1 files changed, 87 insertions, 90 deletions

diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c
index c0b26b6badba..190b6c4a6f2b 100644
--- a/fs/ceph/snap.c
+++ b/fs/ceph/snap.c
@@ -119,6 +119,7 @@ static struct ceph_snap_realm *ceph_create_snap_realm(
 	INIT_LIST_HEAD(&realm->children);
 	INIT_LIST_HEAD(&realm->child_item);
 	INIT_LIST_HEAD(&realm->empty_item);
+	INIT_LIST_HEAD(&realm->dirty_item);
 	INIT_LIST_HEAD(&realm->inodes_with_caps);
 	spin_lock_init(&realm->inodes_with_caps_lock);
 	__insert_snap_realm(&mdsc->snap_realms, realm);
@@ -435,7 +436,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 {
 	struct inode *inode = &ci->vfs_inode;
 	struct ceph_cap_snap *capsnap;
-	int used;
+	int used, dirty;
 
 	capsnap = kzalloc(sizeof(*capsnap), GFP_NOFS);
 	if (!capsnap) {
@@ -445,6 +446,7 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 
 	spin_lock(&inode->i_lock);
 	used = __ceph_caps_used(ci);
+	dirty = __ceph_caps_dirty(ci);
 	if (__ceph_have_pending_cap_snap(ci)) {
 		/* there is no point in queuing multiple "pending" cap_snaps,
 		   as no new writes are allowed to start when pending, so any
@@ -452,27 +454,37 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 		   cap_snap.  lucky us. */
 		dout("queue_cap_snap %p already pending\n", inode);
 		kfree(capsnap);
-	} else if (ci->i_wrbuffer_ref_head || (used & CEPH_CAP_FILE_WR)) {
+	} else if (ci->i_wrbuffer_ref_head || (used & CEPH_CAP_FILE_WR) ||
+		   (dirty & (CEPH_CAP_AUTH_EXCL|CEPH_CAP_XATTR_EXCL|
+			     CEPH_CAP_FILE_EXCL|CEPH_CAP_FILE_WR))) {
 		struct ceph_snap_context *snapc = ci->i_head_snapc;
 
+		dout("queue_cap_snap %p cap_snap %p queuing under %p\n", inode,
+		     capsnap, snapc);
 		igrab(inode);
-
+		
 		atomic_set(&capsnap->nref, 1);
 		capsnap->ci = ci;
 		INIT_LIST_HEAD(&capsnap->ci_item);
 		INIT_LIST_HEAD(&capsnap->flushing_item);
 
-		capsnap->follows = snapc->seq - 1;
+		capsnap->follows = snapc->seq;
 		capsnap->issued = __ceph_caps_issued(ci, NULL);
-		capsnap->dirty = __ceph_caps_dirty(ci);
+		capsnap->dirty = dirty;
 
 		capsnap->mode = inode->i_mode;
 		capsnap->uid = inode->i_uid;
 		capsnap->gid = inode->i_gid;
 
-		/* fixme? */
-		capsnap->xattr_blob = NULL;
-		capsnap->xattr_len = 0;
+		if (dirty & CEPH_CAP_XATTR_EXCL) {
+			__ceph_build_xattrs_blob(ci);
+			capsnap->xattr_blob =
+				ceph_buffer_get(ci->i_xattrs.blob);
+			capsnap->xattr_version = ci->i_xattrs.version;
+		} else {
+			capsnap->xattr_blob = NULL;
+			capsnap->xattr_version = 0;
+		}
 
 		/* dirty page count moved from _head to this cap_snap;
 		   all subsequent writes page dirties occur _after_ this
@@ -480,7 +492,9 @@ void ceph_queue_cap_snap(struct ceph_inode_info *ci)
 		capsnap->dirty_pages = ci->i_wrbuffer_ref_head;
 		ci->i_wrbuffer_ref_head = 0;
 		capsnap->context = snapc;
-		ci->i_head_snapc = NULL;
+		ci->i_head_snapc =
+			ceph_get_snap_context(ci->i_snap_realm->cached_context);
+		dout(" new snapc is %p\n", ci->i_head_snapc);
 		list_add_tail(&capsnap->ci_item, &ci->i_cap_snaps);
 
 		if (used & CEPH_CAP_FILE_WR) {
@@ -539,6 +553,41 @@ int __ceph_finish_cap_snap(struct ceph_inode_info *ci,
 	return 1;  /* caller may want to ceph_flush_snaps */
 }
 
