Merge branch 'misc' into for-linus

Signed-off-by: James Bottomley <JBottomley@Parallels.com>

20 files changed, 1617 insertions, 437 deletions

diff --git a/drivers/scsi/lpfc/lpfc.h b/drivers/scsi/lpfc/lpfc.h
index 7706c99ec8bb..bcc56cac4fd8 100644
--- a/drivers/scsi/lpfc/lpfc.h
+++ b/drivers/scsi/lpfc/lpfc.h
@@ -46,10 +46,15 @@ struct lpfc_sli2_slim;
 #define LPFC_DEFAULT_MENLO_SG_SEG_CNT 128	/* sg element count per scsi
 		cmnd for menlo needs nearly twice as for firmware
 		downloads using bsg */
-#define LPFC_DEFAULT_PROT_SG_SEG_CNT 4096 /* sg protection elements count */
+
+#define LPFC_MIN_SG_SLI4_BUF_SZ	0x800	/* based on LPFC_DEFAULT_SG_SEG_CNT */
+#define LPFC_MAX_SG_SLI4_SEG_CNT_DIF 128 /* sg element count per scsi cmnd */
+#define LPFC_MAX_SG_SEG_CNT_DIF 512	/* sg element count per scsi cmnd  */
 #define LPFC_MAX_SG_SEG_CNT	4096	/* sg element count per scsi cmnd */
+#define LPFC_MAX_SGL_SEG_CNT	512	/* SGL element count per scsi cmnd */
+#define LPFC_MAX_BPL_SEG_CNT	4096	/* BPL element count per scsi cmnd */
+
 #define LPFC_MAX_SGE_SIZE       0x80000000 /* Maximum data allowed in a SGE */
-#define LPFC_MAX_PROT_SG_SEG_CNT 4096	/* prot sg element count per scsi cmd*/
 #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 */
@@ -66,8 +71,10 @@ struct lpfc_sli2_slim;
  * queue depths when there are driver resource error or Firmware
  * resource error.
  */
-#define QUEUE_RAMP_DOWN_INTERVAL	(1 * HZ)   /* 1 Second */
-#define QUEUE_RAMP_UP_INTERVAL		(300 * HZ) /* 5 minutes */
+/* 1 Second */
+#define QUEUE_RAMP_DOWN_INTERVAL	(msecs_to_jiffies(1000 * 1))
+/* 5 minutes */
+#define QUEUE_RAMP_UP_INTERVAL		(msecs_to_jiffies(1000 * 300))
 
 /* Number of exchanges reserved for discovery to complete */
 #define LPFC_DISC_IOCB_BUFF_COUNT 20
@@ -671,6 +678,7 @@ struct lpfc_hba {
 	uint32_t lmt;
 
 	uint32_t fc_topology;	/* link topology, from LINK INIT */
+	uint32_t fc_topology_changed;	/* link topology, from LINK INIT */
 
 	struct lpfc_stats fc_stat;
 
@@ -701,9 +709,11 @@ struct lpfc_hba {
 	uint32_t cfg_poll_tmo;
 	uint32_t cfg_use_msi;
 	uint32_t cfg_fcp_imax;
+	uint32_t cfg_fcp_cpu_map;
 	uint32_t cfg_fcp_wq_count;
 	uint32_t cfg_fcp_eq_count;
 	uint32_t cfg_fcp_io_channel;
+	uint32_t cfg_total_seg_cnt;
 	uint32_t cfg_sg_seg_cnt;
 	uint32_t cfg_prot_sg_seg_cnt;
 	uint32_t cfg_sg_dma_buf_size;
@@ -804,8 +814,10 @@ struct lpfc_hba {
 	uint64_t bg_reftag_err_cnt;
 
 	/* fastpath list. */
-	spinlock_t scsi_buf_list_lock;
-	struct list_head lpfc_scsi_buf_list;
+	spinlock_t scsi_buf_list_get_lock;  /* SCSI buf alloc list lock */
+	spinlock_t scsi_buf_list_put_lock;  /* SCSI buf free list lock */
+	struct list_head lpfc_scsi_buf_list_get;
+	struct list_head lpfc_scsi_buf_list_put;
 	uint32_t total_scsi_bufs;
 	struct list_head lpfc_iocb_list;
 	uint32_t total_iocbq_bufs;
diff --git a/drivers/scsi/lpfc/lpfc_attr.c b/drivers/scsi/lpfc/lpfc_attr.c
index 9290713af253..3c5625b8b1f4 100644
--- a/drivers/scsi/lpfc/lpfc_attr.c
+++ b/drivers/scsi/lpfc/lpfc_attr.c
@@ -674,6 +674,9 @@ lpfc_do_offline(struct lpfc_hba *phba, uint32_t type)
 	int i;
 	int rc;
 
+	if (phba->pport->fc_flag & FC_OFFLINE_MODE)
+		return 0;
+
 	init_completion(&online_compl);
 	rc = lpfc_workq_post_event(phba, &status, &online_compl,
 			      LPFC_EVT_OFFLINE_PREP);
@@ -741,7 +744,8 @@ lpfc_selective_reset(struct lpfc_hba *phba)
 	int status = 0;
 	int rc;
 
-	if (!phba->cfg_enable_hba_reset)
+	if ((!phba->cfg_enable_hba_reset) ||
+	    (phba->pport->fc_flag & FC_OFFLINE_MODE))
 		return -EACCES;
 
 	status = lpfc_do_offline(phba, LPFC_EVT_OFFLINE);
@@ -895,6 +899,7 @@ lpfc_sli4_pdev_reg_request(struct lpfc_hba *phba, uint32_t opcode)
 		pci_disable_sriov(pdev);
 		phba->cfg_sriov_nr_virtfn = 0;
 	}
+
 	status = lpfc_do_offline(phba, LPFC_EVT_OFFLINE);
 
 	if (status != 0)
@@ -2801,6 +2806,8 @@ lpfc_topology_store(struct device *dev, struct device_attribute *attr,
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_INIT,
 			"3054 lpfc_topology changed from %d to %d\n",
 			prev_val, val);
+		if (prev_val != val && phba->sli_rev == LPFC_SLI_REV4)
+			phba->fc_topology_changed = 1;
 		err = lpfc_issue_lip(lpfc_shost_from_vport(phba->pport));
 		if (err) {
 			phba->cfg_topology = prev_val;
@@ -3792,6 +3799,141 @@ lpfc_fcp_imax_init(struct lpfc_hba *phba, int val)
 static DEVICE_ATTR(lpfc_fcp_imax, S_IRUGO | S_IWUSR,
 		   lpfc_fcp_imax_show, lpfc_fcp_imax_store);
 
