From: Alexey Lyahkov Subject: Re: page eviction from the buddy cache Date: Mon, 22 Apr 2013 15:14:55 +0300 Message-ID: <128E4C4E-555C-43B6-9BA4-7914CBAF5B62@gmail.com> References: <51504A40.6020604@ya.ru> <20130327150743.GC14900@thunk.org> <3C8EEEF8-C1EB-4E3D-8DE6-198AB1BEA8C0@gmail.com> <515CD665.9000300@gmail.com> <239AD30A-2A31-4346-A4C7-8A6EB8247990@gmail.com> <51730619.3030204@fastmail.fm> <20130420235718.GA28789@thunk.org> Mime-Version: 1.0 (Apple Message framework v1283) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8BIT Cc: Bernd Schubert , Will Huck , Hugh Dickins , Andrew Perepechko , linux-ext4@vger.kernel.org, akpm@linux-foundation.org, linux-mm@kvack.org, mgorman@suse.de To: Theodore Ts'o Return-path: Received: from mail-wi0-f171.google.com ([209.85.212.171]:40041 "EHLO mail-wi0-f171.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753099Ab3DVMPF convert rfc822-to-8bit (ORCPT ); Mon, 22 Apr 2013 08:15:05 -0400 Received: by mail-wi0-f171.google.com with SMTP id l13so4241420wie.16 for ; Mon, 22 Apr 2013 05:15:03 -0700 (PDT) In-Reply-To: <20130420235718.GA28789@thunk.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: Theodore, May you provide more details about discussion and new interface? someone from mm think we need to implement mark_acccessed_force() with avoiding is in LRU checks? and move single page directly without waiting a LRU drain ? On Apr 21, 2013, at 02:57, Theodore Ts'o wrote: > On Sat, Apr 20, 2013 at 11:18:17PM +0200, Bernd Schubert wrote: >> Alex, Andrew, >> >> did you notice the patch Ted just sent? >> ("ext4: mark all metadata I/O with REQ_META") > > This patch was sent to fix another issue that was brought up at Linux > Storage, Filesystem, and MM workshop. I did bring up this issue with > Mel Gorman while at LSF/MM, and as a result, tThe mm folks are going > to look into making mark_page_accessed() do the right thing, or > perhaps provide us with new interface. The problem with forcing the > page to be marked as activated is this would cause a TLB flush, which > would be pointless since this these buddy bitmap pages aren't actually > mapped in anywhere. > > - Ted