From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from cs1p51.dial.cistron.nl ([62.216.3.52] helo=test.augan) by smtp.cistron.nl with esmtp (Exim 3.13 #1 (Debian)) id 139QgM-0006ch-00 for ; Tue, 04 Jul 2000 13:16:02 +0200 Received: from augan.com (IDENT:roman@serv2.augan [130.1.1.31]) by test.augan (8.9.3/8.8.7) with ESMTP id NAA24458 for ; Tue, 4 Jul 2000 13:15:37 +0200 Message-ID: <3961C759.DA95744E@augan.com> Date: Tue, 04 Jul 2000 13:15:37 +0200 From: Roman Zippel MIME-Version: 1.0 Subject: how does the fs cache work? Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org Return-Path: To: linux-mm List-ID: Hi, I'm currently trying to understand how the fs cache in 2.4 works, but I'm slightly confused, how a few things are supposed to work. Am I seeing it correctly, that if we're looking for a free page we're basically looking into three caches? We check for mapped pages (swap_out), we shrink the lru_cache (shrink_mmap) and we clean the inode cache (shrink_icache_memory). Furthermore the only policy that I see that prevents a page being freed is "was it used last time we checked?", how is that supposed to work under load? Is there any way to share a block that is read from a file and the same read from the block device? A possible user might be e2fsck, but currently it doesn't seem to be problem, as ext2 doesn't use the page cache for meta data. Hmm, I think, that's enough questions for now, I come up with more later. :) bye, Roman -- 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/