1 files changed, 65 insertions, 18 deletions

diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index d91b0de7c502..741666a7676a 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -137,11 +137,20 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 
         mutex_lock(&root->log_mutex);
         if (root->log_root) {
+                if (!root->log_start_pid) {
+                        root->log_start_pid = current->pid;
+                        root->log_multiple_pids = false;
+                } else if (root->log_start_pid != current->pid) {
+                        root->log_multiple_pids = true;
+                }
+
                 root->log_batch++;
                 atomic_inc(&root->log_writers);
                 mutex_unlock(&root->log_mutex);
                 return 0;
         }
+        root->log_multiple_pids = false;
+        root->log_start_pid = current->pid;
         mutex_lock(&root->fs_info->tree_log_mutex);
         if (!root->fs_info->log_root_tree) {
                 ret = btrfs_init_log_root_tree(trans, root->fs_info);
@@ -263,8 +272,8 @@ static int process_one_buffer(struct btrfs_root *log,
                               struct walk_control *wc, u64 gen)
 {
         if (wc->pin)
-                btrfs_update_pinned_extents(log->fs_info->extent_root,
-                                            eb->start, eb->len, 1);
+                btrfs_pin_extent(log->fs_info->extent_root,
+                                 eb->start, eb->len, 0);
 
         if (btrfs_buffer_uptodate(eb, gen)) {
                 if (wc->write)
@@ -534,7 +543,7 @@ 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 ||
@@ -1971,6 +1980,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
         int ret;
         struct btrfs_root *log = root->log_root;
         struct btrfs_root *log_root_tree = root->fs_info->log_root_tree;
+        u64 log_transid = 0;
 
         mutex_lock(&root->log_mutex);
         index1 = root->log_transid % 2;
@@ -1987,10 +1997,11 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
         while (1) {
                 unsigned long batch = root->log_batch;
-                mutex_unlock(&root->log_mutex);
-                schedule_timeout_uninterruptible(1);
-                mutex_lock(&root->log_mutex);
-
+                if (root->log_multiple_pids) {
+                        mutex_unlock(&root->log_mutex);
+                        schedule_timeout_uninterruptible(1);
+                        mutex_lock(&root->log_mutex);
+                }
                 wait_for_writer(trans, root);
                 if (batch == root->log_batch)
                         break;
@@ -2003,14 +2014,19 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
                 goto out;
         }
 
-        ret = btrfs_write_and_wait_marked_extents(log, &log->dirty_log_pages);
+        /* we start IO on  all the marked extents here, but we don't actually
+         * wait for them until later.
+         */
+        ret = btrfs_write_marked_extents(log, &log->dirty_log_pages);
         BUG_ON(ret);
 
         btrfs_set_root_node(&log->root_item, log->node);
 
         root->log_batch = 0;
+        log_transid = root->log_transid;
         root->log_transid++;
         log->log_transid = root->log_transid;
+        root->log_start_pid = 0;
         smp_mb();
         /*
          * log tree has been flushed to disk, new modifications of
@@ -2036,6 +2052,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 
         index2 = log_root_tree->log_transid % 2;
         if (atomic_read(&log_root_tree->log_commit[index2])) {
+                btrfs_wait_marked_extents(log, &log->dirty_log_pages);
                 wait_log_commit(trans, log_root_tree,
                                 log_root_tree->log_transid);
                 mutex_unlock(&log_root_tree->log_mutex);
@@ -2055,6 +2072,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
          * check the full commit flag again
          */
         if (root->fs_info->last_trans_log_full_commit == trans->transid) {
+                btrfs_wait_marked_extents(log, &log->dirty_log_pages);
                 mutex_unlock(&log_root_tree->log_mutex);
                 ret = -EAGAIN;
                 goto out_wake_log_root;
@@ -2063,6 +2081,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
         ret = btrfs_write_and_wait_marked_extents(log_root_tree,
                                 &log_root_tree->dirty_log_pages);
         BUG_ON(ret);
+        btrfs_wait_marked_extents(log, &log->dirty_log_pages);
 
         btrfs_set_super_log_root(&root->fs_info->super_for_commit,
                                 log_root_tree->node->start);
@@ -2082,9 +2101,14 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
          * the running transaction open, so a full commit can't hop
          * in and cause problems either.
          */
-        write_ctree_super(trans, root->fs_info->tree_root, 2);
+        write_ctree_super(trans, root->fs_info->tree_root, 1);
         ret = 0;
 
