From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 20343 invoked by alias); 18 Jan 2007 09:25:32 -0000 Received: (qmail 20242 invoked by uid 22791); 18 Jan 2007 09:25:31 -0000 X-Spam-Check-By: sourceware.org Received: from mxout.hispeed.ch (HELO smtp.hispeed.ch) (62.2.95.247) by sourceware.org (qpsmtpd/0.31) with ESMTP; Thu, 18 Jan 2007 09:25:25 +0000 Received: from indel.ch (84-73-11-232.dclient.hispeed.ch [84.73.11.232]) by smtp.hispeed.ch (8.12.11.20060308/8.12.11/taifun-1.0) with SMTP id l0I9PE3f008104 for ; Thu, 18 Jan 2007 10:25:21 +0100 Received: from FABI.indel.ch [192.168.1.91] by indel.ch [127.0.0.1] with SMTP (MDaemon.v2.7.SP5.R) for ; Thu, 18 Jan 2007 10:25:08 +0100 Message-Id: <5.2.0.9.1.20070118102453.018662f0@NT_SERVER> X-Sender: cenedese@NT_SERVER (Unverified) X-Mailer: QUALCOMM Windows Eudora Version 5.2.0.9 Date: Thu, 18 Jan 2007 09:25:00 -0000 To: gdb@sourceware.org From: Fabian Cenedese Subject: "Preprocess" dwarf info? Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" X-MDaemon-Deliver-To: gdb@sourceware.org X-Return-Path: cenedese@indel.ch X-Virus-Status: Clean X-DCC-spamcheck-01.tornado.cablecom.ch-Metrics: smtp-04.tornado.cablecom.ch 1377; Body=2 Fuz1=2 Fuz2=2 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: 2007-01/txt/msg00279.txt.bz2 Hi I'm now trying out stabs and dwarf debug info. With stabs all gdb symbols get filled with data like type (LOC_STATIC, LOC_REGISTER etc) and values (SYMBOL_VALUE, SYMBOL_VALUE_ADDRESS). If the symbol gets evaluated only few target accesses are needed (registers, memory) for the final value. With dwarf this info is not stored expressively anymore if a symbol is of type LOC_COMPUTED or LOC_COMPUTED_ARG. Instead gdb will evaluate the dwarf debug "code" to return the value every time. I know that there may be symbol "locations" that are not representable with just 1 or 2 values. But still values such as register numbers or offsets will always be the same. So is there a way to somehow "preprocess" the dwarf info to get out the same info gdb stores when using the stabs info? Thanks bye Fabi