[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <Pine.LNX.4.64.0709171501520.29993@schroedinger.engr.sgi.com>
Date: Mon, 17 Sep 2007 15:03:07 -0700 (PDT)
From: Christoph Lameter <clameter@....com>
To: Jörn Engel <joern@...fs.org>
cc: Andrew Morton <akpm@...ux-foundation.org>,
Nick Piggin <nickpiggin@...oo.com.au>, andrea@...e.de,
torvalds@...ux-foundation.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, Christoph Hellwig <hch@....de>,
Mel Gorman <mel@...net.ie>,
William Lee Irwin III <wli@...omorphy.com>,
David Chinner <dgc@....com>,
Jens Axboe <jens.axboe@...cle.com>,
Badari Pulavarty <pbadari@...il.com>,
Maxim Levitsky <maximlevitsky@...il.com>,
Fengguang Wu <fengguang.wu@...il.com>,
swin wang <wangswin@...il.com>, totty.lu@...il.com,
hugh@...itas.com
Subject: Re: [00/41] Large Blocksize Support V7 (adds memmap support)
On Sun, 16 Sep 2007, Jörn Engel wrote:
> I bet! My (false) assumption was the same as Goswin's. If non-movable
> pages are clearly seperated from movable ones and will evict movable
> ones before polluting further mixed superpages, Nick's scenario would be
> nearly infinitely impossible.
>
> Assumption doesn't reflect current code. Enforcing this assumption
> would cost extra overhead. The amount of effort to make Christoph's
> approach work reliably seems substantial and I have no idea whether it
> would be worth it.
My approach is based on Mel's code and is already working the way you
describe. Page cache allocs are marked __GFP_MOVABLE by Mel's work.
Powered by blists - more mailing lists