From mboxrd@z Thu Jan 1 00:00:00 1970 From: Damian Hobson-Garcia Subject: Re: [RFC][PATCH] fs/buffer.c: Revoke LRU when trying to drop buffers Date: Fri, 05 Jul 2013 11:01:33 +0900 Message-ID: <51D628FD.2010806@igel.co.jp> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Marek Szyprowski , viro@zeniv.linux.org.uk, linux-fsdevel@vger.kernel.org, "linux-kernel@vger.kernel.org" To: lauraa@codeaurora.org Return-path: Received: from mail-pd0-f182.google.com ([209.85.192.182]:54895 "EHLO mail-pd0-f182.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755500Ab3GECBg (ORCPT ); Thu, 4 Jul 2013 22:01:36 -0400 Received: by mail-pd0-f182.google.com with SMTP id r10so1533888pdi.41 for ; Thu, 04 Jul 2013 19:01:36 -0700 (PDT) Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Hi, I've come across a problem allocating chunks of CMA memory on a system under heavy disk access load which seems very similar to the situation described as the rationale for this patch: https://lkml.org/lkml/2012/9/11/335 Basically buffer headers sitting on the LRU list preventing them from being dropped during page migration. I've tried the patch and it seems to solve my problems, so I'm wondering if it is worth resubmitting. The discussion on the original thread seems to have faded out. Does anyone know if it was continued somewhere else? Thank you, Damian -- Damian Hobson-Garcia IGEL Co.,Ltd http://www.igel.co.jp