From 40431d6c1288793a682fc6f5e5b5c9d5cac34608 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Wed, 5 Aug 2009 12:57:59 -0400
Subject: Btrfs: optimize set extent bit

The Btrfs set_extent_bit call currently searches the rbtree
every time it needs to find more extent_state objects to fill
the requested operation.

This adds a simple test with rb_next to see if the next object
in the tree was adjacent to the one we just found.  If so,
we skip the search and just use the next object.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/file.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 7c3cd248d8d..a760d97279a 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -136,8 +136,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
 	btrfs_set_trans_block_group(trans, inode);
 	hint_byte = 0;
 
-	set_extent_uptodate(io_tree, start_pos, end_of_last_block, GFP_NOFS);
-
 	/* check for reserved extents on each page, we don't want
 	 * to reset the delalloc bit on things that already have
 	 * extents reserved.
-- 
cgit v1.2.3-70-g09d2


From 890871be854b5f5e43e7ba2475f706209906cc24 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Wed, 2 Sep 2009 16:24:52 -0400
Subject: Btrfs: switch extent_map to a rw lock

There are two main users of the extent_map tree.  The
first is regular file inodes, where it is evenly spread
between readers and writers.

The second is the chunk allocation tree, which maps blocks from
logical addresses to phyiscal ones, and it is 99.99% reads.

The mapping tree is a point of lock contention during heavy IO
workloads, so this commit switches things to a rw lock.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/compression.c |  8 ++++----
 fs/btrfs/disk-io.c     | 14 +++++++-------
 fs/btrfs/extent-tree.c |  4 ++--
 fs/btrfs/extent_io.c   |  8 ++++----
 fs/btrfs/extent_map.c  |  5 +----
 fs/btrfs/extent_map.h  |  2 +-
 fs/btrfs/file.c        |  8 ++++----
 fs/btrfs/inode.c       | 24 ++++++++++++------------
 fs/btrfs/relocation.c  |  4 ++--
 fs/btrfs/volumes.c     | 40 ++++++++++++++++++++--------------------
 10 files changed, 57 insertions(+), 60 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
index de1e2fd3208..78451a58f20 100644
--- a/fs/btrfs/compression.c
+++ b/fs/btrfs/compression.c
@@ -507,10 +507,10 @@ static noinline int add_ra_bio_pages(struct inode *inode,
 		 */
 		set_page_extent_mapped(page);
 		lock_extent(tree, last_offset, end, GFP_NOFS);
-		spin_lock(&em_tree->lock);
+		read_lock(&em_tree->lock);
 		em = lookup_extent_mapping(em_tree, last_offset,
 					   PAGE_CACHE_SIZE);
-		spin_unlock(&em_tree->lock);
+		read_unlock(&em_tree->lock);
 
 		if (!em || last_offset < em->start ||
 		    (last_offset + PAGE_CACHE_SIZE > extent_map_end(em)) ||
@@ -594,11 +594,11 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	em_tree = &BTRFS_I(inode)->extent_tree;
 
 	/* we need the actual starting offset of this extent in the file */
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree,
 				   page_offset(bio->bi_io_vec->bv_page),
 				   PAGE_CACHE_SIZE);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	compressed_len = em->block_len;
 	cb = kmalloc(compressed_bio_size(root, compressed_len), GFP_NOFS);
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 20cefc6f22c..b6cfdd9164e 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -121,15 +121,15 @@ static struct extent_map *btree_get_extent(struct inode *inode,
 	struct extent_map *em;
 	int ret;
 
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, start, len);
 	if (em) {
 		em->bdev =
 			BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
-		spin_unlock(&em_tree->lock);
+		read_unlock(&em_tree->lock);
 		goto out;
 	}
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	em = alloc_extent_map(GFP_NOFS);
 	if (!em) {
@@ -142,7 +142,7 @@ static struct extent_map *btree_get_extent(struct inode *inode,
 	em->block_start = 0;
 	em->bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev;
 
-	spin_lock(&em_tree->lock);
+	write_lock(&em_tree->lock);
 	ret = add_extent_mapping(em_tree, em);
 	if (ret == -EEXIST) {
 		u64 failed_start = em->start;
@@ -161,7 +161,7 @@ static struct extent_map *btree_get_extent(struct inode *inode,
 		free_extent_map(em);
 		em = NULL;
 	}
-	spin_unlock(&em_tree->lock);
+	write_unlock(&em_tree->lock);
 
 	if (ret)
 		em = ERR_PTR(ret);
@@ -1323,9 +1323,9 @@ static void btrfs_unplug_io_fn(struct backing_dev_info *bdi, struct page *page)
 	offset = page_offset(page);
 
 	em_tree = &BTRFS_I(inode)->extent_tree;
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, offset, PAGE_CACHE_SIZE);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 	if (!em) {
 		__unplug_io_fn(bdi, page);
 		return;
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 72a2b9c28e9..edd86ae9e14 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -5396,9 +5396,9 @@ static noinline int relocate_data_extent(struct inode *reloc_inode,
 	lock_extent(&BTRFS_I(reloc_inode)->io_tree, start, end, GFP_NOFS);
 	while (1) {
 		int ret;
-		spin_lock(&em_tree->lock);
+		write_lock(&em_tree->lock);
 		ret = add_extent_mapping(em_tree, em);
-		spin_unlock(&em_tree->lock);
+		write_unlock(&em_tree->lock);
 		if (ret != -EEXIST) {
 			free_extent_map(em);
 			break;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 8d7a152a90c..41cf1b451b4 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -2786,15 +2786,15 @@ int try_release_extent_mapping(struct extent_map_tree *map,
 		u64 len;
 		while (start <= end) {
 			len = end - start + 1;
-			spin_lock(&map->lock);
+			write_lock(&map->lock);
 			em = lookup_extent_mapping(map, start, len);
 			if (!em || IS_ERR(em)) {
-				spin_unlock(&map->lock);
+				write_unlock(&map->lock);
 				break;
 			}
 			if (test_bit(EXTENT_FLAG_PINNED, &em->flags) ||
 			    em->start != start) {
-				spin_unlock(&map->lock);
+				write_unlock(&map->lock);
 				free_extent_map(em);
 				break;
 			}
@@ -2808,7 +2808,7 @@ int try_release_extent_mapping(struct extent_map_tree *map,
 				free_extent_map(em);
 			}
 			start = extent_map_end(em);
-			spin_unlock(&map->lock);
+			write_unlock(&map->lock);
 
 			/* once for us */
 			free_extent_map(em);
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index 30c9365861e..72e9fa3c31f 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -36,7 +36,7 @@ void extent_map_exit(void)
 void extent_map_tree_init(struct extent_map_tree *tree, gfp_t mask)
 {
 	tree->map.rb_node = NULL;
-	spin_lock_init(&tree->lock);
+	rwlock_init(&tree->lock);
 }
 
 /**
@@ -222,7 +222,6 @@ int add_extent_mapping(struct extent_map_tree *tree,
 		ret = -EEXIST;
 		goto out;
 	}
-	assert_spin_locked(&tree->lock);
 	rb = tree_insert(&tree->map, em->start, &em->rb_node);
 	if (rb) {
 		ret = -EEXIST;
@@ -285,7 +284,6 @@ struct extent_map *lookup_extent_mapping(struct extent_map_tree *tree,
 	struct rb_node *next = NULL;
 	u64 end = range_end(start, len);
 
-	assert_spin_locked(&tree->lock);
 	rb_node = __tree_search(&tree->map, start, &prev, &next);
 	if (!rb_node && prev) {
 		em = rb_entry(prev, struct extent_map, rb_node);
@@ -331,7 +329,6 @@ int remove_extent_mapping(struct extent_map_tree *tree, struct extent_map *em)
 	int ret = 0;
 
 	WARN_ON(test_bit(EXTENT_FLAG_PINNED, &em->flags));
-	assert_spin_locked(&tree->lock);
 	rb_erase(&em->rb_node, &tree->map);
 	em->in_tree = 0;
 	return ret;
diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h
index fb6eeef06bb..6216dfbcf9b 100644
--- a/fs/btrfs/extent_map.h
+++ b/fs/btrfs/extent_map.h
@@ -31,7 +31,7 @@ struct extent_map {
 
 struct extent_map_tree {
 	struct rb_root map;
-	spinlock_t lock;
+	rwlock_t lock;
 };
 
 static inline u64 extent_map_end(struct extent_map *em)
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index a760d97279a..8a9c76aecdf 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -188,15 +188,15 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 		if (!split2)
 			split2 = alloc_extent_map(GFP_NOFS);
 
-		spin_lock(&em_tree->lock);
+		write_lock(&em_tree->lock);
 		em = lookup_extent_mapping(em_tree, start, len);
 		if (!em) {
-			spin_unlock(&em_tree->lock);
+			write_unlock(&em_tree->lock);
 			break;
 		}
 		flags = em->flags;
 		if (skip_pinned && test_bit(EXTENT_FLAG_PINNED, &em->flags)) {
-			spin_unlock(&em_tree->lock);
+			write_unlock(&em_tree->lock);
 			if (em->start <= start &&
 			    (!testend || em->start + em->len >= start + len)) {
 				free_extent_map(em);
@@ -259,7 +259,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			free_extent_map(split);
 			split = NULL;
 		}
-		spin_unlock(&em_tree->lock);
+		write_unlock(&em_tree->lock);
 
 		/* once for us */
 		free_extent_map(em);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 04b53b5ebe5..f1df1171861 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -612,9 +612,9 @@ static noinline int submit_compressed_extents(struct inode *inode,
 		set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
 
 		while (1) {
-			spin_lock(&em_tree->lock);
+			write_lock(&em_tree->lock);
 			ret = add_extent_mapping(em_tree, em);
-			spin_unlock(&em_tree->lock);
+			write_unlock(&em_tree->lock);
 			if (ret != -EEXIST) {
 				free_extent_map(em);
 				break;
@@ -748,9 +748,9 @@ static noinline int cow_file_range(struct inode *inode,
 		set_bit(EXTENT_FLAG_PINNED, &em->flags);
 
 		while (1) {
-			spin_lock(&em_tree->lock);
+			write_lock(&em_tree->lock);
 			ret = add_extent_mapping(em_tree, em);
-			spin_unlock(&em_tree->lock);
+			write_unlock(&em_tree->lock);
 			if (ret != -EEXIST) {
 				free_extent_map(em);
 				break;
@@ -1081,9 +1081,9 @@ out_check:
 			em->bdev = root->fs_info->fs_devices->latest_bdev;
 			set_bit(EXTENT_FLAG_PINNED, &em->flags);
 			while (1) {
-				spin_lock(&em_tree->lock);
+				write_lock(&em_tree->lock);
 				ret = add_extent_mapping(em_tree, em);
-				spin_unlock(&em_tree->lock);
+				write_unlock(&em_tree->lock);
 				if (ret != -EEXIST) {
 					free_extent_map(em);
 					break;
@@ -1670,13 +1670,13 @@ static int btrfs_io_failed_hook(struct bio *failed_bio,
 		failrec->last_mirror = 0;
 		failrec->bio_flags = 0;
 
-		spin_lock(&em_tree->lock);
+		read_lock(&em_tree->lock);
 		em = lookup_extent_mapping(em_tree, start, failrec->len);
 		if (em->start > start || em->start + em->len < start) {
 			free_extent_map(em);
 			em = NULL;
 		}
-		spin_unlock(&em_tree->lock);
+		read_unlock(&em_tree->lock);
 
 		if (!em || IS_ERR(em)) {
 			kfree(failrec);
@@ -4069,11 +4069,11 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
 	int compressed;
 
 again:
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, start, len);
 	if (em)
 		em->bdev = root->fs_info->fs_devices->latest_bdev;
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	if (em) {
 		if (em->start > start || em->start + em->len <= start)
@@ -4264,7 +4264,7 @@ insert:
 	}
 
 	err = 0;
-	spin_lock(&em_tree->lock);
+	write_lock(&em_tree->lock);
 	ret = add_extent_mapping(em_tree, em);
 	/* it is possible that someone inserted the extent into the tree
 	 * while we had the lock dropped.  It is also possible that
@@ -4304,7 +4304,7 @@ insert:
 			err = 0;
 		}
 	}
-	spin_unlock(&em_tree->lock);
+	write_unlock(&em_tree->lock);
 out:
 	if (path)
 		btrfs_free_path(path);
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index c04f7f21260..4adab903fc2 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2646,9 +2646,9 @@ int relocate_data_extent(struct inode *inode, struct btrfs_key *extent_key)
 	lock_extent(&BTRFS_I(inode)->io_tree, start, end, GFP_NOFS);
 	while (1) {
 		int ret;
-		spin_lock(&em_tree->lock);
+		write_lock(&em_tree->lock);
 		ret = add_extent_mapping(em_tree, em);
-		spin_unlock(&em_tree->lock);
+		write_unlock(&em_tree->lock);
 		if (ret != -EEXIST) {
 			free_extent_map(em);
 			break;
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index a7e53773e74..d2358c06bbd 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -1749,9 +1749,9 @@ static int btrfs_relocate_chunk(struct btrfs_root *root,
 	 * step two, delete the device extents and the
 	 * chunk tree entries
 	 */
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, chunk_offset, 1);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	BUG_ON(em->start > chunk_offset ||
 	       em->start + em->len < chunk_offset);
@@ -1780,9 +1780,9 @@ static int btrfs_relocate_chunk(struct btrfs_root *root,
 	ret = btrfs_remove_block_group(trans, extent_root, chunk_offset);
 	BUG_ON(ret);
 
-	spin_lock(&em_tree->lock);
+	write_lock(&em_tree->lock);
 	remove_extent_mapping(em_tree, em);
-	spin_unlock(&em_tree->lock);
+	write_unlock(&em_tree->lock);
 
 	kfree(map);
 	em->bdev = NULL;
@@ -2294,9 +2294,9 @@ again:
 	em->block_len = em->len;
 
 	em_tree = &extent_root->fs_info->mapping_tree.map_tree;
-	spin_lock(&em_tree->lock);
+	write_lock(&em_tree->lock);
 	ret = add_extent_mapping(em_tree, em);
-	spin_unlock(&em_tree->lock);
+	write_unlock(&em_tree->lock);
 	BUG_ON(ret);
 	free_extent_map(em);
 
@@ -2491,9 +2491,9 @@ int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset)
 	int readonly = 0;
 	int i;
 
-	spin_lock(&map_tree->map_tree.lock);
+	read_lock(&map_tree->map_tree.lock);
 	em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1);
-	spin_unlock(&map_tree->map_tree.lock);
+	read_unlock(&map_tree->map_tree.lock);
 	if (!em)
 		return 1;
 
@@ -2518,11 +2518,11 @@ void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree)
 	struct extent_map *em;
 
 	while (1) {
-		spin_lock(&tree->map_tree.lock);
+		write_lock(&tree->map_tree.lock);
 		em = lookup_extent_mapping(&tree->map_tree, 0, (u64)-1);
 		if (em)
 			remove_extent_mapping(&tree->map_tree, em);
-		spin_unlock(&tree->map_tree.lock);
+		write_unlock(&tree->map_tree.lock);
 		if (!em)
 			break;
 		kfree(em->bdev);
@@ -2540,9 +2540,9 @@ int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len)
 	struct extent_map_tree *em_tree = &map_tree->map_tree;
 	int ret;
 
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, logical, len);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 	BUG_ON(!em);
 
 	BUG_ON(em->start > logical || em->start + em->len < logical);
@@ -2604,9 +2604,9 @@ again:
 		atomic_set(&multi->error, 0);
 	}
 
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, logical, *length);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	if (!em && unplug_page)
 		return 0;
@@ -2763,9 +2763,9 @@ int btrfs_rmap_block(struct btrfs_mapping_tree *map_tree,
 	u64 stripe_nr;
 	int i, j, nr = 0;
 
-	spin_lock(&em_tree->lock);
+	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, chunk_start, 1);
-	spin_unlock(&em_tree->lock);
+	read_unlock(&em_tree->lock);
 
 	BUG_ON(!em || em->start != chunk_start);
 	map = (struct map_lookup *)em->bdev;
@@ -3053,9 +3053,9 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 	logical = key->offset;
 	length = btrfs_chunk_length(leaf, chunk);
 
-	spin_lock(&map_tree->map_tree.lock);
+	read_lock(&map_tree->map_tree.lock);
 	em = lookup_extent_mapping(&map_tree->map_tree, logical, 1);
-	spin_unlock(&map_tree->map_tree.lock);
+	read_unlock(&map_tree->map_tree.lock);
 
 	/* already mapped? */
 	if (em && em->start <= logical && em->start + em->len > logical) {
@@ -3114,9 +3114,9 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key,
 		map->stripes[i].dev->in_fs_metadata = 1;
 	}
 
-	spin_lock(&map_tree->map_tree.lock);
+	write_lock(&map_tree->map_tree.lock);
 	ret = add_extent_mapping(&map_tree->map_tree, em);
-	spin_unlock(&map_tree->map_tree.lock);
+	write_unlock(&map_tree->map_tree.lock);
 	BUG_ON(ret);
 	free_extent_map(em);
 
-- 
cgit v1.2.3-70-g09d2


From 1edbb734b4e010974c41d2859d22a43d04f5f1cf Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Wed, 2 Sep 2009 13:24:36 -0400
Subject: Btrfs: reduce CPU usage in the extent_state tree

Btrfs is currently mirroring some of the page state bits into
its extent state tree.  The goal behind this was to use it in supporting
blocksizes other than the page size.

But, we don't currently support that, and we're using quite a lot of CPU
on the rb tree and its spin lock.  This commit starts a series of
cleanups to reduce the amount of work done in the extent state tree as
part of each IO.

This commit:

* Adds the ability to lock an extent in the state tree and also set
other bits.  The idea is to do locking and delalloc in one call

* Removes the EXTENT_WRITEBACK and EXTENT_DIRTY bits.  Btrfs is using
a combination of the page bits and the ordered write code for this
instead.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/extent_io.c | 75 ++++++++++++++++++----------------------------------
 fs/btrfs/extent_io.h |  2 ++
 fs/btrfs/file.c      | 19 -------------
 3 files changed, 28 insertions(+), 68 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 8e168a457a3..7c70613eb72 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -654,25 +654,24 @@ static void set_state_bits(struct extent_io_tree *tree,
 }
 
 /*
- * set some bits on a range in the tree.  This may require allocations
- * or sleeping, so the gfp mask is used to indicate what is allowed.
+ * set some bits on a range in the tree.  This may require allocations or
+ * sleeping, so the gfp mask is used to indicate what is allowed.
  *
- * If 'exclusive' == 1, this will fail with -EEXIST if some part of the
- * range already has the desired bits set.  The start of the existing
- * range is returned in failed_start in this case.
+ * If any of the exclusive bits are set, this will fail with -EEXIST if some
+ * part of the range already has the desired bits set.  The start of the
+ * existing range is returned in failed_start in this case.
  *
- * [start, end] is inclusive
- * This takes the tree lock.
+ * [start, end] is inclusive This takes the tree lock.
  */
+
 static int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end,
-			  int bits, int exclusive, u64 *failed_start,
+			  int bits, int exclusive_bits, u64 *failed_start,
 			  gfp_t mask)
 {
 	struct extent_state *state;
 	struct extent_state *prealloc = NULL;
 	struct rb_node *node;
 	int err = 0;
-	int set;
 	u64 last_start;
 	u64 last_end;
 again:
@@ -707,8 +706,7 @@ hit_next:
 	 */
 	if (state->start == start && state->end <= end) {
 		struct rb_node *next_node;
-		set = state->state & bits;
-		if (set && exclusive) {
+		if (state->state & exclusive_bits) {
 			*failed_start = state->start;
 			err = -EEXIST;
 			goto out;
@@ -748,8 +746,7 @@ hit_next:
 	 * desired bit on it.
 	 */
 	if (state->start < start) {
-		set = state->state & bits;
-		if (exclusive && set) {
+		if (state->state & exclusive_bits) {
 			*failed_start = start;
 			err = -EEXIST;
 			goto out;
@@ -799,8 +796,7 @@ hit_next:
 	 * on the first half
 	 */
 	if (state->start <= end && state->end > end) {
-		set = state->state & bits;
-		if (exclusive && set) {
+		if (state->state & exclusive_bits) {
 			*failed_start = start;
 			err = -EEXIST;
 			goto out;
@@ -906,19 +902,6 @@ static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start,
 	return clear_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, 0, mask);
 }
 
-static int set_extent_writeback(struct extent_io_tree *tree, u64 start, u64 end,
-			 gfp_t mask)
-{
-	return set_extent_bit(tree, start, end, EXTENT_WRITEBACK,
-			      0, NULL, mask);
-}
-
-static int clear_extent_writeback(struct extent_io_tree *tree, u64 start,
-				  u64 end, gfp_t mask)
-{
-	return clear_extent_bit(tree, start, end, EXTENT_WRITEBACK, 1, 0, mask);
-}
-
 int wait_on_extent_writeback(struct extent_io_tree *tree, u64 start, u64 end)
 {
 	return wait_extent_bit(tree, start, end, EXTENT_WRITEBACK);
@@ -928,13 +911,14 @@ int wait_on_extent_writeback(struct extent_io_tree *tree, u64 start, u64 end)
  * either insert or lock state struct between start and end use mask to tell
  * us if waiting is desired.
  */
-int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask)
+int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+		     int bits, gfp_t mask)
 {
 	int err;
 	u64 failed_start;
 	while (1) {
-		err = set_extent_bit(tree, start, end, EXTENT_LOCKED, 1,
-				     &failed_start, mask);
+		err = set_extent_bit(tree, start, end, EXTENT_LOCKED | bits,
+				     EXTENT_LOCKED, &failed_start, mask);
 		if (err == -EEXIST && (mask & __GFP_WAIT)) {
 			wait_extent_bit(tree, failed_start, end, EXTENT_LOCKED);
 			start = failed_start;
@@ -946,6 +930,11 @@ int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask)
 	return err;
 }
 
+int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask)
+{
+	return lock_extent_bits(tree, start, end, 0, mask);
+}
+
 int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end,
 		    gfp_t mask)
 {
@@ -985,7 +974,6 @@ int set_range_dirty(struct extent_io_tree *tree, u64 start, u64 end)
 		page_cache_release(page);
 		index++;
 	}
-	set_extent_dirty(tree, start, end, GFP_NOFS);
 	return 0;
 }
 
@@ -1005,7 +993,6 @@ static int set_range_writeback(struct extent_io_tree *tree, u64 start, u64 end)
 		page_cache_release(page);
 		index++;
 	}
-	set_extent_writeback(tree, start, end, GFP_NOFS);
 	return 0;
 }
 
@@ -1563,10 +1550,7 @@ static int check_page_locked(struct extent_io_tree *tree,
 static int check_page_writeback(struct extent_io_tree *tree,
 			     struct page *page)
 {
-	u64 start = (u64)page->index << PAGE_CACHE_SHIFT;
-	u64 end = start + PAGE_CACHE_SIZE - 1;
-	if (!test_range_bit(tree, start, end, EXTENT_WRITEBACK, 0))
-		end_page_writeback(page);
+	end_page_writeback(page);
 	return 0;
 }
 
@@ -1624,13 +1608,11 @@ static void end_bio_extent_writepage(struct bio *bio, int err)
 		}
 
 		if (!uptodate) {
-			clear_extent_uptodate(tree, start, end, GFP_ATOMIC);
+			clear_extent_uptodate(tree, start, end, GFP_NOFS);
 			ClearPageUptodate(page);
 			SetPageError(page);
 		}
 
-		clear_extent_writeback(tree, start, end, GFP_ATOMIC);
-
 		if (whole_page)
 			end_page_writeback(page);
 		else
@@ -2208,8 +2190,9 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 		printk(KERN_ERR "btrfs delalloc bits after lock_extent\n");
 
 	if (last_byte <= start) {
-		clear_extent_dirty(tree, start, page_end, GFP_NOFS);
-		unlock_extent(tree, start, page_end, GFP_NOFS);
+		clear_extent_bit(tree, start, page_end,
+				 EXTENT_LOCKED | EXTENT_DIRTY,
+				 1, 0, GFP_NOFS);
 		if (tree->ops && tree->ops->writepage_end_io_hook)
 			tree->ops->writepage_end_io_hook(page, start,
 							 page_end, NULL, 1);
@@ -2217,12 +2200,10 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 		goto done;
 	}
 
