Merge tag 'dlm-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/…

…teigland/linux-dlm

Pull dlm updates from David Teigland:
 "There are two main patches in this set, both related to the userland
  dlm_controld daemon.

  The first fixes a deadlock between dlm_controld and the dlm_send
  workqueue when both access configfs data simultaneously.

  The second reworks some code to get around a long standing, but
  intentional, unlock balance warning.  The userland daemon no longer
  takes a lock that is later released from the kernel.

  The other commits are minor fixes and changes."

* tag 'dlm-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm:
  dlm: check the maximum size of a request from user
  dlm: cleanup send_to_sock routine
  dlm: convert add_sock routine return value type to void
  dlm: remove redundant variable assignments
  dlm: fix unlock balance warnings
  dlm: fix uninitialized spinlock
  dlm: fix deadlock between dlm_send and dlm_controld

fs/dlm/ast.c

@@ -15,8 +15,8 @@
 #include "lock.h"
 #include "user.h"
 
-static uint64_t			dlm_cb_seq;
-static spinlock_t		dlm_cb_seq_spin;
+static uint64_t dlm_cb_seq;
+static DEFINE_SPINLOCK(dlm_cb_seq_spin);
 
 static void dlm_dump_lkb_callbacks(struct dlm_lkb *lkb)
 {

fs/dlm/config.c

@@ -750,6 +750,7 @@ static ssize_t comm_local_write(struct dlm_comm *cm, const char *buf,
 static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf, size_t len)
 {
 	struct sockaddr_storage *addr;
+	int rv;
 
 	if (len != sizeof(struct sockaddr_storage))
 		return -EINVAL;
@@ -762,6 +763,13 @@ static ssize_t comm_addr_write(struct dlm_comm *cm, const char *buf, size_t len)
 		return -ENOMEM;
 
 	memcpy(addr, buf, len);
+
+	rv = dlm_lowcomms_addr(cm->nodeid, addr, len);
+	if (rv) {
+		kfree(addr);
+		return rv;
+	}
+
 	cm->addr[cm->addr_count++] = addr;
 	return len;
 }
@@ -878,34 +886,7 @@ static void put_space(struct dlm_space *sp)
 	config_item_put(&sp->group.cg_item);
 }
 
-static int addr_compare(struct sockaddr_storage *x, struct sockaddr_storage *y)
-{
-	switch (x->ss_family) {
-	case AF_INET: {
-		struct sockaddr_in *sinx = (struct sockaddr_in *)x;
-		struct sockaddr_in *siny = (struct sockaddr_in *)y;
-		if (sinx->sin_addr.s_addr != siny->sin_addr.s_addr)
-			return 0;
-		if (sinx->sin_port != siny->sin_port)
-			return 0;
-		break;
-	}
-	case AF_INET6: {
-		struct sockaddr_in6 *sinx = (struct sockaddr_in6 *)x;
-		struct sockaddr_in6 *siny = (struct sockaddr_in6 *)y;
-		if (!ipv6_addr_equal(&sinx->sin6_addr, &siny->sin6_addr))
-			return 0;
-		if (sinx->sin6_port != siny->sin6_port)
-			return 0;
-		break;
-	}
-	default:
-		return 0;
-	}
-	return 1;
-}
-
-static struct dlm_comm *get_comm(int nodeid, struct sockaddr_storage *addr)
+static struct dlm_comm *get_comm(int nodeid)
 {
 	struct config_item *i;
 	struct dlm_comm *cm = NULL;
@@ -919,19 +900,11 @@ static struct dlm_comm *get_comm(int nodeid, struct sockaddr_storage *addr)
 	list_for_each_entry(i, &comm_list->cg_children, ci_entry) {
 		cm = config_item_to_comm(i);
 
-		if (nodeid) {
-			if (cm->nodeid != nodeid)
-				continue;
-			found = 1;
-			config_item_get(i);
-			break;
-		} else {
-			if (!cm->addr_count || !addr_compare(cm->addr[0], addr))
-				continue;
-			found = 1;
-			config_item_get(i);
-			break;
-		}
+		if (cm->nodeid != nodeid)
+			continue;
+		found = 1;
+		config_item_get(i);
+		break;
 	}
 	mutex_unlock(&clusters_root.subsys.su_mutex);
 
@@ -995,36 +968,14 @@ int dlm_config_nodes(char *lsname, struct dlm_config_node **nodes_out,
 
 int dlm_comm_seq(int nodeid, uint32_t *seq)
 {
-	struct dlm_comm *cm = get_comm(nodeid, NULL);
+	struct dlm_comm *cm = get_comm(nodeid);
 	if (!cm)
 		return -EEXIST;
 	*seq = cm->seq;
 	put_comm(cm);
 	return 0;
 }
 
