From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 7777 invoked by alias); 11 Dec 2007 18:44:14 -0000 Received: (qmail 7769 invoked by uid 22791); 11 Dec 2007 18:44:14 -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; Tue, 11 Dec 2007 18:44:08 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id B631298020 for ; Tue, 11 Dec 2007 18:44:06 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id A13489801F for ; Tue, 11 Dec 2007 18:44:06 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.68) (envelope-from ) id 1J2A5N-0002vL-Mf for gdb@sourceware.org; Tue, 11 Dec 2007 13:44:05 -0500 Date: Tue, 11 Dec 2007 18:44:00 -0000 From: Daniel Jacobowitz To: gdb@sourceware.org Subject: Re: Stepping off breakpoints in non-stop debugging mode Message-ID: <20071211184405.GA10876@caradoc.them.org> Mail-Followup-To: gdb@sourceware.org References: <20071208170359.GA8777@caradoc.them.org> <20071210231711.GA15675@caradoc.them.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.15 (2007-04-09) 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: 2007-12/txt/msg00077.txt.bz2 On Tue, Dec 11, 2007 at 10:24:44AM -0800, Jim Blandy wrote: > As far as PTRACE_SINGLESTEP is concerned, isn't pushfl the only > user instruction that can observe the state of TF anyway? I can't answer that. Maybe Paul could. The focus of the kernel code I was referring to is instructions which modify the state of TF; you can observe changes by telling whether you get a trap, if your debugger's signal handling is good enough. The debugger can do a software single step over the pushfl instruction if it cares about accuracy enough to bother, and/or fix up the resulting stack value. -- Daniel Jacobowitz CodeSourcery