13642@nevyn.them.org> <6541ed4c0601230745o6e063cb6he3ec808e544cd887@mail.gmail.com> <20060123160220.GB16524@nevyn.them.org> <6541ed4c0601230905q50e4a6b1rc8e4659b63589633@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <6541ed4c0601230905q50e4a6b1rc8e4659b63589633@mail.gmail.com> User-Agent: Mutt/1.5.8i X-IsSubscribed: yes Mailing-List: contact gdb-help@sourceware.org; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-owner@sourceware.org X-SW-Source: 2006-01/txt/msg00221.txt.bz2 On Mon, Jan 23, 2006 at 06:05:45PM +0100, David Lamy-Charrier wrote: > Thanks Daniel for the idea, I am going to try it. > > But if I am right, mi_execute_command is called to handle MI commands > from the front-end, so it is already too late to execute the commands > associated with the breakpoint. > The commands should have been executed before and the front-end should > even not be notified that GDB stopped and continued, no? That would require a much bigger change to the way GDB works. We've been talking about some related changes; the general question is whether the MI front end should know that something other than itself has caused the inferior to change state. This is vital to have multiple interpreters active at the same time. Given how MI works and is defined today, I don't think breakpoint commands make a whole lot of sense: your front end should supply the commands when it sees the breakpoint. Is there a reason you can't make CDT do that? -- Daniel Jacobowitz CodeSourcery