1 files changed, 112 insertions, 159 deletions

diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 5440bab23635..2bfdc641d4e3 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -36,6 +36,7 @@
 #include <linux/xattr.h>
 #include <linux/posix_acl.h>
 #include <linux/falloc.h>
+#include <linux/slab.h>
 #include "compat.h"
 #include "ctree.h"
 #include "disk-io.h"
@@ -379,7 +380,8 @@ again:
          * change at any time if we discover bad compression ratios.
          */
         if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS) &&
-            btrfs_test_opt(root, COMPRESS)) {
+            (btrfs_test_opt(root, COMPRESS) ||
+             (BTRFS_I(inode)->force_compress))) {
                 WARN_ON(pages);
                 pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS);
 
@@ -483,7 +485,10 @@ again:
                 nr_pages_ret = 0;
 
                 /* flag the file so we don't compress in the future */
-                BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
+                if (!btrfs_test_opt(root, FORCE_COMPRESS) &&
+                    !(BTRFS_I(inode)->force_compress)) {
+                        BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
+                }
         }
         if (will_compress) {
                 *num_added += 1;
@@ -569,8 +574,8 @@ retry:
                         unsigned long nr_written = 0;
 
                         lock_extent(io_tree, async_extent->start,
-                                    async_extent->start +
-                                    async_extent->ram_size - 1, GFP_NOFS);
+                                         async_extent->start +
+                                         async_extent->ram_size - 1, GFP_NOFS);
 
                         /* allocate blocks */
                         ret = cow_file_range(inode, async_cow->locked_page,
@@ -792,7 +797,7 @@ static noinline int cow_file_range(struct inode *inode,
         while (disk_num_bytes > 0) {
                 unsigned long op;
 
-                cur_alloc_size = min(disk_num_bytes, root->fs_info->max_extent);
+                cur_alloc_size = disk_num_bytes;
                 ret = btrfs_reserve_extent(trans, root, cur_alloc_size,
                                            root->sectorsize, 0, alloc_hint,
                                            (u64)-1, &ins, 1);
@@ -1210,7 +1215,8 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
         else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC)
                 ret = run_delalloc_nocow(inode, locked_page, start, end,
                                          page_started, 0, nr_written);
-        else if (!btrfs_test_opt(root, COMPRESS))
+        else if (!btrfs_test_opt(root, COMPRESS) &&
+                 !(BTRFS_I(inode)->force_compress))
                 ret = cow_file_range(inode, locked_page, start, end,
                                       page_started, nr_written, 1);
         else
@@ -1222,30 +1228,9 @@ static int run_delalloc_range(struct inode *inode, struct page *locked_page,
 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 oustanding_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;
-        }
-
         spin_lock(&BTRFS_I(inode)->accounting_lock);
         BTRFS_I(inode)->outstanding_extents++;
         spin_unlock(&BTRFS_I(inode)->accounting_lock);
@@ -1263,38 +1248,10 @@ 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) {
-                spin_lock(&BTRFS_I(inode)->accounting_lock);
-                BTRFS_I(inode)->outstanding_extents--;
-                spin_unlock(&BTRFS_I(inode)->accounting_lock);
-                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;
-
         spin_lock(&BTRFS_I(inode)->accounting_lock);
         BTRFS_I(inode)->outstanding_extents--;
         spin_unlock(&BTRFS_I(inode)->accounting_lock);
@@ -1323,6 +1280,7 @@ static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
                 BTRFS_I(inode)->outstanding_extents++;
                 spin_unlock(&BTRFS_I(inode)->accounting_lock);
                 btrfs_delalloc_reserve_space(root, inode, end - start + 1);
+
                 spin_lock(&root->fs_info->delalloc_lock);
                 BTRFS_I(inode)->delalloc_bytes += end - start + 1;
                 root->fs_info->delalloc_bytes += end - start + 1;
@@ -1351,6 +1309,7 @@ static int btrfs_clear_bit_hook(struct inode *inode,
 
                 if (bits & EXTENT_DO_ACCOUNTING) {
                         spin_lock(&BTRFS_I(inode)->accounting_lock);
+                        WARN_ON(!BTRFS_I(inode)->outstanding_extents);
                         BTRFS_I(inode)->outstanding_extents--;
                         spin_unlock(&BTRFS_I(inode)->accounting_lock);
                         btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
@@ -1507,12 +1466,13 @@ static noinline int add_pending_csums(struct btrfs_trans_handle *trans,
         return 0;
 }
 