-	set_extent_uptodate(tree, start, page_end, GFP_NOFS);
 	blocksize = inode->i_sb->s_blocksize;
 
 	while (cur <= end) {
 		if (cur >= last_byte) {
-			clear_extent_dirty(tree, cur, page_end, GFP_NOFS);
 			unlock_extent(tree, unlock_start, page_end, GFP_NOFS);
 			if (tree->ops && tree->ops->writepage_end_io_hook)
 				tree->ops->writepage_end_io_hook(page, cur,
@@ -2255,9 +2236,6 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 		 */
 		if (compressed || block_start == EXTENT_MAP_HOLE ||
 		    block_start == EXTENT_MAP_INLINE) {
-			clear_extent_dirty(tree, cur,
-					   cur + iosize - 1, GFP_NOFS);
-
 			unlock_extent(tree, unlock_start, cur + iosize - 1,
 				      GFP_NOFS);
 
@@ -2291,7 +2269,6 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc,
 			continue;
 		}
 
-		clear_extent_dirty(tree, cur, cur + iosize - 1, GFP_NOFS);
 		if (tree->ops && tree->ops->writepage_io_hook) {
 			ret = tree->ops->writepage_io_hook(page, cur,
 						cur + iosize - 1);
@@ -2619,7 +2596,7 @@ int extent_invalidatepage(struct extent_io_tree *tree,
 		return 0;
 
 	lock_extent(tree, start, end, GFP_NOFS);
-	wait_on_extent_writeback(tree, start, end);
+	wait_on_page_writeback(page);
 	clear_extent_bit(tree, start, end,
 			 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC,
 			 1, 1, GFP_NOFS);
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 5bc20abf3f3..88d134d01fb 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -142,6 +142,8 @@ int try_release_extent_state(struct extent_map_tree *map,
 			     struct extent_io_tree *tree, struct page *page,
 			     gfp_t mask);
 int lock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask);
+int lock_extent_bits(struct extent_io_tree *tree, u64 start, u64 end,
+		     int bits, gfp_t mask);
 int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask);
 int try_lock_extent(struct extent_io_tree *tree, u64 start, u64 end,
 		    gfp_t mask);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 8a9c76aecdf..ef66c3d989b 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -113,8 +113,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
 	int err = 0;
 	int i;
 	struct inode *inode = fdentry(file)->d_inode;
-	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
-	u64 hint_byte;
 	u64 num_bytes;
 	u64 start_pos;
 	u64 end_of_last_block;
@@ -126,20 +124,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
 		    root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
 
 	end_of_last_block = start_pos + num_bytes - 1;
-
-	lock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
-	trans = btrfs_join_transaction(root, 1);
-	if (!trans) {
-		err = -ENOMEM;
-		goto out_unlock;
-	}
-	btrfs_set_trans_block_group(trans, inode);
-	hint_byte = 0;
-
-	/* check for reserved extents on each page, we don't want
-	 * to reset the delalloc bit on things that already have
-	 * extents reserved.
-	 */
 	btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
 	for (i = 0; i < num_pages; i++) {
 		struct page *p = pages[i];
@@ -154,9 +138,6 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
 		 * at this time.
 		 */
 	}
-	err = btrfs_end_transaction(trans, root);
-out_unlock:
-	unlock_extent(io_tree, start_pos, end_of_last_block, GFP_NOFS);
 	return err;
 }
 
-- 
cgit v1.2.3-70-g09d2


From a1ed835e1ab5795f91b198d08c43e2f56848dcf3 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Fri, 11 Sep 2009 12:27:37 -0400
Subject: Btrfs: Fix extent replacment race

Data COW means that whenever we write to a file, we replace any old
extent pointers with new ones.  There was a window where a readpage
might find the old extent pointers on disk and cache them in the
extent_map tree in ram in the middle of a given write replacing them.

Even though both the readpage and the write had their respective bytes
in the file locked, the extent readpage inserts may cover more bytes than
it had locked down.

This commit closes the race by keeping the new extent pinned in the extent
map tree until after the on-disk btree is properly setup with the new
extent pointers.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/ctree.h      |  2 +-
 fs/btrfs/extent_map.c | 50 ++++++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/extent_map.h |  1 +
 fs/btrfs/file.c       |  8 +++++---
 fs/btrfs/inode.c      | 25 ++++++++++++++++++++-----
 fs/btrfs/ioctl.c      |  5 ++---
 fs/btrfs/tree-log.c   |  2 +-
 7 files changed, 80 insertions(+), 13 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 38eeb6c49c8..1ceab8b4d6d 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2292,7 +2292,7 @@ extern struct file_operations btrfs_file_operations;
 int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root, struct inode *inode,
 		       u64 start, u64 end, u64 locked_end,
-		       u64 inline_limit, u64 *hint_block);
+		       u64 inline_limit, u64 *hint_block, int drop_cache);
 int btrfs_mark_extent_written(struct btrfs_trans_handle *trans,
 			      struct btrfs_root *root,
 			      struct inode *inode, u64 start, u64 end);
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index 72e9fa3c31f..5bc7a0d325e 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -198,6 +198,56 @@ static int mergable_maps(struct extent_map *prev, struct extent_map *next)
 	return 0;
 }
 
+int unpin_extent_cache(struct extent_map_tree *tree, u64 start, u64 len)
+{
+	int ret = 0;
+	struct extent_map *merge = NULL;
+	struct rb_node *rb;
+	struct extent_map *em;
+
+	write_lock(&tree->lock);
+	em = lookup_extent_mapping(tree, start, len);
+
+	WARN_ON(em->start != start || !em);
+
+	if (!em)
+		goto out;
+
+	clear_bit(EXTENT_FLAG_PINNED, &em->flags);
+
+	if (em->start != 0) {
+		rb = rb_prev(&em->rb_node);
+		if (rb)
+			merge = rb_entry(rb, struct extent_map, rb_node);
+		if (rb && mergable_maps(merge, em)) {
+			em->start = merge->start;
+			em->len += merge->len;
+			em->block_len += merge->block_len;
+			em->block_start = merge->block_start;
+			merge->in_tree = 0;
+			rb_erase(&merge->rb_node, &tree->map);
+			free_extent_map(merge);
+		}
+	}
+
+	rb = rb_next(&em->rb_node);
+	if (rb)
+		merge = rb_entry(rb, struct extent_map, rb_node);
+	if (rb && mergable_maps(em, merge)) {
+		em->len += merge->len;
+		em->block_len += merge->len;
+		rb_erase(&merge->rb_node, &tree->map);
+		merge->in_tree = 0;
+		free_extent_map(merge);
+	}
+
+	free_extent_map(em);
+out:
+	write_unlock(&tree->lock);
+	return ret;
+
+}
+
 /**
  * add_extent_mapping - add new extent map to the extent tree
  * @tree:	tree to insert new map in
diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h
index 6216dfbcf9b..d3d442f4bbb 100644
--- a/fs/btrfs/extent_map.h
+++ b/fs/btrfs/extent_map.h
@@ -59,4 +59,5 @@ struct extent_map *alloc_extent_map(gfp_t mask);
 void free_extent_map(struct extent_map *em);
 int __init extent_map_init(void);
 void extent_map_exit(void);
+int unpin_extent_cache(struct extent_map_tree *tree, u64 start, u64 len);
 #endif
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index ef66c3d989b..4123db9d514 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -177,10 +177,10 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 		}
 		flags = em->flags;
 		if (skip_pinned && test_bit(EXTENT_FLAG_PINNED, &em->flags)) {
-			write_unlock(&em_tree->lock);
 			if (em->start <= start &&
 			    (!testend || em->start + em->len >= start + len)) {
 				free_extent_map(em);
+				write_unlock(&em_tree->lock);
 				break;
 			}
 			if (start < em->start) {
@@ -190,6 +190,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 				start = em->start + em->len;
 			}
 			free_extent_map(em);
+			write_unlock(&em_tree->lock);
 			continue;
 		}
 		compressed = test_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
@@ -269,7 +270,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root, struct inode *inode,
 		       u64 start, u64 end, u64 locked_end,
-		       u64 inline_limit, u64 *hint_byte)
+		       u64 inline_limit, u64 *hint_byte, int drop_cache)
 {
 	u64 extent_end = 0;
 	u64 search_start = start;
@@ -294,7 +295,8 @@ noinline int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 	int ret;
 
 	inline_limit = 0;
-	btrfs_drop_extent_cache(inode, start, end - 1, 0);
+	if (drop_cache)
+		btrfs_drop_extent_cache(inode, start, end - 1, 0);
 
 	path = btrfs_alloc_path();
 	if (!path)
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 739a245e25d..233fe6f2612 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -232,7 +232,8 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 	}
 
 	ret = btrfs_drop_extents(trans, root, inode, start,
-				 aligned_end, aligned_end, start, &hint_byte);
+				 aligned_end, aligned_end, start,
+				 &hint_byte, 1);
 	BUG_ON(ret);
 
 	if (isize > actual_end)
@@ -241,7 +242,7 @@ static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
 				   inline_len, compressed_size,
 				   compressed_pages);
 	BUG_ON(ret);
-	btrfs_drop_extent_cache(inode, start, aligned_end, 0);
+	btrfs_drop_extent_cache(inode, start, aligned_end - 1, 0);
 	return 0;
 }
 
@@ -1455,9 +1456,19 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	BUG_ON(!path);
 
 	path->leave_spinning = 1;
+
+	/*
+	 * we may be replacing one extent in the tree with another.
+	 * The new extent is pinned in the extent map, and we don't want
+	 * to drop it from the cache until it is completely in the btree.
+	 *
+	 * So, tell btrfs_drop_extents to leave this extent in the cache.
+	 * the caller is expected to unpin it and allow it to be merged
+	 * with the others.
+	 */
 	ret = btrfs_drop_extents(trans, root, inode, file_pos,
 				 file_pos + num_bytes, locked_end,
-				 file_pos, &hint);
+				 file_pos, &hint, 0);
 	BUG_ON(ret);
 
 	ins.objectid = inode->i_ino;
@@ -1485,7 +1496,6 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
 	btrfs_mark_buffer_dirty(leaf);
 
 	inode_add_bytes(inode, num_bytes);
-	btrfs_drop_extent_cache(inode, file_pos, file_pos + num_bytes - 1, 0);
 
 	ins.objectid = disk_bytenr;
 	ins.offset = disk_num_bytes;
@@ -1596,6 +1606,9 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
 						ordered_extent->len,
 						compressed, 0, 0,
 						BTRFS_FILE_EXTENT_REG);
+		unpin_extent_cache(&BTRFS_I(inode)->extent_tree,
+				   ordered_extent->file_offset,
+				   ordered_extent->len);
 		BUG_ON(ret);
 	}
 	unlock_extent(io_tree, ordered_extent->file_offset,
@@ -2940,7 +2953,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
 						 cur_offset,
 						 cur_offset + hole_size,
 						 block_end,
-						 cur_offset, &hint_byte);
+						 cur_offset, &hint_byte, 1);
 			if (err)
 				break;
 			err = btrfs_insert_file_extent(trans, root,
@@ -5086,6 +5099,8 @@ static int prealloc_file_range(struct btrfs_trans_handle *trans,
 						  0, 0, 0,
 						  BTRFS_FILE_EXTENT_PREALLOC);
 		BUG_ON(ret);
+		btrfs_drop_extent_cache(inode, cur_offset,
+					cur_offset + ins.offset -1, 0);
 		num_bytes -= ins.offset;
 		cur_offset += ins.offset;
 		alloc_hint = ins.objectid + ins.offset;
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index 9f4db848db1..e2d8e90259b 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -597,9 +597,8 @@ again:
 		clear_page_dirty_for_io(page);
 
 		btrfs_set_extent_delalloc(inode, page_start, page_end);
-
-		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
 		set_page_dirty(page);
+		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
 		unlock_page(page);
 		page_cache_release(page);
 		balance_dirty_pages_ratelimited_nr(inode->i_mapping, 1);
@@ -977,7 +976,7 @@ static long btrfs_ioctl_clone(struct file *file, unsigned long srcfd,
 
 	/* punch hole in destination first */
 	btrfs_drop_extents(trans, root, inode, off, off + len,
-			   off + len, 0, &hint_byte);
+			   off + len, 0, &hint_byte, 1);
 
 	/* clone data */
 	key.objectid = src->i_ino;
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index d91b0de7c50..8661a7381b3 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -534,7 +534,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans,
 	saved_nbytes = inode_get_bytes(inode);
 	/* drop any overlapping extents */
 	ret = btrfs_drop_extents(trans, root, inode,
-			 start, extent_end, extent_end, start, &alloc_hint);
+			 start, extent_end, extent_end, start, &alloc_hint, 1);
 	BUG_ON(ret);
 
 	if (found_type == BTRFS_FILE_EXTENT_REG ||
-- 
cgit v1.2.3-70-g09d2


From f0f37e2f77731b3473fa6bd5ee53255d9a9cdb40 Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Sun, 27 Sep 2009 22:29:37 +0400
Subject: const: mark struct vm_struct_operations

* mark struct vm_area_struct::vm_ops as const
* mark vm_ops in AGP code

But leave TTM code alone, something is fishy there with global vm_ops
being used.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/ia64/ia32/binfmt_elf32.c                |  4 ++--
 arch/powerpc/platforms/cell/spufs/file.c     | 14 +++++++-------
 arch/x86/pci/i386.c                          |  2 +-
 drivers/char/agp/agp.h                       |  2 +-
 drivers/char/agp/alpha-agp.c                 |  2 +-
 drivers/char/mem.c                           |  2 +-
 drivers/char/mspec.c                         |  2 +-
 drivers/gpu/drm/drm_vm.c                     |  8 ++++----
 drivers/gpu/drm/radeon/radeon_ttm.c          |  2 +-
 drivers/gpu/drm/ttm/ttm_bo_vm.c              |  2 +-
 drivers/ieee1394/dma.c                       |  2 +-
 drivers/infiniband/hw/ehca/ehca_uverbs.c     |  2 +-
 drivers/infiniband/hw/ipath/ipath_file_ops.c |  2 +-
 drivers/infiniband/hw/ipath/ipath_mmap.c     |  2 +-
 drivers/media/video/cafe_ccic.c              |  2 +-
 drivers/media/video/et61x251/et61x251_core.c |  2 +-
 drivers/media/video/gspca/gspca.c            |  2 +-
 drivers/media/video/meye.c                   |  2 +-
 drivers/media/video/sn9c102/sn9c102_core.c   |  2 +-
 drivers/media/video/stk-webcam.c             |  2 +-
 drivers/media/video/uvc/uvc_v4l2.c           |  2 +-
 drivers/media/video/videobuf-dma-contig.c    |  2 +-
 drivers/media/video/videobuf-dma-sg.c        |  2 +-
 drivers/media/video/videobuf-vmalloc.c       |  2 +-
 drivers/media/video/vino.c                   |  2 +-
 drivers/media/video/zc0301/zc0301_core.c     |  2 +-
 drivers/media/video/zoran/zoran_driver.c     |  2 +-
 drivers/misc/sgi-gru/grufile.c               |  2 +-
 drivers/misc/sgi-gru/grutables.h             |  2 +-
 drivers/scsi/sg.c                            |  2 +-
 drivers/uio/uio.c                            |  2 +-
 drivers/usb/mon/mon_bin.c                    |  2 +-
 drivers/video/fb_defio.c                     |  2 +-
 drivers/video/omap/dispc.c                   |  2 +-
 fs/btrfs/file.c                              |  2 +-
 fs/ext4/file.c                               |  2 +-
 fs/fuse/file.c                               |  2 +-
 fs/gfs2/file.c                               |  2 +-
 fs/ncpfs/mmap.c                              |  2 +-
 fs/nfs/file.c                                |  4 ++--
 fs/nilfs2/file.c                             |  2 +-
 fs/ocfs2/mmap.c                              |  2 +-
 fs/sysfs/bin.c                               |  4 ++--
 fs/ubifs/file.c                              |  2 +-
 fs/xfs/linux-2.6/xfs_file.c                  |  4 ++--
 include/linux/agp_backend.h                  |  2 +-
 include/linux/hugetlb.h                      |  2 +-
 include/linux/mm_types.h                     |  2 +-
 include/linux/ramfs.h                        |  2 +-
 ipc/shm.c                                    |  4 ++--
 kernel/perf_event.c                          |  2 +-
 kernel/relay.c                               |  2 +-
 mm/filemap.c                                 |  2 +-
 mm/filemap_xip.c                             |  2 +-
 mm/hugetlb.c                                 |  2 +-
 mm/mmap.c                                    |  2 +-
 mm/nommu.c                                   |  2 +-
 mm/shmem.c                                   |  4 ++--
 net/packet/af_packet.c                       |  2 +-
 sound/core/pcm_native.c                      |  8 ++++----
 sound/usb/usx2y/us122l.c                     |  2 +-
 sound/usb/usx2y/usX2Yhwdep.c                 |  2 +-
 sound/usb/usx2y/usx2yhwdeppcm.c              |  2 +-
 virt/kvm/kvm_main.c                          |  4 ++--
 64 files changed, 83 insertions(+), 83 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/arch/ia64/ia32/binfmt_elf32.c b/arch/ia64/ia32/binfmt_elf32.c
index f92bdaac897..c69552bf893 100644
--- a/arch/ia64/ia32/binfmt_elf32.c
+++ b/arch/ia64/ia32/binfmt_elf32.c
@@ -69,11 +69,11 @@ ia32_install_gate_page (struct vm_area_struct *vma, struct vm_fault *vmf)
 }
 
 
-static struct vm_operations_struct ia32_shared_page_vm_ops = {
+static const struct vm_operations_struct ia32_shared_page_vm_ops = {
 	.fault = ia32_install_shared_page
 };
 
-static struct vm_operations_struct ia32_gate_page_vm_ops = {
+static const struct vm_operations_struct ia32_gate_page_vm_ops = {
 	.fault = ia32_install_gate_page
 };
 
diff --git a/arch/powerpc/platforms/cell/spufs/file.c b/arch/powerpc/platforms/cell/spufs/file.c
index 8f079b865ad..96130944617 100644
--- a/arch/powerpc/platforms/cell/spufs/file.c
+++ b/arch/powerpc/platforms/cell/spufs/file.c
@@ -309,7 +309,7 @@ static int spufs_mem_mmap_access(struct vm_area_struct *vma,
 	return len;
 }
 
-static struct vm_operations_struct spufs_mem_mmap_vmops = {
+static const struct vm_operations_struct spufs_mem_mmap_vmops = {
 	.fault = spufs_mem_mmap_fault,
 	.access = spufs_mem_mmap_access,
 };
@@ -436,7 +436,7 @@ static int spufs_cntl_mmap_fault(struct vm_area_struct *vma,
 	return spufs_ps_fault(vma, vmf, 0x4000, SPUFS_CNTL_MAP_SIZE);
 }
 
-static struct vm_operations_struct spufs_cntl_mmap_vmops = {
+static const struct vm_operations_struct spufs_cntl_mmap_vmops = {
 	.fault = spufs_cntl_mmap_fault,
 };
 
@@ -1143,7 +1143,7 @@ spufs_signal1_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 #endif
 }
 
-static struct vm_operations_struct spufs_signal1_mmap_vmops = {
+static const struct vm_operations_struct spufs_signal1_mmap_vmops = {
 	.fault = spufs_signal1_mmap_fault,
 };
 
@@ -1279,7 +1279,7 @@ spufs_signal2_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 #endif
 }
 
-static struct vm_operations_struct spufs_signal2_mmap_vmops = {
+static const struct vm_operations_struct spufs_signal2_mmap_vmops = {
 	.fault = spufs_signal2_mmap_fault,
 };
 
@@ -1397,7 +1397,7 @@ spufs_mss_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return spufs_ps_fault(vma, vmf, 0x0000, SPUFS_MSS_MAP_SIZE);
 }
 
-static struct vm_operations_struct spufs_mss_mmap_vmops = {
+static const struct vm_operations_struct spufs_mss_mmap_vmops = {
 	.fault = spufs_mss_mmap_fault,
 };
 
@@ -1458,7 +1458,7 @@ spufs_psmap_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return spufs_ps_fault(vma, vmf, 0x0000, SPUFS_PS_MAP_SIZE);
 }
 
-static struct vm_operations_struct spufs_psmap_mmap_vmops = {
+static const struct vm_operations_struct spufs_psmap_mmap_vmops = {
 	.fault = spufs_psmap_mmap_fault,
 };
 
@@ -1517,7 +1517,7 @@ spufs_mfc_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return spufs_ps_fault(vma, vmf, 0x3000, SPUFS_MFC_MAP_SIZE);
 }
 
-static struct vm_operations_struct spufs_mfc_mmap_vmops = {
+static const struct vm_operations_struct spufs_mfc_mmap_vmops = {
 	.fault = spufs_mfc_mmap_fault,
 };
 
diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c
index 52e62e57fed..b22d13b0c71 100644
--- a/arch/x86/pci/i386.c
+++ b/arch/x86/pci/i386.c
@@ -266,7 +266,7 @@ void pcibios_set_master(struct pci_dev *dev)
 	pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat);
 }
 
-static struct vm_operations_struct pci_mmap_ops = {
+static const struct vm_operations_struct pci_mmap_ops = {
 	.access = generic_access_phys,
 };
 
diff --git a/drivers/char/agp/agp.h b/drivers/char/agp/agp.h
index d6f36c004d9..870f12cfed9 100644
--- a/drivers/char/agp/agp.h
+++ b/drivers/char/agp/agp.h
@@ -131,7 +131,7 @@ struct agp_bridge_driver {
 struct agp_bridge_data {
 	const struct agp_version *version;
 	const struct agp_bridge_driver *driver;
-	struct vm_operations_struct *vm_ops;
+	const struct vm_operations_struct *vm_ops;
 	void *previous_size;
 	void *current_size;
 	void *dev_private_data;
diff --git a/drivers/char/agp/alpha-agp.c b/drivers/char/agp/alpha-agp.c
index 5ea4da8e995..dd84af4d4f7 100644
--- a/drivers/char/agp/alpha-agp.c
+++ b/drivers/char/agp/alpha-agp.c
@@ -40,7 +40,7 @@ static struct aper_size_info_fixed alpha_core_agp_sizes[] =
 	{ 0, 0, 0 }, /* filled in by alpha_core_agp_setup */
 };
 
-struct vm_operations_struct alpha_core_agp_vm_ops = {
+static const struct vm_operations_struct alpha_core_agp_vm_ops = {
 	.fault = alpha_core_agp_vm_fault,
 };
 
diff --git a/drivers/char/mem.c b/drivers/char/mem.c
index 6c8b65d069e..a074fceb67d 100644
--- a/drivers/char/mem.c
+++ b/drivers/char/mem.c
@@ -301,7 +301,7 @@ static inline int private_mapping_ok(struct vm_area_struct *vma)
 }
 #endif
 
-static struct vm_operations_struct mmap_mem_ops = {
+static const struct vm_operations_struct mmap_mem_ops = {
 #ifdef CONFIG_HAVE_IOREMAP_PROT
 	.access = generic_access_phys
 #endif
diff --git a/drivers/char/mspec.c b/drivers/char/mspec.c
index 30f095a8c2d..1997270bb6f 100644
--- a/drivers/char/mspec.c
+++ b/drivers/char/mspec.c
@@ -239,7 +239,7 @@ mspec_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return VM_FAULT_NOPAGE;
 }
 
-static struct vm_operations_struct mspec_vm_ops = {
+static const struct vm_operations_struct mspec_vm_ops = {
 	.open = mspec_open,
 	.close = mspec_close,
 	.fault = mspec_fault,
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 7e1fbe5d477..4ac900f4647 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -369,28 +369,28 @@ static int drm_vm_sg_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 }
 
