Btrfs: rescan for qgroups

If qgroup tracking is out of sync, a rescan operation can be started. It
iterates the complete extent tree and recalculates all qgroup tracking data.
This is an expensive operation and should not be used unless required.

A filesystem under rescan can still be umounted. The rescan continues on the
next mount.  Status information is provided with a separate ioctl while a
rescan operation is in progress.

Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>

fs/btrfs/ctree.h

@@ -1021,9 +1021,9 @@ struct btrfs_block_group_item {
  */
 #define BTRFS_QGROUP_STATUS_FLAG_ON		(1ULL << 0)
 /*
- * SCANNING is set during the initialization phase
+ * RESCAN is set during the initialization phase
  */
-#define BTRFS_QGROUP_STATUS_FLAG_SCANNING	(1ULL << 1)
+#define BTRFS_QGROUP_STATUS_FLAG_RESCAN		(1ULL << 1)
 /*
  * Some qgroup entries are known to be out of date,
  * either because the configuration has changed in a way that
@@ -1052,7 +1052,7 @@ struct btrfs_qgroup_status_item {
 	 * only used during scanning to record the progress
 	 * of the scan. It contains a logical address
 	 */
-	__le64 scan;
+	__le64 rescan;
 } __attribute__ ((__packed__));
 
 struct btrfs_qgroup_info_item {
@@ -1603,6 +1603,11 @@ struct btrfs_fs_info {
 	/* used by btrfs_qgroup_record_ref for an efficient tree traversal */
 	u64 qgroup_seq;
 
+	/* qgroup rescan items */
+	struct mutex qgroup_rescan_lock; /* protects the progress item */
+	struct btrfs_key qgroup_rescan_progress;
+	struct btrfs_workers qgroup_rescan_workers;
+
 	/* filesystem state */
 	unsigned long fs_state;
 
@@ -2886,8 +2891,8 @@ BTRFS_SETGET_FUNCS(qgroup_status_version, struct btrfs_qgroup_status_item,
 		   version, 64);
 BTRFS_SETGET_FUNCS(qgroup_status_flags, struct btrfs_qgroup_status_item,
 		   flags, 64);
-BTRFS_SETGET_FUNCS(qgroup_status_scan, struct btrfs_qgroup_status_item,
-		   scan, 64);
+BTRFS_SETGET_FUNCS(qgroup_status_rescan, struct btrfs_qgroup_status_item,
+		   rescan, 64);
 
 /* btrfs_qgroup_info_item */
 BTRFS_SETGET_FUNCS(qgroup_info_generation, struct btrfs_qgroup_info_item,
@@ -3828,7 +3833,7 @@ int btrfs_quota_enable(struct btrfs_trans_handle *trans,
 		       struct btrfs_fs_info *fs_info);
 int btrfs_quota_disable(struct btrfs_trans_handle *trans,
 			struct btrfs_fs_info *fs_info);
-int btrfs_quota_rescan(struct btrfs_fs_info *fs_info);
+int btrfs_qgroup_rescan(struct btrfs_fs_info *fs_info);
 int btrfs_add_qgroup_relation(struct btrfs_trans_handle *trans,
 			      struct btrfs_fs_info *fs_info, u64 src, u64 dst);
 int btrfs_del_qgroup_relation(struct btrfs_trans_handle *trans,

fs/btrfs/disk-io.c

@@ -1976,6 +1976,7 @@ static void btrfs_stop_all_workers(struct btrfs_fs_info *fs_info)
 	btrfs_stop_workers(&fs_info->caching_workers);
 	btrfs_stop_workers(&fs_info->readahead_workers);
 	btrfs_stop_workers(&fs_info->flush_workers);
+	btrfs_stop_workers(&fs_info->qgroup_rescan_workers);
 }
 
 /* helper to cleanup tree roots */
@@ -2267,6 +2268,7 @@ int open_ctree(struct super_block *sb,
 	fs_info->qgroup_seq = 1;
 	fs_info->quota_enabled = 0;
 	fs_info->pending_quota_state = 0;
+	mutex_init(&fs_info->qgroup_rescan_lock);
 
 	btrfs_init_free_cluster(&fs_info->meta_alloc_cluster);
 	btrfs_init_free_cluster(&fs_info->data_alloc_cluster);
@@ -2476,6 +2478,8 @@ int open_ctree(struct super_block *sb,
 	btrfs_init_workers(&fs_info->readahead_workers, "readahead",
 			   fs_info->thread_pool_size,
 			   &fs_info->generic_worker);
+	btrfs_init_workers(&fs_info->qgroup_rescan_workers, "qgroup-rescan", 1,
+			   &fs_info->generic_worker);
 
