Mailing-List: contact cygwin-developers-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-developers-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-developers/>
List-Post: <mailto:cygwin-developers AT cygwin DOT com>
List-Help: <mailto:cygwin-developers-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-developers-owner AT cygwin DOT com
Delivered-To: mailing list cygwin-developers AT cygwin DOT com
X-WM-Posted-At: avacado.atomice.net; Sun, 30 Jun 02 12:36:32 +0100
Message-ID: <003301c2202a$61f9c7b0$0100a8c0@advent02>
From: "Chris January" <chris AT atomice DOT net>
To: <cygwin-developers AT cygwin DOT com>
References: <3D1B4043 DOT EF2CA37A AT ieee DOT org> <3D1B4043 DOT EF2CA37A AT ieee DOT org> <3 DOT 0 DOT 5 DOT 32 DOT 20020627224059 DOT 0080d5b0 AT mail DOT attbi DOT com> <20020629031639 DOT GA14536 AT redhat DOT com> <005801c21f59$e116e680$0100a8c0 AT advent02> <20020629172702 DOT GA29746 AT redhat DOT com>  <20020630002759 DOT GA20659 AT redhat DOT com>
Subject: Re: Slowdown
Date: Sun, 30 Jun 2002 12:36:31 +0100
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
X-Priority: 3
X-MSMail-Priority: Normal
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000

> >>If this still happens to you after today's checkin, then I'd really
> >>appreciate if you could track the problem down.
> >
> >I'm doing so now.  I have already provided a full stack backtrace, but
> >I will attempt to track the error done myself.
>
> I saw the full stack backtrace.  It made no sense to me or I would have
> investigated further.  If I was reading it correctly, cygwin was trying
> to free some memory during initialization that should have been NULL.
> AFAICT, it was trying to allocate the fd table during startup.  The fd
> table should have been NULL at that point.
It seems to be heap corruption as far as I can tell. The problem moved
around when I inserted some printf's to see what where it was failing. So
far I've seen it fail in cfree, _close, and fork_child.
My CYGWIN environment variable is set to ntsec check_case:adjust ntea
glob:noignorecase tty binmode.
I will try to investigate further, but it might be better if someone more
familiar with the recent changes has a look.

Chris