Date: Wed, 16 Sep 2009 16:03:58 -0700
From: Andrew Morton <akpm@linux-foundation.org>
To: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Cc: mingo@elte.hu, rusty@rustcorp.com.au, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] stop_machine: disable preempt in stop machine path
Message-Id: <20090916160358.7faadb5f.akpm@linux-foundation.org>
In-Reply-To: <4AB0933E.8020901@cn.fujitsu.com>
References: <4AB0933E.8020901@cn.fujitsu.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 754
Lines: 20

On Wed, 16 Sep 2009 15:26:54 +0800
Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> wrote:

> We can disable preempt in stop_cpu(), It can reduce the cpu's waiting time
> and make stop_machine path faster.
> 

"it can"?  But does it?  Are there any measurements or observations to
back this up?

The kstop threads are already created with create_rt_workqueue() (which
has super-secret undocumented semantics which I have cleverly
discovered mean "use SCHED_FIFO").

So we won't be getting preempted here anyway?
--
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/