From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 24238 invoked by alias); 15 Jul 2003 17:41:29 -0000 Mailing-List: contact gdb-patches-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: gdb-patches-owner@sources.redhat.com Received: (qmail 24231 invoked from network); 15 Jul 2003 17:41:29 -0000 Received: from unknown (HELO localhost.redhat.com) (66.30.197.194) by sources.redhat.com with SMTP; 15 Jul 2003 17:41:29 -0000 Received: from redhat.com (localhost [127.0.0.1]) by localhost.redhat.com (Postfix) with ESMTP id C77E22B7F; Tue, 15 Jul 2003 13:41:26 -0400 (EDT) Message-ID: <3F143CC6.5060304@redhat.com> Date: Tue, 15 Jul 2003 17:41:00 -0000 From: Andrew Cagney User-Agent: Mozilla/5.0 (X11; U; NetBSD macppc; en-US; rv:1.0.2) Gecko/20030223 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Andrew Cagney , gdb-patches@sources.redhat.com Subject: Re: [patch, rfc, 6.0] Change frame predicates to sniffers References: <3F0F32C4.9060700@redhat.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-SW-Source: 2003-07/txt/msg00298.txt.bz2 > Hello, > > This patch adds frame sniffers (name stolen from other sniffer code) to the frame code. Unlike the existing frame predicate methods that take a PC parameter, these take a full NEXT_FRAME. It leaves the old frame predicates alone (making this safe for the branch). > > I figure that should be enough to keep any unwinder, trying to decide if the frame is good, happy :-) > > I'll look to commit this to both the trunk and 6.0 branch early next week. I'll then follow up with patches to: > > - rips out the old predicate methods (mainline only) > > - replace the dwarf2 predicate with a dwarf2 sniffer > Will apply to mainline and branch. Should fix weird edge case where a function has no cfi but is doing a tail call to a no-return function, but the code following the function does have CFI. > > - (I guess) something to work around the dwarf2 unwind info (need to look at Mark's old code) > > I've so far tested it on a RH9 system (with the sniffer enabled) and it hasn't shown a sign of regressing. > > Can anyone point me at testcases that will magically start passing with this fixed? I've checked this part in Andres ;-)