From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263464AbTHWBdH (ORCPT ); Fri, 22 Aug 2003 21:33:07 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264423AbTHWBdG (ORCPT ); Fri, 22 Aug 2003 21:33:06 -0400 Received: from remt26.cluster1.charter.net ([209.225.8.36]:28665 "EHLO remt26.cluster1.charter.net") by vger.kernel.org with ESMTP id S264353AbTHWBc4 (ORCPT ); Fri, 22 Aug 2003 21:32:56 -0400 Subject: cramfs crc? From: Josh Boyer To: linux-kernel@vger.kernel.org Content-Type: text/plain Organization: Message-Id: <1061602374.2003.15.camel@yoda> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.2 (1.2.2-5) Date: 22 Aug 2003 20:32:54 -0500 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org the cramfs_info structure contains a u32 crc field, but it is never used in the kernel. the mkcramfs tool calculates a crc on the whole fs image and then embeddeds this crc into the image file. why is this never checked in the kernel? it could be recalculated at mount time and compared the the embedded crc. if a mismatch is detected, the mount could simply fail. is there a reason this isn't done, or is it that nobody thought it would be useful? thx, jb