-int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end)
+int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end,
+                              struct extent_state **cached_state)
 {
         if ((end & (PAGE_CACHE_SIZE - 1)) == 0)
                 WARN_ON(1);
         return set_extent_delalloc(&BTRFS_I(inode)->io_tree, start, end,
-                                   GFP_NOFS);
+                                   cached_state, GFP_NOFS);
 }
 
 /* see btrfs_writepage_start_hook for details on why this is required */
@@ -1525,6 +1485,7 @@ static void btrfs_writepage_fixup_worker(struct btrfs_work *work)
 {
         struct btrfs_writepage_fixup *fixup;
         struct btrfs_ordered_extent *ordered;
+        struct extent_state *cached_state = NULL;
         struct page *page;
         struct inode *inode;
         u64 page_start;
@@ -1543,7 +1504,8 @@ again:
         page_start = page_offset(page);
         page_end = page_offset(page) + PAGE_CACHE_SIZE - 1;
 
-        lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end, GFP_NOFS);
+        lock_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end, 0,
+                         &cached_state, GFP_NOFS);
 
         /* already ordered? We're done */
         if (PagePrivate2(page))
@@ -1551,17 +1513,18 @@ again:
 
         ordered = btrfs_lookup_ordered_extent(inode, page_start);
         if (ordered) {
-                unlock_extent(&BTRFS_I(inode)->io_tree, page_start,
-                              page_end, GFP_NOFS);
+                unlock_extent_cached(&BTRFS_I(inode)->io_tree, page_start,
+                                     page_end, &cached_state, GFP_NOFS);
                 unlock_page(page);
                 btrfs_start_ordered_extent(inode, ordered, 1);
                 goto again;
         }
 
-        btrfs_set_extent_delalloc(inode, page_start, page_end);
+        btrfs_set_extent_delalloc(inode, page_start, page_end, &cached_state);
         ClearPageChecked(page);
 out:
-        unlock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end, GFP_NOFS);
+        unlock_extent_cached(&BTRFS_I(inode)->io_tree, page_start, page_end,
+                             &cached_state, GFP_NOFS);
 out_page:
         unlock_page(page);
         page_cache_release(page);
@@ -1680,24 +1643,6 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
  * before we start the transaction.  It limits the amount of btree
  * reads required while inside the transaction.
  */
-static noinline void reada_csum(struct btrfs_root *root,
-                                struct btrfs_path *path,
-                                struct btrfs_ordered_extent *ordered_extent)
-{
-        struct btrfs_ordered_sum *sum;
-        u64 bytenr;
-
-        sum = list_entry(ordered_extent->list.next, struct btrfs_ordered_sum,
-                         list);
-        bytenr = sum->sums[0].bytenr;
-
-        /*
-         * we don't care about the results, the point of this search is
-         * just to get the btree leaves into ram
-         */
-        btrfs_lookup_csum(NULL, root->fs_info->csum_root, path, bytenr, 0);
-}
-
 /* as ordered data IO finishes, this gets called so we can finish
  * an ordered extent if the range of bytes in the file it covers are
  * fully written.
@@ -1708,40 +1653,16 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
         struct btrfs_trans_handle *trans;
         struct btrfs_ordered_extent *ordered_extent = NULL;
         struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
-        struct btrfs_path *path;
+        struct extent_state *cached_state = NULL;
         int compressed = 0;
         int ret;
 
-        ret = btrfs_dec_test_ordered_pending(inode, start, end - start + 1);
+        ret = btrfs_dec_test_ordered_pending(inode, &ordered_extent, start,
+                                             end - start + 1);
         if (!ret)
                 return 0;
-
-        /*
-         * before we join the transaction, try to do some of our IO.
-         * This will limit the amount of IO that we have to do with
-         * the transaction running.  We're unlikely to need to do any
-         * IO if the file extents are new, the disk_i_size checks
-         * covers the most common case.
-         */
-        if (start < BTRFS_I(inode)->disk_i_size) {
-                path = btrfs_alloc_path();
-                if (path) {
-                        ret = btrfs_lookup_file_extent(NULL, root, path,
-                                                       inode->i_ino,
-                                                       start, 0);
-                        ordered_extent = btrfs_lookup_ordered_extent(inode,
-                                                                     start);
-                        if (!list_empty(&ordered_extent->list)) {
-                                btrfs_release_path(root, path);
-                                reada_csum(root, path, ordered_extent);
-                        }
-                        btrfs_free_path(path);
-                }
-        }
-
-        if (!ordered_extent)
-                ordered_extent = btrfs_lookup_ordered_extent(inode, start);
         BUG_ON(!ordered_extent);
+
         if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) {
                 BUG_ON(!list_empty(&ordered_extent->list));
                 ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent);
@@ -1754,9 +1675,9 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
                 goto out;
         }
 
