udf: Fix data corruption on file type conversion

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 <viro@ZenIV.linux.org.uk>
Signed-off-by: Jan Kara <jack@suse.cz>

diff --git a/fs/udf/file.c b/fs/udf/file.c
index c02a27a..1037637 100644 (file)
--- 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)
@@ -156,6 +157,7 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
                                                pos + count)) {
                        err = udf_expand_file_adinicb(inode);
                        if (err) {
+                               mutex_unlock(&inode->i_mutex);
                                udf_debug("udf_expand_adinicb: err=%d\n", err);
                                return err;
                        }
@@ -169,9 +171,17 @@ 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 062b792..982ce05 100644 (file)
--- 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;