ceph: quota: report root dir quota usage in statfs

This commit changes statfs default behaviour when reporting usage
statistics.  Instead of using the overall filesystem usage, statfs now
reports the quota for the filesystem root, if ceph.quota.max_bytes has
been set for this inode.  If quota hasn't been set, it falls back to the
old statfs behaviour.

A new mount option is also added ('noquotadf') to disable this behaviour.

Signed-off-by: Luis Henriques <[email protected]>
Reviewed-by: "Yan, Zheng" <[email protected]>
Signed-off-by: Ilya Dryomov <[email protected]>

Author: Luis Henriques

Documentation/filesystems/ceph.txt

@@ -149,6 +149,10 @@ Mount Options
   noasyncreaddir
 	Do not use the dcache as above for readdir.
 
+  noquotadf
+        Report overall filesystem usage in statfs instead of using the root
+        directory quota.
+
 More Information
 ================
 

fs/ceph/quota.c

@@ -18,6 +18,8 @@
  * along with this program; if not, see <http://www.gnu.org/licenses/>.
  */
 
+#include <linux/statfs.h>
+
 #include "super.h"
 #include "mds_client.h"
 
@@ -303,3 +305,57 @@ bool ceph_quota_is_max_bytes_approaching(struct inode *inode, loff_t newsize)
 	return check_quota_exceeded(inode, QUOTA_CHECK_MAX_BYTES_APPROACHING_OP,
 				    (newsize - size));
 }
+
+/*
+ * ceph_quota_update_statfs - if root has quota update statfs with quota status
+ * @fsc:	filesystem client instance
+ * @buf:	statfs to update
+ *
+ * If the mounted filesystem root has max_bytes quota set, update the filesystem
+ * statistics with the quota status.
+ *
+ * This function returns true if the stats have been updated, false otherwise.
+ */
+bool ceph_quota_update_statfs(struct ceph_fs_client *fsc, struct kstatfs *buf)
+{
+	struct ceph_mds_client *mdsc = fsc->mdsc;
+	struct ceph_inode_info *ci;
+	struct ceph_snap_realm *realm;
+	struct inode *in;
+	u64 total = 0, used, free;
+	bool is_updated = false;
+
+	down_read(&mdsc->snap_rwsem);
+	realm = get_quota_realm(mdsc, d_inode(fsc->sb->s_root));
+	up_read(&mdsc->snap_rwsem);
+	if (!realm)
+		return false;
+
+	spin_lock(&realm->inodes_with_caps_lock);
+	in = realm->inode ? igrab(realm->inode) : NULL;
+	spin_unlock(&realm->inodes_with_caps_lock);
+	if (in) {
+		ci = ceph_inode(in);
+		spin_lock(&ci->i_ceph_lock);
+		if (ci->i_max_bytes) {
+			total = ci->i_max_bytes >> CEPH_BLOCK_SHIFT;
+			used = ci->i_rbytes >> CEPH_BLOCK_SHIFT;
+			/* It is possible for a quota to be exceeded.
+			 * Report 'zero' in that case
+			 */
+			free = total > used ? total - used : 0;
+		}
+		spin_unlock(&ci->i_ceph_lock);
+		if (total) {
+			buf->f_blocks = total;
+			buf->f_bfree = free;
+			buf->f_bavail = free;
+			is_updated = true;
+		}
+		iput(in);
+	}
+	ceph_put_snap_realm(mdsc, realm);
+
+	return is_updated;
+}
+

fs/ceph/super.c

@@ -76,9 +76,18 @@ static int ceph_statfs(struct dentry *dentry, struct kstatfs *buf)
 	 */
 	buf->f_bsize = 1 << CEPH_BLOCK_SHIFT;
 	buf->f_frsize = 1 << CEPH_BLOCK_SHIFT;
-	buf->f_blocks = le64_to_cpu(st.kb) >> (CEPH_BLOCK_SHIFT-10);
-	buf->f_bfree = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10);
-	buf->f_bavail = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10);
+
+	/*
+	 * By default use root quota for stats; fallback to overall filesystem
+	 * usage if using 'noquotadf' mount option or if the root dir doesn't
+	 * have max_bytes quota set.
+	 */
+	if (ceph_test_mount_opt(fsc, NOQUOTADF) ||
+	    !ceph_quota_update_statfs(fsc, buf)) {
+		buf->f_blocks = le64_to_cpu(st.kb) >> (CEPH_BLOCK_SHIFT-10);
+		buf->f_bfree = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10);
+		buf->f_bavail = le64_to_cpu(st.kb_avail) >> (CEPH_BLOCK_SHIFT-10);
+	}
 
 	buf->f_files = le64_to_cpu(st.num_objects);
 	buf->f_ffree = -1;
@@ -151,6 +160,8 @@ enum {
 	Opt_acl,
 #endif
 	Opt_noacl,
+	Opt_quotadf,
+	Opt_noquotadf,
 };
 
 static match_table_t fsopt_tokens = {
@@ -187,6 +198,8 @@ static match_table_t fsopt_tokens = {
 	{Opt_acl, "acl"},
 #endif
 	{Opt_noacl, "noacl"},
+	{Opt_quotadf, "quotadf"},
+	{Opt_noquotadf, "noquotadf"},
 	{-1, NULL}
 };
 
@@ -334,6 +347,12 @@ static int parse_fsopt_token(char *c, void *private)
 	case Opt_norequire_active_mds:
 		fsopt->flags |= CEPH_MOUNT_OPT_MOUNTWAIT;
 		break;
+	case Opt_quotadf:
+		fsopt->flags &= ~CEPH_MOUNT_OPT_NOQUOTADF;
+		break;
+	case Opt_noquotadf:
+		fsopt->flags |= CEPH_MOUNT_OPT_NOQUOTADF;
+		break;
 #ifdef CONFIG_CEPH_FS_POSIX_ACL
 	case Opt_acl:
 		fsopt->sb_flags |= SB_POSIXACL;
@@ -520,6 +539,8 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root)
 	}
 	if (fsopt->flags & CEPH_MOUNT_OPT_NOPOOLPERM)
 		seq_puts(m, ",nopoolperm");
+	if (fsopt->flags & CEPH_MOUNT_OPT_NOQUOTADF)
+		seq_puts(m, ",noquotadf");
 
 #ifdef CONFIG_CEPH_FS_POSIX_ACL
 	if (fsopt->sb_flags & SB_POSIXACL)

fs/ceph/super.h

@@ -39,6 +39,7 @@
 #define CEPH_MOUNT_OPT_FSCACHE         (1<<10) /* use fscache */
 #define CEPH_MOUNT_OPT_NOPOOLPERM      (1<<11) /* no pool permission check */
 #define CEPH_MOUNT_OPT_MOUNTWAIT       (1<<12) /* mount waits if no mds is up */
+#define CEPH_MOUNT_OPT_NOQUOTADF       (1<<13) /* no root dir quota in statfs */
 
 #define CEPH_MOUNT_OPT_DEFAULT    CEPH_MOUNT_OPT_DCACHE
 
@@ -1104,5 +1105,7 @@ extern bool ceph_quota_is_max_bytes_exceeded(struct inode *inode,
 					     loff_t newlen);
 extern bool ceph_quota_is_max_bytes_approaching(struct inode *inode,
 						loff_t newlen);
+extern bool ceph_quota_update_statfs(struct ceph_fs_client *fsc,
+				     struct kstatfs *buf);
 
 #endif /* _FS_CEPH_SUPER_H */