-        lock_extent(io_tree, ordered_extent->file_offset,
-                    ordered_extent->file_offset + ordered_extent->len - 1,
-                    GFP_NOFS);
+        lock_extent_bits(io_tree, ordered_extent->file_offset,
+                         ordered_extent->file_offset + ordered_extent->len - 1,
+                         0, &cached_state, GFP_NOFS);
 
         trans = btrfs_join_transaction(root, 1);
 
@@ -1783,9 +1704,10 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
                                    ordered_extent->len);
                 BUG_ON(ret);
         }
-        unlock_extent(io_tree, ordered_extent->file_offset,
-                    ordered_extent->file_offset + ordered_extent->len - 1,
-                    GFP_NOFS);
+        unlock_extent_cached(io_tree, ordered_extent->file_offset,
+                             ordered_extent->file_offset +
+                             ordered_extent->len - 1, &cached_state, GFP_NOFS);
+
         add_pending_csums(trans, inode, ordered_extent->file_offset,
                           &ordered_extent->list);
 
@@ -2194,7 +2116,7 @@ void btrfs_orphan_cleanup(struct btrfs_root *root)
                 found_key.objectid = found_key.offset;
                 found_key.type = BTRFS_INODE_ITEM_KEY;
                 found_key.offset = 0;
-                inode = btrfs_iget(root->fs_info->sb, &found_key, root);
+                inode = btrfs_iget(root->fs_info->sb, &found_key, root, NULL);
                 if (IS_ERR(inode))
                         break;
 
@@ -3122,6 +3044,7 @@ static int btrfs_truncate_page(struct address_space *mapping, loff_t from)
         struct btrfs_root *root = BTRFS_I(inode)->root;
         struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
         struct btrfs_ordered_extent *ordered;
+        struct extent_state *cached_state = NULL;
         char *kaddr;
         u32 blocksize = root->sectorsize;
         pgoff_t index = from >> PAGE_CACHE_SHIFT;
@@ -3168,12 +3091,14 @@ again:
         }
         wait_on_page_writeback(page);
 
-        lock_extent(io_tree, page_start, page_end, GFP_NOFS);
+        lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state,
+                         GFP_NOFS);
         set_page_extent_mapped(page);
 
         ordered = btrfs_lookup_ordered_extent(inode, page_start);
         if (ordered) {
-                unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+                unlock_extent_cached(io_tree, page_start, page_end,
+                                     &cached_state, GFP_NOFS);
                 unlock_page(page);
                 page_cache_release(page);
                 btrfs_start_ordered_extent(inode, ordered, 1);
@@ -3181,13 +3106,15 @@ again:
                 goto again;
         }
 
-        clear_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end,
+        clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start, page_end,
                           EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING,
-                          GFP_NOFS);
+                          0, 0, &cached_state, GFP_NOFS);
 
-        ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
+        ret = btrfs_set_extent_delalloc(inode, page_start, page_end,
+                                        &cached_state);
         if (ret) {
-                unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+                unlock_extent_cached(io_tree, page_start, page_end,
+                                     &cached_state, GFP_NOFS);
                 goto out_unlock;
         }
 