-int dlm_nodeid_to_addr(int nodeid, struct sockaddr_storage *addr)
-{
-	struct dlm_comm *cm = get_comm(nodeid, NULL);
-	if (!cm)
-		return -EEXIST;
-	if (!cm->addr_count)
-		return -ENOENT;
-	memcpy(addr, cm->addr[0], sizeof(*addr));
-	put_comm(cm);
-	return 0;
-}
-
-int dlm_addr_to_nodeid(struct sockaddr_storage *addr, int *nodeid)
-{
-	struct dlm_comm *cm = get_comm(0, addr);
-	if (!cm)
-		return -EEXIST;
-	*nodeid = cm->nodeid;
-	put_comm(cm);
-	return 0;
-}
-
 int dlm_our_nodeid(void)
 {
 	return local_comm ? local_comm->nodeid : 0;

fs/dlm/config.h

@@ -46,8 +46,6 @@ void dlm_config_exit(void);
 int dlm_config_nodes(char *lsname, struct dlm_config_node **nodes_out,
 		     int *count_out);
 int dlm_comm_seq(int nodeid, uint32_t *seq);
-int dlm_nodeid_to_addr(int nodeid, struct sockaddr_storage *addr);
-int dlm_addr_to_nodeid(struct sockaddr_storage *addr, int *nodeid);
 int dlm_our_nodeid(void);
 int dlm_our_addr(struct sockaddr_storage *addr, int num);
 

fs/dlm/dlm_internal.h

@@ -604,6 +604,7 @@ struct dlm_ls {
 	struct idr		ls_recover_idr;
 	spinlock_t		ls_recover_idr_lock;
 	wait_queue_head_t	ls_wait_general;
+	wait_queue_head_t	ls_recover_lock_wait;
 	struct mutex		ls_clear_proc_locks;
 
 	struct list_head	ls_root_list;	/* root resources */
@@ -616,15 +617,40 @@ struct dlm_ls {
 	char			ls_name[1];
 };
 
-#define LSFL_WORK		0
-#define LSFL_RUNNING		1
-#define LSFL_RECOVERY_STOP	2
-#define LSFL_RCOM_READY		3
-#define LSFL_RCOM_WAIT		4
-#define LSFL_UEVENT_WAIT	5
-#define LSFL_TIMEWARN		6
-#define LSFL_CB_DELAY		7
-#define LSFL_NODIR		8
+/*
+ * LSFL_RECOVER_STOP - dlm_ls_stop() sets this to tell dlm recovery routines
+ * that they should abort what they're doing so new recovery can be started.
+ *
+ * LSFL_RECOVER_DOWN - dlm_ls_stop() sets this to tell dlm_recoverd that it
+ * should do down_write() on the in_recovery rw_semaphore. (doing down_write
+ * within dlm_ls_stop causes complaints about the lock acquired/released
+ * in different contexts.)
+ *
+ * LSFL_RECOVER_LOCK - dlm_recoverd holds the in_recovery rw_semaphore.
+ * It sets this after it is done with down_write() on the in_recovery
+ * rw_semaphore and clears it after it has released the rw_semaphore.
+ *
+ * LSFL_RECOVER_WORK - dlm_ls_start() sets this to tell dlm_recoverd that it
+ * should begin recovery of the lockspace.
+ *
+ * LSFL_RUNNING - set when normal locking activity is enabled.
+ * dlm_ls_stop() clears this to tell dlm locking routines that they should
+ * quit what they are doing so recovery can run.  dlm_recoverd sets
+ * this after recovery is finished.
+ */
+
+#define LSFL_RECOVER_STOP	0
+#define LSFL_RECOVER_DOWN	1
+#define LSFL_RECOVER_LOCK	2
+#define LSFL_RECOVER_WORK	3
+#define LSFL_RUNNING		4
+
+#define LSFL_RCOM_READY		5
+#define LSFL_RCOM_WAIT		6
+#define LSFL_UEVENT_WAIT	7
+#define LSFL_TIMEWARN		8
+#define LSFL_CB_DELAY		9
+#define LSFL_NODIR		10
 
 /* much of this is just saving user space pointers associated with the
    lock that we pass back to the user lib with an ast */
@@ -667,7 +693,7 @@ static inline int dlm_locking_stopped(struct dlm_ls *ls)
 
 static inline int dlm_recovery_stopped(struct dlm_ls *ls)
 {
-	return test_bit(LSFL_RECOVERY_STOP, &ls->ls_flags);
+	return test_bit(LSFL_RECOVER_STOP, &ls->ls_flags);
 }
 
 static inline int dlm_no_directory(struct dlm_ls *ls)

fs/dlm/lockspace.c

@@ -582,8 +582,6 @@ static int new_lockspace(const char *name, const char *cluster,
 	INIT_LIST_HEAD(&ls->ls_root_list);
 	init_rwsem(&ls->ls_root_sem);
 
-	down_write(&ls->ls_in_recovery);
-
 	spin_lock(&lslist_lock);
 	ls->ls_create_count = 1;
 	list_add(&ls->ls_list, &lslist);
@@ -597,13 +595,24 @@ static int new_lockspace(const char *name, const char *cluster,
 		}
 	}
 
-	/* needs to find ls in lslist */
+	init_waitqueue_head(&ls->ls_recover_lock_wait);
+
+	/*
+	 * Once started, dlm_recoverd first looks for ls in lslist, then
+	 * initializes ls_in_recovery as locked in "down" mode.  We need
+	 * to wait for the wakeup from dlm_recoverd because in_recovery
+	 * has to start out in down mode.
+	 */
+
 	error = dlm_recoverd_start(ls);
 	if (error) {
 		log_error(ls, "can't start dlm_recoverd %d", error);
 		goto out_callback;
 	}
 
+	wait_event(ls->ls_recover_lock_wait,
+		   test_bit(LSFL_RECOVER_LOCK, &ls->ls_flags));
+
 	ls->ls_kobj.kset = dlm_kset;
 	error = kobject_init_and_add(&ls->ls_kobj, &dlm_ktype, NULL,
 				     "%s", ls->ls_name);