Merge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (39 commits)
  [SCSI] sd: fix compile failure with CONFIG_BLK_DEV_INTEGRITY=n
  libiscsi: fix locking in iscsi_eh_device_reset
  libiscsi: check reason why we are stopping iscsi session to determine error value
  [SCSI] iscsi_tcp: return a descriptive error value during connection errors
  [SCSI] libiscsi: rename host reset to target reset
  [SCSI] iscsi class: fix endpoint id handling
  [SCSI] libiscsi: Support drivers initiating session removal
  [SCSI] libiscsi: fix data corruption when target has to resend data-in packets
  [SCSI] sd: Switch kernel printing level for DIF messages
  [SCSI] sd: Correctly handle all combinations of DIF and DIX
  [SCSI] sd: Always print actual protection_type
  [SCSI] sd: Issue correct protection operation
  [SCSI] scsi_error: fix target reset handling
  [SCSI] lpfc 8.2.8 v2 : Add statistical reporting control and additional fc vendor events
  [SCSI] lpfc 8.2.8 v2 : Add sysfs control of target queue depth handling
  [SCSI] lpfc 8.2.8 v2 : Revert target busy in favor of transport disrupted
  [SCSI] scsi_dh_alua: remove REQ_NOMERGE
  [SCSI] lpfc 8.2.8 : update driver version to 8.2.8
  [SCSI] lpfc 8.2.8 : Add MSI-X support
  [SCSI] lpfc 8.2.8 : Update driver to use new Host byte error code DID_TRANSPORT_DISRUPTED
  ...

block/blk-core.c

@@ -1075,8 +1075,15 @@ void init_request_from_bio(struct request *req, struct bio *bio)
 	/*
 	 * inherit FAILFAST from bio (for read-ahead, and explicit FAILFAST)
 	 */
-	if (bio_rw_ahead(bio) || bio_failfast(bio))
-		req->cmd_flags |= REQ_FAILFAST;
+	if (bio_rw_ahead(bio))
+		req->cmd_flags |= (REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+				   REQ_FAILFAST_DRIVER);
+	if (bio_failfast_dev(bio))
+		req->cmd_flags |= REQ_FAILFAST_DEV;
+	if (bio_failfast_transport(bio))
+		req->cmd_flags |= REQ_FAILFAST_TRANSPORT;
+	if (bio_failfast_driver(bio))
+		req->cmd_flags |= REQ_FAILFAST_DRIVER;
 
 	/*
 	 * REQ_BARRIER implies no merging, but lets make it explicit

drivers/infiniband/ulp/iser/iscsi_iser.c

@@ -378,6 +378,7 @@ static void iscsi_iser_session_destroy(struct iscsi_cls_session *cls_session)
 {
 	struct Scsi_Host *shost = iscsi_session_to_shost(cls_session);
 
+	iscsi_session_teardown(cls_session);
 	iscsi_host_remove(shost);
 	iscsi_host_free(shost);
 }
@@ -597,7 +598,7 @@ static struct scsi_host_template iscsi_iser_sht = {
 	.cmd_per_lun            = ISCSI_MAX_CMD_PER_LUN,
 	.eh_abort_handler       = iscsi_eh_abort,
 	.eh_device_reset_handler= iscsi_eh_device_reset,
-	.eh_host_reset_handler	= iscsi_eh_host_reset,
+	.eh_target_reset_handler= iscsi_eh_target_reset,
 	.use_clustering         = DISABLE_CLUSTERING,
 	.proc_name              = "iscsi_iser",
 	.this_id                = -1,

drivers/md/dm-mpath.c

@@ -849,7 +849,7 @@ static int multipath_map(struct dm_target *ti, struct bio *bio,
 	dm_bio_record(&mpio->details, bio);
 
 	map_context->ptr = mpio;
-	bio->bi_rw |= (1 << BIO_RW_FAILFAST);
+	bio->bi_rw |= (1 << BIO_RW_FAILFAST_TRANSPORT);
 	r = map_io(m, bio, mpio, 0);
 	if (r < 0 || r == DM_MAPIO_REQUEUE)
 		mempool_free(mpio, m->mpio_pool);

drivers/md/multipath.c

@@ -167,7 +167,7 @@ static int multipath_make_request (struct request_queue *q, struct bio * bio)
 	mp_bh->bio = *bio;
 	mp_bh->bio.bi_sector += multipath->rdev->data_offset;
 	mp_bh->bio.bi_bdev = multipath->rdev->bdev;
-	mp_bh->bio.bi_rw |= (1 << BIO_RW_FAILFAST);
+	mp_bh->bio.bi_rw |= (1 << BIO_RW_FAILFAST_TRANSPORT);
 	mp_bh->bio.bi_end_io = multipath_end_request;
 	mp_bh->bio.bi_private = mp_bh;
 	generic_make_request(&mp_bh->bio);
@@ -393,7 +393,7 @@ static void multipathd (mddev_t *mddev)
 			*bio = *(mp_bh->master_bio);
 			bio->bi_sector += conf->multipaths[mp_bh->path].rdev->data_offset;
 			bio->bi_bdev = conf->multipaths[mp_bh->path].rdev->bdev;
-			bio->bi_rw |= (1 << BIO_RW_FAILFAST);
+			bio->bi_rw |= (1 << BIO_RW_FAILFAST_TRANSPORT);
 			bio->bi_end_io = multipath_end_request;
 			bio->bi_private = mp_bh;
 			generic_make_request(bio);

drivers/s390/block/dasd_diag.c

@@ -544,7 +544,7 @@ static struct dasd_ccw_req *dasd_diag_build_cp(struct dasd_device *memdev,
 	}
 	cqr->retries = DIAG_MAX_RETRIES;
 	cqr->buildclk = get_clock();
-	if (req->cmd_flags & REQ_FAILFAST)
+	if (blk_noretry_request(req))
 		set_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags);
 	cqr->startdev = memdev;
 	cqr->memdev = memdev;

drivers/s390/block/dasd_eckd.c

@@ -1700,7 +1700,7 @@ static struct dasd_ccw_req *dasd_eckd_build_cp(struct dasd_device *startdev,
 			recid++;
 		}
 	}
-	if (req->cmd_flags & REQ_FAILFAST)
+	if (blk_noretry_request(req))
 		set_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags);
 	cqr->startdev = startdev;
 	cqr->memdev = startdev;

drivers/s390/block/dasd_fba.c

@@ -355,7 +355,7 @@ static struct dasd_ccw_req *dasd_fba_build_cp(struct dasd_device * memdev,
 			recid++;
 		}
 	}
-	if (req->cmd_flags & REQ_FAILFAST)
+	if (blk_noretry_request(req))
 		set_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags);
 	cqr->startdev = memdev;
 	cqr->memdev = memdev;

drivers/scsi/constants.c

@@ -1364,7 +1364,8 @@ EXPORT_SYMBOL(scsi_print_sense);
 static const char * const hostbyte_table[]={
 "DID_OK", "DID_NO_CONNECT", "DID_BUS_BUSY", "DID_TIME_OUT", "DID_BAD_TARGET",
 "DID_ABORT", "DID_PARITY", "DID_ERROR", "DID_RESET", "DID_BAD_INTR",
-"DID_PASSTHROUGH", "DID_SOFT_ERROR", "DID_IMM_RETRY", "DID_REQUEUE"};
+"DID_PASSTHROUGH", "DID_SOFT_ERROR", "DID_IMM_RETRY", "DID_REQUEUE",
+"DID_TRANSPORT_DISRUPTED", "DID_TRANSPORT_FAILFAST" };
 #define NUM_HOSTBYTE_STRS ARRAY_SIZE(hostbyte_table)
 
 static const char * const driverbyte_table[]={

drivers/scsi/device_handler/scsi_dh_alua.c

@@ -109,7 +109,8 @@ static struct request *get_alua_req(struct scsi_device *sdev,
 	}
 
 	rq->cmd_type = REQ_TYPE_BLOCK_PC;
-	rq->cmd_flags |= REQ_FAILFAST | REQ_NOMERGE;
+	rq->cmd_flags |= REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+			 REQ_FAILFAST_DRIVER;
 	rq->retries = ALUA_FAILOVER_RETRIES;
 	rq->timeout = ALUA_FAILOVER_TIMEOUT;
 

drivers/scsi/device_handler/scsi_dh_emc.c

@@ -303,7 +303,8 @@ static struct request *get_req(struct scsi_device *sdev, int cmd,
 
 	rq->cmd[4] = len;
 	rq->cmd_type = REQ_TYPE_BLOCK_PC;
-	rq->cmd_flags |= REQ_FAILFAST;
+	rq->cmd_flags |= REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+			 REQ_FAILFAST_DRIVER;
 	rq->timeout = CLARIION_TIMEOUT;
 	rq->retries = CLARIION_RETRIES;
 

drivers/scsi/device_handler/scsi_dh_hp_sw.c

@@ -112,7 +112,8 @@ static int hp_sw_tur(struct scsi_device *sdev, struct hp_sw_dh_data *h)
 		return SCSI_DH_RES_TEMP_UNAVAIL;
 
 	req->cmd_type = REQ_TYPE_BLOCK_PC;
-	req->cmd_flags |= REQ_FAILFAST;
+	req->cmd_flags |= REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+			  REQ_FAILFAST_DRIVER;
 	req->cmd_len = COMMAND_SIZE(TEST_UNIT_READY);
 	req->cmd[0] = TEST_UNIT_READY;
 	req->timeout = HP_SW_TIMEOUT;
@@ -204,7 +205,8 @@ static int hp_sw_start_stop(struct scsi_device *sdev, struct hp_sw_dh_data *h)
 		return SCSI_DH_RES_TEMP_UNAVAIL;
 
 	req->cmd_type = REQ_TYPE_BLOCK_PC;
-	req->cmd_flags |= REQ_FAILFAST;
+	req->cmd_flags |= REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+			  REQ_FAILFAST_DRIVER;
 	req->cmd_len = COMMAND_SIZE(START_STOP);
 	req->cmd[0] = START_STOP;
 	req->cmd[4] = 1;	/* Start spin cycle */

drivers/scsi/device_handler/scsi_dh_rdac.c

@@ -226,7 +226,8 @@ static struct request *get_rdac_req(struct scsi_device *sdev,
 	}
 
 	rq->cmd_type = REQ_TYPE_BLOCK_PC;
-	rq->cmd_flags |= REQ_FAILFAST | REQ_NOMERGE;
+	rq->cmd_flags |= REQ_FAILFAST_DEV | REQ_FAILFAST_TRANSPORT |
+			 REQ_FAILFAST_DRIVER;
 	rq->retries = RDAC_RETRIES;
 	rq->timeout = RDAC_TIMEOUT;
 

drivers/scsi/ibmvscsi/ibmvfc.c

@@ -2031,8 +2031,6 @@ static void ibmvfc_terminate_rport_io(struct fc_rport *rport)
 		spin_unlock_irqrestore(shost->host_lock, flags);
 	} else
 		ibmvfc_issue_fc_host_lip(shost);
-
-	scsi_target_unblock(&rport->dev);
 	LEAVE;
 }
 

drivers/scsi/iscsi_tcp.c

@@ -523,22 +523,20 @@ iscsi_tcp_cleanup_task(struct iscsi_conn *conn, struct iscsi_task *task)
 }
 
 /**
- * iscsi_data_rsp - SCSI Data-In Response processing
+ * iscsi_data_in - SCSI Data-In Response processing
  * @conn: iscsi connection
  * @task: scsi command task
  **/
 static int
-iscsi_data_rsp(struct iscsi_conn *conn, struct iscsi_task *task)
+iscsi_data_in(struct iscsi_conn *conn, struct iscsi_task *task)
 {
 	struct iscsi_tcp_conn *tcp_conn = conn->dd_data;
 	struct iscsi_tcp_task *tcp_task = task->dd_data;
 	struct iscsi_data_rsp *rhdr = (struct iscsi_data_rsp *)tcp_conn->in.hdr;
-	struct iscsi_session *session = conn->session;
-	struct scsi_cmnd *sc = task->sc;
 	int datasn = be32_to_cpu(rhdr->datasn);
-	unsigned total_in_length = scsi_in(sc)->length;
+	unsigned total_in_length = scsi_in(task->sc)->length;
 
-	iscsi_update_cmdsn(session, (struct iscsi_nopin*)rhdr);
+	iscsi_update_cmdsn(conn->session, (struct iscsi_nopin*)rhdr);
 	if (tcp_conn->in.datalen == 0)
 		return 0;
 
@@ -558,23 +556,6 @@ iscsi_data_rsp(struct iscsi_conn *conn, struct iscsi_task *task)
 		return ISCSI_ERR_DATA_OFFSET;
 	}
 
-	if (rhdr->flags & ISCSI_FLAG_DATA_STATUS) {
-		sc->result = (DID_OK << 16) | rhdr->cmd_status;
-		conn->exp_statsn = be32_to_cpu(rhdr->statsn) + 1;
-		if (rhdr->flags & (ISCSI_FLAG_DATA_UNDERFLOW |
-		                   ISCSI_FLAG_DATA_OVERFLOW)) {
-			int res_count = be32_to_cpu(rhdr->residual_count);
-
-			if (res_count > 0 &&
-			    (rhdr->flags & ISCSI_FLAG_CMD_OVERFLOW ||
-			     res_count <= total_in_length))
-				scsi_in(sc)->resid = res_count;
-			else
-				sc->result = (DID_BAD_TARGET << 16) |
-					rhdr->cmd_status;
-		}
-	}
-
 	conn->datain_pdus_cnt++;
 	return 0;
 }
@@ -774,7 +755,7 @@ iscsi_tcp_hdr_dissect(struct iscsi_conn *conn, struct iscsi_hdr *hdr)
 		if (!task)
 			rc = ISCSI_ERR_BAD_ITT;
 		else
-			rc = iscsi_data_rsp(conn, task);
+			rc = iscsi_data_in(conn, task);
 		if (rc) {
 			spin_unlock(&conn->session->lock);
 			break;
@@ -998,7 +979,7 @@ iscsi_tcp_recv(read_descriptor_t *rd_desc, struct sk_buff *skb,
 
 error:
 	debug_tcp("Error receiving PDU, errno=%d\n", rc);
-	iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
+	iscsi_conn_failure(conn, rc);
 	return 0;
 }
 
@@ -1117,8 +1098,10 @@ iscsi_xmit(struct iscsi_conn *conn)
 
 	while (1) {
 		rc = iscsi_tcp_xmit_segment(tcp_conn, segment);
-		if (rc < 0)
+		if (rc < 0) {
+			rc = ISCSI_ERR_XMIT_FAILED;
 			goto error;
+		}
 		if (rc == 0)
 			break;
 
@@ -1127,7 +1110,7 @@ iscsi_xmit(struct iscsi_conn *conn)
 		if (segment->total_copied >= segment->total_size) {
 			if (segment->done != NULL) {
 				rc = segment->done(tcp_conn, segment);
-				if (rc < 0)
+				if (rc != 0)
 					goto error;
 			}
 		}
@@ -1142,8 +1125,8 @@ error:
 	/* Transmit error. We could initiate error recovery
 	 * here. */
 	debug_tcp("Error sending PDU, errno=%d\n", rc);
-	iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
-	return rc;
+	iscsi_conn_failure(conn, rc);
+	return -EIO;
 }
 
 /**
@@ -1904,6 +1887,7 @@ static void iscsi_tcp_session_destroy(struct iscsi_cls_session *cls_session)
 	struct Scsi_Host *shost = iscsi_session_to_shost(cls_session);
 
 	iscsi_r2tpool_free(cls_session->dd_data);
+	iscsi_session_teardown(cls_session);
 
 	iscsi_host_remove(shost);
 	iscsi_host_free(shost);
@@ -1927,7 +1911,7 @@ static struct scsi_host_template iscsi_sht = {
 	.cmd_per_lun		= ISCSI_DEF_CMD_PER_LUN,
 	.eh_abort_handler       = iscsi_eh_abort,
 	.eh_device_reset_handler= iscsi_eh_device_reset,
-	.eh_host_reset_handler	= iscsi_eh_host_reset,
+	.eh_target_reset_handler= iscsi_eh_target_reset,
 	.use_clustering         = DISABLE_CLUSTERING,
 	.slave_configure        = iscsi_tcp_slave_configure,
 	.proc_name		= "iscsi_tcp",

drivers/scsi/libiscsi.c

@@ -404,11 +404,6 @@ static void fail_command(struct iscsi_conn *conn, struct iscsi_task *task,
 		conn->session->queued_cmdsn--;
 	else
 		conn->session->tt->cleanup_task(conn, task);
-	/*
-	 * Check if cleanup_task dropped the lock and the command completed,
-	 */
-	if (!task->sc)
-		return;
 
 	sc->result = err;
 	if (!scsi_bidi_cmnd(sc))
@@ -633,6 +628,40 @@ out:
 	__iscsi_put_task(task);
 }
 
+/**
+ * iscsi_data_in_rsp - SCSI Data-In Response processing
+ * @conn: iscsi connection
+ * @hdr:  iscsi pdu
+ * @task: scsi command task
+ **/
+static void
+iscsi_data_in_rsp(struct iscsi_conn *conn, struct iscsi_hdr *hdr,
+		  struct iscsi_task *task)
+{
+	struct iscsi_data_rsp *rhdr = (struct iscsi_data_rsp *)hdr;
+	struct scsi_cmnd *sc = task->sc;
+
+	if (!(rhdr->flags & ISCSI_FLAG_DATA_STATUS))
+		return;
+
+	sc->result = (DID_OK << 16) | rhdr->cmd_status;
+	conn->exp_statsn = be32_to_cpu(rhdr->statsn) + 1;
+	if (rhdr->flags & (ISCSI_FLAG_DATA_UNDERFLOW |
+	                   ISCSI_FLAG_DATA_OVERFLOW)) {
+		int res_count = be32_to_cpu(rhdr->residual_count);
+
+		if (res_count > 0 &&
+		    (rhdr->flags & ISCSI_FLAG_CMD_OVERFLOW ||
+		     res_count <= scsi_in(sc)->length))
+			scsi_in(sc)->resid = res_count;
+		else
+			sc->result = (DID_BAD_TARGET << 16) | rhdr->cmd_status;
+	}
+
+	conn->scsirsp_pdus_cnt++;
+	__iscsi_put_task(task);
+}
+
 static void iscsi_tmf_rsp(struct iscsi_conn *conn, struct iscsi_hdr *hdr)
 {
 	struct iscsi_tm_rsp *tmf = (struct iscsi_tm_rsp *)hdr;
@@ -818,12 +847,7 @@ int __iscsi_complete_pdu(struct iscsi_conn *conn, struct iscsi_hdr *hdr,
 		iscsi_scsi_cmd_rsp(conn, hdr, task, data, datalen);
 		break;
 	case ISCSI_OP_SCSI_DATA_IN:
-		if (hdr->flags & ISCSI_FLAG_DATA_STATUS) {
-			conn->scsirsp_pdus_cnt++;
-			iscsi_update_cmdsn(session,
-					   (struct iscsi_nopin*) hdr);
-			__iscsi_put_task(task);
-		}
+		iscsi_data_in_rsp(conn, hdr, task);
 		break;
 	case ISCSI_OP_LOGOUT_RSP:
 		iscsi_update_cmdsn(session, (struct iscsi_nopin*)hdr);
@@ -954,6 +978,38 @@ struct iscsi_task *iscsi_itt_to_ctask(struct iscsi_conn *conn, itt_t itt)
 }
 EXPORT_SYMBOL_GPL(iscsi_itt_to_ctask);
 
+void iscsi_session_failure(struct iscsi_cls_session *cls_session,
+			   enum iscsi_err err)
+{
+	struct iscsi_session *session = cls_session->dd_data;
+	struct iscsi_conn *conn;
+	struct device *dev;
+	unsigned long flags;
+
+	spin_lock_irqsave(&session->lock, flags);
+	conn = session->leadconn;
+	if (session->state == ISCSI_STATE_TERMINATE || !conn) {
+		spin_unlock_irqrestore(&session->lock, flags);
+		return;
+	}
+
+	dev = get_device(&conn->cls_conn->dev);
+	spin_unlock_irqrestore(&session->lock, flags);
+	if (!dev)
+	        return;
+	/*
+	 * if the host is being removed bypass the connection
+	 * recovery initialization because we are going to kill
+	 * the session.
+	 */
+	if (err == ISCSI_ERR_INVALID_HOST)
+		iscsi_conn_error_event(conn->cls_conn, err);
+	else
+		iscsi_conn_failure(conn, err);
+	put_device(dev);
+}
+EXPORT_SYMBOL_GPL(iscsi_session_failure);
+
 void iscsi_conn_failure(struct iscsi_conn *conn, enum iscsi_err err)
 {
 	struct iscsi_session *session = conn->session;
@@ -968,9 +1024,10 @@ void iscsi_conn_failure(struct iscsi_conn *conn, enum iscsi_err err)
 	if (conn->stop_stage == 0)
 		session->state = ISCSI_STATE_FAILED;
 	spin_unlock_irqrestore(&session->lock, flags);
+
 	set_bit(ISCSI_SUSPEND_BIT, &conn->suspend_tx);
 	set_bit(ISCSI_SUSPEND_BIT, &conn->suspend_rx);
-	iscsi_conn_error(conn->cls_conn, err);
+	iscsi_conn_error_event(conn->cls_conn, err);
 }
 EXPORT_SYMBOL_GPL(iscsi_conn_failure);
 
@@ -1194,15 +1251,13 @@ int iscsi_queuecommand(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
 		switch (session->state) {
 		case ISCSI_STATE_IN_RECOVERY:
 			reason = FAILURE_SESSION_IN_RECOVERY;
-			sc->result = DID_IMM_RETRY << 16;
-			break;
+			goto reject;
 		case ISCSI_STATE_LOGGING_OUT:
 			reason = FAILURE_SESSION_LOGGING_OUT;
-			sc->result = DID_IMM_RETRY << 16;
-			break;
+			goto reject;
 		case ISCSI_STATE_RECOVERY_FAILED:
 			reason = FAILURE_SESSION_RECOVERY_TIMEOUT;
-			sc->result = DID_NO_CONNECT << 16;
+			sc->result = DID_TRANSPORT_FAILFAST << 16;
 			break;
 		case ISCSI_STATE_TERMINATE:
 			reason = FAILURE_SESSION_TERMINATE;
@@ -1267,7 +1322,7 @@ reject:
 	spin_unlock(&session->lock);
 	debug_scsi("cmd 0x%x rejected (%d)\n", sc->cmnd[0], reason);
 	spin_lock(host->host_lock);
-	return SCSI_MLQUEUE_HOST_BUSY;
+	return SCSI_MLQUEUE_TARGET_BUSY;
 
 fault:
 	spin_unlock(&session->lock);
@@ -1307,7 +1362,7 @@ void iscsi_session_recovery_timedout(struct iscsi_cls_session *cls_session)
 }
 EXPORT_SYMBOL_GPL(iscsi_session_recovery_timedout);
 
-int iscsi_eh_host_reset(struct scsi_cmnd *sc)
+int iscsi_eh_target_reset(struct scsi_cmnd *sc)
 {
 	struct iscsi_cls_session *cls_session;
 	struct iscsi_session *session;
@@ -1321,7 +1376,7 @@ int iscsi_eh_host_reset(struct scsi_cmnd *sc)
 	spin_lock_bh(&session->lock);
 	if (session->state == ISCSI_STATE_TERMINATE) {
 failed:
-		debug_scsi("failing host reset: session terminated "
+		debug_scsi("failing target reset: session terminated "
 			   "[CID %d age %d]\n", conn->id, session->age);
 		spin_unlock_bh(&session->lock);
 		mutex_unlock(&session->eh_mutex);
@@ -1336,7 +1391,7 @@ failed:
 	 */
 	iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED);
 
-	debug_scsi("iscsi_eh_host_reset wait for relogin\n");
+	debug_scsi("iscsi_eh_target_reset wait for relogin\n");
 	wait_event_interruptible(conn->ehwait,
 				 session->state == ISCSI_STATE_TERMINATE ||
 				 session->state == ISCSI_STATE_LOGGED_IN ||
@@ -1348,14 +1403,14 @@ failed:
 	spin_lock_bh(&session->lock);
 	if (session->state == ISCSI_STATE_LOGGED_IN)
 		iscsi_session_printk(KERN_INFO, session,
-				     "host reset succeeded\n");
+				     "target reset succeeded\n");
 	else
 		goto failed;
 	spin_unlock_bh(&session->lock);
 	mutex_unlock(&session->eh_mutex);
 	return SUCCESS;
 }
-EXPORT_SYMBOL_GPL(iscsi_eh_host_reset);
+EXPORT_SYMBOL_GPL(iscsi_eh_target_reset);
 
 static void iscsi_tmf_timedout(unsigned long data)
 {
@@ -1769,10 +1824,10 @@ int iscsi_eh_device_reset(struct scsi_cmnd *sc)
 
 	iscsi_suspend_tx(conn);
 
-	spin_lock(&session->lock);
+	spin_lock_bh(&session->lock);
 	fail_all_commands(conn, sc->device->lun, DID_ERROR);
 	conn->tmf_state = TMF_INITIAL;
-	spin_unlock(&session->lock);
+	spin_unlock_bh(&session->lock);
 
 	iscsi_start_tx(conn);
 	goto done;
@@ -1878,6 +1933,7 @@ struct Scsi_Host *iscsi_host_alloc(struct scsi_host_template *sht,
 				   int dd_data_size, uint16_t qdepth)
 {
 	struct Scsi_Host *shost;
+	struct iscsi_host *ihost;
 
 	shost = scsi_host_alloc(sht, sizeof(struct iscsi_host) + dd_data_size);
 	if (!shost)
@@ -1892,22 +1948,43 @@ struct Scsi_Host *iscsi_host_alloc(struct scsi_host_template *sht,
 		qdepth = ISCSI_DEF_CMD_PER_LUN;
 	}
 	shost->cmd_per_lun = qdepth;
+
+	ihost = shost_priv(shost);
+	spin_lock_init(&ihost->lock);
+	ihost->state = ISCSI_HOST_SETUP;
+	ihost->num_sessions = 0;
+	init_waitqueue_head(&ihost->session_removal_wq);
 	return shost;
 }
 EXPORT_SYMBOL_GPL(iscsi_host_alloc);
 
+static void iscsi_notify_host_removed(struct iscsi_cls_session *cls_session)
+{
+	iscsi_session_failure(cls_session, ISCSI_ERR_INVALID_HOST);
+}
+
 /**
  * iscsi_host_remove - remove host and sessions
  * @shost: scsi host
  *
- * This will also remove any sessions attached to the host, but if userspace
- * is managing the session at the same time this will break. TODO: add
- * refcounting to the netlink iscsi interface so a rmmod or host hot unplug
- * does not remove the memory from under us.
+ * If there are any sessions left, this will initiate the removal and wait
+ * for the completion.
  */
 void iscsi_host_remove(struct Scsi_Host *shost)
 {
-	iscsi_host_for_each_session(shost, iscsi_session_teardown);
+	struct iscsi_host *ihost = shost_priv(shost);
+	unsigned long flags;
+
+	spin_lock_irqsave(&ihost->lock, flags);
+	ihost->state = ISCSI_HOST_REMOVED;
+	spin_unlock_irqrestore(&ihost->lock, flags);
+
+	iscsi_host_for_each_session(shost, iscsi_notify_host_removed);
+	wait_event_interruptible(ihost->session_removal_wq,
+				 ihost->num_sessions == 0);
+	if (signal_pending(current))
+		flush_signals(current);
+
 	scsi_remove_host(shost);
 }
 EXPORT_SYMBOL_GPL(iscsi_host_remove);
@@ -1923,6 +2000,27 @@ void iscsi_host_free(struct Scsi_Host *shost)
 }
 EXPORT_SYMBOL_GPL(iscsi_host_free);
 
+static void iscsi_host_dec_session_cnt(struct Scsi_Host *shost)
+{
+	struct iscsi_host *ihost = shost_priv(shost);
+	unsigned long flags;
+
+	shost = scsi_host_get(shost);
+	if (!shost) {
+		printk(KERN_ERR "Invalid state. Cannot notify host removal "
+		      "of session teardown event because host already "
+		      "removed.\n");
+		return;
+	}
+
+	spin_lock_irqsave(&ihost->lock, flags);
+	ihost->num_sessions--;
+	if (ihost->num_sessions == 0)
+		wake_up(&ihost->session_removal_wq);
+	spin_unlock_irqrestore(&ihost->lock, flags);
+	scsi_host_put(shost);
+}
+
 /**
  * iscsi_session_setup - create iscsi cls session and host and session
  * @iscsit: iscsi transport template
@@ -1943,9 +2041,19 @@ iscsi_session_setup(struct iscsi_transport *iscsit, struct Scsi_Host *shost,
 		    uint16_t cmds_max, int cmd_task_size,
 		    uint32_t initial_cmdsn, unsigned int id)
 {
+	struct iscsi_host *ihost = shost_priv(shost);
 	struct iscsi_session *session;
 	struct iscsi_cls_session *cls_session;
 	int cmd_i, scsi_cmds, total_cmds = cmds_max;
+	unsigned long flags;
+
+	spin_lock_irqsave(&ihost->lock, flags);
+	if (ihost->state == ISCSI_HOST_REMOVED) {
+		spin_unlock_irqrestore(&ihost->lock, flags);
+		return NULL;
+	}
+	ihost->num_sessions++;
+	spin_unlock_irqrestore(&ihost->lock, flags);
 
 	if (!total_cmds)
 		total_cmds = ISCSI_DEF_XMIT_CMDS_MAX;
@@ -1958,7 +2066,7 @@ iscsi_session_setup(struct iscsi_transport *iscsit, struct Scsi_Host *shost,
 		printk(KERN_ERR "iscsi: invalid can_queue of %d. can_queue "
 		       "must be a power of two that is at least %d.\n",
 		       total_cmds, ISCSI_TOTAL_CMDS_MIN);
-		return NULL;
+		goto dec_session_count;
 	}
 
 	if (total_cmds > ISCSI_TOTAL_CMDS_MAX) {
@@ -1982,7 +2090,7 @@ iscsi_session_setup(struct iscsi_transport *iscsit, struct Scsi_Host *shost,
 	cls_session = iscsi_alloc_session(shost, iscsit,
 					  sizeof(struct iscsi_session));
 	if (!cls_session)
-		return NULL;
+		goto dec_session_count;
 	session = cls_session->dd_data;
 	session->cls_session = cls_session;
 	session->host = shost;
@@ -2021,6 +2129,7 @@ iscsi_session_setup(struct iscsi_transport *iscsit, struct Scsi_Host *shost,
 
 	if (iscsi_add_session(cls_session, id))
 		goto cls_session_fail;
+
 	return cls_session;
 
 cls_session_fail:
@@ -2029,6 +2138,8 @@ module_get_fail:
 	iscsi_pool_free(&session->cmdpool);
 cmdpool_alloc_fail:
 	iscsi_free_session(cls_session);
+dec_session_count:
+	iscsi_host_dec_session_cnt(shost);
 	return NULL;
 }
 EXPORT_SYMBOL_GPL(iscsi_session_setup);
@@ -2044,6 +2155,7 @@ void iscsi_session_teardown(struct iscsi_cls_session *cls_session)
 {
 	struct iscsi_session *session = cls_session->dd_data;
 	struct module *owner = cls_session->transport->owner;
+	struct Scsi_Host *shost = session->host;
 
 	iscsi_pool_free(&session->cmdpool);
 
@@ -2056,6 +2168,7 @@ void iscsi_session_teardown(struct iscsi_cls_session *cls_session)
 	kfree(session->ifacename);
 
 	iscsi_destroy_session(cls_session);
+	iscsi_host_dec_session_cnt(shost);
 	module_put(owner);
 }
 EXPORT_SYMBOL_GPL(iscsi_session_teardown);
@@ -2335,8 +2448,10 @@ static void iscsi_start_session_recovery(struct iscsi_session *session,
 	 * flush queues.
 	 */
 	spin_lock_bh(&session->lock);
-	fail_all_commands(conn, -1,
-			STOP_CONN_RECOVER ? DID_BUS_BUSY : DID_ERROR);
+	if (flag == STOP_CONN_RECOVER)
+		fail_all_commands(conn, -1, DID_TRANSPORT_DISRUPTED);
+	else
+		fail_all_commands(conn, -1, DID_ERROR);
 	flush_control_queues(session, conn);
 	spin_unlock_bh(&session->lock);
 	mutex_unlock(&session->eh_mutex);

drivers/scsi/lpfc/lpfc.h

@@ -34,7 +34,14 @@ struct lpfc_sli2_slim;
 #define LPFC_IOCB_LIST_CNT	2250	/* list of IOCBs for fast-path usage. */
 #define LPFC_Q_RAMP_UP_INTERVAL 120     /* lun q_depth ramp up interval */
 #define LPFC_VNAME_LEN		100	/* vport symbolic name length */
-
+#define LPFC_TGTQ_INTERVAL	40000	/* Min amount of time between tgt
+					   queue depth change in millisecs */
+#define LPFC_TGTQ_RAMPUP_PCENT	5	/* Target queue rampup in percentage */
+#define LPFC_MIN_TGT_QDEPTH	100
+#define LPFC_MAX_TGT_QDEPTH	0xFFFF
+
+#define  LPFC_MAX_BUCKET_COUNT 20	/* Maximum no. of buckets for stat data
+					   collection. */
 /*
  * Following time intervals are used of adjusting SCSI device
  * queue depths when there are driver resource error or Firmware
@@ -49,6 +56,9 @@ struct lpfc_sli2_slim;
 #define LPFC_HB_MBOX_INTERVAL   5	/* Heart beat interval in seconds. */
 #define LPFC_HB_MBOX_TIMEOUT    30	/* Heart beat timeout  in seconds. */
 