+/**
+ * lpfc_state_show - Display current driver CPU affinity
+ * @dev: class converted to a Scsi_host structure.
+ * @attr: device attribute, not used.
+ * @buf: on return contains text describing the state of the link.
+ *
+ * Returns: size of formatted string.
+ **/
+static ssize_t
+lpfc_fcp_cpu_map_show(struct device *dev, struct device_attribute *attr,
+		      char *buf)
+{
+	struct Scsi_Host  *shost = class_to_shost(dev);
+	struct lpfc_vport *vport = (struct lpfc_vport *)shost->hostdata;
+	struct lpfc_hba   *phba = vport->phba;
+	struct lpfc_vector_map_info *cpup;
+	int  idx, len = 0;
+
+	if ((phba->sli_rev != LPFC_SLI_REV4) ||
+	    (phba->intr_type != MSIX))
+		return len;
+
+	switch (phba->cfg_fcp_cpu_map) {
+	case 0:
+		len += snprintf(buf + len, PAGE_SIZE-len,
+				"fcp_cpu_map: No mapping (%d)\n",
+				phba->cfg_fcp_cpu_map);
+		return len;
+	case 1:
+		len += snprintf(buf + len, PAGE_SIZE-len,
+				"fcp_cpu_map: HBA centric mapping (%d): "
+				"%d online CPUs\n",
+				phba->cfg_fcp_cpu_map,
+				phba->sli4_hba.num_online_cpu);
+		break;
+	case 2:
+		len += snprintf(buf + len, PAGE_SIZE-len,
+				"fcp_cpu_map: Driver centric mapping (%d): "
+				"%d online CPUs\n",
+				phba->cfg_fcp_cpu_map,
+				phba->sli4_hba.num_online_cpu);
+		break;
+	}
+
+	cpup = phba->sli4_hba.cpu_map;
+	for (idx = 0; idx < phba->sli4_hba.num_present_cpu; idx++) {
+		if (cpup->irq == LPFC_VECTOR_MAP_EMPTY)
+			len += snprintf(buf + len, PAGE_SIZE-len,
+					"CPU %02d io_chan %02d "
+					"physid %d coreid %d\n",
+					idx, cpup->channel_id, cpup->phys_id,
+					cpup->core_id);
+		else
+			len += snprintf(buf + len, PAGE_SIZE-len,
+					"CPU %02d io_chan %02d "
+					"physid %d coreid %d IRQ %d\n",
+					idx, cpup->channel_id, cpup->phys_id,
+					cpup->core_id, cpup->irq);
+
+		cpup++;
+	}
+	return len;
+}
+
+/**
+ * lpfc_fcp_cpu_map_store - Change CPU affinity of driver vectors
+ * @dev: class device that is converted into a Scsi_host.
+ * @attr: device attribute, not used.
+ * @buf: one or more lpfc_polling_flags values.
+ * @count: not used.
+ *
+ * Returns:
+ * -EINVAL  - Not implemented yet.
+ **/
+static ssize_t
+lpfc_fcp_cpu_map_store(struct device *dev, struct device_attribute *attr,
+		       const char *buf, size_t count)
+{
+	int status = -EINVAL;
+	return status;
+}
+
+/*
+# lpfc_fcp_cpu_map: Defines how to map CPUs to IRQ vectors
+# for the HBA.
+#
+# Value range is [0 to 2]. Default value is LPFC_DRIVER_CPU_MAP (2).
+#	0 - Do not affinitze IRQ vectors
+#	1 - Affintize HBA vectors with respect to each HBA
+#	    (start with CPU0 for each HBA)
+#	2 - Affintize HBA vectors with respect to the entire driver
+#	    (round robin thru all CPUs across all HBAs)
+*/
+static int lpfc_fcp_cpu_map = LPFC_DRIVER_CPU_MAP;
+module_param(lpfc_fcp_cpu_map, int, S_IRUGO|S_IWUSR);
+MODULE_PARM_DESC(lpfc_fcp_cpu_map,
+		 "Defines how to map CPUs to IRQ vectors per HBA");
+
+/**
+ * lpfc_fcp_cpu_map_init - Set the initial sr-iov virtual function enable
+ * @phba: lpfc_hba pointer.
+ * @val: link speed value.
+ *
+ * Description:
+ * If val is in a valid range [0-2], then affinitze the adapter's
+ * MSIX vectors.
+ *
+ * Returns:
+ * zero if val saved.
+ * -EINVAL val out of range
+ **/
+static int
+lpfc_fcp_cpu_map_init(struct lpfc_hba *phba, int val)
+{
+	if (phba->sli_rev != LPFC_SLI_REV4) {
+		phba->cfg_fcp_cpu_map = 0;
+		return 0;
+	}
+
+	if (val >= LPFC_MIN_CPU_MAP && val <= LPFC_MAX_CPU_MAP) {
+		phba->cfg_fcp_cpu_map = val;
+		return 0;
+	}
+
+	lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+			"3326 fcp_cpu_map: %d out of range, using default\n",
+			val);
+	phba->cfg_fcp_cpu_map = LPFC_DRIVER_CPU_MAP;
+
+	return 0;
+}
+
+static DEVICE_ATTR(lpfc_fcp_cpu_map, S_IRUGO | S_IWUSR,
+		   lpfc_fcp_cpu_map_show, lpfc_fcp_cpu_map_store);
+
 /*
 # lpfc_fcp_class:  Determines FC class to use for the FCP protocol.
 # Value range is [2,3]. Default value is 3.
@@ -4009,12 +4151,11 @@ LPFC_ATTR_R(enable_bg, 0, 0, 1, "Enable BlockGuard Support");
 #       0  = disabled (default)
 #       1  = enabled
 # Value range is [0,1]. Default value is 0.
+#
+# This feature in under investigation and may be supported in the future.
 */
 unsigned int lpfc_fcp_look_ahead = LPFC_LOOK_AHEAD_OFF;
 
-module_param(lpfc_fcp_look_ahead, uint, S_IRUGO);
-MODULE_PARM_DESC(lpfc_fcp_look_ahead, "Look ahead for completions");
-
 /*
 # lpfc_prot_mask: i
 #	- Bit mask of host protection capabilities used to register with the
@@ -4071,16 +4212,23 @@ MODULE_PARM_DESC(lpfc_delay_discovery,
 
 /*
  * lpfc_sg_seg_cnt - Initial Maximum DMA Segment Count
- * This value can be set to values between 64 and 256. The default value is
+ * This value can be set to values between 64 and 4096. The default value is
  * 64, but may be increased to allow for larger Max I/O sizes. The scsi layer
  * will be allowed to request I/Os of sizes up to (MAX_SEG_COUNT * SEG_SIZE).
+ * Because of the additional overhead involved in setting up T10-DIF,
+ * this parameter will be limited to 128 if BlockGuard is enabled under SLI4
+ * and will be limited to 512 if BlockGuard is enabled under SLI3.
  */
 LPFC_ATTR_R(sg_seg_cnt, LPFC_DEFAULT_SG_SEG_CNT, LPFC_DEFAULT_SG_SEG_CNT,
 	    LPFC_MAX_SG_SEG_CNT, "Max Scatter Gather Segment Count");
 
-LPFC_ATTR_R(prot_sg_seg_cnt, LPFC_DEFAULT_PROT_SG_SEG_CNT,
-		LPFC_DEFAULT_PROT_SG_SEG_CNT, LPFC_MAX_PROT_SG_SEG_CNT,
-		"Max Protection Scatter Gather Segment Count");
+/*
+ * This parameter will be depricated, the driver cannot limit the
+ * protection data s/g list.
+ */
+LPFC_ATTR_R(prot_sg_seg_cnt, LPFC_DEFAULT_SG_SEG_CNT,
+	    LPFC_DEFAULT_SG_SEG_CNT, LPFC_MAX_SG_SEG_CNT,
+	    "Max Protection Scatter Gather Segment Count");
 
 struct device_attribute *lpfc_hba_attrs[] = {
 	&dev_attr_bg_info,
@@ -4141,6 +4289,7 @@ struct device_attribute *lpfc_hba_attrs[] = {
 	&dev_attr_lpfc_poll_tmo,
 	&dev_attr_lpfc_use_msi,
 	&dev_attr_lpfc_fcp_imax,
+	&dev_attr_lpfc_fcp_cpu_map,
 	&dev_attr_lpfc_fcp_wq_count,
 	&dev_attr_lpfc_fcp_eq_count,
 	&dev_attr_lpfc_fcp_io_channel,
@@ -5123,6 +5272,7 @@ lpfc_get_cfgparam(struct lpfc_hba *phba)
 	lpfc_enable_rrq_init(phba, lpfc_enable_rrq);
 	lpfc_use_msi_init(phba, lpfc_use_msi);
 	lpfc_fcp_imax_init(phba, lpfc_fcp_imax);
+	lpfc_fcp_cpu_map_init(phba, lpfc_fcp_cpu_map);
 	lpfc_fcp_wq_count_init(phba, lpfc_fcp_wq_count);
 	lpfc_fcp_eq_count_init(phba, lpfc_fcp_eq_count);
 	lpfc_fcp_io_channel_init(phba, lpfc_fcp_io_channel);
diff --git a/drivers/scsi/lpfc/lpfc_bsg.c b/drivers/scsi/lpfc/lpfc_bsg.c
index 888666892004..094be2cad65b 100644
--- a/drivers/scsi/lpfc/lpfc_bsg.c
+++ b/drivers/scsi/lpfc/lpfc_bsg.c
@@ -219,26 +219,35 @@ lpfc_bsg_copy_data(struct lpfc_dmabuf *dma_buffers,
 	unsigned int transfer_bytes, bytes_copied = 0;
 	unsigned int sg_offset, dma_offset;
 	unsigned char *dma_address, *sg_address;
-	struct scatterlist *sgel;
 	LIST_HEAD(temp_list);
-
+	struct sg_mapping_iter miter;
+	unsigned long flags;
+	unsigned int sg_flags = SG_MITER_ATOMIC;
+	bool sg_valid;
 