 /** AGP virtual memory operations */
-static struct vm_operations_struct drm_vm_ops = {
+static const struct vm_operations_struct drm_vm_ops = {
 	.fault = drm_vm_fault,
 	.open = drm_vm_open,
 	.close = drm_vm_close,
 };
 
 /** Shared virtual memory operations */
-static struct vm_operations_struct drm_vm_shm_ops = {
+static const struct vm_operations_struct drm_vm_shm_ops = {
 	.fault = drm_vm_shm_fault,
 	.open = drm_vm_open,
 	.close = drm_vm_shm_close,
 };
 
 /** DMA virtual memory operations */
-static struct vm_operations_struct drm_vm_dma_ops = {
+static const struct vm_operations_struct drm_vm_dma_ops = {
 	.fault = drm_vm_dma_fault,
 	.open = drm_vm_open,
 	.close = drm_vm_close,
 };
 
 /** Scatter-gather virtual memory operations */
-static struct vm_operations_struct drm_vm_sg_ops = {
+static const struct vm_operations_struct drm_vm_sg_ops = {
 	.fault = drm_vm_sg_fault,
 	.open = drm_vm_open,
 	.close = drm_vm_close,
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index acd889c9454..5b1cf04a011 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -530,7 +530,7 @@ void radeon_ttm_fini(struct radeon_device *rdev)
 }
 
 static struct vm_operations_struct radeon_ttm_vm_ops;
-static struct vm_operations_struct *ttm_vm_ops = NULL;
+static const struct vm_operations_struct *ttm_vm_ops = NULL;
 
 static int radeon_ttm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 33de7637c0c..1c040d04033 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -228,7 +228,7 @@ static void ttm_bo_vm_close(struct vm_area_struct *vma)
 	vma->vm_private_data = NULL;
 }
 
-static struct vm_operations_struct ttm_bo_vm_ops = {
+static const struct vm_operations_struct ttm_bo_vm_ops = {
 	.fault = ttm_bo_vm_fault,
 	.open = ttm_bo_vm_open,
 	.close = ttm_bo_vm_close
diff --git a/drivers/ieee1394/dma.c b/drivers/ieee1394/dma.c
index 1aba8c13fe8..8e7e3344c4b 100644
--- a/drivers/ieee1394/dma.c
+++ b/drivers/ieee1394/dma.c
@@ -247,7 +247,7 @@ static int dma_region_pagefault(struct vm_area_struct *vma,
 	return 0;
 }
 
-static struct vm_operations_struct dma_region_vm_ops = {
+static const struct vm_operations_struct dma_region_vm_ops = {
 	.fault = dma_region_pagefault,
 };
 
diff --git a/drivers/infiniband/hw/ehca/ehca_uverbs.c b/drivers/infiniband/hw/ehca/ehca_uverbs.c
index 3cb688d2913..f1565cae8ec 100644
--- a/drivers/infiniband/hw/ehca/ehca_uverbs.c
+++ b/drivers/infiniband/hw/ehca/ehca_uverbs.c
@@ -95,7 +95,7 @@ static void ehca_mm_close(struct vm_area_struct *vma)
 		     vma->vm_start, vma->vm_end, *count);
 }
 
-static struct vm_operations_struct vm_ops = {
+static const struct vm_operations_struct vm_ops = {
 	.open =	ehca_mm_open,
 	.close = ehca_mm_close,
 };
diff --git a/drivers/infiniband/hw/ipath/ipath_file_ops.c b/drivers/infiniband/hw/ipath/ipath_file_ops.c
index 38a28700661..40dbe54056c 100644
--- a/drivers/infiniband/hw/ipath/ipath_file_ops.c
+++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c
@@ -1151,7 +1151,7 @@ static int ipath_file_vma_fault(struct vm_area_struct *vma,
 	return 0;
 }
 
-static struct vm_operations_struct ipath_file_vm_ops = {
+static const struct vm_operations_struct ipath_file_vm_ops = {
 	.fault = ipath_file_vma_fault,
 };
 
diff --git a/drivers/infiniband/hw/ipath/ipath_mmap.c b/drivers/infiniband/hw/ipath/ipath_mmap.c
index fa830e22002..b28865faf43 100644
--- a/drivers/infiniband/hw/ipath/ipath_mmap.c
+++ b/drivers/infiniband/hw/ipath/ipath_mmap.c
@@ -74,7 +74,7 @@ static void ipath_vma_close(struct vm_area_struct *vma)
 	kref_put(&ip->ref, ipath_release_mmap_info);
 }
 
-static struct vm_operations_struct ipath_vm_ops = {
+static const struct vm_operations_struct ipath_vm_ops = {
 	.open =     ipath_vma_open,
 	.close =    ipath_vma_close,
 };
diff --git a/drivers/media/video/cafe_ccic.c b/drivers/media/video/cafe_ccic.c
index 657c481d255..10230cb3d21 100644
--- a/drivers/media/video/cafe_ccic.c
+++ b/drivers/media/video/cafe_ccic.c
@@ -1325,7 +1325,7 @@ static void cafe_v4l_vm_close(struct vm_area_struct *vma)
 	mutex_unlock(&sbuf->cam->s_mutex);
 }
 
-static struct vm_operations_struct cafe_v4l_vm_ops = {
+static const struct vm_operations_struct cafe_v4l_vm_ops = {
 	.open = cafe_v4l_vm_open,
 	.close = cafe_v4l_vm_close
 };
diff --git a/drivers/media/video/et61x251/et61x251_core.c b/drivers/media/video/et61x251/et61x251_core.c
index 74092f436be..88987a57cf7 100644
--- a/drivers/media/video/et61x251/et61x251_core.c
+++ b/drivers/media/video/et61x251/et61x251_core.c
@@ -1496,7 +1496,7 @@ static void et61x251_vm_close(struct vm_area_struct* vma)
 }
 
 
-static struct vm_operations_struct et61x251_vm_ops = {
+static const struct vm_operations_struct et61x251_vm_ops = {
 	.open = et61x251_vm_open,
 	.close = et61x251_vm_close,
 };
diff --git a/drivers/media/video/gspca/gspca.c b/drivers/media/video/gspca/gspca.c
index cf6540da1e4..23d3fb77691 100644
--- a/drivers/media/video/gspca/gspca.c
+++ b/drivers/media/video/gspca/gspca.c
@@ -99,7 +99,7 @@ static void gspca_vm_close(struct vm_area_struct *vma)
 		frame->v4l2_buf.flags &= ~V4L2_BUF_FLAG_MAPPED;
 }
 
-static struct vm_operations_struct gspca_vm_ops = {
+static const struct vm_operations_struct gspca_vm_ops = {
 	.open		= gspca_vm_open,
 	.close		= gspca_vm_close,
 };
diff --git a/drivers/media/video/meye.c b/drivers/media/video/meye.c
index d0765bed79c..4b1bc05a462 100644
--- a/drivers/media/video/meye.c
+++ b/drivers/media/video/meye.c
@@ -1589,7 +1589,7 @@ static void meye_vm_close(struct vm_area_struct *vma)
 	meye.vma_use_count[idx]--;
 }
 
-static struct vm_operations_struct meye_vm_ops = {
+static const struct vm_operations_struct meye_vm_ops = {
 	.open		= meye_vm_open,
 	.close		= meye_vm_close,
 };
diff --git a/drivers/media/video/sn9c102/sn9c102_core.c b/drivers/media/video/sn9c102/sn9c102_core.c
index 9d84c94e8a4..4a7711c3e74 100644
--- a/drivers/media/video/sn9c102/sn9c102_core.c
+++ b/drivers/media/video/sn9c102/sn9c102_core.c
@@ -2077,7 +2077,7 @@ static void sn9c102_vm_close(struct vm_area_struct* vma)
 }
 
 
-static struct vm_operations_struct sn9c102_vm_ops = {
+static const struct vm_operations_struct sn9c102_vm_ops = {
 	.open = sn9c102_vm_open,
 	.close = sn9c102_vm_close,
 };
diff --git a/drivers/media/video/stk-webcam.c b/drivers/media/video/stk-webcam.c
index 0b996ea4134..6b41865f42b 100644
--- a/drivers/media/video/stk-webcam.c
+++ b/drivers/media/video/stk-webcam.c
@@ -790,7 +790,7 @@ static void stk_v4l_vm_close(struct vm_area_struct *vma)
 	if (sbuf->mapcount == 0)
 		sbuf->v4lbuf.flags &= ~V4L2_BUF_FLAG_MAPPED;
 }
-static struct vm_operations_struct stk_v4l_vm_ops = {
+static const struct vm_operations_struct stk_v4l_vm_ops = {
 	.open = stk_v4l_vm_open,
 	.close = stk_v4l_vm_close
 };
diff --git a/drivers/media/video/uvc/uvc_v4l2.c b/drivers/media/video/uvc/uvc_v4l2.c
index 9e7351569b5..a2bdd806efa 100644
--- a/drivers/media/video/uvc/uvc_v4l2.c
+++ b/drivers/media/video/uvc/uvc_v4l2.c
@@ -1069,7 +1069,7 @@ static void uvc_vm_close(struct vm_area_struct *vma)
 	buffer->vma_use_count--;
 }
 
-static struct vm_operations_struct uvc_vm_ops = {
+static const struct vm_operations_struct uvc_vm_ops = {
 	.open		= uvc_vm_open,
 	.close		= uvc_vm_close,
 };
diff --git a/drivers/media/video/videobuf-dma-contig.c b/drivers/media/video/videobuf-dma-contig.c
index d09ce83a942..635ffc7b039 100644
--- a/drivers/media/video/videobuf-dma-contig.c
+++ b/drivers/media/video/videobuf-dma-contig.c
@@ -105,7 +105,7 @@ static void videobuf_vm_close(struct vm_area_struct *vma)
 	}
 }
 
-static struct vm_operations_struct videobuf_vm_ops = {
+static const struct vm_operations_struct videobuf_vm_ops = {
 	.open     = videobuf_vm_open,
 	.close    = videobuf_vm_close,
 };
diff --git a/drivers/media/video/videobuf-dma-sg.c b/drivers/media/video/videobuf-dma-sg.c
index a8dd22ace3f..53cdd67cebe 100644
--- a/drivers/media/video/videobuf-dma-sg.c
+++ b/drivers/media/video/videobuf-dma-sg.c
@@ -394,7 +394,7 @@ videobuf_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct videobuf_vm_ops =
+static const struct vm_operations_struct videobuf_vm_ops =
 {
 	.open     = videobuf_vm_open,
 	.close    = videobuf_vm_close,
diff --git a/drivers/media/video/videobuf-vmalloc.c b/drivers/media/video/videobuf-vmalloc.c
index 30ae30f99cc..35f3900c563 100644
--- a/drivers/media/video/videobuf-vmalloc.c
+++ b/drivers/media/video/videobuf-vmalloc.c
@@ -116,7 +116,7 @@ static void videobuf_vm_close(struct vm_area_struct *vma)
 	return;
 }
 
-static struct vm_operations_struct videobuf_vm_ops =
+static const struct vm_operations_struct videobuf_vm_ops =
 {
 	.open     = videobuf_vm_open,
 	.close    = videobuf_vm_close,
diff --git a/drivers/media/video/vino.c b/drivers/media/video/vino.c
index cd6a3446ab7..b034a81d2b1 100644
--- a/drivers/media/video/vino.c
+++ b/drivers/media/video/vino.c
@@ -3857,7 +3857,7 @@ static void vino_vm_close(struct vm_area_struct *vma)
 	dprintk("vino_vm_close(): count = %d\n", fb->map_count);
 }
 
-static struct vm_operations_struct vino_vm_ops = {
+static const struct vm_operations_struct vino_vm_ops = {
 	.open	= vino_vm_open,
 	.close	= vino_vm_close,
 };
diff --git a/drivers/media/video/zc0301/zc0301_core.c b/drivers/media/video/zc0301/zc0301_core.c
index b3c6436b33b..312a71336fd 100644
--- a/drivers/media/video/zc0301/zc0301_core.c
+++ b/drivers/media/video/zc0301/zc0301_core.c
@@ -935,7 +935,7 @@ static void zc0301_vm_close(struct vm_area_struct* vma)
 }
 
 
-static struct vm_operations_struct zc0301_vm_ops = {
+static const struct vm_operations_struct zc0301_vm_ops = {
 	.open = zc0301_vm_open,
 	.close = zc0301_vm_close,
 };
diff --git a/drivers/media/video/zoran/zoran_driver.c b/drivers/media/video/zoran/zoran_driver.c
index bcdefb1bcb3..47137deafcf 100644
--- a/drivers/media/video/zoran/zoran_driver.c
+++ b/drivers/media/video/zoran/zoran_driver.c
@@ -3172,7 +3172,7 @@ zoran_vm_close (struct vm_area_struct *vma)
 	mutex_unlock(&zr->resource_lock);
 }
 
-static struct vm_operations_struct zoran_vm_ops = {
+static const struct vm_operations_struct zoran_vm_ops = {
 	.open = zoran_vm_open,
 	.close = zoran_vm_close,
 };
diff --git a/drivers/misc/sgi-gru/grufile.c b/drivers/misc/sgi-gru/grufile.c
index aed609832bc..300e7ba391a 100644
--- a/drivers/misc/sgi-gru/grufile.c
+++ b/drivers/misc/sgi-gru/grufile.c
@@ -438,7 +438,7 @@ static struct miscdevice gru_miscdev = {
 	.fops		= &gru_fops,
 };
 
-struct vm_operations_struct gru_vm_ops = {
+const struct vm_operations_struct gru_vm_ops = {
 	.close		= gru_vma_close,
 	.fault		= gru_fault,
 };
diff --git a/drivers/misc/sgi-gru/grutables.h b/drivers/misc/sgi-gru/grutables.h
index 34ab3d45391..46990bcfa53 100644
--- a/drivers/misc/sgi-gru/grutables.h
+++ b/drivers/misc/sgi-gru/grutables.h
@@ -624,7 +624,7 @@ static inline int is_kernel_context(struct gru_thread_state *gts)
  */
 struct gru_unload_context_req;
 
-extern struct vm_operations_struct gru_vm_ops;
+extern const struct vm_operations_struct gru_vm_ops;
 extern struct device *grudev;
 
 extern struct gru_vma_data *gru_alloc_vma_data(struct vm_area_struct *vma,
diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c
index 848b5946685..0cb049f5cc5 100644
--- a/drivers/scsi/sg.c
+++ b/drivers/scsi/sg.c
@@ -1185,7 +1185,7 @@ sg_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return VM_FAULT_SIGBUS;
 }
 
-static struct vm_operations_struct sg_mmap_vm_ops = {
+static const struct vm_operations_struct sg_mmap_vm_ops = {
 	.fault = sg_vma_fault,
 };
 
diff --git a/drivers/uio/uio.c b/drivers/uio/uio.c
index 03efb065455..a9d70704720 100644
--- a/drivers/uio/uio.c
+++ b/drivers/uio/uio.c
@@ -658,7 +658,7 @@ static int uio_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct uio_vm_ops = {
+static const struct vm_operations_struct uio_vm_ops = {
 	.open = uio_vma_open,
 	.close = uio_vma_close,
 	.fault = uio_vma_fault,
diff --git a/drivers/usb/mon/mon_bin.c b/drivers/usb/mon/mon_bin.c
index dfdc43e2e00..9ed3e741bee 100644
--- a/drivers/usb/mon/mon_bin.c
+++ b/drivers/usb/mon/mon_bin.c
@@ -1174,7 +1174,7 @@ static int mon_bin_vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct mon_bin_vm_ops = {
+static const struct vm_operations_struct mon_bin_vm_ops = {
 	.open =     mon_bin_vma_open,
 	.close =    mon_bin_vma_close,
 	.fault =    mon_bin_vma_fault,
diff --git a/drivers/video/fb_defio.c b/drivers/video/fb_defio.c
index 0a7a6679ee6..c27ab1ed960 100644
--- a/drivers/video/fb_defio.c
+++ b/drivers/video/fb_defio.c
@@ -125,7 +125,7 @@ page_already_added:
 	return 0;
 }
 
-static struct vm_operations_struct fb_deferred_io_vm_ops = {
+static const struct vm_operations_struct fb_deferred_io_vm_ops = {
 	.fault		= fb_deferred_io_fault,
 	.page_mkwrite	= fb_deferred_io_mkwrite,
 };
diff --git a/drivers/video/omap/dispc.c b/drivers/video/omap/dispc.c
index 80a11d078df..f16e4215422 100644
--- a/drivers/video/omap/dispc.c
+++ b/drivers/video/omap/dispc.c
@@ -1035,7 +1035,7 @@ static void mmap_user_close(struct vm_area_struct *vma)
 	atomic_dec(&dispc.map_count[plane]);
 }
 
-static struct vm_operations_struct mmap_user_ops = {
+static const struct vm_operations_struct mmap_user_ops = {
 	.open = mmap_user_open,
 	.close = mmap_user_close,
 };
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 571ad3c13b4..a3492a3ad96 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1184,7 +1184,7 @@ out:
 	return ret > 0 ? EIO : ret;
 }
 
-static struct vm_operations_struct btrfs_file_vm_ops = {
+static const struct vm_operations_struct btrfs_file_vm_ops = {
 	.fault		= filemap_fault,
 	.page_mkwrite	= btrfs_page_mkwrite,
 };
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 5ca3eca70a1..9630583cef2 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -81,7 +81,7 @@ ext4_file_write(struct kiocb *iocb, const struct iovec *iov,
 	return generic_file_aio_write(iocb, iov, nr_segs, pos);
 }
 
-static struct vm_operations_struct ext4_file_vm_ops = {
+static const struct vm_operations_struct ext4_file_vm_ops = {
 	.fault		= filemap_fault,
 	.page_mkwrite   = ext4_page_mkwrite,
 };
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index cbc464043b6..a3492f7d207 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -1313,7 +1313,7 @@ static int fuse_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct fuse_file_vm_ops = {
+static const struct vm_operations_struct fuse_file_vm_ops = {
 	.close		= fuse_vma_close,
 	.fault		= filemap_fault,
 	.page_mkwrite	= fuse_page_mkwrite,
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index 166f38fbd24..4eb308aa323 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -418,7 +418,7 @@ out:
 	return ret;
 }
 
-static struct vm_operations_struct gfs2_vm_ops = {
+static const struct vm_operations_struct gfs2_vm_ops = {
 	.fault = filemap_fault,
 	.page_mkwrite = gfs2_page_mkwrite,
 };
diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c
index 5d8dcb9ee32..15458decdb8 100644
--- a/fs/ncpfs/mmap.c
+++ b/fs/ncpfs/mmap.c
@@ -95,7 +95,7 @@ static int ncp_file_mmap_fault(struct vm_area_struct *area,
 	return VM_FAULT_MAJOR;
 }
 
-static struct vm_operations_struct ncp_file_mmap =
+static const struct vm_operations_struct ncp_file_mmap =
 {
 	.fault = ncp_file_mmap_fault,
 };
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 86d6b4db109..f5fdd39e037 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -59,7 +59,7 @@ static int nfs_lock(struct file *filp, int cmd, struct file_lock *fl);
 static int nfs_flock(struct file *filp, int cmd, struct file_lock *fl);
 static int nfs_setlease(struct file *file, long arg, struct file_lock **fl);
 
-static struct vm_operations_struct nfs_file_vm_ops;
+static const struct vm_operations_struct nfs_file_vm_ops;
 
 const struct file_operations nfs_file_operations = {
 	.llseek		= nfs_file_llseek,
@@ -572,7 +572,7 @@ out_unlock:
 	return VM_FAULT_SIGBUS;
 }
 
-static struct vm_operations_struct nfs_file_vm_ops = {
+static const struct vm_operations_struct nfs_file_vm_ops = {
 	.fault = filemap_fault,
 	.page_mkwrite = nfs_vm_page_mkwrite,
 };
diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c
index fc8278c77cd..7d7b4983dee 100644
--- a/fs/nilfs2/file.c
+++ b/fs/nilfs2/file.c
@@ -117,7 +117,7 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-struct vm_operations_struct nilfs_file_vm_ops = {
+static const struct vm_operations_struct nilfs_file_vm_ops = {
 	.fault		= filemap_fault,
 	.page_mkwrite	= nilfs_page_mkwrite,
 };
diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c
index b606496b72e..39737613424 100644
--- a/fs/ocfs2/mmap.c
+++ b/fs/ocfs2/mmap.c
@@ -202,7 +202,7 @@ out:
 	return ret;
 }
 
-static struct vm_operations_struct ocfs2_file_vm_ops = {
+static const struct vm_operations_struct ocfs2_file_vm_ops = {
 	.fault		= ocfs2_fault,
 	.page_mkwrite	= ocfs2_page_mkwrite,
 };
diff --git a/fs/sysfs/bin.c b/fs/sysfs/bin.c
index 2524714bece..60c702bc10a 100644
--- a/fs/sysfs/bin.c
+++ b/fs/sysfs/bin.c
@@ -40,7 +40,7 @@ struct bin_buffer {
 	struct mutex			mutex;
 	void				*buffer;
 	int				mmapped;
-	struct vm_operations_struct 	*vm_ops;
+	const struct vm_operations_struct *vm_ops;
 	struct file			*file;
 	struct hlist_node		list;
 };
@@ -331,7 +331,7 @@ static int bin_migrate(struct vm_area_struct *vma, const nodemask_t *from,
 }
 #endif
 
-static struct vm_operations_struct bin_vm_ops = {
+static const struct vm_operations_struct bin_vm_ops = {
 	.open		= bin_vma_open,
 	.close		= bin_vma_close,
 	.fault		= bin_fault,
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 2e6481a7701..1009adc8d60 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -1534,7 +1534,7 @@ out_unlock:
 	return err;
 }
 
-static struct vm_operations_struct ubifs_file_vm_ops = {
+static const struct vm_operations_struct ubifs_file_vm_ops = {
 	.fault        = filemap_fault,
 	.page_mkwrite = ubifs_vm_page_mkwrite,
 };
diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c
index 988d8f87bc0..629370974e5 100644
--- a/fs/xfs/linux-2.6/xfs_file.c
+++ b/fs/xfs/linux-2.6/xfs_file.c
@@ -42,7 +42,7 @@
 
 #include <linux/dcache.h>
 
-static struct vm_operations_struct xfs_file_vm_ops;
+static const struct vm_operations_struct xfs_file_vm_ops;
 
 STATIC ssize_t
 xfs_file_aio_read(
@@ -280,7 +280,7 @@ const struct file_operations xfs_dir_file_operations = {
 	.fsync		= xfs_file_fsync,
 };
 
-static struct vm_operations_struct xfs_file_vm_ops = {
+static const struct vm_operations_struct xfs_file_vm_ops = {
 	.fault		= filemap_fault,
 	.page_mkwrite	= xfs_vm_page_mkwrite,
 };
diff --git a/include/linux/agp_backend.h b/include/linux/agp_backend.h
index 880130f7311..9101ed64f80 100644
--- a/include/linux/agp_backend.h
+++ b/include/linux/agp_backend.h
@@ -53,7 +53,7 @@ struct agp_kern_info {
 	int current_memory;
 	bool cant_use_aperture;
 	unsigned long page_mask;
-	struct vm_operations_struct *vm_ops;
+	const struct vm_operations_struct *vm_ops;
 };
 
 /*
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 16937995abd..41a59afc70f 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -163,7 +163,7 @@ static inline struct hugetlbfs_sb_info *HUGETLBFS_SB(struct super_block *sb)
 }
 
 extern const struct file_operations hugetlbfs_file_operations;
-extern struct vm_operations_struct hugetlb_vm_ops;
+extern const struct vm_operations_struct hugetlb_vm_ops;
 struct file *hugetlb_file_setup(const char *name, size_t size, int acct,
 				struct user_struct **user, int creat_flags);
 int hugetlb_get_quota(struct address_space *mapping, long delta);
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index 21d6aa45206..84a524afb3d 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -171,7 +171,7 @@ struct vm_area_struct {
 	struct anon_vma *anon_vma;	/* Serialized by page_table_lock */
 
