4 files changed, 50 insertions, 2 deletions

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 9d7a6c38f0b1..8ee5645ef9e1 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -3458,6 +3458,7 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes, u64 write_bytes)
 void btrfs_free_reserved_data_space(struct inode *inode, u64 bytes);
 void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans,
                                 struct btrfs_root *root);
+void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans);
 int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans,
                                   struct inode *inode);
 void btrfs_orphan_release_metadata(struct inode *inode);
diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 1cbc71d8cb96..4e08e47ace30 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -4116,11 +4116,19 @@ static void check_system_chunk(struct btrfs_trans_handle *trans,
         struct btrfs_space_info *info;
         u64 left;
         u64 thresh;
+        int ret = 0;
+
+        /*
+         * Needed because we can end up allocating a system chunk and for an
+         * atomic and race free space reservation in the chunk block reserve.
+         */
+        ASSERT(mutex_is_locked(&root->fs_info->chunk_mutex));
 
         info = __find_space_info(root->fs_info, BTRFS_BLOCK_GROUP_SYSTEM);
         spin_lock(&info->lock);
         left = info->total_bytes - info->bytes_used - info->bytes_pinned -
-                info->bytes_reserved - info->bytes_readonly;
+                info->bytes_reserved - info->bytes_readonly -
+                info->bytes_may_use;
         spin_unlock(&info->lock);
 
         thresh = get_system_chunk_thresh(root, type);
@@ -4134,7 +4142,21 @@ static void check_system_chunk(struct btrfs_trans_handle *trans,
                 u64 flags;
 
                 flags = btrfs_get_alloc_profile(root->fs_info->chunk_root, 0);
-                btrfs_alloc_chunk(trans, root, flags);
+                /*
+                 * Ignore failure to create system chunk. We might end up not
+                 * needing it, as we might not need to COW all nodes/leafs from
+                 * the paths we visit in the chunk tree (they were already COWed
+                 * or created in the current transaction for example).
+                 */
+                ret = btrfs_alloc_chunk(trans, root, flags);
+        }
+
+        if (!ret) {
+                ret = btrfs_block_rsv_add(root->fs_info->chunk_root,
+                                          &root->fs_info->chunk_block_rsv,
+                                          thresh, BTRFS_RESERVE_NO_FLUSH);
+                if (!ret)
+                        trans->chunk_bytes_reserved += thresh;
         }
 }
 
@@ -5192,6 +5214,24 @@ void btrfs_trans_release_metadata(struct btrfs_trans_handle *trans,
         trans->bytes_reserved = 0;
 }
 
+/*
+ * To be called after all the new block groups attached to the transaction
+ * handle have been created (btrfs_create_pending_block_groups()).
+ */
+void btrfs_trans_release_chunk_metadata(struct btrfs_trans_handle *trans)
+{
+        struct btrfs_fs_info *fs_info = trans->root->fs_info;
+
+        if (!trans->chunk_bytes_reserved)
+                return;
+
+        WARN_ON_ONCE(!list_empty(&trans->new_bgs));
+
+        block_rsv_release_bytes(fs_info, &fs_info->chunk_block_rsv, NULL,
+                                trans->chunk_bytes_reserved);
+        trans->chunk_bytes_reserved = 0;
+}
+
 /* Can only return 0 or -ENOSPC */
 int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans,
                                   struct inode *inode)
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 5628e25250c0..03a3ec7e31ea 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -509,6 +509,7 @@ again:
         h->transaction = cur_trans;
         h->blocks_used = 0;
         h->bytes_reserved = 0;
+        h->chunk_bytes_reserved = 0;
         h->root = root;
         h->delayed_ref_updates = 0;
         h->use_count = 1;
@@ -792,6 +793,8 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
         if (!list_empty(&trans->new_bgs))
                 btrfs_create_pending_block_groups(trans, root);
 
+        btrfs_trans_release_chunk_metadata(trans);
+
         if (lock && !atomic_read(&root->fs_info->open_ioctl_trans) &&
             should_end_transaction(trans, root) &&
             ACCESS_ONCE(cur_trans->state) == TRANS_STATE_RUNNING) {
@@ -2054,6 +2057,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
         clear_bit(BTRFS_INODE_BTREE_LOG1_ERR, &btree_ino->runtime_flags);
         clear_bit(BTRFS_INODE_BTREE_LOG2_ERR, &btree_ino->runtime_flags);
 
+        btrfs_trans_release_chunk_metadata(trans);
+
         spin_lock(&root->fs_info->trans_lock);
         cur_trans->state = TRANS_STATE_UNBLOCKED;
         root->fs_info->running_transaction = NULL;
@@ -2123,6 +2128,7 @@ scrub_continue:
         btrfs_scrub_continue(root);
 cleanup_transaction:
         btrfs_trans_release_metadata(trans, root);
+        btrfs_trans_release_chunk_metadata(trans);
         trans->block_rsv = NULL;
         if (trans->qgroup_reserved) {
                 btrfs_qgroup_free(root, trans->qgroup_reserved);
diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h
index 0b24755596ba..036fa83d6ccb 100644
--- a/fs/btrfs/transaction.h
+++ b/fs/btrfs/transaction.h
@@ -102,6 +102,7 @@ struct btrfs_transaction {
 struct btrfs_trans_handle {
         u64 transid;
         u64 bytes_reserved;
+        u64 chunk_bytes_reserved;
         u64 qgroup_reserved;
         unsigned long use_count;
         unsigned long blocks_reserved;