From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Tue, 11 Jul 2000 21:32:30 +0200 (CEST) From: Andrea Arcangeli Subject: Re: [PATCH] 2.2.17pre7 VM enhancement Re: I/O performance on 2.4.0-test2 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-linux-mm@kvack.org Return-Path: To: "Juan J. Quintela" Cc: Rik van Riel , "Stephen C. Tweedie" , Marcelo Tosatti , Jens Axboe , Alan Cox , Derek Martin , Linux Kernel , linux-mm@kvack.org, "David S. Miller" List-ID: On 11 Jul 2000, Juan J. Quintela wrote: >If you are copying in the background a cp and you don't touch your >vi/emacs/whatever pages in 2 hours (i.e. age = 0) then I think that it >is ok for that pages to be swaped out. Notice that the cage pages >will have _initial age_ and the pages of the binaries will have an >_older_ age. If we want to do that we can do that. My design doesn't forbid this. I only avoid the overhead of the inactive list. Also note that what I was really complaining is to threat the lru_cached and lru_mapped list equally. If you threat them equally you get in troubles as I pointed out. I just want to say that lru_mapped have much more priority than lru_cache. If you give the higher priority with a aging factor, or I give higher priority with a different falling back behaviour it doesn't matter (with the difference that I avoid overhead of refiling between lru lists and I avoid to roll ex-mapped-pages in the lru_cache list just to decrease their age). Andrea -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux.eu.org/Linux-MM/