+/* Error Attention event polling interval */
+#define LPFC_ERATT_POLL_INTERVAL	5 /* EATT poll interval in seconds */
+
 /* Define macros for 64 bit support */
 #define putPaddrLow(addr)    ((uint32_t) (0xffffffff & (u64)(addr)))
 #define putPaddrHigh(addr)   ((uint32_t) (0xffffffff & (((u64)(addr))>>32)))
@@ -60,6 +70,9 @@ struct lpfc_sli2_slim;
 
 #define MAX_HBAEVT	32
 
+/* Number of MSI-X vectors the driver uses */
+#define LPFC_MSIX_VECTORS	2
+
 /* lpfc wait event data ready flag */
 #define LPFC_DATA_READY		(1<<0)
 
@@ -357,6 +370,7 @@ struct lpfc_vport {
 	uint32_t cfg_log_verbose;
 	uint32_t cfg_max_luns;
 	uint32_t cfg_enable_da_id;
+	uint32_t cfg_max_scsicmpl_time;
 
 	uint32_t dev_loss_tmo_changed;
 
@@ -369,6 +383,8 @@ struct lpfc_vport {
 	struct lpfc_debugfs_trc *disc_trc;
 	atomic_t disc_trc_cnt;
 #endif
+	uint8_t stat_data_enabled;
+	uint8_t stat_data_blocked;
 };
 
 struct hbq_s {
@@ -407,10 +423,11 @@ struct lpfc_hba {
 	struct lpfc_sli sli;
 	uint32_t sli_rev;		/* SLI2 or SLI3 */
 	uint32_t sli3_options;		/* Mask of enabled SLI3 options */
-#define LPFC_SLI3_ENABLED	 0x01
-#define LPFC_SLI3_HBQ_ENABLED	 0x02
-#define LPFC_SLI3_NPIV_ENABLED	 0x04
-#define LPFC_SLI3_VPORT_TEARDOWN 0x08
+#define LPFC_SLI3_HBQ_ENABLED		0x01
+#define LPFC_SLI3_NPIV_ENABLED		0x02
+#define LPFC_SLI3_VPORT_TEARDOWN	0x04
+#define LPFC_SLI3_CRP_ENABLED		0x08
+#define LPFC_SLI3_INB_ENABLED		0x10
 	uint32_t iocb_cmd_size;
 	uint32_t iocb_rsp_size;
 
@@ -422,10 +439,20 @@ struct lpfc_hba {
 #define LS_NPIV_FAB_SUPPORTED 0x2	/* Fabric supports NPIV */
 #define LS_IGNORE_ERATT       0x4	/* intr handler should ignore ERATT */
 
-	struct lpfc_sli2_slim *slim2p;
-	struct lpfc_dmabuf hbqslimp;
+	uint32_t hba_flag;	/* hba generic flags */
+#define HBA_ERATT_HANDLED	0x1 /* This flag is set when eratt handled */
+
+	struct lpfc_dmabuf slim2p;
 
-	dma_addr_t slim2p_mapping;
+	MAILBOX_t *mbox;
+	uint32_t *inb_ha_copy;
+	uint32_t *inb_counter;
+	uint32_t inb_last_counter;
+	uint32_t ha_copy;
+	struct _PCB *pcb;
+	struct _IOCB *IOCBs;
+
+	struct lpfc_dmabuf hbqslimp;
 
 	uint16_t pci_cfg_value;
 
@@ -492,7 +519,7 @@ struct lpfc_hba {
 
 	wait_queue_head_t    work_waitq;
 	struct task_struct   *worker_thread;
-	long data_flags;
+	unsigned long data_flags;
 
 	uint32_t hbq_in_use;		/* HBQs in use flag */
 	struct list_head hbqbuf_in_list;  /* in-fly hbq buffer list */
@@ -514,6 +541,7 @@ struct lpfc_hba {
 	void __iomem *HCregaddr;	/* virtual address for host ctl reg */
 
 	struct lpfc_hgp __iomem *host_gp; /* Host side get/put pointers */
+	struct lpfc_pgp   *port_gp;
 	uint32_t __iomem  *hbq_put;     /* Address in SLIM to HBQ put ptrs */
 	uint32_t          *hbq_get;     /* Host mem address of HBQ get ptrs */
 
@@ -536,6 +564,7 @@ struct lpfc_hba {
 	uint8_t soft_wwn_enable;
 
 	struct timer_list fcp_poll_timer;
+	struct timer_list eratt_poll;
 
 	/*
 	 * stat  counters
@@ -565,7 +594,7 @@ struct lpfc_hba {
 
 	struct fc_host_statistics link_stats;
 	enum intr_type_t intr_type;
-	struct msix_entry msix_entries[1];
+	struct msix_entry msix_entries[LPFC_MSIX_VECTORS];
 
 	struct list_head port_list;
 	struct lpfc_vport *pport;	/* physical lpfc_vport pointer */
@@ -605,6 +634,7 @@ struct lpfc_hba {
 	unsigned long last_completion_time;
 	struct timer_list hb_tmofunc;
 	uint8_t hb_outstanding;
+	enum hba_temp_state over_temp_state;
 	/* ndlp reference management */
 	spinlock_t ndlp_lock;
 	/*
@@ -613,7 +643,19 @@ struct lpfc_hba {
 	 */
 #define QUE_BUFTAG_BIT  (1<<31)
 	uint32_t buffer_tag_count;
-	enum hba_temp_state over_temp_state;
+	int wait_4_mlo_maint_flg;
+	wait_queue_head_t wait_4_mlo_m_q;
+	/* data structure used for latency data collection */
+#define LPFC_NO_BUCKET	   0
+#define LPFC_LINEAR_BUCKET 1
+#define LPFC_POWER2_BUCKET 2
+	uint8_t  bucket_type;
+	uint32_t bucket_base;
+	uint32_t bucket_step;
+
+/* Maximum number of events that can be outstanding at any time*/
+#define LPFC_MAX_EVT_COUNT 512
+	atomic_t fast_event_count;
 };
 
 static inline struct Scsi_Host *
@@ -650,15 +692,25 @@ lpfc_worker_wake_up(struct lpfc_hba *phba)
 	return;
 }
 
-#define FC_REG_DUMP_EVENT		0x10	/* Register for Dump events */
-#define FC_REG_TEMPERATURE_EVENT	0x20    /* Register for temperature
-						   event */
+static inline void
+lpfc_sli_read_hs(struct lpfc_hba *phba)
+{
+	/*
+	 * There was a link/board error. Read the status register to retrieve
+	 * the error event and process it.
+	 */
+	phba->sli.slistat.err_attn_event++;
+
+	/* Save status info */
+	phba->work_hs = readl(phba->HSregaddr);
+	phba->work_status[0] = readl(phba->MBslimaddr + 0xa8);
+	phba->work_status[1] = readl(phba->MBslimaddr + 0xac);
+
+	/* Clear chip Host Attention error bit */
+	writel(HA_ERATT, phba->HAregaddr);
+	readl(phba->HAregaddr); /* flush */
+	phba->pport->stopped = 1;
+
+	return;
+}
 
-struct temp_event {
-	uint32_t event_type;
-	uint32_t event_code;
-	uint32_t data;
-};
-#define LPFC_CRIT_TEMP		0x1
-#define LPFC_THRESHOLD_TEMP	0x2
-#define LPFC_NORMAL_TEMP	0x3

drivers/scsi/lpfc/lpfc_crtn.h

@@ -18,7 +18,7 @@
  * included with this package.                                     *
  *******************************************************************/
 
-typedef int (*node_filter)(struct lpfc_nodelist *ndlp, void *param);
+typedef int (*node_filter)(struct lpfc_nodelist *, void *);
 
 struct fc_rport;
 void lpfc_dump_mem(struct lpfc_hba *, LPFC_MBOXQ_t *, uint16_t);
@@ -26,11 +26,11 @@ void lpfc_read_nv(struct lpfc_hba *, LPFC_MBOXQ_t *);
 void lpfc_config_async(struct lpfc_hba *, LPFC_MBOXQ_t *, uint32_t);
 
 void lpfc_heart_beat(struct lpfc_hba *, LPFC_MBOXQ_t *);
-int lpfc_read_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb,
-		 struct lpfc_dmabuf *mp);
+int lpfc_read_la(struct lpfc_hba *, LPFC_MBOXQ_t *, struct lpfc_dmabuf *);
 void lpfc_clear_la(struct lpfc_hba *, LPFC_MBOXQ_t *);
-void lpfc_issue_clear_la(struct lpfc_hba *phba, struct lpfc_vport *vport);
+void lpfc_issue_clear_la(struct lpfc_hba *, struct lpfc_vport *);
 void lpfc_config_link(struct lpfc_hba *, LPFC_MBOXQ_t *);
+int lpfc_config_msi(struct lpfc_hba *, LPFC_MBOXQ_t *);
 int lpfc_read_sparam(struct lpfc_hba *, LPFC_MBOXQ_t *, int);
 void lpfc_read_config(struct lpfc_hba *, LPFC_MBOXQ_t *);
 void lpfc_read_lnk_stat(struct lpfc_hba *, LPFC_MBOXQ_t *);
@@ -43,7 +43,7 @@ void lpfc_unreg_vpi(struct lpfc_hba *, uint16_t, LPFC_MBOXQ_t *);
 void lpfc_init_link(struct lpfc_hba *, LPFC_MBOXQ_t *, uint32_t, uint32_t);
 
 struct lpfc_vport *lpfc_find_vport_by_did(struct lpfc_hba *, uint32_t);
-void lpfc_cleanup_rpis(struct lpfc_vport *vport, int remove);
+void lpfc_cleanup_rpis(struct lpfc_vport *, int);
 int lpfc_linkdown(struct lpfc_hba *);
 void lpfc_port_link_failure(struct lpfc_vport *);
 void lpfc_mbx_cmpl_read_la(struct lpfc_hba *, LPFC_MBOXQ_t *);
@@ -135,7 +135,7 @@ void lpfc_ct_unsol_event(struct lpfc_hba *, struct lpfc_sli_ring *,
 int lpfc_ns_cmd(struct lpfc_vport *, int, uint8_t, uint32_t);
 int lpfc_fdmi_cmd(struct lpfc_vport *, struct lpfc_nodelist *, int);
 void lpfc_fdmi_tmo(unsigned long);
-void lpfc_fdmi_timeout_handler(struct lpfc_vport *vport);
+void lpfc_fdmi_timeout_handler(struct lpfc_vport *);
 
 int lpfc_config_port_prep(struct lpfc_hba *);
 int lpfc_config_port_post(struct lpfc_hba *);
@@ -155,6 +155,8 @@ int lpfc_sli_queue_setup(struct lpfc_hba *);
 void lpfc_handle_eratt(struct lpfc_hba *);
 void lpfc_handle_latt(struct lpfc_hba *);
 irqreturn_t lpfc_intr_handler(int, void *);
+irqreturn_t lpfc_sp_intr_handler(int, void *);
+irqreturn_t lpfc_fp_intr_handler(int, void *);
 
 void lpfc_read_rev(struct lpfc_hba *, LPFC_MBOXQ_t *);
 void lpfc_config_ring(struct lpfc_hba *, int, LPFC_MBOXQ_t *);
@@ -175,11 +177,12 @@ void lpfc_mem_free(struct lpfc_hba *);
 void lpfc_stop_vport_timers(struct lpfc_vport *);
 
 void lpfc_poll_timeout(unsigned long ptr);
-void lpfc_poll_start_timer(struct lpfc_hba * phba);
-void lpfc_sli_poll_fcp_ring(struct lpfc_hba * hba);
+void lpfc_poll_start_timer(struct lpfc_hba *);
+void lpfc_poll_eratt(unsigned long);
+void lpfc_sli_poll_fcp_ring(struct lpfc_hba *);
 struct lpfc_iocbq * lpfc_sli_get_iocbq(struct lpfc_hba *);
-void lpfc_sli_release_iocbq(struct lpfc_hba * phba, struct lpfc_iocbq * iocb);
-uint16_t lpfc_sli_next_iotag(struct lpfc_hba * phba, struct lpfc_iocbq * iocb);
+void lpfc_sli_release_iocbq(struct lpfc_hba *, struct lpfc_iocbq *);
+uint16_t lpfc_sli_next_iotag(struct lpfc_hba *, struct lpfc_iocbq *);
 
 void lpfc_reset_barrier(struct lpfc_hba * phba);
 int lpfc_sli_brdready(struct lpfc_hba *, uint32_t);
@@ -187,11 +190,13 @@ int lpfc_sli_brdkill(struct lpfc_hba *);
 int lpfc_sli_brdreset(struct lpfc_hba *);
 int lpfc_sli_brdrestart(struct lpfc_hba *);
 int lpfc_sli_hba_setup(struct lpfc_hba *);
+int lpfc_sli_config_port(struct lpfc_hba *, int);
 int lpfc_sli_host_down(struct lpfc_vport *);
 int lpfc_sli_hba_down(struct lpfc_hba *);
 int lpfc_sli_issue_mbox(struct lpfc_hba *, LPFC_MBOXQ_t *, uint32_t);
 int lpfc_sli_handle_mb_event(struct lpfc_hba *);
 int lpfc_sli_flush_mbox_queue(struct lpfc_hba *);
+int lpfc_sli_check_eratt(struct lpfc_hba *);
 int lpfc_sli_handle_slow_ring_event(struct lpfc_hba *,
 				    struct lpfc_sli_ring *, uint32_t);
 void lpfc_sli_def_mbox_cmpl(struct lpfc_hba *, LPFC_MBOXQ_t *);
@@ -199,6 +204,7 @@ int lpfc_sli_issue_iocb(struct lpfc_hba *, struct lpfc_sli_ring *,
 			struct lpfc_iocbq *, uint32_t);
 void lpfc_sli_pcimem_bcopy(void *, void *, uint32_t);
 void lpfc_sli_abort_iocb_ring(struct lpfc_hba *, struct lpfc_sli_ring *);
+void lpfc_sli_flush_fcp_rings(struct lpfc_hba *);
 int lpfc_sli_ringpostbuf_put(struct lpfc_hba *, struct lpfc_sli_ring *,
 			     struct lpfc_dmabuf *);
 struct lpfc_dmabuf *lpfc_sli_ringpostbuf_get(struct lpfc_hba *,
@@ -226,17 +232,13 @@ struct lpfc_nodelist *lpfc_findnode_did(struct lpfc_vport *, uint32_t);
 struct lpfc_nodelist *lpfc_findnode_wwpn(struct lpfc_vport *,
 					 struct lpfc_name *);
 
-int lpfc_sli_issue_mbox_wait(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmboxq,
-			     uint32_t timeout);
+int lpfc_sli_issue_mbox_wait(struct lpfc_hba *, LPFC_MBOXQ_t *, uint32_t);
 
-int lpfc_sli_issue_iocb_wait(struct lpfc_hba * phba,
-			     struct lpfc_sli_ring * pring,
-			     struct lpfc_iocbq * piocb,
-			     struct lpfc_iocbq * prspiocbq,
-			     uint32_t timeout);
-void lpfc_sli_abort_fcp_cmpl(struct lpfc_hba * phba,
-			     struct lpfc_iocbq * cmdiocb,
-			     struct lpfc_iocbq * rspiocb);
+int lpfc_sli_issue_iocb_wait(struct lpfc_hba *, struct lpfc_sli_ring *,
+			     struct lpfc_iocbq *, struct lpfc_iocbq *,
+			     uint32_t);
+void lpfc_sli_abort_fcp_cmpl(struct lpfc_hba *, struct lpfc_iocbq *,
+			     struct lpfc_iocbq *);
 
 void lpfc_sli_free_hbq(struct lpfc_hba *, struct hbq_dmabuf *);
 
@@ -269,7 +271,7 @@ void lpfc_dev_loss_tmo_callbk(struct fc_rport *rport);
 
 struct lpfc_vport *lpfc_create_port(struct lpfc_hba *, int, struct device *);
 int  lpfc_vport_disable(struct fc_vport *fc_vport, bool disable);
-void lpfc_mbx_unreg_vpi(struct lpfc_vport *);
+int lpfc_mbx_unreg_vpi(struct lpfc_vport *);
 void destroy_port(struct lpfc_vport *);
 int lpfc_get_instance(void);
 void lpfc_host_attrib_init(struct Scsi_Host *);
@@ -290,6 +292,13 @@ void lpfc_unblock_fabric_iocbs(struct lpfc_hba *);
 void lpfc_adjust_queue_depth(struct lpfc_hba *);
 void lpfc_ramp_down_queue_handler(struct lpfc_hba *);
 void lpfc_ramp_up_queue_handler(struct lpfc_hba *);
+void lpfc_scsi_dev_block(struct lpfc_hba *);
+
+void
+lpfc_send_els_failure_event(struct lpfc_hba *, struct lpfc_iocbq *,
+				struct lpfc_iocbq *);
+struct lpfc_fast_path_event *lpfc_alloc_fast_evt(struct lpfc_hba *);
+void lpfc_free_fast_evt(struct lpfc_hba *, struct lpfc_fast_path_event *);
 
 #define ScsiResult(host_code, scsi_code) (((host_code) << 16) | scsi_code)
 #define HBA_EVENT_RSCN                   5

drivers/scsi/lpfc/lpfc_ct.c

@@ -34,6 +34,7 @@
 
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -134,25 +135,24 @@ lpfc_ct_unsol_event(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 		}
 		list_del(&head);
 	} else {
-		struct lpfc_iocbq  *next;
-
-		list_for_each_entry_safe(iocbq, next, &piocbq->list, list) {
+		INIT_LIST_HEAD(&head);
+		list_add_tail(&head, &piocbq->list);
+		list_for_each_entry(iocbq, &head, list) {
 			icmd = &iocbq->iocb;
 			if (icmd->ulpBdeCount == 0)
-				lpfc_ct_unsol_buffer(phba, piocbq, NULL, 0);
+				lpfc_ct_unsol_buffer(phba, iocbq, NULL, 0);
 			for (i = 0; i < icmd->ulpBdeCount; i++) {
 				paddr = getPaddr(icmd->un.cont64[i].addrHigh,
 						 icmd->un.cont64[i].addrLow);
 				mp = lpfc_sli_ringpostbuf_get(phba, pring,
 							      paddr);
 				size = icmd->un.cont64[i].tus.f.bdeSize;
-				lpfc_ct_unsol_buffer(phba, piocbq, mp, size);
+				lpfc_ct_unsol_buffer(phba, iocbq, mp, size);
 				lpfc_in_buf_free(phba, mp);
 			}
-			list_del(&iocbq->list);
-			lpfc_sli_release_iocbq(phba, iocbq);
 			lpfc_post_buffer(phba, pring, i);
 		}
+		list_del(&head);
 	}
 }
 
@@ -212,7 +212,7 @@ lpfc_alloc_ct_rsp(struct lpfc_hba *phba, int cmdcode, struct ulp_bde64 *bpl,
 		else
 			list_add_tail(&mp->list, &mlist->list);
 
-		bpl->tus.f.bdeFlags = BUFF_USE_RCV;
+		bpl->tus.f.bdeFlags = BUFF_TYPE_BDE_64I;
 		/* build buffer ptr list for IOCB */
 		bpl->addrLow = le32_to_cpu(putPaddrLow(mp->phys) );
 		bpl->addrHigh = le32_to_cpu(putPaddrHigh(mp->phys) );
@@ -283,7 +283,7 @@ lpfc_gen_req(struct lpfc_vport *vport, struct lpfc_dmabuf *bmp,
 	icmd->un.genreq64.bdl.ulpIoTag32 = 0;
 	icmd->un.genreq64.bdl.addrHigh = putPaddrHigh(bmp->phys);
 	icmd->un.genreq64.bdl.addrLow = putPaddrLow(bmp->phys);
-	icmd->un.genreq64.bdl.bdeFlags = BUFF_TYPE_BDL;
+	icmd->un.genreq64.bdl.bdeFlags = BUFF_TYPE_BLP_64;
 	icmd->un.genreq64.bdl.bdeSize = (num_entry * sizeof (struct ulp_bde64));
 
 	if (usr_flg)
@@ -861,7 +861,7 @@ lpfc_cmpl_ct(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 
 		retry++;
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
-				 "0216 Retrying NS cmd %x\n", cmdcode);
+				 "0250 Retrying NS cmd %x\n", cmdcode);
 		rc = lpfc_ns_cmd(vport, cmdcode, retry, 0);
 		if (rc == 0)
 			goto out;

drivers/scsi/lpfc/lpfc_debugfs.c

@@ -1,7 +1,7 @@
 /*******************************************************************
  * This file is part of the Emulex Linux Device Driver for         *
  * Fibre Channel Host Bus Adapters.                                *
- * Copyright (C) 2007 Emulex.  All rights reserved.                *
+ * Copyright (C) 2007-2008 Emulex.  All rights reserved.           *
  * EMULEX and SLI are trademarks of Emulex.                        *
  * www.emulex.com                                                  *
  *                                                                 *
@@ -35,6 +35,7 @@
 
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -46,13 +47,14 @@
 #include "lpfc_debugfs.h"
 
 #ifdef CONFIG_LPFC_DEBUG_FS
-/* debugfs interface
+/**
+ * debugfs interface
  *
  * To access this interface the user should:
  * # mkdir /debug
  * # mount -t debugfs none /debug
  *
- * The lpfc debugfs directory hierachy is:
+ * The lpfc debugfs directory hierarchy is:
  * lpfc/lpfcX/vportY
  * where X is the lpfc hba unique_id
  * where Y is the vport VPI on that hba
@@ -61,14 +63,21 @@
  * discovery_trace
  * This is an ACSII readable file that contains a trace of the last
  * lpfc_debugfs_max_disc_trc events that happened on a specific vport.
- * See lpfc_debugfs.h for different categories of
- * discovery events. To enable the discovery trace, the following
- * module parameters must be set:
+ * See lpfc_debugfs.h for different categories of  discovery events.
+ * To enable the discovery trace, the following module parameters must be set:
  * lpfc_debugfs_enable=1         Turns on lpfc debugfs filesystem support
  * lpfc_debugfs_max_disc_trc=X   Where X is the event trace depth for
  *                               EACH vport. X MUST also be a power of 2.
  * lpfc_debugfs_mask_disc_trc=Y  Where Y is an event mask as defined in
  *                               lpfc_debugfs.h .
+ *
+ * slow_ring_trace
+ * This is an ACSII readable file that contains a trace of the last
+ * lpfc_debugfs_max_slow_ring_trc events that happened on a specific HBA.
+ * To enable the slow ring trace, the following module parameters must be set:
+ * lpfc_debugfs_enable=1         Turns on lpfc debugfs filesystem support
+ * lpfc_debugfs_max_slow_ring_trc=X   Where X is the event trace depth for
+ *                               the HBA. X MUST also be a power of 2.
  */
 static int lpfc_debugfs_enable = 1;
 module_param(lpfc_debugfs_enable, int, 0);
@@ -117,6 +126,25 @@ struct lpfc_debug {
 static atomic_t lpfc_debugfs_seq_trc_cnt = ATOMIC_INIT(0);
 static unsigned long lpfc_debugfs_start_time = 0L;
 
+/**
+ * lpfc_debugfs_disc_trc_data - Dump discovery logging to a buffer.
+ * @vport: The vport to gather the log info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine gathers the lpfc discovery debugfs data from the @vport and
+ * dumps it to @buf up to @size number of bytes. It will start at the next entry
+ * in the log and process the log until the end of the buffer. Then it will
+ * gather from the beginning of the log and process until the current entry.
+ *
+ * Notes:
+ * Discovery logging will be disabled while while this routine dumps the log.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_disc_trc_data(struct lpfc_vport *vport, char *buf, int size)
 {
@@ -125,7 +153,6 @@ lpfc_debugfs_disc_trc_data(struct lpfc_vport *vport, char *buf, int size)
 	struct lpfc_debugfs_trc *dtp;
 	char buffer[LPFC_DEBUG_TRC_ENTRY_SIZE];
 
-
 	enable = lpfc_debugfs_enable;
 	lpfc_debugfs_enable = 0;
 
@@ -159,6 +186,25 @@ lpfc_debugfs_disc_trc_data(struct lpfc_vport *vport, char *buf, int size)
 	return len;
 }
 
+/**
+ * lpfc_debugfs_slow_ring_trc_data - Dump slow ring logging to a buffer.
+ * @phba: The HBA to gather the log info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine gathers the lpfc slow ring debugfs data from the @phba and
+ * dumps it to @buf up to @size number of bytes. It will start at the next entry
+ * in the log and process the log until the end of the buffer. Then it will
+ * gather from the beginning of the log and process until the current entry.
+ *
+ * Notes:
+ * Slow ring logging will be disabled while while this routine dumps the log.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_slow_ring_trc_data(struct lpfc_hba *phba, char *buf, int size)
 {
@@ -203,6 +249,25 @@ lpfc_debugfs_slow_ring_trc_data(struct lpfc_hba *phba, char *buf, int size)
 
 static int lpfc_debugfs_last_hbq = -1;
 
+/**
+ * lpfc_debugfs_hbqinfo_data - Dump host buffer queue info to a buffer.
+ * @phba: The HBA to gather host buffer info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine dumps the host buffer queue info from the @phba to @buf up to
+ * @size number of bytes. A header that describes the current hbq state will be
+ * dumped to @buf first and then info on each hbq entry will be dumped to @buf
+ * until @size bytes have been dumped or all the hbq info has been dumped.
+ *
+ * Notes:
+ * This routine will rotate through each configured HBQ each time called.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_hbqinfo_data(struct lpfc_hba *phba, char *buf, int size)
 {
@@ -303,6 +368,24 @@ skipit:
 
 static int lpfc_debugfs_last_hba_slim_off;
 
+/**
+ * lpfc_debugfs_dumpHBASlim_data - Dump HBA SLIM info to a buffer.
+ * @phba: The HBA to gather SLIM info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine dumps the current contents of HBA SLIM for the HBA associated
+ * with @phba to @buf up to @size bytes of data. This is the raw HBA SLIM data.
+ *
+ * Notes:
+ * This routine will only dump up to 1024 bytes of data each time called and
+ * should be called multiple times to dump the entire HBA SLIM.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_dumpHBASlim_data(struct lpfc_hba *phba, char *buf, int size)
 {
@@ -342,6 +425,21 @@ lpfc_debugfs_dumpHBASlim_data(struct lpfc_hba *phba, char *buf, int size)
 	return len;
 }
 
+/**
+ * lpfc_debugfs_dumpHostSlim_data - Dump host SLIM info to a buffer.
+ * @phba: The HBA to gather Host SLIM info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine dumps the current contents of host SLIM for the host associated
+ * with @phba to @buf up to @size bytes of data. The dump will contain the
+ * Mailbox, PCB, Rings, and Registers that are located in host memory.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_dumpHostSlim_data(struct lpfc_hba *phba, char *buf, int size)
 {
@@ -357,7 +455,7 @@ lpfc_debugfs_dumpHostSlim_data(struct lpfc_hba *phba, char *buf, int size)
 	spin_lock_irq(&phba->hbalock);
 
 	len +=  snprintf(buf+len, size-len, "SLIM Mailbox\n");
-	ptr = (uint32_t *)phba->slim2p;
+	ptr = (uint32_t *)phba->slim2p.virt;
 	i = sizeof(MAILBOX_t);
 	while (i > 0) {
 		len +=  snprintf(buf+len, size-len,
@@ -370,7 +468,7 @@ lpfc_debugfs_dumpHostSlim_data(struct lpfc_hba *phba, char *buf, int size)
 	}
 
 	len +=  snprintf(buf+len, size-len, "SLIM PCB\n");
-	ptr = (uint32_t *)&phba->slim2p->pcb;
+	ptr = (uint32_t *)phba->pcb;
 	i = sizeof(PCB_t);
 	while (i > 0) {
 		len +=  snprintf(buf+len, size-len,
@@ -382,44 +480,16 @@ lpfc_debugfs_dumpHostSlim_data(struct lpfc_hba *phba, char *buf, int size)
 		off += (8 * sizeof(uint32_t));
 	}
 
-	pgpp = (struct lpfc_pgp *)&phba->slim2p->mbx.us.s3_pgp.port;
-	pring = &psli->ring[0];
-	len +=  snprintf(buf+len, size-len,
-		"Ring 0: CMD GetInx:%d (Max:%d Next:%d Local:%d flg:x%x)  "
-		"RSP PutInx:%d Max:%d\n",
-		pgpp->cmdGetInx, pring->numCiocb,
-		pring->next_cmdidx, pring->local_getidx, pring->flag,
-		pgpp->rspPutInx, pring->numRiocb);
-	pgpp++;
-
-	pring = &psli->ring[1];
-	len +=  snprintf(buf+len, size-len,
-		"Ring 1: CMD GetInx:%d (Max:%d Next:%d Local:%d flg:x%x)  "
-		"RSP PutInx:%d Max:%d\n",
-		pgpp->cmdGetInx, pring->numCiocb,
-		pring->next_cmdidx, pring->local_getidx, pring->flag,
-		pgpp->rspPutInx, pring->numRiocb);
-	pgpp++;
-
-	pring = &psli->ring[2];
-	len +=  snprintf(buf+len, size-len,
-		"Ring 2: CMD GetInx:%d (Max:%d Next:%d Local:%d flg:x%x)  "
-		"RSP PutInx:%d Max:%d\n",
-		pgpp->cmdGetInx, pring->numCiocb,
-		pring->next_cmdidx, pring->local_getidx, pring->flag,
-		pgpp->rspPutInx, pring->numRiocb);
-	pgpp++;
-
-	pring = &psli->ring[3];
-	len +=  snprintf(buf+len, size-len,
-		"Ring 3: CMD GetInx:%d (Max:%d Next:%d Local:%d flg:x%x)  "
-		"RSP PutInx:%d Max:%d\n",
-		pgpp->cmdGetInx, pring->numCiocb,
-		pring->next_cmdidx, pring->local_getidx, pring->flag,
-		pgpp->rspPutInx, pring->numRiocb);
-
-
-	ptr = (uint32_t *)&phba->slim2p->mbx.us.s3_pgp.hbq_get;
+	for (i = 0; i < 4; i++) {
+		pgpp = &phba->port_gp[i];
+		pring = &psli->ring[i];
+		len +=  snprintf(buf+len, size-len,
+				 "Ring %d: CMD GetInx:%d (Max:%d Next:%d "
+				 "Local:%d flg:x%x)  RSP PutInx:%d Max:%d\n",
+				 i, pgpp->cmdGetInx, pring->numCiocb,
+				 pring->next_cmdidx, pring->local_getidx,
+				 pring->flag, pgpp->rspPutInx, pring->numRiocb);
+	}
 	word0 = readl(phba->HAregaddr);
 	word1 = readl(phba->CAregaddr);
 	word2 = readl(phba->HSregaddr);
@@ -430,6 +500,21 @@ lpfc_debugfs_dumpHostSlim_data(struct lpfc_hba *phba, char *buf, int size)
 	return len;
 }
 
+/**
+ * lpfc_debugfs_nodelist_data - Dump target node list to a buffer.
+ * @vport: The vport to gather target node info from.
+ * @buf: The buffer to dump log into.
+ * @size: The maximum amount of data to process.
+ *
+ * Description:
+ * This routine dumps the current target node list associated with @vport to
+ * @buf up to @size bytes of data. Each node entry in the dump will contain a
+ * node state, DID, WWPN, WWNN, RPI, flags, type, and other useful fields.
+ *
+ * Return Value:
+ * This routine returns the amount of bytes that were dumped into @buf and will
+ * not exceed @size.
+ **/
 static int
 lpfc_debugfs_nodelist_data(struct lpfc_vport *vport, char *buf, int size)
 {
@@ -513,7 +598,22 @@ lpfc_debugfs_nodelist_data(struct lpfc_vport *vport, char *buf, int size)
 }
 #endif
 
