From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 843 invoked by alias); 2 Oct 2011 02:04:48 -0000 Received: (qmail 798 invoked by uid 22791); 2 Oct 2011 02:04:47 -0000 X-SWARE-Spam-Status: No, hits=-2.5 required=5.0 tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,FREEMAIL_FROM,RCVD_IN_DNSWL_LOW X-Spam-Check-By: sourceware.org Received: from mail-ey0-f169.google.com (HELO mail-ey0-f169.google.com) (209.85.215.169) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Sun, 02 Oct 2011 02:04:33 +0000 Received: by eye13 with SMTP id 13so2268124eye.0 for ; Sat, 01 Oct 2011 19:04:31 -0700 (PDT) MIME-Version: 1.0 Received: by 10.14.10.208 with SMTP id 56mr4656784eev.125.1317521071694; Sat, 01 Oct 2011 19:04:31 -0700 (PDT) Received: by 10.14.127.137 with HTTP; Sat, 1 Oct 2011 19:04:31 -0700 (PDT) In-Reply-To: <4E85D3DF.7070307@twiddle.net> References: <1317339968-16212-1-git-send-email-rth@twiddle.net> <4E856BDE.80904@web.de> <4E85D3DF.7070307@twiddle.net> Date: Sun, 02 Oct 2011 02:04:00 -0000 Message-ID: Subject: Re: [PATCH] tcg: Use the GDB JIT debugging interface. From: Daniel Jacobowitz To: Richard Henderson Cc: Jan Kiszka , qemu-devel@nongnu.org, gdb@sourceware.org Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable 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: 2011-10/txt/msg00001.txt.bz2 On Fri, Sep 30, 2011 at 10:36 AM, Richard Henderson wrote: > On 09/30/2011 12:12 AM, Jan Kiszka wrote: >>> Breakpoint 1, __ldb_mmu (addr=3D1001716, mmu_idx=3D0) >>> =A0 =A0 at /home/rth/work/qemu/qemu/softmmu_template.h:86 >>> 86 =A0 { >>> (gdb) where >>> #0 =A0__ldb_mmu (addr=3D1001716, mmu_idx=3D0) >>> =A0 =A0 at /home/rth/work/qemu/qemu/softmmu_template.h:86 >>> #1 =A00x0000000040000afc in ?? () > ... >> Would it be possible to tag translated code as well? At TB or maybe even >> input instruction level? Only in debugging mode of course. > > I don't know. =A0A couple of points on that: > > (1) I was rather expecting frame 1 above to be marked "code_gen_buffer" > =A0 =A0rather than "??". =A0I *did* install that symbol in the generated > =A0 =A0ELF image as a function covering the entire buffer. =A0So I'm not > =A0 =A0quite sure why GDB didn't use it. Totally wild guess: associated with the wrong section? I don't remember what we do w.r.t. ABS in shared libraries, but for non-ABS we do validate what section we think things are in. And I'm not sure if we load section info in the JIT loader, either. --=20 Thanks, Daniel