Date: Sat, 26 Sep 2009 05:49:36 +0200
From: Andi Kleen <andi@firstfloor.org>
To: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
       Hugh Dickins <hugh.dickins@tiscali.co.uk>,
       Nick Piggin <npiggin@suse.de>, Andi Kleen <andi@firstfloor.org>,
       linux-mm@kvack.org, LKML <linux-kernel@vger.kernel.org>
Subject: Re: [RFC][PATCH] HWPOISON: remove the unsafe __set_page_locked()
Message-ID: <20090926034936.GK30185@one.firstfloor.org>
References: <20090926031537.GA10176@localhost>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20090926031537.GA10176@localhost>
User-Agent: Mutt/1.4.2.2i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1002
Lines: 22

On Sat, Sep 26, 2009 at 11:15:37AM +0800, Wu Fengguang wrote:
> The swap cache and page cache code assume that they 'own' the newly
> allocated page and therefore can disregard the locking rules. However
> now hwpoison can hit any time on any page.
> 
> So use the safer lock_page()/trylock_page(). The main intention is not
> to close such a small time window of memory corruption. But to avoid
> kernel oops that may result from such races, and also avoid raising
> false alerts in hwpoison stress tests.
> 
> This in theory will slightly increase page cache/swap cache overheads,
> however it seems to be too small to be measurable in benchmark.

Thanks. Can you please describe what benchmarks you used?

Acked-by: Andi Kleen <ak@linux.intel.com>
-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/