iscsi/iser-target: Fix isert_conn->state hung shutdown issues

This patch addresses a couple of different hug shutdown issues
related to wait_event() + isert_conn->state.  First, it changes
isert_conn->conn_wait + isert_conn->conn_wait_comp_err from
waitqueues to completions, and sets ISER_CONN_TERMINATING from
within isert_disconnect_work().

Second, it splits isert_free_conn() into isert_wait_conn() that
is called earlier in iscsit_close_connection() to ensure that
all outstanding commands have completed before continuing.

Finally, it breaks isert_cq_comp_err() into seperate TX / RX
related code, and adds logic in isert_cq_rx_comp_err() to wait
for outstanding commands to complete before setting ISER_CONN_DOWN
and calling complete(&isert_conn->conn_wait_comp_err).

Acked-by: Sagi Grimberg <sagig@mellanox.com>
Cc: Or Gerlitz <ogerlitz@mellanox.com>
Cc: <stable@vger.kernel.org> #3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

4 files changed, 55 insertions, 60 deletions

diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c
index 5eb2b88cf947..862d7b4b0411 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.c
+++ b/drivers/infiniband/ulp/isert/ib_isert.c
@@ -492,8 +492,8 @@ isert_connect_request(struct rdma_cm_id *cma_id, struct rdma_cm_event *event)
         isert_conn->state = ISER_CONN_INIT;
         INIT_LIST_HEAD(&isert_conn->conn_accept_node);
         init_completion(&isert_conn->conn_login_comp);
-        init_waitqueue_head(&isert_conn->conn_wait);
-        init_waitqueue_head(&isert_conn->conn_wait_comp_err);
+        init_completion(&isert_conn->conn_wait);
+        init_completion(&isert_conn->conn_wait_comp_err);
         kref_init(&isert_conn->conn_kref);
         kref_get(&isert_conn->conn_kref);
         mutex_init(&isert_conn->conn_mutex);
@@ -688,11 +688,11 @@ isert_disconnect_work(struct work_struct *work)
 
         pr_debug("isert_disconnect_work(): >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n");
         mutex_lock(&isert_conn->conn_mutex);
-        isert_conn->state = ISER_CONN_DOWN;
+        if (isert_conn->state == ISER_CONN_UP)
+                isert_conn->state = ISER_CONN_TERMINATING;
 
         if (isert_conn->post_recv_buf_count == 0 &&
             atomic_read(&isert_conn->post_send_buf_count) == 0) {
-                pr_debug("Calling wake_up(&isert_conn->conn_wait);\n");
                 mutex_unlock(&isert_conn->conn_mutex);
                 goto wake_up;
         }
@@ -712,7 +712,7 @@ isert_disconnect_work(struct work_struct *work)
         mutex_unlock(&isert_conn->conn_mutex);
 
 wake_up:
-        wake_up(&isert_conn->conn_wait);
+        complete(&isert_conn->conn_wait);
         isert_put_conn(isert_conn);
 }
 
@@ -1589,7 +1589,7 @@ isert_do_control_comp(struct work_struct *work)
                 pr_debug("Calling iscsit_logout_post_handler >>>>>>>>>>>>>>\n");
                 /*
                  * Call atomic_dec(&isert_conn->post_send_buf_count)
-                 * from isert_free_conn()
+                 * from isert_wait_conn()
                  */
                 isert_conn->logout_posted = true;
                 iscsit_logout_post_handler(cmd, cmd->conn);
@@ -1691,31 +1691,39 @@ isert_send_completion(struct iser_tx_desc *tx_desc,
 }
 
 static void
-isert_cq_comp_err(struct iser_tx_desc *tx_desc, struct isert_conn *isert_conn)
+isert_cq_tx_comp_err(struct iser_tx_desc *tx_desc, struct isert_conn *isert_conn)
 {
         struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
+        struct isert_cmd *isert_cmd = tx_desc->isert_cmd;
+
+        if (!isert_cmd)
+                isert_unmap_tx_desc(tx_desc, ib_dev);
+        else
+                isert_completion_put(tx_desc, isert_cmd, ib_dev);
+}
+
+static void
+isert_cq_rx_comp_err(struct isert_conn *isert_conn)
+{
+        struct ib_device *ib_dev = isert_conn->conn_cm_id->device;
+        struct iscsi_conn *conn = isert_conn->conn;
 