-
+/**
+ * lpfc_debugfs_disc_trc - Store discovery trace log.
+ * @vport: The vport to associate this trace string with for retrieval.
+ * @mask: Log entry classification.
+ * @fmt: Format string to be displayed when dumping the log.
+ * @data1: 1st data parameter to be applied to @fmt.
+ * @data2: 2nd data parameter to be applied to @fmt.
+ * @data3: 3rd data parameter to be applied to @fmt.
+ *
+ * Description:
+ * This routine is used by the driver code to add a debugfs log entry to the
+ * discovery trace buffer associated with @vport. Only entries with a @mask that
+ * match the current debugfs discovery mask will be saved. Entries that do not
+ * match will be thrown away. @fmt, @data1, @data2, and @data3 are used like
+ * printf when displaying the log.
+ **/
 inline void
 lpfc_debugfs_disc_trc(struct lpfc_vport *vport, int mask, char *fmt,
 	uint32_t data1, uint32_t data2, uint32_t data3)
@@ -542,6 +642,19 @@ lpfc_debugfs_disc_trc(struct lpfc_vport *vport, int mask, char *fmt,
 	return;
 }
 
+/**
+ * lpfc_debugfs_slow_ring_trc - Store slow ring trace log.
+ * @phba: The phba to associate this trace string with for retrieval.
+ * @fmt: Format string to be displayed when dumping the log.
+ * @data1: 1st data parameter to be applied to @fmt.
+ * @data2: 2nd data parameter to be applied to @fmt.
+ * @data3: 3rd data parameter to be applied to @fmt.
+ *
+ * Description:
+ * This routine is used by the driver code to add a debugfs log entry to the
+ * discovery trace buffer associated with @vport. @fmt, @data1, @data2, and
+ * @data3 are used like printf when displaying the log.
+ **/
 inline void
 lpfc_debugfs_slow_ring_trc(struct lpfc_hba *phba, char *fmt,
 	uint32_t data1, uint32_t data2, uint32_t data3)
@@ -568,6 +681,21 @@ lpfc_debugfs_slow_ring_trc(struct lpfc_hba *phba, char *fmt,
 }
 
 #ifdef CONFIG_LPFC_DEBUG_FS
+/**
+ * lpfc_debugfs_disc_trc_open - Open the discovery trace log.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_disc_trc_open(struct inode *inode, struct file *file)
 {
@@ -585,7 +713,7 @@ lpfc_debugfs_disc_trc_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	size =  (lpfc_debugfs_max_disc_trc * LPFC_DEBUG_TRC_ENTRY_SIZE);
 	size = PAGE_ALIGN(size);
 
@@ -603,6 +731,21 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_slow_ring_trc_open - Open the Slow Ring trace log.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_slow_ring_trc_open(struct inode *inode, struct file *file)
 {
@@ -620,7 +763,7 @@ lpfc_debugfs_slow_ring_trc_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	size =  (lpfc_debugfs_max_slow_ring_trc * LPFC_DEBUG_TRC_ENTRY_SIZE);
 	size = PAGE_ALIGN(size);
 
@@ -638,6 +781,21 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_hbqinfo_open - Open the hbqinfo debugfs buffer.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_hbqinfo_open(struct inode *inode, struct file *file)
 {
@@ -649,7 +807,7 @@ lpfc_debugfs_hbqinfo_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	debug->buffer = kmalloc(LPFC_HBQINFO_SIZE, GFP_KERNEL);
 	if (!debug->buffer) {
 		kfree(debug);
@@ -665,6 +823,21 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_dumpHBASlim_open - Open the Dump HBA SLIM debugfs buffer.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_dumpHBASlim_open(struct inode *inode, struct file *file)
 {
@@ -676,7 +849,7 @@ lpfc_debugfs_dumpHBASlim_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	debug->buffer = kmalloc(LPFC_DUMPHBASLIM_SIZE, GFP_KERNEL);
 	if (!debug->buffer) {
 		kfree(debug);
@@ -692,6 +865,21 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_dumpHostSlim_open - Open the Dump Host SLIM debugfs buffer.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_dumpHostSlim_open(struct inode *inode, struct file *file)
 {
@@ -703,7 +891,7 @@ lpfc_debugfs_dumpHostSlim_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	debug->buffer = kmalloc(LPFC_DUMPHOSTSLIM_SIZE, GFP_KERNEL);
 	if (!debug->buffer) {
 		kfree(debug);
@@ -719,6 +907,21 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_nodelist_open - Open the nodelist debugfs file.
+ * @inode: The inode pointer that contains a vport pointer.
+ * @file: The file pointer to attach the log output.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs open file operation. It gets
+ * the vport from the i_private field in @inode, allocates the necessary buffer
+ * for the log, fills the buffer from the in-memory log for this vport, and then
+ * returns a pointer to that log in the private_data field in @file.
+ *
+ * Returns:
+ * This function returns zero if successful. On error it will return an negative
+ * error value.
+ **/
 static int
 lpfc_debugfs_nodelist_open(struct inode *inode, struct file *file)
 {
@@ -730,7 +933,7 @@ lpfc_debugfs_nodelist_open(struct inode *inode, struct file *file)
 	if (!debug)
 		goto out;
 
-	/* Round to page boundry */
+	/* Round to page boundary */
 	debug->buffer = kmalloc(LPFC_NODELIST_SIZE, GFP_KERNEL);
 	if (!debug->buffer) {
 		kfree(debug);
@@ -746,6 +949,23 @@ out:
 	return rc;
 }
 
+/**
+ * lpfc_debugfs_lseek - Seek through a debugfs file.
+ * @file: The file pointer to seek through.
+ * @off: The offset to seek to or the amount to seek by.
+ * @whence: Indicates how to seek.
+ *
+ * Description:
+ * This routine is the entry point for the debugfs lseek file operation. The
+ * @whence parameter indicates whether @off is the offset to directly seek to,
+ * or if it is a value to seek forward or reverse by. This function figures out
+ * what the new offset of the debugfs file will be and assigns that value to the
+ * f_pos field of @file.
+ *
+ * Returns:
+ * This function returns the new offset if successful and returns a negative
+ * error if unable to process the seek.
+ **/
 static loff_t
 lpfc_debugfs_lseek(struct file *file, loff_t off, int whence)
 {
@@ -767,6 +987,22 @@ lpfc_debugfs_lseek(struct file *file, loff_t off, int whence)
 	return (pos < 0 || pos > debug->len) ? -EINVAL : (file->f_pos = pos);
 }
 
+/**
+ * lpfc_debugfs_read - Read a debugfs file.
+ * @file: The file pointer to read from.
+ * @buf: The buffer to copy the data to.
+ * @nbytes: The number of bytes to read.
+ * @ppos: The position in the file to start reading from.
+ *
+ * Description:
+ * This routine reads data from from the buffer indicated in the private_data
+ * field of @file. It will start reading at @ppos and copy up to @nbytes of
+ * data to @buf.
+ *
+ * Returns:
+ * This function returns the amount of data that was read (this could be less
+ * than @nbytes if the end of the file was reached) or a negative error value.
+ **/
 static ssize_t
 lpfc_debugfs_read(struct file *file, char __user *buf,
 		  size_t nbytes, loff_t *ppos)
@@ -776,6 +1012,18 @@ lpfc_debugfs_read(struct file *file, char __user *buf,
 				       debug->len);
 }
 
+/**
+ * lpfc_debugfs_release - Release the buffer used to store debugfs file data.
+ * @inode: The inode pointer that contains a vport pointer. (unused)
+ * @file: The file pointer that contains the buffer to release.
+ *
+ * Description:
+ * This routine frees the buffer that was allocated when the debugfs file was
+ * opened.
+ *
+ * Returns:
+ * This function returns zero.
+ **/
 static int
 lpfc_debugfs_release(struct inode *inode, struct file *file)
 {
@@ -845,6 +1093,16 @@ static struct dentry *lpfc_debugfs_root = NULL;
 static atomic_t lpfc_debugfs_hba_count;
 #endif
 
+/**
+ * lpfc_debugfs_initialize - Initialize debugfs for a vport.
+ * @vport: The vport pointer to initialize.
+ *
+ * Description:
+ * When Debugfs is configured this routine sets up the lpfc debugfs file system.
+ * If not already created, this routine will create the lpfc directory, and
+ * lpfcX directory (for this HBA), and vportX directory for this vport. It will
+ * also create each file used to access lpfc specific debugfs information.
+ **/
 inline void
 lpfc_debugfs_initialize(struct lpfc_vport *vport)
 {
@@ -862,7 +1120,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 		atomic_set(&lpfc_debugfs_hba_count, 0);
 		if (!lpfc_debugfs_root) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-					 "0409 Cannot create debugfs root\n");
+					 "0408 Cannot create debugfs root\n");
 			goto debug_failed;
 		}
 	}
@@ -876,7 +1134,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 			debugfs_create_dir(name, lpfc_debugfs_root);
 		if (!phba->hba_debugfs_root) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-					 "0409 Cannot create debugfs hba\n");
+					 "0412 Cannot create debugfs hba\n");
 			goto debug_failed;
 		}
 		atomic_inc(&lpfc_debugfs_hba_count);
@@ -890,7 +1148,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				 phba, &lpfc_debugfs_op_hbqinfo);
 		if (!phba->debug_hbqinfo) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-				"0409 Cannot create debugfs hbqinfo\n");
+				"0411 Cannot create debugfs hbqinfo\n");
 			goto debug_failed;
 		}
 
@@ -902,7 +1160,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				 phba, &lpfc_debugfs_op_dumpHBASlim);
 		if (!phba->debug_dumpHBASlim) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-				"0409 Cannot create debugfs dumpHBASlim\n");
+				"0413 Cannot create debugfs dumpHBASlim\n");
 			goto debug_failed;
 		}
 
@@ -914,7 +1172,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				 phba, &lpfc_debugfs_op_dumpHostSlim);
 		if (!phba->debug_dumpHostSlim) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-				"0409 Cannot create debugfs dumpHostSlim\n");
+				"0414 Cannot create debugfs dumpHostSlim\n");
 			goto debug_failed;
 		}
 
@@ -944,7 +1202,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				 phba, &lpfc_debugfs_op_slow_ring_trc);
 		if (!phba->debug_slow_ring_trc) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-					 "0409 Cannot create debugfs "
+					 "0415 Cannot create debugfs "
 					 "slow_ring_trace\n");
 			goto debug_failed;
 		}
@@ -955,7 +1213,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				GFP_KERNEL);
 			if (!phba->slow_ring_trc) {
 				lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-						 "0409 Cannot create debugfs "
+						 "0416 Cannot create debugfs "
 						 "slow_ring buffer\n");
 				goto debug_failed;
 			}
@@ -972,7 +1230,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 			debugfs_create_dir(name, phba->hba_debugfs_root);
 		if (!vport->vport_debugfs_root) {
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-					 "0409 Cant create debugfs");
+					 "0417 Cant create debugfs");
 			goto debug_failed;
 		}
 		atomic_inc(&phba->debugfs_vport_count);
@@ -1001,7 +1259,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 
 	if (!vport->disc_trc) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-				 "0409 Cannot create debugfs disc trace "
+				 "0418 Cannot create debugfs disc trace "
 				 "buffer\n");
 		goto debug_failed;
 	}
@@ -1014,7 +1272,7 @@ lpfc_debugfs_initialize(struct lpfc_vport *vport)
 				 vport, &lpfc_debugfs_op_disc_trc);
 	if (!vport->debug_disc_trc) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
-				 "0409 Cannot create debugfs "
+				 "0419 Cannot create debugfs "
 				 "discovery_trace\n");
 		goto debug_failed;
 	}
@@ -1033,7 +1291,17 @@ debug_failed:
 #endif
 }
 
-
+/**
+ * lpfc_debugfs_terminate -  Tear down debugfs infrastructure for this vport.
+ * @vport: The vport pointer to remove from debugfs.
+ *
+ * Description:
+ * When Debugfs is configured this routine removes debugfs file system elements
+ * that are specific to this vport. It also checks to see if there are any
+ * users left for the debugfs directories associated with the HBA and driver. If
+ * this is the last user of the HBA directory or driver directory then it will
+ * remove those from the debugfs infrastructure as well.
+ **/
 inline void
 lpfc_debugfs_terminate(struct lpfc_vport *vport)
 {
@@ -1096,5 +1364,3 @@ lpfc_debugfs_terminate(struct lpfc_vport *vport)
 #endif
 	return;
 }
-
-

drivers/scsi/lpfc/lpfc_disc.h

@@ -37,6 +37,7 @@ enum lpfc_work_type {
 	LPFC_EVT_KILL,
 	LPFC_EVT_ELS_RETRY,
 	LPFC_EVT_DEV_LOSS,
+	LPFC_EVT_FASTPATH_MGMT_EVT,
 };
 
 /* structure used to queue event to the discovery tasklet */
@@ -47,6 +48,24 @@ struct lpfc_work_evt {
 	enum lpfc_work_type   evt;
 };
 
+struct lpfc_scsi_check_condition_event;
+struct lpfc_scsi_varqueuedepth_event;
+struct lpfc_scsi_event_header;
+struct lpfc_fabric_event_header;
+struct lpfc_fcprdchkerr_event;
+
+/* structure used for sending events from fast path */
+struct lpfc_fast_path_event {
+	struct lpfc_work_evt work_evt;
+	struct lpfc_vport     *vport;
+	union {
+		struct lpfc_scsi_check_condition_event check_cond_evt;
+		struct lpfc_scsi_varqueuedepth_event queue_depth_evt;
+		struct lpfc_scsi_event_header scsi_evt;
+		struct lpfc_fabric_event_header fabric_evt;
+		struct lpfc_fcprdchkerr_event read_check_error;
+	} un;
+};
 
 struct lpfc_nodelist {
 	struct list_head nlp_listp;
@@ -88,6 +107,10 @@ struct lpfc_nodelist {
 	unsigned long last_ramp_up_time;        /* jiffy of last ramp up */
 	unsigned long last_q_full_time;		/* jiffy of last queue full */
 	struct kref     kref;
+	atomic_t cmd_pending;
+	uint32_t cmd_qdepth;
+	unsigned long last_change_time;
+	struct lpfc_scsicmd_bkt *lat_data;	/* Latency data */
 };
 
 /* Defines for nlp_flag (uint32) */

drivers/scsi/lpfc/lpfc_hbadisc.c

@@ -30,6 +30,7 @@
 #include <scsi/scsi_transport_fc.h>
 
 #include "lpfc_hw.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_sli.h"
 #include "lpfc_scsi.h"
@@ -88,14 +89,6 @@ lpfc_terminate_rport_io(struct fc_rport *rport)
 			&phba->sli.ring[phba->sli.fcp_ring],
 			ndlp->nlp_sid, 0, LPFC_CTX_TGT);
 	}
-
-	/*
-	 * A device is normally blocked for rediscovery and unblocked when
-	 * devloss timeout happens.  In case a vport is removed or driver
-	 * unloaded before devloss timeout happens, we need to unblock here.
-	 */
-	scsi_target_unblock(&rport->dev);
-	return;
 }
 
 /*
@@ -215,8 +208,16 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 		return;
 	}
 
-	if (ndlp->nlp_state == NLP_STE_MAPPED_NODE)
+	if (ndlp->nlp_state == NLP_STE_MAPPED_NODE) {
+		lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
+				 "0284 Devloss timeout Ignored on "
+				 "WWPN %x:%x:%x:%x:%x:%x:%x:%x "
+				 "NPort x%x\n",
+				 *name, *(name+1), *(name+2), *(name+3),
+				 *(name+4), *(name+5), *(name+6), *(name+7),
+				 ndlp->nlp_DID);
 		return;
+	}
 
 	if (ndlp->nlp_type & NLP_FABRIC) {
 		/* We will clean up these Nodes in linkup */
@@ -237,8 +238,6 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 		lpfc_sli_abort_iocb(vport, &phba->sli.ring[phba->sli.fcp_ring],
 				    ndlp->nlp_sid, 0, LPFC_CTX_TGT);
 	}
-	if (vport->load_flag & FC_UNLOADING)
-		warn_on = 0;
 
 	if (warn_on) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_DISCOVERY,
@@ -276,6 +275,124 @@ lpfc_dev_loss_tmo_handler(struct lpfc_nodelist *ndlp)
 		lpfc_disc_state_machine(vport, ndlp, NULL, NLP_EVT_DEVICE_RM);
 }
 
+/**
+ * lpfc_alloc_fast_evt: Allocates data structure for posting event.
+ * @phba: Pointer to hba context object.
+ *
+ * This function is called from the functions which need to post
+ * events from interrupt context. This function allocates data
+ * structure required for posting event. It also keeps track of
+ * number of events pending and prevent event storm when there are
+ * too many events.
+ **/
+struct lpfc_fast_path_event *
+lpfc_alloc_fast_evt(struct lpfc_hba *phba) {
+	struct lpfc_fast_path_event *ret;
+
+	/* If there are lot of fast event do not exhaust memory due to this */
+	if (atomic_read(&phba->fast_event_count) > LPFC_MAX_EVT_COUNT)
+		return NULL;
+
+	ret = kzalloc(sizeof(struct lpfc_fast_path_event),
+			GFP_ATOMIC);
+	if (ret)
+		atomic_inc(&phba->fast_event_count);
+	INIT_LIST_HEAD(&ret->work_evt.evt_listp);
+	ret->work_evt.evt = LPFC_EVT_FASTPATH_MGMT_EVT;
+	return ret;
+}
+
+/**
+ * lpfc_free_fast_evt: Frees event data structure.
+ * @phba: Pointer to hba context object.
+ * @evt:  Event object which need to be freed.
+ *
+ * This function frees the data structure required for posting
+ * events.
+ **/
+void
+lpfc_free_fast_evt(struct lpfc_hba *phba,
+		struct lpfc_fast_path_event *evt) {
+
+	atomic_dec(&phba->fast_event_count);
+	kfree(evt);
+}
+
+/**
+ * lpfc_send_fastpath_evt: Posts events generated from fast path.
+ * @phba: Pointer to hba context object.
+ * @evtp: Event data structure.
+ *
+ * This function is called from worker thread, when the interrupt
+ * context need to post an event. This function posts the event
+ * to fc transport netlink interface.
+ **/
+static void
+lpfc_send_fastpath_evt(struct lpfc_hba *phba,
+		struct lpfc_work_evt *evtp)
+{
+	unsigned long evt_category, evt_sub_category;
+	struct lpfc_fast_path_event *fast_evt_data;
+	char *evt_data;
+	uint32_t evt_data_size;
+	struct Scsi_Host *shost;
+
+	fast_evt_data = container_of(evtp, struct lpfc_fast_path_event,
+		work_evt);
+
+	evt_category = (unsigned long) fast_evt_data->un.fabric_evt.event_type;
+	evt_sub_category = (unsigned long) fast_evt_data->un.
+			fabric_evt.subcategory;
+	shost = lpfc_shost_from_vport(fast_evt_data->vport);
+	if (evt_category == FC_REG_FABRIC_EVENT) {
+		if (evt_sub_category == LPFC_EVENT_FCPRDCHKERR) {
+			evt_data = (char *) &fast_evt_data->un.read_check_error;
+			evt_data_size = sizeof(fast_evt_data->un.
+				read_check_error);
+		} else if ((evt_sub_category == LPFC_EVENT_FABRIC_BUSY) ||
+			(evt_sub_category == IOSTAT_NPORT_BSY)) {
+			evt_data = (char *) &fast_evt_data->un.fabric_evt;
+			evt_data_size = sizeof(fast_evt_data->un.fabric_evt);
+		} else {
+			lpfc_free_fast_evt(phba, fast_evt_data);
+			return;
+		}
+	} else if (evt_category == FC_REG_SCSI_EVENT) {
+		switch (evt_sub_category) {
+		case LPFC_EVENT_QFULL:
+		case LPFC_EVENT_DEVBSY:
+			evt_data = (char *) &fast_evt_data->un.scsi_evt;
+			evt_data_size = sizeof(fast_evt_data->un.scsi_evt);
+			break;
+		case LPFC_EVENT_CHECK_COND:
+			evt_data = (char *) &fast_evt_data->un.check_cond_evt;
+			evt_data_size =  sizeof(fast_evt_data->un.
+				check_cond_evt);
+			break;
+		case LPFC_EVENT_VARQUEDEPTH:
+			evt_data = (char *) &fast_evt_data->un.queue_depth_evt;
+			evt_data_size = sizeof(fast_evt_data->un.
+				queue_depth_evt);
+			break;
+		default:
+			lpfc_free_fast_evt(phba, fast_evt_data);
+			return;
+		}
+	} else {
+		lpfc_free_fast_evt(phba, fast_evt_data);
+		return;
+	}
+
+	fc_host_post_vendor_event(shost,
+		fc_get_event_number(),
+		evt_data_size,
+		evt_data,
+		SCSI_NL_VID_TYPE_PCI | PCI_VENDOR_ID_EMULEX);
+
+	lpfc_free_fast_evt(phba, fast_evt_data);
+	return;
+}
+
 static void
 lpfc_work_list_done(struct lpfc_hba *phba)
 {
@@ -347,6 +464,10 @@ lpfc_work_list_done(struct lpfc_hba *phba)
 			lpfc_unblock_mgmt_io(phba);
 			complete((struct completion *)(evtp->evt_arg2));
 			break;
+		case LPFC_EVT_FASTPATH_MGMT_EVT:
+			lpfc_send_fastpath_evt(phba, evtp);
+			free_evt = 0;
+			break;
 		}
 		if (free_evt)
 			kfree(evtp);
@@ -371,6 +492,7 @@ lpfc_work_done(struct lpfc_hba *phba)
 	spin_unlock_irq(&phba->hbalock);
 
 	if (ha_copy & HA_ERATT)
+		/* Handle the error attention event */
 		lpfc_handle_eratt(phba);
 
 	if (ha_copy & HA_MBATT)
@@ -378,6 +500,7 @@ lpfc_work_done(struct lpfc_hba *phba)
 
 	if (ha_copy & HA_LATT)
 		lpfc_handle_latt(phba);
+
 	vports = lpfc_create_vport_work_array(phba);
 	if (vports != NULL)
 		for(i = 0; i <= phba->max_vpi; i++) {
@@ -1013,14 +1136,10 @@ out:
 }
 
 static void
-lpfc_mbx_issue_link_down(struct lpfc_hba *phba)
+lpfc_enable_la(struct lpfc_hba *phba)
 {
 	uint32_t control;
 	struct lpfc_sli *psli = &phba->sli;
-
-	lpfc_linkdown(phba);
-
-	/* turn on Link Attention interrupts - no CLEAR_LA needed */
 	spin_lock_irq(&phba->hbalock);
 	psli->sli_flag |= LPFC_PROCESS_LA;
 	control = readl(phba->HCregaddr);
@@ -1030,6 +1149,15 @@ lpfc_mbx_issue_link_down(struct lpfc_hba *phba)
 	spin_unlock_irq(&phba->hbalock);
 }
 
+static void
+lpfc_mbx_issue_link_down(struct lpfc_hba *phba)
+{
+	lpfc_linkdown(phba);
+	lpfc_enable_la(phba);
+	/* turn on Link Attention interrupts - no CLEAR_LA needed */
+}
+
+
 /*
  * This routine handles processing a READ_LA mailbox
  * command upon completion. It is setup in the LPFC_MBOXQ
@@ -1077,8 +1205,12 @@ lpfc_mbx_cmpl_read_la(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	}
 
 	phba->fc_eventTag = la->eventTag;
+	if (la->mm)
+		phba->sli.sli_flag |= LPFC_MENLO_MAINT;
+	else
+		phba->sli.sli_flag &= ~LPFC_MENLO_MAINT;
 
-	if (la->attType == AT_LINK_UP) {
+	if (la->attType == AT_LINK_UP && (!la->mm)) {
 		phba->fc_stat.LinkUp++;
 		if (phba->link_flag & LS_LOOPBACK_MODE) {
 			lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
@@ -1090,13 +1222,15 @@ lpfc_mbx_cmpl_read_la(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 		} else {
 			lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
 					"1303 Link Up Event x%x received "
-					"Data: x%x x%x x%x x%x\n",
+					"Data: x%x x%x x%x x%x x%x x%x %d\n",
 					la->eventTag, phba->fc_eventTag,
 					la->granted_AL_PA, la->UlnkSpeed,
-					phba->alpa_map[0]);
+					phba->alpa_map[0],
+					la->mm, la->fa,
+					phba->wait_4_mlo_maint_flg);
 		}
 		lpfc_mbx_process_link_up(phba, la);
-	} else {
+	} else if (la->attType == AT_LINK_DOWN) {
 		phba->fc_stat.LinkDown++;
 		if (phba->link_flag & LS_LOOPBACK_MODE) {
 			lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
@@ -1109,11 +1243,46 @@ lpfc_mbx_cmpl_read_la(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 		else {
 			lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
 				"1305 Link Down Event x%x received "
+				"Data: x%x x%x x%x x%x x%x\n",
+				la->eventTag, phba->fc_eventTag,
+				phba->pport->port_state, vport->fc_flag,
+				la->mm, la->fa);
+		}
+		lpfc_mbx_issue_link_down(phba);
+	}
+	if (la->mm && la->attType == AT_LINK_UP) {
+		if (phba->link_state != LPFC_LINK_DOWN) {
+			phba->fc_stat.LinkDown++;
+			lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
+				"1312 Link Down Event x%x received "
+				"Data: x%x x%x x%x\n",
+				la->eventTag, phba->fc_eventTag,
+				phba->pport->port_state, vport->fc_flag);
+			lpfc_mbx_issue_link_down(phba);
+		} else
+			lpfc_enable_la(phba);
+
+		lpfc_printf_log(phba, KERN_ERR, LOG_LINK_EVENT,
+				"1310 Menlo Maint Mode Link up Event x%x rcvd "
 				"Data: x%x x%x x%x\n",
 				la->eventTag, phba->fc_eventTag,
 				phba->pport->port_state, vport->fc_flag);
+		/*
+		 * The cmnd that triggered this will be waiting for this
+		 * signal.
+		 */
+		/* WAKEUP for MENLO_SET_MODE or MENLO_RESET command. */
+		if (phba->wait_4_mlo_maint_flg) {
+			phba->wait_4_mlo_maint_flg = 0;
+			wake_up_interruptible(&phba->wait_4_mlo_m_q);
 		}
-		lpfc_mbx_issue_link_down(phba);
+	}
+
+	if (la->fa) {
+		if (la->mm)
+			lpfc_issue_clear_la(phba, vport);
+		lpfc_printf_log(phba, KERN_INFO, LOG_LINK_EVENT,
+				"1311 fa %d\n", la->fa);
 	}
 
 lpfc_mbx_cmpl_read_la_free_mbuf:
@@ -1177,7 +1346,7 @@ lpfc_mbx_cmpl_unreg_vpi(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 		scsi_host_put(shost);
 }
 
-void
+int
 lpfc_mbx_unreg_vpi(struct lpfc_vport *vport)
 {
 	struct lpfc_hba  *phba = vport->phba;
@@ -1186,7 +1355,7 @@ lpfc_mbx_unreg_vpi(struct lpfc_vport *vport)
 
 	mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
 	if (!mbox)
-		return;
+		return 1;
 
 	lpfc_unreg_vpi(phba, vport->vpi, mbox);
 	mbox->vport = vport;
@@ -1197,7 +1366,9 @@ lpfc_mbx_unreg_vpi(struct lpfc_vport *vport)
 				 "1800 Could not issue unreg_vpi\n");
 		mempool_free(mbox, phba->mbox_mem_pool);
 		vport->unreg_vpi_cmpl = VPORT_ERROR;
+		return rc;
 	}
+	return 0;
 }
 
 static void
@@ -1553,6 +1724,22 @@ lpfc_nlp_state_cleanup(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		 */
 		lpfc_register_remote_port(vport, ndlp);
 	}
+	if ((new_state ==  NLP_STE_MAPPED_NODE) &&
+		(vport->stat_data_enabled)) {
+		/*
+		 * A new target is discovered, if there is no buffer for
+		 * statistical data collection allocate buffer.
+		 */
+		ndlp->lat_data = kcalloc(LPFC_MAX_BUCKET_COUNT,
+					 sizeof(struct lpfc_scsicmd_bkt),
+					 GFP_KERNEL);
+
+		if (!ndlp->lat_data)
+			lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE,
+				"0286 lpfc_nlp_state_cleanup failed to "
+				"allocate statistical data buffer DID "
+				"0x%x\n", ndlp->nlp_DID);
+	}
 	/*
 	 * if we added to Mapped list, but the remote port
 	 * registration failed or assigned a target id outside
@@ -2786,7 +2973,7 @@ restart_disc:
 
 	default:
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_DISCOVERY,
-				 "0229 Unexpected discovery timeout, "
+				 "0273 Unexpected discovery timeout, "
 				 "vport State x%x\n", vport->port_state);
 		break;
 	}
@@ -2940,6 +3127,8 @@ lpfc_nlp_init(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 	INIT_LIST_HEAD(&ndlp->nlp_listp);
 	kref_init(&ndlp->kref);
 	NLP_INT_NODE_ACT(ndlp);
+	atomic_set(&ndlp->cmd_pending, 0);
+	ndlp->cmd_qdepth = LPFC_MAX_TGT_QDEPTH;
 
 	lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_NODE,
 		"node init:       did:x%x",
@@ -2979,8 +3168,10 @@ lpfc_nlp_release(struct kref *kref)
 	spin_unlock_irqrestore(&phba->ndlp_lock, flags);
 
 	/* free ndlp memory for final ndlp release */
-	if (NLP_CHK_FREE_REQ(ndlp))
+	if (NLP_CHK_FREE_REQ(ndlp)) {
+		kfree(ndlp->lat_data);
 		mempool_free(ndlp, ndlp->vport->phba->nlp_mem_pool);
+	}
 }
 
 /* This routine bumps the reference count for a ndlp structure to ensure

drivers/scsi/lpfc/lpfc_hw.h

@@ -1107,6 +1107,8 @@ typedef struct {
 /* Start FireFly Register definitions */
 #define PCI_VENDOR_ID_EMULEX        0x10df
 #define PCI_DEVICE_ID_FIREFLY       0x1ae5
+#define PCI_DEVICE_ID_PROTEUS_VF    0xe100
+#define PCI_DEVICE_ID_PROTEUS_PF    0xe180
 #define PCI_DEVICE_ID_SAT_SMB       0xf011
 #define PCI_DEVICE_ID_SAT_MID       0xf015
 #define PCI_DEVICE_ID_RFLY          0xf095
@@ -1133,10 +1135,12 @@ typedef struct {
 #define PCI_DEVICE_ID_LP11000S      0xfc10
 #define PCI_DEVICE_ID_LPE11000S     0xfc20
 #define PCI_DEVICE_ID_SAT_S         0xfc40
+#define PCI_DEVICE_ID_PROTEUS_S     0xfc50
 #define PCI_DEVICE_ID_HELIOS        0xfd00
 #define PCI_DEVICE_ID_HELIOS_SCSP   0xfd11
 #define PCI_DEVICE_ID_HELIOS_DCSP   0xfd12
 #define PCI_DEVICE_ID_ZEPHYR        0xfe00
+#define PCI_DEVICE_ID_HORNET        0xfe05
 #define PCI_DEVICE_ID_ZEPHYR_SCSP   0xfe11
 #define PCI_DEVICE_ID_ZEPHYR_DCSP   0xfe12
 
@@ -1154,6 +1158,7 @@ typedef struct {
 #define ZEPHYR_JEDEC_ID             0x0577
 #define VIPER_JEDEC_ID              0x4838
 #define SATURN_JEDEC_ID             0x1004
+#define HORNET_JDEC_ID              0x2057706D
 
 #define JEDEC_ID_MASK               0x0FFFF000
 #define JEDEC_ID_SHIFT              12
@@ -1198,6 +1203,18 @@ typedef struct {		/* FireFly BIU registers */
 #define HA_RXATT       0x00000008	/* Bit  3 */
 #define HA_RXMASK      0x0000000f
 
+#define HA_R0_CLR_MSK	(HA_R0RE_REQ | HA_R0CE_RSP | HA_R0ATT)
+#define HA_R1_CLR_MSK	(HA_R1RE_REQ | HA_R1CE_RSP | HA_R1ATT)
+#define HA_R2_CLR_MSK	(HA_R2RE_REQ | HA_R2CE_RSP | HA_R2ATT)
+#define HA_R3_CLR_MSK	(HA_R3RE_REQ | HA_R3CE_RSP | HA_R3ATT)
+
+#define HA_R0_POS	3
+#define HA_R1_POS	7
+#define HA_R2_POS	11
+#define HA_R3_POS	15
+#define HA_LE_POS	29
+#define HA_MB_POS	30
+#define HA_ER_POS	31
 /* Chip Attention Register */
 
 #define CA_REG_OFFSET  4	/* Byte offset from register base address */
@@ -1235,7 +1252,7 @@ typedef struct {		/* FireFly BIU registers */
 
 /* Host Control Register */
 
-#define HC_REG_OFFSET  12	/* Word offset from register base address */
+#define HC_REG_OFFSET  12	/* Byte offset from register base address */
 
 #define HC_MBINT_ENA   0x00000001	/* Bit  0 */
 #define HC_R0INT_ENA   0x00000002	/* Bit  1 */
@@ -1248,6 +1265,19 @@ typedef struct {		/* FireFly BIU registers */
 #define HC_LAINT_ENA   0x20000000	/* Bit 29 */
 #define HC_ERINT_ENA   0x80000000	/* Bit 31 */
 
+/* Message Signaled Interrupt eXtension (MSI-X) message identifiers */
+#define MSIX_DFLT_ID	0
+#define MSIX_RNG0_ID	0
+#define MSIX_RNG1_ID	1
+#define MSIX_RNG2_ID	2
+#define MSIX_RNG3_ID	3
+
+#define MSIX_LINK_ID	4
+#define MSIX_MBOX_ID	5
+
+#define MSIX_SPARE0_ID	6
+#define MSIX_SPARE1_ID	7
+
 /* Mailbox Commands */
 #define MBX_SHUTDOWN        0x00	/* terminate testing */
 #define MBX_LOAD_SM         0x01
@@ -1285,10 +1315,14 @@ typedef struct {		/* FireFly BIU registers */
 #define MBX_KILL_BOARD      0x24
 #define MBX_CONFIG_FARP     0x25
 #define MBX_BEACON          0x2A
+#define MBX_CONFIG_MSI      0x30
 #define MBX_HEARTBEAT       0x31
 #define MBX_WRITE_VPARMS    0x32
 #define MBX_ASYNCEVT_ENABLE 0x33
 
+#define MBX_PORT_CAPABILITIES 0x3B
+#define MBX_PORT_IOV_CONTROL 0x3C
+
 #define MBX_CONFIG_HBQ	    0x7C
 #define MBX_LOAD_AREA       0x81
 #define MBX_RUN_BIU_DIAG64  0x84
@@ -1474,24 +1508,18 @@ struct ulp_bde64 {	/* SLI-2 */
 			uint32_t bdeFlags:8;	/* BDE Flags 0 IS A SUPPORTED
 						   VALUE !! */
 #endif
-
-#define BUFF_USE_RSVD       0x01	/* bdeFlags */
-#define BUFF_USE_INTRPT     0x02	/* Not Implemented with LP6000 */
-#define BUFF_USE_CMND       0x04	/* Optional, 1=cmd/rsp 0=data buffer */
-#define BUFF_USE_RCV        0x08	/*  "" "", 1=rcv buffer, 0=xmit
-					    buffer */
-#define BUFF_TYPE_32BIT     0x10	/*  "" "", 1=32 bit addr 0=64 bit
-					    addr */
-#define BUFF_TYPE_SPECIAL   0x20	/* Not Implemented with LP6000  */
-#define BUFF_TYPE_BDL       0x40	/* Optional,  may be set in BDL */
-#define BUFF_TYPE_INVALID   0x80	/*  ""  "" */
+#define BUFF_TYPE_BDE_64    0x00	/* BDE (Host_resident) */
+#define BUFF_TYPE_BDE_IMMED 0x01	/* Immediate Data BDE */
+#define BUFF_TYPE_BDE_64P   0x02	/* BDE (Port-resident) */
+#define BUFF_TYPE_BDE_64I   0x08	/* Input BDE (Host-resident) */
+#define BUFF_TYPE_BDE_64IP  0x0A	/* Input BDE (Port-resident) */
+#define BUFF_TYPE_BLP_64    0x40	/* BLP (Host-resident) */
+#define BUFF_TYPE_BLP_64P   0x42	/* BLP (Port-resident) */
 		} f;
 	} tus;
 	uint32_t addrLow;
 	uint32_t addrHigh;
 };
-#define BDE64_SIZE_WORD 0
-#define BPL64_SIZE_WORD 0x40
 
 typedef struct ULP_BDL {	/* SLI-2 */
 #ifdef __BIG_ENDIAN_BITFIELD
@@ -2201,7 +2229,10 @@ typedef struct {
 typedef struct {
 	uint32_t eventTag;	/* Event tag */
 #ifdef __BIG_ENDIAN_BITFIELD
-	uint32_t rsvd1:22;
+	uint32_t rsvd1:19;
+	uint32_t fa:1;
+	uint32_t mm:1;		/* Menlo Maintenance mode enabled */
+	uint32_t rx:1;
 	uint32_t pb:1;
 	uint32_t il:1;
 	uint32_t attType:8;
@@ -2209,7 +2240,10 @@ typedef struct {
 	uint32_t attType:8;
 	uint32_t il:1;
 	uint32_t pb:1;
-	uint32_t rsvd1:22;
+	uint32_t rx:1;
+	uint32_t mm:1;
+	uint32_t fa:1;
+	uint32_t rsvd1:19;
 #endif
 
 #define AT_RESERVED    0x00	/* Reserved - attType */
@@ -2230,6 +2264,7 @@ typedef struct {
 
 #define TOPOLOGY_PT_PT 0x01	/* Topology is pt-pt / pt-fabric */
 #define TOPOLOGY_LOOP  0x02	/* Topology is FC-AL */
+#define TOPOLOGY_LNK_MENLO_MAINTENANCE 0x05 /* maint mode zephtr to menlo */
 
 	union {
 		struct ulp_bde lilpBde; /* This BDE points to a 128 byte buffer
@@ -2324,6 +2359,36 @@ typedef struct {
 #define  DMP_RSP_OFFSET          0x14   /* word 5 contains first word of rsp */
 #define  DMP_RSP_SIZE            0x6C   /* maximum of 27 words of rsp data */
 
+/* Structure for MB Command UPDATE_CFG (0x1B) */
+
+struct update_cfg_var {
+#ifdef __BIG_ENDIAN_BITFIELD
+	uint32_t rsvd2:16;
+	uint32_t type:8;
+	uint32_t rsvd:1;
+	uint32_t ra:1;
+	uint32_t co:1;
+	uint32_t cv:1;
+	uint32_t req:4;
+	uint32_t entry_length:16;
+	uint32_t region_id:16;
+#else  /*  __LITTLE_ENDIAN_BITFIELD */
+	uint32_t req:4;
+	uint32_t cv:1;
+	uint32_t co:1;
+	uint32_t ra:1;
+	uint32_t rsvd:1;
+	uint32_t type:8;
+	uint32_t rsvd2:16;
+	uint32_t region_id:16;
+	uint32_t entry_length:16;
+#endif
+
+	uint32_t resp_info;
+	uint32_t byte_cnt;
+	uint32_t data_offset;
+};
+
 struct hbq_mask {
 #ifdef __BIG_ENDIAN_BITFIELD
 	uint8_t tmatch;
@@ -2560,6 +2625,40 @@ typedef struct {
 
 } CONFIG_PORT_VAR;
 
+/* Structure for MB Command CONFIG_MSI (0x30) */
+struct config_msi_var {
+#ifdef __BIG_ENDIAN_BITFIELD
+	uint32_t dfltMsgNum:8;	/* Default message number            */
+	uint32_t rsvd1:11;	/* Reserved                          */
+	uint32_t NID:5;		/* Number of secondary attention IDs */
+	uint32_t rsvd2:5;	/* Reserved                          */
+	uint32_t dfltPresent:1;	/* Default message number present    */
+	uint32_t addFlag:1;	/* Add association flag              */
+	uint32_t reportFlag:1;	/* Report association flag           */
+#else	/*  __LITTLE_ENDIAN_BITFIELD */
+	uint32_t reportFlag:1;	/* Report association flag           */
+	uint32_t addFlag:1;	/* Add association flag              */
+	uint32_t dfltPresent:1;	/* Default message number present    */
+	uint32_t rsvd2:5;	/* Reserved                          */
+	uint32_t NID:5;		/* Number of secondary attention IDs */
+	uint32_t rsvd1:11;	/* Reserved                          */
+	uint32_t dfltMsgNum:8;	/* Default message number            */
+#endif
+	uint32_t attentionConditions[2];
+	uint8_t  attentionId[16];
+	uint8_t  messageNumberByHA[64];
+	uint8_t  messageNumberByID[16];
+	uint32_t autoClearHA[2];
+#ifdef __BIG_ENDIAN_BITFIELD
+	uint32_t rsvd3:16;
+	uint32_t autoClearID:16;
+#else	/*  __LITTLE_ENDIAN_BITFIELD */
+	uint32_t autoClearID:16;
+	uint32_t rsvd3:16;
+#endif
+	uint32_t rsvd4;
+};
+
 /* SLI-2 Port Control Block */
 
 /* SLIM POINTER */
@@ -2678,10 +2777,12 @@ typedef union {
 					 * NEW_FEATURE
 					 */
 	struct config_hbq_var varCfgHbq;/* cmd = 0x7c (CONFIG_HBQ)  */
+	struct update_cfg_var varUpdateCfg; /* cmd = 0x1B (UPDATE_CFG)*/
 	CONFIG_PORT_VAR varCfgPort;	/* cmd = 0x88 (CONFIG_PORT)  */
 	REG_VPI_VAR varRegVpi;		/* cmd = 0x96 (REG_VPI) */
 	UNREG_VPI_VAR varUnregVpi;	/* cmd = 0x97 (UNREG_VPI) */
 	ASYNCEVT_ENABLE_VAR varCfgAsyncEvent; /*cmd = x33 (CONFIG_ASYNC) */
+	struct config_msi_var varCfgMSI;/* cmd = x30 (CONFIG_MSI)     */
 } MAILVARIANTS;
 
 /*
@@ -2715,11 +2816,19 @@ struct sli3_pgp {
 	uint32_t hbq_get[16];
 };
 
-typedef union {
-	struct sli2_desc s2;
-	struct sli3_desc s3;
-	struct sli3_pgp  s3_pgp;
-} SLI_VAR;
+struct sli3_inb_pgp {
+	uint32_t ha_copy;
+	uint32_t counter;
+	struct lpfc_pgp port[MAX_RINGS];
+	uint32_t hbq_get[16];
+};
+
+union sli_var {
+	struct sli2_desc	s2;
+	struct sli3_desc	s3;
+	struct sli3_pgp		s3_pgp;
+	struct sli3_inb_pgp	s3_inb_pgp;
+};
 
 typedef struct {
 #ifdef __BIG_ENDIAN_BITFIELD
@@ -2737,7 +2846,7 @@ typedef struct {
 #endif
 
 	MAILVARIANTS un;
-	SLI_VAR us;
+	union sli_var us;
 } MAILBOX_t;
 
 /*
@@ -3105,6 +3214,27 @@ struct que_xri64cx_ext_fields {
 	struct lpfc_hbq_entry	buff[5];
 };
 
+#define LPFC_EXT_DATA_BDE_COUNT 3
+struct fcp_irw_ext {
+	uint32_t	io_tag64_low;
+	uint32_t	io_tag64_high;
+#ifdef __BIG_ENDIAN_BITFIELD
+	uint8_t		reserved1;
+	uint8_t		reserved2;
+	uint8_t		reserved3;
+	uint8_t		ebde_count;
+#else  /* __LITTLE_ENDIAN */
+	uint8_t		ebde_count;
+	uint8_t		reserved3;
+	uint8_t		reserved2;
+	uint8_t		reserved1;
+#endif
+	uint32_t	reserved4;
+	struct ulp_bde64 rbde;		/* response bde */
+	struct ulp_bde64 dbde[LPFC_EXT_DATA_BDE_COUNT];	/* data BDE or BPL */
+	uint8_t icd[32];		/* immediate command data (32 bytes) */
+};
+
 typedef struct _IOCB {	/* IOCB structure */
 	union {
 		GENERIC_RSP grsp;	/* Generic response */
@@ -3190,7 +3320,7 @@ typedef struct _IOCB {	/* IOCB structure */
 
 		/* words 8-31 used for que_xri_cx iocb */
 		struct que_xri64cx_ext_fields que_xri64cx_ext_words;
-
+		struct fcp_irw_ext fcp_ext;
 		uint32_t sli3Words[24]; /* 96 extra bytes for SLI-3 */
 	} unsli3;
 
@@ -3292,3 +3422,10 @@ lpfc_error_lost_link(IOCB_t *iocbp)
 		 iocbp->un.ulpWord[4] == IOERR_LINK_DOWN ||
 		 iocbp->un.ulpWord[4] == IOERR_SLI_DOWN));
 }
+
+#define MENLO_TRANSPORT_TYPE 0xfe
+#define MENLO_CONTEXT 0
+#define MENLO_PU 3
+#define MENLO_TIMEOUT 30
+#define SETVAR_MLOMNT 0x103107
+#define SETVAR_MLORST 0x103007

drivers/scsi/lpfc/lpfc_mbox.c

@@ -1,7 +1,7 @@
 /*******************************************************************
  * This file is part of the Emulex Linux Device Driver for         *
  * Fibre Channel Host Bus Adapters.                                *
- * Copyright (C) 2004-2007 Emulex.  All rights reserved.           *
+ * Copyright (C) 2004-2008 Emulex.  All rights reserved.           *
  * EMULEX and SLI are trademarks of Emulex.                        *
  * www.emulex.com                                                  *
  * Portions Copyright (C) 2004-2005 Christoph Hellwig              *
@@ -30,6 +30,7 @@
 
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -37,10 +38,20 @@
 #include "lpfc_crtn.h"
 #include "lpfc_compat.h"
 
-/**********************************************/
-
-/*                mailbox command             */
-/**********************************************/
+/**
+ * lpfc_dump_mem: Prepare a mailbox command for retrieving HBA's VPD memory.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @offset: offset for dumping VPD memory mailbox command.
+ *
+ * The dump mailbox command provides a method for the device driver to obtain
+ * various types of information from the HBA device.
+ *
+ * This routine prepares the mailbox command for dumping HBA Vital Product
+ * Data (VPD) memory. This mailbox command is to be used for retrieving a
+ * portion (DMP_RSP_SIZE bytes) of a HBA's VPD from the HBA at an address
+ * offset specified by the offset parameter.
+ **/
 void
 lpfc_dump_mem(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb, uint16_t offset)
 {
@@ -65,10 +76,17 @@ lpfc_dump_mem(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb, uint16_t offset)
 	return;
 }
 
-/**********************************************/
-/*  lpfc_read_nv  Issue a READ NVPARAM        */
-/*                mailbox command             */
-/**********************************************/
+/**
+ * lpfc_read_nv: Prepare a mailbox command for reading HBA's NVRAM param.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The read NVRAM mailbox command returns the HBA's non-volatile parameters
+ * that are used as defaults when the Fibre Channel link is brought on-line.
+ *
+ * This routine prepares the mailbox command for reading information stored
+ * in the HBA's NVRAM. Specifically, the HBA's WWNN and WWPN.
+ **/
 void
 lpfc_read_nv(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -81,10 +99,19 @@ lpfc_read_nv(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/**********************************************/
-/*  lpfc_config_async  Issue a                */
-/*  MBX_ASYNC_EVT_ENABLE mailbox command      */
-/**********************************************/
+/**
+ * lpfc_config_async: Prepare a mailbox command for enabling HBA async event.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @ring: ring number for the asynchronous event to be configured.
+ *
+ * The asynchronous event enable mailbox command is used to enable the
+ * asynchronous event posting via the ASYNC_STATUS_CN IOCB response and
+ * specifies the default ring to which events are posted.
+ *
+ * This routine prepares the mailbox command for enabling HBA asynchronous
+ * event support on a IOCB ring.
+ **/
 void
 lpfc_config_async(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb,
 		uint32_t ring)
@@ -99,10 +126,19 @@ lpfc_config_async(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb,
 	return;
 }
 
-/**********************************************/
-/*  lpfc_heart_beat  Issue a HEART_BEAT       */
-/*                mailbox command             */
-/**********************************************/
+/**
+ * lpfc_heart_beat: Prepare a mailbox command for heart beat.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The heart beat mailbox command is used to detect an unresponsive HBA, which
+ * is defined as any device where no error attention is sent and both mailbox
+ * and rings are not processed.
+ *
+ * This routine prepares the mailbox command for issuing a heart beat in the
+ * form of mailbox command to the HBA. The timely completion of the heart
+ * beat mailbox command indicates the health of the HBA.
+ **/
 void
 lpfc_heart_beat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -115,10 +151,26 @@ lpfc_heart_beat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/**********************************************/
-/*  lpfc_read_la  Issue a READ LA             */
-/*                mailbox command             */
-/**********************************************/
+/**
+ * lpfc_read_la: Prepare a mailbox command for reading HBA link attention.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @mp: DMA buffer memory for reading the link attention information into.
+ *
+ * The read link attention mailbox command is issued to read the Link Event
+ * Attention information indicated by the HBA port when the Link Event bit
+ * of the Host Attention (HSTATT) register is set to 1. A Link Event
+ * Attention occurs based on an exception detected at the Fibre Channel link
+ * interface.
+ *
+ * This routine prepares the mailbox command for reading HBA link attention
+ * information. A DMA memory has been set aside and address passed to the
+ * HBA through @mp for the HBA to DMA link attention information into the
+ * memory as part of the execution of the mailbox command.
+ *
+ * Return codes
+ *    0 - Success (currently always return 0)
+ **/
 int
 lpfc_read_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb, struct lpfc_dmabuf *mp)
 {
@@ -143,10 +195,21 @@ lpfc_read_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb, struct lpfc_dmabuf *mp)
 	return (0);
 }
 
-/**********************************************/
-/*  lpfc_clear_la  Issue a CLEAR LA           */
-/*                 mailbox command            */
-/**********************************************/
+/**
+ * lpfc_clear_la: Prepare a mailbox command for clearing HBA link attention.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The clear link attention mailbox command is issued to clear the link event
+ * attention condition indicated by the Link Event bit of the Host Attention
+ * (HSTATT) register. The link event attention condition is cleared only if
+ * the event tag specified matches that of the current link event counter.
+ * The current event tag is read using the read link attention event mailbox
+ * command.
+ *
+ * This routine prepares the mailbox command for clearing HBA link attention
+ * information.
+ **/
 void
 lpfc_clear_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -161,10 +224,20 @@ lpfc_clear_la(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/**************************************************/
-/*  lpfc_config_link  Issue a CONFIG LINK         */
-/*                    mailbox command             */
-/**************************************************/
+/**
+ * lpfc_config_link: Prepare a mailbox command for configuring link on a HBA.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The configure link mailbox command is used before the initialize link
+ * mailbox command to override default value and to configure link-oriented
+ * parameters such as DID address and various timers. Typically, this
+ * command would be used after an F_Port login to set the returned DID address
+ * and the fabric timeout values. This command is not valid before a configure
+ * port command has configured the HBA port.
+ *
+ * This routine prepares the mailbox command for configuring link on a HBA.
+ **/
 void
 lpfc_config_link(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -199,10 +272,98 @@ lpfc_config_link(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/**********************************************/
-/*  lpfc_init_link  Issue an INIT LINK        */
-/*                  mailbox command           */
-/**********************************************/
+/**
+ * lpfc_config_msi: Prepare a mailbox command for configuring msi-x.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The configure MSI-X mailbox command is used to configure the HBA's SLI-3
+ * MSI-X multi-message interrupt vector association to interrupt attention
+ * conditions.
+ *
+ * Return codes
+ *    0 - Success
+ *    -EINVAL - Failure
+ **/
+int
+lpfc_config_msi(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
+{
+	MAILBOX_t *mb = &pmb->mb;
+	uint32_t attentionConditions[2];
+
+	/* Sanity check */
+	if (phba->cfg_use_msi != 2) {
+		lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+				"0475 Not configured for supporting MSI-X "
+				"cfg_use_msi: 0x%x\n", phba->cfg_use_msi);
+		return -EINVAL;
+	}
+
+	if (phba->sli_rev < 3) {
+		lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+				"0476 HBA not supporting SLI-3 or later "
+				"SLI Revision: 0x%x\n", phba->sli_rev);
+		return -EINVAL;
+	}
+
+	/* Clear mailbox command fields */
+	memset(pmb, 0, sizeof(LPFC_MBOXQ_t));
+
+	/*
+	 * SLI-3, Message Signaled Interrupt Fearure.
+	 */
+
+	/* Multi-message attention configuration */
+	attentionConditions[0] = (HA_R0ATT | HA_R1ATT | HA_R2ATT | HA_ERATT |
+				  HA_LATT | HA_MBATT);
+	attentionConditions[1] = 0;
+
+	mb->un.varCfgMSI.attentionConditions[0] = attentionConditions[0];
+	mb->un.varCfgMSI.attentionConditions[1] = attentionConditions[1];
+
+	/*
+	 * Set up message number to HA bit association
+	 */
+#ifdef __BIG_ENDIAN_BITFIELD
+	/* RA0 (FCP Ring) */
+	mb->un.varCfgMSI.messageNumberByHA[HA_R0_POS] = 1;
+	/* RA1 (Other Protocol Extra Ring) */
+	mb->un.varCfgMSI.messageNumberByHA[HA_R1_POS] = 1;
+#else   /*  __LITTLE_ENDIAN_BITFIELD */
+	/* RA0 (FCP Ring) */
+	mb->un.varCfgMSI.messageNumberByHA[HA_R0_POS^3] = 1;
+	/* RA1 (Other Protocol Extra Ring) */
+	mb->un.varCfgMSI.messageNumberByHA[HA_R1_POS^3] = 1;
+#endif
+	/* Multi-message interrupt autoclear configuration*/
+	mb->un.varCfgMSI.autoClearHA[0] = attentionConditions[0];
+	mb->un.varCfgMSI.autoClearHA[1] = attentionConditions[1];
+
+	/* For now, HBA autoclear does not work reliably, disable it */
+	mb->un.varCfgMSI.autoClearHA[0] = 0;
+	mb->un.varCfgMSI.autoClearHA[1] = 0;
+
+	/* Set command and owner bit */
+	mb->mbxCommand = MBX_CONFIG_MSI;
+	mb->mbxOwner = OWN_HOST;
+
+	return 0;
+}
+
+/**
+ * lpfc_init_link: Prepare a mailbox command for initialize link on a HBA.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @topology: the link topology for the link to be initialized to.
+ * @linkspeed: the link speed for the link to be initialized to.
+ *
+ * The initialize link mailbox command is used to initialize the Fibre
+ * Channel link. This command must follow a configure port command that
+ * establishes the mode of operation.
+ *
+ * This routine prepares the mailbox command for initializing link on a HBA
+ * with the specified link topology and speed.
+ **/
 void
 lpfc_init_link(struct lpfc_hba * phba,
 	       LPFC_MBOXQ_t * pmb, uint32_t topology, uint32_t linkspeed)
@@ -269,10 +430,27 @@ lpfc_init_link(struct lpfc_hba * phba,
 	return;
 }
 
-/**********************************************/
-/*  lpfc_read_sparam  Issue a READ SPARAM     */
-/*                    mailbox command         */
-/**********************************************/
+/**
+ * lpfc_read_sparam: Prepare a mailbox command for reading HBA parameters.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @vpi: virtual N_Port identifier.
+ *
+ * The read service parameter mailbox command is used to read the HBA port
+ * service parameters. The service parameters are read into the buffer
+ * specified directly by a BDE in the mailbox command. These service
+ * parameters may then be used to build the payload of an N_Port/F_POrt
+ * login request and reply (LOGI/ACC).
+ *
+ * This routine prepares the mailbox command for reading HBA port service
+ * parameters. The DMA memory is allocated in this function and the addresses
+ * are populated into the mailbox command for the HBA to DMA the service
+ * parameters into.
+ *
+ * Return codes
+ *    0 - Success
+ *    1 - DMA memory allocation failed
+ **/
 int
 lpfc_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb, int vpi)
 {
@@ -312,10 +490,21 @@ lpfc_read_sparam(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb, int vpi)
 	return (0);
 }
 
-/********************************************/
-/*  lpfc_unreg_did  Issue a UNREG_DID       */
-/*                  mailbox command         */
-/********************************************/
+/**
+ * lpfc_unreg_did: Prepare a mailbox command for unregistering DID.
+ * @phba: pointer to lpfc hba data structure.
+ * @vpi: virtual N_Port identifier.
+ * @did: remote port identifier.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The unregister DID mailbox command is used to unregister an N_Port/F_Port
+ * login for an unknown RPI by specifying the DID of a remote port. This
+ * command frees an RPI context in the HBA port. This has the effect of
+ * performing an implicit N_Port/F_Port logout.
+ *
+ * This routine prepares the mailbox command for unregistering a remote
+ * N_Port/F_Port (DID) login.
+ **/
 void
 lpfc_unreg_did(struct lpfc_hba * phba, uint16_t vpi, uint32_t did,
 	       LPFC_MBOXQ_t * pmb)
@@ -333,10 +522,19 @@ lpfc_unreg_did(struct lpfc_hba * phba, uint16_t vpi, uint32_t did,
 	return;
 }
 
-/**********************************************/
-/*  lpfc_read_nv  Issue a READ CONFIG         */
-/*                mailbox command             */
-/**********************************************/
+/**
+ * lpfc_read_config: Prepare a mailbox command for reading HBA configuration.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The read configuration mailbox command is used to read the HBA port
+ * configuration parameters. This mailbox command provides a method for
+ * seeing any parameters that may have changed via various configuration
+ * mailbox commands.
+ *
+ * This routine prepares the mailbox command for reading out HBA configuration
+ * parameters.
+ **/
 void
 lpfc_read_config(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -350,10 +548,18 @@ lpfc_read_config(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/*************************************************/
-/*  lpfc_read_lnk_stat  Issue a READ LINK STATUS */
-/*                mailbox command                */
-/*************************************************/
+/**
+ * lpfc_read_lnk_stat: Prepare a mailbox command for reading HBA link stats.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The read link status mailbox command is used to read the link status from
+ * the HBA. Link status includes all link-related error counters. These
+ * counters are maintained by the HBA and originated in the link hardware
+ * unit. Note that all of these counters wrap.
+ *
+ * This routine prepares the mailbox command for reading out HBA link status.
+ **/
 void
 lpfc_read_lnk_stat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -367,10 +573,30 @@ lpfc_read_lnk_stat(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
-/********************************************/
-/*  lpfc_reg_login  Issue a REG_LOGIN       */
-/*                  mailbox command         */
-/********************************************/
+/**
+ * lpfc_reg_login: Prepare a mailbox command for registering remote login.
+ * @phba: pointer to lpfc hba data structure.
+ * @vpi: virtual N_Port identifier.
+ * @did: remote port identifier.
+ * @param: pointer to memory holding the server parameters.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ * @flag: action flag to be passed back for the complete function.
+ *
+ * The registration login mailbox command is used to register an N_Port or
+ * F_Port login. This registration allows the HBA to cache the remote N_Port
+ * service parameters internally and thereby make the appropriate FC-2
+ * decisions. The remote port service parameters are handed off by the driver
+ * to the HBA using a descriptor entry that directly identifies a buffer in
+ * host memory. In exchange, the HBA returns an RPI identifier.
+ *
+ * This routine prepares the mailbox command for registering remote port login.
+ * The function allocates DMA buffer for passing the service parameters to the
+ * HBA with the mailbox command.
+ *
+ * Return codes
+ *    0 - Success
+ *    1 - DMA memory allocation failed
+ **/
 int
 lpfc_reg_login(struct lpfc_hba *phba, uint16_t vpi, uint32_t did,
 	       uint8_t *param, LPFC_MBOXQ_t *pmb, uint32_t flag)
@@ -418,10 +644,20 @@ lpfc_reg_login(struct lpfc_hba *phba, uint16_t vpi, uint32_t did,
 	return (0);
 }
 
-/**********************************************/
-/*  lpfc_unreg_login  Issue a UNREG_LOGIN     */
-/*                    mailbox command         */
-/**********************************************/
+/**
+ * lpfc_unreg_login: Prepare a mailbox command for unregistering remote login.
+ * @phba: pointer to lpfc hba data structure.
+ * @vpi: virtual N_Port identifier.
+ * @rpi: remote port identifier
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The unregistration login mailbox command is used to unregister an N_Port
+ * or F_Port login. This command frees an RPI context in the HBA. It has the
+ * effect of performing an implicit N_Port/F_Port logout.
+ *
+ * This routine prepares the mailbox command for unregistering remote port
+ * login.
+ **/
 void
 lpfc_unreg_login(struct lpfc_hba *phba, uint16_t vpi, uint32_t rpi,
 		 LPFC_MBOXQ_t * pmb)
@@ -440,10 +676,21 @@ lpfc_unreg_login(struct lpfc_hba *phba, uint16_t vpi, uint32_t rpi,
 	return;
 }
 
-/**************************************************/
-/*  lpfc_reg_vpi   Issue a REG_VPI                */
-/*                    mailbox command             */
-/**************************************************/
+/**
+ * lpfc_reg_vpi: Prepare a mailbox command for registering vport identifier.
+ * @phba: pointer to lpfc hba data structure.
+ * @vpi: virtual N_Port identifier.
+ * @sid: Fibre Channel S_ID (N_Port_ID assigned to a virtual N_Port).
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The registration vport identifier mailbox command is used to activate a
+ * virtual N_Port after it has acquired an N_Port_ID. The HBA validates the
+ * N_Port_ID against the information in the selected virtual N_Port context
+ * block and marks it active to allow normal processing of IOCB commands and
+ * received unsolicited exchanges.
+ *
+ * This routine prepares the mailbox command for registering a virtual N_Port.
+ **/
 void
 lpfc_reg_vpi(struct lpfc_hba *phba, uint16_t vpi, uint32_t sid,
 	     LPFC_MBOXQ_t *pmb)
@@ -461,10 +708,22 @@ lpfc_reg_vpi(struct lpfc_hba *phba, uint16_t vpi, uint32_t sid,
 
 }
 
-/**************************************************/
-/*  lpfc_unreg_vpi   Issue a UNREG_VNPI           */
-/*                    mailbox command             */
-/**************************************************/
+/**
+ * lpfc_unreg_vpi: Prepare a mailbox command for unregistering vport id.
+ * @phba: pointer to lpfc hba data structure.
+ * @vpi: virtual N_Port identifier.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The unregistration vport identifier mailbox command is used to inactivate
+ * a virtual N_Port. The driver must have logged out and unregistered all
+ * remote N_Ports to abort any activity on the virtual N_Port. The HBA will
+ * unregisters any default RPIs associated with the specified vpi, aborting
+ * any active exchanges. The HBA will post the mailbox response after making
+ * the virtual N_Port inactive.
+ *
+ * This routine prepares the mailbox command for unregistering a virtual
+ * N_Port.
+ **/
 void
 lpfc_unreg_vpi(struct lpfc_hba *phba, uint16_t vpi, LPFC_MBOXQ_t *pmb)
 {
@@ -479,12 +738,19 @@ lpfc_unreg_vpi(struct lpfc_hba *phba, uint16_t vpi, LPFC_MBOXQ_t *pmb)
 
 }
 
+/**
+ * lpfc_config_pcb_setup: Set up IOCB rings in the Port Control Block (PCB)
+ * @phba: pointer to lpfc hba data structure.
+ *
+ * This routine sets up and initializes the IOCB rings in the Port Control
+ * Block (PCB).
+ **/
 static void
 lpfc_config_pcb_setup(struct lpfc_hba * phba)
 {
 	struct lpfc_sli *psli = &phba->sli;
 	struct lpfc_sli_ring *pring;
-	PCB_t *pcbp = &phba->slim2p->pcb;
+	PCB_t *pcbp = phba->pcb;
 	dma_addr_t pdma_addr;
 	uint32_t offset;
 	uint32_t iocbCnt = 0;
@@ -513,29 +779,43 @@ lpfc_config_pcb_setup(struct lpfc_hba * phba)
 			continue;
 		}
 		/* Command ring setup for ring */
-		pring->cmdringaddr = (void *) &phba->slim2p->IOCBs[iocbCnt];
+		pring->cmdringaddr = (void *)&phba->IOCBs[iocbCnt];
 		pcbp->rdsc[i].cmdEntries = pring->numCiocb;
 
-		offset = (uint8_t *) &phba->slim2p->IOCBs[iocbCnt] -
-			 (uint8_t *) phba->slim2p;
-		pdma_addr = phba->slim2p_mapping + offset;
+		offset = (uint8_t *) &phba->IOCBs[iocbCnt] -
+			 (uint8_t *) phba->slim2p.virt;
+		pdma_addr = phba->slim2p.phys + offset;
 		pcbp->rdsc[i].cmdAddrHigh = putPaddrHigh(pdma_addr);
 		pcbp->rdsc[i].cmdAddrLow = putPaddrLow(pdma_addr);
 		iocbCnt += pring->numCiocb;
 
 		/* Response ring setup for ring */
-		pring->rspringaddr = (void *) &phba->slim2p->IOCBs[iocbCnt];
+		pring->rspringaddr = (void *) &phba->IOCBs[iocbCnt];
 
 		pcbp->rdsc[i].rspEntries = pring->numRiocb;
-		offset = (uint8_t *)&phba->slim2p->IOCBs[iocbCnt] -
-			 (uint8_t *)phba->slim2p;
-		pdma_addr = phba->slim2p_mapping + offset;
+		offset = (uint8_t *)&phba->IOCBs[iocbCnt] -
+			 (uint8_t *)phba->slim2p.virt;
+		pdma_addr = phba->slim2p.phys + offset;
 		pcbp->rdsc[i].rspAddrHigh = putPaddrHigh(pdma_addr);
 		pcbp->rdsc[i].rspAddrLow = putPaddrLow(pdma_addr);
 		iocbCnt += pring->numRiocb;
 	}
 }
 
+/**
+ * lpfc_read_rev: Prepare a mailbox command for reading HBA revision.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The read revision mailbox command is used to read the revision levels of
+ * the HBA components. These components include hardware units, resident
+ * firmware, and available firmware. HBAs that supports SLI-3 mode of
+ * operation provide different response information depending on the version
+ * requested by the driver.
+ *
+ * This routine prepares the mailbox command for reading HBA revision
+ * information.
+ **/
 void
 lpfc_read_rev(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -548,6 +828,16 @@ lpfc_read_rev(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
+/**
+ * lpfc_build_hbq_profile2: Set up the HBQ Selection Profile 2.
+ * @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
+ * @hbq_desc: pointer to the HBQ selection profile descriptor.
+ *
+ * The Host Buffer Queue (HBQ) Selection Profile 2 specifies that the HBA
+ * tests the incoming frames' R_CTL/TYPE fields with works 10:15 and performs
+ * the Sequence Length Test using the fields in the Selection Profile 2
+ * extension in words 20:31.
+ **/
 static void
 lpfc_build_hbq_profile2(struct config_hbq_var *hbqmb,
 			struct lpfc_hbq_init  *hbq_desc)
@@ -557,6 +847,16 @@ lpfc_build_hbq_profile2(struct config_hbq_var *hbqmb,
 	hbqmb->profiles.profile2.seqlenoff  = hbq_desc->seqlenoff;
 }
 
+/**
+ * lpfc_build_hbq_profile3: Set up the HBQ Selection Profile 3.
+ * @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
+ * @hbq_desc: pointer to the HBQ selection profile descriptor.
+ *
+ * The Host Buffer Queue (HBQ) Selection Profile 3 specifies that the HBA
+ * tests the incoming frame's R_CTL/TYPE fields with words 10:15 and performs
+ * the Sequence Length Test and Byte Field Test using the fields in the
+ * Selection Profile 3 extension in words 20:31.
+ **/
 static void
 lpfc_build_hbq_profile3(struct config_hbq_var *hbqmb,
 			struct lpfc_hbq_init  *hbq_desc)
@@ -569,6 +869,17 @@ lpfc_build_hbq_profile3(struct config_hbq_var *hbqmb,
 	       sizeof(hbqmb->profiles.profile3.cmdmatch));
 }
 
+/**
+ * lpfc_build_hbq_profile5: Set up the HBQ Selection Profile 5.
+ * @hbqmb: pointer to the HBQ configuration data structure in mailbox command.
+ * @hbq_desc: pointer to the HBQ selection profile descriptor.
+ *
+ * The Host Buffer Queue (HBQ) Selection Profile 5 specifies a header HBQ. The
+ * HBA tests the initial frame of an incoming sequence using the frame's
+ * R_CTL/TYPE fields with words 10:15 and performs the Sequence Length Test
+ * and Byte Field Test using the fields in the Selection Profile 5 extension
+ * words 20:31.
+ **/
 static void
 lpfc_build_hbq_profile5(struct config_hbq_var *hbqmb,
 			struct lpfc_hbq_init  *hbq_desc)
@@ -581,6 +892,20 @@ lpfc_build_hbq_profile5(struct config_hbq_var *hbqmb,
 	       sizeof(hbqmb->profiles.profile5.cmdmatch));
 }
 
+/**
+ * lpfc_config_hbq: Prepare a mailbox command for configuring an HBQ.
+ * @phba: pointer to lpfc hba data structure.
+ * @id: HBQ identifier.
+ * @hbq_desc: pointer to the HBA descriptor data structure.
+ * @hbq_entry_index: index of the HBQ entry data structures.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The configure HBQ (Host Buffer Queue) mailbox command is used to configure
+ * an HBQ. The configuration binds events that require buffers to a particular
+ * ring and HBQ based on a selection profile.
+ *
+ * This routine prepares the mailbox command for configuring an HBQ.
+ **/
 void
 lpfc_config_hbq(struct lpfc_hba *phba, uint32_t id,
 		 struct lpfc_hbq_init *hbq_desc,
@@ -641,8 +966,23 @@ lpfc_config_hbq(struct lpfc_hba *phba, uint32_t id,
 	return;
 }
 
-
-
+/**
+ * lpfc_config_ring: Prepare a mailbox command for configuring an IOCB ring.
+ * @phba: pointer to lpfc hba data structure.
+ * @ring:
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The configure ring mailbox command is used to configure an IOCB ring. This
+ * configuration binds from one to six of HBA RC_CTL/TYPE mask entries to the
+ * ring. This is used to map incoming sequences to a particular ring whose
+ * RC_CTL/TYPE mask entry matches that of the sequence. The driver should not
+ * attempt to configure a ring whose number is greater than the number
+ * specified in the Port Control Block (PCB). It is an error to issue the
+ * configure ring command more than once with the same ring number. The HBA
+ * returns an error if the driver attempts this.
+ *
+ * This routine prepares the mailbox command for configuring IOCB ring.
+ **/
 void
 lpfc_config_ring(struct lpfc_hba * phba, int ring, LPFC_MBOXQ_t * pmb)
 {
@@ -684,6 +1024,20 @@ lpfc_config_ring(struct lpfc_hba * phba, int ring, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
+/**
+ * lpfc_config_port: Prepare a mailbox command for configuring port.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The configure port mailbox command is used to identify the Port Control
+ * Block (PCB) in the driver memory. After this command is issued, the
+ * driver must not access the mailbox in the HBA without first resetting
+ * the HBA. The HBA may copy the PCB information to internal storage for
+ * subsequent use; the driver can not change the PCB information unless it
+ * resets the HBA.
+ *
+ * This routine prepares the mailbox command for configuring port.
+ **/
 void
 lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 {
@@ -702,8 +1056,8 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 
 	mb->un.varCfgPort.pcbLen = sizeof(PCB_t);
 
-	offset = (uint8_t *)&phba->slim2p->pcb - (uint8_t *)phba->slim2p;
-	pdma_addr = phba->slim2p_mapping + offset;
+	offset = (uint8_t *)phba->pcb - (uint8_t *)phba->slim2p.virt;
+	pdma_addr = phba->slim2p.phys + offset;
 	mb->un.varCfgPort.pcbLow = putPaddrLow(pdma_addr);
 	mb->un.varCfgPort.pcbHigh = putPaddrHigh(pdma_addr);
 
@@ -711,12 +1065,13 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 
 	if (phba->sli_rev == 3 && phba->vpd.sli3Feat.cerbm) {
 		mb->un.varCfgPort.cerbm = 1; /* Request HBQs */
+		mb->un.varCfgPort.ccrp = 1; /* Command Ring Polling */
+		mb->un.varCfgPort.cinb = 1; /* Interrupt Notification Block */
 		mb->un.varCfgPort.max_hbq = lpfc_sli_hbq_count();
 		if (phba->max_vpi && phba->cfg_enable_npiv &&
 		    phba->vpd.sli3Feat.cmv) {
 			mb->un.varCfgPort.max_vpi = phba->max_vpi;
 			mb->un.varCfgPort.cmv = 1;
-			phba->sli3_options |= LPFC_SLI3_NPIV_ENABLED;
 		} else
 			mb->un.varCfgPort.max_vpi = phba->max_vpi = 0;
 	} else
@@ -724,16 +1079,15 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	mb->un.varCfgPort.sli_mode = phba->sli_rev;
 
 	/* Now setup pcb */
-	phba->slim2p->pcb.type = TYPE_NATIVE_SLI2;
-	phba->slim2p->pcb.feature = FEATURE_INITIAL_SLI2;
+	phba->pcb->type = TYPE_NATIVE_SLI2;
+	phba->pcb->feature = FEATURE_INITIAL_SLI2;
 
 	/* Setup Mailbox pointers */
-	phba->slim2p->pcb.mailBoxSize = offsetof(MAILBOX_t, us) +
-		sizeof(struct sli2_desc);
-	offset = (uint8_t *)&phba->slim2p->mbx - (uint8_t *)phba->slim2p;
-	pdma_addr = phba->slim2p_mapping + offset;
-	phba->slim2p->pcb.mbAddrHigh = putPaddrHigh(pdma_addr);
-	phba->slim2p->pcb.mbAddrLow = putPaddrLow(pdma_addr);
+	phba->pcb->mailBoxSize = sizeof(MAILBOX_t);
+	offset = (uint8_t *)phba->mbox - (uint8_t *)phba->slim2p.virt;
+	pdma_addr = phba->slim2p.phys + offset;
+	phba->pcb->mbAddrHigh = putPaddrHigh(pdma_addr);
+	phba->pcb->mbAddrLow = putPaddrLow(pdma_addr);
 
 	/*
 	 * Setup Host Group ring pointer.
@@ -794,13 +1148,13 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	}
 
 	/* mask off BAR0's flag bits 0 - 3 */
-	phba->slim2p->pcb.hgpAddrLow = (bar_low & PCI_BASE_ADDRESS_MEM_MASK) +
-		(void __iomem *) phba->host_gp -
+	phba->pcb->hgpAddrLow = (bar_low & PCI_BASE_ADDRESS_MEM_MASK) +
+		(void __iomem *)phba->host_gp -
 		(void __iomem *)phba->MBslimaddr;
 	if (bar_low & PCI_BASE_ADDRESS_MEM_TYPE_64)
-		phba->slim2p->pcb.hgpAddrHigh = bar_high;
+		phba->pcb->hgpAddrHigh = bar_high;
 	else
-		phba->slim2p->pcb.hgpAddrHigh = 0;
+		phba->pcb->hgpAddrHigh = 0;
 	/* write HGP data to SLIM at the required longword offset */
 	memset(&hgp, 0, sizeof(struct lpfc_hgp));
 
@@ -810,17 +1164,19 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	}
 
 	/* Setup Port Group ring pointer */
-	if (phba->sli_rev == 3)
-		pgp_offset = (uint8_t *)&phba->slim2p->mbx.us.s3_pgp.port -
-			(uint8_t *)phba->slim2p;
-	else
-		pgp_offset = (uint8_t *)&phba->slim2p->mbx.us.s2.port -
-			(uint8_t *)phba->slim2p;
-
-	pdma_addr = phba->slim2p_mapping + pgp_offset;
-	phba->slim2p->pcb.pgpAddrHigh = putPaddrHigh(pdma_addr);
-	phba->slim2p->pcb.pgpAddrLow = putPaddrLow(pdma_addr);
-	phba->hbq_get = &phba->slim2p->mbx.us.s3_pgp.hbq_get[0];
+	if (phba->sli3_options & LPFC_SLI3_INB_ENABLED) {
+		pgp_offset = offsetof(struct lpfc_sli2_slim,
+				      mbx.us.s3_inb_pgp.port);
+		phba->hbq_get = phba->mbox->us.s3_inb_pgp.hbq_get;
+	} else if (phba->sli_rev == 3) {
+		pgp_offset = offsetof(struct lpfc_sli2_slim,
+				      mbx.us.s3_pgp.port);
+		phba->hbq_get = phba->mbox->us.s3_pgp.hbq_get;
+	} else
+		pgp_offset = offsetof(struct lpfc_sli2_slim, mbx.us.s2.port);
+	pdma_addr = phba->slim2p.phys + pgp_offset;
+	phba->pcb->pgpAddrHigh = putPaddrHigh(pdma_addr);
+	phba->pcb->pgpAddrLow = putPaddrLow(pdma_addr);
 
 	/* Use callback routine to setp rings in the pcb */
 	lpfc_config_pcb_setup(phba);
@@ -835,10 +1191,24 @@ lpfc_config_port(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	}
 
 	/* Swap PCB if needed */
-	lpfc_sli_pcimem_bcopy(&phba->slim2p->pcb, &phba->slim2p->pcb,
-			      sizeof(PCB_t));
+	lpfc_sli_pcimem_bcopy(phba->pcb, phba->pcb, sizeof(PCB_t));
 }
 
+/**
+ * lpfc_kill_board: Prepare a mailbox command for killing board.
+ * @phba: pointer to lpfc hba data structure.
+ * @pmb: pointer to the driver internal queue element for mailbox command.
+ *
+ * The kill board mailbox command is used to tell firmware to perform a
+ * graceful shutdown of a channel on a specified board to prepare for reset.
+ * When the kill board mailbox command is received, the ER3 bit is set to 1
+ * in the Host Status register and the ER Attention bit is set to 1 in the
+ * Host Attention register of the HBA function that received the kill board
+ * command.
+ *
+ * This routine prepares the mailbox command for killing the board in
+ * preparation for a graceful shutdown.
+ **/
 void
 lpfc_kill_board(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 {
@@ -850,6 +1220,16 @@ lpfc_kill_board(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmb)
 	return;
 }
 
+/**
+ * lpfc_mbox_put: Put a mailbox cmd into the tail of driver's mailbox queue.
+ * @phba: pointer to lpfc hba data structure.
+ * @mbq: pointer to the driver internal queue element for mailbox command.
+ *
+ * Driver maintains a internal mailbox command queue implemented as a linked
+ * list. When a mailbox command is issued, it shall be put into the mailbox
+ * command queue such that they shall be processed orderly as HBA can process
+ * one mailbox command at a time.
+ **/
 void
 lpfc_mbox_put(struct lpfc_hba * phba, LPFC_MBOXQ_t * mbq)
 {
@@ -864,6 +1244,20 @@ lpfc_mbox_put(struct lpfc_hba * phba, LPFC_MBOXQ_t * mbq)
 	return;
 }
 
+/**
+ * lpfc_mbox_get: Remove a mailbox cmd from the head of driver's mailbox queue.
+ * @phba: pointer to lpfc hba data structure.
+ *
+ * Driver maintains a internal mailbox command queue implemented as a linked
+ * list. When a mailbox command is issued, it shall be put into the mailbox
+ * command queue such that they shall be processed orderly as HBA can process
+ * one mailbox command at a time. After HBA finished processing a mailbox
+ * command, the driver will remove a pending mailbox command from the head of
+ * the mailbox command queue and send to the HBA for processing.
+ *
+ * Return codes
+ *    pointer to the driver internal queue element for mailbox command.
+ **/
 LPFC_MBOXQ_t *
 lpfc_mbox_get(struct lpfc_hba * phba)
 {
@@ -877,6 +1271,17 @@ lpfc_mbox_get(struct lpfc_hba * phba)
 	return mbq;
 }
 
+/**
+ * lpfc_mbox_cmpl_put: Put mailbox command into mailbox command complete list.
+ * @phba: pointer to lpfc hba data structure.
+ * @mbq: pointer to the driver internal queue element for mailbox command.
+ *
+ * This routine put the completed mailbox command into the mailbox command
+ * complete list. This routine is called from driver interrupt handler
+ * context.The mailbox complete list is used by the driver worker thread
+ * to process mailbox complete callback functions outside the driver interrupt
+ * handler.
+ **/
 void
 lpfc_mbox_cmpl_put(struct lpfc_hba * phba, LPFC_MBOXQ_t * mbq)
 {
@@ -887,6 +1292,17 @@ lpfc_mbox_cmpl_put(struct lpfc_hba * phba, LPFC_MBOXQ_t * mbq)
 	return;
 }
 
+/**
+ * lpfc_mbox_tmo_val: Retrieve mailbox command timeout value.
+ * @phba: pointer to lpfc hba data structure.
+ * @cmd: mailbox command code.
+ *
+ * This routine retrieves the proper timeout value according to the mailbox
+ * command code.
+ *
+ * Return codes
+ *    Timeout value to be used for the given mailbox command
+ **/
 int
 lpfc_mbox_tmo_val(struct lpfc_hba *phba, int cmd)
 {

drivers/scsi/lpfc/lpfc_mem.c

@@ -1,7 +1,7 @@
 /*******************************************************************
  * This file is part of the Emulex Linux Device Driver for         *
  * Fibre Channel Host Bus Adapters.                                *
- * Copyright (C) 2004-2006 Emulex.  All rights reserved.           *
+ * Copyright (C) 2004-2008 Emulex.  All rights reserved.           *
  * EMULEX and SLI are trademarks of Emulex.                        *
  * www.emulex.com                                                  *
  * Portions Copyright (C) 2004-2005 Christoph Hellwig              *
@@ -30,6 +30,7 @@
 
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -39,7 +40,21 @@
 #define LPFC_MEM_POOL_SIZE      64      /* max elem in non-DMA safety pool */
 
 
-
+/**
+ * lpfc_mem_alloc: create and allocate all PCI and memory pools
+ * @phba: HBA to allocate pools for
+ *
+ * Description: Creates and allocates PCI pools lpfc_scsi_dma_buf_pool,
+ * lpfc_mbuf_pool, lpfc_hbq_pool.  Creates and allocates kmalloc-backed mempools
+ * for LPFC_MBOXQ_t and lpfc_nodelist.  Also allocates the VPI bitmask.
+ *
+ * Notes: Not interrupt-safe.  Must be called with no locks held.  If any
+ * allocation fails, frees all successfully allocated memory before returning.
+ *
+ * Returns:
+ *   0 on success
+ *   -ENOMEM on failure (if any memory allocations fail)
+ **/
 int
 lpfc_mem_alloc(struct lpfc_hba * phba)
 {
@@ -120,6 +135,16 @@ lpfc_mem_alloc(struct lpfc_hba * phba)
 	return -ENOMEM;
 }
 
+/**
+ * lpfc_mem_free: Frees all PCI and memory allocated by lpfc_mem_alloc
+ * @phba: HBA to free memory for
+ *
+ * Description: Frees PCI pools lpfc_scsi_dma_buf_pool, lpfc_mbuf_pool,
+ * lpfc_hbq_pool.  Frees kmalloc-backed mempools for LPFC_MBOXQ_t and
+ * lpfc_nodelist.  Also frees the VPI bitmask.
+ *
+ * Returns: None
+ **/
 void
 lpfc_mem_free(struct lpfc_hba * phba)
 {
@@ -181,12 +206,29 @@ lpfc_mem_free(struct lpfc_hba * phba)
 	phba->lpfc_scsi_dma_buf_pool = NULL;
 	phba->lpfc_mbuf_pool = NULL;
 
-				/* Free the iocb lookup array */
+	/* Free the iocb lookup array */
 	kfree(psli->iocbq_lookup);
 	psli->iocbq_lookup = NULL;
-
 }
 
+/**
+ * lpfc_mbuf_alloc: Allocate an mbuf from the lpfc_mbuf_pool PCI pool
+ * @phba: HBA which owns the pool to allocate from
+ * @mem_flags: indicates if this is a priority (MEM_PRI) allocation
+ * @handle: used to return the DMA-mapped address of the mbuf
+ *
+ * Description: Allocates a DMA-mapped buffer from the lpfc_mbuf_pool PCI pool.
+ * Allocates from generic pci_pool_alloc function first and if that fails and
+ * mem_flags has MEM_PRI set (the only defined flag), returns an mbuf from the
+ * HBA's pool.
+ *
+ * Notes: Not interrupt-safe.  Must be called with no locks held.  Takes
+ * phba->hbalock.
+ *
+ * Returns:
+ *   pointer to the allocated mbuf on success
+ *   NULL on failure
+ **/
 void *
 lpfc_mbuf_alloc(struct lpfc_hba *phba, int mem_flags, dma_addr_t *handle)
 {
@@ -206,6 +248,20 @@ lpfc_mbuf_alloc(struct lpfc_hba *phba, int mem_flags, dma_addr_t *handle)
 	return ret;
 }
 
+/**
+ * __lpfc_mem_free: Free an mbuf from the lpfc_mbuf_pool PCI pool (locked)
+ * @phba: HBA which owns the pool to return to
+ * @virt: mbuf to free
+ * @dma: the DMA-mapped address of the lpfc_mbuf_pool to be freed
+ *
+ * Description: Returns an mbuf lpfc_mbuf_pool to the lpfc_mbuf_safety_pool if
+ * it is below its max_count, frees the mbuf otherwise.
+ *
+ * Notes: Must be called with phba->hbalock held to synchronize access to
+ * lpfc_mbuf_safety_pool.
+ *
+ * Returns: None
+ **/
 void
 __lpfc_mbuf_free(struct lpfc_hba * phba, void *virt, dma_addr_t dma)
 {
@@ -221,7 +277,21 @@ __lpfc_mbuf_free(struct lpfc_hba * phba, void *virt, dma_addr_t dma)
 	return;
 }
 
+/**
+ * lpfc_mem_free: Free an mbuf from the lpfc_mbuf_pool PCI pool (unlocked)
+ * @phba: HBA which owns the pool to return to
+ * @virt: mbuf to free
+ * @dma: the DMA-mapped address of the lpfc_mbuf_pool to be freed
+ *
+ * Description: Returns an mbuf lpfc_mbuf_pool to the lpfc_mbuf_safety_pool if
+ * it is below its max_count, frees the mbuf otherwise.
+ *
+ * Notes: Takes phba->hbalock.  Can be called with or without other locks held.
+ *
+ * Returns: None
+ **/
 void
+
 lpfc_mbuf_free(struct lpfc_hba * phba, void *virt, dma_addr_t dma)
 {
 	unsigned long iflags;
@@ -232,6 +302,19 @@ lpfc_mbuf_free(struct lpfc_hba * phba, void *virt, dma_addr_t dma)
 	return;
 }
 
+/**
+ * lpfc_els_hbq_alloc: Allocate an HBQ buffer
+ * @phba: HBA to allocate HBQ buffer for
+ *
+ * Description: Allocates a DMA-mapped HBQ buffer from the lpfc_hbq_pool PCI
+ * pool along a non-DMA-mapped container for it.
+ *
+ * Notes: Not interrupt-safe.  Must be called with no locks held.
+ *
+ * Returns:
+ *   pointer to HBQ on success
+ *   NULL on failure
+ **/
 struct hbq_dmabuf *
 lpfc_els_hbq_alloc(struct lpfc_hba *phba)
 {
@@ -251,6 +334,18 @@ lpfc_els_hbq_alloc(struct lpfc_hba *phba)
 	return hbqbp;
 }
 
+/**
+ * lpfc_mem_hbq_free: Frees an HBQ buffer allocated with lpfc_els_hbq_alloc
+ * @phba: HBA buffer was allocated for
+ * @hbqbp: HBQ container returned by lpfc_els_hbq_alloc
+ *
+ * Description: Frees both the container and the DMA-mapped buffer returned by
+ * lpfc_els_hbq_alloc.
+ *
+ * Notes: Can be called with or without locks held.
+ *
+ * Returns: None
+ **/
 void
 lpfc_els_hbq_free(struct lpfc_hba *phba, struct hbq_dmabuf *hbqbp)
 {
@@ -259,7 +354,18 @@ lpfc_els_hbq_free(struct lpfc_hba *phba, struct hbq_dmabuf *hbqbp)
 	return;
 }
 
-/* This is ONLY called for the LPFC_ELS_HBQ */
+/**
+ * lpfc_in_buf_free: Free a DMA buffer
+ * @phba: HBA buffer is associated with
+ * @mp: Buffer to free
+ *
+ * Description: Frees the given DMA buffer in the appropriate way given if the
+ * HBA is running in SLI3 mode with HBQs enabled.
+ *
+ * Notes: Takes phba->hbalock.  Can be called with or without other locks held.
+ *
+ * Returns: None
+ **/
 void
 lpfc_in_buf_free(struct lpfc_hba *phba, struct lpfc_dmabuf *mp)
 {

drivers/scsi/lpfc/lpfc_nl.h

@@ -0,0 +1,163 @@
+/*******************************************************************
+ * This file is part of the Emulex Linux Device Driver for         *
+ * Fibre Channel Host Bus Adapters.                                *
+ * Copyright (C) 2008 Emulex.  All rights reserved.                *
+ * EMULEX and SLI are trademarks of Emulex.                        *
+ * www.emulex.com                                                  *
+ *                                                                 *
+ * This program is free software; you can redistribute it and/or   *
+ * modify it under the terms of version 2 of the GNU General       *
+ * Public License as published by the Free Software Foundation.    *
+ * This program is distributed in the hope that it will be useful. *
+ * ALL EXPRESS OR IMPLIED CONDITIONS, REPRESENTATIONS AND          *
+ * WARRANTIES, INCLUDING ANY IMPLIED WARRANTY OF MERCHANTABILITY,  *
+ * FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT, ARE      *
+ * DISCLAIMED, EXCEPT TO THE EXTENT THAT SUCH DISCLAIMERS ARE HELD *
+ * TO BE LEGALLY INVALID.  See the GNU General Public License for  *
+ * more details, a copy of which can be found in the file COPYING  *
+ * included with this package.                                     *
+ *******************************************************************/
+
+/* Event definitions for RegisterForEvent */
+#define FC_REG_LINK_EVENT		0x0001	/* link up / down events */
+#define FC_REG_RSCN_EVENT		0x0002	/* RSCN events */
+#define FC_REG_CT_EVENT			0x0004	/* CT request events */
+#define FC_REG_DUMP_EVENT		0x0008	/* Dump events */
+#define FC_REG_TEMPERATURE_EVENT	0x0010	/* temperature events */
+#define FC_REG_ELS_EVENT		0x0020	/* lpfc els events */
+#define FC_REG_FABRIC_EVENT		0x0040	/* lpfc fabric events */
+#define FC_REG_SCSI_EVENT		0x0080	/* lpfc scsi events */
+#define FC_REG_BOARD_EVENT		0x0100	/* lpfc board events */
+#define FC_REG_ADAPTER_EVENT		0x0200	/* lpfc adapter events */
+#define FC_REG_EVENT_MASK		(FC_REG_LINK_EVENT | \
+						FC_REG_RSCN_EVENT | \
+						FC_REG_CT_EVENT | \
+						FC_REG_DUMP_EVENT | \
+						FC_REG_TEMPERATURE_EVENT | \
+						FC_REG_ELS_EVENT | \
+						FC_REG_FABRIC_EVENT | \
+						FC_REG_SCSI_EVENT | \
+						FC_REG_BOARD_EVENT | \
+						FC_REG_ADAPTER_EVENT)
+/* Temperature events */
+#define LPFC_CRIT_TEMP		0x1
+#define LPFC_THRESHOLD_TEMP	0x2
+#define LPFC_NORMAL_TEMP	0x3
+/*
+ * All net link event payloads will begin with and event type
+ * and subcategory. The event type must come first.
+ * The subcategory further defines the data that follows in the rest
+ * of the payload. Each category will have its own unique header plus
+ * any addtional data unique to the subcategory.
+ * The payload sent via the fc transport is one-way driver->application.
+ */
+
+/* els event header */
+struct lpfc_els_event_header {
+	uint32_t event_type;
+	uint32_t subcategory;
+	uint8_t wwpn[8];
+	uint8_t wwnn[8];
+};
+
+/* subcategory codes for FC_REG_ELS_EVENT */
+#define LPFC_EVENT_PLOGI_RCV		0x01
+#define LPFC_EVENT_PRLO_RCV		0x02
+#define LPFC_EVENT_ADISC_RCV		0x04
+#define LPFC_EVENT_LSRJT_RCV		0x08
+
+/* special els lsrjt event */
+struct lpfc_lsrjt_event {
+	struct lpfc_els_event_header header;
+	uint32_t command;
+	uint32_t reason_code;
+	uint32_t explanation;
+};
+
+
+/* fabric event header */
+struct lpfc_fabric_event_header {
+	uint32_t event_type;
+	uint32_t subcategory;
+	uint8_t wwpn[8];
+	uint8_t wwnn[8];
+};
+
+/* subcategory codes for FC_REG_FABRIC_EVENT */
+#define LPFC_EVENT_FABRIC_BUSY		0x01
+#define LPFC_EVENT_PORT_BUSY		0x02
+#define LPFC_EVENT_FCPRDCHKERR		0x04
+
+/* special case fabric fcprdchkerr event */
+struct lpfc_fcprdchkerr_event {
+	struct lpfc_fabric_event_header header;
+	uint32_t lun;
+	uint32_t opcode;
+	uint32_t fcpiparam;
+};
+
+
+/* scsi event header */
+struct lpfc_scsi_event_header {
+	uint32_t event_type;
+	uint32_t subcategory;
+	uint32_t lun;
+	uint8_t wwpn[8];
+	uint8_t wwnn[8];
+};
+
+/* subcategory codes for FC_REG_SCSI_EVENT */
+#define LPFC_EVENT_QFULL	0x0001
+#define LPFC_EVENT_DEVBSY	0x0002
+#define LPFC_EVENT_CHECK_COND	0x0004
+#define LPFC_EVENT_LUNRESET	0x0008
+#define LPFC_EVENT_TGTRESET	0x0010
+#define LPFC_EVENT_BUSRESET	0x0020
+#define LPFC_EVENT_VARQUEDEPTH	0x0040
+
+/* special case scsi varqueuedepth event */
+struct lpfc_scsi_varqueuedepth_event {
+	struct lpfc_scsi_event_header scsi_event;
+	uint32_t oldval;
+	uint32_t newval;
+};
+
+/* special case scsi check condition event */
+struct lpfc_scsi_check_condition_event {
+	struct lpfc_scsi_event_header scsi_event;
+	uint8_t sense_key;
+	uint8_t asc;
+	uint8_t ascq;
+};
+
+/* event codes for FC_REG_BOARD_EVENT */
+#define LPFC_EVENT_PORTINTERR		0x01
+
+/* board event header */
+struct lpfc_board_event_header {
+	uint32_t event_type;
+	uint32_t subcategory;
+};
+
+
+/* event codes for FC_REG_ADAPTER_EVENT */
+#define LPFC_EVENT_ARRIVAL	0x01
+
+/* adapter event header */
+struct lpfc_adapter_event_header {
+	uint32_t event_type;
+	uint32_t subcategory;
+};
+
+
+/* event codes for temp_event */
+#define LPFC_CRIT_TEMP		0x1
+#define LPFC_THRESHOLD_TEMP	0x2
+#define LPFC_NORMAL_TEMP	0x3
+
+struct temp_event {
+	uint32_t event_type;
+	uint32_t event_code;
+	uint32_t data;
+};
+

drivers/scsi/lpfc/lpfc_nportdisc.c

@@ -30,6 +30,7 @@
 
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -1003,20 +1004,8 @@ lpfc_rcv_plogi_adisc_issue(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 			spin_lock_irq(shost->host_lock);
 			ndlp->nlp_flag &= ~NLP_NPR_2B_DISC;
 			spin_unlock_irq(shost->host_lock);
-
-			if (vport->num_disc_nodes) {
+			if (vport->num_disc_nodes)
 				lpfc_more_adisc(vport);
-				if ((vport->num_disc_nodes == 0) &&
-				    (vport->fc_npr_cnt))
-					lpfc_els_disc_plogi(vport);
-				if (vport->num_disc_nodes == 0) {
-					spin_lock_irq(shost->host_lock);
-					vport->fc_flag &= ~FC_NDISC_ACTIVE;
-					spin_unlock_irq(shost->host_lock);
-					lpfc_can_disctmo(vport);
-					lpfc_end_rscn(vport);
-				}
-			}
 		}
 		return ndlp->nlp_state;
 	}
@@ -1865,8 +1854,13 @@ static uint32_t
 lpfc_cmpl_logo_npr_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 			void *arg, uint32_t evt)
 {
+	struct Scsi_Host *shost = lpfc_shost_from_vport(vport);
+	if (ndlp->nlp_DID == Fabric_DID) {
+		spin_lock_irq(shost->host_lock);
+		vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
+		spin_unlock_irq(shost->host_lock);
+	}
 	lpfc_unreg_rpi(vport, ndlp);
-	/* This routine does nothing, just return the current state */
 	return ndlp->nlp_state;
 }
 
@@ -2155,7 +2149,7 @@ lpfc_disc_state_machine(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		lpfc_nlp_put(ndlp);
 	} else {
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
-			"0212 DSM out state %d on NPort free\n", rc);
+			"0213 DSM out state %d on NPort free\n", rc);
 
 		lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_DSM,
 			"DSM out:         ste:%d did:x%x flg:x%x",

drivers/scsi/lpfc/lpfc_scsi.c

@@ -32,6 +32,7 @@
 #include "lpfc_version.h"
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -42,6 +43,111 @@
 #define LPFC_RESET_WAIT  2
 #define LPFC_ABORT_WAIT  2
 
