Merge tag 'dlm-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm

Pull dlm updates from David Teigland:
 "The dlm posix lock handling (for gfs2) has three notable changes:

   - Local pids returned from GETLK are no longer negated. A previous
     patch negating remote pids mistakenly changed local pids also.

   - SETLKW operations can now be interrupted only when the process is
     killed, and not from other signals. General interruption was
     resulting in previously acquired locks being cleared, not just the
     in-progress lock. Handling this correctly will require extending a
     cancel capability to user space (a future feature.)

   - If multiple threads are requesting posix locks (with SETLKW), fix
     incorrect matching of results to the requests.

  The dlm networking has several minor cleanups, and one notable change:

   - Avoid delaying ack messages for too long (used for message
     reliability), resulting in a backlog of un-acked messages. These
     could previously be delayed as a result of either too many or too
     few other messages being sent. Now an upper and lower threshold is
     used to determine when an ack should be sent"

* tag 'dlm-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/linux-dlm:
  fs: dlm: remove filter local comms on close
  fs: dlm: add send ack threshold and append acks to msgs
  fs: dlm: handle sequence numbers as atomic
  fs: dlm: handle lkb wait count as atomic_t
  fs: dlm: filter ourself midcomms calls
  fs: dlm: warn about messages from left nodes
  fs: dlm: move dlm_purge_lkb_callbacks to user module
  fs: dlm: cleanup STOP_IO bitflag set when stop io
  fs: dlm: don't check othercon twice
  fs: dlm: unregister memory at the very last
  fs: dlm: fix missing pending to false
  fs: dlm: clear pending bit when queue was empty
  fs: dlm: revert check required context while close
  fs: dlm: fix mismatch of plock results from userspace
  fs: dlm: make F_SETLK use unkillable wait_event
  fs: dlm: interrupt posix locks only when process is killed
  fs: dlm: fix cleanup pending ops when interrupted
  fs: dlm: return positive pid value for F_GETLK
  dlm: Replace all non-returning strlcpy with strscpy

Author: torvalds

fs/dlm/ast.c

@@ -36,23 +36,6 @@ void dlm_callback_set_last_ptr(struct dlm_callback **from,
 	*from = to;
 }
 
-void dlm_purge_lkb_callbacks(struct dlm_lkb *lkb)
-{
-	struct dlm_callback *cb, *safe;
-
-	list_for_each_entry_safe(cb, safe, &lkb->lkb_callbacks, list) {
-		list_del(&cb->list);
-		kref_put(&cb->ref, dlm_release_callback);
-	}
-
-	clear_bit(DLM_IFL_CB_PENDING_BIT, &lkb->lkb_iflags);
-
-	/* invalidate */
-	dlm_callback_set_last_ptr(&lkb->lkb_last_cast, NULL);
-	dlm_callback_set_last_ptr(&lkb->lkb_last_cb, NULL);
-	lkb->lkb_last_bast_mode = -1;
-}
-
 int dlm_enqueue_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode,
 			     int status, uint32_t sbflags)
 {
@@ -181,10 +164,12 @@ void dlm_callback_work(struct work_struct *work)
 
 	spin_lock(&lkb->lkb_cb_lock);
 	rv = dlm_dequeue_lkb_callback(lkb, &cb);
-	spin_unlock(&lkb->lkb_cb_lock);
-
-	if (WARN_ON_ONCE(rv == DLM_DEQUEUE_CALLBACK_EMPTY))
+	if (WARN_ON_ONCE(rv == DLM_DEQUEUE_CALLBACK_EMPTY)) {
+		clear_bit(DLM_IFL_CB_PENDING_BIT, &lkb->lkb_iflags);
+		spin_unlock(&lkb->lkb_cb_lock);
 		goto out;
+	}
+	spin_unlock(&lkb->lkb_cb_lock);
 
 	for (;;) {
 		castfn = lkb->lkb_astfn;

fs/dlm/ast.h

@@ -26,7 +26,6 @@ void dlm_callback_set_last_ptr(struct dlm_callback **from,
 			       struct dlm_callback *to);
 
 void dlm_release_callback(struct kref *ref);
-void dlm_purge_lkb_callbacks(struct dlm_lkb *lkb);
 void dlm_callback_work(struct work_struct *work);
 int dlm_callback_start(struct dlm_ls *ls);
 void dlm_callback_stop(struct dlm_ls *ls);

fs/dlm/dlm_internal.h

@@ -246,7 +246,7 @@ struct dlm_lkb {
 	int8_t			lkb_highbast;	/* highest mode bast sent for */
 
 	int8_t			lkb_wait_type;	/* type of reply waiting for */
-	int8_t			lkb_wait_count;
+	atomic_t		lkb_wait_count;
 	int			lkb_wait_nodeid; /* for debugging */
 
 	struct list_head	lkb_statequeue;	/* rsb g/c/w list */

fs/dlm/lock.c

@@ -1407,6 +1407,7 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid)
 {
 	struct dlm_ls *ls = lkb->lkb_resource->res_ls;
 	int error = 0;
+	int wc;
 
 	mutex_lock(&ls->ls_waiters_mutex);
 
