4 files changed, 134 insertions, 139 deletions

diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c
index a8020293f34..ee71ea26777 100644
--- a/net/ceph/ceph_common.c
+++ b/net/ceph/ceph_common.c
@@ -305,7 +305,6 @@ ceph_parse_options(char *options, const char *dev_name,
 
 	/* start with defaults */
 	opt->flags = CEPH_OPT_DEFAULT;
-	opt->osd_timeout = CEPH_OSD_TIMEOUT_DEFAULT;
 	opt->osd_keepalive_timeout = CEPH_OSD_KEEPALIVE_DEFAULT;
 	opt->mount_timeout = CEPH_MOUNT_TIMEOUT_DEFAULT; /* seconds */
 	opt->osd_idle_ttl = CEPH_OSD_IDLE_TTL_DEFAULT;   /* seconds */
@@ -391,7 +390,7 @@ ceph_parse_options(char *options, const char *dev_name,
 
 			/* misc */
 		case Opt_osdtimeout:
-			opt->osd_timeout = intval;
+			pr_warning("ignoring deprecated osdtimeout option\n");
 			break;
 		case Opt_osdkeepalivetimeout:
 			opt->osd_keepalive_timeout = intval;
diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c
index 3ef1759403b..5ccf87ed8d6 100644
--- a/net/ceph/messenger.c
+++ b/net/ceph/messenger.c
@@ -506,6 +506,7 @@ static void reset_connection(struct ceph_connection *con)
 {
 	/* reset connection, out_queue, msg_ and connect_seq */
 	/* discard existing out_queue and msg_seq */
+	dout("reset_connection %p\n", con);
 	ceph_msg_remove_list(&con->out_queue);
 	ceph_msg_remove_list(&con->out_sent);
 
@@ -561,7 +562,7 @@ void ceph_con_open(struct ceph_connection *con,
 	mutex_lock(&con->mutex);
 	dout("con_open %p %s\n", con, ceph_pr_addr(&addr->in_addr));
 
-	BUG_ON(con->state != CON_STATE_CLOSED);
+	WARN_ON(con->state != CON_STATE_CLOSED);
 	con->state = CON_STATE_PREOPEN;
 
 	con->peer_name.type = (__u8) entity_type;
@@ -1506,13 +1507,6 @@ static int process_banner(struct ceph_connection *con)
 	return 0;
 }
 
-static void fail_protocol(struct ceph_connection *con)
-{
-	reset_connection(con);
-	BUG_ON(con->state != CON_STATE_NEGOTIATING);
-	con->state = CON_STATE_CLOSED;
-}
-
 static int process_connect(struct ceph_connection *con)
 {
 	u64 sup_feat = con->msgr->supported_features;
@@ -1530,7 +1524,7 @@ static int process_connect(struct ceph_connection *con)
 		       ceph_pr_addr(&con->peer_addr.in_addr),
 		       sup_feat, server_feat, server_feat & ~sup_feat);
 		con->error_msg = "missing required protocol features";
-		fail_protocol(con);
+		reset_connection(con);
 		return -1;
 
 	case CEPH_MSGR_TAG_BADPROTOVER:
@@ -1541,7 +1535,7 @@ static int process_connect(struct ceph_connection *con)
 		       le32_to_cpu(con->out_connect.protocol_version),
 		       le32_to_cpu(con->in_reply.protocol_version));
 		con->error_msg = "protocol version mismatch";
-		fail_protocol(con);
+		reset_connection(con);
 		return -1;
 
 	case CEPH_MSGR_TAG_BADAUTHORIZER:
@@ -1631,11 +1625,11 @@ static int process_connect(struct ceph_connection *con)
 			       ceph_pr_addr(&con->peer_addr.in_addr),
 			       req_feat, server_feat, req_feat & ~server_feat);
 			con->error_msg = "missing required protocol features";
-			fail_protocol(con);
+			reset_connection(con);
 			return -1;
 		}
 
-		BUG_ON(con->state != CON_STATE_NEGOTIATING);
+		WARN_ON(con->state != CON_STATE_NEGOTIATING);
 		con->state = CON_STATE_OPEN;
 
 		con->peer_global_seq = le32_to_cpu(con->in_reply.global_seq);
@@ -2132,7 +2126,6 @@ more:
 		if (ret < 0)
 			goto out;
 
