Merge branch 'master'

11 files changed, 395 insertions, 205 deletions

diff --git a/fs/block_dev.c b/fs/block_dev.c
index af88c43043d..f5958f413bd 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -1104,6 +1104,8 @@ const struct file_operations def_blk_fops = {
 	.readv		= generic_file_readv,
 	.writev		= generic_file_write_nolock,
 	.sendfile	= generic_file_sendfile,
+	.splice_read	= generic_file_splice_read,
+	.splice_write	= generic_file_splice_write,
 };
 
 int ioctl_by_bdev(struct block_device *bdev, unsigned cmd, unsigned long arg)
diff --git a/fs/compat.c b/fs/compat.c
index 2e32bd34047..970888aad84 100644
--- a/fs/compat.c
+++ b/fs/compat.c
@@ -1317,6 +1317,26 @@ out:
 	return ret;
 }
 
+asmlinkage long
+compat_sys_vmsplice(int fd, const struct compat_iovec __user *iov32,
+		    unsigned int nr_segs, unsigned int flags)
+{
+	unsigned i;
+	struct iovec *iov;
+	if (nr_segs > UIO_MAXIOV)
+		return -EINVAL;
+	iov = compat_alloc_user_space(nr_segs * sizeof(struct iovec));
+	for (i = 0; i < nr_segs; i++) {
+		struct compat_iovec v;
+		if (get_user(v.iov_base, &iov32[i].iov_base) ||
+		    get_user(v.iov_len, &iov32[i].iov_len) ||
+		    put_user(compat_ptr(v.iov_base), &iov[i].iov_base) ||
+		    put_user(v.iov_len, &iov[i].iov_len))
+			return -EFAULT;
+	}
+	return sys_vmsplice(fd, iov, nr_segs, flags);
+}
+
 /*
  * Exactly like fs/open.c:sys_open(), except that it doesn't set the
  * O_LARGEFILE flag.
diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c
index 48ae0339af1..2edd7eec88f 100644
--- a/fs/ext3/inode.c
+++ b/fs/ext3/inode.c
@@ -711,7 +711,7 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode,
 	 * direct blocks blocks
 	 */
 	if (num == 0 && blks > 1) {
-		current_block = le32_to_cpu(where->key + 1);
+		current_block = le32_to_cpu(where->key) + 1;
 		for (i = 1; i < blks; i++)
 			*(where->p + i ) = cpu_to_le32(current_block++);
 	}
@@ -724,7 +724,7 @@ static int ext3_splice_branch(handle_t *handle, struct inode *inode,
 	if (block_i) {
 		block_i->last_alloc_logical_block = block + blks - 1;
 		block_i->last_alloc_physical_block =
-				le32_to_cpu(where[num].key + blks - 1);
+				le32_to_cpu(where[num].key) + blks - 1;
 	}
 
 	/* We are done with atomic stuff, now do the rest of housekeeping */
@@ -814,11 +814,13 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode,
 
 	/* Simplest case - block found, no allocation needed */
 	if (!partial) {
-		first_block = chain[depth - 1].key;
+		first_block = le32_to_cpu(chain[depth - 1].key);
 		clear_buffer_new(bh_result);
 		count++;
 		/*map more blocks*/
 		while (count < maxblocks && count <= blocks_to_boundary) {
+			unsigned long blk;
+
 			if (!verify_chain(chain, partial)) {
 				/*
 				 * Indirect block might be removed by
@@ -831,8 +833,9 @@ int ext3_get_blocks_handle(handle_t *handle, struct inode *inode,
 				count = 0;
 				break;
 			}
-			if (le32_to_cpu(*(chain[depth-1].p+count) ==
-					(first_block + count)))
+			blk = le32_to_cpu(*(chain[depth-1].p + count));
+
+			if (blk == first_block + count)
 				count++;
 			else
 				break;
diff --git a/fs/locks.c b/fs/locks.c
index efad798824d..6f99c0a6f83 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -446,15 +446,14 @@ static struct lock_manager_operations lease_manager_ops = {
  */
 static int lease_init(struct file *filp, int type, struct file_lock *fl)
  {
+	if (assign_type(fl, type) != 0)
+		return -EINVAL;
+
 	fl->fl_owner = current->files;
 	fl->fl_pid = current->tgid;
 
 	fl->fl_file = filp;
 	fl->fl_flags = FL_LEASE;
-	if (assign_type(fl, type) != 0) {
-		locks_free_lock(fl);
-		return -EINVAL;
-	}
 	fl->fl_start = 0;
 	fl->fl_end = OFFSET_MAX;
 	fl->fl_ops = NULL;
@@ -466,16 +465,19 @@ static int lease_init(struct file *filp, int type, struct file_lock *fl)
 static int lease_alloc(struct file *filp, int type, struct file_lock **flp)
 {
 	struct file_lock *fl = locks_alloc_lock();
-	int error;
+	int error = -ENOMEM;
 
 	if (fl == NULL)
-		return -ENOMEM;
+		goto out;
 
 	error = lease_init(filp, type, fl);
-	if (error)
-		return error;
+	if (error) {
+		locks_free_lock(fl);
+		fl = NULL;
+	}
+out:
 	*flp = fl;
-	return 0;
+	return error;
 }
 
