From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 22904 invoked by alias); 22 Oct 2007 04:38:40 -0000 Received: (qmail 22896 invoked by uid 22791); 22 Oct 2007 04:38:39 -0000 X-Spam-Check-By: sourceware.org Received: from rock.gnat.com (HELO rock.gnat.com) (205.232.38.15) by sourceware.org (qpsmtpd/0.31) with ESMTP; Mon, 22 Oct 2007 04:38:36 +0000 Received: from localhost (localhost.localdomain [127.0.0.1]) by filtered-rock.gnat.com (Postfix) with ESMTP id CA8732AA603 for ; Mon, 22 Oct 2007 00:38:34 -0400 (EDT) Received: from rock.gnat.com ([127.0.0.1]) by localhost (rock.gnat.com [127.0.0.1]) (amavisd-new, port 10024) with LMTP id BTTWPMaFwCa0 for ; Mon, 22 Oct 2007 00:38:34 -0400 (EDT) Received: from joel.gnat.com (localhost.localdomain [127.0.0.1]) by rock.gnat.com (Postfix) with ESMTP id 8C6072AA5E6 for ; Mon, 22 Oct 2007 00:38:34 -0400 (EDT) Received: by joel.gnat.com (Postfix, from userid 1000) id CE93AE7AC8; Sun, 21 Oct 2007 21:38:31 -0700 (PDT) Date: Mon, 22 Oct 2007 06:01:00 -0000 From: Joel Brobecker To: gdb-patches@sourceware.org Subject: Re: [RFC] Support exec tracing on GNU/Linux and HP-UX Message-ID: <20071022043831.GD764@adacore.com> References: <20071019175920.GA548@caradoc.them.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20071019175920.GA548@caradoc.them.org> User-Agent: Mutt/1.4.2.2i 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: 2007-10/txt/msg00487.txt.bz2 > The user interface is the same as it was when this was originally > added on HP-UX long ago. When GDB sees an exec event, it > automatically loads the new executable as the main symbol file. > This is a little disconcerting if you try to use "run" again > later, but otherwise seems pretty convenient. > > Any comments? Mark (or someone else), could you offer to test this > on HP-UX? I suspect the tests will fail for cosmetic reasons due to > the changed messages but I hope exec tracing will generally work. I have tested this patch, and it unfortunately creates some issues: (gdb) run Starting program: /[...]/p Executing new program: /usr/local/bin/zsh (no debugging symbols found) Unable to find __dld_flags symbol in object file. What seems to happen is that we're enabling exec-following during the startup sequence, and that causes us to do the follow-exec routine on it. Eventually, this fails because the shell binary that I use apparently does not have any symbols, thus causing the solib-som hook to error out. Do we do the same on Linux? (exec-follow during the startup) -- Joel