1 files changed, 129 insertions, 64 deletions

diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c
index 580899bdb991..808370ada888 100644
--- a/fs/btrfs/qgroup.c
+++ b/fs/btrfs/qgroup.c
@@ -11,6 +11,7 @@
 #include <linux/slab.h>
 #include <linux/workqueue.h>
 #include <linux/btrfs.h>
+#include <linux/sched/mm.h>
 
 #include "ctree.h"
 #include "transaction.h"
@@ -497,13 +498,13 @@ next2:
 			break;
 	}
 out:
+	btrfs_free_path(path);
 	fs_info->qgroup_flags |= flags;
 	if (!(fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON))
 		clear_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags);
 	else if (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN &&
 		 ret >= 0)
 		ret = qgroup_rescan_init(fs_info, rescan_progress, 0);
-	btrfs_free_path(path);
 
 	if (ret < 0) {
 		ulist_free(fs_info->qgroup_ulist);
@@ -893,8 +894,6 @@ static int btrfs_clean_quota_tree(struct btrfs_trans_handle *trans,
 	if (!path)
 		return -ENOMEM;
 
-	path->leave_spinning = 1;
-
 	key.objectid = 0;
 	key.offset = 0;
 	key.type = 0;
@@ -936,6 +935,7 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 	struct btrfs_key found_key;
 	struct btrfs_qgroup *qgroup = NULL;
 	struct btrfs_trans_handle *trans = NULL;
+	struct ulist *ulist = NULL;
 	int ret = 0;
 	int slot;
 
@@ -943,8 +943,8 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 	if (fs_info->quota_root)
 		goto out;
 
-	fs_info->qgroup_ulist = ulist_alloc(GFP_KERNEL);
-	if (!fs_info->qgroup_ulist) {
+	ulist = ulist_alloc(GFP_KERNEL);
+	if (!ulist) {
 		ret = -ENOMEM;
 		goto out;
 	}
@@ -952,6 +952,22 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 	ret = btrfs_sysfs_add_qgroups(fs_info);
 	if (ret < 0)
 		goto out;
+
+	/*
+	 * Unlock qgroup_ioctl_lock before starting the transaction. This is to
+	 * avoid lock acquisition inversion problems (reported by lockdep) between
+	 * qgroup_ioctl_lock and the vfs freeze semaphores, acquired when we
+	 * start a transaction.
+	 * After we started the transaction lock qgroup_ioctl_lock again and
+	 * check if someone else created the quota root in the meanwhile. If so,
+	 * just return success and release the transaction handle.
+	 *
+	 * Also we don't need to worry about someone else calling
+	 * btrfs_sysfs_add_qgroups() after we unlock and getting an error because
+	 * that function returns 0 (success) when the sysfs entries already exist.
+	 */
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
+
 	/*
 	 * 1 for quota root item
 	 * 1 for BTRFS_QGROUP_STATUS item
@@ -961,12 +977,20 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 	 * would be a lot of overkill.
 	 */
 	trans = btrfs_start_transaction(tree_root, 2);
+
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
 		trans = NULL;
 		goto out;
 	}
 
+	if (fs_info->quota_root)
+		goto out;
+
+	fs_info->qgroup_ulist = ulist;
+	ulist = NULL;
+
 	/*
 	 * initially create the quota tree
 	 */
@@ -1026,6 +1050,10 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 		btrfs_item_key_to_cpu(leaf, &found_key, slot);
 
 		if (found_key.type == BTRFS_ROOT_REF_KEY) {
+
+			/* Release locks on tree_root before we access quota_root */
+			btrfs_release_path(path);
+
 			ret = add_qgroup_item(trans, quota_root,
 					      found_key.offset);
 			if (ret) {
@@ -1044,6 +1072,20 @@ int btrfs_quota_enable(struct btrfs_fs_info *fs_info)
 				btrfs_abort_transaction(trans, ret);
 				goto out_free_path;
 			}
+			ret = btrfs_search_slot_for_read(tree_root, &found_key,
+							 path, 1, 0);
+			if (ret < 0) {
+				btrfs_abort_transaction(trans, ret);
+				goto out_free_path;
+			}
+			if (ret > 0) {
+				/*
+				 * Shouldn't happen, but in case it does we
+				 * don't need to do the btrfs_next_item, just
+				 * continue.
+				 */
+				continue;
+			}
 		}
 		ret = btrfs_next_item(tree_root, path);
 		if (ret < 0) {
@@ -1106,11 +1148,14 @@ out:
 	if (ret) {
 		ulist_free(fs_info->qgroup_ulist);
 		fs_info->qgroup_ulist = NULL;
-		if (trans)
-			btrfs_end_transaction(trans);
 		btrfs_sysfs_del_qgroups(fs_info);
 	}
 	mutex_unlock(&fs_info->qgroup_ioctl_lock);
+	if (ret && trans)
+		btrfs_end_transaction(trans);
+	else if (trans)
+		ret = btrfs_end_transaction(trans);
+	ulist_free(ulist);
 	return ret;
 }
 