 /* Check if two locks overlap each other.
@@ -1372,6 +1374,7 @@ static int __setlease(struct file *filp, long arg, struct file_lock **flp)
 		goto out;
 
 	if (my_before != NULL) {
+		*flp = *my_before;
 		error = lease->fl_lmops->fl_change(my_before, arg);
 		goto out;
 	}
diff --git a/fs/pipe.c b/fs/pipe.c
index 7fefb10db8d..5acd8954aaa 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -55,7 +55,8 @@ void pipe_wait(struct pipe_inode_info *pipe)
 }
 
 static int
-pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len)
+pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len,
+			int atomic)
 {
 	unsigned long copy;
 
@@ -64,8 +65,13 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len)
 			iov++;
 		copy = min_t(unsigned long, len, iov->iov_len);
 
-		if (copy_from_user(to, iov->iov_base, copy))
-			return -EFAULT;
+		if (atomic) {
+			if (__copy_from_user_inatomic(to, iov->iov_base, copy))
+				return -EFAULT;
+		} else {
+			if (copy_from_user(to, iov->iov_base, copy))
+				return -EFAULT;
+		}
 		to += copy;
 		len -= copy;
 		iov->iov_base += copy;
@@ -75,7 +81,8 @@ pipe_iov_copy_from_user(void *to, struct iovec *iov, unsigned long len)
 }
 
 static int
-pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len)
+pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len,
+		      int atomic)
 {
 	unsigned long copy;
 
@@ -84,8 +91,13 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len)
 			iov++;
 		copy = min_t(unsigned long, len, iov->iov_len);
 
-		if (copy_to_user(iov->iov_base, from, copy))
-			return -EFAULT;
+		if (atomic) {
+			if (__copy_to_user_inatomic(iov->iov_base, from, copy))
+				return -EFAULT;
+		} else {
+			if (copy_to_user(iov->iov_base, from, copy))
+				return -EFAULT;
+		}
 		from += copy;
 		len -= copy;
 		iov->iov_base += copy;
@@ -94,13 +106,52 @@ pipe_iov_copy_to_user(struct iovec *iov, const void *from, unsigned long len)
 	return 0;
 }
 
+/*
+ * Attempt to pre-fault in the user memory, so we can use atomic copies.
+ * Returns the number of bytes not faulted in.
+ */
+static int iov_fault_in_pages_write(struct iovec *iov, unsigned long len)
+{
+	while (!iov->iov_len)
+		iov++;
+
+	while (len > 0) {
+		unsigned long this_len;
+
+		this_len = min_t(unsigned long, len, iov->iov_len);
+		if (fault_in_pages_writeable(iov->iov_base, this_len))
+			break;
+
+		len -= this_len;
+		iov++;
+	}
+
+	return len;
+}
+
+/*
+ * Pre-fault in the user memory, so we can use atomic copies.
+ */
+static void iov_fault_in_pages_read(struct iovec *iov, unsigned long len)
+{
+	while (!iov->iov_len)
+		iov++;
+
+	while (len > 0) {
+		unsigned long this_len;
+
+		this_len = min_t(unsigned long, len, iov->iov_len);
+		fault_in_pages_readable(iov->iov_base, this_len);
+		len -= this_len;
+		iov++;
+	}
+}
+
 static void anon_pipe_buf_release(struct pipe_inode_info *pipe,
 				  struct pipe_buffer *buf)
 {
 	struct page *page = buf->page;
 
-	buf->flags &= ~PIPE_BUF_FLAG_STOLEN;
-
 	/*
 	 * If nobody else uses this page, and we don't already have a
 	 * temporary page, let's keep track of it as a one-deep
@@ -112,38 +163,58 @@ static void anon_pipe_buf_release(struct pipe_inode_info *pipe,
 		page_cache_release(page);
 }
 
-static void * anon_pipe_buf_map(struct file *file, struct pipe_inode_info *pipe,
-				struct pipe_buffer *buf)
+void *generic_pipe_buf_map(struct pipe_inode_info *pipe,
+			   struct pipe_buffer *buf, int atomic)
 {
+	if (atomic) {
+		buf->flags |= PIPE_BUF_FLAG_ATOMIC;
+		return kmap_atomic(buf->page, KM_USER0);
+	}
+
 	return kmap(buf->page);
 }
 