@@ -3200,7 +3127,8 @@ again:
         }
         ClearPageChecked(page);
         set_page_dirty(page);
-        unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+        unlock_extent_cached(io_tree, page_start, page_end, &cached_state,
+                             GFP_NOFS);
 
 out_unlock:
         if (ret)
@@ -3218,6 +3146,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
         struct btrfs_root *root = BTRFS_I(inode)->root;
         struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
         struct extent_map *em;
+        struct extent_state *cached_state = NULL;
         u64 mask = root->sectorsize - 1;
         u64 hole_start = (inode->i_size + mask) & ~mask;
         u64 block_end = (size + mask) & ~mask;
@@ -3233,11 +3162,13 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
                 struct btrfs_ordered_extent *ordered;
                 btrfs_wait_ordered_range(inode, hole_start,
                                          block_end - hole_start);
-                lock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
+                lock_extent_bits(io_tree, hole_start, block_end - 1, 0,
+                                 &cached_state, GFP_NOFS);
                 ordered = btrfs_lookup_ordered_extent(inode, hole_start);
                 if (!ordered)
                         break;
-                unlock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
+                unlock_extent_cached(io_tree, hole_start, block_end - 1,
+                                     &cached_state, GFP_NOFS);
                 btrfs_put_ordered_extent(ordered);
         }
 
@@ -3282,7 +3213,8 @@ int btrfs_cont_expand(struct inode *inode, loff_t size)
                         break;
         }
 
-        unlock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
+        unlock_extent_cached(io_tree, hole_start, block_end - 1, &cached_state,
+                             GFP_NOFS);
         return err;
 }
 
@@ -3680,6 +3612,7 @@ static noinline void init_btrfs_i(struct inode *inode)
         bi->index_cnt = (u64)-1;
         bi->last_unlink_trans = 0;
         bi->ordered_data_close = 0;
+        bi->force_compress = 0;
         extent_map_tree_init(&BTRFS_I(inode)->extent_tree, GFP_NOFS);
         extent_io_tree_init(&BTRFS_I(inode)->io_tree,
                              inode->i_mapping, GFP_NOFS);
@@ -3728,7 +3661,7 @@ static struct inode *btrfs_iget_locked(struct super_block *s,
  * Returns in *is_new if the inode was read from disk
  */
 struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
-                         struct btrfs_root *root)
+                         struct btrfs_root *root, int *new)
 {
         struct inode *inode;
 
@@ -3743,6 +3676,8 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
 
                 inode_tree_add(inode);
                 unlock_new_inode(inode);
+                if (new)
+                        *new = 1;
         }
 
         return inode;
@@ -3795,7 +3730,7 @@ struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry)
                 return NULL;
 
         if (location.type == BTRFS_INODE_ITEM_KEY) {
-                inode = btrfs_iget(dir->i_sb, &location, root);
+                inode = btrfs_iget(dir->i_sb, &location, root, NULL);
                 return inode;
         }
 
@@ -3810,7 +3745,7 @@ struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry)
                 else
                         inode = new_simple_dir(dir->i_sb, &location, sub_root);
         } else {
-                inode = btrfs_iget(dir->i_sb, &location, sub_root);
+                inode = btrfs_iget(dir->i_sb, &location, sub_root, NULL);
         }
         srcu_read_unlock(&root->fs_info->subvol_srcu, index);
 
@@ -3995,7 +3930,11 @@ skip:
 
         /* Reached end of directory/root. Bump pos past the last item. */
         if (key_type == BTRFS_DIR_INDEX_KEY)
-                filp->f_pos = INT_LIMIT(off_t);
+                /*
+                 * 32-bit glibc will use getdents64, but then strtol -
+                 * so the last number we can serve is this.
+                 */
+                filp->f_pos = 0x7fffffff;
         else
                 filp->f_pos++;
 nopos:
@@ -4005,7 +3944,7 @@ err:
         return ret;
 }
 
-int btrfs_write_inode(struct inode *inode, int wait)
+int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc)
 {
         struct btrfs_root *root = BTRFS_I(inode)->root;
         struct btrfs_trans_handle *trans;
@@ -4014,7 +3953,7 @@ int btrfs_write_inode(struct inode *inode, int wait)
         if (root->fs_info->btree_inode == inode)
                 return 0;
 
-        if (wait) {
+        if (wbc->sync_mode == WB_SYNC_ALL) {
                 trans = btrfs_join_transaction(root, 1);
                 btrfs_set_trans_block_group(trans, inode);
                 ret = btrfs_commit_transaction(trans, root);
@@ -4538,7 +4477,7 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
         err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
         if (err) {
                 err = -ENOSPC;
-                goto out_unlock;
+                goto out_fail;
         }
 
         inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
@@ -5016,6 +4955,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
 {
         struct extent_io_tree *tree;
         struct btrfs_ordered_extent *ordered;
+        struct extent_state *cached_state = NULL;
         u64 page_start = page_offset(page);
         u64 page_end = page_start + PAGE_CACHE_SIZE - 1;
 
