1 files changed, 58 insertions, 38 deletions

diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index bd3511c5ca81..ab408a23ba32 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -378,6 +378,18 @@ static int check_xflags(unsigned int flags)
 	return 0;
 }
 
+bool btrfs_exclop_start(struct btrfs_fs_info *fs_info,
+			enum btrfs_exclusive_operation type)
+{
+	return !cmpxchg(&fs_info->exclusive_operation, BTRFS_EXCLOP_NONE, type);
+}
+
+void btrfs_exclop_finish(struct btrfs_fs_info *fs_info)
+{
+	WRITE_ONCE(fs_info->exclusive_operation, BTRFS_EXCLOP_NONE);
+	sysfs_notify(&fs_info->fs_devices->fsid_kobj, NULL, "exclusive_operation");
+}
+
 /*
  * Set the xflags from the internal inode flags. The remaining items of fsxattr
  * are zeroed.
@@ -618,7 +630,7 @@ static noinline int create_subvol(struct inode *dir,
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
-		btrfs_subvolume_release_metadata(fs_info, &block_rsv);
+		btrfs_subvolume_release_metadata(root, &block_rsv);
 		goto fail_free;
 	}
 	trans->block_rsv = &block_rsv;
@@ -628,7 +640,8 @@ static noinline int create_subvol(struct inode *dir,
 	if (ret)
 		goto fail;
 
-	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0);
+	leaf = btrfs_alloc_tree_block(trans, root, 0, objectid, NULL, 0, 0, 0,
+				      BTRFS_NESTING_NORMAL);
 	if (IS_ERR(leaf)) {
 		ret = PTR_ERR(leaf);
 		goto fail;
@@ -742,7 +755,7 @@ fail:
 	kfree(root_item);
 	trans->block_rsv = NULL;
 	trans->bytes_reserved = 0;
-	btrfs_subvolume_release_metadata(fs_info, &block_rsv);
+	btrfs_subvolume_release_metadata(root, &block_rsv);
 
 	err = btrfs_commit_transaction(trans);
 	if (err && !ret)
@@ -856,7 +869,7 @@ fail:
 	if (ret && pending_snapshot->snap)
 		pending_snapshot->snap->anon_dev = 0;
 	btrfs_put_root(pending_snapshot->snap);
-	btrfs_subvolume_release_metadata(fs_info, &pending_snapshot->block_rsv);
+	btrfs_subvolume_release_metadata(root, &pending_snapshot->block_rsv);
 free_pending:
 	if (pending_snapshot->anon_dev)
 		free_anon_bdev(pending_snapshot->anon_dev);
@@ -1306,7 +1319,7 @@ again:
 				break;
 
 			unlock_page(page);
-			btrfs_start_ordered_extent(inode, ordered, 1);
+			btrfs_start_ordered_extent(ordered, 1);
 			btrfs_put_ordered_extent(ordered);
 			lock_page(page);
 			/*
@@ -1638,7 +1651,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 	if (ret)
 		return ret;
 
-	if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) {
+	if (!btrfs_exclop_start(fs_info, BTRFS_EXCLOP_RESIZE)) {
 		mnt_drop_write_file(file);
 		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 	}
@@ -1752,7 +1765,7 @@ static noinline int btrfs_ioctl_resize(struct file *file,
 out_free:
 	kfree(vol_args);
 out:
-	clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+	btrfs_exclop_finish(fs_info);
 	mnt_drop_write_file(file);
 	return ret;
 }
@@ -2086,9 +2099,14 @@ static noinline int copy_to_sk(struct btrfs_path *path,
 		sh.len = item_len;
 		sh.transid = found_transid;
 
-		/* copy search result header */
-		if (copy_to_user(ubuf + *sk_offset, &sh, sizeof(sh))) {
-			ret = -EFAULT;
+		/*
+		 * Copy search result header. If we fault then loop again so we
+		 * can fault in the pages and -EFAULT there if there's a
+		 * problem. Otherwise we'll fault and then copy the buffer in
+		 * properly this next time through
+		 */
+		if (copy_to_user_nofault(ubuf + *sk_offset, &sh, sizeof(sh))) {
+			ret = 0;
 			goto out;
 		}
 
@@ -2096,10 +2114,14 @@ static noinline int copy_to_sk(struct btrfs_path *path,
 
 		if (item_len) {
 			char __user *up = ubuf + *sk_offset;
-			/* copy the item */
-			if (read_extent_buffer_to_user(leaf, up,
-						       item_off, item_len)) {
-				ret = -EFAULT;
+			/*
+			 * Copy the item, same behavior as above, but reset the
+			 * * sk_offset so we copy the full thing again.
+			 */
+			if (read_extent_buffer_to_user_nofault(leaf, up,
+						item_off, item_len)) {
+				ret = 0;
+				*sk_offset -= sizeof(sh);
 				goto out;
 			}
 
@@ -2184,6 +2206,11 @@ static noinline int search_ioctl(struct inode *inode,
 	key.offset = sk->min_offset;
 
 	while (1) {
+		ret = fault_in_pages_writeable(ubuf + sk_offset,
+					       *buf_size - sk_offset);
+		if (ret)
+			break;
+
 		ret = btrfs_search_forward(root, &key, path, sk->min_transid);
 		if (ret != 0) {
 			if (ret > 0)
@@ -3112,7 +3139,7 @@ static long btrfs_ioctl_add_dev(struct btrfs_fs_info *fs_info, void __user *arg)
 	if (!capable(CAP_SYS_ADMIN))
 		return -EPERM;
 
-	if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags))
+	if (!btrfs_exclop_start(fs_info, BTRFS_EXCLOP_DEV_ADD))
 		return BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 
 	vol_args = memdup_user(arg, sizeof(*vol_args));
@@ -3129,7 +3156,7 @@ static long btrfs_ioctl_add_dev(struct btrfs_fs_info *fs_info, void __user *arg)
 
 	kfree(vol_args);
 out:
-	clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+	btrfs_exclop_finish(fs_info);
 	return ret;
 }
 
@@ -3158,7 +3185,7 @@ static long btrfs_ioctl_rm_dev_v2(struct file *file, void __user *arg)
 		goto out;
 	}
 
-	if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) {
+	if (!btrfs_exclop_start(fs_info, BTRFS_EXCLOP_DEV_REMOVE)) {
 		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 		goto out;
 	}
@@ -3169,7 +3196,7 @@ static long btrfs_ioctl_rm_dev_v2(struct file *file, void __user *arg)
 		vol_args->name[BTRFS_SUBVOL_NAME_MAX] = '\0';
 		ret = btrfs_rm_device(fs_info, vol_args->name, 0);
 	}
-	clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+	btrfs_exclop_finish(fs_info);
 
 	if (!ret) {
 		if (vol_args->flags & BTRFS_DEVICE_SPEC_BY_ID)
@@ -3200,7 +3227,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
 	if (ret)
 		return ret;
 
-	if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) {
+	if (!btrfs_exclop_start(fs_info, BTRFS_EXCLOP_DEV_REMOVE)) {
 		ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 		goto out_drop_write;
 	}
@@ -3218,7 +3245,7 @@ static long btrfs_ioctl_rm_dev(struct file *file, void __user *arg)
 		btrfs_info(fs_info, "disk deleted %s", vol_args->name);
 	kfree(vol_args);
 out:
-	clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+	btrfs_exclop_finish(fs_info);
 out_drop_write:
 	mnt_drop_write_file(file);
 