+/*
+ * Queue cap_snaps for snap writeback for this realm and its children.
+ * Called under snap_rwsem, so realm topology won't change.
+ */
+static void queue_realm_cap_snaps(struct ceph_snap_realm *realm)
+{
+	struct ceph_inode_info *ci;
+	struct inode *lastinode = NULL;
+	struct ceph_snap_realm *child;
+
+	dout("queue_realm_cap_snaps %p %llx inodes\n", realm, realm->ino);
+
+	spin_lock(&realm->inodes_with_caps_lock);
+	list_for_each_entry(ci, &realm->inodes_with_caps,
+			    i_snap_realm_item) {
+		struct inode *inode = igrab(&ci->vfs_inode);
+		if (!inode)
+			continue;
+		spin_unlock(&realm->inodes_with_caps_lock);
+		if (lastinode)
+			iput(lastinode);
+		lastinode = inode;
+		ceph_queue_cap_snap(ci);
+		spin_lock(&realm->inodes_with_caps_lock);
+	}
+	spin_unlock(&realm->inodes_with_caps_lock);
+	if (lastinode)
+		iput(lastinode);
+
+	dout("queue_realm_cap_snaps %p %llx children\n", realm, realm->ino);
+	list_for_each_entry(child, &realm->children, child_item)
+		queue_realm_cap_snaps(child);
+
+	dout("queue_realm_cap_snaps %p %llx done\n", realm, realm->ino);
+}
 
 /*
  * Parse and apply a snapblob "snap trace" from the MDS.  This specifies
@@ -556,6 +605,7 @@ int ceph_update_snap_trace(struct ceph_mds_client *mdsc,
 	struct ceph_snap_realm *realm;
 	int invalidate = 0;
 	int err = -ENOMEM;
+	LIST_HEAD(dirty_realms);
 
 	dout("update_snap_trace deletion=%d\n", deletion);
 more:
@@ -578,45 +628,6 @@ more:
 		}
 	}
 
-	if (le64_to_cpu(ri->seq) > realm->seq) {
-		dout("update_snap_trace updating %llx %p %lld -> %lld\n",
-		     realm->ino, realm, realm->seq, le64_to_cpu(ri->seq));
-		/*
-		 * if the realm seq has changed, queue a cap_snap for every
-		 * inode with open caps.  we do this _before_ we update
-		 * the realm info so that we prepare for writeback under the
-		 * _previous_ snap context.
-		 *
-		 * ...unless it's a snap deletion!
-		 */
-		if (!deletion) {
-			struct ceph_inode_info *ci;
-			struct inode *lastinode = NULL;
-
-			spin_lock(&realm->inodes_with_caps_lock);
-			list_for_each_entry(ci, &realm->inodes_with_caps,
-					    i_snap_realm_item) {
-				struct inode *inode = igrab(&ci->vfs_inode);
-				if (!inode)
-					continue;
-				spin_unlock(&realm->inodes_with_caps_lock);
-				if (lastinode)
-					iput(lastinode);
-				lastinode = inode;
-				ceph_queue_cap_snap(ci);
-				spin_lock(&realm->inodes_with_caps_lock);
-			}
-			spin_unlock(&realm->inodes_with_caps_lock);
-			if (lastinode)
-				iput(lastinode);
-			dout("update_snap_trace cap_snaps queued\n");
-		}
-
-	} else {
-		dout("update_snap_trace %llx %p seq %lld unchanged\n",
-		     realm->ino, realm, realm->seq);
-	}
-
 	/* ensure the parent is correct */
 	err = adjust_snap_realm_parent(mdsc, realm, le64_to_cpu(ri->parent));
 	if (err < 0)
@@ -624,6 +635,8 @@ more:
 	invalidate += err;
 
 	if (le64_to_cpu(ri->seq) > realm->seq) {
+		dout("update_snap_trace updating %llx %p %lld -> %lld\n",
+		     realm->ino, realm, realm->seq, le64_to_cpu(ri->seq));
 		/* update realm parameters, snap lists */
 		realm->seq = le64_to_cpu(ri->seq);
 		realm->created = le64_to_cpu(ri->created);
@@ -641,9 +654,17 @@ more:
 		if (err < 0)
 			goto fail;
 
