Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/josef/btrfs-next into for-linus-3.9

Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Conflicts:
	fs/btrfs/disk-io.c

38 files changed, 1630 insertions, 702 deletions

diff --git a/fs/btrfs/backref.h b/fs/btrfs/backref.h
index d61feca79455..310a7f6d09b1 100644
--- a/fs/btrfs/backref.h
+++ b/fs/btrfs/backref.h
@@ -19,7 +19,7 @@
 #ifndef __BTRFS_BACKREF__
 #define __BTRFS_BACKREF__
 
-#include "ioctl.h"
+#include <linux/btrfs.h>
 #include "ulist.h"
 #include "extent_io.h"
 
diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h
index 2a8c242bc4f5..d9b97d4960e6 100644
--- a/fs/btrfs/btrfs_inode.h
+++ b/fs/btrfs/btrfs_inode.h
@@ -40,6 +40,8 @@
 #define BTRFS_INODE_HAS_ASYNC_EXTENT            6
 #define BTRFS_INODE_NEEDS_FULL_SYNC             7
 #define BTRFS_INODE_COPY_EVERYTHING             8
+#define BTRFS_INODE_IN_DELALLOC_LIST            9
+#define BTRFS_INODE_READDIO_NEED_LOCK           10
 
 /* in memory btrfs inode */
 struct btrfs_inode {
@@ -216,4 +218,22 @@ static inline int btrfs_inode_in_log(struct inode *inode, u64 generation)
         return 0;
 }
 
+/*
+ * Disable DIO read nolock optimization, so new dio readers will be forced
+ * to grab i_mutex. It is used to avoid the endless truncate due to
+ * nonlocked dio read.
+ */
+static inline void btrfs_inode_block_unlocked_dio(struct inode *inode)
+{
+        set_bit(BTRFS_INODE_READDIO_NEED_LOCK, &BTRFS_I(inode)->runtime_flags);
+        smp_mb();
+}
+
+static inline void btrfs_inode_resume_unlocked_dio(struct inode *inode)
+{
+        smp_mb__before_clear_bit();
+        clear_bit(BTRFS_INODE_READDIO_NEED_LOCK,
+                  &BTRFS_I(inode)->runtime_flags);
+}
+
 #endif
diff --git a/fs/btrfs/check-integrity.c b/fs/btrfs/check-integrity.c
index 11d47bfb62b4..18af6f48781a 100644
--- a/fs/btrfs/check-integrity.c
+++ b/fs/btrfs/check-integrity.c
@@ -813,8 +813,7 @@ static int btrfsic_process_superblock_dev_mirror(
             (bh->b_data + (dev_bytenr & 4095));
 
         if (btrfs_super_bytenr(super_tmp) != dev_bytenr ||
-            strncmp((char *)(&(super_tmp->magic)), BTRFS_MAGIC,
-                    sizeof(super_tmp->magic)) ||
+            super_tmp->magic != cpu_to_le64(BTRFS_MAGIC) ||
             memcmp(device->uuid, super_tmp->dev_item.uuid, BTRFS_UUID_SIZE) ||
             btrfs_super_nodesize(super_tmp) != state->metablock_size ||
             btrfs_super_leafsize(super_tmp) != state->metablock_size ||
diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c
index eea5da7a2b9a..ecd25a1b4e51 100644
--- a/fs/btrfs/ctree.c
+++ b/fs/btrfs/ctree.c
@@ -1138,6 +1138,7 @@ __tree_mod_log_rewind(struct extent_buffer *eb, u64 time_seq,
                 switch (tm->op) {
                 case MOD_LOG_KEY_REMOVE_WHILE_FREEING:
                         BUG_ON(tm->slot < n);
+                        /* Fallthrough */
                 case MOD_LOG_KEY_REMOVE_WHILE_MOVING:
                 case MOD_LOG_KEY_REMOVE:
                         btrfs_set_node_key(eb, &tm->key, tm->slot);
@@ -1222,7 +1223,7 @@ tree_mod_log_rewind(struct btrfs_fs_info *fs_info, struct extent_buffer *eb,
 
         __tree_mod_log_rewind(eb_rewin, time_seq, tm);
         WARN_ON(btrfs_header_nritems(eb_rewin) >
-                BTRFS_NODEPTRS_PER_BLOCK(fs_info->fs_root));
+                BTRFS_NODEPTRS_PER_BLOCK(fs_info->tree_root));
 
         return eb_rewin;
 }
@@ -1441,7 +1442,7 @@ int btrfs_comp_cpu_keys(struct btrfs_key *k1, struct btrfs_key *k2)
  */
 int btrfs_realloc_node(struct btrfs_trans_handle *trans,
                        struct btrfs_root *root, struct extent_buffer *parent,
-                       int start_slot, int cache_only, u64 *last_ret,
+                       int start_slot, u64 *last_ret,
                        struct btrfs_key *progress)
 {
         struct extent_buffer *cur;
@@ -1461,8 +1462,6 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
         struct btrfs_disk_key disk_key;
 
         parent_level = btrfs_header_level(parent);
-        if (cache_only && parent_level != 1)
-                return 0;
 
         WARN_ON(trans->transaction != root->fs_info->running_transaction);
         WARN_ON(trans->transid != root->fs_info->generation);
@@ -1508,10 +1507,6 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans,
                 else
                         uptodate = 0;
                 if (!cur || !uptodate) {
-                        if (cache_only) {
-                                free_extent_buffer(cur);
-                                continue;
-                        }
                         if (!cur) {
                                 cur = read_tree_block(root, blocknr,
                                                          blocksize, gen);
@@ -4825,8 +4820,8 @@ int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path)
 
 /*
  * A helper function to walk down the tree starting at min_key, and looking
- * for nodes or leaves that are either in cache or have a minimum
- * transaction id.  This is used by the btree defrag code, and tree logging
+ * for nodes or leaves that are have a minimum transaction id.
+ * This is used by the btree defrag code, and tree logging
  *
  * This does not cow, but it does stuff the starting key it finds back