[SCSI] qla4xxx: set device state as per Link UP and LINK DOWN

Link Down -> Mark all devices missing
 Previously, the driver took no action on a Link Down,
 and waited for the I/O on a dead connection to timeout
 in the firmware before marking the DDB missing.

Link Up -> Mark all devices online
 F/W will do auto login to all the devices only once.
 After that its the responsibility of the driver to
 relogin to devices whenever there is :
        * Any sort of connection failure or
        * KATO expires indicating target has logged out or
        * I/O times out etc.

Signed-off-by: Vikas Chaudhary <vikas.chaudhary@qlogic.com>
Signed-off-by: Karen Higgins <karen.higgins@qlogic.com>
Signed-off-by: Ravi Anand <ravi.anand@qlogic.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>

4 files changed, 70 insertions, 7 deletions

diff --git a/drivers/scsi/qla4xxx/ql4_def.h b/drivers/scsi/qla4xxx/ql4_def.h
index 3175709d486..347a7c8b53e 100644
--- a/drivers/scsi/qla4xxx/ql4_def.h
+++ b/drivers/scsi/qla4xxx/ql4_def.h
@@ -310,6 +310,7 @@ struct scsi_qla_host {
 #define DPC_ISNS_RESTART                7 /* 0x00000080 */
 #define DPC_AEN                         9 /* 0x00000200 */
 #define DPC_GET_DHCP_IP_ADDR            15 /* 0x00008000 */
+#define DPC_LINK_CHANGED                18 /* 0x00040000 */
 
         struct Scsi_Host *host; /* pointer to host data */
         uint32_t tot_ddbs;
diff --git a/drivers/scsi/qla4xxx/ql4_init.c b/drivers/scsi/qla4xxx/ql4_init.c
index 36ec02c49a1..7f6cc2ebf46 100644
--- a/drivers/scsi/qla4xxx/ql4_init.c
+++ b/drivers/scsi/qla4xxx/ql4_init.c
@@ -1279,6 +1279,7 @@ int qla4xxx_initialize_adapter(struct scsi_qla_host *ha,
         int status = QLA_ERROR;
         int8_t ip_address[IP_ADDR_LEN] = {0} ;
 
+        clear_bit(AF_ONLINE, &ha->flags);
         ha->eeprom_cmd_data = 0;
 
         qla4x00_pci_config(ha);
@@ -1456,10 +1457,15 @@ int qla4xxx_process_ddb_changed(struct scsi_qla_host *ha,
                  * the device came back.
                  */
         } else {
-                /* Device went away, try to relogin. */
-                /* Mark device missing */
-                if (atomic_read(&ddb_entry->state) == DDB_STATE_ONLINE)
+                /* Device went away, mark device missing */
+                if (atomic_read(&ddb_entry->state) == DDB_STATE_ONLINE) {
+                        DEBUG2(dev_info(&ha->pdev->dev, "%s mark missing "
+                                        "ddb_entry 0x%p sess 0x%p conn 0x%p\n",
+                                        __func__, ddb_entry,
+                                        ddb_entry->sess, ddb_entry->conn));
                         qla4xxx_mark_device_missing(ha, ddb_entry);
+                }
+
                 /*
                  * Relogin if device state changed to a not active state.
                  * However, do not relogin if this aen is a result of an IOCTL
diff --git a/drivers/scsi/qla4xxx/ql4_isr.c b/drivers/scsi/qla4xxx/ql4_isr.c
index c196d55eae3..9db286df7ca 100644
--- a/drivers/scsi/qla4xxx/ql4_isr.c
+++ b/drivers/scsi/qla4xxx/ql4_isr.c
@@ -498,15 +498,22 @@ static void qla4xxx_isr_decode_mailbox(struct scsi_qla_host * ha,
                         break;
 
                 case MBOX_ASTS_LINK_UP:
-                        DEBUG2(printk("scsi%ld: AEN %04x Adapter LINK UP\n",
-                                      ha->host_no, mbox_status));
                         set_bit(AF_LINK_UP, &ha->flags);
+                        if (test_bit(AF_INIT_DONE, &ha->flags))
+                                set_bit(DPC_LINK_CHANGED, &ha->dpc_flags);
+
+                        DEBUG2(printk(KERN_INFO "scsi%ld: AEN %04x Adapter"
+                                        " LINK UP\n", ha->host_no,
+                                        mbox_status));
                         break;
 
