From: Jan Kara Subject: [PATCH] udf: Fix data corruption on file type conversion Date: Tue, 18 Feb 2014 13:20:51 +0100 Message-ID: <1392726051-22734-1-git-send-email-jack@suse.cz> Cc: Al Viro , Jan Kara To: linux-fsdevel@vger.kernel.org Return-path: Received: from cantor2.suse.de ([195.135.220.15]:33019 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755040AbaBRMU7 (ORCPT ); Tue, 18 Feb 2014 07:20:59 -0500 Sender: linux-fsdevel-owner@vger.kernel.org List-ID: UDF has two types of files - files with data stored in inode (ICB in UDF terminology) and files with data stored in external data blocks. We convert file from in-inode format to external format in udf_file_aio_write() when we find out data won't fit into inode any longer. However the following race between two O_APPEND writes can happen: CPU1 CPU2 udf_file_aio_write() udf_file_aio_write() down_write(&iinfo->i_data_sem); checks that i_size + count1 fits within inode => no need to convert up_write(&iinfo->i_data_sem); down_write(&iinfo->i_data_sem); checks that i_size + count2 fits within inode => no need to convert up_write(&iinfo->i_data_sem); generic_file_aio_write() - extends file by count1 bytes generic_file_aio_write() - extends file by count2 bytes Clearly if count1 + count2 doesn't fit into the inode, we overwrite kernel buffers beyond inode, possibly corrupting the filesystem as well. Fix the problem by acquiring i_mutex before checking whether write fits into the inode and using __generic_file_aio_write() afterwards which puts check and write into one critical section. Reported-by: Al Viro Signed-off-by: Jan Kara --- fs/udf/file.c | 13 +++++++++++-- fs/udf/inode.c | 1 + 2 files changed, 12 insertions(+), 2 deletions(-) If noone objects, I'll send this fix to Linus later this week. diff --git a/fs/udf/file.c b/fs/udf/file.c index c02a27a19c6d..942b1c2c64ce 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -144,6 +144,7 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov, size_t count = iocb->ki_nbytes; struct udf_inode_info *iinfo = UDF_I(inode); + mutex_lock(&inode->i_mutex); down_write(&iinfo->i_data_sem); if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { if (file->f_flags & O_APPEND) @@ -169,9 +170,17 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov, } else up_write(&iinfo->i_data_sem); - retval = generic_file_aio_write(iocb, iov, nr_segs, ppos); - if (retval > 0) + retval = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); + mutex_unlock(&inode->i_mutex); + + if (retval > 0) { + ssize_t err; + mark_inode_dirty(inode); + err = generic_write_sync(file, iocb->ki_pos - retval, retval); + if (err < 0) + retval = err; + } return retval; } diff --git a/fs/udf/inode.c b/fs/udf/inode.c index 062b7925bca0..982ce05c87ed 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -265,6 +265,7 @@ int udf_expand_file_adinicb(struct inode *inode) .nr_to_write = 1, }; + WARN_ON_ONCE(!mutex_is_locked(&inode->i_mutex)); if (!iinfo->i_lenAlloc) { if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_USE_SHORT_AD)) iinfo->i_alloc_type = ICBTAG_FLAG_AD_SHORT; -- 1.8.1.4