@@ -1428,20 +1429,17 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid)
 			error = -EBUSY;
 			goto out;
 		}
-		lkb->lkb_wait_count++;
+		wc = atomic_inc_return(&lkb->lkb_wait_count);
 		hold_lkb(lkb);
 
 		log_debug(ls, "addwait %x cur %d overlap %d count %d f %x",
-			  lkb->lkb_id, lkb->lkb_wait_type, mstype,
-			  lkb->lkb_wait_count, dlm_iflags_val(lkb));
+			  lkb->lkb_id, lkb->lkb_wait_type, mstype, wc,
+			  dlm_iflags_val(lkb));
 		goto out;
 	}
 
-	DLM_ASSERT(!lkb->lkb_wait_count,
-		   dlm_print_lkb(lkb);
-		   printk("wait_count %d\n", lkb->lkb_wait_count););
-
-	lkb->lkb_wait_count++;
+	wc = atomic_fetch_inc(&lkb->lkb_wait_count);
+	DLM_ASSERT(!wc, dlm_print_lkb(lkb); printk("wait_count %d\n", wc););
 	lkb->lkb_wait_type = mstype;
 	lkb->lkb_wait_nodeid = to_nodeid; /* for debugging */
 	hold_lkb(lkb);
@@ -1504,7 +1502,7 @@ static int _remove_from_waiters(struct dlm_lkb *lkb, int mstype,
 		log_debug(ls, "remwait %x convert_reply zap overlap_cancel",
 			  lkb->lkb_id);
 		lkb->lkb_wait_type = 0;
-		lkb->lkb_wait_count--;
+		atomic_dec(&lkb->lkb_wait_count);
 		unhold_lkb(lkb);
 		goto out_del;
 	}
@@ -1531,16 +1529,15 @@ static int _remove_from_waiters(struct dlm_lkb *lkb, int mstype,
 	if (overlap_done && lkb->lkb_wait_type) {
 		log_error(ls, "remwait error %x reply %d wait_type %d overlap",
 			  lkb->lkb_id, mstype, lkb->lkb_wait_type);
-		lkb->lkb_wait_count--;
+		atomic_dec(&lkb->lkb_wait_count);
 		unhold_lkb(lkb);
 		lkb->lkb_wait_type = 0;
 	}
 
-	DLM_ASSERT(lkb->lkb_wait_count, dlm_print_lkb(lkb););
+	DLM_ASSERT(atomic_read(&lkb->lkb_wait_count), dlm_print_lkb(lkb););
 
 	clear_bit(DLM_IFL_RESEND_BIT, &lkb->lkb_iflags);
-	lkb->lkb_wait_count--;
-	if (!lkb->lkb_wait_count)
+	if (atomic_dec_and_test(&lkb->lkb_wait_count))
 		list_del_init(&lkb->lkb_wait_reply);
 	unhold_lkb(lkb);
 	return 0;
@@ -2669,7 +2666,7 @@ static int validate_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb,
 			goto out;
 
 		/* lock not allowed if there's any op in progress */
-		if (lkb->lkb_wait_type || lkb->lkb_wait_count)
+		if (lkb->lkb_wait_type || atomic_read(&lkb->lkb_wait_count))
 			goto out;
 
 		if (is_overlap(lkb))
@@ -2731,7 +2728,7 @@ static int validate_unlock_args(struct dlm_lkb *lkb, struct dlm_args *args)
 
 	/* normal unlock not allowed if there's any op in progress */
 	if (!(args->flags & (DLM_LKF_CANCEL | DLM_LKF_FORCEUNLOCK)) &&
-	    (lkb->lkb_wait_type || lkb->lkb_wait_count))
+	    (lkb->lkb_wait_type || atomic_read(&lkb->lkb_wait_count)))
 		goto out;
 
 	/* an lkb may be waiting for an rsb lookup to complete where the
@@ -4616,7 +4613,7 @@ static void _receive_message(struct dlm_ls *ls, struct dlm_message *ms,
 {
 	int error = 0, noent = 0;
 