 	/* Function pointers to deal with this struct. */
-	struct vm_operations_struct * vm_ops;
+	const struct vm_operations_struct *vm_ops;
 
 	/* Information about our backing store: */
 	unsigned long vm_pgoff;		/* Offset (within vm_file) in PAGE_SIZE
diff --git a/include/linux/ramfs.h b/include/linux/ramfs.h
index 37aaf2b3986..4e768dda87b 100644
--- a/include/linux/ramfs.h
+++ b/include/linux/ramfs.h
@@ -17,7 +17,7 @@ extern int ramfs_nommu_mmap(struct file *file, struct vm_area_struct *vma);
 #endif
 
 extern const struct file_operations ramfs_file_operations;
-extern struct vm_operations_struct generic_file_vm_ops;
+extern const struct vm_operations_struct generic_file_vm_ops;
 extern int __init init_rootfs(void);
 
 #endif
diff --git a/ipc/shm.c b/ipc/shm.c
index 9eb1488b543..464694e0aa4 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -55,7 +55,7 @@ struct shm_file_data {
 #define shm_file_data(file) (*((struct shm_file_data **)&(file)->private_data))
 
 static const struct file_operations shm_file_operations;
-static struct vm_operations_struct shm_vm_ops;
+static const struct vm_operations_struct shm_vm_ops;
 
 #define shm_ids(ns)	((ns)->ids[IPC_SHM_IDS])
 
@@ -312,7 +312,7 @@ static const struct file_operations shm_file_operations = {
 	.get_unmapped_area	= shm_get_unmapped_area,
 };
 
-static struct vm_operations_struct shm_vm_ops = {
+static const struct vm_operations_struct shm_vm_ops = {
 	.open	= shm_open,	/* callback for a new vm-area open */
 	.close	= shm_close,	/* callback for when the vm-area is released */
 	.fault	= shm_fault,
diff --git a/kernel/perf_event.c b/kernel/perf_event.c
index 76ac4db405e..0f86feb6db0 100644
--- a/kernel/perf_event.c
+++ b/kernel/perf_event.c
@@ -2253,7 +2253,7 @@ static void perf_mmap_close(struct vm_area_struct *vma)
 	}
 }
 
-static struct vm_operations_struct perf_mmap_vmops = {
+static const struct vm_operations_struct perf_mmap_vmops = {
 	.open		= perf_mmap_open,
 	.close		= perf_mmap_close,
 	.fault		= perf_mmap_fault,
diff --git a/kernel/relay.c b/kernel/relay.c
index bc188549788..760c26209a3 100644
--- a/kernel/relay.c
+++ b/kernel/relay.c
@@ -60,7 +60,7 @@ static int relay_buf_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 /*
  * vm_ops for relay file mappings.
  */
-static struct vm_operations_struct relay_file_mmap_ops = {
+static const struct vm_operations_struct relay_file_mmap_ops = {
 	.fault = relay_buf_fault,
 	.close = relay_file_mmap_close,
 };
diff --git a/mm/filemap.c b/mm/filemap.c
index 6c84e598b4a..ef169f37156 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -1611,7 +1611,7 @@ page_not_uptodate:
 }
 EXPORT_SYMBOL(filemap_fault);
 
-struct vm_operations_struct generic_file_vm_ops = {
+const struct vm_operations_struct generic_file_vm_ops = {
 	.fault		= filemap_fault,
 };
 
diff --git a/mm/filemap_xip.c b/mm/filemap_xip.c
index 427dfe3ce78..1888b2d71bb 100644
--- a/mm/filemap_xip.c
+++ b/mm/filemap_xip.c
@@ -296,7 +296,7 @@ out:
 	}
 }
 
-static struct vm_operations_struct xip_file_vm_ops = {
+static const struct vm_operations_struct xip_file_vm_ops = {
 	.fault	= xip_file_fault,
 };
 
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 6f048fcc749..5d7601b0287 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1721,7 +1721,7 @@ static int hugetlb_vm_op_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-struct vm_operations_struct hugetlb_vm_ops = {
+const struct vm_operations_struct hugetlb_vm_ops = {
 	.fault = hugetlb_vm_op_fault,
 	.open = hugetlb_vm_op_open,
 	.close = hugetlb_vm_op_close,
diff --git a/mm/mmap.c b/mm/mmap.c
index 21d4029a07b..73f5e4b6401 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -2282,7 +2282,7 @@ static void special_mapping_close(struct vm_area_struct *vma)
 {
 }
 
-static struct vm_operations_struct special_mapping_vmops = {
+static const struct vm_operations_struct special_mapping_vmops = {
 	.close = special_mapping_close,
 	.fault = special_mapping_fault,
 };
diff --git a/mm/nommu.c b/mm/nommu.c
index c73aa4753d7..5189b5aed8c 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -79,7 +79,7 @@ static struct kmem_cache *vm_region_jar;
 struct rb_root nommu_region_tree = RB_ROOT;
 DECLARE_RWSEM(nommu_region_sem);
 
-struct vm_operations_struct generic_file_vm_ops = {
+const struct vm_operations_struct generic_file_vm_ops = {
 };
 
 /*
diff --git a/mm/shmem.c b/mm/shmem.c
index ccf446a9faa..356dd99566e 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -218,7 +218,7 @@ static const struct file_operations shmem_file_operations;
 static const struct inode_operations shmem_inode_operations;
 static const struct inode_operations shmem_dir_inode_operations;
 static const struct inode_operations shmem_special_inode_operations;
-static struct vm_operations_struct shmem_vm_ops;
+static const struct vm_operations_struct shmem_vm_ops;
 
 static struct backing_dev_info shmem_backing_dev_info  __read_mostly = {
 	.ra_pages	= 0,	/* No readahead */
@@ -2498,7 +2498,7 @@ static const struct super_operations shmem_ops = {
 	.put_super	= shmem_put_super,
 };
 
-static struct vm_operations_struct shmem_vm_ops = {
+static const struct vm_operations_struct shmem_vm_ops = {
 	.fault		= shmem_fault,
 #ifdef CONFIG_NUMA
 	.set_policy     = shmem_set_policy,
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index d3d52c66cdc..103d5611b81 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -2084,7 +2084,7 @@ static void packet_mm_close(struct vm_area_struct *vma)
 		atomic_dec(&pkt_sk(sk)->mapped);
 }
 
-static struct vm_operations_struct packet_mmap_ops = {
+static const struct vm_operations_struct packet_mmap_ops = {
 	.open	=	packet_mm_open,
 	.close	=	packet_mm_close,
 };
diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c
index 561d6d95a2d..ab73edf2c89 100644
--- a/sound/core/pcm_native.c
+++ b/sound/core/pcm_native.c
@@ -2985,7 +2985,7 @@ static int snd_pcm_mmap_status_fault(struct vm_area_struct *area,
 	return 0;
 }
 
-static struct vm_operations_struct snd_pcm_vm_ops_status =
+static const struct vm_operations_struct snd_pcm_vm_ops_status =
 {
 	.fault =	snd_pcm_mmap_status_fault,
 };
@@ -3024,7 +3024,7 @@ static int snd_pcm_mmap_control_fault(struct vm_area_struct *area,
 	return 0;
 }
 
-static struct vm_operations_struct snd_pcm_vm_ops_control =
+static const struct vm_operations_struct snd_pcm_vm_ops_control =
 {
 	.fault =	snd_pcm_mmap_control_fault,
 };
@@ -3094,7 +3094,7 @@ static int snd_pcm_mmap_data_fault(struct vm_area_struct *area,
 	return 0;
 }
 
-static struct vm_operations_struct snd_pcm_vm_ops_data =
+static const struct vm_operations_struct snd_pcm_vm_ops_data =
 {
 	.open =		snd_pcm_mmap_data_open,
 	.close =	snd_pcm_mmap_data_close,
@@ -3118,7 +3118,7 @@ static int snd_pcm_default_mmap(struct snd_pcm_substream *substream,
  * mmap the DMA buffer on I/O memory area
  */
 #if SNDRV_PCM_INFO_MMAP_IOMEM
-static struct vm_operations_struct snd_pcm_vm_ops_data_mmio =
+static const struct vm_operations_struct snd_pcm_vm_ops_data_mmio =
 {
 	.open =		snd_pcm_mmap_data_open,
 	.close =	snd_pcm_mmap_data_close,
diff --git a/sound/usb/usx2y/us122l.c b/sound/usb/usx2y/us122l.c
index fd44946ce4b..99f33766cd5 100644
--- a/sound/usb/usx2y/us122l.c
+++ b/sound/usb/usx2y/us122l.c
@@ -154,7 +154,7 @@ static void usb_stream_hwdep_vm_close(struct vm_area_struct *area)
 	snd_printdd(KERN_DEBUG "%i\n", atomic_read(&us122l->mmap_count));
 }
 
-static struct vm_operations_struct usb_stream_hwdep_vm_ops = {
+static const struct vm_operations_struct usb_stream_hwdep_vm_ops = {
 	.open = usb_stream_hwdep_vm_open,
 	.fault = usb_stream_hwdep_vm_fault,
 	.close = usb_stream_hwdep_vm_close,
diff --git a/sound/usb/usx2y/usX2Yhwdep.c b/sound/usb/usx2y/usX2Yhwdep.c
index f3d8f71265d..52e04b2f35d 100644
--- a/sound/usb/usx2y/usX2Yhwdep.c
+++ b/sound/usb/usx2y/usX2Yhwdep.c
@@ -53,7 +53,7 @@ static int snd_us428ctls_vm_fault(struct vm_area_struct *area,
 	return 0;
 }
 
-static struct vm_operations_struct us428ctls_vm_ops = {
+static const struct vm_operations_struct us428ctls_vm_ops = {
 	.fault = snd_us428ctls_vm_fault,
 };
 
diff --git a/sound/usb/usx2y/usx2yhwdeppcm.c b/sound/usb/usx2y/usx2yhwdeppcm.c
index 117946f2deb..4b2304c2e02 100644
--- a/sound/usb/usx2y/usx2yhwdeppcm.c
+++ b/sound/usb/usx2y/usx2yhwdeppcm.c
@@ -697,7 +697,7 @@ static int snd_usX2Y_hwdep_pcm_vm_fault(struct vm_area_struct *area,
 }
 
 
-static struct vm_operations_struct snd_usX2Y_hwdep_pcm_vm_ops = {
+static const struct vm_operations_struct snd_usX2Y_hwdep_pcm_vm_ops = {
 	.open = snd_usX2Y_hwdep_pcm_vm_open,
 	.close = snd_usX2Y_hwdep_pcm_vm_close,
 	.fault = snd_usX2Y_hwdep_pcm_vm_fault,
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 034a798b043..b5e7e3f1183 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1713,7 +1713,7 @@ static int kvm_vcpu_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct kvm_vcpu_vm_ops = {
+static const struct vm_operations_struct kvm_vcpu_vm_ops = {
 	.fault = kvm_vcpu_fault,
 };
 
@@ -2317,7 +2317,7 @@ static int kvm_vm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	return 0;
 }
 
-static struct vm_operations_struct kvm_vm_vm_ops = {
+static const struct vm_operations_struct kvm_vm_vm_ops = {
 	.fault = kvm_vm_fault,
 };
 
-- 
cgit v1.2.3-70-g09d2


From 9ed74f2dba6ebf9f30b80554290bfc73cc3ef083 Mon Sep 17 00:00:00 2001
From: Josef Bacik <josef@redhat.com>
Date: Fri, 11 Sep 2009 16:12:44 -0400
Subject: Btrfs: proper -ENOSPC handling

At the start of a transaction we do a btrfs_reserve_metadata_space() and
specify how many items we plan on modifying.  Then once we've done our
modifications and such, just call btrfs_unreserve_metadata_space() for
the same number of items we reserved.

For keeping track of metadata needed for data I've had to add an extent_io op
for when we merge extents.  This lets us track space properly when we are doing
sequential writes, so we don't end up reserving way more metadata space than
what we need.

The only place where the metadata space accounting is not done is in the
relocation code.  This is because Yan is going to be reworking that code in the
near future, so running btrfs-vol -b could still possibly result in a ENOSPC
related panic.  This patch also turns off the metadata_ratio stuff in order to
allow users to more efficiently use their disk space.

This patch makes it so we track how much metadata we need for an inode's
delayed allocation extents by tracking how many extents are currently
waiting for allocation.  It introduces two new callbacks for the
extent_io tree's, merge_extent_hook and split_extent_hook.  These help
us keep track of when we merge delalloc extents together and split them
up.  Reservations are handled prior to any actually dirty'ing occurs,
and then we unreserve after we dirty.

btrfs_unreserve_metadata_for_delalloc() will make the appropriate
unreservations as needed based on the number of reservations we
currently have and the number of extents we currently have.  Doing the
reservation outside of doing any of the actual dirty'ing lets us do
things like filemap_flush() the inode to try and force delalloc to
happen, or as a last resort actually start allocation on all delalloc
inodes in the fs.  This has survived dbench, fs_mark and an fsx torture
test.

Signed-off-by: Josef Bacik <jbacik@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/btrfs_inode.h |   8 +
 fs/btrfs/ctree.h       |  23 ++-
 fs/btrfs/disk-io.c     |   2 +-
 fs/btrfs/extent-tree.c | 389 +++++++++++++++++++++++++++++++++++++++++++------
 fs/btrfs/extent_io.c   |  92 +++++++++---
 fs/btrfs/extent_io.h   |  13 +-
 fs/btrfs/file.c        |  11 +-
 fs/btrfs/inode.c       | 224 ++++++++++++++++++++++++----
 fs/btrfs/ioctl.c       |  21 ++-
 fs/btrfs/transaction.c |  10 ++
 10 files changed, 678 insertions(+), 115 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 82ee56bba29..a54d354cefc 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -127,6 +127,14 @@ struct btrfs_inode {
 	 */
 	u64 last_unlink_trans;
 
+	/*
+	 * These two counters are for delalloc metadata reservations.  We keep
+	 * track of how many extents we've accounted for vs how many extents we
+	 * have.
+	 */
+	int delalloc_reserved_extents;
+	int delalloc_extents;
+
 	/*
 	 * ordered_data_close is set by truncate when a file that used
 	 * to have good data has been truncated to zero.  When it is set
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 80599b4e42b..b3959a150c3 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -675,18 +675,19 @@ struct btrfs_space_info {
 				   current allocations */
 	u64 bytes_readonly;	/* total bytes that are read only */
 	u64 bytes_super;	/* total bytes reserved for the super blocks */
-
-	/* delalloc accounting */
-	u64 bytes_delalloc;	/* number of bytes reserved for allocation,
-				   this space is not necessarily reserved yet
-				   by the allocator */
+	u64 bytes_root;		/* the number of bytes needed to commit a
+				   transaction */
 	u64 bytes_may_use;	/* number of bytes that may be used for
-				   delalloc */
+				   delalloc/allocations */
+	u64 bytes_delalloc;	/* number of bytes currently reserved for
+				   delayed allocation */
 
 	int full;		/* indicates that we cannot allocate any more
 				   chunks for this space */
 	int force_alloc;	/* set if we need to force a chunk alloc for
 				   this space */
+	int force_delalloc;	/* make people start doing filemap_flush until
+				   we're under a threshold */
 
 	struct list_head list;
 
@@ -695,6 +696,9 @@ struct btrfs_space_info {
 	spinlock_t lock;
 	struct rw_semaphore groups_sem;
 	atomic_t caching_threads;
+
+	int allocating_chunk;
+	wait_queue_head_t wait;
 };
 
 /*
@@ -2022,7 +2026,12 @@ u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags);
 void btrfs_set_inode_space_info(struct btrfs_root *root, struct inode *ionde);
 void btrfs_clear_space_info_full(struct btrfs_fs_info *info);
 
-int btrfs_check_metadata_free_space(struct btrfs_root *root);
+int btrfs_reserve_metadata_space(struct btrfs_root *root, int num_items);
+int btrfs_unreserve_metadata_space(struct btrfs_root *root, int num_items);
+int btrfs_unreserve_metadata_for_delalloc(struct btrfs_root *root,
+					  struct inode *inode, int num_items);
+int btrfs_reserve_metadata_for_delalloc(struct btrfs_root *root,
+					struct inode *inode, int num_items);
 int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode,
 				u64 bytes);
 void btrfs_free_reserved_data_space(struct btrfs_root *root,
diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index f1e905f7e70..ece8d1e26b5 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1629,7 +1629,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 	fs_info->sb = sb;
 	fs_info->max_extent = (u64)-1;
 	fs_info->max_inline = 8192 * 1024;
-	fs_info->metadata_ratio = 8;
+	fs_info->metadata_ratio = 0;
 
 	fs_info->thread_pool_size = min_t(unsigned long,
 					  num_online_cpus() + 2, 8);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 90d314eeff6..a4b2b03cd68 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -68,6 +68,8 @@ static int pin_down_bytes(struct btrfs_trans_handle *trans,
 			  struct extent_buffer **must_clean);
 static int find_next_key(struct btrfs_path *path, int level,
 			 struct btrfs_key *key);
+static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
+			    int dump_block_groups);
 
 static noinline int
 block_group_cache_done(struct btrfs_block_group_cache *cache)
@@ -2764,67 +2766,346 @@ void btrfs_set_inode_space_info(struct btrfs_root *root, struct inode *inode)
 						       alloc_target);
 }
 
+static u64 calculate_bytes_needed(struct btrfs_root *root, int num_items)
+{
+	u64 num_bytes;
+	int level;
+
+	level = BTRFS_MAX_LEVEL - 2;
+	/*
+	 * NOTE: these calculations are absolutely the worst possible case.
+	 * This assumes that _every_ item we insert will require a new leaf, and
+	 * that the tree has grown to its maximum level size.
+	 */
+
+	/*
+	 * for every item we insert we could insert both an extent item and a
+	 * extent ref item.  Then for ever item we insert, we will need to cow
+	 * both the original leaf, plus the leaf to the left and right of it.
+	 *
+	 * Unless we are talking about the extent root, then we just want the
+	 * number of items * 2, since we just need the extent item plus its ref.
+	 */
+	if (root == root->fs_info->extent_root)
+		num_bytes = num_items * 2;
+	else
+		num_bytes = (num_items + (2 * num_items)) * 3;
+
+	/*
+	 * num_bytes is total number of leaves we could need times the leaf
+	 * size, and then for every leaf we could end up cow'ing 2 nodes per
+	 * level, down to the leaf level.
+	 */
+	num_bytes = (num_bytes * root->leafsize) +
+		(num_bytes * (level * 2)) * root->nodesize;
+
+	return num_bytes;
+}
+
 /*
- * for now this just makes sure we have at least 5% of our metadata space free
- * for use.
+ * Unreserve metadata space for delalloc.  If we have less reserved credits than
+ * we have extents, this function does nothing.
  */
-int btrfs_check_metadata_free_space(struct btrfs_root *root)
+int btrfs_unreserve_metadata_for_delalloc(struct btrfs_root *root,
+					  struct inode *inode, int num_items)
 {
 	struct btrfs_fs_info *info = root->fs_info;
 	struct btrfs_space_info *meta_sinfo;
-	u64 alloc_target, thresh;
-	int committed = 0, ret;
+	u64 num_bytes;
+	u64 alloc_target;
+	bool bug = false;
 
 	/* get the space info for where the metadata will live */
 	alloc_target = btrfs_get_alloc_profile(root, 0);
 	meta_sinfo = __find_space_info(info, alloc_target);
-	if (!meta_sinfo)
-		goto alloc;
 
-again:
+	num_bytes = calculate_bytes_needed(root->fs_info->extent_root,
+					   num_items);
+
 	spin_lock(&meta_sinfo->lock);
-	if (!meta_sinfo->full)
-		thresh = meta_sinfo->total_bytes * 80;
-	else
-		thresh = meta_sinfo->total_bytes * 95;
+	if (BTRFS_I(inode)->delalloc_reserved_extents <=
+	    BTRFS_I(inode)->delalloc_extents) {
+		spin_unlock(&meta_sinfo->lock);
+		return 0;
+	}
+
+	BTRFS_I(inode)->delalloc_reserved_extents--;
+	BUG_ON(BTRFS_I(inode)->delalloc_reserved_extents < 0);
+
+	if (meta_sinfo->bytes_delalloc < num_bytes) {
+		bug = true;
+		meta_sinfo->bytes_delalloc = 0;
+	} else {
+		meta_sinfo->bytes_delalloc -= num_bytes;
+	}
+	spin_unlock(&meta_sinfo->lock);
 
+	BUG_ON(bug);
+
+	return 0;
+}
+
+static void check_force_delalloc(struct btrfs_space_info *meta_sinfo)
+{
+	u64 thresh;
+
+	thresh = meta_sinfo->bytes_used + meta_sinfo->bytes_reserved +
+		meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly +
+		meta_sinfo->bytes_super + meta_sinfo->bytes_root +
+		meta_sinfo->bytes_may_use;
+
+	thresh = meta_sinfo->total_bytes - thresh;
+	thresh *= 80;
 	do_div(thresh, 100);
+	if (thresh <= meta_sinfo->bytes_delalloc)
+		meta_sinfo->force_delalloc = 1;
+	else
+		meta_sinfo->force_delalloc = 0;
+}
 
-	if (meta_sinfo->bytes_used + meta_sinfo->bytes_reserved +
-	    meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly +
-	    meta_sinfo->bytes_super > thresh) {
-		struct btrfs_trans_handle *trans;
-		if (!meta_sinfo->full) {
-			meta_sinfo->force_alloc = 1;
+static int maybe_allocate_chunk(struct btrfs_root *root,
+				 struct btrfs_space_info *info)
+{
+	struct btrfs_super_block *disk_super = &root->fs_info->super_copy;
+	struct btrfs_trans_handle *trans;
+	bool wait = false;
+	int ret = 0;
+	u64 min_metadata;
+	u64 free_space;
+
+	free_space = btrfs_super_total_bytes(disk_super);
+	/*
+	 * we allow the metadata to grow to a max of either 5gb or 5% of the
+	 * space in the volume.
+	 */
+	min_metadata = min((u64)5 * 1024 * 1024 * 1024,
+			     div64_u64(free_space * 5, 100));
+	if (info->total_bytes >= min_metadata) {
+		spin_unlock(&info->lock);
+		return 0;
+	}
+
+	if (info->full) {
+		spin_unlock(&info->lock);
+		return 0;
+	}
+
+	if (!info->allocating_chunk) {
+		info->force_alloc = 1;
+		info->allocating_chunk = 1;
+		init_waitqueue_head(&info->wait);
+	} else {
+		wait = true;
+	}
+
+	spin_unlock(&info->lock);
+
+	if (wait) {
+		wait_event(info->wait,
+			   !info->allocating_chunk);
+		return 1;
+	}
+
+	trans = btrfs_start_transaction(root, 1);
+	if (!trans) {
+		ret = -ENOMEM;
+		goto out;
+	}
+
+	ret = do_chunk_alloc(trans, root->fs_info->extent_root,
+			     4096 + 2 * 1024 * 1024,
+			     info->flags, 0);
+	btrfs_end_transaction(trans, root);
+	if (ret)
+		goto out;
+out:
+	spin_lock(&info->lock);
+	info->allocating_chunk = 0;
+	spin_unlock(&info->lock);
+	wake_up(&info->wait);
+
+	if (ret)
+		return 0;
+	return 1;
+}
+
+/*
+ * Reserve metadata space for delalloc.
+ */
+int btrfs_reserve_metadata_for_delalloc(struct btrfs_root *root,
+					struct inode *inode, int num_items)
+{
+	struct btrfs_fs_info *info = root->fs_info;
+	struct btrfs_space_info *meta_sinfo;
+	u64 num_bytes;
+	u64 used;
+	u64 alloc_target;
+	int flushed = 0;
+	int force_delalloc;
+
+	/* get the space info for where the metadata will live */
+	alloc_target = btrfs_get_alloc_profile(root, 0);
+	meta_sinfo = __find_space_info(info, alloc_target);
+
+	num_bytes = calculate_bytes_needed(root->fs_info->extent_root,
+					   num_items);
+again:
+	spin_lock(&meta_sinfo->lock);
+
+	force_delalloc = meta_sinfo->force_delalloc;
+
+	if (unlikely(!meta_sinfo->bytes_root))
+		meta_sinfo->bytes_root = calculate_bytes_needed(root, 6);
+
+	if (!flushed)
+		meta_sinfo->bytes_delalloc += num_bytes;
+
+	used = meta_sinfo->bytes_used + meta_sinfo->bytes_reserved +
+		meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly +
+		meta_sinfo->bytes_super + meta_sinfo->bytes_root +
+		meta_sinfo->bytes_may_use + meta_sinfo->bytes_delalloc;
+
+	if (used > meta_sinfo->total_bytes) {
+		flushed++;
+
+		if (flushed == 1) {
+			if (maybe_allocate_chunk(root, meta_sinfo))
+				goto again;
+			flushed++;
+		} else {
 			spin_unlock(&meta_sinfo->lock);
-alloc:
-			trans = btrfs_start_transaction(root, 1);
-			if (!trans)
-				return -ENOMEM;
+		}
 
