From: Toshiyuki Okajima Subject: [Q] ext4: the max file size of each case is correct? Date: Tue, 10 Aug 2010 23:23:21 +0900 Message-ID: <20100810232321.0eb72247.toshi.okajima@jp.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: linux-ext4@vger.kernel.org To: tytso@mit.edu, adilger@sun.com Return-path: Received: from fgwmail7.fujitsu.co.jp ([192.51.44.37]:48668 "EHLO fgwmail7.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751537Ab0HJFl5 (ORCPT ); Tue, 10 Aug 2010 01:41:57 -0400 Received: from m6.gw.fujitsu.co.jp ([10.0.50.76]) by fgwmail7.fujitsu.co.jp (Fujitsu Gateway) with ESMTP id o7A5fskA016270 for (envelope-from toshi.okajima@jp.fujitsu.com); Tue, 10 Aug 2010 14:41:54 +0900 Received: from smail (m6 [127.0.0.1]) by outgoing.m6.gw.fujitsu.co.jp (Postfix) with ESMTP id 3AAB945DE4E for ; Tue, 10 Aug 2010 14:41:54 +0900 (JST) Received: from s6.gw.fujitsu.co.jp (s6.gw.fujitsu.co.jp [10.0.50.96]) by m6.gw.fujitsu.co.jp (Postfix) with ESMTP id 1B7B145DE4C for ; Tue, 10 Aug 2010 14:41:54 +0900 (JST) Received: from s6.gw.fujitsu.co.jp (localhost.localdomain [127.0.0.1]) by s6.gw.fujitsu.co.jp (Postfix) with ESMTP id 056271DB8013 for ; Tue, 10 Aug 2010 14:41:54 +0900 (JST) Received: from m108.s.css.fujitsu.com (m108.s.css.fujitsu.com [10.249.87.108]) by s6.gw.fujitsu.co.jp (Postfix) with ESMTP id AF76C1DB8012 for ; Tue, 10 Aug 2010 14:41:50 +0900 (JST) Sender: linux-ext4-owner@vger.kernel.org List-ID: Hi. I have tested the following cases in order to confirm the maximum file size. For the tests, I selected two parameters: (These parameters relate to the max file size.) 1) Filesystem Feature "extent" 2) File(Inode) Flag "EXT4_EXTENTS_FL" (This parameter corresponds to the case where people shifts from ext3 into ext4. (Files which are created with ext3 have no "EXT4_EXTENTS_FL" flag.)) Table. the max file size which we can write or seek at each filesystem feature tuning and file flag setting +============+===============================+===============================+ | \ File flag| | | | \ | !EXT4_EXTENTS_FL | EXT4_EXTETNS_FL | |Fs Features\| | | +------------+-------------------------------+-------------------------------+ | !extent | write: 2194719883264 | write: -------------- | | | seek: 2199023251456 | seek: -------------- | +------------+-------------------------------+-------------------------------+ | extent | write: 4402345721856 | write: 17592186044415 | | | seek: 17592186044415 | seek: 17592186044415 | +------------+-------------------------------+-------------------------------+ ( The symbols (!extent, extent) mean: !extent: The filesystem feature "extent" is not set. ex. mkfs.ext3 ; mount -t ext4 extent: The filesystem feature "extent" is set. ex. mkfs.ext3 ; tune2fs -Oextent,huge_file ; mount -t ext4 The symbols ("!EXT4_EXTENTS_FL","EXT4_EXTENTS_FL") mean: !EXT4_EXTENS_FL: The file flag, "EXT4_EXTENTS_FL" is not set. EXT4_EXTENS_FL: The file flag, "EXT4_EXTENTS_FL" is set. ) According to the table, if EXT4_EXTETNS_FL flag is not set, the max file size of write() is different from the one of seek(). These differences don't include in the ext4-specification, do they? I think the differences are not the ext4-specification. So, I made a fix patch of the differences. ================================================================================ Subject: [PATCH] ext4: create own llseek function to handle the max file size correctly. From: Toshiyuki Okajima If the file has no EXT4_EXTENTS_FL flag, the maximum size which can be written (write systemcall) is different from the maximum size which can be seeked (lseek systemcall). For example, the following 2 cases show