struct page *page = buf->page;
struct address_space *mapping = page_mapping(page);
- WARN_ON(!PageLocked(page));
+ lock_page(page);
+
WARN_ON(!PageUptodate(page));
/*
if (PagePrivate(page))
try_to_release_page(page, mapping_gfp_mask(mapping));
- if (!remove_mapping(mapping, page))
+ if (!remove_mapping(mapping, page)) {
+ unlock_page(page);
return 1;
+ }
buf->flags |= PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU;
return 0;
kunmap(buf->page);
}
+static void page_cache_pipe_buf_get(struct pipe_inode_info *info,
+ struct pipe_buffer *buf)
+{
+ page_cache_get(buf->page);
+}
+
static struct pipe_buf_operations page_cache_pipe_buf_ops = {
.can_merge = 0,
.map = page_cache_pipe_buf_map,
.unmap = page_cache_pipe_buf_unmap,
.release = page_cache_pipe_buf_release,
.steal = page_cache_pipe_buf_steal,
+ .get = page_cache_pipe_buf_get,
};
/*
* pipe buffer operations. Otherwise very similar to the regular pipe_writev().
*/
static ssize_t move_to_pipe(struct pipe_inode_info *pipe, struct page **pages,
- int nr_pages, unsigned long offset,
- unsigned long len, unsigned int flags)
+ int nr_pages, unsigned long len,
+ unsigned int offset, unsigned int flags)
{
int ret, do_wakeup, i;
}
static int
-__generic_file_splice_read(struct file *in, struct pipe_inode_info *pipe,
- size_t len, unsigned int flags)
+__generic_file_splice_read(struct file *in, loff_t *ppos,
+ struct pipe_inode_info *pipe, size_t len,
+ unsigned int flags)
{
struct address_space *mapping = in->f_mapping;
- unsigned int offset, nr_pages;
+ unsigned int loff, offset, nr_pages;
struct page *pages[PIPE_BUFFERS];
struct page *page;
- pgoff_t index;
+ pgoff_t index, end_index;
+ loff_t isize;
+ size_t bytes;
int i, error;
- index = in->f_pos >> PAGE_CACHE_SHIFT;
- offset = in->f_pos & ~PAGE_CACHE_MASK;
+ index = *ppos >> PAGE_CACHE_SHIFT;
+ loff = offset = *ppos & ~PAGE_CACHE_MASK;
nr_pages = (len + offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
if (nr_pages > PIPE_BUFFERS)
* Now fill in the holes:
*/
error = 0;
+ bytes = 0;
for (i = 0; i < nr_pages; i++, index++) {
+ unsigned int this_len;
+
+ if (!len)
+ break;
+
+ /*
+ * this_len is the max we'll use from this page
+ */
+ this_len = min(len, PAGE_CACHE_SIZE - loff);
find_page:
/*
* lookup the page for this index
*/
page = find_get_page(mapping, index);
if (!page) {
- /*
- * If in nonblock mode then dont block on
- * readpage (we've kicked readahead so there
- * will be asynchronous progress):
- */
- if (flags & SPLICE_F_NONBLOCK)
- break;
-
/*
* page didn't exist, allocate one
*/
* If the page isn't uptodate, we may need to start io on it
*/
if (!PageUptodate(page)) {
+ /*
+ * If in nonblock mode then dont block on waiting
+ * for an in-flight io page
+ */
+ if (flags & SPLICE_F_NONBLOCK)
+ break;
+
lock_page(page);
/*
goto find_page;
break;
}
+
+ /*
+ * i_size must be checked after ->readpage().
+ */
+ isize = i_size_read(mapping->host);
+ end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
+ if (unlikely(!isize || index > end_index)) {
+ page_cache_release(page);
+ break;
+ }
+
+ /*
+ * if this is the last page, see if we need to shrink
+ * the length and stop
+ */
+ if (end_index == index) {
+ loff = PAGE_CACHE_SIZE - (isize & ~PAGE_CACHE_MASK);
+ if (bytes + loff > isize) {
+ page_cache_release(page);
+ break;
+ }
+ /*
+ * force quit after adding this page
+ */
+ nr_pages = i;
+ this_len = min(this_len, loff);
+ }
}
fill_it:
pages[i] = page;
+ bytes += this_len;
+ len -= this_len;
+ loff = 0;
}
if (i)
- return move_to_pipe(pipe, pages, i, offset, len, flags);
+ return move_to_pipe(pipe, pages, i, bytes, offset, flags);
return error;
}
*
* Will read pages from given file and fill them into a pipe.
