From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 15891 invoked by alias); 24 Feb 2003 07:07:19 -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 15884 invoked from network); 24 Feb 2003 07:07:18 -0000 Received: from unknown (HELO dberlin.org) (69.3.5.6) by 172.16.49.205 with SMTP; 24 Feb 2003 07:07:18 -0000 Received: from [127.0.0.1] (HELO dberlin.org) by dberlin.org (CommuniGate Pro SMTP 4.0.6) with ESMTP-TLS id 2980447; Mon, 24 Feb 2003 02:07:18 -0500 Date: Mon, 24 Feb 2003 07:07:00 -0000 From: Daniel Berlin To: Daniel Jacobowitz cc: Michael Elizabeth Chastain , , Subject: Re: [rfa] new test, pr-1090.exp, multi-register variables In-Reply-To: <20030224041749.GB20002@nevyn.them.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-SW-Source: 2003-02/txt/msg00573.txt.bz2 On Sun, 23 Feb 2003, Daniel Jacobowitz wrote: > On Sun, Feb 23, 2003 at 09:55:44PM -0600, Michael Elizabeth Chastain wrote: > > This is a new test script for pr gdb/1090, which is about register > > variables which occupy several registers. Both gdb 5.3 and gdb > > HEAD%20030223 get this wrong. In my testbed, this happens with > > gcc 2.95.3; gcc v3 compilers don't allocate multi-register variables. > > > > The symptom of this bug is that gdb prints the first word of a > > structure correctly but botches the second word. The test has a nice > > KFAIL for this. > > > > Okay to commit? > > You may want to see gdb/214. Mark posted a proposal for solving it to > gdb@ some weeks ago and it never went further; I don't remember what > the verdict was. > > Some day very soon we'll have GDB support to describe this explicitly; > some day after that GCC will output the appropriate debug info > (DW_OP_piece). May be a little while. As I mentioned, GCC already outputs DW_OP_piece in one specific case. If you'd like it to do it more often, let me know when you want it outputted, and i'll make it so. > >