-		BUG_ON(con->state != CON_STATE_CONNECTING);
 		con->state = CON_STATE_NEGOTIATING;
 
 		/*
@@ -2160,7 +2153,7 @@ more:
 		goto more;
 	}
 
-	BUG_ON(con->state != CON_STATE_OPEN);
+	WARN_ON(con->state != CON_STATE_OPEN);
 
 	if (con->in_base_pos < 0) {
 		/*
@@ -2244,22 +2237,62 @@ bad_tag:
 
 
 /*
- * Atomically queue work on a connection.  Bump @con reference to
- * avoid races with connection teardown.
+ * Atomically queue work on a connection after the specified delay.
+ * Bump @con reference to avoid races with connection teardown.
+ * Returns 0 if work was queued, or an error code otherwise.
  */
-static void queue_con(struct ceph_connection *con)
+static int queue_con_delay(struct ceph_connection *con, unsigned long delay)
 {
 	if (!con->ops->get(con)) {
-		dout("queue_con %p ref count 0\n", con);
-		return;
+		dout("%s %p ref count 0\n", __func__, con);
+
+		return -ENOENT;
 	}
 
-	if (!queue_delayed_work(ceph_msgr_wq, &con->work, 0)) {
-		dout("queue_con %p - already queued\n", con);
+	if (!queue_delayed_work(ceph_msgr_wq, &con->work, delay)) {
+		dout("%s %p - already queued\n", __func__, con);
 		con->ops->put(con);
-	} else {
-		dout("queue_con %p\n", con);
+
+		return -EBUSY;
+	}
+
+	dout("%s %p %lu\n", __func__, con, delay);
+
+	return 0;
+}
+
+static void queue_con(struct ceph_connection *con)
+{
+	(void) queue_con_delay(con, 0);
+}
+
+static bool con_sock_closed(struct ceph_connection *con)
+{
+	if (!test_and_clear_bit(CON_FLAG_SOCK_CLOSED, &con->flags))
+		return false;
+
+#define CASE(x)								\
+	case CON_STATE_ ## x:						\
+		con->error_msg = "socket closed (con state " #x ")";	\
+		break;
+
+	switch (con->state) {
+	CASE(CLOSED);
+	CASE(PREOPEN);
+	CASE(CONNECTING);
+	CASE(NEGOTIATING);
+	CASE(OPEN);
+	CASE(STANDBY);
+	default:
+		pr_warning("%s con %p unrecognized state %lu\n",
+			__func__, con, con->state);
+		con->error_msg = "unrecognized con state";
+		BUG();
+		break;
 	}
+#undef CASE
+
+	return true;
 }
 
 /*
@@ -2273,35 +2306,16 @@ static void con_work(struct work_struct *work)
 
 	mutex_lock(&con->mutex);
 restart:
-	if (test_and_clear_bit(CON_FLAG_SOCK_CLOSED, &con->flags)) {
-		switch (con->state) {
-		case CON_STATE_CONNECTING:
-			con->error_msg = "connection failed";
-			break;
-		case CON_STATE_NEGOTIATING:
-			con->error_msg = "negotiation failed";
-			break;
-		case CON_STATE_OPEN:
-			con->error_msg = "socket closed";
-			break;
-		default:
-			dout("unrecognized con state %d\n", (int)con->state);
-			con->error_msg = "unrecognized con state";
-			BUG();
-		}
+	if (con_sock_closed(con))
 		goto fault;
-	}
 