*/
-ssize_t generic_file_splice_read(struct file *in, struct pipe_inode_info *pipe,
- size_t len, unsigned int flags)
+ssize_t generic_file_splice_read(struct file *in, loff_t *ppos,
+ struct pipe_inode_info *pipe, size_t len,
+ unsigned int flags)
{
ssize_t spliced;
int ret;
spliced = 0;
while (len) {
- ret = __generic_file_splice_read(in, pipe, len, flags);
+ ret = __generic_file_splice_read(in, ppos, pipe, len, flags);
- if (ret <= 0)
+ if (ret < 0)
break;
+ else if (!ret) {
+ if (spliced)
+ break;
+ if (flags & SPLICE_F_NONBLOCK) {
+ ret = -EAGAIN;
+ break;
+ }
+ }
- in->f_pos += ret;
+ *ppos += ret;
len -= ret;
spliced += ret;
-
- if (!(flags & SPLICE_F_NONBLOCK))
- continue;
- ret = -EAGAIN;
- break;
}
if (spliced)
if (sd->flags & SPLICE_F_MOVE) {
/*
* If steal succeeds, buf->page is now pruned from the vm
- * side (LRU and page cache) and we can reuse it.
+ * side (LRU and page cache) and we can reuse it. The page
+ * will also be looked on successful return.
*/
if (buf->ops->steal(info, buf))
goto find_page;
- /*
- * this will also set the page locked
- */
page = buf->page;
if (add_to_page_cache(page, mapping, index, gfp_mask))
goto find_page;
lru_cache_add(page);
} else {
find_page:
- ret = -ENOMEM;
- page = find_or_create_page(mapping, index, gfp_mask);
- if (!page)
- goto out_nomem;
+ page = find_lock_page(mapping, index);
+ if (!page) {
+ ret = -ENOMEM;
+ page = page_cache_alloc_cold(mapping);
+ if (unlikely(!page))
+ goto out_nomem;
+
+ /*
+ * This will also lock the page
+ */
+ ret = add_to_page_cache_lru(page, mapping, index,
+ gfp_mask);
+ if (unlikely(ret))
+ goto out;
+ }
/*
- * If the page is uptodate, it is also locked. If it isn't
- * uptodate, we can mark it uptodate if we are filling the
- * full page. Otherwise we need to read it in first...
+ * We get here with the page locked. If the page is also
+ * uptodate, we don't need to do more. If it isn't, we
+ * may need to bring it in if we are not going to overwrite
+ * the full page.
*/
if (!PageUptodate(page)) {
if (sd->len < PAGE_CACHE_SIZE) {
ret = -EIO;
goto out;
}
- } else {
- WARN_ON(!PageLocked(page));
+ } else
SetPageUptodate(page);
- }
}
}
mark_page_accessed(page);
balance_dirty_pages_ratelimited(mapping);
out:
- if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) {
+ if (!(buf->flags & PIPE_BUF_FLAG_STOLEN))
page_cache_release(page);
- unlock_page(page);
- }
+
+ unlock_page(page);
out_nomem:
buf->ops->unmap(info, buf);
return ret;
* to the wanted destination. See pipe_to_file/pipe_to_sendpage above.