 	list_splice_init(&dma_buffers->list, &temp_list);
 	list_add(&dma_buffers->list, &temp_list);
 	sg_offset = 0;
-	sgel = bsg_buffers->sg_list;
+	if (to_buffers)
+		sg_flags |= SG_MITER_FROM_SG;
+	else
+		sg_flags |= SG_MITER_TO_SG;
+	sg_miter_start(&miter, bsg_buffers->sg_list, bsg_buffers->sg_cnt,
+		       sg_flags);
+	local_irq_save(flags);
+	sg_valid = sg_miter_next(&miter);
 	list_for_each_entry(mp, &temp_list, list) {
 		dma_offset = 0;
-		while (bytes_to_transfer && sgel &&
+		while (bytes_to_transfer && sg_valid &&
 		       (dma_offset < LPFC_BPL_SIZE)) {
 			dma_address = mp->virt + dma_offset;
 			if (sg_offset) {
 				/* Continue previous partial transfer of sg */
-				sg_address = sg_virt(sgel) + sg_offset;
-				transfer_bytes = sgel->length - sg_offset;
+				sg_address = miter.addr + sg_offset;
+				transfer_bytes = miter.length - sg_offset;
 			} else {
-				sg_address = sg_virt(sgel);
-				transfer_bytes = sgel->length;
+				sg_address = miter.addr;
+				transfer_bytes = miter.length;
 			}
 			if (bytes_to_transfer < transfer_bytes)
 				transfer_bytes = bytes_to_transfer;
@@ -252,12 +261,14 @@ lpfc_bsg_copy_data(struct lpfc_dmabuf *dma_buffers,
 			sg_offset += transfer_bytes;
 			bytes_to_transfer -= transfer_bytes;
 			bytes_copied += transfer_bytes;
-			if (sg_offset >= sgel->length) {
+			if (sg_offset >= miter.length) {
 				sg_offset = 0;
-				sgel = sg_next(sgel);
+				sg_valid = sg_miter_next(&miter);
 			}
 		}
 	}
+	sg_miter_stop(&miter);
+	local_irq_restore(flags);
 	list_del_init(&dma_buffers->list);
 	list_splice(&temp_list, &dma_buffers->list);
 	return bytes_copied;
@@ -471,6 +482,7 @@ lpfc_bsg_send_mgmt_cmd(struct fc_bsg_job *job)
 	cmdiocbq->context1 = dd_data;
 	cmdiocbq->context2 = cmp;
 	cmdiocbq->context3 = bmp;
+	cmdiocbq->context_un.ndlp = ndlp;
 	dd_data->type = TYPE_IOCB;
 	dd_data->set_job = job;
 	dd_data->context_un.iocb.cmdiocbq = cmdiocbq;
@@ -1508,6 +1520,7 @@ lpfc_issue_ct_rsp(struct lpfc_hba *phba, struct fc_bsg_job *job, uint32_t tag,
 	ctiocb->context1 = dd_data;
 	ctiocb->context2 = cmp;
 	ctiocb->context3 = bmp;
+	ctiocb->context_un.ndlp = ndlp;
 	ctiocb->iocb_cmpl = lpfc_issue_ct_rsp_cmp;
 
 	dd_data->type = TYPE_IOCB;
@@ -2576,7 +2589,8 @@ static int lpfcdiag_loop_get_xri(struct lpfc_hba *phba, uint16_t rpi,
 	evt->wait_time_stamp = jiffies;
 	time_left = wait_event_interruptible_timeout(
 		evt->wq, !list_empty(&evt->events_to_see),
-		((phba->fc_ratov * 2) + LPFC_DRVR_TIMEOUT) * HZ);
+		msecs_to_jiffies(1000 *
+			((phba->fc_ratov * 2) + LPFC_DRVR_TIMEOUT)));
 	if (list_empty(&evt->events_to_see))
 		ret_val = (time_left) ? -EINTR : -ETIMEDOUT;
 	else {
@@ -3151,7 +3165,8 @@ lpfc_bsg_diag_loopback_run(struct fc_bsg_job *job)
 	evt->waiting = 1;
 	time_left = wait_event_interruptible_timeout(
 		evt->wq, !list_empty(&evt->events_to_see),
-		((phba->fc_ratov * 2) + LPFC_DRVR_TIMEOUT) * HZ);
+		msecs_to_jiffies(1000 *
+			((phba->fc_ratov * 2) + LPFC_DRVR_TIMEOUT)));
 	evt->waiting = 0;
 	if (list_empty(&evt->events_to_see)) {
 		rc = (time_left) ? -EINTR : -ETIMEDOUT;
diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index 7631893ae005..d41456e5f814 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -470,3 +470,4 @@ int lpfc_sli4_xri_sgl_update(struct lpfc_hba *);
 void lpfc_free_sgl_list(struct lpfc_hba *, struct list_head *);
 uint32_t lpfc_sli_port_speed_get(struct lpfc_hba *);
 int lpfc_sli4_request_firmware_update(struct lpfc_hba *, uint8_t);
+void lpfc_sli4_offline_eratt(struct lpfc_hba *);
diff --git a/drivers/scsi/lpfc/lpfc_ct.c b/drivers/scsi/lpfc/lpfc_ct.c
index 7bff3a19af56..ae1a07c57cae 100644
--- a/drivers/scsi/lpfc/lpfc_ct.c
+++ b/drivers/scsi/lpfc/lpfc_ct.c
@@ -1811,7 +1811,8 @@ lpfc_fdmi_timeout_handler(struct lpfc_vport *vport)
 		if (init_utsname()->nodename[0] != '\0')
 			lpfc_fdmi_cmd(vport, ndlp, SLI_MGMT_DHBA);
 		else
-			mod_timer(&vport->fc_fdmitmo, jiffies + HZ * 60);
+			mod_timer(&vport->fc_fdmitmo, jiffies +
+				  msecs_to_jiffies(1000 * 60));
 	}
 	return;
 }
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index bbed8471bf0b..3cae0a92e8bd 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -29,6 +29,7 @@
 #include <scsi/scsi_host.h>
 #include <scsi/scsi_transport_fc.h>
 
+
 #include "lpfc_hw4.h"
 #include "lpfc_hw.h"
 #include "lpfc_sli.h"
@@ -238,7 +239,10 @@ lpfc_prep_els_iocb(struct lpfc_vport *vport, uint8_t expectRsp,
 
 		icmd->un.elsreq64.remoteID = did;		/* DID */
 		icmd->ulpCommand = CMD_ELS_REQUEST64_CR;
-		icmd->ulpTimeout = phba->fc_ratov * 2;
+		if (elscmd == ELS_CMD_FLOGI)
+			icmd->ulpTimeout = FF_DEF_RATOV * 2;
+		else
+			icmd->ulpTimeout = phba->fc_ratov * 2;
 	} else {
 		icmd->un.xseq64.bdl.addrHigh = putPaddrHigh(pbuflist->phys);
 		icmd->un.xseq64.bdl.addrLow = putPaddrLow(pbuflist->phys);
@@ -308,16 +312,20 @@ lpfc_prep_els_iocb(struct lpfc_vport *vport, uint8_t expectRsp,
 		/* Xmit ELS command <elsCmd> to remote NPORT <did> */
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
 				 "0116 Xmit ELS command x%x to remote "
-				 "NPORT x%x I/O tag: x%x, port state: x%x\n",
+				 "NPORT x%x I/O tag: x%x, port state:x%x"
+				 " fc_flag:x%x\n",
 				 elscmd, did, elsiocb->iotag,
-				 vport->port_state);
+				 vport->port_state,
+				 vport->fc_flag);
 	} else {
 		/* Xmit ELS response <elsCmd> to remote NPORT <did> */
 		lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
 				 "0117 Xmit ELS response x%x to remote "
-				 "NPORT x%x I/O tag: x%x, size: x%x\n",
+				 "NPORT x%x I/O tag: x%x, size: x%x "
+				 "port_state x%x fc_flag x%x\n",
 				 elscmd, ndlp->nlp_DID, elsiocb->iotag,
-				 cmdSize);
+				 cmdSize, vport->port_state,
+				 vport->fc_flag);
 	}
 	return elsiocb;
 
