* that transfers data buffers to or from a pipe buffer.
*
* Named by Larry McVoy, original implementation from Linus, extended by
- * Jens to support splicing to files and fixing the initial implementation
- * bugs.
+ * Jens to support splicing to files, network, direct splicing, etc and
+ * fixing lots of bugs.
*
- * Copyright (C) 2005 Jens Axboe <axboe@suse.de>
- * Copyright (C) 2005 Linus Torvalds <torvalds@osdl.org>
+ * Copyright (C) 2005-2006 Jens Axboe <axboe@suse.de>
+ * Copyright (C) 2005-2006 Linus Torvalds <torvalds@osdl.org>
+ * Copyright (C) 2006 Ingo Molnar <mingo@elte.hu>
*
*/
#include <linux/fs.h>
/*
* Page got truncated/unhashed. This will cause a 0-byte
- * splice, if this is the first page
+ * splice, if this is the first page.
*/
if (!page->mapping) {
err = -ENODATA;
}
/*
- * uh oh, read-error from disk
+ * Uh oh, read-error from disk.
*/
if (!PageUptodate(page)) {
err = -EIO;
}
/*
- * page is ok afterall, fall through to mapping
+ * Page is ok afterall, fall through to mapping.
*/
unlock_page(page);
}
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,
};
/*
mutex_lock(&pipe->inode->i_mutex);
for (;;) {
- int bufs;
-
if (!pipe->readers) {
send_sig(SIGPIPE, current, 0);
if (!ret)
break;
}
- bufs = pipe->nrbufs;
- if (bufs < PIPE_BUFFERS) {
- int newbuf = (pipe->curbuf + bufs) & (PIPE_BUFFERS - 1);
+ if (pipe->nrbufs < PIPE_BUFFERS) {
+ int newbuf = (pipe->curbuf + pipe->nrbufs) & (PIPE_BUFFERS - 1);
struct pipe_buffer *buf = pipe->bufs + newbuf;
struct page *page = pages[i++];
unsigned long this_len;
buf->offset = offset;
buf->len = this_len;
buf->ops = &page_cache_pipe_buf_ops;
- pipe->nrbufs = ++bufs;
- do_wakeup = 1;
+ pipe->nrbufs++;
+ if (pipe->inode)
+ do_wakeup = 1;
ret += this_len;
len -= this_len;
break;
if (!len)
break;
- if (bufs < PIPE_BUFFERS)
+ if (pipe->nrbufs < PIPE_BUFFERS)
continue;
break;
}
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;
struct page *pages[PIPE_BUFFERS];
struct page *page;
pgoff_t index;
- int i;
+ int i, error;
- index = in->f_pos >> PAGE_CACHE_SHIFT;
- offset = in->f_pos & ~PAGE_CACHE_MASK;
+ index = *ppos >> PAGE_CACHE_SHIFT;
+ offset = *ppos & ~PAGE_CACHE_MASK;
nr_pages = (len + offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
if (nr_pages > PIPE_BUFFERS)
nr_pages = PIPE_BUFFERS;
/*
- * initiate read-ahead on this page range. however, don't call into
+ * Initiate read-ahead on this page range. however, don't call into
* read-ahead if this is a non-zero offset (we are likely doing small
* chunk splice and the page is already there) for a single page.
*/
do_page_cache_readahead(mapping, in, index, nr_pages);
/*
- * now fill in the holes
+ * Now fill in the holes:
*/
+ error = 0;
for (i = 0; i < nr_pages; i++, index++) {
+find_page:
/*
- * no page there, look one up / create it
+ * lookup the page for this index
*/
- page = find_or_create_page(mapping, index,
- mapping_gfp_mask(mapping));
- if (!page)
- break;
+ 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;
- if (PageUptodate(page))
- unlock_page(page);
- else {
- int error = mapping->a_ops->readpage(in, page);
+ /*
+ * page didn't exist, allocate one
+ */
+ page = page_cache_alloc_cold(mapping);
+ if (!page)
+ break;
+ error = add_to_page_cache_lru(page, mapping, index,
+ mapping_gfp_mask(mapping));
if (unlikely(error)) {
page_cache_release(page);
break;
}
+
+ goto readpage;
}
+
+ /*
+ * If the page isn't uptodate, we may need to start io on it
+ */
+ if (!PageUptodate(page)) {
+ lock_page(page);
+
+ /*
+ * page was truncated, stop here. if this isn't the
+ * first page, we'll just complete what we already
+ * added
+ */
+ if (!page->mapping) {
+ unlock_page(page);
+ page_cache_release(page);
+ break;
+ }
+ /*
+ * page was already under io and is now done, great
+ */
+ if (PageUptodate(page)) {
+ unlock_page(page);
+ goto fill_it;
+ }
+
+readpage:
+ /*
+ * need to read in the page
+ */
+ error = mapping->a_ops->readpage(in, page);
+
+ if (unlikely(error)) {
+ page_cache_release(page);
+ if (error == AOP_TRUNCATED_PAGE)
+ goto find_page;
+ break;
+ }
+ }
+fill_it:
pages[i] = page;
}
if (i)
return move_to_pipe(pipe, pages, i, offset, len, flags);
- return 0;
+ return error;
}
/**
* @flags: splice modifier flags
*
* 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)
break;
- in->f_pos += ret;
+ *ppos += ret;
len -= ret;
spliced += ret;
int more;
/*
- * sub-optimal, but we are limited by the pipe ->map. we don't
+ * Sub-optimal, but we are limited by the pipe ->map. We don't
* need a kmap'ed buffer here, we just want to make sure we
* have the page pinned if the pipe page originates from the
- * page cache
+ * page cache.
