Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Date: Fri, 2 Feb 2001 09:36:57 +0100
From: Corinna Vinschen <cygwin AT cygwin DOT com>
To: cygwin AT cygwin DOT com
Subject: Re: Some basic questions
Message-ID: <20010202093657.M19867@cygbert.vinschen.de>
Mail-Followup-To: cygwin AT cygwin DOT com
References: <3A7A25FF DOT 2A2BDB7 AT nc DOT rr DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
In-Reply-To: <3A7A25FF.2A2BDB7@nc.rr.com>; from gsmith@nc.rr.com on Thu, Feb 01, 2001 at 10:14:07PM -0500

On Thu, Feb 01, 2001 at 10:14:07PM -0500, Greg Smith wrote:
> q1) The program obtains the emulated storage in a single malloc().  Under linux,
>      we allow the size to range up to 256M, but under cygwin, the most storage
>      we can malloc at a single time is 96M.  Is there some way we can adjust this limit ?
>      By the same token, we seem to get malloc() failures a lot sooner than we do
>      under linux; I am pretty sure the total amount of space we've tried to obtain is
>      under 256M; again, is this some limit we can bypass ?

Try the following: Open the registry, goto
HKEY_CURRENT_USER\Software\Cygnus Solutions\Cygwin
and add a value "heap_chunk_in_mb" of type DWORD with an appropriate
high value, say, 512 or so. Be sure to stop _all_ Cygwin processes
first.

> q2) What is the state of pthreads under cygwin ?  (The emulator is heavily multi-threaded).
>      Currently we are using an implementation of pthreads from
>      ftp://sources.redhat.com/pub/pthreads-win32, and using the snapshot from August 13
>      (later instances seem to require a dll that I haven't been successful in creating).

Cygwin itself is not fully thead-aware. Contributions are welcome.

> q3) I am having a difficult time debugging the multi-threaded program both under
>      cygwin and linux (Redhat 7.0 + the 2.4 kernel and associated requisites).  I
>      assume my problems under cygwin is due to using a foreign pthreads implementation ?
>      I would also be extremely grateful if some kind person could direct me to
>      some instructions for debugging under linux, too.  By the way, the cygwin debugger
>      is impressive, is there a linux equivalent too ?  (Afraid my ignorance is showing
>      here, I'm actually a mainframer type myself).

gdb is part of all Linux distributions. If you are talking about the
GUI, try `gdb -w'. On Windows, the GUI is default and you get the
CLI version by calling `gdb -nw'. On most Linuces it's the other way around.

> q4) One reason I upgraded to linux kernel 2.4 is that the program works, in my
>      personal experience, 20% faster under cygwin than under linux.  Even with the
>      new kernel, I can not recoup the performance difference.  Does anyone have
>      a plausible explanation ?

Nope. I would _never_ expect that assuming the hardware is identical.
The application should be way faster under Linux. I fear you have to debug
this problem intensively.

However, just as a hint, this is not a Linux mailing list. Please don't
begin posting Linux specific problems here.

Thanks,
Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Developer                                mailto:cygwin AT cygwin DOT com
Red Hat, Inc.

--
Want to unsubscribe from this list?
Check out: http://cygwin.com/ml/#unsubscribe-simple