@@ -1123,19 +1168,29 @@ int btrfs_quota_disable(struct btrfs_fs_info *fs_info)
 	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (!fs_info->quota_root)
 		goto out;
+	mutex_unlock(&fs_info->qgroup_ioctl_lock);
 
 	/*
 	 * 1 For the root item
 	 *
 	 * We should also reserve enough items for the quota tree deletion in
 	 * btrfs_clean_quota_tree but this is not done.
+	 *
+	 * Also, we must always start a transaction without holding the mutex
+	 * qgroup_ioctl_lock, see btrfs_quota_enable().
 	 */
 	trans = btrfs_start_transaction(fs_info->tree_root, 1);
+
+	mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
+		trans = NULL;
 		goto out;
 	}
 
+	if (!fs_info->quota_root)
+		goto out;
+
 	clear_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags);
 	btrfs_qgroup_wait_for_completion(fs_info, false);
 	spin_lock(&fs_info->qgroup_lock);
@@ -1149,13 +1204,13 @@ int btrfs_quota_disable(struct btrfs_fs_info *fs_info)
 	ret = btrfs_clean_quota_tree(trans, quota_root);
 	if (ret) {
 		btrfs_abort_transaction(trans, ret);
-		goto end_trans;
+		goto out;
 	}
 
 	ret = btrfs_del_root(trans, &quota_root->root_key);
 	if (ret) {
 		btrfs_abort_transaction(trans, ret);
-		goto end_trans;
+		goto out;
 	}
 
 	list_del(&quota_root->dirty_list);
@@ -1167,10 +1222,13 @@ int btrfs_quota_disable(struct btrfs_fs_info *fs_info)
 
 	btrfs_put_root(quota_root);
 
-end_trans:
-	ret = btrfs_end_transaction(trans);
 out:
 	mutex_unlock(&fs_info->qgroup_ioctl_lock);
+	if (ret && trans)
+		btrfs_end_transaction(trans);
+	else if (trans)
+		ret = btrfs_end_transaction(trans);
+
 	return ret;
 }
 
@@ -1306,13 +1364,17 @@ int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans, u64 src,
 	struct btrfs_qgroup *member;
 	struct btrfs_qgroup_list *list;
 	struct ulist *tmp;
+	unsigned int nofs_flag;
 	int ret = 0;
 
 	/* Check the level of src and dst first */
 	if (btrfs_qgroup_level(src) >= btrfs_qgroup_level(dst))
 		return -EINVAL;
 
+	/* We hold a transaction handle open, must do a NOFS allocation. */
+	nofs_flag = memalloc_nofs_save();
 	tmp = ulist_alloc(GFP_KERNEL);
+	memalloc_nofs_restore(nofs_flag);
 	if (!tmp)
 		return -ENOMEM;
 
@@ -1369,10 +1431,14 @@ static int __del_qgroup_relation(struct btrfs_trans_handle *trans, u64 src,
 	struct btrfs_qgroup_list *list;
 	struct ulist *tmp;
 	bool found = false;
+	unsigned int nofs_flag;
 	int ret = 0;
 	int ret2;
 
+	/* We hold a transaction handle open, must do a NOFS allocation. */
+	nofs_flag = memalloc_nofs_save();
 	tmp = ulist_alloc(GFP_KERNEL);
+	memalloc_nofs_restore(nofs_flag);
 	if (!tmp)
 		return -ENOMEM;
 
@@ -1876,34 +1942,22 @@ static int qgroup_trace_extent_swap(struct btrfs_trans_handle* trans,
 		struct btrfs_key dst_key;
 
 		if (src_path->nodes[cur_level] == NULL) {
-			struct btrfs_key first_key;
 			struct extent_buffer *eb;
 			int parent_slot;
-			u64 child_gen;
-			u64 child_bytenr;
 
 			eb = src_path->nodes[cur_level + 1];
 			parent_slot = src_path->slots[cur_level + 1];
-			child_bytenr = btrfs_node_blockptr(eb, parent_slot);
-			child_gen = btrfs_node_ptr_generation(eb, parent_slot);
-			btrfs_node_key_to_cpu(eb, &first_key, parent_slot);
 