 	/*
 	 * endios are largely parallel and should have a very
@@ -2510,6 +2514,7 @@ int open_ctree(struct super_block *sb,
 	ret |= btrfs_start_workers(&fs_info->caching_workers);
 	ret |= btrfs_start_workers(&fs_info->readahead_workers);
 	ret |= btrfs_start_workers(&fs_info->flush_workers);
+	ret |= btrfs_start_workers(&fs_info->qgroup_rescan_workers);
 	if (ret) {
 		err = -ENOMEM;
 		goto fail_sb_buffer;

fs/btrfs/ioctl.c

@@ -3701,12 +3701,10 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 	}
 
 	down_write(&root->fs_info->subvol_sem);
-	if (sa->cmd != BTRFS_QUOTA_CTL_RESCAN) {
-		trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
-		if (IS_ERR(trans)) {
-			ret = PTR_ERR(trans);
-			goto out;
-		}
+	trans = btrfs_start_transaction(root->fs_info->tree_root, 2);
+	if (IS_ERR(trans)) {
+		ret = PTR_ERR(trans);
+		goto out;
 	}
 
 	switch (sa->cmd) {
@@ -3716,9 +3714,6 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 	case BTRFS_QUOTA_CTL_DISABLE:
 		ret = btrfs_quota_disable(trans, root->fs_info);
 		break;
-	case BTRFS_QUOTA_CTL_RESCAN:
-		ret = btrfs_quota_rescan(root->fs_info);
-		break;
 	default:
 		ret = -EINVAL;
 		break;
@@ -3727,11 +3722,9 @@ static long btrfs_ioctl_quota_ctl(struct file *file, void __user *arg)
 	if (copy_to_user(arg, sa, sizeof(*sa)))
 		ret = -EFAULT;
 
-	if (trans) {
-		err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
-		if (err && !ret)
-			ret = err;
-	}
+	err = btrfs_commit_transaction(trans, root->fs_info->tree_root);
+	if (err && !ret)
+		ret = err;
 out:
 	kfree(sa);
 	up_write(&root->fs_info->subvol_sem);
@@ -3886,6 +3879,64 @@ static long btrfs_ioctl_qgroup_limit(struct file *file, void __user *arg)
 	return ret;
 }
 
+static long btrfs_ioctl_quota_rescan(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+	struct btrfs_ioctl_quota_rescan_args *qsa;
+	int ret;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	ret = mnt_want_write_file(file);
+	if (ret)
+		return ret;
+
+	qsa = memdup_user(arg, sizeof(*qsa));
+	if (IS_ERR(qsa)) {
+		ret = PTR_ERR(qsa);
+		goto drop_write;
+	}
+
+	if (qsa->flags) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ret = btrfs_qgroup_rescan(root->fs_info);
+
+out:
+	kfree(qsa);
+drop_write:
+	mnt_drop_write_file(file);
+	return ret;
+}
+
+static long btrfs_ioctl_quota_rescan_status(struct file *file, void __user *arg)
+{
+	struct btrfs_root *root = BTRFS_I(fdentry(file)->d_inode)->root;
+	struct btrfs_ioctl_quota_rescan_args *qsa;
+	int ret = 0;
+
+	if (!capable(CAP_SYS_ADMIN))
+		return -EPERM;
+
+	qsa = kzalloc(sizeof(*qsa), GFP_NOFS);
+	if (!qsa)
+		return -ENOMEM;
+
+	if (root->fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_RESCAN) {
+		qsa->flags = 1;
+		qsa->progress = root->fs_info->qgroup_rescan_progress.objectid;
+	}
+
+	if (copy_to_user(arg, qsa, sizeof(*qsa)))
+		ret = -EFAULT;
+
+	kfree(qsa);
+	return ret;
+}
+
 static long btrfs_ioctl_set_received_subvol(struct file *file,
 					    void __user *arg)
 {
@@ -4124,6 +4175,10 @@ long btrfs_ioctl(struct file *file, unsigned int
 		return btrfs_ioctl_qgroup_create(file, argp);
 	case BTRFS_IOC_QGROUP_LIMIT:
 		return btrfs_ioctl_qgroup_limit(file, argp);
+	case BTRFS_IOC_QUOTA_RESCAN:
+		return btrfs_ioctl_quota_rescan(file, argp);
+	case BTRFS_IOC_QUOTA_RESCAN_STATUS:
+		return btrfs_ioctl_quota_rescan_status(file, argp);
 	case BTRFS_IOC_DEV_REPLACE:
 		return btrfs_ioctl_dev_replace(root, argp);
 	case BTRFS_IOC_GET_FSLABEL: