From: Andi Kleen <ak@suse.de>
To: Arjan van de Ven <arjan@intel.linux.com>
Subject: Re: [Patch 1/3] prefetch the mmap_sem in the fault path
Date: Thu, 23 Feb 2006 10:39:35 +0100
User-Agent: KMail/1.9.1
Cc: linux-kernel@vger.kernel.org, akpm@osdl.org
References: <1140686152.2972.25.camel@laptopd505.fenrus.org> <1140687007.4672.6.camel@laptopd505.fenrus.org>
In-Reply-To: <1140687007.4672.6.camel@laptopd505.fenrus.org>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200602231039.36338.ak@suse.de>
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 893
Lines: 19

On Thursday 23 February 2006 10:30, Arjan van de Ven wrote:
> In a micro-benchmark that stresses the pagefault path, the down_read_trylock
> on the mmap_sem showed up quite high on the profile. Turns out this lock is
> bouncing between cpus quite a bit and thus is cache-cold a lot. This patch
> prefetches the lock (for write) as early as possible (and before some other
> somewhat expensive operations). With this patch, the down_read_trylock
> basically fell out of the top of profile.

It is hard to believe because you effectively didn't do the prefetch
very early

(e.g. the patch from your prefetch to taking the lock is quite short) 

-Andi
-
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/