[PATCH] splice: add direct fd <-> fd splicing support

It's more efficient for sendfile() emulation. Basically we cache an
internal private pipe and just use that as the intermediate area for
pages. Direct splicing is not available from sys_splice(), it is only
meant to be used for sendfile() emulation.

Additional patch from Ingo Molnar to avoid the PIPE_BUFFERS loop at
exit for the normal fast path.

Signed-off-by: Jens Axboe <axboe@suse.de>

6 files changed, 150 insertions, 21 deletions

diff --git a/fs/pipe.c b/fs/pipe.c
index 705b48692627..036536f072c9 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -691,12 +691,10 @@ struct pipe_inode_info * alloc_pipe_info(struct inode *inode)
         return info;
 }
 
-void free_pipe_info(struct inode *inode)
+void __free_pipe_info(struct pipe_inode_info *info)
 {
         int i;
-        struct pipe_inode_info *info = inode->i_pipe;
 
-        inode->i_pipe = NULL;
         for (i = 0; i < PIPE_BUFFERS; i++) {
                 struct pipe_buffer *buf = info->bufs + i;
                 if (buf->ops)
@@ -707,6 +705,12 @@ void free_pipe_info(struct inode *inode)
         kfree(info);
 }
 
+void free_pipe_info(struct inode *inode)
+{
+        __free_pipe_info(inode->i_pipe);
+        inode->i_pipe = NULL;
+}
+
 static struct vfsmount *pipe_mnt __read_mostly;
 static int pipefs_delete_dentry(struct dentry *dentry)
 {
diff --git a/fs/splice.c b/fs/splice.c
index a5326127aad5..c47b561edac0 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -680,8 +680,7 @@ 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)
+                           size_t len, unsigned int flags)
 {
         loff_t pos;
         int ret;
@@ -692,9 +691,6 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out,
         if (!(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);
@@ -707,9 +703,8 @@ static long do_splice_from(struct pipe_inode_info *pipe, struct file *out,
 /*
  * Attempt to initiate a splice from a file to a pipe.
  */
-static long do_splice_to(struct file *in, loff_t __user *off_in,
-                         struct pipe_inode_info *pipe, size_t len,
-                         unsigned int flags)
+static long do_splice_to(struct file *in, struct pipe_inode_info *pipe,
+                         size_t len, unsigned int flags)
 {
         loff_t pos, isize, left;
         int ret;
@@ -720,9 +715,6 @@ static long do_splice_to(struct file *in, loff_t __user *off_in,
         if (!(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);
@@ -740,6 +732,118 @@ static long do_splice_to(struct file *in, loff_t __user *off_in,
         return in->f_op->splice_read(in, pipe, len, flags);
 }
 
+long do_splice_direct(struct file *in, struct file *out, size_t len,
+                      unsigned int flags)
+{
+        struct pipe_inode_info *pipe;
+        long ret, bytes;
+        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 (!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;
+
+        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, 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, 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.
  */
@@ -749,25 +853,33 @@ static long do_splice(struct file *in, loff_t __user *off_in,
 {
         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;
-
         pipe = in->f_dentry->d_inode->i_pipe;
         if (pipe) {
                 if (off_in)
                         return -ESPIPE;
+                if (off_out) {
+                        if (out->f_op->llseek == no_llseek)
+                                return -EINVAL;
+                        if (copy_from_user(&out->f_pos, off_out,
+                                           sizeof(loff_t)))
+                                return -EFAULT;
+                }
 
-                return do_splice_from(pipe, out, off_out, len, flags);
+                return do_splice_from(pipe, out, len, flags);
         }
 
         pipe = out->f_dentry->d_inode->i_pipe;
         if (pipe) {
                 if (off_out)
                         return -ESPIPE;
+                if (off_in) {
+                        if (in->f_op->llseek == no_llseek)
+                                return -EINVAL;
+                        if (copy_from_user(&in->f_pos, off_in, sizeof(loff_t)))
+                                return -EFAULT;
+                }
 
-                return do_splice_to(in, off_in, pipe, len, flags);
+                return do_splice_to(in, pipe, len, flags);
         }
 
         return -EINVAL;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 7e6454454fbd..9e8e2ee353b4 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -1613,6 +1613,8 @@ extern void do_generic_mapping_read(struct address_space *mapping,
                                     loff_t *, read_descriptor_t *, read_actor_t);
 extern ssize_t generic_file_splice_read(struct file *, struct pipe_inode_info *, size_t, unsigned int);
 extern ssize_t generic_file_splice_write(struct pipe_inode_info *, struct file *, size_t, unsigned int);
+extern long do_splice_direct(struct file *in, struct file *out,
+                             size_t len, unsigned int flags);
 extern void
 file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping);
 extern ssize_t generic_file_readv(struct file *filp, const struct iovec *iov, 
diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h
index 9cf99cb34c15..660e9d866e5d 100644
--- a/include/linux/pipe_fs_i.h
+++ b/include/linux/pipe_fs_i.h
@@ -58,6 +58,7 @@ void pipe_wait(struct pipe_inode_info *pipe);
 
 struct pipe_inode_info * alloc_pipe_info(struct inode * inode);
 void free_pipe_info(struct inode * inode);
+void __free_pipe_info(struct pipe_inode_info *);
 
 /*
  * splice is tied to pipes as a transport (at least for now), so we'll just
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 541f4828f5e7..e194ec75833d 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -684,6 +684,7 @@ static inline void prefetch_stack(struct task_struct *t) { }
 
 struct audit_context;           /* See audit.c */
 struct mempolicy;
+struct pipe_inode_info;
 
 enum sleep_type {
         SLEEP_NORMAL,
@@ -882,6 +883,11 @@ struct task_struct {
 
         atomic_t fs_excl;       /* holding fs exclusive resources */
         struct rcu_head rcu;
+
+        /*
+         * cache last used pipe for splice
+         */
+        struct pipe_inode_info *splice_pipe;
 };
 
 static inline pid_t process_group(struct task_struct *tsk)
diff --git a/kernel/exit.c b/kernel/exit.c
index 6c2eeb8f6390..1a9787ac6173 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -34,6 +34,7 @@
 #include <linux/mutex.h>
 #include <linux/futex.h>
 #include <linux/compat.h>
+#include <linux/pipe_fs_i.h>
 
 #include <asm/uaccess.h>
 #include <asm/unistd.h>
@@ -941,6 +942,9 @@ fastcall NORET_TYPE void do_exit(long code)
         if (tsk->io_context)
                 exit_io_context();
 
+        if (tsk->splice_pipe)
+                __free_pipe_info(tsk->splice_pipe);
+
         /* PF_DEAD causes final put_task_struct after we schedule. */
         preempt_disable();
         BUG_ON(tsk->flags & PF_DEAD);