-			ret = do_chunk_alloc(trans, root->fs_info->extent_root,
-					     2 * 1024 * 1024, alloc_target, 0);
-			btrfs_end_transaction(trans, root);
-			if (!meta_sinfo) {
-				meta_sinfo = __find_space_info(info,
-							       alloc_target);
-			}
+		if (flushed == 2) {
+			filemap_flush(inode->i_mapping);
+			goto again;
+		} else if (flushed == 3) {
+			btrfs_start_delalloc_inodes(root);
+			btrfs_wait_ordered_extents(root, 0);
 			goto again;
 		}
+		spin_lock(&meta_sinfo->lock);
+		meta_sinfo->bytes_delalloc -= num_bytes;
 		spin_unlock(&meta_sinfo->lock);
+		printk(KERN_ERR "enospc, has %d, reserved %d\n",
+		       BTRFS_I(inode)->delalloc_extents,
+		       BTRFS_I(inode)->delalloc_reserved_extents);
+		dump_space_info(meta_sinfo, 0, 0);
+		return -ENOSPC;
+	}
 
-		if (!committed) {
-			committed = 1;
-			trans = btrfs_join_transaction(root, 1);
-			if (!trans)
-				return -ENOMEM;
-			ret = btrfs_commit_transaction(trans, root);
-			if (ret)
-				return ret;
+	BTRFS_I(inode)->delalloc_reserved_extents++;
+	check_force_delalloc(meta_sinfo);
+	spin_unlock(&meta_sinfo->lock);
+
+	if (!flushed && force_delalloc)
+		filemap_flush(inode->i_mapping);
+
+	return 0;
+}
+
+/*
+ * unreserve num_items number of items worth of metadata space.  This needs to
+ * be paired with btrfs_reserve_metadata_space.
+ *
+ * NOTE: if you have the option, run this _AFTER_ you do a
+ * btrfs_end_transaction, since btrfs_end_transaction will run delayed ref
+ * oprations which will result in more used metadata, so we want to make sure we
+ * can do that without issue.
+ */
+int btrfs_unreserve_metadata_space(struct btrfs_root *root, int num_items)
+{
+	struct btrfs_fs_info *info = root->fs_info;
+	struct btrfs_space_info *meta_sinfo;
+	u64 num_bytes;
+	u64 alloc_target;
+	bool bug = false;
+
+	/* get the space info for where the metadata will live */
+	alloc_target = btrfs_get_alloc_profile(root, 0);
+	meta_sinfo = __find_space_info(info, alloc_target);
+
+	num_bytes = calculate_bytes_needed(root, num_items);
+
+	spin_lock(&meta_sinfo->lock);
+	if (meta_sinfo->bytes_may_use < num_bytes) {
+		bug = true;
+		meta_sinfo->bytes_may_use = 0;
+	} else {
+		meta_sinfo->bytes_may_use -= num_bytes;
+	}
+	spin_unlock(&meta_sinfo->lock);
+
+	BUG_ON(bug);
+
+	return 0;
+}
+
+/*
+ * Reserve some metadata space for use.  We'll calculate the worste case number
+ * of bytes that would be needed to modify num_items number of items.  If we
+ * have space, fantastic, if not, you get -ENOSPC.  Please call
+ * btrfs_unreserve_metadata_space when you are done for the _SAME_ number of
+ * items you reserved, since whatever metadata you needed should have already
+ * been allocated.
+ *
+ * This will commit the transaction to make more space if we don't have enough
+ * metadata space.  THe only time we don't do this is if we're reserving space
+ * inside of a transaction, then we will just return -ENOSPC and it is the
+ * callers responsibility to handle it properly.
+ */
+int btrfs_reserve_metadata_space(struct btrfs_root *root, int num_items)
+{
+	struct btrfs_fs_info *info = root->fs_info;
+	struct btrfs_space_info *meta_sinfo;
+	u64 num_bytes;
+	u64 used;
+	u64 alloc_target;
+	int retries = 0;
+
+	/* get the space info for where the metadata will live */
+	alloc_target = btrfs_get_alloc_profile(root, 0);
+	meta_sinfo = __find_space_info(info, alloc_target);
+
+	num_bytes = calculate_bytes_needed(root, num_items);
+again:
+	spin_lock(&meta_sinfo->lock);
+
+	if (unlikely(!meta_sinfo->bytes_root))
+		meta_sinfo->bytes_root = calculate_bytes_needed(root, 6);
+
+	if (!retries)
+		meta_sinfo->bytes_may_use += num_bytes;
+
+	used = meta_sinfo->bytes_used + meta_sinfo->bytes_reserved +
+		meta_sinfo->bytes_pinned + meta_sinfo->bytes_readonly +
+		meta_sinfo->bytes_super + meta_sinfo->bytes_root +
+		meta_sinfo->bytes_may_use + meta_sinfo->bytes_delalloc;
+
+	if (used > meta_sinfo->total_bytes) {
+		retries++;
+		if (retries == 1) {
+			if (maybe_allocate_chunk(root, meta_sinfo))
+				goto again;
+			retries++;
+		} else {
+			spin_unlock(&meta_sinfo->lock);
+		}
+
+		if (retries == 2) {
+			btrfs_start_delalloc_inodes(root);
+			btrfs_wait_ordered_extents(root, 0);
 			goto again;
 		}
+		spin_lock(&meta_sinfo->lock);
+		meta_sinfo->bytes_may_use -= num_bytes;
+		spin_unlock(&meta_sinfo->lock);
+
+		dump_space_info(meta_sinfo, 0, 0);
 		return -ENOSPC;
 	}
+
+	check_force_delalloc(meta_sinfo);
 	spin_unlock(&meta_sinfo->lock);
 
 	return 0;
@@ -2915,7 +3196,7 @@ alloc:
 	BTRFS_I(inode)->reserved_bytes += bytes;
 	spin_unlock(&data_sinfo->lock);
 
-	return btrfs_check_metadata_free_space(root);
+	return 0;
 }
 
 /*
@@ -3014,17 +3295,15 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 	BUG_ON(!space_info);
 
 	spin_lock(&space_info->lock);
-	if (space_info->force_alloc) {
+	if (space_info->force_alloc)
 		force = 1;
-		space_info->force_alloc = 0;
-	}
 	if (space_info->full) {
 		spin_unlock(&space_info->lock);
 		goto out;
 	}
 
 	thresh = space_info->total_bytes - space_info->bytes_readonly;
-	thresh = div_factor(thresh, 6);
+	thresh = div_factor(thresh, 8);
 	if (!force &&
 	   (space_info->bytes_used + space_info->bytes_pinned +
 	    space_info->bytes_reserved + alloc_bytes) < thresh) {
@@ -3038,7 +3317,7 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 	 * we keep a reasonable number of metadata chunks allocated in the
 	 * FS as well.
 	 */
-	if (flags & BTRFS_BLOCK_GROUP_DATA) {
+	if (flags & BTRFS_BLOCK_GROUP_DATA && fs_info->metadata_ratio) {
 		fs_info->data_chunk_allocations++;
 		if (!(fs_info->data_chunk_allocations %
 		      fs_info->metadata_ratio))
@@ -3046,8 +3325,11 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans,
 	}
 
 	ret = btrfs_alloc_chunk(trans, extent_root, flags);
+	spin_lock(&space_info->lock);
 	if (ret)
 		space_info->full = 1;
+	space_info->force_alloc = 0;
+	spin_unlock(&space_info->lock);
 out:
 	mutex_unlock(&extent_root->fs_info->chunk_mutex);
 	return ret;
@@ -4062,21 +4344,32 @@ loop:
 	return ret;
 }
 
-static void dump_space_info(struct btrfs_space_info *info, u64 bytes)
+static void dump_space_info(struct btrfs_space_info *info, u64 bytes,
+			    int dump_block_groups)
 {
 	struct btrfs_block_group_cache *cache;
 
+	spin_lock(&info->lock);
 	printk(KERN_INFO "space_info has %llu free, is %sfull\n",
 	       (unsigned long long)(info->total_bytes - info->bytes_used -
-				    info->bytes_pinned - info->bytes_reserved),
+				    info->bytes_pinned - info->bytes_reserved -
+				    info->bytes_super),
 	       (info->full) ? "" : "not ");
 	printk(KERN_INFO "space_info total=%llu, pinned=%llu, delalloc=%llu,"
-	       " may_use=%llu, used=%llu\n",
+	       " may_use=%llu, used=%llu, root=%llu, super=%llu, reserved=%llu"
+	       "\n",
 	       (unsigned long long)info->total_bytes,
 	       (unsigned long long)info->bytes_pinned,
 	       (unsigned long long)info->bytes_delalloc,
 	       (unsigned long long)info->bytes_may_use,
-	       (unsigned long long)info->bytes_used);
+	       (unsigned long long)info->bytes_used,
+	       (unsigned long long)info->bytes_root,
+	       (unsigned long long)info->bytes_super,
+	       (unsigned long long)info->bytes_reserved);
+	spin_unlock(&info->lock);
+
+	if (!dump_block_groups)
+		return;
 
 	down_read(&info->groups_sem);
 	list_for_each_entry(cache, &info->block_groups, list) {
@@ -4144,7 +4437,7 @@ again:
 		printk(KERN_ERR "btrfs allocation failed flags %llu, "
 		       "wanted %llu\n", (unsigned long long)data,
 		       (unsigned long long)num_bytes);
-		dump_space_info(sinfo, num_bytes);
+		dump_space_info(sinfo, num_bytes, 1);
 	}
 
 	return ret;
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 0cb88f8146e..de1793ba004 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -280,6 +280,14 @@ static struct extent_buffer *buffer_search(struct extent_io_tree *tree,
 	return NULL;
 }
 
+static void merge_cb(struct extent_io_tree *tree, struct extent_state *new,
+		     struct extent_state *other)
+{
+	if (tree->ops && tree->ops->merge_extent_hook)
+		tree->ops->merge_extent_hook(tree->mapping->host, new,
+					     other);
+}
+
 /*
  * utility function to look for merge candidates inside a given range.
  * Any extents with matching state are merged together into a single
@@ -303,6 +311,7 @@ static int merge_state(struct extent_io_tree *tree,
 		other = rb_entry(other_node, struct extent_state, rb_node);
 		if (other->end == state->start - 1 &&
 		    other->state == state->state) {
+			merge_cb(tree, state, other);
 			state->start = other->start;
 			other->tree = NULL;
 			rb_erase(&other->rb_node, &tree->state);
@@ -314,33 +323,37 @@ static int merge_state(struct extent_io_tree *tree,
 		other = rb_entry(other_node, struct extent_state, rb_node);
 		if (other->start == state->end + 1 &&
 		    other->state == state->state) {
+			merge_cb(tree, state, other);
 			other->start = state->start;
 			state->tree = NULL;
 			rb_erase(&state->rb_node, &tree->state);
 			free_extent_state(state);
+			state = NULL;
 		}
 	}
+
 	return 0;
 }
 
-static void set_state_cb(struct extent_io_tree *tree,
+static int set_state_cb(struct extent_io_tree *tree,
 			 struct extent_state *state,
 			 unsigned long bits)
 {
 	if (tree->ops && tree->ops->set_bit_hook) {
-		tree->ops->set_bit_hook(tree->mapping->host, state->start,
-					state->end, state->state, bits);
+		return tree->ops->set_bit_hook(tree->mapping->host,
+					       state->start, state->end,
+					       state->state, bits);
 	}
+
+	return 0;
 }
 
 static void clear_state_cb(struct extent_io_tree *tree,
 			   struct extent_state *state,
 			   unsigned long bits)
 {
-	if (tree->ops && tree->ops->clear_bit_hook) {
-		tree->ops->clear_bit_hook(tree->mapping->host, state->start,
-					  state->end, state->state, bits);
-	}
+	if (tree->ops && tree->ops->clear_bit_hook)
+		tree->ops->clear_bit_hook(tree->mapping->host, state, bits);
 }
 
 /*
@@ -358,6 +371,7 @@ static int insert_state(struct extent_io_tree *tree,
 			int bits)
 {
 	struct rb_node *node;
+	int ret;
 
 	if (end < start) {
 		printk(KERN_ERR "btrfs end < start %llu %llu\n",
@@ -365,11 +379,14 @@ static int insert_state(struct extent_io_tree *tree,
 		       (unsigned long long)start);
 		WARN_ON(1);
 	}
-	if (bits & EXTENT_DIRTY)
-		tree->dirty_bytes += end - start + 1;
 	state->start = start;
 	state->end = end;
-	set_state_cb(tree, state, bits);
+	ret = set_state_cb(tree, state, bits);
+	if (ret)
+		return ret;
+
+	if (bits & EXTENT_DIRTY)
+		tree->dirty_bytes += end - start + 1;
 	state->state |= bits;
 	node = tree_insert(&tree->state, end, &state->rb_node);
 	if (node) {
@@ -387,6 +404,15 @@ static int insert_state(struct extent_io_tree *tree,
 	return 0;
 }
 
+static int split_cb(struct extent_io_tree *tree, struct extent_state *orig,
+		     u64 split)
+{
+	if (tree->ops && tree->ops->split_extent_hook)
+		return tree->ops->split_extent_hook(tree->mapping->host,
+						    orig, split);
+	return 0;
+}
+
 /*
  * split a given extent state struct in two, inserting the preallocated
  * struct 'prealloc' as the newly created second half.  'split' indicates an
@@ -405,6 +431,9 @@ static int split_state(struct extent_io_tree *tree, struct extent_state *orig,
 		       struct extent_state *prealloc, u64 split)
 {
 	struct rb_node *node;
+
+	split_cb(tree, orig, split);
+
 	prealloc->start = orig->start;
 	prealloc->end = split - 1;
 	prealloc->state = orig->state;
@@ -542,8 +571,8 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			set |= clear_state_bit(tree, state, bits,
-					wake, delete);
+			set |= clear_state_bit(tree, state, bits, wake,
+					       delete);
 			if (last_end == (u64)-1)
 				goto out;
 			start = last_end + 1;
@@ -561,12 +590,11 @@ hit_next:
 			prealloc = alloc_extent_state(GFP_ATOMIC);
 		err = split_state(tree, state, prealloc, end + 1);
 		BUG_ON(err == -EEXIST);
-
 		if (wake)
 			wake_up(&state->wq);
 
-		set |= clear_state_bit(tree, prealloc, bits,
-				       wake, delete);
+		set |= clear_state_bit(tree, prealloc, bits, wake, delete);
+
 		prealloc = NULL;
 		goto out;
 	}
@@ -667,16 +695,23 @@ out:
 	return 0;
 }
 
-static void set_state_bits(struct extent_io_tree *tree,
+static int set_state_bits(struct extent_io_tree *tree,
 			   struct extent_state *state,
 			   int bits)
 {
+	int ret;
+
+	ret = set_state_cb(tree, state, bits);
+	if (ret)
+		return ret;
+
 	if ((bits & EXTENT_DIRTY) && !(state->state & EXTENT_DIRTY)) {
 		u64 range = state->end - state->start + 1;
 		tree->dirty_bytes += range;
 	}
-	set_state_cb(tree, state, bits);
 	state->state |= bits;
+
+	return 0;
 }
 
 static void cache_state(struct extent_state *state,
@@ -758,7 +793,10 @@ hit_next:
 			goto out;
 		}
 
-		set_state_bits(tree, state, bits);
+		err = set_state_bits(tree, state, bits);
+		if (err)
+			goto out;
+
 		cache_state(state, cached_state);
 		merge_state(tree, state);
 		if (last_end == (u64)-1)
@@ -805,7 +843,9 @@ hit_next:
 		if (err)
 			goto out;
 		if (state->end <= end) {
-			set_state_bits(tree, state, bits);
+			err = set_state_bits(tree, state, bits);
+			if (err)
+				goto out;
 			cache_state(state, cached_state);
 			merge_state(tree, state);
 			if (last_end == (u64)-1)
@@ -829,11 +869,13 @@ hit_next:
 			this_end = last_start - 1;
 		err = insert_state(tree, prealloc, start, this_end,
 				   bits);
-		cache_state(prealloc, cached_state);
-		prealloc = NULL;
 		BUG_ON(err == -EEXIST);
-		if (err)
+		if (err) {
+			prealloc = NULL;
 			goto out;
+		}
+		cache_state(prealloc, cached_state);
+		prealloc = NULL;
 		start = this_end + 1;
 		goto search_again;
 	}
@@ -852,7 +894,11 @@ hit_next:
 		err = split_state(tree, state, prealloc, end + 1);
 		BUG_ON(err == -EEXIST);
 
-		set_state_bits(tree, prealloc, bits);
+		err = set_state_bits(tree, prealloc, bits);
+		if (err) {
+			prealloc = NULL;
+			goto out;
+		}
 		cache_state(prealloc, cached_state);
 		merge_state(tree, prealloc);
 		prealloc = NULL;
diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h
index 14ed16fd862..4794ec891fe 100644
--- a/fs/btrfs/extent_io.h
+++ b/fs/btrfs/extent_io.h
@@ -60,8 +60,13 @@ struct extent_io_ops {
 				      struct extent_state *state, int uptodate);
 	int (*set_bit_hook)(struct inode *inode, u64 start, u64 end,
 			    unsigned long old, unsigned long bits);
-	int (*clear_bit_hook)(struct inode *inode, u64 start, u64 end,
-			    unsigned long old, unsigned long bits);
+	int (*clear_bit_hook)(struct inode *inode, struct extent_state *state,
+			      unsigned long bits);
+	int (*merge_extent_hook)(struct inode *inode,
+				 struct extent_state *new,
+				 struct extent_state *other);
+	int (*split_extent_hook)(struct inode *inode,
+				 struct extent_state *orig, u64 split);
 	int (*write_cache_pages_lock_hook)(struct page *page);
 };
 
@@ -79,10 +84,14 @@ struct extent_state {
 	u64 start;
 	u64 end; /* inclusive */
 	struct rb_node rb_node;
+
+	/* ADD NEW ELEMENTS AFTER THIS */
 	struct extent_io_tree *tree;
 	wait_queue_head_t wq;
 	atomic_t refs;
 	unsigned long state;
+	u64 split_start;
+	u64 split_end;
 
 	/* for use by the FS */
 	u64 private;
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 571ad3c13b4..1be96ba6f6b 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -123,7 +123,10 @@ static noinline int dirty_and_release_pages(struct btrfs_trans_handle *trans,
 		    root->sectorsize - 1) & ~((u64)root->sectorsize - 1);
 
 	end_of_last_block = start_pos + num_bytes - 1;
-	btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
+	err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block);
+	if (err)
+		return err;
+
 	for (i = 0; i < num_pages; i++) {
 		struct page *p = pages[i];
 		SetPageUptodate(p);
@@ -927,6 +930,11 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf,
 	err = file_remove_suid(file);
 	if (err)
 		goto out_nolock;
+
+	err = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+	if (err)
+		goto out_nolock;
+
 	file_update_time(file);
 
 	pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
@@ -1028,6 +1036,7 @@ out:
 	mutex_unlock(&inode->i_mutex);
 	if (ret)
 		err = ret;
+	btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
 
 out_nolock:
 	kfree(pages);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 01c5f8b5a34..3cc5677f544 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -1159,6 +1159,83 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
 	return ret;
 }
 
+static int btrfs_split_extent_hook(struct inode *inode,
+				    struct extent_state *orig, u64 split)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	u64 size;
+
+	if (!(orig->state & EXTENT_DELALLOC))
+		return 0;
+
+	size = orig->end - orig->start + 1;
+	if (size > root->fs_info->max_extent) {
+		u64 num_extents;
+		u64 new_size;
+
+		new_size = orig->end - split + 1;
+		num_extents = div64_u64(size + root->fs_info->max_extent - 1,
+					root->fs_info->max_extent);
+
+		/*
+		 * if we break a large extent up then leave delalloc_extents be,
+		 * since we've already accounted for the large extent.
+		 */
+		if (div64_u64(new_size + root->fs_info->max_extent - 1,
+			      root->fs_info->max_extent) < num_extents)
+			return 0;
+	}
+
+	BTRFS_I(inode)->delalloc_extents++;
+
+	return 0;
+}
+
+/*
+ * extent_io.c merge_extent_hook, used to track merged delayed allocation
+ * extents so we can keep track of new extents that are just merged onto old
+ * extents, such as when we are doing sequential writes, so we can properly
+ * account for the metadata space we'll need.
+ */
+static int btrfs_merge_extent_hook(struct inode *inode,
+				   struct extent_state *new,
+				   struct extent_state *other)
+{
+	struct btrfs_root *root = BTRFS_I(inode)->root;
+	u64 new_size, old_size;
+	u64 num_extents;
+
+	/* not delalloc, ignore it */
+	if (!(other->state & EXTENT_DELALLOC))
+		return 0;
+
+	old_size = other->end - other->start + 1;
+	if (new->start < other->start)
+		new_size = other->end - new->start + 1;
+	else
+		new_size = new->end - other->start + 1;
+
+	/* we're not bigger than the max, unreserve the space and go */
+	if (new_size <= root->fs_info->max_extent) {
+		BTRFS_I(inode)->delalloc_extents--;
+		return 0;
+	}
+
+	/*
+	 * If we grew by another max_extent, just return, we want to keep that
+	 * reserved amount.
+	 */
+	num_extents = div64_u64(old_size + root->fs_info->max_extent - 1,
+				root->fs_info->max_extent);
+	if (div64_u64(new_size + root->fs_info->max_extent - 1,
+		      root->fs_info->max_extent) > num_extents)
+		return 0;
+
+	BTRFS_I(inode)->delalloc_extents--;
+
+	return 0;
+}
+
 /*
  * extent_io.c set_bit_hook, used to track delayed allocation
  * bytes in this file, and to maintain the list of inodes that
@@ -1167,6 +1244,7 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
 static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
 		       unsigned long old, unsigned long bits)
 {
+
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
 	 * but in this case, we are only testeing for the DELALLOC
@@ -1174,6 +1252,8 @@ static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
 	 */
 	if (!(old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
 		struct btrfs_root *root = BTRFS_I(inode)->root;
+
+		BTRFS_I(inode)->delalloc_extents++;
 		btrfs_delalloc_reserve_space(root, inode, end - start + 1);
 		spin_lock(&root->fs_info->delalloc_lock);
 		BTRFS_I(inode)->delalloc_bytes += end - start + 1;
@@ -1190,22 +1270,27 @@ static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
 /*
  * extent_io.c clear_bit_hook, see set_bit_hook for why
  */
-static int btrfs_clear_bit_hook(struct inode *inode, u64 start, u64 end,
-			 unsigned long old, unsigned long bits)
+static int btrfs_clear_bit_hook(struct inode *inode,
+				struct extent_state *state, unsigned long bits)
 {
 	/*
 	 * set_bit and clear bit hooks normally require _irqsave/restore
 	 * but in this case, we are only testeing for the DELALLOC
 	 * bit, which is only set or cleared with irqs on
 	 */
-	if ((old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
+	if ((state->state & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
 		struct btrfs_root *root = BTRFS_I(inode)->root;
 