@@ -5034,7 +4974,8 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
                 btrfs_releasepage(page, GFP_NOFS);
                 return;
         }
-        lock_extent(tree, page_start, page_end, GFP_NOFS);
+        lock_extent_bits(tree, page_start, page_end, 0, &cached_state,
+                         GFP_NOFS);
         ordered = btrfs_lookup_ordered_extent(page->mapping->host,
                                            page_offset(page));
         if (ordered) {
@@ -5045,7 +4986,7 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
                 clear_extent_bit(tree, page_start, page_end,
                                  EXTENT_DIRTY | EXTENT_DELALLOC |
                                  EXTENT_LOCKED | EXTENT_DO_ACCOUNTING, 1, 0,
-                                 NULL, GFP_NOFS);
+                                 &cached_state, GFP_NOFS);
                 /*
                  * whoever cleared the private bit is responsible
                  * for the finish_ordered_io
@@ -5055,11 +4996,13 @@ static void btrfs_invalidatepage(struct page *page, unsigned long offset)
                                                 page_start, page_end);
                 }
                 btrfs_put_ordered_extent(ordered);
-                lock_extent(tree, page_start, page_end, GFP_NOFS);
+                cached_state = NULL;
+                lock_extent_bits(tree, page_start, page_end, 0, &cached_state,
+                                 GFP_NOFS);
         }
         clear_extent_bit(tree, page_start, page_end,
                  EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
-                 EXTENT_DO_ACCOUNTING, 1, 1, NULL, GFP_NOFS);
+                 EXTENT_DO_ACCOUNTING, 1, 1, &cached_state, GFP_NOFS);
         __btrfs_releasepage(page, GFP_NOFS);
 
         ClearPageChecked(page);
@@ -5092,6 +5035,7 @@ int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
         struct btrfs_root *root = BTRFS_I(inode)->root;
         struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
         struct btrfs_ordered_extent *ordered;
+        struct extent_state *cached_state = NULL;
         char *kaddr;
         unsigned long zero_start;
         loff_t size;
@@ -5130,7 +5074,8 @@ again:
         }
         wait_on_page_writeback(page);
 
-        lock_extent(io_tree, page_start, page_end, GFP_NOFS);
+        lock_extent_bits(io_tree, page_start, page_end, 0, &cached_state,
+                         GFP_NOFS);
         set_page_extent_mapped(page);
 
         /*
@@ -5139,7 +5084,8 @@ again:
          */
         ordered = btrfs_lookup_ordered_extent(inode, page_start);
         if (ordered) {
-                unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+                unlock_extent_cached(io_tree, page_start, page_end,
+                                     &cached_state, GFP_NOFS);
                 unlock_page(page);
                 btrfs_start_ordered_extent(inode, ordered, 1);
                 btrfs_put_ordered_extent(ordered);
@@ -5153,13 +5099,15 @@ again:
          * is probably a better way to do this, but for now keep consistent with
          * prepare_pages in the normal write path.
          */
-        clear_extent_bits(&BTRFS_I(inode)->io_tree, page_start, page_end,
+        clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start, page_end,
                           EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING,
-                          GFP_NOFS);
+                          0, 0, &cached_state, GFP_NOFS);
 
-        ret = btrfs_set_extent_delalloc(inode, page_start, page_end);
+        ret = btrfs_set_extent_delalloc(inode, page_start, page_end,
+                                        &cached_state);
         if (ret) {
-                unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+                unlock_extent_cached(io_tree, page_start, page_end,
+                                     &cached_state, GFP_NOFS);
                 ret = VM_FAULT_SIGBUS;
                 btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE);
                 goto out_unlock;
@@ -5185,7 +5133,7 @@ again:
         BTRFS_I(inode)->last_trans = root->fs_info->generation;
         BTRFS_I(inode)->last_sub_trans = BTRFS_I(inode)->root->log_transid;
 