@@ -909,6 +917,23 @@ lpfc_cmpl_els_flogi_nport(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 	spin_lock_irq(shost->host_lock);
 	vport->fc_flag |= FC_PT2PT;
 	spin_unlock_irq(shost->host_lock);
+	/* If physical FC port changed, unreg VFI and ALL VPIs / RPIs */
+	if ((phba->sli_rev == LPFC_SLI_REV4) && phba->fc_topology_changed) {
+		lpfc_unregister_fcf_prep(phba);
+
+		/* The FC_VFI_REGISTERED flag will get clear in the cmpl
+		 * handler for unreg_vfi, but if we don't force the
+		 * FC_VFI_REGISTERED flag then the reg_vfi mailbox could be
+		 * built with the update bit set instead of just the vp bit to
+		 * change the Nport ID.  We need to have the vp set and the
+		 * Upd cleared on topology changes.
+		 */
+		spin_lock_irq(shost->host_lock);
+		vport->fc_flag &= ~FC_VFI_REGISTERED;
+		spin_unlock_irq(shost->host_lock);
+		phba->fc_topology_changed = 0;
+		lpfc_issue_reg_vfi(vport);
+	}
 
 	/* Start discovery - this should just do CLEAR_LA */
 	lpfc_disc_start(vport);
@@ -1030,9 +1055,19 @@ stop_rr_fcf_flogi:
 			vport->cfg_discovery_threads = LPFC_MAX_DISC_THREADS;
 		if ((phba->sli_rev == LPFC_SLI_REV4) &&
 		    (!(vport->fc_flag & FC_VFI_REGISTERED) ||
-		     (vport->fc_prevDID != vport->fc_myDID))) {
-			if (vport->fc_flag & FC_VFI_REGISTERED)
-				lpfc_sli4_unreg_all_rpis(vport);
+		     (vport->fc_prevDID != vport->fc_myDID) ||
+			phba->fc_topology_changed)) {
+			if (vport->fc_flag & FC_VFI_REGISTERED) {
+				if (phba->fc_topology_changed) {
+					lpfc_unregister_fcf_prep(phba);
+					spin_lock_irq(shost->host_lock);
+					vport->fc_flag &= ~FC_VFI_REGISTERED;
+					spin_unlock_irq(shost->host_lock);
+					phba->fc_topology_changed = 0;
+				} else {
+					lpfc_sli4_unreg_all_rpis(vport);
+				}
+			}
 			lpfc_issue_reg_vfi(vport);
 			lpfc_nlp_put(ndlp);
 			goto out;
@@ -1054,10 +1089,11 @@ stop_rr_fcf_flogi:
 
 	/* FLOGI completes successfully */
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
-			 "0101 FLOGI completes successfully "
-			 "Data: x%x x%x x%x x%x\n",
+			 "0101 FLOGI completes successfully, I/O tag:x%x, "
+			 "Data: x%x x%x x%x x%x x%x x%x\n", cmdiocb->iotag,
 			 irsp->un.ulpWord[4], sp->cmn.e_d_tov,
-			 sp->cmn.w2.r_a_tov, sp->cmn.edtovResolution);
+			 sp->cmn.w2.r_a_tov, sp->cmn.edtovResolution,
+			 vport->port_state, vport->fc_flag);
 
 	if (vport->port_state == LPFC_FLOGI) {
 		/*
@@ -5047,6 +5083,8 @@ lpfc_els_rcv_flogi(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
 	struct ls_rjt stat;
 	uint32_t cmd, did;
 	int rc;
+	uint32_t fc_flag = 0;
+	uint32_t port_state = 0;
 
 	cmd = *lp++;
 	sp = (struct serv_parm *) lp;
@@ -5113,16 +5151,25 @@ lpfc_els_rcv_flogi(struct lpfc_vport *vport, struct lpfc_iocbq *cmdiocb,
 			 * will be.
 			 */
 			vport->fc_myDID = PT2PT_LocalID;
-		}
+		} else
+			vport->fc_myDID = PT2PT_RemoteID;
 
 		/*
 		 * The vport state should go to LPFC_FLOGI only
 		 * AFTER we issue a FLOGI, not receive one.
 		 */
 		spin_lock_irq(shost->host_lock);
+		fc_flag = vport->fc_flag;
+		port_state = vport->port_state;
 		vport->fc_flag |= FC_PT2PT;
 		vport->fc_flag &= ~(FC_FABRIC | FC_PUBLIC_LOOP);
+		vport->port_state = LPFC_FLOGI;
 		spin_unlock_irq(shost->host_lock);
+		lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
+				 "3311 Rcv Flogi PS x%x new PS x%x "
+				 "fc_flag x%x new fc_flag x%x\n",
+				 port_state, vport->port_state,
+				 fc_flag, vport->fc_flag);
 
 		/*
 		 * We temporarily set fc_myDID to make it look like we are
@@ -6241,7 +6288,8 @@ lpfc_els_timeout_handler(struct lpfc_vport *vport)
 	}
 
 	if (!list_empty(&phba->sli.ring[LPFC_ELS_RING].txcmplq))
-		mod_timer(&vport->els_tmofunc, jiffies + HZ * timeout);
+		mod_timer(&vport->els_tmofunc,
+			  jiffies + msecs_to_jiffies(1000 * timeout));
 }
 
 /**
@@ -6612,7 +6660,9 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 	/* ELS command <elsCmd> received from NPORT <did> */
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
 			 "0112 ELS command x%x received from NPORT x%x "
-			 "Data: x%x\n", cmd, did, vport->port_state);
+			 "Data: x%x x%x x%x x%x\n",
+			cmd, did, vport->port_state, vport->fc_flag,
+			vport->fc_myDID, vport->fc_prevDID);
 	switch (cmd) {
 	case ELS_CMD_PLOGI:
 		lpfc_debugfs_disc_trc(vport, LPFC_DISC_TRC_ELS_UNSOL,
@@ -6621,6 +6671,19 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 
 		phba->fc_stat.elsRcvPLOGI++;
 		ndlp = lpfc_plogi_confirm_nport(phba, payload, ndlp);
+		if (phba->sli_rev == LPFC_SLI_REV4 &&
+		    (phba->pport->fc_flag & FC_PT2PT)) {
+			vport->fc_prevDID = vport->fc_myDID;
+			/* Our DID needs to be updated before registering
+			 * the vfi. This is done in lpfc_rcv_plogi but
+			 * that is called after the reg_vfi.
+			 */
+			vport->fc_myDID = elsiocb->iocb.un.rcvels.parmRo;
+			lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
+					 "3312 Remote port assigned DID x%x "
+					 "%x\n", vport->fc_myDID,
+					 vport->fc_prevDID);
+		}
 
 		lpfc_send_els_event(vport, ndlp, payload);
 
@@ -6630,6 +6693,7 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 			rjt_exp = LSEXP_NOTHING_MORE;
 			break;
 		}
+		shost = lpfc_shost_from_vport(vport);
 		if (vport->port_state < LPFC_DISC_AUTH) {
 			if (!(phba->pport->fc_flag & FC_PT2PT) ||
 				(phba->pport->fc_flag & FC_PT2PT_PLOGI)) {
@@ -6641,9 +6705,18 @@ lpfc_els_unsol_buffer(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 			 * another NPort and the other side has initiated
 			 * the PLOGI before responding to our FLOGI.
 			 */
+			if (phba->sli_rev == LPFC_SLI_REV4 &&
+			    (phba->fc_topology_changed ||
+			     vport->fc_myDID != vport->fc_prevDID)) {
+				lpfc_unregister_fcf_prep(phba);
+				spin_lock_irq(shost->host_lock);
+				vport->fc_flag &= ~FC_VFI_REGISTERED;
+				spin_unlock_irq(shost->host_lock);
+				phba->fc_topology_changed = 0;
+				lpfc_issue_reg_vfi(vport);
+			}
 		}
 