-	if (!dlm_is_member(ls, le32_to_cpu(ms->m_header.h_nodeid))) {
+	if (WARN_ON_ONCE(!dlm_is_member(ls, le32_to_cpu(ms->m_header.h_nodeid)))) {
 		log_limit(ls, "receive %d from non-member %d %x %x %d",
 			  le32_to_cpu(ms->m_type),
 			  le32_to_cpu(ms->m_header.h_nodeid),
@@ -4754,7 +4751,7 @@ static void dlm_receive_message(struct dlm_ls *ls, struct dlm_message *ms,
 		/* If we were a member of this lockspace, left, and rejoined,
 		   other nodes may still be sending us messages from the
 		   lockspace generation before we left. */
-		if (!ls->ls_generation) {
+		if (WARN_ON_ONCE(!ls->ls_generation)) {
 			log_limit(ls, "receive %d from %d ignore old gen",
 				  le32_to_cpu(ms->m_type), nodeid);
 			return;
@@ -5066,10 +5063,9 @@ int dlm_recover_waiters_post(struct dlm_ls *ls)
 		/* drop all wait_count references we still
 		 * hold a reference for this iteration.
 		 */
-		while (lkb->lkb_wait_count) {
-			lkb->lkb_wait_count--;
+		while (!atomic_dec_and_test(&lkb->lkb_wait_count))
 			unhold_lkb(lkb);
-		}
+
 		mutex_lock(&ls->ls_waiters_mutex);
 		list_del_init(&lkb->lkb_wait_reply);
 		mutex_unlock(&ls->ls_waiters_mutex);

fs/dlm/lockspace.c

@@ -935,15 +935,3 @@ void dlm_stop_lockspaces(void)
 		log_print("dlm user daemon left %d lockspaces", count);
 }
 
-void dlm_stop_lockspaces_check(void)
-{
-	struct dlm_ls *ls;
-
-	spin_lock(&lslist_lock);
-	list_for_each_entry(ls, &lslist, ls_list) {
-		if (WARN_ON(!rwsem_is_locked(&ls->ls_in_recovery) ||
-			    !dlm_locking_stopped(ls)))
-			break;
-	}
-	spin_unlock(&lslist_lock);
-}

fs/dlm/lockspace.h

@@ -27,7 +27,6 @@ struct dlm_ls *dlm_find_lockspace_local(void *id);
 struct dlm_ls *dlm_find_lockspace_device(int minor);
 void dlm_put_lockspace(struct dlm_ls *ls);
 void dlm_stop_lockspaces(void);
-void dlm_stop_lockspaces_check(void);
 int dlm_new_user_lockspace(const char *name, const char *cluster,
 			   uint32_t flags, int lvblen,
 			   const struct dlm_lockspace_ops *ops,

fs/dlm/lowcomms.c

@@ -546,9 +546,6 @@ int dlm_lowcomms_connect_node(int nodeid)
 	struct connection *con;
 	int idx;
 
-	if (nodeid == dlm_our_nodeid())
-		return 0;
-
 	idx = srcu_read_lock(&connections_srcu);
 	con = nodeid2con(nodeid, 0);
 	if (WARN_ON_ONCE(!con)) {
@@ -735,19 +732,15 @@ static void stop_connection_io(struct connection *con)
 	if (con->othercon)
 		stop_connection_io(con->othercon);
 
+	spin_lock_bh(&con->writequeue_lock);
+	set_bit(CF_IO_STOP, &con->flags);
+	spin_unlock_bh(&con->writequeue_lock);
+
 	down_write(&con->sock_lock);
 	if (con->sock) {
 		lock_sock(con->sock->sk);
 		restore_callbacks(con->sock->sk);
-
-		spin_lock_bh(&con->writequeue_lock);
-		set_bit(CF_IO_STOP, &con->flags);
-		spin_unlock_bh(&con->writequeue_lock);
 		release_sock(con->sock->sk);
-	} else {
-		spin_lock_bh(&con->writequeue_lock);
-		set_bit(CF_IO_STOP, &con->flags);
-		spin_unlock_bh(&con->writequeue_lock);
 	}
 	up_write(&con->sock_lock);
 
@@ -867,37 +860,16 @@ struct dlm_processed_nodes {
 	struct list_head list;
 };
 
-static void add_processed_node(int nodeid, struct list_head *processed_nodes)
-{
-	struct dlm_processed_nodes *n;
-
-	list_for_each_entry(n, processed_nodes, list) {
-		/* we already remembered this node */
-		if (n->nodeid == nodeid)
-			return;
-	}
-
-	/* if it's fails in worst case we simple don't send an ack back.
-	 * We try it next time.
-	 */
-	n = kmalloc(sizeof(*n), GFP_NOFS);
-	if (!n)
-		return;
-
-	n->nodeid = nodeid;
-	list_add(&n->list, processed_nodes);
-}
-
 static void process_dlm_messages(struct work_struct *work)
 {
-	struct dlm_processed_nodes *n, *n_tmp;
 	struct processqueue_entry *pentry;
 	LIST_HEAD(processed_nodes);
 
