From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 26 Mar 2003 00:27:57 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 26 Mar 2003 00:27:57 -0500 Received: from rwcrmhc53.attbi.com ([204.127.198.39]:29098 "EHLO rwcrmhc53.attbi.com") by vger.kernel.org with ESMTP id ; Wed, 26 Mar 2003 00:27:55 -0500 Message-ID: <3E813E66.1060802@on-demand-tech.com> Date: Tue, 25 Mar 2003 23:45:10 -0600 From: Adam Kelly User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20020826 X-Accept-Language: en-us, en MIME-Version: 1.0 To: viro@math.psu.edu, linux-kernel@vger.kernel.org, torvalds@transmeta.com Subject: [PATCH] linux-2.5.66/fs/cramfs/inode.c typecheck - int vs. struct Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Very minor change. GCC now does more type checking, so This just zeros a timespec struct so that i_mtime, etc will match types. gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7) compiling linux-2.5.66 -Adam --- inode.c.orig 2003-03-25 22:50:17.000000000 -0600 +++ inode.c 2003-03-25 23:00:50.000000000 -0600 @@ -44,6 +44,8 @@ { struct inode * inode = new_inode(sb); + const struct timespec zero_timespec = {0}; + if (inode) { inode->i_mode = cramfs_inode->mode; inode->i_uid = cramfs_inode->uid; @@ -51,7 +53,7 @@ inode->i_blocks = (cramfs_inode->size - 1) / 512 + 1; inode->i_blksize = PAGE_CACHE_SIZE; inode->i_gid = cramfs_inode->gid; - inode->i_mtime = inode->i_atime = inode->i_ctime = 0; + inode->i_mtime = inode->i_atime = inode->i_ctime = zero_timespec; inode->i_ino = CRAMINO(cramfs_inode); /* inode->i_nlink is left 1 - arguably wrong for directories, but it's the best we can do without reading the directory