-		shost = lpfc_shost_from_vport(vport);
 		spin_lock_irq(shost->host_lock);
 		ndlp->nlp_flag &= ~NLP_TARGET_REMOVE;
 		spin_unlock_irq(shost->host_lock);
@@ -7002,8 +7075,11 @@ lpfc_do_scr_ns_plogi(struct lpfc_hba *phba, struct lpfc_vport *vport)
 	spin_lock_irq(shost->host_lock);
 	if (vport->fc_flag & FC_DISC_DELAYED) {
 		spin_unlock_irq(shost->host_lock);
+		lpfc_printf_log(phba, KERN_ERR, LOG_DISCOVERY,
+				"3334 Delay fc port discovery for %d seconds\n",
+				phba->fc_ratov);
 		mod_timer(&vport->delayed_disc_tmo,
-			jiffies + HZ * phba->fc_ratov);
+			jiffies + msecs_to_jiffies(1000 * phba->fc_ratov));
 		return;
 	}
 	spin_unlock_irq(shost->host_lock);
@@ -7287,7 +7363,7 @@ lpfc_retry_pport_discovery(struct lpfc_hba *phba)
 		return;
 
 	shost = lpfc_shost_from_vport(phba->pport);
-	mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ);
+	mod_timer(&ndlp->nlp_delayfunc, jiffies + msecs_to_jiffies(1000));
 	spin_lock_irq(shost->host_lock);
 	ndlp->nlp_flag |= NLP_DELAY_TMO;
 	spin_unlock_irq(shost->host_lock);
@@ -7791,7 +7867,8 @@ lpfc_block_fabric_iocbs(struct lpfc_hba *phba)
 	blocked = test_and_set_bit(FABRIC_COMANDS_BLOCKED, &phba->bit_flags);
 	/* Start a timer to unblock fabric iocbs after 100ms */
 	if (!blocked)
-		mod_timer(&phba->fabric_block_timer, jiffies + HZ/10 );
+		mod_timer(&phba->fabric_block_timer,
+			  jiffies + msecs_to_jiffies(100));
 
 	return;
 }
diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index 326e05a65a73..0f6e2548f35d 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -160,11 +160,12 @@ lpfc_dev_loss_tmo_callbk(struct fc_rport *rport)
 	if (!list_empty(&evtp->evt_listp))
 		return;
 
+	evtp->evt_arg1  = lpfc_nlp_get(ndlp);
+
 	spin_lock_irq(&phba->hbalock);
 	/* We need to hold the node by incrementing the reference
 	 * count until this queued work is done
 	 */
-	evtp->evt_arg1  = lpfc_nlp_get(ndlp);
 	if (evtp->evt_arg1) {
 		evtp->evt = LPFC_EVT_DEV_LOSS;
 		list_add_tail(&evtp->evt_listp, &phba->work_list);
@@ -1008,9 +1009,6 @@ lpfc_linkup(struct lpfc_hba *phba)
 		for (i = 0; i <= phba->max_vports && vports[i] != NULL; i++)
 			lpfc_linkup_port(vports[i]);
 	lpfc_destroy_vport_work_array(phba, vports);
-	if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
-	    (phba->sli_rev < LPFC_SLI_REV4))
-		lpfc_issue_clear_la(phba, phba->pport);
 
 	return 0;
 }
@@ -1436,7 +1434,8 @@ lpfc_register_fcf(struct lpfc_hba *phba)
 	if (phba->fcf.fcf_flag & FCF_REGISTERED) {
 		phba->fcf.fcf_flag |= (FCF_SCAN_DONE | FCF_IN_USE);
 		phba->hba_flag &= ~FCF_TS_INPROG;
-		if (phba->pport->port_state != LPFC_FLOGI) {
+		if (phba->pport->port_state != LPFC_FLOGI &&
+		    phba->pport->fc_flag & FC_FABRIC) {
 			phba->hba_flag |= FCF_RR_INPROG;
 			spin_unlock_irq(&phba->hbalock);
 			lpfc_initial_flogi(phba->pport);
@@ -2270,8 +2269,11 @@ lpfc_mbx_cmpl_fcf_scan_read_fcf_rec(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
 				spin_unlock_irq(&phba->hbalock);
 				lpfc_printf_log(phba, KERN_INFO, LOG_FIP,
 						"2836 New FCF matches in-use "
-						"FCF (x%x)\n",
-						phba->fcf.current_rec.fcf_indx);
+						"FCF (x%x), port_state:x%x, "
+						"fc_flag:x%x\n",
+						phba->fcf.current_rec.fcf_indx,
+						phba->pport->port_state,
+						phba->pport->fc_flag);
 				goto out;
 			} else
 				lpfc_printf_log(phba, KERN_ERR, LOG_FIP,
@@ -2796,7 +2798,19 @@ void
 lpfc_issue_init_vpi(struct lpfc_vport *vport)
 {
 	LPFC_MBOXQ_t *mboxq;
-	int rc;
+	int rc, vpi;
+
+	if ((vport->port_type != LPFC_PHYSICAL_PORT) && (!vport->vpi)) {
+		vpi = lpfc_alloc_vpi(vport->phba);
+		if (!vpi) {
+			lpfc_printf_vlog(vport, KERN_ERR,
+					 LOG_MBOX,
+					 "3303 Failed to obtain vport vpi\n");
+			lpfc_vport_set_state(vport, FC_VPORT_FAILED);
+			return;
+		}
+		vport->vpi = vpi;
+	}
 
 	mboxq = mempool_alloc(vport->phba->mbox_mem_pool, GFP_KERNEL);
 	if (!mboxq) {
@@ -2894,9 +2908,14 @@ lpfc_mbx_cmpl_reg_vfi(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
 		goto out_free_mem;
 	}
 
-	/* If the VFI is already registered, there is nothing else to do */
+	/* If the VFI is already registered, there is nothing else to do
+	 * Unless this was a VFI update and we are in PT2PT mode, then
+	 * we should drop through to set the port state to ready.
+	 */
 	if (vport->fc_flag & FC_VFI_REGISTERED)
-		goto out_free_mem;
+		if (!(phba->sli_rev == LPFC_SLI_REV4 &&
+		      vport->fc_flag & FC_PT2PT))
+			goto out_free_mem;
 
 	/* The VPI is implicitly registered when the VFI is registered */
 	spin_lock_irq(shost->host_lock);
@@ -2913,6 +2932,13 @@ lpfc_mbx_cmpl_reg_vfi(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
 		goto out_free_mem;
 	}
 
+	lpfc_printf_vlog(vport, KERN_INFO, LOG_SLI,
+			 "3313 cmpl reg vfi  port_state:%x fc_flag:%x myDid:%x "
+			 "alpacnt:%d LinkState:%x topology:%x\n",
+			 vport->port_state, vport->fc_flag, vport->fc_myDID,
+			 vport->phba->alpa_map[0],
+			 phba->link_state, phba->fc_topology);
+
 	if (vport->port_state == LPFC_FABRIC_CFG_LINK) {
 		/*
 		 * For private loop or for NPort pt2pt,
@@ -2925,7 +2951,10 @@ lpfc_mbx_cmpl_reg_vfi(struct lpfc_hba *phba, LPFC_MBOXQ_t *mboxq)
 			/* Use loop map to make discovery list */
 			lpfc_disc_list_loopmap(vport);
 			/* Start discovery */
-			lpfc_disc_start(vport);
+			if (vport->fc_flag & FC_PT2PT)
+				vport->port_state = LPFC_VPORT_READY;
+			else
+				lpfc_disc_start(vport);
 		} else {
 			lpfc_start_fdiscs(phba);
 			lpfc_do_scr_ns_plogi(phba, vport);
@@ -3007,6 +3036,15 @@ lpfc_mbx_process_link_up(struct lpfc_hba *phba, struct lpfc_mbx_read_top *la)
 		break;
 	}
 
