From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 3998 invoked by alias); 4 Feb 2008 21:45:31 -0000 Received: (qmail 3989 invoked by uid 22791); 4 Feb 2008 21:45:31 -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; Mon, 04 Feb 2008 21:45:06 +0000 Received: from nan.false.org (localhost [127.0.0.1]) by nan.false.org (Postfix) with ESMTP id 56E769829D; Mon, 4 Feb 2008 21:45:05 +0000 (GMT) Received: from caradoc.them.org (22.svnf5.xdsl.nauticom.net [209.195.183.55]) by nan.false.org (Postfix) with ESMTP id 3547998216; Mon, 4 Feb 2008 21:45:05 +0000 (GMT) Received: from drow by caradoc.them.org with local (Exim 4.68) (envelope-from ) id 1JM97g-0006gS-1C; Mon, 04 Feb 2008 16:45:04 -0500 Date: Mon, 04 Feb 2008 21:45:00 -0000 From: Daniel Jacobowitz To: Greg Law Cc: gdb-patches@sourceware.org Subject: Re: SIGSEGV on gdb 6.7* Message-ID: <20080204214504.GA25564@caradoc.them.org> Mail-Followup-To: Greg Law , gdb-patches@sourceware.org References: <47A77A6C.8050007@undo-software.com> <20080204210333.GA23250@caradoc.them.org> <47A7850B.10202@undo-software.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <47A7850B.10202@undo-software.com> User-Agent: Mutt/1.5.17 (2007-12-11) 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: 2008-02/txt/msg00090.txt.bz2 [Please reply to the list, thanks!] On Mon, Feb 04, 2008 at 09:35:07PM +0000, Greg Law wrote: > When a register is examined, we (eventually) get to > frame_register_unwind. This does: > > frame->unwind->prev_register (frame->next, &frame->prologue_cache, regnum, > optimizedp, lvalp, addrp, realnump, bufferp); > > which is actually a function pointer to (on plain old x86 Linux) > sentinel_frame_prev_register, which goes: Ah yes. That's the only prologue cache which has any business accessing a regcache directly, none of the others do. flushregs should invalidate the frame cache and current/selected frame. I have been meaning to fix that for, roughly, ever. Does it work better if you do that? > Yeah, I guess this more important on remote targets over slow > links, right? In that case, do we not need to take care from > registers_changed() to go and invalidate all the pointers to cached > registers in the frame structures? Hmm, although thinking about it now, I > guess the frame structures all need > throwing away/recalculating if the registers have changed anyway? I can't > find what causes that to happen though. That's the key point :-) -- Daniel Jacobowitz CodeSourcery