-			eb = read_tree_block(fs_info, child_bytenr, child_gen,
-					     cur_level, &first_key);
+			eb = btrfs_read_node_slot(eb, parent_slot);
 			if (IS_ERR(eb)) {
 				ret = PTR_ERR(eb);
 				goto out;
-			} else if (!extent_buffer_uptodate(eb)) {
-				free_extent_buffer(eb);
-				ret = -EIO;
-				goto out;
 			}
 
 			src_path->nodes[cur_level] = eb;
 
 			btrfs_tree_read_lock(eb);
-			btrfs_set_lock_blocking_read(eb);
-			src_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING;
+			src_path->locks[cur_level] = BTRFS_READ_LOCK;
 		}
 
 		src_path->slots[cur_level] = dst_path->slots[cur_level];
@@ -1998,10 +2052,8 @@ static int qgroup_trace_new_subtree_blocks(struct btrfs_trans_handle* trans,
 
 	/* Read the tree block if needed */
 	if (dst_path->nodes[cur_level] == NULL) {
-		struct btrfs_key first_key;
 		int parent_slot;
 		u64 child_gen;
-		u64 child_bytenr;
 
 		/*
 		 * dst_path->nodes[root_level] must be initialized before
@@ -2020,31 +2072,23 @@ static int qgroup_trace_new_subtree_blocks(struct btrfs_trans_handle* trans,
 		  */
 		eb = dst_path->nodes[cur_level + 1];
 		parent_slot = dst_path->slots[cur_level + 1];
-		child_bytenr = btrfs_node_blockptr(eb, parent_slot);
 		child_gen = btrfs_node_ptr_generation(eb, parent_slot);
-		btrfs_node_key_to_cpu(eb, &first_key, parent_slot);
 
 		/* This node is old, no need to trace */
 		if (child_gen < last_snapshot)
 			goto out;
 
-		eb = read_tree_block(fs_info, child_bytenr, child_gen,
-				     cur_level, &first_key);
+		eb = btrfs_read_node_slot(eb, parent_slot);
 		if (IS_ERR(eb)) {
 			ret = PTR_ERR(eb);
 			goto out;
-		} else if (!extent_buffer_uptodate(eb)) {
-			free_extent_buffer(eb);
-			ret = -EIO;
-			goto out;
 		}
 
 		dst_path->nodes[cur_level] = eb;
 		dst_path->slots[cur_level] = 0;
 
 		btrfs_tree_read_lock(eb);
-		btrfs_set_lock_blocking_read(eb);
-		dst_path->locks[cur_level] = BTRFS_READ_LOCK_BLOCKING;
+		dst_path->locks[cur_level] = BTRFS_READ_LOCK;
 		need_cleanup = true;
 	}
 
@@ -2188,38 +2232,28 @@ walk_down:
 	level = root_level;
 	while (level >= 0) {
 		if (path->nodes[level] == NULL) {
-			struct btrfs_key first_key;
 			int parent_slot;
-			u64 child_gen;
 			u64 child_bytenr;
 
 			/*
-			 * We need to get child blockptr/gen from parent before
-			 * we can read it.
+			 * We need to get child blockptr from parent before we
+			 * can read it.
 			  */
 			eb = path->nodes[level + 1];
 			parent_slot = path->slots[level + 1];
 			child_bytenr = btrfs_node_blockptr(eb, parent_slot);
-			child_gen = btrfs_node_ptr_generation(eb, parent_slot);
-			btrfs_node_key_to_cpu(eb, &first_key, parent_slot);
 
-			eb = read_tree_block(fs_info, child_bytenr, child_gen,
-					     level, &first_key);
+			eb = btrfs_read_node_slot(eb, parent_slot);
 			if (IS_ERR(eb)) {
 				ret = PTR_ERR(eb);
 				goto out;
-			} else if (!extent_buffer_uptodate(eb)) {
-				free_extent_buffer(eb);
-				ret = -EIO;
-				goto out;
 			}
 
 			path->nodes[level] = eb;
 			path->slots[level] = 0;
 
 			btrfs_tree_read_lock(eb);
-			btrfs_set_lock_blocking_read(eb);
-			path->locks[level] = BTRFS_READ_LOCK_BLOCKING;
+			path->locks[level] = BTRFS_READ_LOCK;
 
 			ret = btrfs_qgroup_trace_extent(trans, child_bytenr,
 							fs_info->nodesize,
@@ -3156,6 +3190,12 @@ out:
 	return ret;
 }
 