+        mutex_lock(&root->log_mutex);
+        if (root->last_log_commit < log_transid)
+                root->last_log_commit = log_transid;
+        mutex_unlock(&root->log_mutex);
+
 out_wake_log_root:
         atomic_set(&log_root_tree->log_commit[index2], 0);
         smp_mb();
@@ -2605,7 +2629,7 @@ static noinline int copy_items(struct btrfs_trans_handle *trans,
                                                                 extent);
                                 cs = btrfs_file_extent_offset(src, extent);
                                 cl = btrfs_file_extent_num_bytes(src,
-                                                                extent);;
+                                                                extent);
                                 if (btrfs_file_extent_compression(src,
                                                                   extent)) {
                                         cs = 0;
@@ -2841,7 +2865,7 @@ static noinline int check_parent_dirs_for_sync(struct btrfs_trans_handle *trans,
                 if (!parent || !parent->d_inode || sb != parent->d_inode->i_sb)
                         break;
 
-                if (parent == sb->s_root)
+                if (IS_ROOT(parent))
                         break;
 
                 parent = parent->d_parent;
@@ -2852,6 +2876,21 @@ out:
         return ret;
 }
 
+static int inode_in_log(struct btrfs_trans_handle *trans,
+                 struct inode *inode)
+{
+        struct btrfs_root *root = BTRFS_I(inode)->root;
+        int ret = 0;
+
+        mutex_lock(&root->log_mutex);
+        if (BTRFS_I(inode)->logged_trans == trans->transid &&
+            BTRFS_I(inode)->last_sub_trans <= root->last_log_commit)
+                ret = 1;
+        mutex_unlock(&root->log_mutex);
+        return ret;
+}
+
+
 /*
  * helper function around btrfs_log_inode to make sure newly created
  * parent directories also end up in the log.  A minimal inode and backref
@@ -2880,11 +2919,22 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
                 goto end_no_trans;
         }
 
+        if (root != BTRFS_I(inode)->root ||
+            btrfs_root_refs(&root->root_item) == 0) {
+                ret = 1;
+                goto end_no_trans;
+        }
+
         ret = check_parent_dirs_for_sync(trans, inode, parent,
                                          sb, last_committed);
         if (ret)
                 goto end_no_trans;
 
+        if (inode_in_log(trans, inode)) {
+                ret = BTRFS_NO_LOG_SYNC;
+                goto end_no_trans;
+        }
+
         start_log_trans(trans, root);
 
         ret = btrfs_log_inode(trans, root, inode, inode_only);
@@ -2907,12 +2957,15 @@ int btrfs_log_inode_parent(struct btrfs_trans_handle *trans,
                         break;
 
                 inode = parent->d_inode;
+                if (root != BTRFS_I(inode)->root)
+                        break;
+
                 if (BTRFS_I(inode)->generation >
                     root->fs_info->last_trans_committed) {
                         ret = btrfs_log_inode(trans, root, inode, inode_only);
                         BUG_ON(ret);
                 }
-                if (parent == sb->s_root)
+                if (IS_ROOT(parent))
                         break;
 
                 parent = parent->d_parent;
@@ -2951,7 +3004,6 @@ int btrfs_recover_log_trees(struct btrfs_root *log_root_tree)
         struct btrfs_key tmp_key;
         struct btrfs_root *log;
         struct btrfs_fs_info *fs_info = log_root_tree->fs_info;
-        u64 highest_inode;
         struct walk_control wc = {
                 .process_func = process_one_buffer,
                 .stage = 0,
@@ -3010,11 +3062,6 @@ again:
                                                       path);
                         BUG_ON(ret);
                 }
-                ret = btrfs_find_highest_inode(wc.replay_dest, &highest_inode);
-                if (ret == 0) {
-                        wc.replay_dest->highest_inode = highest_inode;
-                        wc.replay_dest->last_inode_alloc = highest_inode;
-                }
 
                 key.offset = found_key.offset - 1;
                 wc.replay_dest->log_root = NULL;