 	spin_lock(&processqueue_lock);
 	pentry = list_first_entry_or_null(&processqueue,
 					  struct processqueue_entry, list);
 	if (WARN_ON_ONCE(!pentry)) {
+		process_dlm_messages_pending = false;
 		spin_unlock(&processqueue_lock);
 		return;
 	}
@@ -908,7 +880,6 @@ static void process_dlm_messages(struct work_struct *work)
 	for (;;) {
 		dlm_process_incoming_buffer(pentry->nodeid, pentry->buf,
 					    pentry->buflen);
-		add_processed_node(pentry->nodeid, &processed_nodes);
 		free_processqueue_entry(pentry);
 
 		spin_lock(&processqueue_lock);
@@ -923,13 +894,6 @@ static void process_dlm_messages(struct work_struct *work)
 		list_del(&pentry->list);
 		spin_unlock(&processqueue_lock);
 	}
-
-	/* send ack back after we processed couple of messages */
-	list_for_each_entry_safe(n, n_tmp, &processed_nodes, list) {
-		list_del(&n->list);
-		dlm_midcomms_receive_done(n->nodeid);
-		kfree(n);
-	}
 }
 
 /* Data received from remote end */
@@ -1500,8 +1464,7 @@ int dlm_lowcomms_close(int nodeid)
 	call_srcu(&connections_srcu, &con->rcu, connection_release);
 	if (con->othercon) {
 		clean_one_writequeue(con->othercon);
-		if (con->othercon)
-			call_srcu(&connections_srcu, &con->othercon->rcu, connection_release);
+		call_srcu(&connections_srcu, &con->othercon->rcu, connection_release);
 	}
 	srcu_read_unlock(&connections_srcu, idx);
 

fs/dlm/main.c

@@ -73,10 +73,10 @@ static void __exit exit_dlm(void)
 	dlm_plock_exit();
 	dlm_user_exit();
 	dlm_config_exit();
-	dlm_memory_exit();
 	dlm_lockspace_exit();
 	dlm_midcomms_exit();
 	dlm_unregister_debugfs();
+	dlm_memory_exit();
 }
 
 module_init(init_dlm);

fs/dlm/member.c

@@ -307,6 +307,21 @@ static void add_ordered_member(struct dlm_ls *ls, struct dlm_member *new)
 	}
 }
 
+static int add_remote_member(int nodeid)
+{
+	int error;
+
+	if (nodeid == dlm_our_nodeid())
+		return 0;
+
+	error = dlm_lowcomms_connect_node(nodeid);
+	if (error < 0)
+		return error;
+
+	dlm_midcomms_add_member(nodeid);
+	return 0;
+}
+
 static int dlm_add_member(struct dlm_ls *ls, struct dlm_config_node *node)
 {
 	struct dlm_member *memb;
@@ -316,16 +331,16 @@ static int dlm_add_member(struct dlm_ls *ls, struct dlm_config_node *node)
 	if (!memb)
 		return -ENOMEM;
 
-	error = dlm_lowcomms_connect_node(node->nodeid);
+	memb->nodeid = node->nodeid;
+	memb->weight = node->weight;
+	memb->comm_seq = node->comm_seq;
+
+	error = add_remote_member(node->nodeid);
 	if (error < 0) {
 		kfree(memb);
 		return error;
 	}
 
-	memb->nodeid = node->nodeid;
-	memb->weight = node->weight;
-	memb->comm_seq = node->comm_seq;
-	dlm_midcomms_add_member(node->nodeid);
 	add_ordered_member(ls, memb);
 	ls->ls_num_nodes++;
 	return 0;
@@ -370,11 +385,19 @@ static void clear_memb_list(struct list_head *head,
 	}
 }
 
-static void clear_members_cb(int nodeid)
+static void remove_remote_member(int nodeid)
 {
+	if (nodeid == dlm_our_nodeid())
+		return;
+
 	dlm_midcomms_remove_member(nodeid);
 }
 
+static void clear_members_cb(int nodeid)
+{
+	remove_remote_member(nodeid);
+}
+
 void dlm_clear_members(struct dlm_ls *ls)
 {
 	clear_memb_list(&ls->ls_nodes, clear_members_cb);
@@ -562,7 +585,7 @@ int dlm_recover_members(struct dlm_ls *ls, struct dlm_recover *rv, int *neg_out)
 
 		neg++;
 		list_move(&memb->list, &ls->ls_nodes_gone);
-		dlm_midcomms_remove_member(memb->nodeid);
+		remove_remote_member(memb->nodeid);
 		ls->ls_num_nodes--;
 		dlm_lsop_recover_slot(ls, memb);
 	}