+	if (phba->fc_topology &&
+	    phba->fc_topology != bf_get(lpfc_mbx_read_top_topology, la)) {
+		lpfc_printf_log(phba, KERN_WARNING, LOG_SLI,
+				"3314 Toplogy changed was 0x%x is 0x%x\n",
+				phba->fc_topology,
+				bf_get(lpfc_mbx_read_top_topology, la));
+		phba->fc_topology_changed = 1;
+	}
+
 	phba->fc_topology = bf_get(lpfc_mbx_read_top_topology, la);
 	phba->link_flag &= ~LS_NPIV_FAB_SUPPORTED;
 
@@ -4235,7 +4273,7 @@ lpfc_set_disctmo(struct lpfc_vport *vport)
 			tmo, vport->port_state, vport->fc_flag);
 	}
 
-	mod_timer(&vport->fc_disctmo, jiffies + HZ * tmo);
+	mod_timer(&vport->fc_disctmo, jiffies + msecs_to_jiffies(1000 * tmo));
 	spin_lock_irq(shost->host_lock);
 	vport->fc_flag |= FC_DISC_TMO;
 	spin_unlock_irq(shost->host_lock);
@@ -4949,8 +4987,12 @@ lpfc_disc_start(struct lpfc_vport *vport)
 	uint32_t clear_la_pending;
 	int did_changed;
 
-	if (!lpfc_is_link_up(phba))
+	if (!lpfc_is_link_up(phba)) {
+		lpfc_printf_vlog(vport, KERN_INFO, LOG_SLI,
+				 "3315 Link is not up %x\n",
+				 phba->link_state);
 		return;
+	}
 
 	if (phba->link_state == LPFC_CLEAR_LA)
 		clear_la_pending = 1;
@@ -4983,11 +5025,13 @@ lpfc_disc_start(struct lpfc_vport *vport)
 	if (num_sent)
 		return;
 
-	/* Register the VPI for SLI3, NON-NPIV only. */
+	/* Register the VPI for SLI3, NPIV only. */
 	if ((phba->sli3_options & LPFC_SLI3_NPIV_ENABLED) &&
 	    !(vport->fc_flag & FC_PT2PT) &&
 	    !(vport->fc_flag & FC_RSCN_MODE) &&
 	    (phba->sli_rev < LPFC_SLI_REV4)) {
+		if (vport->port_type == LPFC_PHYSICAL_PORT)
+			lpfc_issue_clear_la(phba, vport);
 		lpfc_issue_reg_vpi(phba, vport);
 		return;
 	}
@@ -5410,7 +5454,8 @@ lpfc_mbx_cmpl_fdmi_reg_login(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	if (vport->cfg_fdmi_on == 1)
 		lpfc_fdmi_cmd(vport, ndlp, SLI_MGMT_DHBA);
 	else
-		mod_timer(&vport->fc_fdmitmo, jiffies + HZ * 60);
+		mod_timer(&vport->fc_fdmitmo,
+			  jiffies + msecs_to_jiffies(1000 * 60));
 
 	/* decrement the node reference count held for this callback
 	 * function.
@@ -5855,7 +5900,7 @@ lpfc_unregister_fcf_prep(struct lpfc_hba *phba)
 	struct lpfc_vport **vports;
 	struct lpfc_nodelist *ndlp;
 	struct Scsi_Host *shost;
-	int i, rc;
+	int i = 0, rc;
 
 	/* Unregister RPIs */
 	if (lpfc_fcf_inuse(phba))
@@ -5883,6 +5928,20 @@ lpfc_unregister_fcf_prep(struct lpfc_hba *phba)
 			spin_unlock_irq(shost->host_lock);
 		}
 	lpfc_destroy_vport_work_array(phba, vports);
+	if (i == 0 && (!(phba->sli3_options & LPFC_SLI3_NPIV_ENABLED))) {
+		ndlp = lpfc_findnode_did(phba->pport, Fabric_DID);
+		if (ndlp)
+			lpfc_cancel_retry_delay_tmo(phba->pport, ndlp);
+		lpfc_cleanup_pending_mbox(phba->pport);
+		if (phba->sli_rev == LPFC_SLI_REV4)
+			lpfc_sli4_unreg_all_rpis(phba->pport);
+		lpfc_mbx_unreg_vpi(phba->pport);
+		shost = lpfc_shost_from_vport(phba->pport);
+		spin_lock_irq(shost->host_lock);
+		phba->pport->fc_flag |= FC_VPORT_NEEDS_INIT_VPI;
+		phba->pport->vpi_state &= ~LPFC_VPI_REGISTERED;
+		spin_unlock_irq(shost->host_lock);
+	}
 
 	/* Cleanup any outstanding ELS commands */
 	lpfc_els_flush_all_cmd(phba);
diff --git a/drivers/scsi/lpfc/lpfc_hw.h b/drivers/scsi/lpfc/lpfc_hw.h
index e8c476031703..83700c18f468 100644
--- a/drivers/scsi/lpfc/lpfc_hw.h
+++ b/drivers/scsi/lpfc/lpfc_hw.h
@@ -1667,6 +1667,7 @@ enum lpfc_protgrp_type {
 #define	BG_OP_IN_CSUM_OUT_CSUM		0x5
 #define	BG_OP_IN_CRC_OUT_CSUM		0x6
 #define	BG_OP_IN_CSUM_OUT_CRC		0x7
+#define	BG_OP_RAW_MODE			0x8
 
 struct lpfc_pde5 {
 	uint32_t word0;
diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 1dd2f6f0a127..713a4613ec3a 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -200,6 +200,11 @@ struct lpfc_sli_intf {
 #define LPFC_MAX_IMAX          5000000
 #define LPFC_DEF_IMAX          50000
 
+#define LPFC_MIN_CPU_MAP       0
+#define LPFC_MAX_CPU_MAP       2
+#define LPFC_HBA_CPU_MAP       1
+#define LPFC_DRIVER_CPU_MAP    2  /* Default */
+
 /* PORT_CAPABILITIES constants. */
 #define LPFC_MAX_SUPPORTED_PAGES	8
 
@@ -621,7 +626,7 @@ struct lpfc_register {
 #define lpfc_sliport_status_rdy_SHIFT	23
 #define lpfc_sliport_status_rdy_MASK	0x1
 #define lpfc_sliport_status_rdy_WORD	word0
-#define MAX_IF_TYPE_2_RESETS	1000
+#define MAX_IF_TYPE_2_RESETS		6
 
 #define LPFC_CTL_PORT_CTL_OFFSET	0x408
 #define lpfc_sliport_ctrl_end_SHIFT	30
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 90b8b0515e23..cb465b253910 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -33,6 +33,7 @@
 #include <linux/slab.h>
 #include <linux/firmware.h>
 #include <linux/miscdevice.h>
+#include <linux/percpu.h>
 
 #include <scsi/scsi.h>
 #include <scsi/scsi_device.h>
@@ -58,6 +59,9 @@ char *_dump_buf_dif;
 unsigned long _dump_buf_dif_order;
 spinlock_t _dump_buf_lock;
 
+/* Used when mapping IRQ vectors in a driver centric manner */
+uint16_t lpfc_used_cpu[LPFC_MAX_CPU];
+
 static void lpfc_get_hba_model_desc(struct lpfc_hba *, uint8_t *, uint8_t *);
 static int lpfc_post_rcv_buf(struct lpfc_hba *);
 static int lpfc_sli4_queue_verify(struct lpfc_hba *);
@@ -541,13 +545,16 @@ lpfc_config_port_post(struct lpfc_hba *phba)
 