-static void anon_pipe_buf_unmap(struct pipe_inode_info *pipe,
-				struct pipe_buffer *buf)
+void generic_pipe_buf_unmap(struct pipe_inode_info *pipe,
+			    struct pipe_buffer *buf, void *map_data)
 {
-	kunmap(buf->page);
+	if (buf->flags & PIPE_BUF_FLAG_ATOMIC) {
+		buf->flags &= ~PIPE_BUF_FLAG_ATOMIC;
+		kunmap_atomic(map_data, KM_USER0);
+	} else
+		kunmap(buf->page);
 }
 
-static int anon_pipe_buf_steal(struct pipe_inode_info *pipe,
-			       struct pipe_buffer *buf)
+int generic_pipe_buf_steal(struct pipe_inode_info *pipe,
+			   struct pipe_buffer *buf)
 {
-	buf->flags |= PIPE_BUF_FLAG_STOLEN;
-	return 0;
+	struct page *page = buf->page;
+
+	if (page_count(page) == 1) {
+		lock_page(page);
+		return 0;
+	}
+
+	return 1;
 }
 
-static void anon_pipe_buf_get(struct pipe_inode_info *info,
-			      struct pipe_buffer *buf)
+void generic_pipe_buf_get(struct pipe_inode_info *info, struct pipe_buffer *buf)
 {
 	page_cache_get(buf->page);
 }
 
+int generic_pipe_buf_pin(struct pipe_inode_info *info, struct pipe_buffer *buf)
+{
+	return 0;
+}
+
 static struct pipe_buf_operations anon_pipe_buf_ops = {
 	.can_merge = 1,
-	.map = anon_pipe_buf_map,
-	.unmap = anon_pipe_buf_unmap,
+	.map = generic_pipe_buf_map,
+	.unmap = generic_pipe_buf_unmap,
+	.pin = generic_pipe_buf_pin,
 	.release = anon_pipe_buf_release,
-	.steal = anon_pipe_buf_steal,
-	.get = anon_pipe_buf_get,
+	.steal = generic_pipe_buf_steal,
+	.get = generic_pipe_buf_get,
 };
 
 static ssize_t
@@ -174,22 +245,33 @@ pipe_readv(struct file *filp, const struct iovec *_iov,
 			struct pipe_buf_operations *ops = buf->ops;
 			void *addr;
 			size_t chars = buf->len;
-			int error;
+			int error, atomic;
 
 			if (chars > total_len)
 				chars = total_len;
 
-			addr = ops->map(filp, pipe, buf);
-			if (IS_ERR(addr)) {
+			error = ops->pin(pipe, buf);
+			if (error) {
 				if (!ret)
-					ret = PTR_ERR(addr);
+					error = ret;
 				break;
 			}
-			error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars);
-			ops->unmap(pipe, buf);
+
+			atomic = !iov_fault_in_pages_write(iov, chars);
+redo:
+			addr = ops->map(pipe, buf, atomic);
+			error = pipe_iov_copy_to_user(iov, addr + buf->offset, chars, atomic);
+			ops->unmap(pipe, buf, addr);
 			if (unlikely(error)) {
+				/*
+				 * Just retry with the slow path if we failed.
+				 */
+				if (atomic) {
+					atomic = 0;
+					goto redo;
+				}
 				if (!ret)
-					ret = -EFAULT;
+					ret = error;
 				break;
 			}
 			ret += chars;
@@ -293,21 +375,28 @@ pipe_writev(struct file *filp, const struct iovec *_iov,
 		int offset = buf->offset + buf->len;
 
 		if (ops->can_merge && offset + chars <= PAGE_SIZE) {
+			int error, atomic = 1;
 			void *addr;
-			int error;
 
-			addr = ops->map(filp, pipe, buf);
-			if (IS_ERR(addr)) {
-				error = PTR_ERR(addr);
+			error = ops->pin(pipe, buf);
+			if (error)
 				goto out;
-			}
+
+			iov_fault_in_pages_read(iov, chars);
+redo1:
+			addr = ops->map(pipe, buf, atomic);
 			error = pipe_iov_copy_from_user(offset + addr, iov,
-							chars);
-			ops->unmap(pipe, buf);
+							chars, atomic);
+			ops->unmap(pipe, buf, addr);
 			ret = error;
 			do_wakeup = 1;
-			if (error)
+			if (error) {
+				if (atomic) {
+					atomic = 0;
+					goto redo1;
+				}
 				goto out;
+			}
 			buf->len += chars;
 			total_len -= chars;
 			ret = chars;
@@ -330,7 +419,8 @@ pipe_writev(struct file *filp, const struct iovec *_iov,
 			int newbuf = (pipe->curbuf + bufs) & (PIPE_BUFFERS-1);
 			struct pipe_buffer *buf = pipe->bufs + newbuf;
 			struct page *page = pipe->tmp_page;
-			int error;
+			char *src;
+			int error, atomic = 1;
 
 			if (!page) {
 				page = alloc_page(GFP_HIGHUSER);
@@ -350,11 +440,27 @@ pipe_writev(struct file *filp, const struct iovec *_iov,
 			if (chars > total_len)
 				chars = total_len;
 
