From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 13788 invoked by alias); 20 Aug 2007 14:21:57 -0000 Received: (qmail 13670 invoked by uid 22791); 20 Aug 2007 14:21:55 -0000 X-Spam-Check-By: sourceware.org Received: from NaN.false.org (HELO nan.false.org) (208.75.86.248) by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 20 Aug 2007 14:21:45 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id 48ABF980C0; Mon, 20 Aug 2007 14:21:44 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id 09366980BF; Mon, 20 Aug 2007 14:21:44 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.67) (envelope-from ) id 1IN88T-0007TW-Ff; Mon, 20 Aug 2007 10:21:41 -0400 Date: Mon, 20 Aug 2007 14:21:00 -0000 From: Daniel Jacobowitz To: Luis Machado Cc: Eli Zaretskii , gdb-patches@sourceware.org Subject: Re: [patch] Watchpoints: support for thread parameters Message-ID: <20070820142141.GA18034@caradoc.them.org> Mail-Followup-To: Luis Machado , Eli Zaretskii , gdb-patches@sourceware.org References: <1187298178.5853.11.camel@localhost> <1187365616.4520.14.camel@localhost> <20070817184953.GA1747@caradoc.them.org> <1187618630.4974.5.camel@localhost> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1187618630.4974.5.camel@localhost> User-Agent: Mutt/1.5.15 (2007-04-09) X-IsSubscribed: yes Mailing-List: contact gdb-patches-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sourceware.org X-SW-Source: 2007-08/txt/msg00392.txt.bz2 On Mon, Aug 20, 2007 at 11:03:49AM -0300, Luis Machado wrote: > On Fri, 2007-08-17 at 14:49 -0400, Daniel Jacobowitz wrote: > > On Fri, Aug 17, 2007 at 09:41:38PM +0300, Eli Zaretskii wrote: > > > Is this all that needs to be done to have thread-specific watchpoints? > > > Do they work after this patch as you want them to? > > > > In my opinion, not really. On some systems, you can tell the target > > layer which threads the watchpoint should be inserted for and spare > > yourself extra stops in "wrong" threads. > > What do you have in mind regarding the whole feature? Will we head > towards a different solution? The work you've done will still be necessary. I have no plans to work on the bigger picture, myself, but thread specificity needs to be communicated to lower layers somehow. Consider: in the current status quo the watchpoint would be inserted in every thread, using up precious hardware debug resources and triggering unnecessarily. We'll just resume in threads other than the indicated one, so we shouldn't set up the watchpoint anywhere else. -- Daniel Jacobowitz CodeSourcery