Date: Thu, 22 Jan 2009 09:34:57 +0100
From: Ingo Molnar <mingo@elte.hu>
To: Mandeep Singh Baines <msb@google.com>
Cc: fweisbec@gmail.com, linux-kernel@vger.kernel.org, rientjes@google.com,
       mbligh@google.com, thockin@google.com
Subject: Re: [PATCH v2] softlockup: remove hung_task_check_count
Message-ID: <20090122083457.GC7438@elte.hu>
References: <20090121111314.GA23469@elte.hu> <20090122005405.GA6067@google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20090122005405.GA6067@google.com>
User-Agent: Mutt/1.5.18 (2008-05-17)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1182
Lines: 34


* Mandeep Singh Baines <msb@google.com> wrote:

> > do the need_resched() check first (it's very lighweight) - and thus 
> > only do the heavy ops (get-task-struct & tasklist_lock unlock) if that 
> > is set?
> 
> Wanted to upper-bound the amount of time the lock is held. In order to 
> give others a chance to write_lock the tasklist, released the lock 
> regardless of whether a re-schedule is need.

but this:

> +static void check_hung_reschedule(struct task_struct *t)
> +{
> +	get_task_struct(t);
> +	read_unlock(&tasklist_lock);
> +	if (need_resched())
> +		schedule();
> +	read_lock(&tasklist_lock);
> +	put_task_struct(t);
> +}

does not actually achieve that. Releasing a lock does not mean that other 
CPUs will immediately be able to get it - if the ex-owner quickly 
re-acquires it then it will often succeed in doing so. Perhaps adding a 
cpu_relax() would increase the chance ... but still, it looks a bit weird.

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/