-			error = pipe_iov_copy_from_user(kmap(page), iov, chars);
-			kunmap(page);
+			iov_fault_in_pages_read(iov, chars);
+redo2:
+			if (atomic)
+				src = kmap_atomic(page, KM_USER0);
+			else
+				src = kmap(page);
+
+			error = pipe_iov_copy_from_user(src, iov, chars,
+							atomic);
+			if (atomic)
+				kunmap_atomic(src, KM_USER0);
+			else
+				kunmap(page);
+
 			if (unlikely(error)) {
+				if (atomic) {
+					atomic = 0;
+					goto redo2;
+				}
 				if (!ret)
-					ret = -EFAULT;
+					ret = error;
 				break;
 			}
 			ret += chars;
diff --git a/fs/splice.c b/fs/splice.c
index 447ebc0a37f..a285fd746dc 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -51,7 +51,7 @@ struct splice_pipe_desc {
  * addition of remove_mapping(). If success is returned, the caller may
  * attempt to reuse this page for another destination.
  */
-static int page_cache_pipe_buf_steal(struct pipe_inode_info *info,
+static int page_cache_pipe_buf_steal(struct pipe_inode_info *pipe,
 				     struct pipe_buffer *buf)
 {
 	struct page *page = buf->page;
@@ -78,21 +78,19 @@ static int page_cache_pipe_buf_steal(struct pipe_inode_info *info,
 		return 1;
 	}
 
-	buf->flags |= PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU;
+	buf->flags |= PIPE_BUF_FLAG_LRU;
 	return 0;
 }
 
-static void page_cache_pipe_buf_release(struct pipe_inode_info *info,
+static void page_cache_pipe_buf_release(struct pipe_inode_info *pipe,
 					struct pipe_buffer *buf)
 {
 	page_cache_release(buf->page);
-	buf->page = NULL;
-	buf->flags &= ~(PIPE_BUF_FLAG_STOLEN | PIPE_BUF_FLAG_LRU);
+	buf->flags &= ~PIPE_BUF_FLAG_LRU;
 }
 
-static void *page_cache_pipe_buf_map(struct file *file,
-				     struct pipe_inode_info *info,
-				     struct pipe_buffer *buf)
+static int page_cache_pipe_buf_pin(struct pipe_inode_info *pipe,
+				   struct pipe_buffer *buf)
 {
 	struct page *page = buf->page;
 	int err;
@@ -118,64 +116,45 @@ static void *page_cache_pipe_buf_map(struct file *file,
 		}
 
 		/*
-		 * Page is ok afterall, fall through to mapping.
+		 * Page is ok afterall, we are done.
 		 */
 		unlock_page(page);
 	}
 
-	return kmap(page);
+	return 0;
 error:
 	unlock_page(page);
-	return ERR_PTR(err);
-}
-
-static void page_cache_pipe_buf_unmap(struct pipe_inode_info *info,
-				      struct pipe_buffer *buf)
-{
-	kunmap(buf->page);
-}
-
-static void *user_page_pipe_buf_map(struct file *file,
-				    struct pipe_inode_info *pipe,
-				    struct pipe_buffer *buf)
-{
-	return kmap(buf->page);
-}
-
-static void user_page_pipe_buf_unmap(struct pipe_inode_info *pipe,
-				     struct pipe_buffer *buf)
-{
-	kunmap(buf->page);
-}
-
-static void page_cache_pipe_buf_get(struct pipe_inode_info *info,
-				    struct pipe_buffer *buf)
-{
-	page_cache_get(buf->page);
+	return err;
 }
 
 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,
+	.map = generic_pipe_buf_map,
+	.unmap = generic_pipe_buf_unmap,
+	.pin = page_cache_pipe_buf_pin,
 	.release = page_cache_pipe_buf_release,
 	.steal = page_cache_pipe_buf_steal,
-	.get = page_cache_pipe_buf_get,
+	.get = generic_pipe_buf_get,
 };
 
 static int user_page_pipe_buf_steal(struct pipe_inode_info *pipe,
 				    struct pipe_buffer *buf)
 {
-	return 1;
+	if (!(buf->flags & PIPE_BUF_FLAG_GIFT))
+		return 1;
+
+	buf->flags |= PIPE_BUF_FLAG_LRU;
+	return generic_pipe_buf_steal(pipe, buf);
 }
 
 static struct pipe_buf_operations user_page_pipe_buf_ops = {
 	.can_merge = 0,
-	.map = user_page_pipe_buf_map,
-	.unmap = user_page_pipe_buf_unmap,
+	.map = generic_pipe_buf_map,
+	.unmap = generic_pipe_buf_unmap,
+	.pin = generic_pipe_buf_pin,
 	.release = page_cache_pipe_buf_release,
 	.steal = user_page_pipe_buf_steal,
-	.get = page_cache_pipe_buf_get,
+	.get = generic_pipe_buf_get,
 };
 
