Date: Tue, 26 Apr 2011 14:25:35 +0800
From: Wu Fengguang <fengguang.wu@intel.com>
To: Dave Young <hidave.darkstar@gmail.com>
Cc: linux-mm <linux-mm@kvack.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: readahead and oom
Message-ID: <20110426062535.GB19717@localhost>
References: <BANLkTin8mE=DLWma=U+CdJaQW03X2M2W1w@mail.gmail.com>
 <20110426055521.GA18473@localhost>
 <BANLkTik8k9A8N8CPk+eXo9c_syxJFRyFCA@mail.gmail.com>
 <BANLkTim0MNgqeh1KTfvpVFuAvebKyQV8Hg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <BANLkTim0MNgqeh1KTfvpVFuAvebKyQV8Hg@mail.gmail.com>
User-Agent: Mutt/1.5.20 (2009-06-14)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1515
Lines: 37

On Tue, Apr 26, 2011 at 02:07:17PM +0800, Dave Young wrote:
> On Tue, Apr 26, 2011 at 2:05 PM, Dave Young <hidave.darkstar@gmail.com> wrote:
> > On Tue, Apr 26, 2011 at 1:55 PM, Wu Fengguang <fengguang.wu@intel.com> wrote:
> >> On Tue, Apr 26, 2011 at 01:49:25PM +0800, Dave Young wrote:
> >>> Hi,
> >>>
> >>> When memory pressure is high, readahead could cause oom killing.
> >>> IMHO we should stop readaheading under such circumstances。If it's true
> >>> how to fix it?
> >>
> >> Good question. Before OOM there will be readahead thrashings, which
> >> can be addressed by this patch:
> >>
> >> http://lkml.org/lkml/2010/2/2/229
> >
> > Hi, I'm not clear about the patch, could be regard as below cases?
> > 1) readahead alloc fail due to low memory such as other large allocation
> 
> For example vm balloon allocate lots of memory, then readahead could
> fail immediately and then oom

If true, that would be the problem of vm balloon. It's not good to
consume lots of memory all of a sudden, which will likely impact lots
of kernel subsystems.

btw readahead page allocations are completely optional. They are OK to
fail and in theory shall not trigger OOM on themselves. We may
consider passing __GFP_NORETRY for readahead page allocations.

Thanks,
Fengguang

--
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/