-        if (tx_desc) {
-                struct isert_cmd *isert_cmd = tx_desc->isert_cmd;
+        if (isert_conn->post_recv_buf_count)
+                return;
 
-                if (!isert_cmd)
-                        isert_unmap_tx_desc(tx_desc, ib_dev);
-                else
-                        isert_completion_put(tx_desc, isert_cmd, ib_dev);
+        if (conn->sess) {
+                target_sess_cmd_list_set_waiting(conn->sess->se_sess);
+                target_wait_for_sess_cmds(conn->sess->se_sess);
         }
 
-        if (isert_conn->post_recv_buf_count == 0 &&
-            atomic_read(&isert_conn->post_send_buf_count) == 0) {
-                pr_debug("isert_cq_comp_err >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n");
-                pr_debug("Calling wake_up from isert_cq_comp_err\n");
+        while (atomic_read(&isert_conn->post_send_buf_count))
+                msleep(3000);
 
-                mutex_lock(&isert_conn->conn_mutex);
-                if (isert_conn->state != ISER_CONN_DOWN)
-                        isert_conn->state = ISER_CONN_TERMINATING;
-                mutex_unlock(&isert_conn->conn_mutex);
+        mutex_lock(&isert_conn->conn_mutex);
+        isert_conn->state = ISER_CONN_DOWN;
+        mutex_unlock(&isert_conn->conn_mutex);
 
-                wake_up(&isert_conn->conn_wait_comp_err);
-        }
+        complete(&isert_conn->conn_wait_comp_err);
 }
 
 static void
@@ -1740,8 +1748,9 @@ isert_cq_tx_work(struct work_struct *work)
                         pr_debug("TX wc.status != IB_WC_SUCCESS >>>>>>>>>>>>>>\n");
                         pr_debug("TX wc.status: 0x%08x\n", wc.status);
                         pr_debug("TX wc.vendor_err: 0x%08x\n", wc.vendor_err);
+
                         atomic_dec(&isert_conn->post_send_buf_count);
-                        isert_cq_comp_err(tx_desc, isert_conn);
+                        isert_cq_tx_comp_err(tx_desc, isert_conn);
                 }
         }
 
@@ -1784,7 +1793,7 @@ isert_cq_rx_work(struct work_struct *work)
                                          wc.vendor_err);
                         }
                         isert_conn->post_recv_buf_count--;
-                        isert_cq_comp_err(NULL, isert_conn);
+                        isert_cq_rx_comp_err(isert_conn);
                 }
         }
 
@@ -2702,22 +2711,11 @@ isert_free_np(struct iscsi_np *np)
         kfree(isert_np);
 }
 
-static int isert_check_state(struct isert_conn *isert_conn, int state)
-{
-        int ret;
-
-        mutex_lock(&isert_conn->conn_mutex);
-        ret = (isert_conn->state == state);
-        mutex_unlock(&isert_conn->conn_mutex);
-
-        return ret;
-}
-
-static void isert_free_conn(struct iscsi_conn *conn)
+static void isert_wait_conn(struct iscsi_conn *conn)
 {
         struct isert_conn *isert_conn = conn->context;
 
-        pr_debug("isert_free_conn: Starting \n");
+        pr_debug("isert_wait_conn: Starting \n");
         /*
          * Decrement post_send_buf_count for special case when called
          * from isert_do_control_comp() -> iscsit_logout_post_handler()
@@ -2727,38 +2725,29 @@ static void isert_free_conn(struct iscsi_conn *conn)
                 atomic_dec(&isert_conn->post_send_buf_count);
 