+/**
+ * lpfc_update_stats: Update statistical data for the command completion.
+ * @phba: Pointer to HBA object.
+ * @lpfc_cmd: lpfc scsi command object pointer.
+ *
+ * This function is called when there is a command completion and this
+ * function updates the statistical data for the command completion.
+ **/
+static void
+lpfc_update_stats(struct lpfc_hba *phba, struct  lpfc_scsi_buf *lpfc_cmd)
+{
+	struct lpfc_rport_data *rdata = lpfc_cmd->rdata;
+	struct lpfc_nodelist *pnode = rdata->pnode;
+	struct scsi_cmnd *cmd = lpfc_cmd->pCmd;
+	unsigned long flags;
+	struct Scsi_Host  *shost = cmd->device->host;
+	struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
+	unsigned long latency;
+	int i;
+
+	if (cmd->result)
+		return;
+
+	spin_lock_irqsave(shost->host_lock, flags);
+	if (!vport->stat_data_enabled ||
+		vport->stat_data_blocked ||
+		!pnode->lat_data ||
+		(phba->bucket_type == LPFC_NO_BUCKET)) {
+		spin_unlock_irqrestore(shost->host_lock, flags);
+		return;
+	}
+	latency = jiffies_to_msecs(jiffies - lpfc_cmd->start_time);
+
+	if (phba->bucket_type == LPFC_LINEAR_BUCKET) {
+		i = (latency + phba->bucket_step - 1 - phba->bucket_base)/
+			phba->bucket_step;
+		if (i >= LPFC_MAX_BUCKET_COUNT)
+			i = LPFC_MAX_BUCKET_COUNT;
+	} else {
+		for (i = 0; i < LPFC_MAX_BUCKET_COUNT-1; i++)
+			if (latency <= (phba->bucket_base +
+				((1<<i)*phba->bucket_step)))
+				break;
+	}
+
+	pnode->lat_data[i].cmd_count++;
+	spin_unlock_irqrestore(shost->host_lock, flags);
+}
+
+
+/**
+ * lpfc_send_sdev_queuedepth_change_event: Posts a queuedepth change
+ *                   event.
+ * @phba: Pointer to HBA context object.
+ * @vport: Pointer to vport object.
+ * @ndlp: Pointer to FC node associated with the target.
+ * @lun: Lun number of the scsi device.
+ * @old_val: Old value of the queue depth.
+ * @new_val: New value of the queue depth.
+ *
+ * This function sends an event to the mgmt application indicating
+ * there is a change in the scsi device queue depth.
+ **/
+static void
+lpfc_send_sdev_queuedepth_change_event(struct lpfc_hba *phba,
+		struct lpfc_vport  *vport,
+		struct lpfc_nodelist *ndlp,
+		uint32_t lun,
+		uint32_t old_val,
+		uint32_t new_val)
+{
+	struct lpfc_fast_path_event *fast_path_evt;
+	unsigned long flags;
+
+	fast_path_evt = lpfc_alloc_fast_evt(phba);
+	if (!fast_path_evt)
+		return;
+
+	fast_path_evt->un.queue_depth_evt.scsi_event.event_type =
+		FC_REG_SCSI_EVENT;
+	fast_path_evt->un.queue_depth_evt.scsi_event.subcategory =
+		LPFC_EVENT_VARQUEDEPTH;
+
+	/* Report all luns with change in queue depth */
+	fast_path_evt->un.queue_depth_evt.scsi_event.lun = lun;
+	if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
+		memcpy(&fast_path_evt->un.queue_depth_evt.scsi_event.wwpn,
+			&ndlp->nlp_portname, sizeof(struct lpfc_name));
+		memcpy(&fast_path_evt->un.queue_depth_evt.scsi_event.wwnn,
+			&ndlp->nlp_nodename, sizeof(struct lpfc_name));
+	}
+
+	fast_path_evt->un.queue_depth_evt.oldval = old_val;
+	fast_path_evt->un.queue_depth_evt.newval = new_val;
+	fast_path_evt->vport = vport;
+
+	fast_path_evt->work_evt.evt = LPFC_EVT_FASTPATH_MGMT_EVT;
+	spin_lock_irqsave(&phba->hbalock, flags);
+	list_add_tail(&fast_path_evt->work_evt.evt_listp, &phba->work_list);
+	spin_unlock_irqrestore(&phba->hbalock, flags);
+	lpfc_worker_wake_up(phba);
+
+	return;
+}
+
 /*
  * This function is called with no lock held when there is a resource
  * error in driver or in firmware.
@@ -117,9 +223,10 @@ lpfc_ramp_down_queue_handler(struct lpfc_hba *phba)
 	struct lpfc_vport **vports;
 	struct Scsi_Host  *shost;
 	struct scsi_device *sdev;
-	unsigned long new_queue_depth;
+	unsigned long new_queue_depth, old_queue_depth;
 	unsigned long num_rsrc_err, num_cmd_success;
 	int i;
+	struct lpfc_rport_data *rdata;
 
 	num_rsrc_err = atomic_read(&phba->num_rsrc_err);
 	num_cmd_success = atomic_read(&phba->num_cmd_success);
@@ -137,6 +244,7 @@ lpfc_ramp_down_queue_handler(struct lpfc_hba *phba)
 				else
 					new_queue_depth = sdev->queue_depth -
 								new_queue_depth;
+				old_queue_depth = sdev->queue_depth;
 				if (sdev->ordered_tags)
 					scsi_adjust_queue_depth(sdev,
 							MSG_ORDERED_TAG,
@@ -145,6 +253,13 @@ lpfc_ramp_down_queue_handler(struct lpfc_hba *phba)
 					scsi_adjust_queue_depth(sdev,
 							MSG_SIMPLE_TAG,
 							new_queue_depth);
+				rdata = sdev->hostdata;
+				if (rdata)
+					lpfc_send_sdev_queuedepth_change_event(
+						phba, vports[i],
+						rdata->pnode,
+						sdev->lun, old_queue_depth,
+						new_queue_depth);
 			}
 		}
 	lpfc_destroy_vport_work_array(phba, vports);
@@ -159,6 +274,7 @@ lpfc_ramp_up_queue_handler(struct lpfc_hba *phba)
 	struct Scsi_Host  *shost;
 	struct scsi_device *sdev;
 	int i;
+	struct lpfc_rport_data *rdata;
 
 	vports = lpfc_create_vport_work_array(phba);
 	if (vports != NULL)
@@ -176,6 +292,14 @@ lpfc_ramp_up_queue_handler(struct lpfc_hba *phba)
 					scsi_adjust_queue_depth(sdev,
 							MSG_SIMPLE_TAG,
 							sdev->queue_depth+1);
+				rdata = sdev->hostdata;
+				if (rdata)
+					lpfc_send_sdev_queuedepth_change_event(
+						phba, vports[i],
+						rdata->pnode,
+						sdev->lun,
+						sdev->queue_depth - 1,
+						sdev->queue_depth);
 			}
 		}
 	lpfc_destroy_vport_work_array(phba, vports);
@@ -183,6 +307,35 @@ lpfc_ramp_up_queue_handler(struct lpfc_hba *phba)
 	atomic_set(&phba->num_cmd_success, 0);
 }
 
+/**
+ * lpfc_scsi_dev_block: set all scsi hosts to block state.
+ * @phba: Pointer to HBA context object.
+ *
+ * This function walks vport list and set each SCSI host to block state
+ * by invoking fc_remote_port_delete() routine. This function is invoked
+ * with EEH when device's PCI slot has been permanently disabled.
+ **/
+void
+lpfc_scsi_dev_block(struct lpfc_hba *phba)
+{
+	struct lpfc_vport **vports;
+	struct Scsi_Host  *shost;
+	struct scsi_device *sdev;
+	struct fc_rport *rport;
+	int i;
+
+	vports = lpfc_create_vport_work_array(phba);
+	if (vports != NULL)
+		for (i = 0; i <= phba->max_vpi && vports[i] != NULL; i++) {
+			shost = lpfc_shost_from_vport(vports[i]);
+			shost_for_each_device(sdev, shost) {
+				rport = starget_to_rport(scsi_target(sdev));
+				fc_remote_port_delete(rport);
+			}
+		}
+	lpfc_destroy_vport_work_array(phba, vports);
+}
+
 /*
  * This routine allocates a scsi buffer, which contains all the necessary
  * information needed to initiate a SCSI I/O.  The non-DMAable buffer region
@@ -198,7 +351,9 @@ lpfc_new_scsi_buf(struct lpfc_vport *vport)
 	struct lpfc_scsi_buf *psb;
 	struct ulp_bde64 *bpl;
 	IOCB_t *iocb;
-	dma_addr_t pdma_phys;
+	dma_addr_t pdma_phys_fcp_cmd;
+	dma_addr_t pdma_phys_fcp_rsp;
+	dma_addr_t pdma_phys_bpl;
 	uint16_t iotag;
 
 	psb = kzalloc(sizeof(struct lpfc_scsi_buf), GFP_KERNEL);
@@ -238,40 +393,60 @@ lpfc_new_scsi_buf(struct lpfc_vport *vport)
 
 	/* Initialize local short-hand pointers. */
 	bpl = psb->fcp_bpl;
-	pdma_phys = psb->dma_handle;
+	pdma_phys_fcp_cmd = psb->dma_handle;
+	pdma_phys_fcp_rsp = psb->dma_handle + sizeof(struct fcp_cmnd);
+	pdma_phys_bpl = psb->dma_handle + sizeof(struct fcp_cmnd) +
+			sizeof(struct fcp_rsp);
 
 	/*
 	 * The first two bdes are the FCP_CMD and FCP_RSP.  The balance are sg
 	 * list bdes.  Initialize the first two and leave the rest for
 	 * queuecommand.
 	 */
-	bpl->addrHigh = le32_to_cpu(putPaddrHigh(pdma_phys));
-	bpl->addrLow = le32_to_cpu(putPaddrLow(pdma_phys));
-	bpl->tus.f.bdeSize = sizeof (struct fcp_cmnd);
-	bpl->tus.f.bdeFlags = BUFF_USE_CMND;
-	bpl->tus.w = le32_to_cpu(bpl->tus.w);
-	bpl++;
+	bpl[0].addrHigh = le32_to_cpu(putPaddrHigh(pdma_phys_fcp_cmd));
+	bpl[0].addrLow = le32_to_cpu(putPaddrLow(pdma_phys_fcp_cmd));
+	bpl[0].tus.f.bdeSize = sizeof(struct fcp_cmnd);
+	bpl[0].tus.f.bdeFlags = BUFF_TYPE_BDE_64;
+	bpl[0].tus.w = le32_to_cpu(bpl->tus.w);
 
 	/* Setup the physical region for the FCP RSP */
-	pdma_phys += sizeof (struct fcp_cmnd);
-	bpl->addrHigh = le32_to_cpu(putPaddrHigh(pdma_phys));
-	bpl->addrLow = le32_to_cpu(putPaddrLow(pdma_phys));
-	bpl->tus.f.bdeSize = sizeof (struct fcp_rsp);
-	bpl->tus.f.bdeFlags = (BUFF_USE_CMND | BUFF_USE_RCV);
-	bpl->tus.w = le32_to_cpu(bpl->tus.w);
+	bpl[1].addrHigh = le32_to_cpu(putPaddrHigh(pdma_phys_fcp_rsp));
+	bpl[1].addrLow = le32_to_cpu(putPaddrLow(pdma_phys_fcp_rsp));
+	bpl[1].tus.f.bdeSize = sizeof(struct fcp_rsp);
+	bpl[1].tus.f.bdeFlags = BUFF_TYPE_BDE_64;
+	bpl[1].tus.w = le32_to_cpu(bpl->tus.w);
 
 	/*
 	 * Since the IOCB for the FCP I/O is built into this lpfc_scsi_buf,
 	 * initialize it with all known data now.
 	 */
-	pdma_phys += (sizeof (struct fcp_rsp));
 	iocb = &psb->cur_iocbq.iocb;
 	iocb->un.fcpi64.bdl.ulpIoTag32 = 0;
-	iocb->un.fcpi64.bdl.addrHigh = putPaddrHigh(pdma_phys);
-	iocb->un.fcpi64.bdl.addrLow = putPaddrLow(pdma_phys);
-	iocb->un.fcpi64.bdl.bdeSize = (2 * sizeof (struct ulp_bde64));
-	iocb->un.fcpi64.bdl.bdeFlags = BUFF_TYPE_BDL;
-	iocb->ulpBdeCount = 1;
+	if (phba->sli_rev == 3) {
+		/* fill in immediate fcp command BDE */
+		iocb->un.fcpi64.bdl.bdeFlags = BUFF_TYPE_BDE_IMMED;
+		iocb->un.fcpi64.bdl.bdeSize = sizeof(struct fcp_cmnd);
+		iocb->un.fcpi64.bdl.addrLow = offsetof(IOCB_t,
+						       unsli3.fcp_ext.icd);
+		iocb->un.fcpi64.bdl.addrHigh = 0;
+		iocb->ulpBdeCount = 0;
+		iocb->ulpLe = 0;
+		/* fill in responce BDE */
+		iocb->unsli3.fcp_ext.rbde.tus.f.bdeFlags = BUFF_TYPE_BDE_64;
+		iocb->unsli3.fcp_ext.rbde.tus.f.bdeSize =
+						sizeof(struct fcp_rsp);
+		iocb->unsli3.fcp_ext.rbde.addrLow =
+						putPaddrLow(pdma_phys_fcp_rsp);
+		iocb->unsli3.fcp_ext.rbde.addrHigh =
+						putPaddrHigh(pdma_phys_fcp_rsp);
+	} else {
+		iocb->un.fcpi64.bdl.bdeFlags = BUFF_TYPE_BLP_64;
+		iocb->un.fcpi64.bdl.bdeSize = (2 * sizeof(struct ulp_bde64));
+		iocb->un.fcpi64.bdl.addrLow = putPaddrLow(pdma_phys_bpl);
+		iocb->un.fcpi64.bdl.addrHigh = putPaddrHigh(pdma_phys_bpl);
+		iocb->ulpBdeCount = 1;
+		iocb->ulpLe = 1;
+	}
 	iocb->ulpClass = CLASS3;
 
 	return psb;
@@ -313,8 +488,9 @@ lpfc_scsi_prep_dma_buf(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
 	struct fcp_cmnd *fcp_cmnd = lpfc_cmd->fcp_cmnd;
 	struct ulp_bde64 *bpl = lpfc_cmd->fcp_bpl;
 	IOCB_t *iocb_cmd = &lpfc_cmd->cur_iocbq.iocb;
+	struct ulp_bde64 *data_bde = iocb_cmd->unsli3.fcp_ext.dbde;
 	dma_addr_t physaddr;
-	uint32_t i, num_bde = 0;
+	uint32_t num_bde = 0;
 	int nseg, datadir = scsi_cmnd->sc_data_direction;
 
 	/*
@@ -352,37 +528,159 @@ lpfc_scsi_prep_dma_buf(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
 		 * during probe that limits the number of sg elements in any
 		 * single scsi command.  Just run through the seg_cnt and format
 		 * the bde's.
+		 * When using SLI-3 the driver will try to fit all the BDEs into
+		 * the IOCB. If it can't then the BDEs get added to a BPL as it
+		 * does for SLI-2 mode.
 		 */
-		scsi_for_each_sg(scsi_cmnd, sgel, nseg, i) {
+		scsi_for_each_sg(scsi_cmnd, sgel, nseg, num_bde) {
 			physaddr = sg_dma_address(sgel);
-			bpl->addrLow = le32_to_cpu(putPaddrLow(physaddr));
-			bpl->addrHigh = le32_to_cpu(putPaddrHigh(physaddr));
-			bpl->tus.f.bdeSize = sg_dma_len(sgel);
-			if (datadir == DMA_TO_DEVICE)
-				bpl->tus.f.bdeFlags = 0;
-			else
-				bpl->tus.f.bdeFlags = BUFF_USE_RCV;
-			bpl->tus.w = le32_to_cpu(bpl->tus.w);
-			bpl++;
-			num_bde++;
+			if (phba->sli_rev == 3 &&
+			    nseg <= LPFC_EXT_DATA_BDE_COUNT) {
+				data_bde->tus.f.bdeFlags = BUFF_TYPE_BDE_64;
+				data_bde->tus.f.bdeSize = sg_dma_len(sgel);
+				data_bde->addrLow = putPaddrLow(physaddr);
+				data_bde->addrHigh = putPaddrHigh(physaddr);
+				data_bde++;
+			} else {
+				bpl->tus.f.bdeFlags = BUFF_TYPE_BDE_64;
+				bpl->tus.f.bdeSize = sg_dma_len(sgel);
+				bpl->tus.w = le32_to_cpu(bpl->tus.w);
+				bpl->addrLow =
+					le32_to_cpu(putPaddrLow(physaddr));
+				bpl->addrHigh =
+					le32_to_cpu(putPaddrHigh(physaddr));
+				bpl++;
+			}
 		}
 	}
 
 	/*
 	 * Finish initializing those IOCB fields that are dependent on the
-	 * scsi_cmnd request_buffer.  Note that the bdeSize is explicitly
-	 * reinitialized since all iocb memory resources are used many times
-	 * for transmit, receive, and continuation bpl's.
+	 * scsi_cmnd request_buffer.  Note that for SLI-2 the bdeSize is
+	 * explicitly reinitialized and for SLI-3 the extended bde count is
+	 * explicitly reinitialized since all iocb memory resources are reused.
 	 */
-	iocb_cmd->un.fcpi64.bdl.bdeSize = (2 * sizeof (struct ulp_bde64));
-	iocb_cmd->un.fcpi64.bdl.bdeSize +=
-		(num_bde * sizeof (struct ulp_bde64));
-	iocb_cmd->ulpBdeCount = 1;
-	iocb_cmd->ulpLe = 1;
+	if (phba->sli_rev == 3) {
+		if (num_bde > LPFC_EXT_DATA_BDE_COUNT) {
+			/*
+			 * The extended IOCB format can only fit 3 BDE or a BPL.
+			 * This I/O has more than 3 BDE so the 1st data bde will
+			 * be a BPL that is filled in here.
+			 */
+			physaddr = lpfc_cmd->dma_handle;
+			data_bde->tus.f.bdeFlags = BUFF_TYPE_BLP_64;
+			data_bde->tus.f.bdeSize = (num_bde *
+						   sizeof(struct ulp_bde64));
+			physaddr += (sizeof(struct fcp_cmnd) +
+				     sizeof(struct fcp_rsp) +
+				     (2 * sizeof(struct ulp_bde64)));
+			data_bde->addrHigh = putPaddrHigh(physaddr);
+			data_bde->addrLow = putPaddrLow(physaddr);
+			/* ebde count includes the responce bde and data bpl */
+			iocb_cmd->unsli3.fcp_ext.ebde_count = 2;
+		} else {
+			/* ebde count includes the responce bde and data bdes */
+			iocb_cmd->unsli3.fcp_ext.ebde_count = (num_bde + 1);
+		}
+	} else {
+		iocb_cmd->un.fcpi64.bdl.bdeSize =
+			((num_bde + 2) * sizeof(struct ulp_bde64));
+	}
 	fcp_cmnd->fcpDl = cpu_to_be32(scsi_bufflen(scsi_cmnd));
 	return 0;
 }
 
+/**
+ * lpfc_send_scsi_error_event: Posts an event when there is SCSI error.
+ * @phba: Pointer to hba context object.
+ * @vport: Pointer to vport object.
+ * @lpfc_cmd: Pointer to lpfc scsi command which reported the error.
+ * @rsp_iocb: Pointer to response iocb object which reported error.
+ *
+ * This function posts an event when there is a SCSI command reporting
+ * error from the scsi device.
+ **/
+static void
+lpfc_send_scsi_error_event(struct lpfc_hba *phba, struct lpfc_vport *vport,
+		struct lpfc_scsi_buf *lpfc_cmd, struct lpfc_iocbq *rsp_iocb) {
+	struct scsi_cmnd *cmnd = lpfc_cmd->pCmd;
+	struct fcp_rsp *fcprsp = lpfc_cmd->fcp_rsp;
+	uint32_t resp_info = fcprsp->rspStatus2;
+	uint32_t scsi_status = fcprsp->rspStatus3;
+	uint32_t fcpi_parm = rsp_iocb->iocb.un.fcpi.fcpi_parm;
+	struct lpfc_fast_path_event *fast_path_evt = NULL;
+	struct lpfc_nodelist *pnode = lpfc_cmd->rdata->pnode;
+	unsigned long flags;
+
+	/* If there is queuefull or busy condition send a scsi event */
+	if ((cmnd->result == SAM_STAT_TASK_SET_FULL) ||
+		(cmnd->result == SAM_STAT_BUSY)) {
+		fast_path_evt = lpfc_alloc_fast_evt(phba);
+		if (!fast_path_evt)
+			return;
+		fast_path_evt->un.scsi_evt.event_type =
+			FC_REG_SCSI_EVENT;
+		fast_path_evt->un.scsi_evt.subcategory =
+		(cmnd->result == SAM_STAT_TASK_SET_FULL) ?
+		LPFC_EVENT_QFULL : LPFC_EVENT_DEVBSY;
+		fast_path_evt->un.scsi_evt.lun = cmnd->device->lun;
+		memcpy(&fast_path_evt->un.scsi_evt.wwpn,
+			&pnode->nlp_portname, sizeof(struct lpfc_name));
+		memcpy(&fast_path_evt->un.scsi_evt.wwnn,
+			&pnode->nlp_nodename, sizeof(struct lpfc_name));
+	} else if ((resp_info & SNS_LEN_VALID) && fcprsp->rspSnsLen &&
+		((cmnd->cmnd[0] == READ_10) || (cmnd->cmnd[0] == WRITE_10))) {
+		fast_path_evt = lpfc_alloc_fast_evt(phba);
+		if (!fast_path_evt)
+			return;
+		fast_path_evt->un.check_cond_evt.scsi_event.event_type =
+			FC_REG_SCSI_EVENT;
+		fast_path_evt->un.check_cond_evt.scsi_event.subcategory =
+			LPFC_EVENT_CHECK_COND;
+		fast_path_evt->un.check_cond_evt.scsi_event.lun =
+			cmnd->device->lun;
+		memcpy(&fast_path_evt->un.check_cond_evt.scsi_event.wwpn,
+			&pnode->nlp_portname, sizeof(struct lpfc_name));
+		memcpy(&fast_path_evt->un.check_cond_evt.scsi_event.wwnn,
+			&pnode->nlp_nodename, sizeof(struct lpfc_name));
+		fast_path_evt->un.check_cond_evt.sense_key =
+			cmnd->sense_buffer[2] & 0xf;
+		fast_path_evt->un.check_cond_evt.asc = cmnd->sense_buffer[12];
+		fast_path_evt->un.check_cond_evt.ascq = cmnd->sense_buffer[13];
+	} else if ((cmnd->sc_data_direction == DMA_FROM_DEVICE) &&
+		     fcpi_parm &&
+		     ((be32_to_cpu(fcprsp->rspResId) != fcpi_parm) ||
+			((scsi_status == SAM_STAT_GOOD) &&
+			!(resp_info & (RESID_UNDER | RESID_OVER))))) {
+		/*
+		 * If status is good or resid does not match with fcp_param and
+		 * there is valid fcpi_parm, then there is a read_check error
+		 */
+		fast_path_evt = lpfc_alloc_fast_evt(phba);
+		if (!fast_path_evt)
+			return;
+		fast_path_evt->un.read_check_error.header.event_type =
+			FC_REG_FABRIC_EVENT;
+		fast_path_evt->un.read_check_error.header.subcategory =
+			LPFC_EVENT_FCPRDCHKERR;
+		memcpy(&fast_path_evt->un.read_check_error.header.wwpn,
+			&pnode->nlp_portname, sizeof(struct lpfc_name));
+		memcpy(&fast_path_evt->un.read_check_error.header.wwnn,
+			&pnode->nlp_nodename, sizeof(struct lpfc_name));
+		fast_path_evt->un.read_check_error.lun = cmnd->device->lun;
+		fast_path_evt->un.read_check_error.opcode = cmnd->cmnd[0];
+		fast_path_evt->un.read_check_error.fcpiparam =
+			fcpi_parm;
+	} else
+		return;
+
+	fast_path_evt->vport = vport;
+	spin_lock_irqsave(&phba->hbalock, flags);
+	list_add_tail(&fast_path_evt->work_evt.evt_listp, &phba->work_list);
+	spin_unlock_irqrestore(&phba->hbalock, flags);
+	lpfc_worker_wake_up(phba);
+	return;
+}
 static void
 lpfc_scsi_unprep_dma_buf(struct lpfc_hba * phba, struct lpfc_scsi_buf * psb)
 {
@@ -411,6 +709,7 @@ lpfc_handle_fcp_err(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd,
 	uint32_t rsplen = 0;
 	uint32_t logit = LOG_FCP | LOG_FCP_ERROR;
 
+
 	/*
 	 *  If this is a task management command, there is no
 	 *  scsi packet associated with this lpfc_cmd.  The driver
@@ -526,6 +825,7 @@ lpfc_handle_fcp_err(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd,
 
  out:
 	cmnd->result = ScsiResult(host_status, scsi_status);
+	lpfc_send_scsi_error_event(vport->phba, vport, lpfc_cmd, rsp_iocb);
 }
 
 static void
@@ -542,9 +842,11 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 	struct scsi_device *sdev, *tmp_sdev;
 	int depth = 0;
 	unsigned long flags;
+	struct lpfc_fast_path_event *fast_path_evt;
 
 	lpfc_cmd->result = pIocbOut->iocb.un.ulpWord[4];
 	lpfc_cmd->status = pIocbOut->iocb.ulpStatus;
+	atomic_dec(&pnode->cmd_pending);
 
 	if (lpfc_cmd->status) {
 		if (lpfc_cmd->status == IOSTAT_LOCAL_REJECT &&
@@ -570,12 +872,36 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 			break;
 		case IOSTAT_NPORT_BSY:
 		case IOSTAT_FABRIC_BSY:
-			cmd->result = ScsiResult(DID_BUS_BUSY, 0);
+			cmd->result = ScsiResult(DID_TRANSPORT_DISRUPTED, 0);
+			fast_path_evt = lpfc_alloc_fast_evt(phba);
+			if (!fast_path_evt)
+				break;
+			fast_path_evt->un.fabric_evt.event_type =
+				FC_REG_FABRIC_EVENT;
+			fast_path_evt->un.fabric_evt.subcategory =
+				(lpfc_cmd->status == IOSTAT_NPORT_BSY) ?
+				LPFC_EVENT_PORT_BUSY : LPFC_EVENT_FABRIC_BUSY;
+			if (pnode && NLP_CHK_NODE_ACT(pnode)) {
+				memcpy(&fast_path_evt->un.fabric_evt.wwpn,
+					&pnode->nlp_portname,
+					sizeof(struct lpfc_name));
+				memcpy(&fast_path_evt->un.fabric_evt.wwnn,
+					&pnode->nlp_nodename,
+					sizeof(struct lpfc_name));
+			}
+			fast_path_evt->vport = vport;
+			fast_path_evt->work_evt.evt =
+				LPFC_EVT_FASTPATH_MGMT_EVT;
+			spin_lock_irqsave(&phba->hbalock, flags);
+			list_add_tail(&fast_path_evt->work_evt.evt_listp,
+				&phba->work_list);
+			spin_unlock_irqrestore(&phba->hbalock, flags);
+			lpfc_worker_wake_up(phba);
 			break;
 		case IOSTAT_LOCAL_REJECT:
-			if (lpfc_cmd->result == RJT_UNAVAIL_PERM ||
+			if (lpfc_cmd->result == IOERR_INVALID_RPI ||
 			    lpfc_cmd->result == IOERR_NO_RESOURCES ||
-			    lpfc_cmd->result == RJT_LOGIN_REQUIRED) {
+			    lpfc_cmd->result == IOERR_ABORT_REQUESTED) {
 				cmd->result = ScsiResult(DID_REQUEUE, 0);
 				break;
 			} /* else: fall through */
@@ -586,7 +912,8 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 
 		if (!pnode || !NLP_CHK_NODE_ACT(pnode)
 		    || (pnode->nlp_state != NLP_STE_MAPPED_NODE))
-			cmd->result = ScsiResult(DID_BUS_BUSY, SAM_STAT_BUSY);
+			cmd->result = ScsiResult(DID_TRANSPORT_DISRUPTED,
+						 SAM_STAT_BUSY);
 	} else {
 		cmd->result = ScsiResult(DID_OK, 0);
 	}
@@ -602,8 +929,32 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 				 scsi_get_resid(cmd));
 	}
 
+	lpfc_update_stats(phba, lpfc_cmd);
 	result = cmd->result;
 	sdev = cmd->device;
+	if (vport->cfg_max_scsicmpl_time &&
+	   time_after(jiffies, lpfc_cmd->start_time +
+		msecs_to_jiffies(vport->cfg_max_scsicmpl_time))) {
+		spin_lock_irqsave(sdev->host->host_lock, flags);
+		if ((pnode->cmd_qdepth > atomic_read(&pnode->cmd_pending) &&
+		    (atomic_read(&pnode->cmd_pending) > LPFC_MIN_TGT_QDEPTH) &&
+		    ((cmd->cmnd[0] == READ_10) || (cmd->cmnd[0] == WRITE_10))))
+			pnode->cmd_qdepth = atomic_read(&pnode->cmd_pending);
+
+		pnode->last_change_time = jiffies;
+		spin_unlock_irqrestore(sdev->host->host_lock, flags);
+	} else if ((pnode->cmd_qdepth < LPFC_MAX_TGT_QDEPTH) &&
+		   time_after(jiffies, pnode->last_change_time +
+			msecs_to_jiffies(LPFC_TGTQ_INTERVAL))) {
+		spin_lock_irqsave(sdev->host->host_lock, flags);
+		pnode->cmd_qdepth += pnode->cmd_qdepth *
+			LPFC_TGTQ_RAMPUP_PCENT / 100;
+		if (pnode->cmd_qdepth > LPFC_MAX_TGT_QDEPTH)
+			pnode->cmd_qdepth = LPFC_MAX_TGT_QDEPTH;
+		pnode->last_change_time = jiffies;
+		spin_unlock_irqrestore(sdev->host->host_lock, flags);
+	}
+
 	lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd);
 	cmd->scsi_done(cmd);
 
@@ -647,6 +998,9 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 				pnode->last_ramp_up_time = jiffies;
 			}
 		}
+		lpfc_send_sdev_queuedepth_change_event(phba, vport, pnode,
+			0xFFFFFFFF,
+			sdev->queue_depth - 1, sdev->queue_depth);
 	}
 
 	/*
@@ -676,6 +1030,9 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 			lpfc_printf_vlog(vport, KERN_WARNING, LOG_FCP,
 					 "0711 detected queue full - lun queue "
 					 "depth adjusted to %d.\n", depth);
+			lpfc_send_sdev_queuedepth_change_event(phba, vport,
+				pnode, 0xFFFFFFFF,
+				depth+1, depth);
 		}
 	}
 
@@ -692,6 +1049,24 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn,
 	lpfc_release_scsi_buf(phba, lpfc_cmd);
 }
 
+/**
+ * lpfc_fcpcmd_to_iocb - copy the fcp_cmd data into the IOCB.
+ * @data: A pointer to the immediate command data portion of the IOCB.
+ * @fcp_cmnd: The FCP Command that is provided by the SCSI layer.
+ *
+ * The routine copies the entire FCP command from @fcp_cmnd to @data while
+ * byte swapping the data to big endian format for transmission on the wire.
+ **/
+static void
+lpfc_fcpcmd_to_iocb(uint8_t *data, struct fcp_cmnd *fcp_cmnd)
+{
+	int i, j;
+	for (i = 0, j = 0; i < sizeof(struct fcp_cmnd);
+	     i += sizeof(uint32_t), j++) {
+		((uint32_t *)data)[j] = cpu_to_be32(((uint32_t *)fcp_cmnd)[j]);
+	}
+}
+
 static void
 lpfc_scsi_prep_cmnd(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd,
 		    struct lpfc_nodelist *pnode)
@@ -758,7 +1133,8 @@ lpfc_scsi_prep_cmnd(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd,
 		fcp_cmnd->fcpCntl3 = 0;
 		phba->fc4ControlRequests++;
 	}
-
+	if (phba->sli_rev == 3)
+		lpfc_fcpcmd_to_iocb(iocb_cmd->unsli3.fcp_ext.icd, fcp_cmnd);
 	/*
 	 * Finish initializing those IOCB fields that are independent
 	 * of the scsi_cmnd request_buffer
@@ -798,11 +1174,13 @@ lpfc_scsi_prep_task_mgmt_cmd(struct lpfc_vport *vport,
 	piocb = &piocbq->iocb;
 
 	fcp_cmnd = lpfc_cmd->fcp_cmnd;
-	int_to_scsilun(lun, &lpfc_cmd->fcp_cmnd->fcp_lun);
+	/* Clear out any old data in the FCP command area */
+	memset(fcp_cmnd, 0, sizeof(struct fcp_cmnd));
+	int_to_scsilun(lun, &fcp_cmnd->fcp_lun);
 	fcp_cmnd->fcpCntl2 = task_mgmt_cmd;
-
+	if (vport->phba->sli_rev == 3)
+		lpfc_fcpcmd_to_iocb(piocb->unsli3.fcp_ext.icd, fcp_cmnd);
 	piocb->ulpCommand = CMD_FCP_ICMND64_CR;
-
 	piocb->ulpContext = ndlp->nlp_rpi;
 	if (ndlp->nlp_fcp_info & NLP_FCP_2_DEVICE) {
 		piocb->ulpFCP2Rcvy = 1;
@@ -967,9 +1345,12 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 	 * transport is still transitioning.
 	 */
 	if (!ndlp || !NLP_CHK_NODE_ACT(ndlp)) {
-		cmnd->result = ScsiResult(DID_BUS_BUSY, 0);
+		cmnd->result = ScsiResult(DID_TRANSPORT_DISRUPTED, 0);
 		goto out_fail_command;
 	}
+	if (atomic_read(&ndlp->cmd_pending) >= ndlp->cmd_qdepth)
+		goto out_host_busy;
+
 	lpfc_cmd = lpfc_get_scsi_buf(phba);
 	if (lpfc_cmd == NULL) {
 		lpfc_adjust_queue_depth(phba);
@@ -980,6 +1361,7 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 		goto out_host_busy;
 	}
 