 	/* Set up ring-0 (ELS) timer */
 	timeout = phba->fc_ratov * 2;
-	mod_timer(&vport->els_tmofunc, jiffies + HZ * timeout);
+	mod_timer(&vport->els_tmofunc,
+		  jiffies + msecs_to_jiffies(1000 * timeout));
 	/* Set up heart beat (HB) timer */
-	mod_timer(&phba->hb_tmofunc, jiffies + HZ * LPFC_HB_MBOX_INTERVAL);
+	mod_timer(&phba->hb_tmofunc,
+		  jiffies + msecs_to_jiffies(1000 * LPFC_HB_MBOX_INTERVAL));
 	phba->hb_outstanding = 0;
 	phba->last_completion_time = jiffies;
 	/* Set up error attention (ERATT) polling timer */
-	mod_timer(&phba->eratt_poll, jiffies + HZ * LPFC_ERATT_POLL_INTERVAL);
+	mod_timer(&phba->eratt_poll,
+		  jiffies + msecs_to_jiffies(1000 * LPFC_ERATT_POLL_INTERVAL));
 
 	if (phba->hba_flag & LINK_DISABLED) {
 		lpfc_printf_log(phba,
@@ -908,9 +915,9 @@ lpfc_hba_down_post_s4(struct lpfc_hba *phba)
 		psb->pCmd = NULL;
 		psb->status = IOSTAT_SUCCESS;
 	}
-	spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag);
-	list_splice(&aborts, &phba->lpfc_scsi_buf_list);
-	spin_unlock_irqrestore(&phba->scsi_buf_list_lock, iflag);
+	spin_lock_irqsave(&phba->scsi_buf_list_put_lock, iflag);
+	list_splice(&aborts, &phba->lpfc_scsi_buf_list_put);
+	spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, iflag);
 	return 0;
 }
 
@@ -1021,7 +1028,8 @@ lpfc_hb_mbox_cmpl(struct lpfc_hba * phba, LPFC_MBOXQ_t * pmboxq)
 		!(phba->link_state == LPFC_HBA_ERROR) &&
 		!(phba->pport->load_flag & FC_UNLOADING))
 		mod_timer(&phba->hb_tmofunc,
-			jiffies + HZ * LPFC_HB_MBOX_INTERVAL);
+			  jiffies +
+			  msecs_to_jiffies(1000 * LPFC_HB_MBOX_INTERVAL));
 	return;
 }
 
@@ -1064,15 +1072,18 @@ lpfc_hb_timeout_handler(struct lpfc_hba *phba)
 
 	spin_lock_irq(&phba->pport->work_port_lock);
 
-	if (time_after(phba->last_completion_time + LPFC_HB_MBOX_INTERVAL * HZ,
-		jiffies)) {
+	if (time_after(phba->last_completion_time +
+			msecs_to_jiffies(1000 * LPFC_HB_MBOX_INTERVAL),
+			jiffies)) {
 		spin_unlock_irq(&phba->pport->work_port_lock);
 		if (!phba->hb_outstanding)
 			mod_timer(&phba->hb_tmofunc,
-				jiffies + HZ * LPFC_HB_MBOX_INTERVAL);
+				jiffies +
+				msecs_to_jiffies(1000 * LPFC_HB_MBOX_INTERVAL));
 		else
 			mod_timer(&phba->hb_tmofunc,
-				jiffies + HZ * LPFC_HB_MBOX_TIMEOUT);
+				jiffies +
+				msecs_to_jiffies(1000 * LPFC_HB_MBOX_TIMEOUT));
 		return;
 	}
 	spin_unlock_irq(&phba->pport->work_port_lock);
@@ -1104,7 +1115,8 @@ lpfc_hb_timeout_handler(struct lpfc_hba *phba)
 				if (!pmboxq) {
 					mod_timer(&phba->hb_tmofunc,
 						 jiffies +
-						 HZ * LPFC_HB_MBOX_INTERVAL);
+						 msecs_to_jiffies(1000 *
+						 LPFC_HB_MBOX_INTERVAL));
 					return;
 				}
 
@@ -1120,7 +1132,8 @@ lpfc_hb_timeout_handler(struct lpfc_hba *phba)
 							phba->mbox_mem_pool);
 					mod_timer(&phba->hb_tmofunc,
 						jiffies +
-						HZ * LPFC_HB_MBOX_INTERVAL);
+						msecs_to_jiffies(1000 *
+						LPFC_HB_MBOX_INTERVAL));
 					return;
 				}
 				phba->skipped_hb = 0;
@@ -1136,7 +1149,8 @@ lpfc_hb_timeout_handler(struct lpfc_hba *phba)
 				phba->skipped_hb = jiffies;
 
 			mod_timer(&phba->hb_tmofunc,
-				  jiffies + HZ * LPFC_HB_MBOX_TIMEOUT);
+				 jiffies +
+				 msecs_to_jiffies(1000 * LPFC_HB_MBOX_TIMEOUT));
 			return;
 		} else {
 			/*
@@ -1150,7 +1164,8 @@ lpfc_hb_timeout_handler(struct lpfc_hba *phba)
 					jiffies_to_msecs(jiffies
 						 - phba->last_completion_time));
 			mod_timer(&phba->hb_tmofunc,
-				  jiffies + HZ * LPFC_HB_MBOX_TIMEOUT);
+				jiffies +
+				msecs_to_jiffies(1000 * LPFC_HB_MBOX_TIMEOUT));
 		}
 	}
 }
@@ -1191,7 +1206,7 @@ lpfc_offline_eratt(struct lpfc_hba *phba)
  * This routine is called to bring a SLI4 HBA offline when HBA hardware error
  * other than Port Error 6 has been detected.
  **/
-static void
+void
 lpfc_sli4_offline_eratt(struct lpfc_hba *phba)
 {
 	lpfc_offline_prep(phba, LPFC_MBX_NO_WAIT);
@@ -2633,6 +2648,7 @@ lpfc_online(struct lpfc_hba *phba)
 	struct lpfc_vport *vport;
 	struct lpfc_vport **vports;
 	int i;
+	bool vpis_cleared = false;
 
 	if (!phba)
 		return 0;
@@ -2656,6 +2672,10 @@ lpfc_online(struct lpfc_hba *phba)
 			lpfc_unblock_mgmt_io(phba);
 			return 1;
 		}
+		spin_lock_irq(&phba->hbalock);
+		if (!phba->sli4_hba.max_cfg_param.vpi_used)
+			vpis_cleared = true;
+		spin_unlock_irq(&phba->hbalock);
 	} else {
 		if (lpfc_sli_hba_setup(phba)) {	/* Initialize SLI2/SLI3 HBA */
 			lpfc_unblock_mgmt_io(phba);
@@ -2672,8 +2692,13 @@ lpfc_online(struct lpfc_hba *phba)
 			vports[i]->fc_flag &= ~FC_OFFLINE_MODE;
 			if (phba->sli3_options & LPFC_SLI3_NPIV_ENABLED)
 				vports[i]->fc_flag |= FC_VPORT_NEEDS_REG_VPI;
-			if (phba->sli_rev == LPFC_SLI_REV4)
+			if (phba->sli_rev == LPFC_SLI_REV4) {
 				vports[i]->fc_flag |= FC_VPORT_NEEDS_INIT_VPI;
+				if ((vpis_cleared) &&
+				    (vports[i]->port_type !=
+					LPFC_PHYSICAL_PORT))
+					vports[i]->vpi = 0;
+			}
 			spin_unlock_irq(shost->host_lock);
 		}
 		lpfc_destroy_vport_work_array(phba, vports);
@@ -2833,16 +2858,30 @@ lpfc_scsi_free(struct lpfc_hba *phba)
 	struct lpfc_iocbq *io, *io_next;
 
 	spin_lock_irq(&phba->hbalock);
+
 	/* Release all the lpfc_scsi_bufs maintained by this host. */
-	spin_lock(&phba->scsi_buf_list_lock);
-	list_for_each_entry_safe(sb, sb_next, &phba->lpfc_scsi_buf_list, list) {
+
+	spin_lock(&phba->scsi_buf_list_put_lock);
+	list_for_each_entry_safe(sb, sb_next, &phba->lpfc_scsi_buf_list_put,
+				 list) {
 		list_del(&sb->list);
 		pci_pool_free(phba->lpfc_scsi_dma_buf_pool, sb->data,
 			      sb->dma_handle);
 		kfree(sb);
 		phba->total_scsi_bufs--;
 	}
-	spin_unlock(&phba->scsi_buf_list_lock);
+	spin_unlock(&phba->scsi_buf_list_put_lock);
+
+	spin_lock(&phba->scsi_buf_list_get_lock);
+	list_for_each_entry_safe(sb, sb_next, &phba->lpfc_scsi_buf_list_get,
+				 list) {
+		list_del(&sb->list);
+		pci_pool_free(phba->lpfc_scsi_dma_buf_pool, sb->data,
+			      sb->dma_handle);
+		kfree(sb);
+		phba->total_scsi_bufs--;
+	}
+	spin_unlock(&phba->scsi_buf_list_get_lock);
 
