Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sat, 30 Dec 2000 18:16:19 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 30 Dec 2000 18:16:09 -0500 Received: from neon-gw.transmeta.com ([209.10.217.66]:39950 "EHLO neon-gw.transmeta.com") by vger.kernel.org with ESMTP id ; Sat, 30 Dec 2000 18:15:53 -0500 Date: Sat, 30 Dec 2000 14:44:56 -0800 (PST) From: Linus Torvalds To: "Eric W. Biederman" cc: Alexander Viro , Daniel Phillips , linux-kernel@vger.kernel.org Subject: Re: [RFC] Generic deferred file writing In-Reply-To: Message-ID: 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 On 30 Dec 2000, Eric W. Biederman wrote: > > One other thing to think about for the VFS/MM layer is limiting the > total number of dirty pages in the system (to what disk pressure shows > the disk can handle), to keep system performance smooth when swapping. This is a separate issue, and I think that it is most closely tied in to the "RSS limit" kind of patches because of the memory mapping issues. If you've seen the RSS rlimit patch (it's been posted a few times this week), then you could think of that modified by a "Resident writable pages Set Size" approach. Not just for shared mappings - this is also an issue with limiting swapout. (I actually don't think that RSS is all that interesting, it's really the "potentially dirty RSS" that counts for VM behaviour - everything else can be dropped easily enough) Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/