Btrfs: fix missing qgroup reservation before fallocating (6113077c) · Commits · 戴 / test

Steps to reproduce: mkfs.btrfs <disk> mount <disk> <mnt> btrfs quota enable <mnt> btrfs sub create <mnt>/subv btrfs qgroup limit 10M <mnt>/subv fallocate --length 20M <mnt>/subv/data For the above example, fallocating will return successfully which is not expected, we try to fix it by doing qgroup reservation before fallocating. Signed-off-by:

Wang Shilong <wangsl-fnst@cn.fujitsu.com> Reviewed-by:

Miao Xie <miaox@cn.fujitsu.com> Signed-off-by:

Josef Bacik <jbacik@fusionio.com> Signed-off-by:

Chris Mason <chris.mason@fusionio.com>

fs/btrfs/file.c

+9 −0

Original line number	Diff line number	Diff line
		@@ -2142,6 +2142,7 @@ static long btrfs_fallocate(struct file *file, int mode,
		{
		struct inode *inode = file->f_path.dentry->d_inode;
		struct extent_state *cached_state = NULL;
		struct btrfs_root *root = BTRFS_I(inode)->root;
		u64 cur_offset;
		u64 last_byte;
		u64 alloc_start;
		@@ -2169,6 +2170,11 @@ static long btrfs_fallocate(struct file *file, int mode,
		ret = btrfs_check_data_free_space(inode, alloc_end - alloc_start);
		if (ret)
		return ret;
		if (root->fs_info->quota_enabled) {
		ret = btrfs_qgroup_reserve(root, alloc_end - alloc_start);
		if (ret)
		goto out_reserve_fail;
		}

		/*
		* wait for ordered IO before we have any locks. We'll loop again
		@@ -2272,6 +2278,9 @@ static long btrfs_fallocate(struct file *file, int mode,
		&cached_state, GFP_NOFS);
		out:
		mutex_unlock(&inode->i_mutex);
		if (root->fs_info->quota_enabled)
		btrfs_qgroup_free(root, alloc_end - alloc_start);
		out_reserve_fail:
		/* Let go of our reservation. */
		btrfs_free_reserved_data_space(inode, alloc_end - alloc_start);
		return ret;

Admin message