From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 18228 invoked by alias); 14 Jun 2010 12:07:47 -0000 Received: (qmail 18217 invoked by uid 22791); 14 Jun 2010 12:07:46 -0000 X-SWARE-Spam-Status: No, hits=-2.0 required=5.0 tests=AWL,BAYES_00,DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FROM X-Spam-Check-By: sourceware.org Received: from mail-wy0-f169.google.com (HELO mail-wy0-f169.google.com) (74.125.82.169) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Mon, 14 Jun 2010 12:07:42 +0000 Received: by wyf28 with SMTP id 28so4082177wyf.0 for ; Mon, 14 Jun 2010 05:07:40 -0700 (PDT) Received: by 10.216.88.144 with SMTP id a16mr2290060wef.77.1276517259789; Mon, 14 Jun 2010 05:07:39 -0700 (PDT) MIME-Version: 1.0 Received: by 10.216.167.144 with HTTP; Mon, 14 Jun 2010 05:07:19 -0700 (PDT) In-Reply-To: References: <20100607223524.GA2897@host0.dyn.jankratochvil.net> <20100608184224.GA6814@host0.dyn.jankratochvil.net> <20100609054948.GA5627@host0.dyn.jankratochvil.net> From: Frederic Riss Date: Mon, 14 Jun 2010 12:07:00 -0000 Message-ID: Subject: Re: RFC: optimized-out pieces To: Tom Tromey Cc: Jan Kratochvil , gdb-patches@sourceware.org Content-Type: text/plain; charset=ISO-8859-1 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: 2010-06/txt/msg00304.txt.bz2 Hi Tom, On 11 June 2010 17:34, Tom Tromey wrote: >>>>>> "Tom" == Tom Tromey writes: > > Tom> New patch appended. > > I am checking in the latest version of this patch. I think this patch regresses on architectures that mark registers as optimized in their CFI. Specifically when executing the 'return' command. When iterating over the unwound registers to gather the values to assign to the registers in pop_frame, frame_register_unwind calls value_contents_all that will error out if it finds an unsaved register. I think it's expected to have such registers when doing a return. I fixed that by preventing frame_register_unwind to call value_contents_all in the optimized case (not providing the one-liner as I don't have the copyright assignment in place anymore). Cheers, Fred