 	if (test_and_clear_bit(CON_FLAG_BACKOFF, &con->flags)) {
 		dout("con_work %p backing off\n", con);
-		if (queue_delayed_work(ceph_msgr_wq, &con->work,
-				       round_jiffies_relative(con->delay))) {
-			dout("con_work %p backoff %lu\n", con, con->delay);
-			mutex_unlock(&con->mutex);
-			return;
-		} else {
+		ret = queue_con_delay(con, round_jiffies_relative(con->delay));
+		if (ret) {
 			dout("con_work %p FAILED to back off %lu\n", con,
 			     con->delay);
+			BUG_ON(ret == -ENOENT);
 			set_bit(CON_FLAG_BACKOFF, &con->flags);
 		}
 		goto done;
@@ -2356,12 +2370,12 @@ fault:
 static void ceph_fault(struct ceph_connection *con)
 	__releases(con->mutex)
 {
-	pr_err("%s%lld %s %s\n", ENTITY_NAME(con->peer_name),
+	pr_warning("%s%lld %s %s\n", ENTITY_NAME(con->peer_name),
 	       ceph_pr_addr(&con->peer_addr.in_addr), con->error_msg);
 	dout("fault %p state %lu to peer %s\n",
 	     con, con->state, ceph_pr_addr(&con->peer_addr.in_addr));
 
-	BUG_ON(con->state != CON_STATE_CONNECTING &&
+	WARN_ON(con->state != CON_STATE_CONNECTING &&
 	       con->state != CON_STATE_NEGOTIATING &&
 	       con->state != CON_STATE_OPEN);
 
@@ -2398,24 +2412,8 @@ static void ceph_fault(struct ceph_connection *con)
 			con->delay = BASE_DELAY_INTERVAL;
 		else if (con->delay < MAX_DELAY_INTERVAL)
 			con->delay *= 2;
-		con->ops->get(con);
-		if (queue_delayed_work(ceph_msgr_wq, &con->work,
-				       round_jiffies_relative(con->delay))) {
-			dout("fault queued %p delay %lu\n", con, con->delay);
-		} else {
-			con->ops->put(con);
-			dout("fault failed to queue %p delay %lu, backoff\n",
-			     con, con->delay);
-			/*
-			 * In many cases we see a socket state change
-			 * while con_work is running and end up
-			 * queuing (non-delayed) work, such that we
-			 * can't backoff with a delay.  Set a flag so
-			 * that when con_work restarts we schedule the
-			 * delay then.
-			 */
-			set_bit(CON_FLAG_BACKOFF, &con->flags);
-		}
+		set_bit(CON_FLAG_BACKOFF, &con->flags);
+		queue_con(con);
 	}
 
 out_unlock:
diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c
index c1d756cc744..eb9a4447876 100644
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -221,6 +221,7 @@ struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
 	kref_init(&req->r_kref);
 	init_completion(&req->r_completion);
 	init_completion(&req->r_safe_completion);
+	RB_CLEAR_NODE(&req->r_node);
 	INIT_LIST_HEAD(&req->r_unsafe_item);
 	INIT_LIST_HEAD(&req->r_linger_item);
 	INIT_LIST_HEAD(&req->r_linger_osd);
@@ -580,7 +581,7 @@ static void __kick_osd_requests(struct ceph_osd_client *osdc,
 
 	dout("__kick_osd_requests osd%d\n", osd->o_osd);
 	err = __reset_osd(osdc, osd);
-	if (err == -EAGAIN)
+	if (err)
 		return;
 
 	list_for_each_entry(req, &osd->o_requests, r_osd_item) {
@@ -607,14 +608,6 @@ static void __kick_osd_requests(struct ceph_osd_client *osdc,
 	}
 }
 