                 case MBOX_ASTS_LINK_DOWN:
-                        DEBUG2(printk("scsi%ld: AEN %04x Adapter LINK DOWN\n",
-                                      ha->host_no, mbox_status));
                         clear_bit(AF_LINK_UP, &ha->flags);
+                        set_bit(DPC_LINK_CHANGED, &ha->dpc_flags);
+
+                        DEBUG2(printk(KERN_INFO "scsi%ld: AEN %04x Adapter"
+                                        " LINK DOWN\n", ha->host_no,
+                                        mbox_status));
                         break;
 
                 case MBOX_ASTS_HEARTBEAT:
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index 2ccad36bee9..d6c8b429a67 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -685,6 +685,7 @@ static void qla4xxx_timer(struct scsi_qla_host *ha)
              test_bit(DPC_RESET_HA_DESTROY_DDB_LIST, &ha->dpc_flags) ||
              test_bit(DPC_RESET_HA_INTR, &ha->dpc_flags) ||
              test_bit(DPC_GET_DHCP_IP_ADDR, &ha->dpc_flags) ||
+             test_bit(DPC_LINK_CHANGED, &ha->dpc_flags) ||
              test_bit(DPC_AEN, &ha->dpc_flags)) &&
              ha->dpc_thread) {
                 DEBUG2(printk("scsi%ld: %s: scheduling dpc routine"
@@ -1069,6 +1070,54 @@ static void qla4xxx_do_dpc(struct work_struct *work)
         if (test_and_clear_bit(DPC_GET_DHCP_IP_ADDR, &ha->dpc_flags))
                 qla4xxx_get_dhcp_ip_address(ha);
 
+        /* ---- link change? --- */
+        if (test_and_clear_bit(DPC_LINK_CHANGED, &ha->dpc_flags)) {
+                if (!test_bit(AF_LINK_UP, &ha->flags)) {
+                        /* ---- link down? --- */
+                        list_for_each_entry_safe(ddb_entry, dtemp,
+                                                 &ha->ddb_list, list) {
+                                if (atomic_read(&ddb_entry->state) ==
+                                                DDB_STATE_ONLINE)
+                                        qla4xxx_mark_device_missing(ha,
+                                                        ddb_entry);
+                        }
+                } else {
+                        /* ---- link up? --- *
+                         * F/W will auto login to all devices ONLY ONCE after
+                         * link up during driver initialization and runtime
+                         * fatal error recovery.  Therefore, the driver must
+                         * manually relogin to devices when recovering from
+                         * connection failures, logouts, expired KATO, etc. */
+
+                        list_for_each_entry_safe(ddb_entry, dtemp,
+                                                        &ha->ddb_list, list) {
+                                if ((atomic_read(&ddb_entry->state) ==
+                                                 DDB_STATE_MISSING) ||
+                                    (atomic_read(&ddb_entry->state) ==
+                                                 DDB_STATE_DEAD)) {
+                                        if (ddb_entry->fw_ddb_device_state ==
+                                            DDB_DS_SESSION_ACTIVE) {
+                                                atomic_set(&ddb_entry->state,
+                                                           DDB_STATE_ONLINE);
+                                                dev_info(&ha->pdev->dev,
+                                                    "scsi%ld: %s: ddb[%d]"
+                                                    " os[%d] marked"
+                                                    " ONLINE\n",
+                                                    ha->host_no, __func__,
+                                                    ddb_entry->fw_ddb_index,
+                                                    ddb_entry->os_target_id);
+
+                                                iscsi_unblock_session(
+                                                    ddb_entry->sess);
+                                        } else
+                                                qla4xxx_relogin_device(
+                                                    ha, ddb_entry);
+                                }
+
+                        }
+                }
+        }
+
         /* ---- relogin device? --- */
         if (adapter_up(ha) &&
             test_and_clear_bit(DPC_RELOGIN_DEVICE, &ha->dpc_flags)) {