*/
ptr = buf->ops->map(file, info, buf);
if (IS_ERR(ptr))
offset = sd->pos & ~PAGE_CACHE_MASK;
/*
- * reuse buf page, if SPLICE_F_MOVE is set
+ * Reuse buf page, if SPLICE_F_MOVE is set.
*/
if (sd->flags & SPLICE_F_MOVE) {
/*
if (!PageUptodate(page)) {
/*
- * page got invalidated, repeat
+ * Page got invalidated, repeat.
*/
if (!page->mapping) {
unlock_page(page);
* 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);
for (;;) {
- int bufs = pipe->nrbufs;
-
- if (bufs) {
- int curbuf = pipe->curbuf;
- struct pipe_buffer *buf = pipe->bufs + curbuf;
+ if (pipe->nrbufs) {
+ struct pipe_buffer *buf = pipe->bufs + pipe->curbuf;
struct pipe_buf_operations *ops = buf->ops;
sd.len = buf->len;
ret += sd.len;
buf->offset += sd.len;
buf->len -= sd.len;
+
if (!buf->len) {
buf->ops = NULL;
ops->release(pipe, buf);
- curbuf = (curbuf + 1) & (PIPE_BUFFERS - 1);
- pipe->curbuf = curbuf;
- pipe->nrbufs = --bufs;
- do_wakeup = 1;
+ pipe->curbuf = (pipe->curbuf + 1) & (PIPE_BUFFERS - 1);
+ pipe->nrbufs--;
+ if (pipe->inode)
+ do_wakeup = 1;
}
sd.pos += sd.len;
break;
}
- if (bufs)
+ if (pipe->nrbufs)
continue;
if (!pipe->writers)
break;
kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
}
- mutex_lock(&out->f_mapping->host->i_mutex);
- out->f_pos = sd.pos;
- mutex_unlock(&out->f_mapping->host->i_mutex);
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);
+ ret = move_from_pipe(pipe, out, ppos, len, flags, pipe_to_file);
/*
- * if file or inode is SYNC and we actually wrote some data, sync it
+ * 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) {
mutex_lock(&inode->i_mutex);
err = generic_osync_inode(mapping->host, mapping,
- OSYNC_METADATA|OSYNC_DATA);
+ OSYNC_METADATA|OSYNC_DATA);
mutex_unlock(&inode->i_mutex);
if (err)
*
*/
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,
- loff_t __user *off_out, size_t len,
- unsigned int flags)
+ loff_t *ppos, size_t len, unsigned int flags)
{
- loff_t pos;
int ret;
- if (!out->f_op || !out->f_op->splice_write)
+ if (unlikely(!out->f_op || !out->f_op->splice_write))
return -EINVAL;
- if (!(out->f_mode & FMODE_WRITE))
+ if (unlikely(!(out->f_mode & FMODE_WRITE)))
return -EBADF;
- if (off_out && copy_from_user(&out->f_pos, off_out, sizeof(loff_t)))
- return -EFAULT;
-
- 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, loff_t __user *off_in,
+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 (!in->f_op || !in->f_op->splice_read)
+ if (unlikely(!in->f_op || !in->f_op->splice_read))
return -EINVAL;
- if (!(in->f_mode & FMODE_READ))
+ if (unlikely(!(in->f_mode & FMODE_READ)))
return -EBADF;
- if (off_in && copy_from_user(&in->f_pos, off_in, sizeof(loff_t)))
- return -EFAULT;
-
- 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;
- if (left < len)
+ 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, 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;
+
+ /*
+ * We require the input being a regular file, as we don't want to
+ * randomly drop data for eg socket -> socket splicing. Use the
+ * piped splicing for that!