*/
static ssize_t move_from_pipe(struct pipe_inode_info *pipe, struct file *out,
- size_t len, unsigned int flags,
+ loff_t *ppos, size_t len, unsigned int flags,
splice_actor *actor)
{
int ret, do_wakeup, err;
sd.total_len = len;
sd.flags = flags;
sd.file = out;
- sd.pos = out->f_pos;
+ sd.pos = *ppos;
if (pipe->inode)
mutex_lock(&pipe->inode->i_mutex);
kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
}
- out->f_pos = sd.pos;
return ret;
-
}
/**
*/
ssize_t
generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out,
- size_t len, unsigned int flags)
+ loff_t *ppos, size_t len, unsigned int flags)
{
struct address_space *mapping = out->f_mapping;
ssize_t ret;
- ret = move_from_pipe(pipe, out, len, flags, pipe_to_file);
-
- /*
- * If file or inode is SYNC and we actually wrote some data, sync it.
- */
- if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(mapping->host))
- && ret > 0) {
+ ret = move_from_pipe(pipe, out, ppos, len, flags, pipe_to_file);
+ if (ret > 0) {
struct inode *inode = mapping->host;
- int err;
- mutex_lock(&inode->i_mutex);
- err = generic_osync_inode(mapping->host, mapping,
- OSYNC_METADATA|OSYNC_DATA);
- mutex_unlock(&inode->i_mutex);
+ *ppos += ret;
- if (err)
- ret = err;
+ /*
+ * 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 = generic_osync_inode(inode, mapping,
+ OSYNC_METADATA|OSYNC_DATA);
+ mutex_unlock(&inode->i_mutex);
+
+ if (err)
+ ret = err;
+ }
}
return ret;
*
*/
ssize_t generic_splice_sendpage(struct pipe_inode_info *pipe, struct file *out,
- size_t len, unsigned int flags)
+ loff_t *ppos, size_t len, unsigned int flags)
{
- return move_from_pipe(pipe, out, len, flags, pipe_to_sendpage);
+ return move_from_pipe(pipe, out, ppos, len, flags, pipe_to_sendpage);
}
EXPORT_SYMBOL(generic_splice_sendpage);
* Attempt to initiate a splice from pipe to file.
*/
static long do_splice_from(struct pipe_inode_info *pipe, struct file *out,
- size_t len, unsigned int flags)
+ loff_t *ppos, size_t len, unsigned int flags)
{
- loff_t pos;
int ret;
if (unlikely(!out->f_op || !out->f_op->splice_write))
if (unlikely(!(out->f_mode & FMODE_WRITE)))
return -EBADF;
- pos = out->f_pos;
-
- ret = rw_verify_area(WRITE, out, &pos, len);
+ ret = rw_verify_area(WRITE, out, ppos, len);
if (unlikely(ret < 0))
return ret;
- return out->f_op->splice_write(pipe, out, len, flags);
+ return out->f_op->splice_write(pipe, out, ppos, len, flags);
}
/*
* Attempt to initiate a splice from a file to a pipe.
*/
-static long do_splice_to(struct file *in, struct pipe_inode_info *pipe,
- size_t len, unsigned int flags)
+static long do_splice_to(struct file *in, loff_t *ppos,
+ struct pipe_inode_info *pipe, size_t len,
+ unsigned int flags)
{
- loff_t pos, isize, left;
+ loff_t isize, left;
int ret;
if (unlikely(!in->f_op || !in->f_op->splice_read))
if (unlikely(!(in->f_mode & FMODE_READ)))
return -EBADF;
- pos = in->f_pos;
-
- ret = rw_verify_area(READ, in, &pos, len);
+ ret = rw_verify_area(READ, in, ppos, len);
if (unlikely(ret < 0))
return ret;
isize = i_size_read(in->f_mapping->host);
- if (unlikely(in->f_pos >= isize))
+ if (unlikely(*ppos >= isize))
return 0;
- left = isize - in->f_pos;
+ left = isize - *ppos;
if (unlikely(left < len))
len = left;
