From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261605AbUBPLmb (ORCPT ); Mon, 16 Feb 2004 06:42:31 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263996AbUBPLmb (ORCPT ); Mon, 16 Feb 2004 06:42:31 -0500 Received: from gate.corvil.net ([213.94.219.177]:20741 "EHLO corvil.com") by vger.kernel.org with ESMTP id S261605AbUBPLm3 (ORCPT ); Mon, 16 Feb 2004 06:42:29 -0500 Message-ID: <4030ACA3.6020009@draigBrady.com> Date: Mon, 16 Feb 2004 11:42:27 +0000 From: P@draigBrady.com User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031016 X-Accept-Language: en-us, en MIME-Version: 1.0 To: Linux Kernel Mailing List Subject: IDE and locking Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I'm working on an embedded system here that basically counts ethernet packets. I'm using 2.4.20 and I'm noticing that when files are read from the compact flash (ext2), ethernet packets are being dropped. So to work around the problem, I'm precaching all the files (12MB) at boot like: find / -type f | while read file; do dd if=$file of=/dev/null; done So questions. 1. Is this due to the BKL around the IDE subsystem? 2. Are writes as susceptible as reads to the problem? I'm guessing if the compact flash had a write cache buffer I would be OK as I only write max 2MB at a time? 3. There was talk about removing the BKL in 2.6 with reference to: http://sourceforge.net/project/showfiles.php?group_id=8875 Has this work been included? 4. Is there a max number of files that can be cached by linux? 5. Will the files be removed at any stage from the cache if there is no memory pressure? 6. Can I reserve memory for the file cache? cheers. -- Pádraig Brady - http://www.pixelbeat.org