Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 21 Nov 2002 19:05:07 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 21 Nov 2002 19:05:07 -0500 Received: from 1-245.ctame701-1.telepar.net.br ([200.181.137.245]:44454 "EHLO 1-245.ctame701-1.telepar.net.br") by vger.kernel.org with ESMTP id ; Thu, 21 Nov 2002 19:05:06 -0500 Date: Thu, 21 Nov 2002 22:12:03 -0200 (BRST) From: Rik van Riel X-X-Sender: riel@imladris.surriel.com To: Kernel Mailing List cc: linux-mm@kvack.org, Arjan van de Ven Subject: [PATCH][RFC] split active lists for rmap 15 Message-ID: X-spambait: aardvark@kernelnewbies.org X-spammeplease: aardvark@nl.linux.org MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1335 Lines: 35 Hi, it turns out that while Arjan's O(1) VM code works fine in most scenarios, it's too swap happy in some cases. In order to try to fix that I've split up the (per zone) active list into an active_anon list (process working set) and an active_cache list (file cache). If the active_cache list is "large" we deactivate those pages preferentially. Of course, if those pages really are referenced they'll just be moved back to the active list and we'll end up swapping out processes ... but in the common case of streaming over lots and lots of file data we'll swap out the processes very slowly, meaning that the machine should stay nicely usable during eg. bitkeeper pulls, backups and apt-get updates. This patch is still experimental, so any testing is welcome. As usual, you can grab it from http://surriel.com/patches/ http://surriel.com/patches/2.4/2.4.19-rmap15-splitactive have fun, Rik -- Bravely reimplemented by the knights who say "NIH". http://www.surriel.com/ http://guru.conectiva.com/ Current spamtrap: october@surriel.com - 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/