From owner-freebsd-threads@FreeBSD.ORG  Fri Jun 17 10:17:28 2005
Return-Path: <owner-freebsd-threads@FreeBSD.ORG>
X-Original-To: freebsd-threads@freebsd.org
Delivered-To: freebsd-threads@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125])
	by hub.freebsd.org (Postfix) with ESMTP id 708AC16A41C;
	Fri, 17 Jun 2005 10:17:28 +0000 (GMT) (envelope-from ant@emict.com)
Received: from mail.emict.com (brig.emict.com [212.90.172.42])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 0023C43D53;
	Fri, 17 Jun 2005 10:17:27 +0000 (GMT) (envelope-from ant@emict.com)
Received: from [10.0.0.232] (unknown [10.0.0.232])
	by mail.emict.com (Postfix) with ESMTP
	id 29A8133ED1; Fri, 17 Jun 2005 13:17:26 +0300 (EEST)
From: Andriy Tkachuk <ant@emict.com>
Organization: eMICT
To: David Xu <davidxu@freebsd.org>
Date: Fri, 17 Jun 2005 13:17:23 +0300
User-Agent: KMail/1.8
References: <200506161309.51431.ant@emict.com>
	<20050616141857.GA63602@stack.nl> <42B1F414.9040203@freebsd.org>
In-Reply-To: <42B1F414.9040203@freebsd.org>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200506171317.24632.ant@emict.com>
Cc: Marc Olzheim <marcolz@stack.nl>, freebsd-threads@freebsd.org
Subject: Re: more than 2k threads with -lpthread or -lthr
X-BeenThere: freebsd-threads@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Threading on FreeBSD <freebsd-threads.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-threads>, 
	<mailto:freebsd-threads-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-threads>
List-Post: <mailto:freebsd-threads@freebsd.org>
List-Help: <mailto:freebsd-threads-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-threads>,
	<mailto:freebsd-threads-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Fri, 17 Jun 2005 10:17:28 -0000

thank you guys, all started working.

actually the problem with sigsegv was not about
stack size - my fault - program was killed with
this signal even when number of threads was not large.

on 5.4-STABLE Fri Jun 17 09:33:54 EEST 2005
kernel this problem disappeared. Note, that with old
kernel (from 5.4-RELEASE) but with new libpthread
(from 5.4-STABLE) the problem still was present, that's why
i untimely sayd that this bug present in RELENG_5_x.
So i suppose, that something in 5.4-STABLE kernel is fixed
regarding threads compared to 5.4-RELEASE kernel.

after this another problem appeared - no more, than
30xx threads where started. But i set stacksize
to 2 pages (8K) and all started working fine.

Again: sorry for misinformation and thank you for help.

Regards,
  Andriy Tkachuk.

On Friday 17 June 2005 00:50, David Xu wrote:
> Marc Olzheim wrote:
> 
> >On Thu, Jun 16, 2005 at 06:52:20PM +0800, David Xu wrote:
> >  
> >
> >>Both libpthread and libthr use 1M bytes stack on 32bits platform
> >>and 2M bytes stack on 64bits platform. I think 2k threads needs
> >>2G stack on i386, this is too large.
> >>    
> >>
> >
> >In which case pthread_attr_setstacksize() might help ?
> >
> >Marc
> >  
> >
> pthread_attr_setstacksize definitively helps, libc_r default uses 64k
> stack for thread, this is very small stack.
> 
> Here two sysctl also have to tuned to numbers larger than 2000 for
> libthr and libpthread:
> kern.threads.max_groups_per_proc
> kern.threads.max_threads_per_proc
> 
> David Xu
> 
>