Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753011AbbBXWJA (ORCPT ); Tue, 24 Feb 2015 17:09:00 -0500 Received: from mx1.redhat.com ([209.132.183.28]:53733 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751973AbbBXWI6 (ORCPT ); Tue, 24 Feb 2015 17:08:58 -0500 Date: Tue, 24 Feb 2015 17:08:44 -0500 From: Rafael Aquini To: Linus Torvalds Cc: linux-mm , Andrew Morton , Johannes Weiner , Rik van Riel , David Rientjes , Linux Kernel Mailing List , loberman@redhat.com, Larry Woodman , Raghavendra K T Subject: Re: [PATCH] mm: readahead: get back a sensible upper limit Message-ID: <20150224220843.GL19014@t510.redhat.com> References: <9cc2b63100622f5fd17fa5e4adc59233a2b41877.1424779443.git.aquini@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1046 Lines: 29 On Tue, Feb 24, 2015 at 01:56:25PM -0800, Linus Torvalds wrote: > On Tue, Feb 24, 2015 at 4:58 AM, Rafael Aquini wrote: > > > > This patch brings back the old behavior of max_sane_readahead() > > Yeah no. > > There was a reason that code was killed. No way in hell are we > bringing back the insanities with node memory etc. > Would you consider bringing it back, but instead of node memory state, utilizing global memory state instead? > Also, we have never actually heard of anything sane that actualyl > depended on this. Last time this came up it was a made-up benchmark, > not an actual real load that cared. > > Who can possibly care about this in real life? > People filing bugs complaining their applications that memory map files are getting hurt by it. -- Rafael -- 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/