+static bool rescan_should_stop(struct btrfs_fs_info *fs_info)
+{
+	return btrfs_fs_closing(fs_info) ||
+		test_bit(BTRFS_FS_STATE_REMOUNTING, &fs_info->fs_state);
+}
+
 static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 {
 	struct btrfs_fs_info *fs_info = container_of(work, struct btrfs_fs_info,
@@ -3164,6 +3204,7 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 	struct btrfs_trans_handle *trans = NULL;
 	int err = -ENOMEM;
 	int ret = 0;
+	bool stopped = false;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -3176,7 +3217,7 @@ static void btrfs_qgroup_rescan_worker(struct btrfs_work *work)
 	path->skip_locking = 1;
 
 	err = 0;
-	while (!err && !btrfs_fs_closing(fs_info)) {
+	while (!err && !(stopped = rescan_should_stop(fs_info))) {
 		trans = btrfs_start_transaction(fs_info->fs_root, 0);
 		if (IS_ERR(trans)) {
 			err = PTR_ERR(trans);
@@ -3219,7 +3260,7 @@ out:
 	}
 
 	mutex_lock(&fs_info->qgroup_rescan_lock);
-	if (!btrfs_fs_closing(fs_info))
+	if (!stopped)
 		fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN;
 	if (trans) {
 		ret = update_qgroup_status_item(trans);
@@ -3238,7 +3279,7 @@ out:
 
 	btrfs_end_transaction(trans);
 
-	if (btrfs_fs_closing(fs_info)) {
+	if (stopped) {
 		btrfs_info(fs_info, "qgroup scan paused");
 	} else if (err >= 0) {
 		btrfs_info(fs_info, "qgroup scan completed%s",
@@ -3417,24 +3458,20 @@ static int qgroup_unreserve_range(struct btrfs_inode *inode,
 {
 	struct rb_node *node;
 	struct rb_node *next;
-	struct ulist_node *entry = NULL;
+	struct ulist_node *entry;
 	int ret = 0;
 
 	node = reserved->range_changed.root.rb_node;
+	if (!node)
+		return 0;
 	while (node) {
 		entry = rb_entry(node, struct ulist_node, rb_node);
 		if (entry->val < start)
 			node = node->rb_right;
-		else if (entry)
-			node = node->rb_left;
 		else
-			break;
+			node = node->rb_left;
 	}
 
-	/* Empty changeset */
-	if (!entry)
-		return 0;
-
 	if (entry->val > start && rb_prev(&entry->rb_node))
 		entry = rb_entry(rb_prev(&entry->rb_node), struct ulist_node,
 				 rb_node);
@@ -3498,12 +3535,37 @@ static int try_flush_qgroup(struct btrfs_root *root)
 {
 	struct btrfs_trans_handle *trans;
 	int ret;
+	bool can_commit = true;
+
+	/*
+	 * If current process holds a transaction, we shouldn't flush, as we
+	 * assume all space reservation happens before a transaction handle is
+	 * held.
+	 *
+	 * But there are cases like btrfs_delayed_item_reserve_metadata() where
+	 * we try to reserve space with one transction handle already held.
+	 * In that case we can't commit transaction, but at least try to end it
+	 * and hope the started data writes can free some space.
+	 */
+	if (current->journal_info &&
+	    current->journal_info != BTRFS_SEND_TRANS_STUB)
+		can_commit = false;
 
 	/*
 	 * We don't want to run flush again and again, so if there is a running
 	 * one, we won't try to start a new flush, but exit directly.
 	 */
 	if (test_and_set_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state)) {
+		/*
+		 * We are already holding a transaction, thus we can block other
+		 * threads from flushing.  So exit right now. This increases
+		 * the chance of EDQUOT for heavy load and near limit cases.
+		 * But we can argue that if we're already near limit, EDQUOT is
+		 * unavoidable anyway.
+		 */
+		if (!can_commit)
+			return 0;
+
 		wait_event(root->qgroup_flush_wait,
 			!test_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state));
 		return 0;
@@ -3520,7 +3582,10 @@ static int try_flush_qgroup(struct btrfs_root *root)
 		goto out;
 	}
 
-	ret = btrfs_commit_transaction(trans);
+	if (can_commit)
+		ret = btrfs_commit_transaction(trans);
+	else
+		ret = btrfs_end_transaction(trans);
 out:
 	clear_bit(BTRFS_ROOT_QGROUP_FLUSHING, &root->state);
 	wake_up(&root->qgroup_flush_wait);
@@ -4160,7 +4225,7 @@ int btrfs_qgroup_trace_subtree_after_cow(struct btrfs_trans_handle *trans,
 	spin_unlock(&blocks->lock);
 
 	/* Read out reloc subtree root */
-	reloc_eb = read_tree_block(fs_info, block->reloc_bytenr,
+	reloc_eb = read_tree_block(fs_info, block->reloc_bytenr, 0,
 				   block->reloc_generation, block->level,
 				   &block->first_key);
 	if (IS_ERR(reloc_eb)) {