- return in->f_op->splice_read(in, pipe, len, flags);
+ return in->f_op->splice_read(in, ppos, pipe, len, flags);
}
-long do_splice_direct(struct file *in, struct file *out, size_t len,
- unsigned int flags)
+long do_splice_direct(struct file *in, loff_t *ppos, struct file *out,
+ size_t len, unsigned int flags)
{
struct pipe_inode_info *pipe;
long ret, bytes;
+ loff_t out_off;
umode_t i_mode;
int i;
*/
ret = 0;
bytes = 0;
+ out_off = 0;
while (len) {
size_t read_len, max_read_len;
*/
max_read_len = min(len, (size_t)(PIPE_BUFFERS*PAGE_SIZE));
- ret = do_splice_to(in, pipe, max_read_len, flags);
+ ret = do_splice_to(in, ppos, pipe, max_read_len, flags);
if (unlikely(ret < 0))
goto out_release;
* must not do the output in nonblocking mode as then we
* could get stuck data in the internal pipe:
*/
- ret = do_splice_from(pipe, out, read_len,
+ ret = do_splice_from(pipe, out, &out_off, read_len,
flags & ~SPLICE_F_NONBLOCK);
if (unlikely(ret < 0))
goto out_release;
size_t len, unsigned int flags)
{
struct pipe_inode_info *pipe;
+ loff_t offset, *off;
+ long ret;
pipe = in->f_dentry->d_inode->i_pipe;
if (pipe) {
if (off_out) {
if (out->f_op->llseek == no_llseek)
return -EINVAL;
- if (copy_from_user(&out->f_pos, off_out,
- sizeof(loff_t)))
+ if (copy_from_user(&offset, off_out, sizeof(loff_t)))
return -EFAULT;
- }
+ off = &offset;
+ } else
+ off = &out->f_pos;
+
+ ret = do_splice_from(pipe, out, off, len, flags);
- return do_splice_from(pipe, out, len, flags);
+ if (off_out && copy_to_user(off_out, off, sizeof(loff_t)))
+ ret = -EFAULT;
+
+ return ret;
}
pipe = out->f_dentry->d_inode->i_pipe;
if (off_in) {
if (in->f_op->llseek == no_llseek)
return -EINVAL;
- if (copy_from_user(&in->f_pos, off_in, sizeof(loff_t)))
+ if (copy_from_user(&offset, off_in, sizeof(loff_t)))
return -EFAULT;
- }
+ off = &offset;
+ } else
+ off = &in->f_pos;
+
+ ret = do_splice_to(in, off, pipe, len, flags);
+
+ if (off_in && copy_to_user(off_in, off, sizeof(loff_t)))
+ ret = -EFAULT;
- return do_splice_to(in, pipe, len, flags);
+ return ret;
}
return -EINVAL;
return error;
}
+
+/*
+ * Link contents of ipipe to opipe.
+ */
+static int link_pipe(struct pipe_inode_info *ipipe,
+ struct pipe_inode_info *opipe,
+ size_t len, unsigned int flags)
+{
+ struct pipe_buffer *ibuf, *obuf;
+ int ret, do_wakeup, i, ipipe_first;
+
+ ret = do_wakeup = ipipe_first = 0;
+
+ /*
+ * Potential ABBA deadlock, work around it by ordering lock
+ * grabbing by inode address. Otherwise two different processes
+ * could deadlock (one doing tee from A -> B, the other from B -> A).
+ */
+ if (ipipe->inode < opipe->inode) {
+ ipipe_first = 1;
+ mutex_lock(&ipipe->inode->i_mutex);
+ mutex_lock(&opipe->inode->i_mutex);
+ } else {
+ mutex_lock(&opipe->inode->i_mutex);
+ mutex_lock(&ipipe->inode->i_mutex);
+ }
+
+ for (i = 0;; i++) {
+ if (!opipe->readers) {
+ send_sig(SIGPIPE, current, 0);
+ if (!ret)
+ ret = -EPIPE;
+ break;
+ }
+ if (ipipe->nrbufs - i) {
+ ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (PIPE_BUFFERS - 1));
+
+ /*
+ * If we have room, fill this buffer
+ */
+ if (opipe->nrbufs < PIPE_BUFFERS) {
+ int nbuf = (opipe->curbuf + opipe->nrbufs) & (PIPE_BUFFERS - 1);
+
+ /*
+ * Get a reference to this pipe buffer,
+ * so we can copy the contents over.
