net/smc: check for healthy link group resp. connections

If a problem for at least one connection of a link group is detected,
the whole link group and all its connections are terminated.
This patch adds a check for healthy link group when trying to reserve
a work request, and checks for healthy connections before starting
a tx worker.

Signed-off-by: Ursula Braun <ubraun@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

4 files changed, 29 insertions, 12 deletions

diff --git a/net/smc/smc_cdc.c b/net/smc/smc_cdc.c
index 51805334e001..6e8f5fbe0f09 100644
--- a/net/smc/smc_cdc.c
+++ b/net/smc/smc_cdc.c
@@ -65,9 +65,14 @@ int smc_cdc_get_free_slot(struct smc_connection *conn,
 			  struct smc_cdc_tx_pend **pend)
 {
 	struct smc_link *link = &conn->lgr->lnk[SMC_SINGLE_LINK];
+	int rc;
 
-	return smc_wr_tx_get_free_slot(link, smc_cdc_tx_handler, wr_buf,
-				       (struct smc_wr_tx_pend_priv **)pend);
+	rc = smc_wr_tx_get_free_slot(link, smc_cdc_tx_handler, wr_buf,
+				     (struct smc_wr_tx_pend_priv **)pend);
+	if (!conn->alert_token_local)
+		/* abnormal termination */
+		rc = -EPIPE;
+	return rc;
 }
 
 static inline void smc_cdc_add_pending_send(struct smc_connection *conn,
diff --git a/net/smc/smc_diag.c b/net/smc/smc_diag.c
index d2d01cf70224..427b91c1c964 100644
--- a/net/smc/smc_diag.c
+++ b/net/smc/smc_diag.c
@@ -86,7 +86,8 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb,
 	if (smc_diag_msg_attrs_fill(sk, skb, r, user_ns))
 		goto errout;
 
-	if ((req->diag_ext & (1 << (SMC_DIAG_CONNINFO - 1))) && smc->conn.lgr) {
+	if ((req->diag_ext & (1 << (SMC_DIAG_CONNINFO - 1))) &&
+	    smc->conn.alert_token_local) {
 		struct smc_connection *conn = &smc->conn;
 		struct smc_diag_conninfo cinfo = {
 			.token = conn->alert_token_local,
@@ -124,7 +125,8 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb,
 			goto errout;
 	}
 
-	if ((req->diag_ext & (1 << (SMC_DIAG_LGRINFO - 1))) && smc->conn.lgr) {
+	if ((req->diag_ext & (1 << (SMC_DIAG_LGRINFO - 1))) && smc->conn.lgr &&
+	    !list_empty(&smc->conn.lgr->list)) {
 		struct smc_diag_lgrinfo linfo = {
 			.role = smc->conn.lgr->role,
 			.lnk[0].ibport = smc->conn.lgr->lnk[0].ibport,
diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c
index 71b7d9f079f0..838bce20c361 100644
--- a/net/smc/smc_tx.c
+++ b/net/smc/smc_tx.c
@@ -408,8 +408,9 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn)
 				goto out_unlock;
 			}
 			rc = 0;
-			schedule_delayed_work(&conn->tx_work,
-					      SMC_TX_WORK_DELAY);
+			if (conn->alert_token_local) /* connection healthy */
+				schedule_delayed_work(&conn->tx_work,
+						      SMC_TX_WORK_DELAY);
 		}
 		goto out_unlock;
 	}
@@ -440,10 +441,17 @@ static void smc_tx_work(struct work_struct *work)
 	int rc;
 
 	lock_sock(&smc->sk);
+	if (smc->sk.sk_err ||
+	    !conn->alert_token_local ||
+	    conn->local_rx_ctrl.conn_state_flags.peer_conn_abort)
+		goto out;
+
 	rc = smc_tx_sndbuf_nonempty(conn);
 	if (!rc && conn->local_rx_ctrl.prod_flags.write_blocked &&
 	    !atomic_read(&conn->bytes_to_rcv))
 		conn->local_rx_ctrl.prod_flags.write_blocked = 0;
+
+out:
 	release_sock(&smc->sk);
 }
 
@@ -464,7 +472,8 @@ void smc_tx_consumer_update(struct smc_connection *conn)
 	    ((to_confirm > conn->rmbe_update_limit) &&
 	     ((to_confirm > (conn->rmbe_size / 2)) ||
 	      conn->local_rx_ctrl.prod_flags.write_blocked))) {
-		if (smc_cdc_get_slot_and_msg_send(conn) < 0) {
+		if ((smc_cdc_get_slot_and_msg_send(conn) < 0) &&
+		    conn->alert_token_local) { /* connection healthy */
 			schedule_delayed_work(&conn->tx_work,
 					      SMC_TX_WORK_DELAY);
 			return;
diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c
index 621c65850a18..1b8af23e6e2b 100644
--- a/net/smc/smc_wr.c
+++ b/net/smc/smc_wr.c
@@ -174,9 +174,9 @@ int smc_wr_tx_get_free_slot(struct smc_link *link,
 			    struct smc_wr_tx_pend_priv **wr_pend_priv)
 {
 	struct smc_wr_tx_pend *wr_pend;
+	u32 idx = link->wr_tx_cnt;
 	struct ib_send_wr *wr_ib;
 	u64 wr_id;
-	u32 idx;
 	int rc;
 
 	*wr_buf = NULL;
@@ -186,16 +186,17 @@ int smc_wr_tx_get_free_slot(struct smc_link *link,
 		if (rc)
 			return rc;
 	} else {
+		struct smc_link_group *lgr;
+
+		lgr = container_of(link, struct smc_link_group,
+				   lnk[SMC_SINGLE_LINK]);
 		rc = wait_event_timeout(
 			link->wr_tx_wait,
+			list_empty(&lgr->list) || /* lgr terminated */
 			(smc_wr_tx_get_free_slot_index(link, &idx) != -EBUSY),
 			SMC_WR_TX_WAIT_FREE_SLOT_TIME);
 		if (!rc) {
 			/* timeout - terminate connections */
-			struct smc_link_group *lgr;
-
-			lgr = container_of(link, struct smc_link_group,
-					   lnk[SMC_SINGLE_LINK]);
 			smc_lgr_terminate(lgr);
 			return -EPIPE;
 		}