From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 1446 invoked by alias); 3 Oct 2008 17:52:29 -0000 Received: (qmail 1398 invoked by uid 22791); 3 Oct 2008 17:52:29 -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; Fri, 03 Oct 2008 17:51:49 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id E51A010D3B; Fri, 3 Oct 2008 17:51:47 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id C3DCD10D36; Fri, 3 Oct 2008 17:51:45 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.69) (envelope-from ) id 1Kloob-00041r-1W; Fri, 03 Oct 2008 13:51:45 -0400 Date: Fri, 03 Oct 2008 17:52:00 -0000 From: Daniel Jacobowitz To: Joel Brobecker Cc: =?iso-8859-1?Q?S=E9rgio_Durigan_J=FAnior?= , gdb-patches@sourceware.org Subject: Re: [PATCH 1/4] 'catch syscall' feature -- Architecture-independent part Message-ID: <20081003175145.GA15469@caradoc.them.org> Mail-Followup-To: Joel Brobecker , =?iso-8859-1?Q?S=E9rgio_Durigan_J=FAnior?= , gdb-patches@sourceware.org References: <1222798409.30389.23.camel@miki> <20081002211256.GO3665@adacore.com> <1223001252.9858.11.camel@miki> <20081003060629.GQ3665@adacore.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20081003060629.GQ3665@adacore.com> User-Agent: Mutt/1.5.17 (2008-05-11) 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: 2008-10/txt/msg00090.txt.bz2 On Thu, Oct 02, 2008 at 11:06:29PM -0700, Joel Brobecker wrote: > That's correct. The short answer is that, if we make your catchpoint > use a more generic type and base the actual implementation of > breakpoint_ops methods, then, later on, when someone decides to > implement a new kind of catchpoint with similar functionality, > then all he should have to do is create a new breakpoint_ops vector > with appropriate methods, and voila. Right. This is what's supposed to be possible, at least. The way I would suggest to do this is to create a new bp_catchpoint. We can either leave existing OS catchpoints alone, or else convert some of them (fork/vfork for example) to the new mechanism as proof that it works. -- Daniel Jacobowitz CodeSourcery