+		BTRFS_I(inode)->delalloc_extents--;
+		btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
+
 		spin_lock(&root->fs_info->delalloc_lock);
-		if (end - start + 1 > root->fs_info->delalloc_bytes) {
+		if (state->end - state->start + 1 >
+		    root->fs_info->delalloc_bytes) {
 			printk(KERN_INFO "btrfs warning: delalloc account "
 			       "%llu %llu\n",
-			       (unsigned long long)end - start + 1,
+			       (unsigned long long)
+			       state->end - state->start + 1,
 			       (unsigned long long)
 			       root->fs_info->delalloc_bytes);
 			btrfs_delalloc_free_space(root, inode, (u64)-1);
@@ -1213,9 +1298,12 @@ static int btrfs_clear_bit_hook(struct inode *inode, u64 start, u64 end,
 			BTRFS_I(inode)->delalloc_bytes = 0;
 		} else {
 			btrfs_delalloc_free_space(root, inode,
-						  end - start + 1);
-			root->fs_info->delalloc_bytes -= end - start + 1;
-			BTRFS_I(inode)->delalloc_bytes -= end - start + 1;
+						  state->end -
+						  state->start + 1);
+			root->fs_info->delalloc_bytes -= state->end -
+				state->start + 1;
+			BTRFS_I(inode)->delalloc_bytes -= state->end -
+				state->start + 1;
 		}
 		if (BTRFS_I(inode)->delalloc_bytes == 0 &&
 		    !list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
@@ -2950,7 +3038,12 @@ again:
 		goto again;
 	}
 
-	btrfs_set_extent_delalloc(inode, page_start, page_end);
+	ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
+	if (ret) {
+		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+		goto out_unlock;
+	}
+
 	ret = 0;
 	if (offset != PAGE_CACHE_SIZE) {
 		kaddr = kmap(page);
@@ -2981,15 +3074,11 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
 	u64 last_byte;
 	u64 cur_offset;
 	u64 hole_size;
-	int err;
+	int err = 0;
 
 	if (size <= hole_start)
 		return 0;
 
-	err = btrfs_check_metadata_free_space(root);
-	if (err)
-		return err;
-
 	btrfs_truncate_page(inode->i_mapping, inode->i_size);
 
 	while (1) {
@@ -3024,12 +3113,18 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
 						 cur_offset, &hint_byte, 1);
 			if (err)
 				break;
+
+			err = btrfs_reserve_metadata_space(root, 1);
+			if (err)
+				break;
+
 			err = btrfs_insert_file_extent(trans, root,
 					inode->i_ino, cur_offset, 0,
 					0, hole_size, 0, hole_size,
 					0, 0, 0);
 			btrfs_drop_extent_cache(inode, hole_start,
 					last_byte - 1, 0);
+			btrfs_unreserve_metadata_space(root, 1);
 		}
 		free_extent_map(em);
 		cur_offset = last_byte;
@@ -3990,11 +4085,18 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
 	if (!new_valid_dev(rdev))
 		return -EINVAL;
 
-	err = btrfs_check_metadata_free_space(root);
+	/*
+	 * 2 for inode item and ref
+	 * 2 for dir items
+	 * 1 for xattr if selinux is on
+	 */
+	err = btrfs_reserve_metadata_space(root, 5);
 	if (err)
-		goto fail;
+		return err;
 
 	trans = btrfs_start_transaction(root, 1);
+	if (!trans)
+		goto fail;
 	btrfs_set_trans_block_group(trans, dir);
 
 	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
@@ -4032,6 +4134,7 @@ out_unlock:
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
 fail:
+	btrfs_unreserve_metadata_space(root, 5);
 	if (drop_inode) {
 		inode_dec_link_count(inode);
 		iput(inode);
@@ -4052,10 +4155,18 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry,
 	u64 objectid;
 	u64 index = 0;
 
-	err = btrfs_check_metadata_free_space(root);
+	/*
+	 * 2 for inode item and ref
+	 * 2 for dir items
+	 * 1 for xattr if selinux is on
+	 */
+	err = btrfs_reserve_metadata_space(root, 5);
 	if (err)
-		goto fail;
+		return err;
+
 	trans = btrfs_start_transaction(root, 1);
+	if (!trans)
+		goto fail;
 	btrfs_set_trans_block_group(trans, dir);
 
 	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
@@ -4096,6 +4207,7 @@ out_unlock:
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
 fail:
+	btrfs_unreserve_metadata_space(root, 5);
 	if (drop_inode) {
 		inode_dec_link_count(inode);
 		iput(inode);
@@ -4118,10 +4230,16 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 	if (inode->i_nlink == 0)
 		return -ENOENT;
 
-	btrfs_inc_nlink(inode);
-	err = btrfs_check_metadata_free_space(root);
+	/*
+	 * 1 item for inode ref
+	 * 2 items for dir items
+	 */
+	err = btrfs_reserve_metadata_space(root, 3);
 	if (err)
-		goto fail;
+		return err;
+
+	btrfs_inc_nlink(inode);
+
 	err = btrfs_set_inode_index(dir, &index);
 	if (err)
 		goto fail;
@@ -4145,6 +4263,7 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
 fail:
+	btrfs_unreserve_metadata_space(root, 3);
 	if (drop_inode) {
 		inode_dec_link_count(inode);
 		iput(inode);
@@ -4164,17 +4283,21 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 	u64 index = 0;
 	unsigned long nr = 1;
 
-	err = btrfs_check_metadata_free_space(root);
+	/*
+	 * 2 items for inode and ref
+	 * 2 items for dir items
+	 * 1 for xattr if selinux is on
+	 */
+	err = btrfs_reserve_metadata_space(root, 5);
 	if (err)
-		goto out_unlock;
+		return err;
 
 	trans = btrfs_start_transaction(root, 1);
-	btrfs_set_trans_block_group(trans, dir);
-
-	if (IS_ERR(trans)) {
-		err = PTR_ERR(trans);
+	if (!trans) {
+		err = -ENOMEM;
 		goto out_unlock;
 	}
+	btrfs_set_trans_block_group(trans, dir);
 
 	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
 	if (err) {
@@ -4223,6 +4346,7 @@ out_fail:
 	btrfs_end_transaction_throttle(trans, root);
 
 out_unlock:
+	btrfs_unreserve_metadata_space(root, 5);
 	if (drop_on_err)
 		iput(inode);
 	btrfs_btree_balance_dirty(root, nr);
@@ -4747,6 +4871,13 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 		goto out;
 	}
 
+	ret = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+	if (ret) {
+		btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE);
+		ret = VM_FAULT_SIGBUS;
+		goto out;
+	}
+
 	ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */
 again:
 	lock_page(page);
@@ -4778,7 +4909,12 @@ again:
 		goto again;
 	}
 
-	btrfs_set_extent_delalloc(inode, page_start, page_end);
+	ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
+	if (ret) {
+		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+		ret = VM_FAULT_SIGBUS;
+		goto out_unlock;
+	}
 	ret = 0;
 
 	/* page is wholly or partially inside EOF */
@@ -4801,6 +4937,7 @@ again:
 	unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
 
 out_unlock:
+	btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
 	if (!ret)
 		return VM_FAULT_LOCKED;
 	unlock_page(page);
@@ -4917,6 +5054,8 @@ struct inode *btrfs_alloc_inode(struct super_block *sb)
 		return NULL;
 	ei->last_trans = 0;
 	ei->logged_trans = 0;
+	ei->delalloc_extents = 0;
+	ei->delalloc_reserved_extents = 0;
 	btrfs_ordered_inode_tree_init(&ei->ordered_tree);
 	INIT_LIST_HEAD(&ei->i_orphan);
 	INIT_LIST_HEAD(&ei->ordered_operations);
@@ -5070,7 +5209,12 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
 	    new_inode->i_size > BTRFS_EMPTY_DIR_SIZE)
 		return -ENOTEMPTY;
 
-	ret = btrfs_check_metadata_free_space(root);
+	/*
+	 * 2 items for dir items
+	 * 1 item for orphan entry
+	 * 1 item for ref
+	 */
+	ret = btrfs_reserve_metadata_space(root, 4);
 	if (ret)
 		return ret;
 
@@ -5185,6 +5329,8 @@ out_fail:
 
 	if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
 		up_read(&root->fs_info->subvol_sem);
+
+	btrfs_unreserve_metadata_space(root, 4);
 	return ret;
 }
 
@@ -5256,11 +5402,18 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry,
 	if (name_len > BTRFS_MAX_INLINE_DATA_SIZE(root))
 		return -ENAMETOOLONG;
 
-	err = btrfs_check_metadata_free_space(root);
+	/*
+	 * 2 items for inode item and ref
+	 * 2 items for dir items
+	 * 1 item for xattr if selinux is on
+	 */
+	err = btrfs_reserve_metadata_space(root, 5);
 	if (err)
-		goto out_fail;
+		return err;
 
 	trans = btrfs_start_transaction(root, 1);
+	if (!trans)
+		goto out_fail;
 	btrfs_set_trans_block_group(trans, dir);
 
 	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
@@ -5341,6 +5494,7 @@ out_unlock:
 	nr = trans->blocks_used;
 	btrfs_end_transaction_throttle(trans, root);
 out_fail:
+	btrfs_unreserve_metadata_space(root, 5);
 	if (drop_inode) {
 		inode_dec_link_count(inode);
 		iput(inode);
@@ -5362,6 +5516,11 @@ static int prealloc_file_range(struct btrfs_trans_handle *trans,
 
 	while (num_bytes > 0) {
 		alloc_size = min(num_bytes, root->fs_info->max_extent);
+
+		ret = btrfs_reserve_metadata_space(root, 1);
+		if (ret)
+			goto out;
+
 		ret = btrfs_reserve_extent(trans, root, alloc_size,
 					   root->sectorsize, 0, alloc_hint,
 					   (u64)-1, &ins, 1);
@@ -5381,6 +5540,7 @@ static int prealloc_file_range(struct btrfs_trans_handle *trans,
 		num_bytes -= ins.offset;
 		cur_offset += ins.offset;
 		alloc_hint = ins.objectid + ins.offset;
+		btrfs_unreserve_metadata_space(root, 1);
 	}
 out:
 	if (cur_offset > start) {
@@ -5566,6 +5726,8 @@ static struct extent_io_ops btrfs_extent_io_ops = {
 	.readpage_io_failed_hook = btrfs_io_failed_hook,
 	.set_bit_hook = btrfs_set_bit_hook,
 	.clear_bit_hook = btrfs_clear_bit_hook,
+	.merge_extent_hook = btrfs_merge_extent_hook,
+	.split_extent_hook = btrfs_split_extent_hook,
 };
 
 /*
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index a8577a7f26a..4de7ef6f860 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -239,7 +239,13 @@ static noinline int create_subvol(struct btrfs_root *root,
 	u64 index = 0;
 	unsigned long nr = 1;
 
-	ret = btrfs_check_metadata_free_space(root);
+	/*
+	 * 1 - inode item
+	 * 2 - refs
+	 * 1 - root item
+	 * 2 - dir items
+	 */
+	ret = btrfs_reserve_metadata_space(root, 6);
 	if (ret)
 		return ret;
 
@@ -340,6 +346,9 @@ fail:
 	err = btrfs_commit_transaction(trans, root);
 	if (err && !ret)
 		ret = err;
+
+	btrfs_unreserve_metadata_space(root, 6);
+	btrfs_btree_balance_dirty(root, nr);
 	return ret;
 }
 
@@ -355,19 +364,27 @@ static int create_snapshot(struct btrfs_root *root, struct dentry *dentry,
 	if (!root->ref_cows)
 		return -EINVAL;
 
-	ret = btrfs_check_metadata_free_space(root);
+	/*
+	 * 1 - inode item
+	 * 2 - refs
+	 * 1 - root item
+	 * 2 - dir items
+	 */
+	ret = btrfs_reserve_metadata_space(root, 6);
 	if (ret)
 		goto fail_unlock;
 
 	pending_snapshot = kzalloc(sizeof(*pending_snapshot), GFP_NOFS);
 	if (!pending_snapshot) {
 		ret = -ENOMEM;
+		btrfs_unreserve_metadata_space(root, 6);
 		goto fail_unlock;
 	}
 	pending_snapshot->name = kmalloc(namelen + 1, GFP_NOFS);
 	if (!pending_snapshot->name) {
 		ret = -ENOMEM;
 		kfree(pending_snapshot);
+		btrfs_unreserve_metadata_space(root, 6);
 		goto fail_unlock;
 	}
 	memcpy(pending_snapshot->name, name, namelen);
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 88f866f85e7..0b8f36d4400 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -186,6 +186,9 @@ static struct btrfs_trans_handle *start_transaction(struct btrfs_root *root,
 	h->alloc_exclude_start = 0;
 	h->delayed_ref_updates = 0;
 
+	if (!current->journal_info)
+		current->journal_info = h;
+
 	root->fs_info->running_transaction->use_count++;
 	record_root_in_trans(h, root);
 	mutex_unlock(&root->fs_info->trans_mutex);
@@ -317,6 +320,9 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 		wake_up(&cur_trans->writer_wait);
 	put_transaction(cur_trans);
 	mutex_unlock(&info->trans_mutex);
+
+	if (current->journal_info == trans)
+		current->journal_info = NULL;
 	memset(trans, 0, sizeof(*trans));
 	kmem_cache_free(btrfs_trans_handle_cachep, trans);
 
@@ -743,6 +749,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans,
 	memcpy(&pending->root_key, &key, sizeof(key));
 fail:
 	kfree(new_root_item);
+	btrfs_unreserve_metadata_space(root, 6);
 	return ret;
 }
 
@@ -1059,6 +1066,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 
 	mutex_unlock(&root->fs_info->trans_mutex);
 
+	if (current->journal_info == trans)
+		current->journal_info = NULL;
+
 	kmem_cache_free(btrfs_trans_handle_cachep, trans);
 	return ret;
 }
-- 
cgit v1.2.3-70-g09d2


From ab93dbecfba72bbc04b7036343d180aaff1b61a3 Mon Sep 17 00:00:00 2001
From: Chris Mason <chris.mason@oracle.com>
Date: Thu, 1 Oct 2009 12:29:10 -0400
Subject: Btrfs: take i_mutex before generic_write_checks

btrfs_file_write was incorrectly calling generic_write_checks without
taking i_mutex.  This lead to problems with racing around i_size when
doing O_APPEND writes.

The fix here is to move i_mutex higher.

Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/file.c | 25 +++++++++++++++++--------
 1 file changed, 17 insertions(+), 8 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 1be96ba6f6b..f155179877a 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -920,26 +920,35 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf,
 	start_pos = pos;
 
 	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+
+	/* do the reserve before the mutex lock in case we have to do some
+	 * flushing.  We wouldn't deadlock, but this is more polite.
+	 */
+	err = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
+	if (err)
+		goto out_nolock;
+
+	mutex_lock(&inode->i_mutex);
+
 	current->backing_dev_info = inode->i_mapping->backing_dev_info;
 	err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode));
 	if (err)
-		goto out_nolock;
+		goto out;
+
 	if (count == 0)
-		goto out_nolock;
+		goto out;
 
 	err = file_remove_suid(file);
 	if (err)
-		goto out_nolock;
-
-	err = btrfs_reserve_metadata_for_delalloc(root, inode, 1);
-	if (err)
-		goto out_nolock;
+		goto out;
 
 	file_update_time(file);
 
 	pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL);
 
-	mutex_lock(&inode->i_mutex);
+	/* generic_write_checks can change our pos */
+	start_pos = pos;
+
 	BTRFS_I(inode)->sequence++;
 	first_index = pos >> PAGE_CACHE_SHIFT;
 	last_index = (pos + count) >> PAGE_CACHE_SHIFT;
-- 
cgit v1.2.3-70-g09d2


From 8aa38c31b7659e338fee4d9af4c3805acbd9806f Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Thu, 1 Oct 2009 12:58:30 -0400
Subject: Btrfs: remove duplicates of filemap_ helpers

Use filemap_fdatawrite_range and filemap_fdatawait_range instead of
local copies of the functions.  For filemap_fdatawait_range that
also means replacing the awkward old wait_on_page_writeback_range
calling convention with the regular filemap byte offsets.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
---
 fs/btrfs/disk-io.c      | 10 +++---
 fs/btrfs/file.c         |  5 ++-
 fs/btrfs/ordered-data.c | 93 +++----------------------------------------------
 fs/btrfs/ordered-data.h |  4 ---
 4 files changed, 10 insertions(+), 102 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index d20dc05208f..af0435f79fa 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -822,16 +822,14 @@ struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
 
 int btrfs_write_tree_block(struct extent_buffer *buf)
 {
-	return btrfs_fdatawrite_range(buf->first_page->mapping, buf->start,
-				      buf->start + buf->len - 1, WB_SYNC_ALL);
+	return filemap_fdatawrite_range(buf->first_page->mapping, buf->start,
+					buf->start + buf->len - 1);
 }
 
 int btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
 {
-	return btrfs_wait_on_page_writeback_range(buf->first_page->mapping,
-				  buf->start >> PAGE_CACHE_SHIFT,
-				  (buf->start + buf->len - 1) >>
-				   PAGE_CACHE_SHIFT);
+	return filemap_fdatawait_range(buf->first_page->mapping,
+				       buf->start, buf->start + buf->len - 1);
 }
 
 struct extent_buffer *read_tree_block(struct btrfs_root *root, u64 bytenr,
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 7351bdbca26..ca784a7fbeb 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1022,9 +1022,8 @@ static ssize_t btrfs_file_write(struct file *file, const char __user *buf,
 		}
 
 		if (will_write) {
-			btrfs_fdatawrite_range(inode->i_mapping, pos,
-					       pos + write_bytes - 1,
-					       WB_SYNC_ALL);
+			filemap_fdatawrite_range(inode->i_mapping, pos,
+						 pos + write_bytes - 1);
 		} else {
 			balance_dirty_pages_ratelimited_nr(inode->i_mapping,
 							   num_pages);
diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c
index b5d6d24726b..897fba835f8 100644
--- a/fs/btrfs/ordered-data.c
+++ b/fs/btrfs/ordered-data.c
@@ -458,7 +458,7 @@ void btrfs_start_ordered_extent(struct inode *inode,
 	 * start IO on any dirty ones so the wait doesn't stall waiting
 	 * for pdflush to find them
 	 */
-	btrfs_fdatawrite_range(inode->i_mapping, start, end, WB_SYNC_ALL);
+	filemap_fdatawrite_range(inode->i_mapping, start, end);
 	if (wait) {
 		wait_event(entry->wait, test_bit(BTRFS_ORDERED_COMPLETE,
 						 &entry->flags));
@@ -488,17 +488,15 @@ again:
 	/* start IO across the range first to instantiate any delalloc
 	 * extents
 	 */
-	btrfs_fdatawrite_range(inode->i_mapping, start, orig_end, WB_SYNC_ALL);
+	filemap_fdatawrite_range(inode->i_mapping, start, orig_end);
 
 	/* The compression code will leave pages locked but return from
 	 * writepage without setting the page writeback.  Starting again
 	 * with WB_SYNC_ALL will end up waiting for the IO to actually start.
 	 */
-	btrfs_fdatawrite_range(inode->i_mapping, start, orig_end, WB_SYNC_ALL);
+	filemap_fdatawrite_range(inode->i_mapping, start, orig_end);
 
-	btrfs_wait_on_page_writeback_range(inode->i_mapping,
-					   start >> PAGE_CACHE_SHIFT,
-					   orig_end >> PAGE_CACHE_SHIFT);
+	filemap_fdatawait_range(inode->i_mapping, start, orig_end);
 
 	end = orig_end;
 	found = 0;
@@ -716,89 +714,6 @@ out:
 }
 
 
-/**
- * taken from mm/filemap.c because it isn't exported
- *
- * __filemap_fdatawrite_range - start writeback on mapping dirty pages in range
- * @mapping:	address space structure to write
- * @start:	offset in bytes where the range starts
- * @end:	offset in bytes where the range ends (inclusive)
- * @sync_mode:	enable synchronous operation
- *
- * Start writeback against all of a mapping's dirty pages that lie
- * within the byte offsets <start, end> inclusive.
- *
- * If sync_mode is WB_SYNC_ALL then this is a "data integrity" operation, as
- * opposed to a regular memory cleansing writeback.  The difference between
- * these two operations is that if a dirty page/buffer is encountered, it must
- * be waited upon, and not just skipped over.
- */
-int btrfs_fdatawrite_range(struct address_space *mapping, loff_t start,
-			   loff_t end, int sync_mode)
-{
-	struct writeback_control wbc = {
-		.sync_mode = sync_mode,
-		.nr_to_write = mapping->nrpages * 2,
-		.range_start = start,
-		.range_end = end,
-	};
-	return btrfs_writepages(mapping, &wbc);
-}
-
-/**
- * taken from mm/filemap.c because it isn't exported
- *
- * wait_on_page_writeback_range - wait for writeback to complete
- * @mapping:	target address_space
- * @start:	beginning page index
- * @end:	ending page index
- *
- * Wait for writeback to complete against pages indexed by start->end
- * inclusive
- */
-int btrfs_wait_on_page_writeback_range(struct address_space *mapping,
-				       pgoff_t start, pgoff_t end)
-{
-	struct pagevec pvec;
-	int nr_pages;
-	int ret = 0;
-	pgoff_t index;
-
-	if (end < start)
-		return 0;
-
-	pagevec_init(&pvec, 0);
-	index = start;
-	while ((index <= end) &&
-			(nr_pages = pagevec_lookup_tag(&pvec, mapping, &index,
-			PAGECACHE_TAG_WRITEBACK,
-			min(end - index, (pgoff_t)PAGEVEC_SIZE-1) + 1)) != 0) {
-		unsigned i;
-
-		for (i = 0; i < nr_pages; i++) {
-			struct page *page = pvec.pages[i];
-
-			/* until radix tree lookup accepts end_index */
-			if (page->index > end)
-				continue;
-
-			wait_on_page_writeback(page);
-			if (PageError(page))
-				ret = -EIO;
-		}
-		pagevec_release(&pvec);
-		cond_resched();
-	}
-
-	/* Check for outstanding write errors */
-	if (test_and_clear_bit(AS_ENOSPC, &mapping->flags))
-		ret = -ENOSPC;
-	if (test_and_clear_bit(AS_EIO, &mapping->flags))
-		ret = -EIO;
-
-	return ret;
-}
-
 /*
  * add a given inode to the list of inodes that must be fully on
  * disk before a transaction commit finishes.
diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h
index 993a7ea45c7..f82e87488ca 100644
--- a/fs/btrfs/ordered-data.h
+++ b/fs/btrfs/ordered-data.h
@@ -153,10 +153,6 @@ btrfs_lookup_first_ordered_extent(struct inode * inode, u64 file_offset);
 int btrfs_ordered_update_i_size(struct inode *inode,
 				struct btrfs_ordered_extent *ordered);
 int btrfs_find_ordered_sum(struct inode *inode, u64 offset, u64 disk_bytenr, u32 *sum);
-int btrfs_wait_on_page_writeback_range(struct address_space *mapping,
-				       pgoff_t start, pgoff_t end);
-int btrfs_fdatawrite_range(struct address_space *mapping, loff_t start,
-			   loff_t end, int sync_mode);
 int btrfs_wait_ordered_extents(struct btrfs_root *root, int nocow_only);
 int btrfs_run_ordered_operations(struct btrfs_root *root, int wait);
 int btrfs_add_ordered_operation(struct btrfs_trans_handle *trans,
-- 
cgit v1.2.3-70-g09d2


From 828c09509b9695271bcbdc53e9fc9a6a737148d2 Mon Sep 17 00:00:00 2001
From: Alexey Dobriyan <adobriyan@gmail.com>
Date: Thu, 1 Oct 2009 15:43:56 -0700
Subject: const: constify remaining file_operations

