From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 26 Apr 2002 09:12:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 26 Apr 2002 09:12:41 -0400 Received: from web12504.mail.yahoo.com ([216.136.173.196]:262 "HELO web12504.mail.yahoo.com") by vger.kernel.org with SMTP id ; Fri, 26 Apr 2002 09:12:40 -0400 Message-ID: <20020426131240.11007.qmail@web12504.mail.yahoo.com> Date: Fri, 26 Apr 2002 06:12:40 -0700 (PDT) From: sanjay kumar To: linux-kernel@vger.kernel.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 Hi, Any file system like ext2, JFS etc are aware about the meta-data of filesystem and file-data, when it writes some file-blocks on disk. I want to know whether it is possible to differentiate between meta-data and file-data in block device layer (I mean without using file-system specific functions). In other words, whether it is possible to know that the given buffer (of buffer cache), contains the meta-data/file-data. Here meta-data means filesystem specific data like superblock, group descriptor, bitmap blocks (if file system supports), blocks containing inode and extent information, information about directories and files, blocks containing indirect pointers for any file (applicable in indirect addressing in ext2), etc. Is it true that all file systems (if not all, then mostly), use block_read_full_page() and block_write_full_page() to read/write the file-data ? And these functions are not used for reading/writing the meta-data. -Sanjay Kumar PS - I am not the member of this group, so please CC me, when you are replying. __________________________________________________ Do You Yahoo!? Yahoo! Games - play chess, backgammon, pool and more http://games.yahoo.com/