-        unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
+        unlock_extent_cached(io_tree, page_start, page_end, &cached_state, GFP_NOFS);
 
 out_unlock:
         btrfs_unreserve_metadata_for_delalloc(root, inode, 1);
@@ -5390,7 +5338,6 @@ free:
 void btrfs_drop_inode(struct inode *inode)
 {
         struct btrfs_root *root = BTRFS_I(inode)->root;
-
         if (inode->i_nlink > 0 && btrfs_root_refs(&root->root_item) == 0)
                 generic_delete_inode(inode);
         else
@@ -5789,22 +5736,20 @@ out_fail:
 }
 
 static int prealloc_file_range(struct inode *inode, u64 start, u64 end,
-                               u64 alloc_hint, int mode)
+                        u64 alloc_hint, int mode, loff_t actual_len)
 {
         struct btrfs_trans_handle *trans;
         struct btrfs_root *root = BTRFS_I(inode)->root;
         struct btrfs_key ins;
-        u64 alloc_size;
         u64 cur_offset = start;
         u64 num_bytes = end - start;
         int ret = 0;
+        u64 i_size;
 
         while (num_bytes > 0) {
-                alloc_size = min(num_bytes, root->fs_info->max_extent);
-
                 trans = btrfs_start_transaction(root, 1);
 
-                ret = btrfs_reserve_extent(trans, root, alloc_size,
+                ret = btrfs_reserve_extent(trans, root, num_bytes,
                                            root->sectorsize, 0, alloc_hint,
                                            (u64)-1, &ins, 1);
                 if (ret) {
@@ -5835,9 +5780,15 @@ static int prealloc_file_range(struct inode *inode, u64 start, u64 end,
                 inode->i_ctime = CURRENT_TIME;
                 BTRFS_I(inode)->flags |= BTRFS_INODE_PREALLOC;
                 if (!(mode & FALLOC_FL_KEEP_SIZE) &&
-                    cur_offset > inode->i_size) {
-                        i_size_write(inode, cur_offset);
-                        btrfs_ordered_update_i_size(inode, cur_offset, NULL);
+                        (actual_len > inode->i_size) &&
+                        (cur_offset > inode->i_size)) {
+
+                        if (cur_offset > actual_len)
+                                i_size  = actual_len;
+                        else
+                                i_size = cur_offset;
+                        i_size_write(inode, i_size);
+                        btrfs_ordered_update_i_size(inode, i_size, NULL);
                 }
 
                 ret = btrfs_update_inode(trans, root, inode);
@@ -5857,6 +5808,7 @@ stop_trans:
 static long btrfs_fallocate(struct inode *inode, int mode,
                             loff_t offset, loff_t len)
 {
+        struct extent_state *cached_state = NULL;
         u64 cur_offset;
         u64 last_byte;
         u64 alloc_start;
@@ -5895,16 +5847,17 @@ static long btrfs_fallocate(struct inode *inode, int mode,
                 /* the extent lock is ordered inside the running
                  * transaction
                  */
-                lock_extent(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
-                            GFP_NOFS);
+                lock_extent_bits(&BTRFS_I(inode)->io_tree, alloc_start,
+                                 locked_end, 0, &cached_state, GFP_NOFS);
                 ordered = btrfs_lookup_first_ordered_extent(inode,
                                                             alloc_end - 1);
                 if (ordered &&
                     ordered->file_offset + ordered->len > alloc_start &&
                     ordered->file_offset < alloc_end) {
                         btrfs_put_ordered_extent(ordered);
-                        unlock_extent(&BTRFS_I(inode)->io_tree,
-                                      alloc_start, locked_end, GFP_NOFS);
+                        unlock_extent_cached(&BTRFS_I(inode)->io_tree,
+                                             alloc_start, locked_end,
+                                             &cached_state, GFP_NOFS);
                         /*
                          * we can't wait on the range with the transaction
                          * running or with the extent lock held
@@ -5930,7 +5883,7 @@ static long btrfs_fallocate(struct inode *inode, int mode,
                      !test_bit(EXTENT_FLAG_PREALLOC, &em->flags))) {
                         ret = prealloc_file_range(inode,
                                                   cur_offset, last_byte,
-                                                  alloc_hint, mode);
+                                                alloc_hint, mode, offset+len);
                         if (ret < 0) {
                                 free_extent_map(em);
                                 break;
@@ -5946,8 +5899,8 @@ static long btrfs_fallocate(struct inode *inode, int mode,
                         break;
                 }
         }
-        unlock_extent(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
-                      GFP_NOFS);
+        unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end,
+                             &cached_state, GFP_NOFS);
 
         btrfs_free_reserved_data_space(BTRFS_I(inode)->root, inode,
                                        alloc_end - alloc_start);