 /*
@@ -210,6 +189,9 @@ static ssize_t splice_to_pipe(struct pipe_inode_info *pipe,
 			buf->offset = spd->partial[page_nr].offset;
 			buf->len = spd->partial[page_nr].len;
 			buf->ops = spd->ops;
+			if (spd->flags & SPLICE_F_GIFT)
+				buf->flags |= PIPE_BUF_FLAG_GIFT;
+
 			pipe->nrbufs++;
 			page_nr++;
 			ret += buf->len;
@@ -279,7 +261,7 @@ __generic_file_splice_read(struct file *in, loff_t *ppos,
 	pgoff_t index, end_index;
 	loff_t isize;
 	size_t total_len;
-	int error;
+	int error, page_nr;
 	struct splice_pipe_desc spd = {
 		.pages = pages,
 		.partial = partial,
@@ -299,47 +281,83 @@ __generic_file_splice_read(struct file *in, loff_t *ppos,
 	 * 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.
 	 */
-	if (!loff || spd.nr_pages > 1)
-		do_page_cache_readahead(mapping, in, index, spd.nr_pages);
+	if (!loff || nr_pages > 1)
+		page_cache_readahead(mapping, &in->f_ra, in, index, nr_pages);
 
 	/*
 	 * Now fill in the holes:
 	 */
 	error = 0;
 	total_len = 0;
-	for (spd.nr_pages = 0; spd.nr_pages < nr_pages; spd.nr_pages++, index++) {
-		unsigned int this_len;
 
-		if (!len)
-			break;
+	/*
+	 * Lookup the (hopefully) full range of pages we need.
+	 */
+	spd.nr_pages = find_get_pages_contig(mapping, index, nr_pages, pages);
 
+	/*
+	 * If find_get_pages_contig() returned fewer pages than we needed,
+	 * allocate the rest.
+	 */
+	index += spd.nr_pages;
+	while (spd.nr_pages < nr_pages) {
 		/*
-		 * this_len is the max we'll use from this page
-		 */
-		this_len = min_t(unsigned long, len, PAGE_CACHE_SIZE - loff);
-find_page:
-		/*
-		 * lookup the page for this index
+		 * Page could be there, find_get_pages_contig() breaks on
+		 * the first hole.
 		 */
 		page = find_get_page(mapping, index);
 		if (!page) {
 			/*
-			 * page didn't exist, allocate one
+			 * Make sure the read-ahead engine is notified
+			 * about this failure.
+			 */
+			handle_ra_miss(mapping, &in->f_ra, index);
+
+			/*
+			 * 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));
+					      mapping_gfp_mask(mapping));
 			if (unlikely(error)) {
 				page_cache_release(page);
+				if (error == -EEXIST)
+					continue;
 				break;
 			}
-
-			goto readpage;
+			/*
+			 * add_to_page_cache() locks the page, unlock it
+			 * to avoid convoluting the logic below even more.
+			 */
+			unlock_page(page);
 		}
 
+		pages[spd.nr_pages++] = page;
+		index++;
+	}
+
+	/*
+	 * Now loop over the map and see if we need to start IO on any
+	 * pages, fill in the partial map, etc.
+	 */
+	index = *ppos >> PAGE_CACHE_SHIFT;
+	nr_pages = spd.nr_pages;
+	spd.nr_pages = 0;
+	for (page_nr = 0; page_nr < nr_pages; page_nr++) {
+		unsigned int this_len;
+
+		if (!len)
+			break;
+
+		/*
+		 * this_len is the max we'll use from this page
+		 */
+		this_len = min_t(unsigned long, len, PAGE_CACHE_SIZE - loff);
+		page = pages[page_nr];
+
 		/*
 		 * If the page isn't uptodate, we may need to start io on it
 		 */
@@ -360,7 +378,6 @@ find_page:
 			 */
 			if (!page->mapping) {
 				unlock_page(page);
-				page_cache_release(page);
 				break;
 			}
 			/*
@@ -371,16 +388,20 @@ find_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);
+				/*
+				 * We really should re-lookup the page here,
+				 * but it complicates things a lot. Instead
+				 * lets just do what we already stored, and
+				 * we'll get it the next time we are called.
+				 */
 				if (error == AOP_TRUNCATED_PAGE)
-					goto find_page;
+					error = 0;
+
 				break;
 			}
 
@@ -389,10 +410,8 @@ readpage:
 			 */
 			isize = i_size_read(mapping->host);
 			end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
-			if (unlikely(!isize || index > end_index)) {
-				page_cache_release(page);
+			if (unlikely(!isize || index > end_index))
 				break;
-			}
 