-static void kick_osd_requests(struct ceph_osd_client *osdc,
-			      struct ceph_osd *kickosd)
-{
-	mutex_lock(&osdc->request_mutex);
-	__kick_osd_requests(osdc, kickosd);
-	mutex_unlock(&osdc->request_mutex);
-}
-
 /*
  * If the osd connection drops, we need to resubmit all requests.
  */
@@ -628,7 +621,9 @@ static void osd_reset(struct ceph_connection *con)
 	dout("osd_reset osd%d\n", osd->o_osd);
 	osdc = osd->o_osdc;
 	down_read(&osdc->map_sem);
-	kick_osd_requests(osdc, osd);
+	mutex_lock(&osdc->request_mutex);
+	__kick_osd_requests(osdc, osd);
+	mutex_unlock(&osdc->request_mutex);
 	send_queued(osdc);
 	up_read(&osdc->map_sem);
 }
@@ -647,6 +642,7 @@ static struct ceph_osd *create_osd(struct ceph_osd_client *osdc, int onum)
 	atomic_set(&osd->o_ref, 1);
 	osd->o_osdc = osdc;
 	osd->o_osd = onum;
+	RB_CLEAR_NODE(&osd->o_node);
 	INIT_LIST_HEAD(&osd->o_requests);
 	INIT_LIST_HEAD(&osd->o_linger_requests);
 	INIT_LIST_HEAD(&osd->o_osd_lru);
@@ -750,6 +746,7 @@ static int __reset_osd(struct ceph_osd_client *osdc, struct ceph_osd *osd)
 	if (list_empty(&osd->o_requests) &&
 	    list_empty(&osd->o_linger_requests)) {
 		__remove_osd(osdc, osd);
+		ret = -ENODEV;
 	} else if (memcmp(&osdc->osdmap->osd_addr[osd->o_osd],
 			  &osd->o_con.peer_addr,
 			  sizeof(osd->o_con.peer_addr)) == 0 &&
@@ -876,9 +873,9 @@ static void __unregister_request(struct ceph_osd_client *osdc,
 			req->r_osd = NULL;
 	}
 
+	list_del_init(&req->r_req_lru_item);
 	ceph_osdc_put_request(req);
 
-	list_del_init(&req->r_req_lru_item);
 	if (osdc->num_requests == 0) {
 		dout(" no requests, canceling timeout\n");
 		__cancel_osd_timeout(osdc);
@@ -910,8 +907,8 @@ static void __unregister_linger_request(struct ceph_osd_client *osdc,
 					struct ceph_osd_request *req)
 {
 	dout("__unregister_linger_request %p\n", req);
+	list_del_init(&req->r_linger_item);
 	if (req->r_osd) {
-		list_del_init(&req->r_linger_item);
 		list_del_init(&req->r_linger_osd);
 
 		if (list_empty(&req->r_osd->o_requests) &&
@@ -1090,12 +1087,10 @@ static void handle_timeout(struct work_struct *work)
 {
 	struct ceph_osd_client *osdc =
 		container_of(work, struct ceph_osd_client, timeout_work.work);
-	struct ceph_osd_request *req, *last_req = NULL;
+	struct ceph_osd_request *req;
 	struct ceph_osd *osd;
-	unsigned long timeout = osdc->client->options->osd_timeout * HZ;
 	unsigned long keepalive =
 		osdc->client->options->osd_keepalive_timeout * HZ;
-	unsigned long last_stamp = 0;
 	struct list_head slow_osds;
 	dout("timeout\n");
 	down_read(&osdc->map_sem);
@@ -1105,37 +1100,6 @@ static void handle_timeout(struct work_struct *work)
 	mutex_lock(&osdc->request_mutex);
 