+ */
+ i_mode = in->f_dentry->d_inode->i_mode;
+ if (unlikely(!S_ISREG(i_mode) && !S_ISBLK(i_mode)))
+ return -EINVAL;
+
+ /*
+ * neither in nor out is a pipe, setup an internal pipe attached to
+ * 'out' and transfer the wanted data from 'in' to 'out' through that
+ */
+ pipe = current->splice_pipe;
+ if (unlikely(!pipe)) {
+ pipe = alloc_pipe_info(NULL);
+ if (!pipe)
+ return -ENOMEM;
+
+ /*
+ * We don't have an immediate reader, but we'll read the stuff
+ * out of the pipe right after the move_to_pipe(). So set
+ * PIPE_READERS appropriately.
+ */
+ pipe->readers = 1;
+
+ current->splice_pipe = pipe;
+ }
+
+ /*
+ * Do the splice.
+ */
+ ret = 0;
+ bytes = 0;
+ out_off = 0;
+
+ while (len) {
+ size_t read_len, max_read_len;
+
+ /*
+ * Do at most PIPE_BUFFERS pages worth of transfer:
+ */
+ max_read_len = min(len, (size_t)(PIPE_BUFFERS*PAGE_SIZE));
+
+ ret = do_splice_to(in, ppos, pipe, max_read_len, flags);
+ if (unlikely(ret < 0))
+ goto out_release;
+
+ read_len = ret;
+
+ /*
+ * NOTE: nonblocking mode only applies to the input. We
+ * 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, &out_off, read_len,
+ flags & ~SPLICE_F_NONBLOCK);
+ if (unlikely(ret < 0))
+ goto out_release;
+
+ bytes += ret;
+ len -= ret;
+
+ /*
+ * In nonblocking mode, if we got back a short read then
+ * that was due to either an IO error or due to the
+ * pagecache entry not being there. In the IO error case
+ * the _next_ splice attempt will produce a clean IO error
+ * return value (not a short read), so in both cases it's
+ * correct to break out of the loop here:
+ */
+ if ((flags & SPLICE_F_NONBLOCK) && (read_len < max_read_len))
+ break;
+ }
+
+ pipe->nrbufs = pipe->curbuf = 0;
+
+ return bytes;
+
+out_release:
+ /*
+ * If we did an incomplete transfer we must release
+ * the pipe buffers in question:
+ */
+ for (i = 0; i < PIPE_BUFFERS; i++) {
+ struct pipe_buffer *buf = pipe->bufs + i;
+
+ if (buf->ops) {
+ buf->ops->release(pipe, buf);
+ buf->ops = NULL;
+ }
+ }
+ pipe->nrbufs = pipe->curbuf = 0;
+
+ /*
+ * If we transferred some data, return the number of bytes:
+ */
+ if (bytes > 0)
+ return bytes;
+
+ return ret;
}
+EXPORT_SYMBOL(do_splice_direct);
+
/*
* Determine where to splice to/from.
*/
size_t len, unsigned int flags)
{
struct pipe_inode_info *pipe;
-
- if (off_out && out->f_op->llseek == no_llseek)
- return -EINVAL;
- if (off_in && in->f_op->llseek == no_llseek)
- return -EINVAL;
+ loff_t offset, *off;
pipe = in->f_dentry->d_inode->i_pipe;
if (pipe) {
if (off_in)
return -ESPIPE;
-
- return do_splice_from(pipe, out, off_out, len, flags);
+ if (off_out) {
+ if (out->f_op->llseek == no_llseek)
+ return -EINVAL;
+ if (copy_from_user(&offset, off_out, sizeof(loff_t)))
+ return -EFAULT;
+ off = &offset;
+ } else
+ off = &out->f_pos;
+
+ return do_splice_from(pipe, out, off, len, flags);
}
pipe = out->f_dentry->d_inode->i_pipe;
if (pipe) {
if (off_out)
return -ESPIPE;
-
- return do_splice_to(in, off_in, pipe, len, flags);
+ if (off_in) {
+ if (in->f_op->llseek == no_llseek)
+ return -EINVAL;
+ if (copy_from_user(&offset, off_in, sizeof(loff_t)))
+ return -EFAULT;
+ off = &offset;
+ } else
+ off = &in->f_pos;
+
+ return do_splice_to(in, off, pipe, len, flags);
}
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 = 0, do_wakeup = 0, i;
+
+ /*
+ * 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) {
+ 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 (flags & SPLICE_F_NONBLOCK) {
+ 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;
+ }
+ if (flags & SPLICE_F_NONBLOCK) {
+ 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;
+}