From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 2809 invoked by alias); 19 Nov 2003 07:08:38 -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 2802 invoked from network); 19 Nov 2003 07:08:38 -0000 Received: from unknown (HELO zenia.home) (12.223.225.216) by sources.redhat.com with SMTP; 19 Nov 2003 07:08:38 -0000 Received: by zenia.home (Postfix, from userid 5433) id A270E20766; Wed, 19 Nov 2003 02:06:05 -0500 (EST) To: Daniel Jacobowitz Cc: gdb-patches@sources.redhat.com, ezannoni@redhat.com Subject: Re: [RFA] Delete another call to dwarf2read.c:decode_locdesc References: <20031119012558.GA4202@nevyn.them.org> From: Jim Blandy Date: Wed, 19 Nov 2003 07:08:00 -0000 In-Reply-To: <20031119012558.GA4202@nevyn.them.org> Message-ID: User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-SW-Source: 2003-11/txt/msg00391.txt.bz2 Daniel Jacobowitz writes: > The global variables set according to the DW_OP_frame_base are no longer > used anywhere except decode_locdesc, and decode_locdesc is never called when > we should expect a frame-relative object; see the updated comments in the > patch below. This function is now only used for member offsets, etc.; > partial symbols (DW_TAG_variable only); and common blocks, which act as > global variables. The partial symbol case should either be eliminated or > changed to use computed locations. The common block case probably likewise. > And the remaining cases expect constants. > > So DW_OP_breg* and DW_OP_fbreg are dead wood at this point and can be > removed from this function. They're handled entirely in the LOC_COMPUTED > support code. > > These happened to be next on my campaign to whack global state from the > dwarf2 reader; marching slowly down the file. Tested on i386-linux/dwarf2. > OK? Superb! Please commit.