 	/*
-	 * reset osds that appear to be _really_ unresponsive.  this
-	 * is a failsafe measure.. we really shouldn't be getting to
-	 * this point if the system is working properly.  the monitors
-	 * should mark the osd as failed and we should find out about
-	 * it from an updated osd map.
-	 */
-	while (timeout && !list_empty(&osdc->req_lru)) {
-		req = list_entry(osdc->req_lru.next, struct ceph_osd_request,
-				 r_req_lru_item);
-
-		/* hasn't been long enough since we sent it? */
-		if (time_before(jiffies, req->r_stamp + timeout))
-			break;
-
-		/* hasn't been long enough since it was acked? */
-		if (req->r_request->ack_stamp == 0 ||
-		    time_before(jiffies, req->r_request->ack_stamp + timeout))
-			break;
-
-		BUG_ON(req == last_req && req->r_stamp == last_stamp);
-		last_req = req;
-		last_stamp = req->r_stamp;
-
-		osd = req->r_osd;
-		BUG_ON(!osd);
-		pr_warning(" tid %llu timed out on osd%d, will reset osd\n",
-			   req->r_tid, osd->o_osd);
-		__kick_osd_requests(osdc, osd);
-	}
-
-	/*
 	 * ping osds that are a bit slow.  this ensures that if there
 	 * is a break in the TCP connection we will notice, and reopen
 	 * a connection with that osd (from the fault callback).
@@ -1306,7 +1270,7 @@ static void reset_changed_osds(struct ceph_osd_client *osdc)
  * Requeue requests whose mapping to an OSD has changed.  If requests map to
  * no osd, request a new map.
  *
- * Caller should hold map_sem for read and request_mutex.
+ * Caller should hold map_sem for read.
  */
 static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 {
@@ -1320,6 +1284,24 @@ static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 	for (p = rb_first(&osdc->requests); p; ) {
 		req = rb_entry(p, struct ceph_osd_request, r_node);
 		p = rb_next(p);
+
+		/*
+		 * For linger requests that have not yet been
+		 * registered, move them to the linger list; they'll
+		 * be sent to the osd in the loop below.  Unregister
+		 * the request before re-registering it as a linger
+		 * request to ensure the __map_request() below
+		 * will decide it needs to be sent.
+		 */
+		if (req->r_linger && list_empty(&req->r_linger_item)) {
+			dout("%p tid %llu restart on osd%d\n",
+			     req, req->r_tid,
+			     req->r_osd ? req->r_osd->o_osd : -1);
+			__unregister_request(osdc, req);
+			__register_linger_request(osdc, req);
+			continue;
+		}
+
 		err = __map_request(osdc, req, force_resend);
 		if (err < 0)
 			continue;  /* error */
@@ -1334,17 +1316,6 @@ static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 				req->r_flags |= CEPH_OSD_FLAG_RETRY;
 			}
 		}
-		if (req->r_linger && list_empty(&req->r_linger_item)) {
-			/*
-			 * register as a linger so that we will
-			 * re-submit below and get a new tid
-			 */
-			dout("%p tid %llu restart on osd%d\n",
-			     req, req->r_tid,
-			     req->r_osd ? req->r_osd->o_osd : -1);
-			__register_linger_request(osdc, req);
-			__unregister_request(osdc, req);
-		}
 	}
 
 	list_for_each_entry_safe(req, nreq, &osdc->req_linger,
@@ -1352,6 +1323,7 @@ static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 		dout("linger req=%p req->r_osd=%p\n", req, req->r_osd);
 
 		err = __map_request(osdc, req, force_resend);
+		dout("__map_request returned %d\n", err);
 		if (err == 0)
 			continue;  /* no change and no osd was specified */
 		if (err < 0)
@@ -1364,8 +1336,8 @@ static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 
 		dout("kicking lingering %p tid %llu osd%d\n", req, req->r_tid,
 		     req->r_osd ? req->r_osd->o_osd : -1);
-		__unregister_linger_request(osdc, req);
 		__register_request(osdc, req);
+		__unregister_linger_request(osdc, req);
 	}
 	mutex_unlock(&osdc->request_mutex);
 
@@ -1373,6 +1345,7 @@ static void kick_requests(struct ceph_osd_client *osdc, int force_resend)
 		dout("%d requests for down osds, need new map\n", needmap);
 		ceph_monc_request_next_osdmap(&osdc->client->monc);
 	}
+	reset_changed_osds(osdc);
 }
 
 
@@ -1429,7 +1402,6 @@ void ceph_osdc_handle_map(struct ceph_osd_client *osdc, struct ceph_msg *msg)
 				osdc->osdmap = newmap;
 			}
 			kick_requests(osdc, 0);
-			reset_changed_osds(osdc);
 		} else {
 			dout("ignoring incremental map %u len %d\n",
 			     epoch, maplen);
@@ -1599,6 +1571,7 @@ int ceph_osdc_create_event(struct ceph_osd_client *osdc,
 	event->data = data;
 	event->osdc = osdc;
 	INIT_LIST_HEAD(&event->osd_node);
+	RB_CLEAR_NODE(&event->node);
 	kref_init(&event->kref);   /* one ref for us */
 	kref_get(&event->kref);    /* one ref for the caller */
 	init_completion(&event->completion);
diff --git a/net/ceph/osdmap.c b/net/ceph/osdmap.c
index 5433fb0eb3c..de73214b5d2 100644
--- a/net/ceph/osdmap.c
+++ b/net/ceph/osdmap.c
@@ -469,6 +469,22 @@ static struct ceph_pg_pool_info *__lookup_pg_pool(struct rb_root *root, int id)
 	return NULL;
 }
 