 			/*
 			 * if this is the last page, see if we need to shrink
@@ -400,27 +419,33 @@ readpage:
 			 */
 			if (end_index == index) {
 				loff = PAGE_CACHE_SIZE - (isize & ~PAGE_CACHE_MASK);
-				if (total_len + loff > isize) {
-					page_cache_release(page);
+				if (total_len + loff > isize)
 					break;
-				}
 				/*
 				 * force quit after adding this page
 				 */
-				nr_pages = spd.nr_pages;
+				len = this_len;
 				this_len = min(this_len, loff);
 				loff = 0;
 			}
 		}
 fill_it:
-		pages[spd.nr_pages] = page;
-		partial[spd.nr_pages].offset = loff;
-		partial[spd.nr_pages].len = this_len;
+		partial[page_nr].offset = loff;
+		partial[page_nr].len = this_len;
 		len -= this_len;
 		total_len += this_len;
 		loff = 0;
+		spd.nr_pages++;
+		index++;
 	}
 
+	/*
+	 * Release any pages at the end, if we quit early. 'i' is how far
+	 * we got, 'nr_pages' is how many pages are in the map.
+	 */
+	while (page_nr < nr_pages)
+		page_cache_release(pages[page_nr++]);
+
 	if (spd.nr_pages)
 		return splice_to_pipe(pipe, &spd);
 
@@ -477,31 +502,21 @@ EXPORT_SYMBOL(generic_file_splice_read);
  * Send 'sd->len' bytes to socket from 'sd->file' at position 'sd->pos'
  * using sendpage(). Return the number of bytes sent.
  */
-static int pipe_to_sendpage(struct pipe_inode_info *info,
+static int pipe_to_sendpage(struct pipe_inode_info *pipe,
 			    struct pipe_buffer *buf, struct splice_desc *sd)
 {
 	struct file *file = sd->file;
 	loff_t pos = sd->pos;
-	ssize_t ret;
-	void *ptr;
-	int more;
-
-	/*
-	 * 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.
-	 */
-	ptr = buf->ops->map(file, info, buf);
-	if (IS_ERR(ptr))
-		return PTR_ERR(ptr);
+	int ret, more;
 
-	more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len;
+	ret = buf->ops->pin(pipe, buf);
+	if (!ret) {
+		more = (sd->flags & SPLICE_F_MORE) || sd->len < sd->total_len;
 
-	ret = file->f_op->sendpage(file, buf->page, buf->offset, sd->len,
-				   &pos, more);
+		ret = file->f_op->sendpage(file, buf->page, buf->offset,
+					   sd->len, &pos, more);
+	}
 
-	buf->ops->unmap(info, buf);
 	return ret;
 }
 
@@ -525,7 +540,7 @@ static int pipe_to_sendpage(struct pipe_inode_info *info,
  * 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 *info, struct pipe_buffer *buf,
+static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
 			struct splice_desc *sd)
 {
 	struct file *file = sd->file;
@@ -534,15 +549,14 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf,
 	unsigned int offset, this_len;
 	struct page *page;
 	pgoff_t index;
-	char *src;
 	int ret;
 
 	/*
 	 * make sure the data in this buffer is uptodate
 	 */
-	src = buf->ops->map(file, info, buf);
-	if (IS_ERR(src))
-		return PTR_ERR(src);
+	ret = buf->ops->pin(pipe, buf);
+	if (unlikely(ret))
+		return ret;
 
 	index = sd->pos >> PAGE_CACHE_SHIFT;
 	offset = sd->pos & ~PAGE_CACHE_MASK;
@@ -552,20 +566,25 @@ static int pipe_to_file(struct pipe_inode_info *info, struct pipe_buffer *buf,
 		this_len = PAGE_CACHE_SIZE - offset;
 
 	/*
-	 * Reuse buf page, if SPLICE_F_MOVE is set.
+	 * Reuse buf page, if SPLICE_F_MOVE is set and we are doing a full
+	 * page.
 	 */
-	if (sd->flags & SPLICE_F_MOVE) {
+	if ((sd->flags & SPLICE_F_MOVE) && this_len == PAGE_CACHE_SIZE) {
 		/*
-		 * If steal succeeds, buf->page is now pruned from the vm
-		 * side (LRU and page cache) and we can reuse it. The page
-		 * will also be looked on successful return.
+		 * If steal succeeds, buf->page is now pruned from the
+		 * pagecache and we can reuse it. The page will also be
+		 * locked on successful return.
 		 */
-		if (buf->ops->steal(info, buf))
+		if (buf->ops->steal(pipe, buf))
 			goto find_page;
 
 		page = buf->page;
-		if (add_to_page_cache(page, mapping, index, gfp_mask))
+		if (add_to_page_cache(page, mapping, index, gfp_mask)) {
+			unlock_page(page);
 			goto find_page;
+		}
+
+		page_cache_get(page);
 
 		if (!(buf->flags & PIPE_BUF_FLAG_LRU))
 			lru_cache_add(page);
@@ -619,40 +638,55 @@ find_page:
 	}
 
 	ret = mapping->a_ops->prepare_write(file, page, offset, offset+this_len);
-	if (ret == AOP_TRUNCATED_PAGE) {
+	if (unlikely(ret)) {
+		loff_t isize = i_size_read(mapping->host);
+
+		if (ret != AOP_TRUNCATED_PAGE)
+			unlock_page(page);
 		page_cache_release(page);
-		goto find_page;
-	} else if (ret)
+		if (ret == AOP_TRUNCATED_PAGE)
+			goto find_page;
+
+		/*
+		 * prepare_write() may have instantiated a few blocks
+		 * outside i_size.  Trim these off again.
+		 */
+		if (sd->pos + this_len > isize)
+			vmtruncate(mapping->host, isize);
+
 		goto out;
+	}
 
-	if (!(buf->flags & PIPE_BUF_FLAG_STOLEN)) {
-		char *dst = kmap_atomic(page, KM_USER0);
+	if (buf->page != page) {
+		/*
+		 * Careful, ->map() uses KM_USER0!
+		 */
+		char *src = buf->ops->map(pipe, buf, 1);
+		char *dst = kmap_atomic(page, KM_USER1);
 
 		memcpy(dst + offset, src + buf->offset, this_len);
 		flush_dcache_page(page);
-		kunmap_atomic(dst, KM_USER0);
+		kunmap_atomic(dst, KM_USER1);
+		buf->ops->unmap(pipe, buf, src);
 	}
 
 	ret = mapping->a_ops->commit_write(file, page, offset, offset+this_len);
-	if (ret == AOP_TRUNCATED_PAGE) {
+	if (!ret) {
+		/*
+		 * Return the number of bytes written and mark page as
+		 * accessed, we are now done!
+		 */
+		ret = this_len;
+		mark_page_accessed(page);
+		balance_dirty_pages_ratelimited(mapping);
+	} else if (ret == AOP_TRUNCATED_PAGE) {
 		page_cache_release(page);
 		goto find_page;
-	} else if (ret)
-		goto out;
-
-	/*
-	 * Return the number of bytes written.
-	 */
-	ret = this_len;
-	mark_page_accessed(page);
-	balance_dirty_pages_ratelimited(mapping);
+	}
 out:
-	if (!(buf->flags & PIPE_BUF_FLAG_STOLEN))
-		page_cache_release(page);
-
+	page_cache_release(page);
 	unlock_page(page);
 out_nomem:
-	buf->ops->unmap(info, buf);
 	return ret;
 }
 
@@ -1060,7 +1094,7 @@ static long do_splice(struct file *in, loff_t __user *off_in,
  */
 static int get_iovec_page_array(const struct iovec __user *iov,
 				unsigned int nr_vecs, struct page **pages,
-				struct partial_page *partial)
+				struct partial_page *partial, int aligned)
 {
 	int buffers = 0, error = 0;
 
@@ -1100,6 +1134,15 @@ static int get_iovec_page_array(const struct iovec __user *iov,
 		 * in the user pages.
 		 */
 		off = (unsigned long) base & ~PAGE_MASK;
+
+		/*
+		 * If asked for alignment, the offset must be zero and the
+		 * length a multiple of the PAGE_SIZE.
+		 */
+		error = -EINVAL;
+		if (aligned && (off || len & ~PAGE_MASK))
+			break;
+
 		npages = (off + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
 		if (npages > PIPE_BUFFERS - buffers)
 			npages = PIPE_BUFFERS - buffers;
@@ -1115,7 +1158,7 @@ static int get_iovec_page_array(const struct iovec __user *iov,
 		 * Fill this contiguous range into the partial page map.
 		 */
 		for (i = 0; i < error; i++) {
-			const int plen = min_t(size_t, len, PAGE_SIZE) - off;
+			const int plen = min_t(size_t, len, PAGE_SIZE - off);
 
 			partial[buffers].offset = off;
 			partial[buffers].len = plen;
@@ -1193,7 +1236,8 @@ static long do_vmsplice(struct file *file, const struct iovec __user *iov,
 	else if (unlikely(!nr_segs))
 		return 0;
 
-	spd.nr_pages = get_iovec_page_array(iov, nr_segs, pages, partial);
+	spd.nr_pages = get_iovec_page_array(iov, nr_segs, pages, partial,
+					    flags & SPLICE_F_GIFT);
 	if (spd.nr_pages <= 0)
 		return spd.nr_pages;
 
@@ -1301,6 +1345,12 @@ static int link_pipe(struct pipe_inode_info *ipipe,
 				obuf = opipe->bufs + nbuf;
 				*obuf = *ibuf;
 
+				/*
+				 * Don't inherit the gift flag, we need to
+				 * prevent multiple steals of this page.
+				 */
+				obuf->flags &= ~PIPE_BUF_FLAG_GIFT;
+
 				if (obuf->len > len)
 					obuf->len = len;
 
diff --git a/fs/stat.c b/fs/stat.c
index 9948cc1685a..0f282face32 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -261,7 +261,7 @@ asmlinkage long sys_newlstat(char __user *filename, struct stat __user *statbuf)
 	return error;
 }
 
-#ifndef __ARCH_WANT_STAT64
+#if !defined(__ARCH_WANT_STAT64) || defined(__ARCH_WANT_SYS_NEWFSTATAT)
 asmlinkage long sys_newfstatat(int dfd, char __user *filename,
 				struct stat __user *statbuf, int flag)
 {
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 64ee07db0d5..8558226281c 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -1942,8 +1942,10 @@ xfs_alloc_fix_freelist(
 		/*
 		 * Allocate as many blocks as possible at once.
 		 */
-		if ((error = xfs_alloc_ag_vextent(&targs)))
+		if ((error = xfs_alloc_ag_vextent(&targs))) {
+			xfs_trans_brelse(tp, agflbp);
 			return error;
+		}
 		/*
 		 * Stop if we run out.  Won't happen if callers are obeying
 		 * the restrictions correctly.  Can happen for free calls
@@ -1960,6 +1962,7 @@ xfs_alloc_fix_freelist(
 				return error;
 		}
 	}
+	xfs_trans_brelse(tp, agflbp);
 	args->agbp = agbp;
 	return 0;
 }
diff --git a/fs/xfs/xfs_rename.c b/fs/xfs/xfs_rename.c
index 81a05cfd77d..1f148762eb2 100644
--- a/fs/xfs/xfs_rename.c
+++ b/fs/xfs/xfs_rename.c
@@ -316,6 +316,18 @@ xfs_rename(
 		}
 	}
 
+	/*
+	 * If we are using project inheritance, we only allow renames
+	 * into our tree when the project IDs are the same; else the
+	 * tree quota mechanism would be circumvented.
+	 */
+	if (unlikely((target_dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
+		     (target_dp->i_d.di_projid != src_ip->i_d.di_projid))) {
+		error = XFS_ERROR(EXDEV);
+		xfs_rename_unlock4(inodes, XFS_ILOCK_SHARED);
+		goto rele_return;
+	}
+
 	new_parent = (src_dp != target_dp);
 	src_is_directory = ((src_ip->i_d.di_mode & S_IFMT) == S_IFDIR);
 
diff --git a/fs/xfs/xfs_vfsops.c b/fs/xfs/xfs_vfsops.c
index f0e09ca1413..36ea1b2094f 100644
--- a/fs/xfs/xfs_vfsops.c
+++ b/fs/xfs/xfs_vfsops.c
@@ -669,31 +669,22 @@ xfs_mntupdate(
 	xfs_mount_t	*mp = XFS_BHVTOM(bdp);
 	int		error;
 
-	if (args->flags & XFSMNT_BARRIER)
-		mp->m_flags |= XFS_MOUNT_BARRIER;
-	else
-		mp->m_flags &= ~XFS_MOUNT_BARRIER;
-
-	if ((vfsp->vfs_flag & VFS_RDONLY) &&
-	    !(*flags & MS_RDONLY)) {
-		vfsp->vfs_flag &= ~VFS_RDONLY;
-
-		if (args->flags & XFSMNT_BARRIER)
+	if (!(*flags & MS_RDONLY)) {			/* rw/ro -> rw */
+		if (vfsp->vfs_flag & VFS_RDONLY)
+			vfsp->vfs_flag &= ~VFS_RDONLY;
+		if (args->flags & XFSMNT_BARRIER) {
+			mp->m_flags |= XFS_MOUNT_BARRIER;
 			xfs_mountfs_check_barriers(mp);
-	}
-
-	if (!(vfsp->vfs_flag & VFS_RDONLY) &&
-	    (*flags & MS_RDONLY)) {
+		} else {
+			mp->m_flags &= ~XFS_MOUNT_BARRIER;
+		}
+	} else if (!(vfsp->vfs_flag & VFS_RDONLY)) {	/* rw -> ro */
 		VFS_SYNC(vfsp, SYNC_FSDATA|SYNC_BDFLUSH|SYNC_ATTR, NULL, error);
-
 		xfs_quiesce_fs(mp);
-
-		/* Ok now write out an unmount record */
 		xfs_log_unmount_write(mp);
 		xfs_unmountfs_writesb(mp);
 		vfsp->vfs_flag |= VFS_RDONLY;
 	}
-
 	return 0;
 }
 
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index fa71b305ba5..7027ae68ee3 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -2663,7 +2663,7 @@ xfs_link(
 	 */
 	if (unlikely((tdp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT) &&
 		     (tdp->i_d.di_projid != sip->i_d.di_projid))) {
-		error = XFS_ERROR(EPERM);
+		error = XFS_ERROR(EXDEV);
 		goto error_return;
 	}