From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 31630 invoked by alias); 5 Nov 2008 19:34:06 -0000 Received: (qmail 31612 invoked by uid 22791); 5 Nov 2008 19:34:06 -0000 X-Spam-Check-By: sourceware.org Received: from mail.codesourcery.com (HELO mail.codesourcery.com) (65.74.133.4) by sourceware.org (qpsmtpd/0.31) with ESMTP; Wed, 05 Nov 2008 19:33:32 +0000 Received: (qmail 948 invoked from network); 5 Nov 2008 19:33:30 -0000 Received: from unknown (HELO orlando.local) (pedro@127.0.0.2) by mail.codesourcery.com with ESMTPA; 5 Nov 2008 19:33:30 -0000 From: Pedro Alves To: gdb-patches@sourceware.org, Eli Zaretskii Subject: Re: [PATCH 1/4] 'catch syscall' feature -- =?utf-8?q?=09Architecture-independent=09part?= Date: Wed, 05 Nov 2008 19:34:00 -0000 User-Agent: KMail/1.9.10 Cc: bauerman@br.ibm.com, sergiodj@linux.vnet.ibm.com References: <1225773079.24532.52.camel@miki> <200811051856.27087.pedro@codesourcery.com> In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200811051933.28792.pedro@codesourcery.com> 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: 2008-11/txt/msg00090.txt.bz2 On Wednesday 05 November 2008 19:08:49, Eli Zaretskii wrote: > > From: Pedro Alves > > Date: Wed, 5 Nov 2008 18:56:26 +0000 > > Cc: bauerman@br.ibm.com, > > sergiodj@linux.vnet.ibm.com > > > > What's wrong with "break ReadFile" ? > > Does it work for you? ... it does. It's a function exported from a regular dll, just like all other functions exported from other dlls. I can only infer that you haven't gone through the trouble of trying it yourself. > And if the breakpoint breaks, can you easily > see the arguments of the call? > No, but that's not different from breaking on any other function in any other system when you don't have debug info for it. To solve this, the real solution is to make GDB make use of MSFT's PDB debug info, not to somehow make up a new method to extract arguments from plain regular code. This is *no different* from breaking on the "read" function in libc.so on linux, without debug info for libc.so. Eventually the libc 'read' function calls the real 'read' syscall, just like on Windows. -- Pedro Alves