us the differences: #1: mkfs.ext3 ; mount -t ext4 #2: mkfs.ext3 ; tune2fs -Oextent,huge_file ; mount -t ext4 Table. the max file size which we can write or seek at each filesystem feature tuning and file flag setting +============+===============================+===============================+ | \ File flag| | | | \ | !EXT4_EXTENTS_FL | EXT4_EXTETNS_FL | |case \| | | +------------+-------------------------------+-------------------------------+ | #1 | write: 2194719883264 | write: -------------- | | | seek: 2199023251456 | seek: -------------- | +------------+-------------------------------+-------------------------------+ | #2 | write: 4402345721856 | write: 17592186044415 | | | seek: 17592186044415 | seek: 17592186044415 | +------------+-------------------------------+-------------------------------+ The differences exist because ext4 has 2 max-file-size (sb->s_maxbytes, EXT4_SB(sb)->s_bitmap_maxbytes) although generic_file_llseek uses only sb->s_maxbytes. (llseek of ext4_file_operations is generic_file_llseek.) Therefore we create own llseek function which uses 2 max-file-size. The new own function originates from generic_file_llseek_nolocked(). If the file flag, "EXT4_EXTENTS_FL" is not set, the function alters inode->i_sb->s_maxbytes into EXT4_SB(inode->i_sb)->s_bitmap_maxbytes. Signed-off-by: Toshiyuki Okajima --- fs/ext4/dir.c | 2 +- fs/ext4/ext4.h | 1 + fs/ext4/file.c | 39 ++++++++++++++++++++++++++++++++++++++- 3 files changed, 40 insertions(+), 2 deletions(-) diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index ea5e6cb..62c9bba 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -39,7 +39,7 @@ static int ext4_release_dir(struct inode *inode, struct file *filp); const struct file_operations ext4_dir_operations = { - .llseek = generic_file_llseek, + .llseek = ext4_llseek, .read = generic_read_dir, .readdir = ext4_readdir, /* we take BKL. needed?*/ .unlocked_ioctl = ext4_ioctl, diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 19a4de5..e7050cd 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -1870,6 +1870,7 @@ extern const struct file_operations ext4_dir_operations; /* file.c */ extern const struct inode_operations ext4_file_inode_operations; extern const struct file_operations ext4_file_operations; +extern loff_t ext4_llseek(struct file *file, loff_t offset, int origin); /* namei.c */ extern const struct inode_operations ext4_dir_inode_operations; diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 5313ae4..f2e2d57 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -129,8 +129,45 @@ static int ext4_file_open(struct inode * inode, struct file * filp) return dquot_file_open(inode, filp); } +loff_t ext4_llseek(struct file *file, loff_t offset, int origin) +{ + struct inode *inode = file->f_mapping->host; + loff_t maxbytes; + + mutex_lock(&inode->i_mutex); + if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) + maxbytes = EXT4_SB(inode->i_sb)->s_bitmap_maxbytes; + else + maxbytes = inode->i_sb->s_maxbytes; + switch (origin) { + case SEEK_END: + offset += inode->i_size; + break; + case SEEK_CUR: + if (offset == 0) { + mutex_unlock(&inode->i_mutex); + return file->f_pos; + } + offset += file->f_pos; + break; + } + + if (offset < 0 || offset > maxbytes) { + mutex_unlock(&inode->i_mutex); + return -EINVAL; + } + + if (offset != file->f_pos) { + file->f_pos = offset; + file->f_version = 0; + } + mutex_unlock(&inode->i_mutex); + + return offset; +} + const struct file_operations ext4_file_operations = { - .llseek = generic_file_llseek, + .llseek = ext4_llseek, .read = do_sync_read, .write = do_sync_write, .aio_read = generic_file_aio_read, -- 1.5.5.6