+ */
+ ibuf->ops->get(ipipe, ibuf);
+
+ obuf = opipe->bufs + nbuf;
+ *obuf = *ibuf;
+
+ if (obuf->len > len)
+ obuf->len = len;
+
+ opipe->nrbufs++;
+ do_wakeup = 1;
+ ret += obuf->len;
+ len -= obuf->len;
+
+ if (!len)
+ break;
+ if (opipe->nrbufs < PIPE_BUFFERS)
+ continue;
+ }
+
+ /*
+ * We have input available, but no output room.
+ * If we already copied data, return that. If we
+ * need to drop the opipe lock, it must be ordered
+ * last to avoid deadlocks.
+ */
+ if ((flags & SPLICE_F_NONBLOCK) || !ipipe_first) {
+ if (!ret)
+ ret = -EAGAIN;
+ break;
+ }
+ if (signal_pending(current)) {
+ if (!ret)
+ ret = -ERESTARTSYS;
+ break;
+ }
+ if (do_wakeup) {
+ smp_mb();
+ if (waitqueue_active(&opipe->wait))
+ wake_up_interruptible(&opipe->wait);
+ kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN);
+ do_wakeup = 0;
+ }
+
+ opipe->waiting_writers++;
+ pipe_wait(opipe);
+ opipe->waiting_writers--;
+ continue;
+ }
+
+ /*
+ * No input buffers, do the usual checks for available
+ * writers and blocking and wait if necessary
+ */
+ if (!ipipe->writers)
+ break;
+ if (!ipipe->waiting_writers) {
+ if (ret)
+ break;
+ }
+ /*
+ * pipe_wait() drops the ipipe mutex. To avoid deadlocks
+ * with another process, we can only safely do that if
+ * the ipipe lock is ordered last.
+ */
+ if ((flags & SPLICE_F_NONBLOCK) || ipipe_first) {
+ if (!ret)
+ ret = -EAGAIN;
+ break;
+ }
+ if (signal_pending(current)) {
+ if (!ret)
+ ret = -ERESTARTSYS;
+ break;
+ }
+
+ if (waitqueue_active(&ipipe->wait))
+ wake_up_interruptible_sync(&ipipe->wait);
+ kill_fasync(&ipipe->fasync_writers, SIGIO, POLL_OUT);
+
+ pipe_wait(ipipe);
+ }
+
+ mutex_unlock(&ipipe->inode->i_mutex);
+ mutex_unlock(&opipe->inode->i_mutex);
+
+ if (do_wakeup) {
+ smp_mb();
+ if (waitqueue_active(&opipe->wait))
+ wake_up_interruptible(&opipe->wait);
+ kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN);
+ }
+
+ return ret;
+}
+
+/*
+ * This is a tee(1) implementation that works on pipes. It doesn't copy
+ * any data, it simply references the 'in' pages on the 'out' pipe.
+ * The 'flags' used are the SPLICE_F_* variants, currently the only
+ * applicable one is SPLICE_F_NONBLOCK.
+ */
+static long do_tee(struct file *in, struct file *out, size_t len,
+ unsigned int flags)
+{
+ struct pipe_inode_info *ipipe = in->f_dentry->d_inode->i_pipe;
+ struct pipe_inode_info *opipe = out->f_dentry->d_inode->i_pipe;
+
+ /*
+ * Link ipipe to the two output pipes, consuming as we go along.
+ */
+ if (ipipe && opipe)
+ return link_pipe(ipipe, opipe, len, flags);
+
+ return -EINVAL;
+}
+
+asmlinkage long sys_tee(int fdin, int fdout, size_t len, unsigned int flags)
+{
+ struct file *in;
+ int error, fput_in;
+
+ if (unlikely(!len))
+ return 0;
+
+ error = -EBADF;
+ in = fget_light(fdin, &fput_in);
+ if (in) {
+ if (in->f_mode & FMODE_READ) {
+ int fput_out;
+ struct file *out = fget_light(fdout, &fput_out);
+
+ if (out) {
+ if (out->f_mode & FMODE_WRITE)
+ error = do_tee(in, out, len, flags);
+ fput_light(out, fput_out);
+ }
+ }
+ fput_light(in, fput_in);
+ }
+
+ return error;
+}