+		/* queue realm for cap_snap creation */
+		list_add(&realm->dirty_item, &dirty_realms);
+
 		invalidate = 1;
 	} else if (!realm->cached_context) {
+		dout("update_snap_trace %llx %p seq %lld new\n",
+		     realm->ino, realm, realm->seq);
 		invalidate = 1;
+	} else {
+		dout("update_snap_trace %llx %p seq %lld unchanged\n",
+		     realm->ino, realm, realm->seq);
 	}
 
 	dout("done with %llx %p, invalidated=%d, %p %p\n", realm->ino,
@@ -656,6 +677,14 @@ more:
 	if (invalidate)
 		rebuild_snap_realms(realm);
 
+	/*
+	 * queue cap snaps _after_ we've built the new snap contexts,
+	 * so that i_head_snapc can be set appropriately.
+	 */
+	list_for_each_entry(realm, &dirty_realms, dirty_item) {
+		queue_realm_cap_snaps(realm);
+	}
+
 	__cleanup_empty_realms(mdsc);
 	return 0;
 
@@ -688,7 +717,7 @@ static void flush_snaps(struct ceph_mds_client *mdsc)
 		igrab(inode);
 		spin_unlock(&mdsc->snap_flush_lock);
 		spin_lock(&inode->i_lock);
-		__ceph_flush_snaps(ci, &session);
+		__ceph_flush_snaps(ci, &session, 0);
 		spin_unlock(&inode->i_lock);
 		iput(inode);
 		spin_lock(&mdsc->snap_flush_lock);
@@ -789,6 +818,7 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 			};
 			struct inode *inode = ceph_find_inode(sb, vino);
 			struct ceph_inode_info *ci;
+			struct ceph_snap_realm *oldrealm;
 
 			if (!inode)
 				continue;
@@ -814,18 +844,19 @@ void ceph_handle_snap(struct ceph_mds_client *mdsc,
 			dout(" will move %p to split realm %llx %p\n",
 			     inode, realm->ino, realm);
 			/*
-			 * Remove the inode from the realm's inode
-			 * list, but don't add it to the new realm
-			 * yet.  We don't want the cap_snap to be
-			 * queued (again) by ceph_update_snap_trace()
-			 * below.  Queue it _now_, under the old context.
+			 * Move the inode to the new realm
 			 */
 			spin_lock(&realm->inodes_with_caps_lock);
 			list_del_init(&ci->i_snap_realm_item);
+			list_add(&ci->i_snap_realm_item,
+				 &realm->inodes_with_caps);
+			oldrealm = ci->i_snap_realm;
+			ci->i_snap_realm = realm;
 			spin_unlock(&realm->inodes_with_caps_lock);
 			spin_unlock(&inode->i_lock);
 
-			ceph_queue_cap_snap(ci);
+			ceph_get_snap_realm(mdsc, realm);
+			ceph_put_snap_realm(mdsc, oldrealm);
 
 			iput(inode);
 			continue;
@@ -853,43 +884,9 @@ skip_inode:
 	ceph_update_snap_trace(mdsc, p, e,
 			       op == CEPH_SNAP_OP_DESTROY);
 
-	if (op == CEPH_SNAP_OP_SPLIT) {
-		/*
-		 * ok, _now_ add the inodes into the new realm.
-		 */
-		for (i = 0; i < num_split_inos; i++) {
-			struct ceph_vino vino = {
-				.ino = le64_to_cpu(split_inos[i]),
-				.snap = CEPH_NOSNAP,
-			};
-			struct inode *inode = ceph_find_inode(sb, vino);
-			struct ceph_inode_info *ci;
-
-			if (!inode)
-				continue;
-			ci = ceph_inode(inode);
-			spin_lock(&inode->i_lock);
-			if (list_empty(&ci->i_snap_realm_item)) {
-				struct ceph_snap_realm *oldrealm =
-					ci->i_snap_realm;
-
-				dout(" moving %p to split realm %llx %p\n",
-				     inode, realm->ino, realm);
-				spin_lock(&realm->inodes_with_caps_lock);
-				list_add(&ci->i_snap_realm_item,
-					 &realm->inodes_with_caps);
-				ci->i_snap_realm = realm;
-				spin_unlock(&realm->inodes_with_caps_lock);
-				ceph_get_snap_realm(mdsc, realm);
-				ceph_put_snap_realm(mdsc, oldrealm);
-			}
-			spin_unlock(&inode->i_lock);
-			iput(inode);
-		}
-
+	if (op == CEPH_SNAP_OP_SPLIT)
 		/* we took a reference when we created the realm, above */
 		ceph_put_snap_realm(mdsc, realm);
-	}
 
 	__cleanup_empty_realms(mdsc);