[akpm@linux-foundation.org: fix KVM]
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
 arch/arm/mach-ns9xxx/clock.c               |  2 +-
 arch/blackfin/mach-bf561/coreb.c           |  2 +-
 arch/cris/arch-v10/drivers/sync_serial.c   |  2 +-
 arch/cris/arch-v32/drivers/mach-fs/gpio.c  |  2 +-
 arch/powerpc/kvm/timing.c                  |  2 +-
 arch/powerpc/platforms/cell/spufs/file.c   |  2 +-
 arch/powerpc/platforms/pseries/dtl.c       |  2 +-
 arch/x86/xen/debugfs.c                     |  2 +-
 drivers/acpi/video.c                       |  2 +-
 drivers/block/cciss.c                      |  2 +-
 drivers/char/apm-emulation.c               |  2 +-
 drivers/char/bfin-otp.c                    |  2 +-
 drivers/char/xilinx_hwicap/xilinx_hwicap.c |  2 +-
 drivers/gpio/gpiolib.c                     |  2 +-
 drivers/hwmon/fschmd.c                     |  2 +-
 drivers/lguest/lguest_user.c               |  2 +-
 drivers/media/dvb/dvb-core/dmxdev.c        |  2 +-
 drivers/media/dvb/firewire/firedtv-ci.c    |  2 +-
 drivers/misc/phantom.c                     |  2 +-
 drivers/misc/sgi-gru/grufile.c             |  3 +--
 drivers/mmc/core/debugfs.c                 |  2 +-
 drivers/s390/cio/qdio_debug.c              |  2 +-
 drivers/s390/cio/qdio_perf.c               |  2 +-
 drivers/scsi/sg.c                          | 43 +++++++++++++++++++++---------
 drivers/spi/spidev.c                       |  2 +-
 drivers/usb/class/usbtmc.c                 |  2 +-
 drivers/usb/gadget/printer.c               |  2 +-
 drivers/usb/host/whci/debug.c              |  6 ++---
 drivers/usb/misc/rio500.c                  |  3 +--
 drivers/uwb/uwb-debug.c                    |  6 ++---
 fs/btrfs/ctree.h                           |  2 +-
 fs/btrfs/file.c                            |  2 +-
 fs/btrfs/inode.c                           |  4 +--
 fs/jbd2/journal.c                          |  2 +-
 fs/nfsd/nfsctl.c                           |  2 +-
 fs/nilfs2/dir.c                            |  2 +-
 fs/nilfs2/file.c                           |  2 +-
 fs/nilfs2/mdt.c                            |  2 +-
 fs/nilfs2/nilfs.h                          |  4 +--
 fs/ocfs2/cluster/heartbeat.c               |  2 +-
 fs/ocfs2/cluster/netdebug.c                |  4 +--
 fs/ocfs2/dlm/dlmdebug.c                    |  8 +++---
 fs/ocfs2/super.c                           |  2 +-
 fs/omfs/dir.c                              |  2 +-
 fs/omfs/file.c                             |  2 +-
 fs/omfs/omfs.h                             |  4 +--
 include/linux/cgroup.h                     |  2 +-
 include/linux/fs.h                         |  2 +-
 kernel/cgroup.c                            | 10 +++----
 kernel/kprobes.c                           |  4 +--
 kernel/rcutree_trace.c                     | 10 +++----
 kernel/sched.c                             |  2 +-
 kernel/time/timer_list.c                   |  2 +-
 kernel/time/timer_stats.c                  |  2 +-
 samples/tracepoints/tracepoint-sample.c    |  2 +-
 security/integrity/ima/ima_fs.c            | 10 +++----
 virt/kvm/kvm_main.c                        |  2 +-
 57 files changed, 110 insertions(+), 95 deletions(-)

(limited to 'fs/btrfs/file.c')

diff --git a/arch/arm/mach-ns9xxx/clock.c b/arch/arm/mach-ns9xxx/clock.c
index 44ed20d4a38..cf81cbc5754 100644
--- a/arch/arm/mach-ns9xxx/clock.c
+++ b/arch/arm/mach-ns9xxx/clock.c
@@ -195,7 +195,7 @@ static int clk_debugfs_open(struct inode *inode, struct file *file)
 	return single_open(file, clk_debugfs_show, NULL);
 }
 
-static struct file_operations clk_debugfs_operations = {
+static const struct file_operations clk_debugfs_operations = {
 	.open = clk_debugfs_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
diff --git a/arch/blackfin/mach-bf561/coreb.c b/arch/blackfin/mach-bf561/coreb.c
index 93635a766f9..1e60a92dd60 100644
--- a/arch/blackfin/mach-bf561/coreb.c
+++ b/arch/blackfin/mach-bf561/coreb.c
@@ -48,7 +48,7 @@ coreb_ioctl(struct inode *inode, struct file *file, unsigned int cmd, unsigned l
 	return ret;
 }
 
-static struct file_operations coreb_fops = {
+static const struct file_operations coreb_fops = {
 	.owner   = THIS_MODULE,
 	.ioctl   = coreb_ioctl,
 };
diff --git a/arch/cris/arch-v10/drivers/sync_serial.c b/arch/cris/arch-v10/drivers/sync_serial.c
index 6cc1a0319a5..562b9a7feae 100644
--- a/arch/cris/arch-v10/drivers/sync_serial.c
+++ b/arch/cris/arch-v10/drivers/sync_serial.c
@@ -244,7 +244,7 @@ static unsigned sync_serial_prescale_shadow;
 
 #define NUMBER_OF_PORTS 2
 
-static struct file_operations sync_serial_fops = {
+static const struct file_operations sync_serial_fops = {
 	.owner   = THIS_MODULE,
 	.write   = sync_serial_write,
 	.read    = sync_serial_read,
diff --git a/arch/cris/arch-v32/drivers/mach-fs/gpio.c b/arch/cris/arch-v32/drivers/mach-fs/gpio.c
index fe1fde89388..d89ab80498e 100644
--- a/arch/cris/arch-v32/drivers/mach-fs/gpio.c
+++ b/arch/cris/arch-v32/drivers/mach-fs/gpio.c
@@ -855,7 +855,7 @@ gpio_leds_ioctl(unsigned int cmd, unsigned long arg)
 	return 0;
 }
 
-struct file_operations gpio_fops = {
+static const struct file_operations gpio_fops = {
 	.owner       = THIS_MODULE,
 	.poll        = gpio_poll,
 	.ioctl       = gpio_ioctl,
diff --git a/arch/powerpc/kvm/timing.c b/arch/powerpc/kvm/timing.c
index 47ee603f558..2aa371e3007 100644
--- a/arch/powerpc/kvm/timing.c
+++ b/arch/powerpc/kvm/timing.c
@@ -201,7 +201,7 @@ static int kvmppc_exit_timing_open(struct inode *inode, struct file *file)
 	return single_open(file, kvmppc_exit_timing_show, inode->i_private);
 }
 
-static struct file_operations kvmppc_exit_timing_fops = {
+static const struct file_operations kvmppc_exit_timing_fops = {
 	.owner   = THIS_MODULE,
 	.open    = kvmppc_exit_timing_open,
 	.read    = seq_read,
diff --git a/arch/powerpc/platforms/cell/spufs/file.c b/arch/powerpc/platforms/cell/spufs/file.c
index 96130944617..884e8bcec49 100644
--- a/arch/powerpc/platforms/cell/spufs/file.c
+++ b/arch/powerpc/platforms/cell/spufs/file.c
@@ -147,7 +147,7 @@ static int __fops ## _open(struct inode *inode, struct file *file)	\
 	__simple_attr_check_format(__fmt, 0ull);			\
 	return spufs_attr_open(inode, file, __get, __set, __fmt);	\
 }									\
-static struct file_operations __fops = {				\
+static const struct file_operations __fops = {				\
 	.owner	 = THIS_MODULE,						\
 	.open	 = __fops ## _open,					\
 	.release = spufs_attr_release,					\
diff --git a/arch/powerpc/platforms/pseries/dtl.c b/arch/powerpc/platforms/pseries/dtl.c
index ab69925d579..937a544a236 100644
--- a/arch/powerpc/platforms/pseries/dtl.c
+++ b/arch/powerpc/platforms/pseries/dtl.c
@@ -209,7 +209,7 @@ static ssize_t dtl_file_read(struct file *filp, char __user *buf, size_t len,
 	return n_read * sizeof(struct dtl_entry);
 }
 
-static struct file_operations dtl_fops = {
+static const struct file_operations dtl_fops = {
 	.open		= dtl_file_open,
 	.release	= dtl_file_release,
 	.read		= dtl_file_read,
diff --git a/arch/x86/xen/debugfs.c b/arch/x86/xen/debugfs.c
index b53225d2cac..e133ce25e29 100644
--- a/arch/x86/xen/debugfs.c
+++ b/arch/x86/xen/debugfs.c
@@ -100,7 +100,7 @@ static int xen_array_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static struct file_operations u32_array_fops = {
+static const struct file_operations u32_array_fops = {
 	.owner	= THIS_MODULE,
 	.open	= u32_array_open,
 	.release= xen_array_release,
diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c
index a4fddb24476..f6e54bf8dd9 100644
--- a/drivers/acpi/video.c
+++ b/drivers/acpi/video.c
@@ -285,7 +285,7 @@ static int acpi_video_device_brightness_open_fs(struct inode *inode,
 						struct file *file);
 static ssize_t acpi_video_device_write_brightness(struct file *file,
 	const char __user *buffer, size_t count, loff_t *data);
-static struct file_operations acpi_video_device_brightness_fops = {
+static const struct file_operations acpi_video_device_brightness_fops = {
 	.owner = THIS_MODULE,
 	.open = acpi_video_device_brightness_open_fs,
 	.read = seq_read,
diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c
index 24c3e21ab26..1ece0b47b58 100644
--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -426,7 +426,7 @@ out:
 	return err;
 }
 
-static struct file_operations cciss_proc_fops = {
+static const struct file_operations cciss_proc_fops = {
 	.owner	 = THIS_MODULE,
 	.open    = cciss_seq_open,
 	.read    = seq_read,
diff --git a/drivers/char/apm-emulation.c b/drivers/char/apm-emulation.c
index aaca40283be..4f568cb9af3 100644
--- a/drivers/char/apm-emulation.c
+++ b/drivers/char/apm-emulation.c
@@ -393,7 +393,7 @@ static int apm_open(struct inode * inode, struct file * filp)
 	return as ? 0 : -ENOMEM;
 }
 
-static struct file_operations apm_bios_fops = {
+static const struct file_operations apm_bios_fops = {
 	.owner		= THIS_MODULE,
 	.read		= apm_read,
 	.poll		= apm_poll,
diff --git a/drivers/char/bfin-otp.c b/drivers/char/bfin-otp.c
index e3dd24bff51..836d4f0a876 100644
--- a/drivers/char/bfin-otp.c
+++ b/drivers/char/bfin-otp.c
@@ -217,7 +217,7 @@ static long bfin_otp_ioctl(struct file *filp, unsigned cmd, unsigned long arg)
 # define bfin_otp_ioctl NULL
 #endif
 
-static struct file_operations bfin_otp_fops = {
+static const struct file_operations bfin_otp_fops = {
 	.owner          = THIS_MODULE,
 	.unlocked_ioctl = bfin_otp_ioctl,
 	.read           = bfin_otp_read,
diff --git a/drivers/char/xilinx_hwicap/xilinx_hwicap.c b/drivers/char/xilinx_hwicap/xilinx_hwicap.c
index f40ab699860..4846d50199f 100644
--- a/drivers/char/xilinx_hwicap/xilinx_hwicap.c
+++ b/drivers/char/xilinx_hwicap/xilinx_hwicap.c
@@ -559,7 +559,7 @@ static int hwicap_release(struct inode *inode, struct file *file)
 	return status;
 }
 
-static struct file_operations hwicap_fops = {
+static const struct file_operations hwicap_fops = {
 	.owner = THIS_MODULE,
 	.write = hwicap_write,
 	.read = hwicap_read,
diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index bb11a429394..662ed923d9e 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -1487,7 +1487,7 @@ static int gpiolib_open(struct inode *inode, struct file *file)
 	return single_open(file, gpiolib_show, NULL);
 }
 
-static struct file_operations gpiolib_operations = {
+static const struct file_operations gpiolib_operations = {
 	.open		= gpiolib_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
diff --git a/drivers/hwmon/fschmd.c b/drivers/hwmon/fschmd.c
index ea955edde87..2a7a85a6dc3 100644
--- a/drivers/hwmon/fschmd.c
+++ b/drivers/hwmon/fschmd.c
@@ -915,7 +915,7 @@ static int watchdog_ioctl(struct inode *inode, struct file *filp,
 	return ret;
 }
 
-static struct file_operations watchdog_fops = {
+static const struct file_operations watchdog_fops = {
 	.owner = THIS_MODULE,
 	.llseek = no_llseek,
 	.open = watchdog_open,
diff --git a/drivers/lguest/lguest_user.c b/drivers/lguest/lguest_user.c
index b4d3f7ca554..bd1632388e4 100644
--- a/drivers/lguest/lguest_user.c
+++ b/drivers/lguest/lguest_user.c
@@ -508,7 +508,7 @@ static int close(struct inode *inode, struct file *file)
  * uses: reading and writing a character device called /dev/lguest.  All the
  * work happens in the read(), write() and close() routines:
  */
-static struct file_operations lguest_fops = {
+static const struct file_operations lguest_fops = {
 	.owner	 = THIS_MODULE,
 	.release = close,
 	.write	 = write,
diff --git a/drivers/media/dvb/dvb-core/dmxdev.c b/drivers/media/dvb/dvb-core/dmxdev.c
index 3750ff48cba..51641498359 100644
--- a/drivers/media/dvb/dvb-core/dmxdev.c
+++ b/drivers/media/dvb/dvb-core/dmxdev.c
@@ -1203,7 +1203,7 @@ static unsigned int dvb_dvr_poll(struct file *file, poll_table *wait)
 	return mask;
 }
 
-static struct file_operations dvb_dvr_fops = {
+static const struct file_operations dvb_dvr_fops = {
 	.owner = THIS_MODULE,
 	.read = dvb_dvr_read,
 	.write = dvb_dvr_write,
diff --git a/drivers/media/dvb/firewire/firedtv-ci.c b/drivers/media/dvb/firewire/firedtv-ci.c
index eeb80d0ea3f..853e04b7cb3 100644
--- a/drivers/media/dvb/firewire/firedtv-ci.c
+++ b/drivers/media/dvb/firewire/firedtv-ci.c
@@ -215,7 +215,7 @@ static unsigned int fdtv_ca_io_poll(struct file *file, poll_table *wait)
 	return POLLIN;
 }
 
-static struct file_operations fdtv_ca_fops = {
+static const struct file_operations fdtv_ca_fops = {
 	.owner		= THIS_MODULE,
 	.ioctl		= dvb_generic_ioctl,
 	.open		= dvb_generic_open,
diff --git a/drivers/misc/phantom.c b/drivers/misc/phantom.c
index fa57b67593a..90a95ce8dc3 100644
--- a/drivers/misc/phantom.c
+++ b/drivers/misc/phantom.c
@@ -271,7 +271,7 @@ static unsigned int phantom_poll(struct file *file, poll_table *wait)
 	return mask;
 }
 
-static struct file_operations phantom_file_ops = {
+static const struct file_operations phantom_file_ops = {
 	.open = phantom_open,
 	.release = phantom_release,
 	.unlocked_ioctl = phantom_ioctl,
diff --git a/drivers/misc/sgi-gru/grufile.c b/drivers/misc/sgi-gru/grufile.c
index 300e7ba391a..41c8fe2a928 100644
--- a/drivers/misc/sgi-gru/grufile.c
+++ b/drivers/misc/sgi-gru/grufile.c
@@ -53,7 +53,6 @@ struct gru_stats_s gru_stats;
 /* Guaranteed user available resources on each node */
 static int max_user_cbrs, max_user_dsr_bytes;
 
-static struct file_operations gru_fops;
 static struct miscdevice gru_miscdev;
 
 
@@ -426,7 +425,7 @@ static void __exit gru_exit(void)
 	gru_proc_exit();
 }
 
-static struct file_operations gru_fops = {
+static const struct file_operations gru_fops = {
 	.owner		= THIS_MODULE,
 	.unlocked_ioctl	= gru_file_unlocked_ioctl,
 	.mmap		= gru_file_mmap,
diff --git a/drivers/mmc/core/debugfs.c b/drivers/mmc/core/debugfs.c
index 610dbd1fcc8..96d10f40fb2 100644
--- a/drivers/mmc/core/debugfs.c
+++ b/drivers/mmc/core/debugfs.c
@@ -240,7 +240,7 @@ static int mmc_ext_csd_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static struct file_operations mmc_dbg_ext_csd_fops = {
+static const struct file_operations mmc_dbg_ext_csd_fops = {
 	.open		= mmc_ext_csd_open,
 	.read		= mmc_ext_csd_read,
 	.release	= mmc_ext_csd_release,
diff --git a/drivers/s390/cio/qdio_debug.c b/drivers/s390/cio/qdio_debug.c
index 1b78f639ead..76769978285 100644
--- a/drivers/s390/cio/qdio_debug.c
+++ b/drivers/s390/cio/qdio_debug.c
@@ -125,7 +125,7 @@ static int qstat_seq_open(struct inode *inode, struct file *filp)
 			   filp->f_path.dentry->d_inode->i_private);
 }
 
-static struct file_operations debugfs_fops = {
+static const struct file_operations debugfs_fops = {
 	.owner	 = THIS_MODULE,
 	.open	 = qstat_seq_open,
 	.read	 = seq_read,
diff --git a/drivers/s390/cio/qdio_perf.c b/drivers/s390/cio/qdio_perf.c
index eff943923c6..968e3c7c263 100644
--- a/drivers/s390/cio/qdio_perf.c
+++ b/drivers/s390/cio/qdio_perf.c
@@ -84,7 +84,7 @@ static int qdio_perf_seq_open(struct inode *inode, struct file *filp)
 	return single_open(filp, qdio_perf_proc_show, NULL);
 }
 
-static struct file_operations qdio_perf_proc_fops = {
+static const struct file_operations qdio_perf_proc_fops = {
 	.owner	 = THIS_MODULE,
 	.open	 = qdio_perf_seq_open,
 	.read	 = seq_read,
diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c
index 0cb049f5cc5..747a5e5c127 100644
--- a/drivers/scsi/sg.c
+++ b/drivers/scsi/sg.c
@@ -1317,7 +1317,7 @@ static void sg_rq_end_io(struct request *rq, int uptodate)
 	}
 }
 
-static struct file_operations sg_fops = {
+static const struct file_operations sg_fops = {
 	.owner = THIS_MODULE,
 	.read = sg_read,
 	.write = sg_write,
@@ -2194,9 +2194,11 @@ static int sg_proc_seq_show_int(struct seq_file *s, void *v);
 static int sg_proc_single_open_adio(struct inode *inode, struct file *file);
 static ssize_t sg_proc_write_adio(struct file *filp, const char __user *buffer,
 			          size_t count, loff_t *off);
-static struct file_operations adio_fops = {
-	/* .owner, .read and .llseek added in sg_proc_init() */
+static const struct file_operations adio_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_single_open_adio,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.write = sg_proc_write_adio,
 	.release = single_release,
 };
@@ -2204,23 +2206,32 @@ static struct file_operations adio_fops = {
 static int sg_proc_single_open_dressz(struct inode *inode, struct file *file);
 static ssize_t sg_proc_write_dressz(struct file *filp, 
 		const char __user *buffer, size_t count, loff_t *off);
-static struct file_operations dressz_fops = {
+static const struct file_operations dressz_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_single_open_dressz,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.write = sg_proc_write_dressz,
 	.release = single_release,
 };
 
 static int sg_proc_seq_show_version(struct seq_file *s, void *v);
 static int sg_proc_single_open_version(struct inode *inode, struct file *file);
-static struct file_operations version_fops = {
+static const struct file_operations version_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_single_open_version,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.release = single_release,
 };
 
 static int sg_proc_seq_show_devhdr(struct seq_file *s, void *v);
 static int sg_proc_single_open_devhdr(struct inode *inode, struct file *file);
-static struct file_operations devhdr_fops = {
+static const struct file_operations devhdr_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_single_open_devhdr,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.release = single_release,
 };
 
@@ -2229,8 +2240,11 @@ static int sg_proc_open_dev(struct inode *inode, struct file *file);
 static void * dev_seq_start(struct seq_file *s, loff_t *pos);
 static void * dev_seq_next(struct seq_file *s, void *v, loff_t *pos);
 static void dev_seq_stop(struct seq_file *s, void *v);
-static struct file_operations dev_fops = {
+static const struct file_operations dev_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_open_dev,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.release = seq_release,
 };
 static const struct seq_operations dev_seq_ops = {
@@ -2242,8 +2256,11 @@ static const struct seq_operations dev_seq_ops = {
 
 static int sg_proc_seq_show_devstrs(struct seq_file *s, void *v);
 static int sg_proc_open_devstrs(struct inode *inode, struct file *file);
-static struct file_operations devstrs_fops = {
+static const struct file_operations devstrs_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_open_devstrs,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.release = seq_release,
 };
 static const struct seq_operations devstrs_seq_ops = {
@@ -2255,8 +2272,11 @@ static const struct seq_operations devstrs_seq_ops = {
 
 static int sg_proc_seq_show_debug(struct seq_file *s, void *v);
 static int sg_proc_open_debug(struct inode *inode, struct file *file);
-static struct file_operations debug_fops = {
+static const struct file_operations debug_fops = {
+	.owner = THIS_MODULE,
 	.open = sg_proc_open_debug,
+	.read = seq_read,
+	.llseek = seq_lseek,
 	.release = seq_release,
 };
 static const struct seq_operations debug_seq_ops = {
@@ -2269,7 +2289,7 @@ static const struct seq_operations debug_seq_ops = {
 
 struct sg_proc_leaf {
 	const char * name;
-	struct file_operations * fops;
+	const struct file_operations * fops;
 };
 
 static struct sg_proc_leaf sg_proc_leaf_arr[] = {
@@ -2295,9 +2315,6 @@ sg_proc_init(void)
 	for (k = 0; k < num_leaves; ++k) {
 		leaf = &sg_proc_leaf_arr[k];
 		mask = leaf->fops->write ? S_IRUGO | S_IWUSR : S_IRUGO;
-		leaf->fops->owner = THIS_MODULE;
-		leaf->fops->read = seq_read;
-		leaf->fops->llseek = seq_lseek;
 		proc_create(leaf->name, mask, sg_proc_sgp, leaf->fops);
 	}
 	return 0;
diff --git a/drivers/spi/spidev.c b/drivers/spi/spidev.c
index f921bd1109e..5d23983f02f 100644
--- a/drivers/spi/spidev.c
+++ b/drivers/spi/spidev.c
@@ -537,7 +537,7 @@ static int spidev_release(struct inode *inode, struct file *filp)
 	return status;
 }
 
-static struct file_operations spidev_fops = {
+static const struct file_operations spidev_fops = {
 	.owner =	THIS_MODULE,
 	/* REVISIT switch to aio primitives, so that userspace
 	 * gets more complete API coverage.  It'll simplify things
diff --git a/drivers/usb/class/usbtmc.c b/drivers/usb/class/usbtmc.c
index 333ee02e7b2..864f0ba6a34 100644
--- a/drivers/usb/class/usbtmc.c
+++ b/drivers/usb/class/usbtmc.c
@@ -993,7 +993,7 @@ skip_io_on_zombie:
 	return retval;
 }
 
-static struct file_operations fops = {
+static const struct file_operations fops = {
 	.owner		= THIS_MODULE,
 	.read		= usbtmc_read,
 	.write		= usbtmc_write,
diff --git a/drivers/usb/gadget/printer.c b/drivers/usb/gadget/printer.c
index 29500154d00..2d867fd2241 100644
--- a/drivers/usb/gadget/printer.c
+++ b/drivers/usb/gadget/printer.c
@@ -875,7 +875,7 @@ printer_ioctl(struct file *fd, unsigned int code, unsigned long arg)
 }
 