@@ -3448,15 +3475,12 @@ static long btrfs_ioctl_space_info(struct btrfs_fs_info *fs_info,
 		struct btrfs_space_info *tmp;
 
 		info = NULL;
-		rcu_read_lock();
-		list_for_each_entry_rcu(tmp, &fs_info->space_info,
-					list) {
+		list_for_each_entry(tmp, &fs_info->space_info, list) {
 			if (tmp->flags == types[i]) {
 				info = tmp;
 				break;
 			}
 		}
-		rcu_read_unlock();
 
 		if (!info)
 			continue;
@@ -3504,15 +3528,12 @@ static long btrfs_ioctl_space_info(struct btrfs_fs_info *fs_info,
 			break;
 
 		info = NULL;
-		rcu_read_lock();
-		list_for_each_entry_rcu(tmp, &fs_info->space_info,
-					list) {
+		list_for_each_entry(tmp, &fs_info->space_info, list) {
 			if (tmp->flags == types[i]) {
 				info = tmp;
 				break;
 			}
 		}
-		rcu_read_unlock();
 
 		if (!info)
 			continue;
@@ -3722,11 +3743,11 @@ static long btrfs_ioctl_dev_replace(struct btrfs_fs_info *fs_info,
 			ret = -EROFS;
 			goto out;
 		}
-		if (test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) {
+		if (!btrfs_exclop_start(fs_info, BTRFS_EXCLOP_DEV_REPLACE)) {
 			ret = BTRFS_ERROR_DEV_EXCL_RUN_IN_PROGRESS;
 		} else {
 			ret = btrfs_dev_replace_by_ioctl(fs_info, p);
-			clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+			btrfs_exclop_finish(fs_info);
 		}
 		break;
 	case BTRFS_IOCTL_DEV_REPLACE_CMD_STATUS:
@@ -3937,7 +3958,7 @@ static long btrfs_ioctl_balance(struct file *file, void __user *arg)
 		return ret;
 
 again:
-	if (!test_and_set_bit(BTRFS_FS_EXCL_OP, &fs_info->flags)) {
+	if (btrfs_exclop_start(fs_info, BTRFS_EXCLOP_BALANCE)) {
 		mutex_lock(&fs_info->balance_mutex);
 		need_unlock = true;
 		goto locked;
@@ -3983,7 +4004,6 @@ again:
 	}
 
 locked:
-	BUG_ON(!test_bit(BTRFS_FS_EXCL_OP, &fs_info->flags));
 
 	if (arg) {
 		bargs = memdup_user(arg, sizeof(*bargs));
@@ -4038,10 +4058,10 @@ locked:
 
 do_balance:
 	/*
-	 * Ownership of bctl and filesystem flag BTRFS_FS_EXCL_OP goes to
-	 * btrfs_balance.  bctl is freed in reset_balance_state, or, if
-	 * restriper was paused all the way until unmount, in free_fs_info.
-	 * The flag should be cleared after reset_balance_state.
+	 * Ownership of bctl and exclusive operation goes to btrfs_balance.
+	 * bctl is freed in reset_balance_state, or, if restriper was paused
+	 * all the way until unmount, in free_fs_info.  The flag should be
+	 * cleared after reset_balance_state.
 	 */
 	need_unlock = false;
 
@@ -4060,7 +4080,7 @@ out_bargs:
 out_unlock:
 	mutex_unlock(&fs_info->balance_mutex);
 	if (need_unlock)
-		clear_bit(BTRFS_FS_EXCL_OP, &fs_info->flags);
+		btrfs_exclop_finish(fs_info);
 out:
 	mnt_drop_write_file(file);
 	return ret;
@@ -4883,7 +4903,7 @@ long btrfs_ioctl(struct file *file, unsigned int
 	case BTRFS_IOC_SYNC: {
 		int ret;
 
-		ret = btrfs_start_delalloc_roots(fs_info, -1);
+		ret = btrfs_start_delalloc_roots(fs_info, U64_MAX);
 		if (ret)
 			return ret;
 		ret = btrfs_sync_fs(inode->i_sb, 1);