+const char *ceph_pg_pool_name_by_id(struct ceph_osdmap *map, u64 id)
+{
+	struct ceph_pg_pool_info *pi;
+
+	if (id == CEPH_NOPOOL)
+		return NULL;
+
+	if (WARN_ON_ONCE(id > (u64) INT_MAX))
+		return NULL;
+
+	pi = __lookup_pg_pool(&map->pg_pools, (int) id);
+
+	return pi ? pi->name : NULL;
+}
+EXPORT_SYMBOL(ceph_pg_pool_name_by_id);
+
 int ceph_pg_poolid_by_name(struct ceph_osdmap *map, const char *name)
 {
 	struct rb_node *rbp;
@@ -645,10 +661,12 @@ struct ceph_osdmap *osdmap_decode(void **p, void *end)
 	ceph_decode_32_safe(p, end, max, bad);
 	while (max--) {
 		ceph_decode_need(p, end, 4 + 1 + sizeof(pi->v), bad);
+		err = -ENOMEM;
 		pi = kzalloc(sizeof(*pi), GFP_NOFS);
 		if (!pi)
 			goto bad;
 		pi->id = ceph_decode_32(p);
+		err = -EINVAL;
 		ev = ceph_decode_8(p); /* encoding version */
 		if (ev > CEPH_PG_POOL_VERSION) {
 			pr_warning("got unknown v %d > %d of ceph_pg_pool\n",
@@ -664,8 +682,13 @@ struct ceph_osdmap *osdmap_decode(void **p, void *end)
 		__insert_pg_pool(&map->pg_pools, pi);
 	}
 
-	if (version >= 5 && __decode_pool_names(p, end, map) < 0)
-		goto bad;
+	if (version >= 5) {
+		err = __decode_pool_names(p, end, map);
+		if (err < 0) {
+			dout("fail to decode pool names");
+			goto bad;
+		}
+	}
 
 	ceph_decode_32_safe(p, end, map->pool_max, bad);
 
@@ -745,7 +768,7 @@ struct ceph_osdmap *osdmap_decode(void **p, void *end)
 	return map;
 
 bad:
-	dout("osdmap_decode fail\n");
+	dout("osdmap_decode fail err %d\n", err);
 	ceph_osdmap_destroy(map);
 	return ERR_PTR(err);
 }
@@ -839,6 +862,7 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
 		if (ev > CEPH_PG_POOL_VERSION) {
 			pr_warning("got unknown v %d > %d of ceph_pg_pool\n",
 				   ev, CEPH_PG_POOL_VERSION);
+			err = -EINVAL;
 			goto bad;
 		}
 		pi = __lookup_pg_pool(&map->pg_pools, pool);
@@ -855,8 +879,11 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
 		if (err < 0)
 			goto bad;
 	}
-	if (version >= 5 && __decode_pool_names(p, end, map) < 0)
-		goto bad;
+	if (version >= 5) {
+		err = __decode_pool_names(p, end, map);
+		if (err < 0)
+			goto bad;
+	}
 
 	/* old_pool */
 	ceph_decode_32_safe(p, end, len, bad);
@@ -932,15 +959,13 @@ struct ceph_osdmap *osdmap_apply_incremental(void **p, void *end,
 			(void) __remove_pg_mapping(&map->pg_temp, pgid);
 
 			/* insert */
-			if (pglen > (UINT_MAX - sizeof(*pg)) / sizeof(u32)) {
-				err = -EINVAL;
+			err = -EINVAL;
+			if (pglen > (UINT_MAX - sizeof(*pg)) / sizeof(u32))
 				goto bad;
-			}
+			err = -ENOMEM;
 			pg = kmalloc(sizeof(*pg) + sizeof(u32)*pglen, GFP_NOFS);
-			if (!pg) {
-				err = -ENOMEM;
+			if (!pg)
 				goto bad;
-			}
 			pg->pgid = pgid;
 			pg->len = pglen;
 			for (j = 0; j < pglen; j++)