 /* used after endpoint configuration */
-static struct file_operations printer_io_operations = {
+static const struct file_operations printer_io_operations = {
 	.owner =	THIS_MODULE,
 	.open =		printer_open,
 	.read =		printer_read,
diff --git a/drivers/usb/host/whci/debug.c b/drivers/usb/host/whci/debug.c
index cf2d45946c5..2273c815941 100644
--- a/drivers/usb/host/whci/debug.c
+++ b/drivers/usb/host/whci/debug.c
@@ -134,7 +134,7 @@ static int pzl_open(struct inode *inode, struct file *file)
 	return single_open(file, pzl_print, inode->i_private);
 }
 
-static struct file_operations di_fops = {
+static const struct file_operations di_fops = {
 	.open    = di_open,
 	.read    = seq_read,
 	.llseek  = seq_lseek,
@@ -142,7 +142,7 @@ static struct file_operations di_fops = {
 	.owner   = THIS_MODULE,
 };
 
-static struct file_operations asl_fops = {
+static const struct file_operations asl_fops = {
 	.open    = asl_open,
 	.read    = seq_read,
 	.llseek  = seq_lseek,
@@ -150,7 +150,7 @@ static struct file_operations asl_fops = {
 	.owner   = THIS_MODULE,
 };
 
-static struct file_operations pzl_fops = {
+static const struct file_operations pzl_fops = {
 	.open    = pzl_open,
 	.read    = seq_read,
 	.llseek  = seq_lseek,
diff --git a/drivers/usb/misc/rio500.c b/drivers/usb/misc/rio500.c
index d645f3899fe..32d0199d0c3 100644
--- a/drivers/usb/misc/rio500.c
+++ b/drivers/usb/misc/rio500.c
@@ -429,8 +429,7 @@ read_rio(struct file *file, char __user *buffer, size_t count, loff_t * ppos)
 	return read_count;
 }
 
-static struct
-file_operations usb_rio_fops = {
+static const struct file_operations usb_rio_fops = {
 	.owner =	THIS_MODULE,
 	.read =		read_rio,
 	.write =	write_rio,
diff --git a/drivers/uwb/uwb-debug.c b/drivers/uwb/uwb-debug.c
index 4a42993700c..2eecec0c13c 100644
--- a/drivers/uwb/uwb-debug.c
+++ b/drivers/uwb/uwb-debug.c
@@ -205,7 +205,7 @@ static ssize_t command_write(struct file *file, const char __user *buf,
 	return ret < 0 ? ret : len;
 }
 
-static struct file_operations command_fops = {
+static const struct file_operations command_fops = {
 	.open   = command_open,
 	.write  = command_write,
 	.read   = NULL,
@@ -255,7 +255,7 @@ static int reservations_open(struct inode *inode, struct file *file)
 	return single_open(file, reservations_print, inode->i_private);
 }
 
-static struct file_operations reservations_fops = {
+static const struct file_operations reservations_fops = {
 	.open    = reservations_open,
 	.read    = seq_read,
 	.llseek  = seq_lseek,
@@ -283,7 +283,7 @@ static int drp_avail_open(struct inode *inode, struct file *file)
 	return single_open(file, drp_avail_print, inode->i_private);
 }
 
-static struct file_operations drp_avail_fops = {
+static const struct file_operations drp_avail_fops = {
 	.open    = drp_avail_open,
 	.read    = seq_read,
 	.llseek  = seq_lseek,
diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 80599b4e42b..4484eb3408a 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -2326,7 +2326,7 @@ int btrfs_sync_file(struct file *file, struct dentry *dentry, int datasync);
 int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end,
 			    int skip_pinned);
 int btrfs_check_file(struct btrfs_root *root, struct inode *inode);
-extern struct file_operations btrfs_file_operations;
+extern const struct file_operations btrfs_file_operations;
 int btrfs_drop_extents(struct btrfs_trans_handle *trans,
 		       struct btrfs_root *root, struct inode *inode,
 		       u64 start, u64 end, u64 locked_end,
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index a3492a3ad96..9ed17dbe5c6 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1196,7 +1196,7 @@ static int btrfs_file_mmap(struct file	*filp, struct vm_area_struct *vma)
 	return 0;
 }
 
-struct file_operations btrfs_file_operations = {
+const struct file_operations btrfs_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= do_sync_read,
 	.aio_read       = generic_file_aio_read,
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index e9b76bcd1c1..b9fe06d751c 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -62,7 +62,7 @@ static const struct inode_operations btrfs_special_inode_operations;
 static const struct inode_operations btrfs_file_inode_operations;
 static const struct address_space_operations btrfs_aops;
 static const struct address_space_operations btrfs_symlink_aops;
-static struct file_operations btrfs_dir_file_operations;
+static const struct file_operations btrfs_dir_file_operations;
 static struct extent_io_ops btrfs_extent_io_ops;
 
 static struct kmem_cache *btrfs_inode_cachep;
@@ -5544,7 +5544,7 @@ static const struct inode_operations btrfs_dir_ro_inode_operations = {
 	.permission	= btrfs_permission,
 };
 
-static struct file_operations btrfs_dir_file_operations = {
+static const struct file_operations btrfs_dir_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
 	.readdir	= btrfs_real_readdir,
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 761af77491f..b0ab5219bec 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -770,7 +770,7 @@ static int jbd2_seq_info_release(struct inode *inode, struct file *file)
 	return seq_release(inode, file);
 }
 
-static struct file_operations jbd2_seq_info_fops = {
+static const struct file_operations jbd2_seq_info_fops = {
 	.owner		= THIS_MODULE,
 	.open           = jbd2_seq_info_open,
 	.read           = seq_read,
diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c
index 00388d2a3c9..5c01fc148ce 100644
--- a/fs/nfsd/nfsctl.c
+++ b/fs/nfsd/nfsctl.c
@@ -176,7 +176,7 @@ static const struct file_operations exports_operations = {
 extern int nfsd_pool_stats_open(struct inode *inode, struct file *file);
 extern int nfsd_pool_stats_release(struct inode *inode, struct file *file);
 
-static struct file_operations pool_stats_operations = {
+static const struct file_operations pool_stats_operations = {
 	.open		= nfsd_pool_stats_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c
index 1a4fa04cf07..e097099bfc8 100644
--- a/fs/nilfs2/dir.c
+++ b/fs/nilfs2/dir.c
@@ -697,7 +697,7 @@ not_empty:
 	return 0;
 }
 
-struct file_operations nilfs_dir_operations = {
+const struct file_operations nilfs_dir_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= generic_read_dir,
 	.readdir	= nilfs_readdir,
diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c
index 7d7b4983dee..30292df443c 100644
--- a/fs/nilfs2/file.c
+++ b/fs/nilfs2/file.c
@@ -134,7 +134,7 @@ static int nilfs_file_mmap(struct file *file, struct vm_area_struct *vma)
  * We have mostly NULL's here: the current defaults are ok for
  * the nilfs filesystem.
  */
-struct file_operations nilfs_file_operations = {
+const struct file_operations nilfs_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read		= do_sync_read,
 	.write		= do_sync_write,
diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c
index b18c4998f8d..f6326112d64 100644
--- a/fs/nilfs2/mdt.c
+++ b/fs/nilfs2/mdt.c
@@ -433,7 +433,7 @@ static const struct address_space_operations def_mdt_aops = {
 };
 
 static const struct inode_operations def_mdt_iops;
-static struct file_operations def_mdt_fops;
+static const struct file_operations def_mdt_fops;
 
 /*
  * NILFS2 uses pseudo inodes for meta data files such as DAT, cpfile, sufile,
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index bad7368782d..4da6f67e9a9 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -294,9 +294,9 @@ void nilfs_clear_gcdat_inode(struct the_nilfs *);
 /*
  * Inodes and files operations
  */
-extern struct file_operations nilfs_dir_operations;
+extern const struct file_operations nilfs_dir_operations;
 extern const struct inode_operations nilfs_file_inode_operations;
-extern struct file_operations nilfs_file_operations;
+extern const struct file_operations nilfs_file_operations;
 extern const struct address_space_operations nilfs_aops;
 extern const struct inode_operations nilfs_dir_inode_operations;
 extern const struct inode_operations nilfs_special_inode_operations;
diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c
index 09cc25d0461..c452d116b89 100644
--- a/fs/ocfs2/cluster/heartbeat.c
+++ b/fs/ocfs2/cluster/heartbeat.c
@@ -966,7 +966,7 @@ static ssize_t o2hb_debug_read(struct file *file, char __user *buf,
 }
 #endif  /* CONFIG_DEBUG_FS */
 
-static struct file_operations o2hb_debug_fops = {
+static const struct file_operations o2hb_debug_fops = {
 	.open =		o2hb_debug_open,
 	.release =	o2hb_debug_release,
 	.read =		o2hb_debug_read,
diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c
index cfb2be708ab..da794bc07a6 100644
--- a/fs/ocfs2/cluster/netdebug.c
+++ b/fs/ocfs2/cluster/netdebug.c
@@ -207,7 +207,7 @@ static int nst_fop_release(struct inode *inode, struct file *file)
 	return seq_release_private(inode, file);
 }
 
-static struct file_operations nst_seq_fops = {
+static const struct file_operations nst_seq_fops = {
 	.open = nst_fop_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
@@ -388,7 +388,7 @@ static int sc_fop_release(struct inode *inode, struct file *file)
 	return seq_release_private(inode, file);
 }
 
-static struct file_operations sc_seq_fops = {
+static const struct file_operations sc_seq_fops = {
 	.open = sc_fop_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c
index ca46002ec10..42b0bad7a61 100644
--- a/fs/ocfs2/dlm/dlmdebug.c
+++ b/fs/ocfs2/dlm/dlmdebug.c
@@ -478,7 +478,7 @@ bail:
 	return -ENOMEM;
 }
 
-static struct file_operations debug_purgelist_fops = {
+static const struct file_operations debug_purgelist_fops = {
 	.open =		debug_purgelist_open,
 	.release =	debug_buffer_release,
 	.read =		debug_buffer_read,
@@ -538,7 +538,7 @@ bail:
 	return -ENOMEM;
 }
 
-static struct file_operations debug_mle_fops = {
+static const struct file_operations debug_mle_fops = {
 	.open =		debug_mle_open,
 	.release =	debug_buffer_release,
 	.read =		debug_buffer_read,
@@ -741,7 +741,7 @@ static int debug_lockres_release(struct inode *inode, struct file *file)
 	return seq_release_private(inode, file);
 }
 
-static struct file_operations debug_lockres_fops = {
+static const struct file_operations debug_lockres_fops = {
 	.open =		debug_lockres_open,
 	.release =	debug_lockres_release,
 	.read =		seq_read,
@@ -925,7 +925,7 @@ bail:
 	return -ENOMEM;
 }
 
-static struct file_operations debug_state_fops = {
+static const struct file_operations debug_state_fops = {
 	.open =		debug_state_open,
 	.release =	debug_buffer_release,
 	.read =		debug_buffer_read,
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index 4cc3c890a2c..c0e48aeebb1 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -373,7 +373,7 @@ static ssize_t ocfs2_debug_read(struct file *file, char __user *buf,
 }
 #endif	/* CONFIG_DEBUG_FS */
 
-static struct file_operations ocfs2_osb_debug_fops = {
+static const struct file_operations ocfs2_osb_debug_fops = {
 	.open =		ocfs2_osb_debug_open,
 	.release =	ocfs2_debug_release,
 	.read =		ocfs2_debug_read,
diff --git a/fs/omfs/dir.c b/fs/omfs/dir.c
index 3680bae335b..b42d6241903 100644
--- a/fs/omfs/dir.c
+++ b/fs/omfs/dir.c
@@ -498,7 +498,7 @@ const struct inode_operations omfs_dir_inops = {
 	.rmdir = omfs_rmdir,
 };
 
-struct file_operations omfs_dir_operations = {
+const struct file_operations omfs_dir_operations = {
 	.read = generic_read_dir,
 	.readdir = omfs_readdir,
 	.llseek = generic_file_llseek,
diff --git a/fs/omfs/file.c b/fs/omfs/file.c
index 4845fbb18e6..399487c0936 100644
--- a/fs/omfs/file.c
+++ b/fs/omfs/file.c
@@ -322,7 +322,7 @@ static sector_t omfs_bmap(struct address_space *mapping, sector_t block)
 	return generic_block_bmap(mapping, block, omfs_get_block);
 }
 
-struct file_operations omfs_file_operations = {
+const struct file_operations omfs_file_operations = {
 	.llseek = generic_file_llseek,
 	.read = do_sync_read,
 	.write = do_sync_write,
diff --git a/fs/omfs/omfs.h b/fs/omfs/omfs.h
index df71039945a..ebe2fdbe535 100644
--- a/fs/omfs/omfs.h
+++ b/fs/omfs/omfs.h
@@ -44,14 +44,14 @@ extern int omfs_allocate_range(struct super_block *sb, int min_request,
 extern int omfs_clear_range(struct super_block *sb, u64 block, int count);
 
 /* dir.c */
-extern struct file_operations omfs_dir_operations;
+extern const struct file_operations omfs_dir_operations;
 extern const struct inode_operations omfs_dir_inops;
 extern int omfs_make_empty(struct inode *inode, struct super_block *sb);
 extern int omfs_is_bad(struct omfs_sb_info *sbi, struct omfs_header *header,
 			u64 fsblock);
 
 /* file.c */
-extern struct file_operations omfs_file_operations;
+extern const struct file_operations omfs_file_operations;
 extern const struct inode_operations omfs_file_inops;
 extern const struct address_space_operations omfs_aops;
 extern void omfs_make_empty_table(struct buffer_head *bh, int offset);
diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h
index b62bb9294d0..0008dee6651 100644
--- a/include/linux/cgroup.h
+++ b/include/linux/cgroup.h
@@ -37,7 +37,7 @@ extern void cgroup_exit(struct task_struct *p, int run_callbacks);
 extern int cgroupstats_build(struct cgroupstats *stats,
 				struct dentry *dentry);
 
-extern struct file_operations proc_cgroup_operations;
+extern const struct file_operations proc_cgroup_operations;
 
 /* Define the enumeration of all cgroup subsystems */
 #define SUBSYS(_x) _x ## _subsys_id,
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 2adaa2529f1..a1e6899d4b6 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -2446,7 +2446,7 @@ static int __fops ## _open(struct inode *inode, struct file *file)	\
 	__simple_attr_check_format(__fmt, 0ull);			\
 	return simple_attr_open(inode, file, __get, __set, __fmt);	\
 }									\
-static struct file_operations __fops = {				\
+static const struct file_operations __fops = {				\
 	.owner	 = THIS_MODULE,						\
 	.open	 = __fops ## _open,					\
 	.release = simple_attr_release,					\
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 7ccba4bc5e3..d2b88596efd 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -703,7 +703,7 @@ static int cgroup_mkdir(struct inode *dir, struct dentry *dentry, int mode);
 static int cgroup_rmdir(struct inode *unused_dir, struct dentry *dentry);
 static int cgroup_populate_dir(struct cgroup *cgrp);
 static const struct inode_operations cgroup_dir_inode_operations;
-static struct file_operations proc_cgroupstats_operations;
+static const struct file_operations proc_cgroupstats_operations;
 
 static struct backing_dev_info cgroup_backing_dev_info = {
 	.name		= "cgroup",
@@ -1863,7 +1863,7 @@ static int cgroup_seqfile_release(struct inode *inode, struct file *file)
 	return single_release(inode, file);
 }
 
-static struct file_operations cgroup_seqfile_operations = {
+static const struct file_operations cgroup_seqfile_operations = {
 	.read = seq_read,
 	.write = cgroup_file_write,
 	.llseek = seq_lseek,
@@ -1922,7 +1922,7 @@ static int cgroup_rename(struct inode *old_dir, struct dentry *old_dentry,
 	return simple_rename(old_dir, old_dentry, new_dir, new_dentry);
 }
 
-static struct file_operations cgroup_file_operations = {
+static const struct file_operations cgroup_file_operations = {
 	.read = cgroup_file_read,
 	.write = cgroup_file_write,
 	.llseek = generic_file_llseek,
@@ -3369,7 +3369,7 @@ static int cgroup_open(struct inode *inode, struct file *file)
 	return single_open(file, proc_cgroup_show, pid);
 }
 
-struct file_operations proc_cgroup_operations = {
+const struct file_operations proc_cgroup_operations = {
 	.open		= cgroup_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
@@ -3398,7 +3398,7 @@ static int cgroupstats_open(struct inode *inode, struct file *file)
 	return single_open(file, proc_cgroupstats_show, NULL);
 }
 
-static struct file_operations proc_cgroupstats_operations = {
+static const struct file_operations proc_cgroupstats_operations = {
 	.open = cgroupstats_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
diff --git a/kernel/kprobes.c b/kernel/kprobes.c
index cfadc1291d0..5240d75f4c6 100644
--- a/kernel/kprobes.c
+++ b/kernel/kprobes.c
@@ -1333,7 +1333,7 @@ static int __kprobes kprobes_open(struct inode *inode, struct file *filp)
 	return seq_open(filp, &kprobes_seq_ops);
 }
 
-static struct file_operations debugfs_kprobes_operations = {
+static const struct file_operations debugfs_kprobes_operations = {
 	.open           = kprobes_open,
 	.read           = seq_read,
 	.llseek         = seq_lseek,
@@ -1515,7 +1515,7 @@ static ssize_t write_enabled_file_bool(struct file *file,
 	return count;
 }
 
-static struct file_operations fops_kp = {
+static const struct file_operations fops_kp = {
 	.read =         read_enabled_file_bool,
 	.write =        write_enabled_file_bool,
 };
diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c
index c89f5e9fd17..179e6ad80dc 100644
--- a/kernel/rcutree_trace.c
+++ b/kernel/rcutree_trace.c
@@ -93,7 +93,7 @@ static int rcudata_open(struct inode *inode, struct file *file)
 	return single_open(file, show_rcudata, NULL);
 }
 
-static struct file_operations rcudata_fops = {
+static const struct file_operations rcudata_fops = {
 	.owner = THIS_MODULE,
 	.open = rcudata_open,
 	.read = seq_read,
@@ -145,7 +145,7 @@ static int rcudata_csv_open(struct inode *inode, struct file *file)
 	return single_open(file, show_rcudata_csv, NULL);
 }
 
-static struct file_operations rcudata_csv_fops = {
+static const struct file_operations rcudata_csv_fops = {
 	.owner = THIS_MODULE,
 	.open = rcudata_csv_open,
 	.read = seq_read,
@@ -196,7 +196,7 @@ static int rcuhier_open(struct inode *inode, struct file *file)
 	return single_open(file, show_rcuhier, NULL);
 }
 
-static struct file_operations rcuhier_fops = {
+static const struct file_operations rcuhier_fops = {
 	.owner = THIS_MODULE,
 	.open = rcuhier_open,
 	.read = seq_read,
@@ -222,7 +222,7 @@ static int rcugp_open(struct inode *inode, struct file *file)
 	return single_open(file, show_rcugp, NULL);
 }
 
-static struct file_operations rcugp_fops = {
+static const struct file_operations rcugp_fops = {
 	.owner = THIS_MODULE,
 	.open = rcugp_open,
 	.read = seq_read,
@@ -276,7 +276,7 @@ static int rcu_pending_open(struct inode *inode, struct file *file)
 	return single_open(file, show_rcu_pending, NULL);
 }
 
-static struct file_operations rcu_pending_fops = {
+static const struct file_operations rcu_pending_fops = {
 	.owner = THIS_MODULE,
 	.open = rcu_pending_open,
 	.read = seq_read,
diff --git a/kernel/sched.c b/kernel/sched.c
index ee61f454a98..1535f3884b8 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -780,7 +780,7 @@ static int sched_feat_open(struct inode *inode, struct file *filp)
 	return single_open(filp, sched_feat_show, NULL);
 }
 
-static struct file_operations sched_feat_fops = {
+static const struct file_operations sched_feat_fops = {
 	.open		= sched_feat_open,
 	.write		= sched_feat_write,
 	.read		= seq_read,
diff --git a/kernel/time/timer_list.c b/kernel/time/timer_list.c
index fddd69d16e0..1b5b7aa2fdf 100644
--- a/kernel/time/timer_list.c
+++ b/kernel/time/timer_list.c
@@ -275,7 +275,7 @@ static int timer_list_open(struct inode *inode, struct file *filp)
 	return single_open(filp, timer_list_show, NULL);
 }
 
-static struct file_operations timer_list_fops = {
+static const struct file_operations timer_list_fops = {
 	.open		= timer_list_open,
 	.read		= seq_read,
 	.llseek		= seq_lseek,
diff --git a/kernel/time/timer_stats.c b/kernel/time/timer_stats.c
index 4cde8b9c716..ee5681f8d7e 100644
--- a/kernel/time/timer_stats.c
+++ b/kernel/time/timer_stats.c
@@ -395,7 +395,7 @@ static int tstats_open(struct inode *inode, struct file *filp)
 	return single_open(filp, tstats_show, NULL);
 }
 
-static struct file_operations tstats_fops = {
+static const struct file_operations tstats_fops = {
 	.open		= tstats_open,
 	.read		= seq_read,
 	.write		= tstats_write,
diff --git a/samples/tracepoints/tracepoint-sample.c b/samples/tracepoints/tracepoint-sample.c
index 9cf80a11e8b..26fab33ffa8 100644
--- a/samples/tracepoints/tracepoint-sample.c
+++ b/samples/tracepoints/tracepoint-sample.c
@@ -28,7 +28,7 @@ static int my_open(struct inode *inode, struct file *file)
 	return -EPERM;
 }
 
-static struct file_operations mark_ops = {
+static const struct file_operations mark_ops = {
 	.open = my_open,
 };
 
diff --git a/security/integrity/ima/ima_fs.c b/security/integrity/ima/ima_fs.c
index 8e9777b7640..0c72c9c3895 100644
--- a/security/integrity/ima/ima_fs.c
+++ b/security/integrity/ima/ima_fs.c
@@ -43,7 +43,7 @@ static ssize_t ima_show_htable_violations(struct file *filp,
 	return ima_show_htable_value(buf, count, ppos, &ima_htable.violations);
 }
 
-static struct file_operations ima_htable_violations_ops = {
+static const struct file_operations ima_htable_violations_ops = {
 	.read = ima_show_htable_violations
 };
 
@@ -55,7 +55,7 @@ static ssize_t ima_show_measurements_count(struct file *filp,
 
 }
 
-static struct file_operations ima_measurements_count_ops = {
+static const struct file_operations ima_measurements_count_ops = {
 	.read = ima_show_measurements_count
 };
 
@@ -158,7 +158,7 @@ static int ima_measurements_open(struct inode *inode, struct file *file)
 	return seq_open(file, &ima_measurments_seqops);
 }
 
-static struct file_operations ima_measurements_ops = {
+static const struct file_operations ima_measurements_ops = {
 	.open = ima_measurements_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
@@ -233,7 +233,7 @@ static int ima_ascii_measurements_open(struct inode *inode, struct file *file)
 	return seq_open(file, &ima_ascii_measurements_seqops);
 }
 
-static struct file_operations ima_ascii_measurements_ops = {
+static const struct file_operations ima_ascii_measurements_ops = {
 	.open = ima_ascii_measurements_open,
 	.read = seq_read,
 	.llseek = seq_lseek,
@@ -313,7 +313,7 @@ static int ima_release_policy(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static struct file_operations ima_measure_policy_ops = {
+static const struct file_operations ima_measure_policy_ops = {
 	.open = ima_open_policy,
 	.write = ima_write_policy,
 	.release = ima_release_policy
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index b5e7e3f1183..e79c54034bc 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2625,7 +2625,7 @@ static int vcpu_stat_get(void *_offset, u64 *val)
 
 DEFINE_SIMPLE_ATTRIBUTE(vcpu_stat_fops, vcpu_stat_get, NULL, "%llu\n");
 
-static struct file_operations *stat_fops[] = {
+static const struct file_operations *stat_fops[] = {
 	[KVM_STAT_VCPU] = &vcpu_stat_fops,
 	[KVM_STAT_VM]   = &vm_stat_fops,
 };
-- 
cgit v1.2.3-70-g09d2