ceph: rados pool namespace support

This patch adds codes that decode pool namespace information in
cap message and request reply. Pool namespace is saved in i_layout,
it will be passed to libceph when doing read/write.

Signed-off-by: Yan, Zheng <zyan@redhat.com>

1 files changed, 52 insertions, 15 deletions

diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 3f8efd866fec..d5b6f959a3c3 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1730,7 +1730,8 @@ enum {
 	POOL_WRITE	= 2,
 };
 
-static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
+static int __ceph_pool_perm_get(struct ceph_inode_info *ci,
+				s64 pool, struct ceph_string *pool_ns)
 {
 	struct ceph_fs_client *fsc = ceph_inode_to_client(&ci->vfs_inode);
 	struct ceph_mds_client *mdsc = fsc->mdsc;
@@ -1738,6 +1739,7 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 	struct rb_node **p, *parent;
 	struct ceph_pool_perm *perm;
 	struct page **pages;
+	size_t pool_ns_len;
 	int err = 0, err2 = 0, have = 0;
 
 	down_read(&mdsc->pool_perm_rwsem);
@@ -1749,17 +1751,31 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 		else if (pool > perm->pool)
 			p = &(*p)->rb_right;
 		else {
-			have = perm->perm;
-			break;
+			int ret = ceph_compare_string(pool_ns,
+						perm->pool_ns,
+						perm->pool_ns_len);
+			if (ret < 0)
+				p = &(*p)->rb_left;
+			else if (ret > 0)
+				p = &(*p)->rb_right;
+			else {
+				have = perm->perm;
+				break;
+			}
 		}
 	}
 	up_read(&mdsc->pool_perm_rwsem);
 	if (*p)
 		goto out;
 
-	dout("__ceph_pool_perm_get pool %lld no perm cached\n", pool);
+	if (pool_ns)
+		dout("__ceph_pool_perm_get pool %lld ns %.*s no perm cached\n",
+		     pool, (int)pool_ns->len, pool_ns->str);
+	else
+		dout("__ceph_pool_perm_get pool %lld no perm cached\n", pool);
 
 	down_write(&mdsc->pool_perm_rwsem);
+	p = &mdsc->pool_perm_tree.rb_node;
 	parent = NULL;
 	while (*p) {
 		parent = *p;
@@ -1769,8 +1785,17 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 		else if (pool > perm->pool)
 			p = &(*p)->rb_right;
 		else {
-			have = perm->perm;
-			break;
+			int ret = ceph_compare_string(pool_ns,
+						perm->pool_ns,
+						perm->pool_ns_len);
+			if (ret < 0)
+				p = &(*p)->rb_left;
+			else if (ret > 0)
+				p = &(*p)->rb_right;
+			else {
+				have = perm->perm;
+				break;
+			}
 		}
 	}
 	if (*p) {
@@ -1788,6 +1813,8 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 	rd_req->r_flags = CEPH_OSD_FLAG_READ;
 	osd_req_op_init(rd_req, 0, CEPH_OSD_OP_STAT, 0);
 	rd_req->r_base_oloc.pool = pool;
+	if (pool_ns)
+		rd_req->r_base_oloc.pool_ns = ceph_get_string(pool_ns);
 	ceph_oid_printf(&rd_req->r_base_oid, "%llx.00000000", ci->i_vino.ino);
 
 	err = ceph_osdc_alloc_messages(rd_req, GFP_NOFS);
@@ -1841,7 +1868,8 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 		goto out_unlock;
 	}
 
-	perm = kmalloc(sizeof(*perm), GFP_NOFS);
+	pool_ns_len = pool_ns ? pool_ns->len : 0;
+	perm = kmalloc(sizeof(*perm) + pool_ns_len + 1, GFP_NOFS);
 	if (!perm) {
 		err = -ENOMEM;
 		goto out_unlock;
@@ -1849,6 +1877,11 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, s64 pool)
 
 	perm->pool = pool;
 	perm->perm = have;
+	perm->pool_ns_len = pool_ns_len;
+	if (pool_ns_len > 0)
+		memcpy(perm->pool_ns, pool_ns->str, pool_ns_len);
+	perm->pool_ns[pool_ns_len] = 0;
+
 	rb_link_node(&perm->node, parent, p);
 	rb_insert_color(&perm->node, &mdsc->pool_perm_tree);
 	err = 0;
@@ -1860,19 +1893,20 @@ out_unlock:
 out:
 	if (!err)
 		err = have;
-	dout("__ceph_pool_perm_get pool %lld result = %d\n", pool, err);
+	if (pool_ns)
+		dout("__ceph_pool_perm_get pool %lld ns %.*s result = %d\n",
+		     pool, (int)pool_ns->len, pool_ns->str, err);
+	else
+		dout("__ceph_pool_perm_get pool %lld result = %d\n", pool, err);
 	return err;
 }
 
 int ceph_pool_perm_check(struct ceph_inode_info *ci, int need)
 {
 	s64 pool;
+	struct ceph_string *pool_ns;
 	int ret, flags;
 
-	/* does not support pool namespace yet */
-	if (ci->i_pool_ns_len)
-		return -EIO;
-
 	if (ceph_test_mount_opt(ceph_inode_to_client(&ci->vfs_inode),
 				NOPOOLPERM))
 		return 0;
@@ -1896,7 +1930,9 @@ check:
 		return 0;
 	}
 
-	ret = __ceph_pool_perm_get(ci, pool);
+	pool_ns = ceph_try_get_string(ci->i_layout.pool_ns);
+	ret = __ceph_pool_perm_get(ci, pool, pool_ns);
+	ceph_put_string(pool_ns);
 	if (ret < 0)
 		return ret;
 
@@ -1907,8 +1943,9 @@ check:
 		flags |= CEPH_I_POOL_WR;
 
 	spin_lock(&ci->i_ceph_lock);
-	if (pool == ci->i_layout.pool_id) {
-		ci->i_ceph_flags = flags;
+	if (pool == ci->i_layout.pool_id &&
+	    pool_ns == rcu_dereference_raw(ci->i_layout.pool_ns)) {
+		ci->i_ceph_flags |= flags;
         } else {
 		pool = ci->i_layout.pool_id;
 		flags = ci->i_ceph_flags;