         if (isert_conn->conn_cm_id && isert_conn->state != ISER_CONN_DOWN) {
-                pr_debug("Calling rdma_disconnect from isert_free_conn\n");
+                pr_debug("Calling rdma_disconnect from isert_wait_conn\n");
                 rdma_disconnect(isert_conn->conn_cm_id);
         }
         /*
          * Only wait for conn_wait_comp_err if the isert_conn made it
          * into full feature phase..
          */
-        if (isert_conn->state == ISER_CONN_UP) {
-                pr_debug("isert_free_conn: Before wait_event comp_err %d\n",
-                         isert_conn->state);
-                mutex_unlock(&isert_conn->conn_mutex);
-
-                wait_event(isert_conn->conn_wait_comp_err,
-                          (isert_check_state(isert_conn, ISER_CONN_TERMINATING)));
-
-                wait_event(isert_conn->conn_wait,
-                          (isert_check_state(isert_conn, ISER_CONN_DOWN)));
-
-                isert_put_conn(isert_conn);
-                return;
-        }
         if (isert_conn->state == ISER_CONN_INIT) {
                 mutex_unlock(&isert_conn->conn_mutex);
-                isert_put_conn(isert_conn);
                 return;
         }
-        pr_debug("isert_free_conn: wait_event conn_wait %d\n",
-                 isert_conn->state);
+        if (isert_conn->state == ISER_CONN_UP)
+                isert_conn->state = ISER_CONN_TERMINATING;
         mutex_unlock(&isert_conn->conn_mutex);
 
-        wait_event(isert_conn->conn_wait,
-                  (isert_check_state(isert_conn, ISER_CONN_DOWN)));
+        wait_for_completion(&isert_conn->conn_wait_comp_err);
+
+        wait_for_completion(&isert_conn->conn_wait);
+}
+
+static void isert_free_conn(struct iscsi_conn *conn)
+{
+        struct isert_conn *isert_conn = conn->context;
 
         isert_put_conn(isert_conn);
 }
@@ -2771,6 +2760,7 @@ static struct iscsit_transport iser_target_transport = {
         .iscsit_setup_np        = isert_setup_np,
         .iscsit_accept_np       = isert_accept_np,
         .iscsit_free_np         = isert_free_np,
+        .iscsit_wait_conn       = isert_wait_conn,
         .iscsit_free_conn       = isert_free_conn,
         .iscsit_get_login_rx    = isert_get_login_rx,
         .iscsit_put_login_tx    = isert_put_login_tx,
diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h
index 708a069002f3..41e799ff8ecc 100644
--- a/drivers/infiniband/ulp/isert/ib_isert.h
+++ b/drivers/infiniband/ulp/isert/ib_isert.h
@@ -116,8 +116,8 @@ struct isert_conn {
         struct isert_device     *conn_device;
         struct work_struct      conn_logout_work;
         struct mutex            conn_mutex;
-        wait_queue_head_t       conn_wait;
-        wait_queue_head_t       conn_wait_comp_err;
+        struct completion       conn_wait;
+        struct completion       conn_wait_comp_err;
         struct kref             conn_kref;
         struct list_head        conn_fr_pool;
         int                     conn_fr_pool_size;
diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c
index ece82b0d8ea3..b83ec378d04f 100644
--- a/drivers/target/iscsi/iscsi_target.c
+++ b/drivers/target/iscsi/iscsi_target.c
@@ -4196,6 +4196,10 @@ int iscsit_close_connection(
         iscsit_stop_timers_for_cmds(conn);
         iscsit_stop_nopin_response_timer(conn);
         iscsit_stop_nopin_timer(conn);
+
+        if (conn->conn_transport->iscsit_wait_conn)
+                conn->conn_transport->iscsit_wait_conn(conn);
+
         iscsit_free_queue_reqs_for_conn(conn);
 
         /*
diff --git a/include/target/iscsi/iscsi_transport.h b/include/target/iscsi/iscsi_transport.h
index ae5a17111968..4483fadfa68d 100644
--- a/include/target/iscsi/iscsi_transport.h
+++ b/include/target/iscsi/iscsi_transport.h
@@ -12,6 +12,7 @@ struct iscsit_transport {
         int (*iscsit_setup_np)(struct iscsi_np *, struct __kernel_sockaddr_storage *);
         int (*iscsit_accept_np)(struct iscsi_np *, struct iscsi_conn *);
         void (*iscsit_free_np)(struct iscsi_np *);
+        void (*iscsit_wait_conn)(struct iscsi_conn *);
         void (*iscsit_free_conn)(struct iscsi_conn *);
         int (*iscsit_get_login_rx)(struct iscsi_conn *, struct iscsi_login *);
         int (*iscsit_put_login_tx)(struct iscsi_conn *, struct iscsi_login *, u32);