+	lpfc_cmd->start_time = jiffies;
 	/*
 	 * Store the midlayer's command structure for the completion phase
 	 * and complete the command initialization.
@@ -987,6 +1369,7 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 	lpfc_cmd->pCmd  = cmnd;
 	lpfc_cmd->rdata = rdata;
 	lpfc_cmd->timeout = 0;
+	lpfc_cmd->start_time = jiffies;
 	cmnd->host_scribble = (unsigned char *)lpfc_cmd;
 	cmnd->scsi_done = done;
 
@@ -996,6 +1379,7 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 
 	lpfc_scsi_prep_cmnd(vport, lpfc_cmd, ndlp);
 
+	atomic_inc(&ndlp->cmd_pending);
 	err = lpfc_sli_issue_iocb(phba, &phba->sli.ring[psli->fcp_ring],
 				  &lpfc_cmd->cur_iocbq, SLI_IOCB_RET_IOCB);
 	if (err)
@@ -1010,6 +1394,7 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 	return 0;
 
  out_host_busy_free_buf:
+	atomic_dec(&ndlp->cmd_pending);
 	lpfc_scsi_unprep_dma_buf(phba, lpfc_cmd);
 	lpfc_release_scsi_buf(phba, lpfc_cmd);
  out_host_busy:
@@ -1145,6 +1530,7 @@ lpfc_device_reset_handler(struct scsi_cmnd *cmnd)
 	int ret = SUCCESS;
 	int status;
 	int cnt;
+	struct lpfc_scsi_event_header scsi_event;
 
 	lpfc_block_error_handler(cmnd);
 	/*
@@ -1163,6 +1549,19 @@ lpfc_device_reset_handler(struct scsi_cmnd *cmnd)
 			break;
 		pnode = rdata->pnode;
 	}
+
+	scsi_event.event_type = FC_REG_SCSI_EVENT;
+	scsi_event.subcategory = LPFC_EVENT_TGTRESET;
+	scsi_event.lun = 0;
+	memcpy(scsi_event.wwpn, &pnode->nlp_portname, sizeof(struct lpfc_name));
+	memcpy(scsi_event.wwnn, &pnode->nlp_nodename, sizeof(struct lpfc_name));
+
+	fc_host_post_vendor_event(shost,
+		fc_get_event_number(),
+		sizeof(scsi_event),
+		(char *)&scsi_event,
+		SCSI_NL_VID_TYPE_PCI | PCI_VENDOR_ID_EMULEX);
+
 	if (!rdata || pnode->nlp_state != NLP_STE_MAPPED_NODE) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
 				 "0721 LUN Reset rport "
@@ -1242,10 +1641,23 @@ lpfc_bus_reset_handler(struct scsi_cmnd *cmnd)
 	struct lpfc_hba   *phba = vport->phba;
 	struct lpfc_nodelist *ndlp = NULL;
 	int match;
-	int ret = SUCCESS, status, i;
+	int ret = SUCCESS, status = SUCCESS, i;
 	int cnt;
 	struct lpfc_scsi_buf * lpfc_cmd;
 	unsigned long later;
+	struct lpfc_scsi_event_header scsi_event;
+
+	scsi_event.event_type = FC_REG_SCSI_EVENT;
+	scsi_event.subcategory = LPFC_EVENT_BUSRESET;
+	scsi_event.lun = 0;
+	memcpy(scsi_event.wwpn, &vport->fc_portname, sizeof(struct lpfc_name));
+	memcpy(scsi_event.wwnn, &vport->fc_nodename, sizeof(struct lpfc_name));
+
+	fc_host_post_vendor_event(shost,
+		fc_get_event_number(),
+		sizeof(scsi_event),
+		(char *)&scsi_event,
+		SCSI_NL_VID_TYPE_PCI | PCI_VENDOR_ID_EMULEX);
 
 	lpfc_block_error_handler(cmnd);
 	/*

drivers/scsi/lpfc/lpfc_scsi.h

@@ -107,6 +107,10 @@ struct fcp_cmnd {
 
 };
 
+struct lpfc_scsicmd_bkt {
+	uint32_t cmd_count;
+};
+
 struct lpfc_scsi_buf {
 	struct list_head list;
 	struct scsi_cmnd *pCmd;
@@ -139,6 +143,7 @@ struct lpfc_scsi_buf {
 	 */
 	struct lpfc_iocbq cur_iocbq;
 	wait_queue_head_t *waitq;
+	unsigned long start_time;
 };
 
 #define LPFC_SCSI_DMA_EXT_SIZE 264

drivers/scsi/lpfc/lpfc_sli.h

@@ -233,6 +233,7 @@ struct lpfc_sli {
 #define LPFC_SLI2_ACTIVE          0x200	/* SLI2 overlay in firmware is active */
 #define LPFC_PROCESS_LA           0x400	/* Able to process link attention */
 #define LPFC_BLOCK_MGMT_IO        0x800	/* Don't allow mgmt mbx or iocb cmds */
+#define LPFC_MENLO_MAINT          0x1000 /* need for menl fw download */
 
 	struct lpfc_sli_ring ring[LPFC_MAX_RING];
 	int fcp_ring;		/* ring used for FCP initiator commands */

drivers/scsi/lpfc/lpfc_version.h

@@ -18,9 +18,11 @@
  * included with this package.                                     *
  *******************************************************************/
 
-#define LPFC_DRIVER_VERSION "8.2.7"
+#define LPFC_DRIVER_VERSION "8.2.8"
 
-#define LPFC_DRIVER_NAME "lpfc"
+#define LPFC_DRIVER_NAME		"lpfc"
+#define LPFC_SP_DRIVER_HANDLER_NAME	"lpfc:sp"
+#define LPFC_FP_DRIVER_HANDLER_NAME	"lpfc:fp"
 
 #define LPFC_MODULE_DESC "Emulex LightPulse Fibre Channel SCSI driver " \
 		LPFC_DRIVER_VERSION

drivers/scsi/lpfc/lpfc_vport.c

@@ -34,6 +34,7 @@
 #include <scsi/scsi_transport_fc.h>
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
+#include "lpfc_nl.h"
 #include "lpfc_disc.h"
 #include "lpfc_scsi.h"
 #include "lpfc.h"
@@ -204,6 +205,77 @@ lpfc_unique_wwpn(struct lpfc_hba *phba, struct lpfc_vport *new_vport)
 	return 1;
 }
 
+/**
+ * lpfc_discovery_wait: Wait for driver discovery to quiesce.
+ * @vport: The virtual port for which this call is being executed.
+ *
+ * This driver calls this routine specifically from lpfc_vport_delete
+ * to enforce a synchronous execution of vport
+ * delete relative to discovery activities.  The
+ * lpfc_vport_delete routine should not return until it
+ * can reasonably guarantee that discovery has quiesced.
+ * Post FDISC LOGO, the driver must wait until its SAN teardown is
+ * complete and all resources recovered before allowing
+ * cleanup.
+ *
+ * This routine does not require any locks held.
+ **/
+static void lpfc_discovery_wait(struct lpfc_vport *vport)
+{
+	struct lpfc_hba *phba = vport->phba;
+	uint32_t wait_flags = 0;
+	unsigned long wait_time_max;
+	unsigned long start_time;
+
+	wait_flags = FC_RSCN_MODE | FC_RSCN_DISCOVERY | FC_NLP_MORE |
+		     FC_RSCN_DEFERRED | FC_NDISC_ACTIVE | FC_DISC_TMO;
+
+	/*
+	 * The time constraint on this loop is a balance between the
+	 * fabric RA_TOV value and dev_loss tmo.  The driver's
+	 * devloss_tmo is 10 giving this loop a 3x multiplier minimally.
+	 */
+	wait_time_max = msecs_to_jiffies(((phba->fc_ratov * 3) + 3) * 1000);
+	wait_time_max += jiffies;
+	start_time = jiffies;
+	while (time_before(jiffies, wait_time_max)) {
+		if ((vport->num_disc_nodes > 0)    ||
+		    (vport->fc_flag & wait_flags)  ||
+		    ((vport->port_state > LPFC_VPORT_FAILED) &&
+		     (vport->port_state < LPFC_VPORT_READY))) {
+			lpfc_printf_log(phba, KERN_INFO, LOG_VPORT,
+					"1833 Vport discovery quiesce Wait:"
+					" vpi x%x state x%x fc_flags x%x"
+					" num_nodes x%x, waiting 1000 msecs"
+					" total wait msecs x%x\n",
+					vport->vpi, vport->port_state,
+					vport->fc_flag, vport->num_disc_nodes,
+					jiffies_to_msecs(jiffies - start_time));
+			msleep(1000);
+		} else {
+			/* Base case.  Wait variants satisfied.  Break out */
+			lpfc_printf_log(phba, KERN_INFO, LOG_VPORT,
+					 "1834 Vport discovery quiesced:"
+					 " vpi x%x state x%x fc_flags x%x"
+					 " wait msecs x%x\n",
+					 vport->vpi, vport->port_state,
+					 vport->fc_flag,
+					 jiffies_to_msecs(jiffies
+						- start_time));
+			break;
+		}
+	}
+
+	if (time_after(jiffies, wait_time_max))
+		lpfc_printf_log(phba, KERN_ERR, LOG_VPORT,
+				"1835 Vport discovery quiesce failed:"
+				" vpi x%x state x%x fc_flags x%x"
+				" wait msecs x%x\n",
+				vport->vpi, vport->port_state,
+				vport->fc_flag,
+				jiffies_to_msecs(jiffies - start_time));
+}
+
 int
 lpfc_vport_create(struct fc_vport *fc_vport, bool disable)
 {
@@ -506,8 +578,12 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 	 * initiated after we've disposed of all other resources associated
 	 * with the port.
 	 */
-	if (!scsi_host_get(shost) || !scsi_host_get(shost))
+	if (!scsi_host_get(shost))
+		return VPORT_INVAL;
+	if (!scsi_host_get(shost)) {
+		scsi_host_put(shost);
 		return VPORT_INVAL;
+	}
 	spin_lock_irq(&phba->hbalock);
 	vport->load_flag |= FC_UNLOADING;
 	spin_unlock_irq(&phba->hbalock);
@@ -597,11 +673,16 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 		}
 		vport->unreg_vpi_cmpl = VPORT_INVAL;
 		timeout = msecs_to_jiffies(phba->fc_ratov * 2000);
+		if (ndlp->nlp_state == NLP_STE_UNUSED_NODE)
+			goto skip_logo;
 		if (!lpfc_issue_els_npiv_logo(vport, ndlp))
 			while (vport->unreg_vpi_cmpl == VPORT_INVAL && timeout)
 				timeout = schedule_timeout(timeout);
 	}
 
+	if (!(phba->pport->load_flag & FC_UNLOADING))
+		lpfc_discovery_wait(vport);
+
 skip_logo:
 	lpfc_cleanup(vport);
 	lpfc_sli_host_down(vport);
@@ -615,8 +696,10 @@ skip_logo:
 		 * Completion of unreg_vpi (lpfc_mbx_cmpl_unreg_vpi)
 		 * does the scsi_host_put() to release the vport.
 		 */
-		lpfc_mbx_unreg_vpi(vport);
-	}
+		if (lpfc_mbx_unreg_vpi(vport))
+			scsi_host_put(shost);
+	} else
+		scsi_host_put(shost);
 
 	lpfc_free_vpi(phba, vport->vpi);
 	vport->work_port_events = 0;
@@ -663,3 +746,82 @@ lpfc_destroy_vport_work_array(struct lpfc_hba *phba, struct lpfc_vport **vports)
 		scsi_host_put(lpfc_shost_from_vport(vports[i]));
 	kfree(vports);
 }
+
+
+/**
+ * lpfc_vport_reset_stat_data: Reset the statistical data for the vport.
+ * @vport: Pointer to vport object.
+ *
+ * This function resets the statistical data for the vport. This function
+ * is called with the host_lock held
+ **/
+void
+lpfc_vport_reset_stat_data(struct lpfc_vport *vport)
+{
+	struct lpfc_nodelist *ndlp = NULL, *next_ndlp = NULL;
+
+	list_for_each_entry_safe(ndlp, next_ndlp, &vport->fc_nodes, nlp_listp) {
+		if (!NLP_CHK_NODE_ACT(ndlp))
+			continue;
+		if (ndlp->lat_data)
+			memset(ndlp->lat_data, 0, LPFC_MAX_BUCKET_COUNT *
+				sizeof(struct lpfc_scsicmd_bkt));
+	}
+}
+
+
+/**
+ * lpfc_alloc_bucket: Allocate data buffer required for collecting
+ *  statistical data.
+ * @vport: Pointer to vport object.
+ *
+ * This function allocates data buffer required for all the FC
+ * nodes of the vport to collect statistical data.
+ **/
+void
+lpfc_alloc_bucket(struct lpfc_vport *vport)
+{
+	struct lpfc_nodelist *ndlp = NULL, *next_ndlp = NULL;
+
+	list_for_each_entry_safe(ndlp, next_ndlp, &vport->fc_nodes, nlp_listp) {
+		if (!NLP_CHK_NODE_ACT(ndlp))
+			continue;
+
+		kfree(ndlp->lat_data);
+		ndlp->lat_data = NULL;
+
+		if (ndlp->nlp_state == NLP_STE_MAPPED_NODE) {
+			ndlp->lat_data = kcalloc(LPFC_MAX_BUCKET_COUNT,
+					 sizeof(struct lpfc_scsicmd_bkt),
+					 GFP_ATOMIC);
+
+			if (!ndlp->lat_data)
+				lpfc_printf_vlog(vport, KERN_ERR, LOG_NODE,
+					"0287 lpfc_alloc_bucket failed to "
+					"allocate statistical data buffer DID "
+					"0x%x\n", ndlp->nlp_DID);
+		}
+	}
+}
+
+/**
+ * lpfc_free_bucket: Free data buffer required for collecting
+ *  statistical data.
+ * @vport: Pointer to vport object.
+ *
+ * Th function frees statistical data buffer of all the FC
+ * nodes of the vport.
+ **/
+void
+lpfc_free_bucket(struct lpfc_vport *vport)
+{
+	struct lpfc_nodelist *ndlp = NULL, *next_ndlp = NULL;
+
+	list_for_each_entry_safe(ndlp, next_ndlp, &vport->fc_nodes, nlp_listp) {
+		if (!NLP_CHK_NODE_ACT(ndlp))
+			continue;
+
+		kfree(ndlp->lat_data);
+		ndlp->lat_data = NULL;
+	}
+}

drivers/scsi/lpfc/lpfc_vport.h

@@ -112,4 +112,8 @@ struct vport_cmd_tag {
 void lpfc_vport_set_state(struct lpfc_vport *vport,
 			  enum fc_vport_state new_state);
 
+void lpfc_vport_reset_stat_data(struct lpfc_vport *);
+void lpfc_alloc_bucket(struct lpfc_vport *);
+void lpfc_free_bucket(struct lpfc_vport *);
+
 #endif /* H_LPFC_VPORT */

drivers/scsi/qla2xxx/qla_attr.c

@@ -1006,7 +1006,6 @@ qla2x00_terminate_rport_io(struct fc_rport *rport)
 	}
 
 	qla2x00_abort_fcport_cmds(fcport);
-	scsi_target_unblock(&rport->dev);
 }
 
 static int

drivers/scsi/qla2xxx/qla_isr.c

@@ -1187,7 +1187,12 @@ qla2x00_status_entry(scsi_qla_host_t *ha, void *pkt)
 		    cp->serial_number, comp_status,
 		    atomic_read(&fcport->state)));
 
-		cp->result = DID_BUS_BUSY << 16;
+		/*
+		 * We are going to have the fc class block the rport
+		 * while we try to recover so instruct the mid layer
+		 * to requeue until the class decides how to handle this.
+		 */
+		cp->result = DID_TRANSPORT_DISRUPTED << 16;
 		if (atomic_read(&fcport->state) == FCS_ONLINE)
 			qla2x00_mark_device_lost(fcport->ha, fcport, 1, 1);
 		break;
@@ -1214,7 +1219,12 @@ qla2x00_status_entry(scsi_qla_host_t *ha, void *pkt)
 		break;
 
 	case CS_TIMEOUT:
-		cp->result = DID_BUS_BUSY << 16;
+		/*
+		 * We are going to have the fc class block the rport
+		 * while we try to recover so instruct the mid layer
+		 * to requeue until the class decides how to handle this.
+		 */
+		cp->result = DID_TRANSPORT_DISRUPTED << 16;
 
 		if (IS_FWI2_CAPABLE(ha)) {
 			DEBUG2(printk(KERN_INFO

drivers/scsi/qla2xxx/qla_os.c

@@ -394,10 +394,8 @@ qla2x00_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 	}
 
 	/* Close window on fcport/rport state-transitioning. */
-	if (fcport->drport) {
-		cmd->result = DID_IMM_RETRY << 16;
-		goto qc_fail_command;
-	}
+	if (fcport->drport)
+		goto qc_target_busy;
 
 	if (atomic_read(&fcport->state) != FCS_ONLINE) {
 		if (atomic_read(&fcport->state) == FCS_DEVICE_DEAD ||
@@ -405,7 +403,7 @@ qla2x00_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 			cmd->result = DID_NO_CONNECT << 16;
 			goto qc_fail_command;
 		}
-		goto qc_host_busy;
+		goto qc_target_busy;
 	}
 
 	spin_unlock_irq(ha->host->host_lock);
@@ -428,10 +426,11 @@ qc_host_busy_free_sp:
 
 qc_host_busy_lock:
 	spin_lock_irq(ha->host->host_lock);
-
-qc_host_busy:
 	return SCSI_MLQUEUE_HOST_BUSY;
 
+qc_target_busy:
+	return SCSI_MLQUEUE_TARGET_BUSY;
+
 qc_fail_command:
 	done(cmd);
 
@@ -461,10 +460,8 @@ qla24xx_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 	}
 
 	/* Close window on fcport/rport state-transitioning. */
-	if (fcport->drport) {
-		cmd->result = DID_IMM_RETRY << 16;
-		goto qc24_fail_command;
-	}
+	if (fcport->drport)
+		goto qc24_target_busy;
 
 	if (atomic_read(&fcport->state) != FCS_ONLINE) {
 		if (atomic_read(&fcport->state) == FCS_DEVICE_DEAD ||
@@ -472,7 +469,7 @@ qla24xx_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 			cmd->result = DID_NO_CONNECT << 16;
 			goto qc24_fail_command;
 		}
-		goto qc24_host_busy;
+		goto qc24_target_busy;
 	}
 
 	spin_unlock_irq(ha->host->host_lock);
@@ -495,10 +492,11 @@ qc24_host_busy_free_sp:
 
 qc24_host_busy_lock:
 	spin_lock_irq(ha->host->host_lock);
-
-qc24_host_busy:
 	return SCSI_MLQUEUE_HOST_BUSY;
 
+qc24_target_busy:
+	return SCSI_MLQUEUE_TARGET_BUSY;
+
 qc24_fail_command:
 	done(cmd);
 

drivers/scsi/qla4xxx/ql4_isr.c

@@ -139,7 +139,7 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 			      ha->host_no, cmd->device->channel,
 			      cmd->device->id, cmd->device->lun));
 
-		cmd->result = DID_BUS_BUSY << 16;
+		cmd->result = DID_TRANSPORT_DISRUPTED << 16;
 
 		/*
 		 * Mark device missing so that we won't continue to send
@@ -243,7 +243,7 @@ static void qla4xxx_status_entry(struct scsi_qla_host *ha,
 		if (atomic_read(&ddb_entry->state) == DDB_STATE_ONLINE)
 			qla4xxx_mark_device_missing(ha, ddb_entry);
 
-		cmd->result = DID_BUS_BUSY << 16;
+		cmd->result = DID_TRANSPORT_DISRUPTED << 16;
 		break;
 
 	case SCS_QUEUE_FULL:

drivers/scsi/qla4xxx/ql4_os.c

@@ -353,7 +353,7 @@ void qla4xxx_mark_device_missing(struct scsi_qla_host *ha,
 		      ha->host_no, ddb_entry->bus, ddb_entry->target,
 		      ddb_entry->fw_ddb_index));
 	iscsi_block_session(ddb_entry->sess);
-	iscsi_conn_error(ddb_entry->conn, ISCSI_ERR_CONN_FAILED);
+	iscsi_conn_error_event(ddb_entry->conn, ISCSI_ERR_CONN_FAILED);
 }
 
 static struct srb* qla4xxx_get_new_srb(struct scsi_qla_host *ha,
@@ -439,7 +439,7 @@ static int qla4xxx_queuecommand(struct scsi_cmnd *cmd,
 			cmd->result = DID_NO_CONNECT << 16;
 			goto qc_fail_command;
 		}
-		goto qc_host_busy;
+		return SCSI_MLQUEUE_TARGET_BUSY;
 	}
 
 	if (test_bit(DPC_RESET_HA_INTR, &ha->dpc_flags))

drivers/scsi/scsi.c

@@ -754,8 +754,12 @@ int scsi_dispatch_cmd(struct scsi_cmnd *cmd)
 	}
 	spin_unlock_irqrestore(host->host_lock, flags);
 	if (rtn) {
-		scsi_queue_insert(cmd, (rtn == SCSI_MLQUEUE_DEVICE_BUSY) ?
-						rtn : SCSI_MLQUEUE_HOST_BUSY);
+		if (rtn != SCSI_MLQUEUE_DEVICE_BUSY &&
+		    rtn != SCSI_MLQUEUE_TARGET_BUSY)
+			rtn = SCSI_MLQUEUE_HOST_BUSY;
+
+		scsi_queue_insert(cmd, rtn);
+
 		SCSI_LOG_MLQUEUE(3,
 		    printk("queuecommand : request rejected\n"));
 	}
@@ -800,6 +804,7 @@ static struct scsi_driver *scsi_cmd_to_driver(struct scsi_cmnd *cmd)
 void scsi_finish_command(struct scsi_cmnd *cmd)
 {
 	struct scsi_device *sdev = cmd->device;
+	struct scsi_target *starget = scsi_target(sdev);
 	struct Scsi_Host *shost = sdev->host;
 	struct scsi_driver *drv;
 	unsigned int good_bytes;
@@ -815,6 +820,7 @@ void scsi_finish_command(struct scsi_cmnd *cmd)
 	 * XXX(hch): What about locking?
          */
         shost->host_blocked = 0;
+	starget->target_blocked = 0;
         sdev->device_blocked = 0;
 
 	/*

drivers/scsi/scsi_error.c

@@ -1065,10 +1065,10 @@ static int scsi_eh_target_reset(struct Scsi_Host *shost,
 				struct list_head *done_q)
 {
 	struct scsi_cmnd *scmd, *tgtr_scmd, *next;
-	unsigned int id;
+	unsigned int id = 0;
 	int rtn;
 
-	for (id = 0; id <= shost->max_id; id++) {
+	do {
 		tgtr_scmd = NULL;
 		list_for_each_entry(scmd, work_q, eh_entry) {
 			if (id == scmd_id(scmd)) {
@@ -1076,8 +1076,18 @@ static int scsi_eh_target_reset(struct Scsi_Host *shost,
 				break;
 			}
 		}
+		if (!tgtr_scmd) {
+			/* not one exactly equal; find the next highest */
+			list_for_each_entry(scmd, work_q, eh_entry) {
+				if (scmd_id(scmd) > id &&
+				    (!tgtr_scmd ||
+				     scmd_id(tgtr_scmd) > scmd_id(scmd)))
+						tgtr_scmd = scmd;
+			}
+		}
 		if (!tgtr_scmd)
-			continue;
+			/* no more commands, that's it */
+			break;
 
 		SCSI_LOG_ERROR_RECOVERY(3, printk("%s: Sending target reset "
 						  "to target %d\n",
@@ -1096,7 +1106,8 @@ static int scsi_eh_target_reset(struct Scsi_Host *shost,
 							  " failed target: "
 							  "%d\n",
 							  current->comm, id));
-	}
+		id++;
+	} while(id != 0);
 
 	return list_empty(work_q);
 }
@@ -1218,6 +1229,40 @@ static void scsi_eh_offline_sdevs(struct list_head *work_q,
 	return;
 }
 
+/**
+ * scsi_noretry_cmd - determinte if command should be failed fast
+ * @scmd:	SCSI cmd to examine.
+ */
+int scsi_noretry_cmd(struct scsi_cmnd *scmd)
+{
+	switch (host_byte(scmd->result)) {
+	case DID_OK:
+		break;
+	case DID_BUS_BUSY:
+		return blk_failfast_transport(scmd->request);
+	case DID_PARITY:
+		return blk_failfast_dev(scmd->request);
+	case DID_ERROR:
+		if (msg_byte(scmd->result) == COMMAND_COMPLETE &&
+		    status_byte(scmd->result) == RESERVATION_CONFLICT)
+			return 0;
+		/* fall through */
+	case DID_SOFT_ERROR:
+		return blk_failfast_driver(scmd->request);
+	}
+
+	switch (status_byte(scmd->result)) {
+	case CHECK_CONDITION:
+		/*
+		 * assume caller has checked sense and determinted
+		 * the check condition was retryable.
+		 */
+		return blk_failfast_dev(scmd->request);
+	}
+
+	return 0;
+}
+
 /**
  * scsi_decide_disposition - Disposition a cmd on return from LLD.
  * @scmd:	SCSI cmd to examine.
@@ -1290,7 +1335,20 @@ int scsi_decide_disposition(struct scsi_cmnd *scmd)
 
 	case DID_REQUEUE:
 		return ADD_TO_MLQUEUE;
-
+	case DID_TRANSPORT_DISRUPTED:
+		/*
+		 * LLD/transport was disrupted during processing of the IO.
+		 * The transport class is now blocked/blocking,
+		 * and the transport will decide what to do with the IO
+		 * based on its timers and recovery capablilities.
+		 */
+		return ADD_TO_MLQUEUE;
+	case DID_TRANSPORT_FAILFAST:
+		/*
+		 * The transport decided to failfast the IO (most likely
+		 * the fast io fail tmo fired), so send IO directly upwards.
+		 */
+		return SUCCESS;
 	case DID_ERROR:
 		if (msg_byte(scmd->result) == COMMAND_COMPLETE &&
 		    status_byte(scmd->result) == RESERVATION_CONFLICT)
@@ -1383,7 +1441,7 @@ int scsi_decide_disposition(struct scsi_cmnd *scmd)
 	 * even if the request is marked fast fail, we still requeue
 	 * for queue congestion conditions (QUEUE_FULL or BUSY) */
 	if ((++scmd->retries) <= scmd->allowed
-	    && !blk_noretry_request(scmd->request)) {
+	    && !scsi_noretry_cmd(scmd)) {
 		return NEEDS_RETRY;
 	} else {
 		/*
@@ -1508,7 +1566,7 @@ void scsi_eh_flush_done_q(struct list_head *done_q)
 	list_for_each_entry_safe(scmd, next, done_q, eh_entry) {
 		list_del_init(&scmd->eh_entry);
 		if (scsi_device_online(scmd->device) &&
-		    !blk_noretry_request(scmd->request) &&
+		    !scsi_noretry_cmd(scmd) &&
 		    (++scmd->retries <= scmd->allowed)) {
 			SCSI_LOG_ERROR_RECOVERY(3, printk("%s: flush"
 							  " retry cmd: %p\n",

drivers/scsi/scsi_lib.c

@@ -114,6 +114,7 @@ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason)
 {
 	struct Scsi_Host *host = cmd->device->host;
 	struct scsi_device *device = cmd->device;
+	struct scsi_target *starget = scsi_target(device);
 	struct request_queue *q = device->request_queue;
 	unsigned long flags;
 
@@ -133,10 +134,17 @@ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason)
 	 * if a command is requeued with no other commands outstanding
 	 * either for the device or for the host.
 	 */
-	if (reason == SCSI_MLQUEUE_HOST_BUSY)
+	switch (reason) {
+	case SCSI_MLQUEUE_HOST_BUSY:
 		host->host_blocked = host->max_host_blocked;
-	else if (reason == SCSI_MLQUEUE_DEVICE_BUSY)
+		break;
+	case SCSI_MLQUEUE_DEVICE_BUSY:
 		device->device_blocked = device->max_device_blocked;
+		break;
+	case SCSI_MLQUEUE_TARGET_BUSY:
+		starget->target_blocked = starget->max_target_blocked;
+		break;
+	}
 
 	/*
 	 * Decrement the counters, since these commands are no longer
@@ -460,10 +468,12 @@ static void scsi_init_cmd_errh(struct scsi_cmnd *cmd)
 void scsi_device_unbusy(struct scsi_device *sdev)
 {
 	struct Scsi_Host *shost = sdev->host;
+	struct scsi_target *starget = scsi_target(sdev);
 	unsigned long flags;
 
 	spin_lock_irqsave(shost->host_lock, flags);
 	shost->host_busy--;
+	starget->target_busy--;
 	if (unlikely(scsi_host_in_recovery(shost) &&
 		     (shost->host_failed || shost->host_eh_scheduled)))
 		scsi_eh_wakeup(shost);
@@ -519,6 +529,13 @@ static void scsi_single_lun_run(struct scsi_device *current_sdev)
 	spin_unlock_irqrestore(shost->host_lock, flags);
 }
 
+static inline int scsi_target_is_busy(struct scsi_target *starget)
+{
+	return ((starget->can_queue > 0 &&
+		 starget->target_busy >= starget->can_queue) ||
+		 starget->target_blocked);
+}
+
 /*
  * Function:	scsi_run_queue()
  *
@@ -533,7 +550,7 @@ static void scsi_single_lun_run(struct scsi_device *current_sdev)
  */
 static void scsi_run_queue(struct request_queue *q)
 {
-	struct scsi_device *sdev = q->queuedata;
+	struct scsi_device *starved_head = NULL, *sdev = q->queuedata;
 	struct Scsi_Host *shost = sdev->host;
 	unsigned long flags;
 
@@ -560,6 +577,21 @@ static void scsi_run_queue(struct request_queue *q)
 		 */
 		sdev = list_entry(shost->starved_list.next,
 					  struct scsi_device, starved_entry);
+		/*
+		 * The *queue_ready functions can add a device back onto the
+		 * starved list's tail, so we must check for a infinite loop.
+		 */
+		if (sdev == starved_head)
+			break;
+		if (!starved_head)
+			starved_head = sdev;
+
+		if (scsi_target_is_busy(scsi_target(sdev))) {
+			list_move_tail(&sdev->starved_entry,
+				       &shost->starved_list);
+			continue;
+		}
+
 		list_del_init(&sdev->starved_entry);
 		spin_unlock(shost->host_lock);
 
@@ -575,13 +607,6 @@ static void scsi_run_queue(struct request_queue *q)
 		spin_unlock(sdev->request_queue->queue_lock);
 
 		spin_lock(shost->host_lock);
-		if (unlikely(!list_empty(&sdev->starved_entry)))
-			/*
-			 * sdev lost a race, and was put back on the
-			 * starved list. This is unlikely but without this
-			 * in theory we could loop forever.
-			 */
-			break;
 	}
 	spin_unlock_irqrestore(shost->host_lock, flags);
 
@@ -681,7 +706,7 @@ static struct scsi_cmnd *scsi_end_request(struct scsi_cmnd *cmd, int error,
 			leftover = req->data_len;
 
 		/* kill remainder if no retrys */
-		if (error && blk_noretry_request(req))
+		if (error && scsi_noretry_cmd(cmd))
 			blk_end_request(req, error, leftover);
 		else {
 			if (requeue) {
@@ -1344,6 +1369,52 @@ static inline int scsi_dev_queue_ready(struct request_queue *q,
 	return 1;
 }
 
+
+/*
+ * scsi_target_queue_ready: checks if there we can send commands to target
+ * @sdev: scsi device on starget to check.
+ *
+ * Called with the host lock held.
+ */
+static inline int scsi_target_queue_ready(struct Scsi_Host *shost,
+					   struct scsi_device *sdev)
+{
+	struct scsi_target *starget = scsi_target(sdev);
+
+	if (starget->single_lun) {
+		if (starget->starget_sdev_user &&
+		    starget->starget_sdev_user != sdev)
+			return 0;
+		starget->starget_sdev_user = sdev;
+	}
+
+	if (starget->target_busy == 0 && starget->target_blocked) {
+		/*
+		 * unblock after target_blocked iterates to zero
+		 */
+		if (--starget->target_blocked == 0) {
+			SCSI_LOG_MLQUEUE(3, starget_printk(KERN_INFO, starget,
+					 "unblocking target at zero depth\n"));
+		} else {
+			blk_plug_device(sdev->request_queue);
+			return 0;
+		}
+	}
+
+	if (scsi_target_is_busy(starget)) {
+		if (list_empty(&sdev->starved_entry)) {
+			list_add_tail(&sdev->starved_entry,
+				      &shost->starved_list);
+			return 0;
+		}
+	}
+
+	/* We're OK to process the command, so we can't be starved */
+	if (!list_empty(&sdev->starved_entry))
+		list_del_init(&sdev->starved_entry);
+	return 1;
+}
+
 /*
  * scsi_host_queue_ready: if we can send requests to shost, return 1 else
  * return 0. We must end up running the queue again whenever 0 is
@@ -1390,6 +1461,7 @@ static void scsi_kill_request(struct request *req, struct request_queue *q)
 {
 	struct scsi_cmnd *cmd = req->special;
 	struct scsi_device *sdev = cmd->device;
+	struct scsi_target *starget = scsi_target(sdev);
 	struct Scsi_Host *shost = sdev->host;
 
 	blkdev_dequeue_request(req);
@@ -1413,6 +1485,7 @@ static void scsi_kill_request(struct request *req, struct request_queue *q)
 	spin_unlock(sdev->request_queue->queue_lock);
 	spin_lock(shost->host_lock);
 	shost->host_busy++;
+	starget->target_busy++;
 	spin_unlock(shost->host_lock);
 	spin_lock(sdev->request_queue->queue_lock);
 
