Commit 328eaaba authored by Miklos Szeredi's avatar Miklos Szeredi Committed by Jens Axboe

ocfs2: fix i_mutex locking in ocfs2_splice_to_file()

Rearrange locking of i_mutex on destination and call to
ocfs2_rw_lock() so locks are only held while buffers are copied with
the pipe_to_file() actor, and not while waiting for more data on the
pipe.
Signed-off-by: default avatarMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: default avatarJens Axboe <jens.axboe@oracle.com>
parent eb443e5a
......@@ -1912,6 +1912,22 @@ out_sems:
return written ? written : ret;
}
static int ocfs2_splice_to_file(struct pipe_inode_info *pipe,
struct file *out,
struct splice_desc *sd)
{
int ret;
ret = ocfs2_prepare_inode_for_write(out->f_path.dentry, &sd->pos,
sd->total_len, 0, NULL);
if (ret < 0) {
mlog_errno(ret);
return ret;
}
return splice_from_pipe_feed(pipe, sd, pipe_to_file);
}
static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe,
struct file *out,
loff_t *ppos,
......@@ -1919,39 +1935,77 @@ static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe,
unsigned int flags)
{
int ret;
struct inode *inode = out->f_path.dentry->d_inode;
struct address_space *mapping = out->f_mapping;
struct inode *inode = mapping->host;
struct splice_desc sd = {
.total_len = len,
.flags = flags,
.pos = *ppos,
.u.file = out,
};
mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", out, pipe,
(unsigned int)len,
out->f_path.dentry->d_name.len,
out->f_path.dentry->d_name.name);
mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
if (pipe->inode)
mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_PARENT);
ret = ocfs2_rw_lock(inode, 1);
if (ret < 0) {
mlog_errno(ret);
goto out;
}
splice_from_pipe_begin(&sd);
do {
ret = splice_from_pipe_next(pipe, &sd);
if (ret <= 0)
break;
ret = ocfs2_prepare_inode_for_write(out->f_path.dentry, ppos, len, 0,
NULL);
if (ret < 0) {
mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
ret = ocfs2_rw_lock(inode, 1);
if (ret < 0)
mlog_errno(ret);
goto out_unlock;
else {
ret = ocfs2_splice_to_file(pipe, out, &sd);
ocfs2_rw_unlock(inode, 1);
}
mutex_unlock(&inode->i_mutex);
} while (ret > 0);
splice_from_pipe_end(pipe, &sd);
if (pipe->inode)
mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_CHILD);
ret = generic_file_splice_write_nolock(pipe, out, ppos, len, flags);
if (pipe->inode)
mutex_unlock(&pipe->inode->i_mutex);
out_unlock:
if (sd.num_spliced)
ret = sd.num_spliced;
if (ret > 0) {
unsigned long nr_pages;
*ppos += ret;
nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
/*
* If file or inode is SYNC and we actually wrote some data,
* sync it.
*/
if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) {
int err;
mutex_lock(&inode->i_mutex);
err = ocfs2_rw_lock(inode, 1);
if (err < 0) {
mlog_errno(err);
} else {
err = generic_osync_inode(inode, mapping,
OSYNC_METADATA|OSYNC_DATA);
ocfs2_rw_unlock(inode, 1);
out:
}
mutex_unlock(&inode->i_mutex);
if (err)
ret = err;
}
balance_dirty_pages_ratelimited_nr(mapping, nr_pages);
}
mlog_exit(ret);
return ret;
}
......
......@@ -555,7 +555,7 @@ static int pipe_to_sendpage(struct pipe_inode_info *pipe,
* SPLICE_F_MOVE isn't set, or we cannot move the page, we simply create
* a new page in the output file page cache and fill/dirty that.
*/
static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
struct splice_desc *sd)
{
struct file *file = sd->u.file;
......@@ -600,6 +600,7 @@ static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
out:
return ret;
}
EXPORT_SYMBOL(pipe_to_file);
static void wakeup_pipe_writers(struct pipe_inode_info *pipe)
{
......
......@@ -75,6 +75,8 @@ extern int splice_from_pipe_next(struct pipe_inode_info *,
extern void splice_from_pipe_begin(struct splice_desc *);
extern void splice_from_pipe_end(struct pipe_inode_info *,
struct splice_desc *);
extern int pipe_to_file(struct pipe_inode_info *, struct pipe_buffer *,
struct splice_desc *);
extern ssize_t splice_to_pipe(struct pipe_inode_info *,
struct splice_pipe_desc *);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment