From mboxrd@z Thu Jan  1 00:00:00 1970
From: Ingo Molnar <mingo@elte.hu>
Subject: Re: [PATCH] kvm tools: Fix pthread mutex error checks
Date: Sat, 9 Apr 2011 15:15:47 +0200
Message-ID: <20110409131547.GA12997@elte.hu>
References: <1302350278-5478-1-git-send-email-penberg@kernel.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: kvm@vger.kernel.org, Asias He <asias.hejun@gmail.com>,
	Cyrill Gorcunov <gorcunov@gmail.com>
To: Pekka Enberg <penberg@kernel.org>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx3.mail.elte.hu ([157.181.1.138]:40818 "EHLO mx3.mail.elte.hu"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1752176Ab1DINPw (ORCPT <rfc822;kvm@vger.kernel.org>);
	Sat, 9 Apr 2011 09:15:52 -0400
Content-Disposition: inline
In-Reply-To: <1302350278-5478-1-git-send-email-penberg@kernel.org>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>


* Pekka Enberg <penberg@kernel.org> wrote:

> The pthread_mutex_{lock|unlock} functions return non-zero, not negative number
> upon error. Fix that wrong assumption in the code.

glibc/pthreads mutex API semantics are pretty silly IMO.

I *think* it would be better to try to match the kernel API here, and provide 
trivial wrappers around mutex_lock()/mutex_unlock(). We wont ever bring down 
threads in a hostile way, so we wont actually need the error returns. CPU 
threads should probably only exit once the kvm process exits, after all cleanup 
has been done.

So mutex_lock() could be implemented as something like:

	void mutex_lock(pthread_mutex_t mutex)
	{
		if (pthread_mutex_lock(mutex) != 0)
			die("unexpected pthread_mutex_lock() failure!");
	}

That way usage would be more obvious and more familar to kernel developers :-) 

[ It would also open up the possibility, in the far future, to bring lockdep to 
  user-space ;-) ]

What do you think?

Thanks,

	Ingo