A glibc bug?

[hjl@lucon.org (H.J. Lu)]

From chris@classnet.med.miami.edu  Mon May 17 11:33:51 1999
Return-Path: <chris@classnet.med.miami.edu>
Delivered-To: hjl@lucon.org
Received: from beezer.med.miami.edu (beezer.med.miami.edu [129.171.143.105])
	by ocean.lucon.org (Postfix) with ESMTP id 5DC3257B7
	for <hjl@lucon.org>; Mon, 17 May 1999 11:33:50 -0700 (PDT)
Received: from beezer.med.miami.edu (chris@beezer.med.miami.edu [129.171.143.105])
	by beezer.med.miami.edu (8.9.3/8.9.3/Debian/GNU) with SMTP id OAA19960
	for <hjl@lucon.org>; Mon, 17 May 1999 14:33:48 -0400
Date: Mon, 17 May 1999 14:33:46 -0400 (EDT)
From: Christopher C Chimelis <chris@classnet.med.miami.edu>
X-Sender: chris@beezer.med.miami.edu
To: "H.J. Lu" <hjl@lucon.org>
Subject: Re: Easy question :-)
In-Reply-To: <19990516002408.64E6A57B7@ocean.lucon.org>
Message-ID: <Pine.LNX.3.96.990517143055.19834B-100000@beezer.med.miami.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII


On Sat, 15 May 1999, H.J. Lu wrote:

> My goal is merge my changes into the official binutils.
> The main problem is the Linux/ARM support in the offical
> binutils is not working. Due to various reasons, they
> cannot take the Linux/ARM support from my binutils. They
> have to start from scratch. Before the offical binutils
> is working on Linux/ARM, I cannot start working on the
> merge. I am waiting for the word from the Linux/ARM
> people.

Ah, ok.  Thanks.  Oh, also btw, have you heard anything about problems
with gprof under glibc 2.1 systems with egcs?  I'm having a helluva time
trying to figure it out.

This is the report I got, fyi...I've confirmed it's malfunction on alpha
and i386 and have also confirmed that things work as they should under
glibc 2.0.7ish libs and egcs 2.91.66.

> When I compile a program with -pg to attempt to do some profiling, and
> run the program, gprof complains that gmon.out is missing call-graph
> data. Not sure if this is a egcs or a glibc problem, but I'm pretty
> sure it's not a gprof problem, as the gmon.out files being produced
> seem to be unusually small. It used to work fine back when I was using
> glibc2.0 & gcc (non-egcs) from hamm.
>
> What's in the actual .c file doesn't seem to matter, it doesn't work
> for any I try. Here's what I am doing:
>
> gcc -Wall -g -pg -o test test.c
> ./test
> gprof test
>
> Which produces the output:
>
> gprof: gmon.out file is missing call-graph data



-- 
H.J. Lu (hjl@gnu.org)