@@ -1550,14 +1623,13 @@ static void scsi_request_fn(struct request_queue *q)
 			goto not_ready;
 		}
 
+		if (!scsi_target_queue_ready(shost, sdev))
+			goto not_ready;
+
 		if (!scsi_host_queue_ready(q, shost, sdev))
 			goto not_ready;
-		if (scsi_target(sdev)->single_lun) {
-			if (scsi_target(sdev)->starget_sdev_user &&
-			    scsi_target(sdev)->starget_sdev_user != sdev)
-				goto not_ready;
-			scsi_target(sdev)->starget_sdev_user = sdev;
-		}
+
+		scsi_target(sdev)->target_busy++;
 		shost->host_busy++;
 
 		/*

drivers/scsi/scsi_priv.h

@@ -59,6 +59,7 @@ void scsi_eh_ready_devs(struct Scsi_Host *shost,
 			struct list_head *done_q);
 int scsi_eh_get_sense(struct list_head *work_q,
 		      struct list_head *done_q);
+int scsi_noretry_cmd(struct scsi_cmnd *scmd);
 
 /* scsi_lib.c */
 extern int scsi_maybe_unblock_host(struct scsi_device *sdev);

drivers/scsi/scsi_scan.c

@@ -419,6 +419,7 @@ static struct scsi_target *scsi_alloc_target(struct device *parent,
 	dev->type = &scsi_target_type;
 	starget->id = id;
 	starget->channel = channel;
+	starget->can_queue = 0;
 	INIT_LIST_HEAD(&starget->siblings);
 	INIT_LIST_HEAD(&starget->devices);
 	starget->state = STARGET_CREATED;

drivers/scsi/scsi_transport_fc.c

@@ -2133,8 +2133,7 @@ fc_attach_transport(struct fc_function_template *ft)
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(roles);
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(port_state);
 	SETUP_PRIVATE_RPORT_ATTRIBUTE_RD(scsi_target_id);
-	if (ft->terminate_rport_io)
-		SETUP_PRIVATE_RPORT_ATTRIBUTE_RW(fast_io_fail_tmo);
+	SETUP_PRIVATE_RPORT_ATTRIBUTE_RW(fast_io_fail_tmo);
 
 	BUG_ON(count > FC_RPORT_NUM_ATTRS);
 
@@ -2328,6 +2327,22 @@ fc_remove_host(struct Scsi_Host *shost)
 }
 EXPORT_SYMBOL(fc_remove_host);
 
+static void fc_terminate_rport_io(struct fc_rport *rport)
+{
+	struct Scsi_Host *shost = rport_to_shost(rport);
+	struct fc_internal *i = to_fc_internal(shost->transportt);
+
+	/* Involve the LLDD if possible to terminate all io on the rport. */
+	if (i->f->terminate_rport_io)
+		i->f->terminate_rport_io(rport);
+
+	/*
+	 * must unblock to flush queued IO. The caller will have set
+	 * the port_state or flags, so that fc_remote_port_chkready will
+	 * fail IO.
+	 */
+	scsi_target_unblock(&rport->dev);
+}
 
 /**
  * fc_starget_delete - called to delete the scsi decendents of an rport
@@ -2340,13 +2355,8 @@ fc_starget_delete(struct work_struct *work)
 {
 	struct fc_rport *rport =
 		container_of(work, struct fc_rport, stgt_delete_work);
-	struct Scsi_Host *shost = rport_to_shost(rport);
-	struct fc_internal *i = to_fc_internal(shost->transportt);
-
-	/* Involve the LLDD if possible to terminate all io on the rport. */
-	if (i->f->terminate_rport_io)
-		i->f->terminate_rport_io(rport);
 
+	fc_terminate_rport_io(rport);
 	scsi_remove_target(&rport->dev);
 }
 
@@ -2372,10 +2382,7 @@ fc_rport_final_delete(struct work_struct *work)
 	if (rport->flags & FC_RPORT_SCAN_PENDING)
 		scsi_flush_work(shost);
 
-	/* involve the LLDD to terminate all pending i/o */
-	if (i->f->terminate_rport_io)
-		i->f->terminate_rport_io(rport);
-
+	fc_terminate_rport_io(rport);
 	/*
 	 * Cancel any outstanding timers. These should really exist
 	 * only when rmmod'ing the LLDD and we're asking for
@@ -2639,7 +2646,8 @@ fc_remote_port_add(struct Scsi_Host *shost, int channel,
 
 				spin_lock_irqsave(shost->host_lock, flags);
 
-				rport->flags &= ~FC_RPORT_DEVLOSS_PENDING;
+				rport->flags &= ~(FC_RPORT_FAST_FAIL_TIMEDOUT |
+						  FC_RPORT_DEVLOSS_PENDING);
 
 				/* if target, initiate a scan */
 				if (rport->scsi_target_id != -1) {
@@ -2702,6 +2710,7 @@ fc_remote_port_add(struct Scsi_Host *shost, int channel,
 			rport->port_id = ids->port_id;
 			rport->roles = ids->roles;
 			rport->port_state = FC_PORTSTATE_ONLINE;
+			rport->flags &= ~FC_RPORT_FAST_FAIL_TIMEDOUT;
 
 			if (fci->f->dd_fcrport_size)
 				memset(rport->dd_data, 0,
@@ -2784,7 +2793,6 @@ void
 fc_remote_port_delete(struct fc_rport  *rport)
 {
 	struct Scsi_Host *shost = rport_to_shost(rport);
-	struct fc_internal *i = to_fc_internal(shost->transportt);
 	int timeout = rport->dev_loss_tmo;
 	unsigned long flags;
 
@@ -2830,7 +2838,7 @@ fc_remote_port_delete(struct fc_rport  *rport)
 
 	/* see if we need to kill io faster than waiting for device loss */
 	if ((rport->fast_io_fail_tmo != -1) &&
-	    (rport->fast_io_fail_tmo < timeout) && (i->f->terminate_rport_io))
+	    (rport->fast_io_fail_tmo < timeout))
 		fc_queue_devloss_work(shost, &rport->fail_io_work,
 					rport->fast_io_fail_tmo * HZ);
 
@@ -2906,7 +2914,8 @@ fc_remote_port_rolechg(struct fc_rport  *rport, u32 roles)
 			fc_flush_devloss(shost);
 
 		spin_lock_irqsave(shost->host_lock, flags);
-		rport->flags &= ~FC_RPORT_DEVLOSS_PENDING;
+		rport->flags &= ~(FC_RPORT_FAST_FAIL_TIMEDOUT |
+				  FC_RPORT_DEVLOSS_PENDING);
 		spin_unlock_irqrestore(shost->host_lock, flags);
 
 		/* ensure any stgt delete functions are done */
@@ -3001,6 +3010,7 @@ fc_timeout_deleted_rport(struct work_struct *work)
 	rport->supported_classes = FC_COS_UNSPECIFIED;
 	rport->roles = FC_PORT_ROLE_UNKNOWN;
 	rport->port_state = FC_PORTSTATE_NOTPRESENT;
+	rport->flags &= ~FC_RPORT_FAST_FAIL_TIMEDOUT;
 
 	/* remove the identifiers that aren't used in the consisting binding */
 	switch (fc_host->tgtid_bind_type) {
@@ -3043,13 +3053,12 @@ fc_timeout_fail_rport_io(struct work_struct *work)
 {
 	struct fc_rport *rport =
 		container_of(work, struct fc_rport, fail_io_work.work);
-	struct Scsi_Host *shost = rport_to_shost(rport);
-	struct fc_internal *i = to_fc_internal(shost->transportt);
 
 	if (rport->port_state != FC_PORTSTATE_BLOCKED)
 		return;
 
-	i->f->terminate_rport_io(rport);
+	rport->flags |= FC_RPORT_FAST_FAIL_TIMEDOUT;
+	fc_terminate_rport_io(rport);
 }
 
 /**

drivers/scsi/scsi_transport_iscsi.c

@@ -138,7 +138,7 @@ static ssize_t
 show_ep_handle(struct device *dev, struct device_attribute *attr, char *buf)
 {
 	struct iscsi_endpoint *ep = iscsi_dev_to_endpoint(dev);
-	return sprintf(buf, "%u\n", ep->id);
+	return sprintf(buf, "%llu\n", (unsigned long long) ep->id);
 }
 static ISCSI_ATTR(ep, handle, S_IRUGO, show_ep_handle, NULL);
 
@@ -156,7 +156,7 @@ static struct attribute_group iscsi_endpoint_group = {
 static int iscsi_match_epid(struct device *dev, void *data)
 {
 	struct iscsi_endpoint *ep = iscsi_dev_to_endpoint(dev);
-	unsigned int *epid = (unsigned int *) data;
+	uint64_t *epid = (uint64_t *) data;
 
 	return *epid == ep->id;
 }
@@ -166,7 +166,7 @@ iscsi_create_endpoint(int dd_size)
 {
 	struct device *dev;
 	struct iscsi_endpoint *ep;
-	unsigned int id;
+	uint64_t id;
 	int err;
 
 	for (id = 1; id < ISCSI_MAX_EPID; id++) {
@@ -187,7 +187,8 @@ iscsi_create_endpoint(int dd_size)
 
 	ep->id = id;
 	ep->dev.class = &iscsi_endpoint_class;
-	snprintf(ep->dev.bus_id, BUS_ID_SIZE, "ep-%u", id);
+	snprintf(ep->dev.bus_id, BUS_ID_SIZE, "ep-%llu",
+		 (unsigned long long) id);
 	err = device_register(&ep->dev);
         if (err)
                 goto free_ep;
@@ -374,10 +375,10 @@ int iscsi_session_chkready(struct iscsi_cls_session *session)
 		err = 0;
 		break;
 	case ISCSI_SESSION_FAILED:
-		err = DID_IMM_RETRY << 16;
+		err = DID_TRANSPORT_DISRUPTED << 16;
 		break;
 	case ISCSI_SESSION_FREE:
-		err = DID_NO_CONNECT << 16;
+		err = DID_TRANSPORT_FAILFAST << 16;
 		break;
 	default:
 		err = DID_NO_CONNECT << 16;
@@ -1010,7 +1011,7 @@ int iscsi_recv_pdu(struct iscsi_cls_conn *conn, struct iscsi_hdr *hdr,
 
 	skb = alloc_skb(len, GFP_ATOMIC);
 	if (!skb) {
-		iscsi_conn_error(conn, ISCSI_ERR_CONN_FAILED);
+		iscsi_conn_error_event(conn, ISCSI_ERR_CONN_FAILED);
 		iscsi_cls_conn_printk(KERN_ERR, conn, "can not deliver "
 				      "control PDU: OOM\n");
 		return -ENOMEM;
@@ -1031,7 +1032,7 @@ int iscsi_recv_pdu(struct iscsi_cls_conn *conn, struct iscsi_hdr *hdr,
 }
 EXPORT_SYMBOL_GPL(iscsi_recv_pdu);
 
-void iscsi_conn_error(struct iscsi_cls_conn *conn, enum iscsi_err error)
+void iscsi_conn_error_event(struct iscsi_cls_conn *conn, enum iscsi_err error)
 {
 	struct nlmsghdr	*nlh;
 	struct sk_buff	*skb;
@@ -1063,7 +1064,7 @@ void iscsi_conn_error(struct iscsi_cls_conn *conn, enum iscsi_err error)
 	iscsi_cls_conn_printk(KERN_INFO, conn, "detected conn error (%d)\n",
 			      error);
 }
-EXPORT_SYMBOL_GPL(iscsi_conn_error);
+EXPORT_SYMBOL_GPL(iscsi_conn_error_event);
 
 static int
 iscsi_if_send_reply(int pid, int seq, int type, int done, int multi,

drivers/scsi/scsi_transport_spi.c

@@ -109,7 +109,9 @@ static int spi_execute(struct scsi_device *sdev, const void *cmd,
 	for(i = 0; i < DV_RETRIES; i++) {
 		result = scsi_execute(sdev, cmd, dir, buffer, bufflen,
 				      sense, DV_TIMEOUT, /* retries */ 1,
-				      REQ_FAILFAST);
+				      REQ_FAILFAST_DEV |
+				      REQ_FAILFAST_TRANSPORT |
+				      REQ_FAILFAST_DRIVER);
 		if (result & DRIVER_SENSE) {
 			struct scsi_sense_hdr sshdr_tmp;
 			if (!sshdr)

drivers/scsi/sd.c

@@ -384,7 +384,7 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq)
 	sector_t block = rq->sector;
 	sector_t threshold;
 	unsigned int this_count = rq->nr_sectors;
-	int ret;
+	int ret, host_dif;
 
 	if (rq->cmd_type == REQ_TYPE_BLOCK_PC) {
 		ret = scsi_setup_blk_pc_cmnd(sdp, rq);
@@ -515,7 +515,8 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq)
 					rq->nr_sectors));
 
 	/* Set RDPROTECT/WRPROTECT if disk is formatted with DIF */
-	if (scsi_host_dif_capable(sdp->host, sdkp->protection_type))
+	host_dif = scsi_host_dif_capable(sdp->host, sdkp->protection_type);
+	if (host_dif)
 		SCpnt->cmnd[1] = 1 << 5;
 	else
 		SCpnt->cmnd[1] = 0;
@@ -573,8 +574,9 @@ static int sd_prep_fn(struct request_queue *q, struct request *rq)
 	SCpnt->sdb.length = this_count * sdp->sector_size;
 
 	/* If DIF or DIX is enabled, tell HBA how to handle request */
-	if (sdkp->protection_type || scsi_prot_sg_count(SCpnt))
-		sd_dif_op(SCpnt, sdkp->protection_type, scsi_prot_sg_count(SCpnt));
+	if (host_dif || scsi_prot_sg_count(SCpnt))
+		sd_dif_op(SCpnt, host_dif, scsi_prot_sg_count(SCpnt),
+			  sdkp->protection_type);
 
 	/*
 	 * We shouldn't disconnect in the middle of a sector, so with a dumb
@@ -1252,14 +1254,12 @@ void sd_read_protection_type(struct scsi_disk *sdkp, unsigned char *buffer)
 	else
 		type = ((buffer[12] >> 1) & 7) + 1; /* P_TYPE 0 = Type 1 */
 
+	sdkp->protection_type = type;
+
 	switch (type) {
 	case SD_DIF_TYPE0_PROTECTION:
-		sdkp->protection_type = 0;
-		break;
-
 	case SD_DIF_TYPE1_PROTECTION:
 	case SD_DIF_TYPE3_PROTECTION:
-		sdkp->protection_type = type;
 		break;
 
 	case SD_DIF_TYPE2_PROTECTION:
@@ -1277,7 +1277,6 @@ void sd_read_protection_type(struct scsi_disk *sdkp, unsigned char *buffer)
 	return;
 
 disable:
-	sdkp->protection_type = 0;
 	sdkp->capacity = 0;
 }
 

drivers/scsi/sd.h

@@ -97,19 +97,28 @@ struct sd_dif_tuple {
        __be32 ref_tag;		/* Target LBA or indirect LBA */
 };
 
-#if defined(CONFIG_BLK_DEV_INTEGRITY)
+#ifdef CONFIG_BLK_DEV_INTEGRITY
 
-extern void sd_dif_op(struct scsi_cmnd *, unsigned int, unsigned int);
+extern void sd_dif_op(struct scsi_cmnd *, unsigned int, unsigned int, unsigned int);
 extern void sd_dif_config_host(struct scsi_disk *);
 extern int sd_dif_prepare(struct request *rq, sector_t, unsigned int);
 extern void sd_dif_complete(struct scsi_cmnd *, unsigned int);
 
 #else /* CONFIG_BLK_DEV_INTEGRITY */
 
-#define sd_dif_op(a, b, c)			do { } while (0)
-#define sd_dif_config_host(a)			do { } while (0)
-#define sd_dif_prepare(a, b, c)			(0)
-#define sd_dif_complete(a, b)			(0)
+static inline void sd_dif_op(struct scsi_cmnd *cmd, unsigned int a, unsigned int b, unsigned int c)
+{
+}
+static inline void sd_dif_config_host(struct scsi_disk *disk)
+{
+}
+static inline int sd_dif_prepare(struct request *rq, sector_t s, unsigned int a)
+{
+	return 0;
+}
+static inline void sd_dif_complete(struct scsi_cmnd *cmd, unsigned int a)
+{
+}
 
 #endif /* CONFIG_BLK_DEV_INTEGRITY */
 

drivers/scsi/sd_dif.c

@@ -311,25 +311,26 @@ void sd_dif_config_host(struct scsi_disk *sdkp)
 	struct scsi_device *sdp = sdkp->device;
 	struct gendisk *disk = sdkp->disk;
 	u8 type = sdkp->protection_type;
+	int dif, dix;
 
-	/* If this HBA doesn't support DIX, resort to normal I/O or DIF */
-	if (scsi_host_dix_capable(sdp->host, type) == 0) {
+	dif = scsi_host_dif_capable(sdp->host, type);
+	dix = scsi_host_dix_capable(sdp->host, type);
 
-		if (type == SD_DIF_TYPE0_PROTECTION)
-			return;
-
-		if (scsi_host_dif_capable(sdp->host, type) == 0) {
-			sd_printk(KERN_INFO, sdkp, "Type %d protection " \
-				  "unsupported by HBA. Disabling DIF.\n", type);
-			sdkp->protection_type = 0;
-			return;
-		}
+	if (!dix && scsi_host_dix_capable(sdp->host, 0)) {
+		dif = 0; dix = 1;
+	}
 
-		sd_printk(KERN_INFO, sdkp, "Enabling DIF Type %d protection\n",
-			  type);
+	if (type) {
+		if (dif)
+			sd_printk(KERN_NOTICE, sdkp,
+				  "Enabling DIF Type %d protection\n", type);
+		else
+			sd_printk(KERN_NOTICE, sdkp,
+				  "Disabling DIF Type %d protection\n", type);
+	}
 
+	if (!dix)
 		return;
-	}
 
 	/* Enable DMA of protection information */
 	if (scsi_host_get_guard(sdkp->device->host) & SHOST_DIX_GUARD_IP)
@@ -343,17 +344,17 @@ void sd_dif_config_host(struct scsi_disk *sdkp)
 		else
 			blk_integrity_register(disk, &dif_type1_integrity_crc);
 
-	sd_printk(KERN_INFO, sdkp,
-		  "Enabling %s integrity protection\n", disk->integrity->name);
+	sd_printk(KERN_NOTICE, sdkp,
+		  "Enabling DIX %s protection\n", disk->integrity->name);
 
 	/* Signal to block layer that we support sector tagging */
-	if (type && sdkp->ATO) {
+	if (dif && type && sdkp->ATO) {
 		if (type == SD_DIF_TYPE3_PROTECTION)
 			disk->integrity->tag_size = sizeof(u16) + sizeof(u32);
 		else
 			disk->integrity->tag_size = sizeof(u16);
 
-		sd_printk(KERN_INFO, sdkp, "DIF application tag size %u\n",
+		sd_printk(KERN_NOTICE, sdkp, "DIF application tag size %u\n",
 			  disk->integrity->tag_size);
 	}
 }
@@ -361,7 +362,7 @@ void sd_dif_config_host(struct scsi_disk *sdkp)
 /*
  * DIF DMA operation magic decoder ring.
  */
-void sd_dif_op(struct scsi_cmnd *scmd, unsigned int dif, unsigned int dix)
+void sd_dif_op(struct scsi_cmnd *scmd, unsigned int dif, unsigned int dix, unsigned int type)
 {
 	int csum_convert, prot_op;
 
@@ -406,7 +407,8 @@ void sd_dif_op(struct scsi_cmnd *scmd, unsigned int dif, unsigned int dix)
 	}
 
 	scsi_set_prot_op(scmd, prot_op);
-	scsi_set_prot_type(scmd, dif);
+	if (dif)
+		scsi_set_prot_type(scmd, type);
 }
 
 /*

include/linux/bio.h

@@ -129,25 +129,30 @@ struct bio {
  * bit 2 -- barrier
  *	Insert a serialization point in the IO queue, forcing previously
  *	submitted IO to be completed before this oen is issued.
- * bit 3 -- fail fast, don't want low level driver retries
- * bit 4 -- synchronous I/O hint: the block layer will unplug immediately
+ * bit 3 -- synchronous I/O hint: the block layer will unplug immediately
  *	Note that this does NOT indicate that the IO itself is sync, just
  *	that the block layer will not postpone issue of this IO by plugging.
- * bit 5 -- metadata request
+ * bit 4 -- metadata request
  *	Used for tracing to differentiate metadata and data IO. May also
  *	get some preferential treatment in the IO scheduler
- * bit 6 -- discard sectors
+ * bit 5 -- discard sectors
  *	Informs the lower level device that this range of sectors is no longer
  *	used by the file system and may thus be freed by the device. Used
  *	for flash based storage.
+ * bit 6 -- fail fast device errors
+ * bit 7 -- fail fast transport errors
+ * bit 8 -- fail fast driver errors
+ *	Don't want driver retries for any fast fail whatever the reason.
  */
 #define BIO_RW		0	/* Must match RW in req flags (blkdev.h) */
 #define BIO_RW_AHEAD	1	/* Must match FAILFAST in req flags */
 #define BIO_RW_BARRIER	2
-#define BIO_RW_FAILFAST	3
-#define BIO_RW_SYNC	4
-#define BIO_RW_META	5
-#define BIO_RW_DISCARD	6
+#define BIO_RW_SYNC	3
+#define BIO_RW_META	4
+#define BIO_RW_DISCARD	5
+#define BIO_RW_FAILFAST_DEV		6
+#define BIO_RW_FAILFAST_TRANSPORT	7
+#define BIO_RW_FAILFAST_DRIVER		8
 
 /*
  * upper 16 bits of bi_rw define the io priority of this bio
@@ -174,7 +179,10 @@ struct bio {
 #define bio_sectors(bio)	((bio)->bi_size >> 9)
 #define bio_barrier(bio)	((bio)->bi_rw & (1 << BIO_RW_BARRIER))
 #define bio_sync(bio)		((bio)->bi_rw & (1 << BIO_RW_SYNC))
-#define bio_failfast(bio)	((bio)->bi_rw & (1 << BIO_RW_FAILFAST))
+#define bio_failfast_dev(bio)	((bio)->bi_rw &	(1 << BIO_RW_FAILFAST_DEV))
+#define bio_failfast_transport(bio)	\
+	((bio)->bi_rw & (1 << BIO_RW_FAILFAST_TRANSPORT))
+#define bio_failfast_driver(bio) ((bio)->bi_rw & (1 << BIO_RW_FAILFAST_DRIVER))
 #define bio_rw_ahead(bio)	((bio)->bi_rw & (1 << BIO_RW_AHEAD))
 #define bio_rw_meta(bio)	((bio)->bi_rw & (1 << BIO_RW_META))
 #define bio_discard(bio)	((bio)->bi_rw & (1 << BIO_RW_DISCARD))

include/linux/blkdev.h

@@ -87,7 +87,9 @@ enum {
  */
 enum rq_flag_bits {
 	__REQ_RW,		/* not set, read. set, write */
-	__REQ_FAILFAST,		/* no low level driver retries */
+	__REQ_FAILFAST_DEV,	/* no driver retries of device errors */
+	__REQ_FAILFAST_TRANSPORT, /* no driver retries of transport errors */
+	__REQ_FAILFAST_DRIVER,	/* no driver retries of driver errors */
 	__REQ_DISCARD,		/* request to discard sectors */
 	__REQ_SORTED,		/* elevator knows about this request */
 	__REQ_SOFTBARRIER,	/* may not be passed by ioscheduler */
@@ -111,8 +113,10 @@ enum rq_flag_bits {
 };
 
 #define REQ_RW		(1 << __REQ_RW)
+#define REQ_FAILFAST_DEV	(1 << __REQ_FAILFAST_DEV)
+#define REQ_FAILFAST_TRANSPORT	(1 << __REQ_FAILFAST_TRANSPORT)
+#define REQ_FAILFAST_DRIVER	(1 << __REQ_FAILFAST_DRIVER)
 #define REQ_DISCARD	(1 << __REQ_DISCARD)
-#define REQ_FAILFAST	(1 << __REQ_FAILFAST)
 #define REQ_SORTED	(1 << __REQ_SORTED)
 #define REQ_SOFTBARRIER	(1 << __REQ_SOFTBARRIER)
 #define REQ_HARDBARRIER	(1 << __REQ_HARDBARRIER)
@@ -560,7 +564,12 @@ enum {
 #define blk_special_request(rq)	((rq)->cmd_type == REQ_TYPE_SPECIAL)
 #define blk_sense_request(rq)	((rq)->cmd_type == REQ_TYPE_SENSE)
 
-#define blk_noretry_request(rq)	((rq)->cmd_flags & REQ_FAILFAST)
+#define blk_failfast_dev(rq)	((rq)->cmd_flags & REQ_FAILFAST_DEV)
+#define blk_failfast_transport(rq) ((rq)->cmd_flags & REQ_FAILFAST_TRANSPORT)
+#define blk_failfast_driver(rq)	((rq)->cmd_flags & REQ_FAILFAST_DRIVER)
+#define blk_noretry_request(rq)	(blk_failfast_dev(rq) ||	\
+				 blk_failfast_transport(rq) ||	\
+				 blk_failfast_driver(rq))
 #define blk_rq_started(rq)	((rq)->cmd_flags & REQ_STARTED)
 
 #define blk_account_rq(rq)	(blk_rq_started(rq) && (blk_fs_request(rq) || blk_discard_rq(rq))) 

include/scsi/iscsi_if.h

@@ -213,6 +213,8 @@ enum iscsi_err {
 	ISCSI_ERR_DATA_DGST		= ISCSI_ERR_BASE + 15,
 	ISCSI_ERR_PARAM_NOT_FOUND	= ISCSI_ERR_BASE + 16,
 	ISCSI_ERR_NO_SCSI_CMD		= ISCSI_ERR_BASE + 17,
+	ISCSI_ERR_INVALID_HOST		= ISCSI_ERR_BASE + 18,
+	ISCSI_ERR_XMIT_FAILED		= ISCSI_ERR_BASE + 19,
 };
 
 /*

include/scsi/libiscsi.h

@@ -287,6 +287,11 @@ struct iscsi_session {
 	struct iscsi_pool	cmdpool;	/* PDU's pool */
 };
 
+enum {
+	ISCSI_HOST_SETUP,
+	ISCSI_HOST_REMOVED,
+};
+
 struct iscsi_host {
 	char			*initiatorname;
 	/* hw address or netdev iscsi connection is bound to */
@@ -295,6 +300,12 @@ struct iscsi_host {
 	/* local address */
 	int			local_port;
 	char			local_address[ISCSI_ADDRESS_BUF_LEN];
+
+	wait_queue_head_t	session_removal_wq;
+	/* protects sessions and state */
+	spinlock_t		lock;
+	int			num_sessions;
+	int			state;
 };
 
 /*
@@ -302,7 +313,7 @@ struct iscsi_host {
  */
 extern int iscsi_change_queue_depth(struct scsi_device *sdev, int depth);
 extern int iscsi_eh_abort(struct scsi_cmnd *sc);
-extern int iscsi_eh_host_reset(struct scsi_cmnd *sc);
+extern int iscsi_eh_target_reset(struct scsi_cmnd *sc);
 extern int iscsi_eh_device_reset(struct scsi_cmnd *sc);
 extern int iscsi_queuecommand(struct scsi_cmnd *sc,
 			      void (*done)(struct scsi_cmnd *));
@@ -351,6 +362,8 @@ extern void iscsi_conn_stop(struct iscsi_cls_conn *, int);
 extern int iscsi_conn_bind(struct iscsi_cls_session *, struct iscsi_cls_conn *,
 			   int);
 extern void iscsi_conn_failure(struct iscsi_conn *conn, enum iscsi_err err);
+extern void iscsi_session_failure(struct iscsi_cls_session *cls_session,
+				  enum iscsi_err err);
 extern int iscsi_conn_get_param(struct iscsi_cls_conn *cls_conn,
 				enum iscsi_param param, char *buf);
 extern void iscsi_suspend_tx(struct iscsi_conn *conn);

include/scsi/scsi.h

@@ -381,6 +381,11 @@ static inline int scsi_is_wlun(unsigned int lun)
 #define DID_IMM_RETRY   0x0c	/* Retry without decrementing retry count  */
 #define DID_REQUEUE	0x0d	/* Requeue command (no immediate retry) also
 				 * without decrementing the retry count	   */
+#define DID_TRANSPORT_DISRUPTED 0x0e /* Transport error disrupted execution
+				      * and the driver blocked the port to
+				      * recover the link. Transport class will
+				      * retry or fail IO */
+#define DID_TRANSPORT_FAILFAST	0x0f /* Transport class fastfailed the io */
 #define DRIVER_OK       0x00	/* Driver status                           */
 
 /*
@@ -426,6 +431,7 @@ static inline int scsi_is_wlun(unsigned int lun)
 #define SCSI_MLQUEUE_HOST_BUSY   0x1055
 #define SCSI_MLQUEUE_DEVICE_BUSY 0x1056
 #define SCSI_MLQUEUE_EH_RETRY    0x1057
+#define SCSI_MLQUEUE_TARGET_BUSY 0x1058
 
 /*
  *  Use these to separate status msg and our bytes

include/scsi/scsi_device.h

@@ -238,6 +238,16 @@ struct scsi_target {
 						 * for the device at a time. */
 	unsigned int		pdt_1f_for_no_lun;	/* PDT = 0x1f */
 						/* means no lun present */
+	/* commands actually active on LLD. protected by host lock. */
+	unsigned int		target_busy;
+	/*
+	 * LLDs should set this in the slave_alloc host template callout.
+	 * If set to zero then there is not limit.
+	 */
+	unsigned int		can_queue;
+	unsigned int		target_blocked;
+	unsigned int		max_target_blocked;
+#define SCSI_DEFAULT_TARGET_BLOCKED	3
 
 	char			scsi_level;
 	struct execute_work	ew;

include/scsi/scsi_transport_fc.h

@@ -357,6 +357,7 @@ struct fc_rport {	/* aka fc_starget_attrs */
 /* bit field values for struct fc_rport "flags" field: */
 #define FC_RPORT_DEVLOSS_PENDING	0x01
 #define FC_RPORT_SCAN_PENDING		0x02
+#define FC_RPORT_FAST_FAIL_TIMEDOUT	0x03
 
 #define	dev_to_rport(d)				\
 	container_of(d, struct fc_rport, dev)
@@ -678,12 +679,15 @@ fc_remote_port_chkready(struct fc_rport *rport)
 		if (rport->roles & FC_PORT_ROLE_FCP_TARGET)
 			result = 0;
 		else if (rport->flags & FC_RPORT_DEVLOSS_PENDING)
-			result = DID_IMM_RETRY << 16;
+			result = DID_TRANSPORT_DISRUPTED << 16;
 		else
 			result = DID_NO_CONNECT << 16;
 		break;
 	case FC_PORTSTATE_BLOCKED:
-		result = DID_IMM_RETRY << 16;
+		if (rport->flags & FC_RPORT_FAST_FAIL_TIMEDOUT)
+			result = DID_TRANSPORT_FAILFAST << 16;
+		else
+			result = DID_TRANSPORT_DISRUPTED << 16;
 		break;
 	default:
 		result = DID_NO_CONNECT << 16;

include/scsi/scsi_transport_iscsi.h

@@ -135,7 +135,8 @@ extern int iscsi_unregister_transport(struct iscsi_transport *tt);
 /*
  * control plane upcalls
  */
-extern void iscsi_conn_error(struct iscsi_cls_conn *conn, enum iscsi_err error);
+extern void iscsi_conn_error_event(struct iscsi_cls_conn *conn,
+				   enum iscsi_err error);
 extern int iscsi_recv_pdu(struct iscsi_cls_conn *conn, struct iscsi_hdr *hdr,
 			  char *data, uint32_t data_size);
 
@@ -207,7 +208,7 @@ extern void iscsi_host_for_each_session(struct Scsi_Host *shost,
 struct iscsi_endpoint {
 	void *dd_data;			/* LLD private data */
 	struct device dev;
-	unsigned int id;
+	uint64_t id;
 };
 
 /*