 	/* Release all the lpfc_iocbq entries maintained by this host. */
 	list_for_each_entry_safe(io, io_next, &phba->lpfc_iocb_list, list) {
@@ -2978,9 +3017,12 @@ lpfc_sli4_xri_sgl_update(struct lpfc_hba *phba)
 			phba->sli4_hba.scsi_xri_cnt,
 			phba->sli4_hba.scsi_xri_max);
 
-	spin_lock_irq(&phba->scsi_buf_list_lock);
-	list_splice_init(&phba->lpfc_scsi_buf_list, &scsi_sgl_list);
-	spin_unlock_irq(&phba->scsi_buf_list_lock);
+	spin_lock_irq(&phba->scsi_buf_list_get_lock);
+	spin_lock_irq(&phba->scsi_buf_list_put_lock);
+	list_splice_init(&phba->lpfc_scsi_buf_list_get, &scsi_sgl_list);
+	list_splice(&phba->lpfc_scsi_buf_list_put, &scsi_sgl_list);
+	spin_unlock_irq(&phba->scsi_buf_list_put_lock);
+	spin_unlock_irq(&phba->scsi_buf_list_get_lock);
 
 	if (phba->sli4_hba.scsi_xri_cnt > phba->sli4_hba.scsi_xri_max) {
 		/* max scsi xri shrinked below the allocated scsi buffers */
@@ -2994,9 +3036,9 @@ lpfc_sli4_xri_sgl_update(struct lpfc_hba *phba)
 				      psb->dma_handle);
 			kfree(psb);
 		}
-		spin_lock_irq(&phba->scsi_buf_list_lock);
+		spin_lock_irq(&phba->scsi_buf_list_get_lock);
 		phba->sli4_hba.scsi_xri_cnt -= scsi_xri_cnt;
-		spin_unlock_irq(&phba->scsi_buf_list_lock);
+		spin_unlock_irq(&phba->scsi_buf_list_get_lock);
 	}
 
 	/* update xris associated to remaining allocated scsi buffers */
@@ -3014,9 +3056,12 @@ lpfc_sli4_xri_sgl_update(struct lpfc_hba *phba)
 		psb->cur_iocbq.sli4_lxritag = lxri;
 		psb->cur_iocbq.sli4_xritag = phba->sli4_hba.xri_ids[lxri];
 	}
-	spin_lock_irq(&phba->scsi_buf_list_lock);
-	list_splice_init(&scsi_sgl_list, &phba->lpfc_scsi_buf_list);
-	spin_unlock_irq(&phba->scsi_buf_list_lock);
+	spin_lock_irq(&phba->scsi_buf_list_get_lock);
+	spin_lock_irq(&phba->scsi_buf_list_put_lock);
+	list_splice_init(&scsi_sgl_list, &phba->lpfc_scsi_buf_list_get);
+	INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list_put);
+	spin_unlock_irq(&phba->scsi_buf_list_put_lock);
+	spin_unlock_irq(&phba->scsi_buf_list_get_lock);
 
 	return 0;
 
@@ -3197,14 +3242,15 @@ int lpfc_scan_finished(struct Scsi_Host *shost, unsigned long time)
 		stat = 1;
 		goto finished;
 	}
-	if (time >= 30 * HZ) {
+	if (time >= msecs_to_jiffies(30 * 1000)) {
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
 				"0461 Scanning longer than 30 "
 				"seconds.  Continuing initialization\n");
 		stat = 1;
 		goto finished;
 	}
-	if (time >= 15 * HZ && phba->link_state <= LPFC_LINK_DOWN) {
+	if (time >= msecs_to_jiffies(15 * 1000) &&
+	    phba->link_state <= LPFC_LINK_DOWN) {
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
 				"0465 Link down longer than 15 "
 				"seconds.  Continuing initialization\n");
@@ -3216,7 +3262,7 @@ int lpfc_scan_finished(struct Scsi_Host *shost, unsigned long time)
 		goto finished;
 	if (vport->num_disc_nodes || vport->fc_prli_sent)
 		goto finished;
-	if (vport->fc_map_cnt == 0 && time < 2 * HZ)
+	if (vport->fc_map_cnt == 0 && time < msecs_to_jiffies(2 * 1000))
 		goto finished;
 	if ((phba->sli.sli_flag & LPFC_SLI_MBOX_ACTIVE) != 0)
 		goto finished;
@@ -4215,7 +4261,8 @@ lpfc_sli4_async_fip_evt(struct lpfc_hba *phba,
 			 * If there are other active VLinks present,
 			 * re-instantiate the Vlink using FDISC.
 			 */
-			mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ);
+			mod_timer(&ndlp->nlp_delayfunc,
+				  jiffies + msecs_to_jiffies(1000));
 			shost = lpfc_shost_from_vport(vport);
 			spin_lock_irq(shost->host_lock);
 			ndlp->nlp_flag |= NLP_DELAY_TMO;
@@ -4707,23 +4754,52 @@ lpfc_sli_driver_resource_setup(struct lpfc_hba *phba)
 		return -ENOMEM;
 
 	/*
-	 * Since the sg_tablesize is module parameter, the sg_dma_buf_size
+	 * Since lpfc_sg_seg_cnt is module parameter, the sg_dma_buf_size
 	 * used to create the sg_dma_buf_pool must be dynamically calculated.
-	 * 2 segments are added since the IOCB needs a command and response bde.
 	 */
-	phba->cfg_sg_dma_buf_size = sizeof(struct fcp_cmnd) +
-		sizeof(struct fcp_rsp) +
-			((phba->cfg_sg_seg_cnt + 2) * sizeof(struct ulp_bde64));
 
+	/* Initialize the host templates the configured values. */
+	lpfc_vport_template.sg_tablesize = phba->cfg_sg_seg_cnt;
+	lpfc_template.sg_tablesize = phba->cfg_sg_seg_cnt;
+
+	/* There are going to be 2 reserved BDEs: 1 FCP cmnd + 1 FCP rsp */
 	if (phba->cfg_enable_bg) {
-		phba->cfg_sg_seg_cnt = LPFC_MAX_SG_SEG_CNT;
-		phba->cfg_sg_dma_buf_size +=
-			phba->cfg_prot_sg_seg_cnt * sizeof(struct ulp_bde64);
+		/*
+		 * The scsi_buf for a T10-DIF I/O will hold the FCP cmnd,
+		 * the FCP rsp, and a BDE for each. Sice we have no control
+		 * over how many protection data segments the SCSI Layer
+		 * will hand us (ie: there could be one for every block
+		 * in the IO), we just allocate enough BDEs to accomidate
+		 * our max amount and we need to limit lpfc_sg_seg_cnt to
+		 * minimize the risk of running out.
+		 */
+		phba->cfg_sg_dma_buf_size = sizeof(struct fcp_cmnd) +
+			sizeof(struct fcp_rsp) +
+			(LPFC_MAX_SG_SEG_CNT * sizeof(struct ulp_bde64));
+
+		if (phba->cfg_sg_seg_cnt > LPFC_MAX_SG_SEG_CNT_DIF)
+			phba->cfg_sg_seg_cnt = LPFC_MAX_SG_SEG_CNT_DIF;
+
+		/* Total BDEs in BPL for scsi_sg_list and scsi_sg_prot_list */
+		phba->cfg_total_seg_cnt = LPFC_MAX_SG_SEG_CNT;
+	} else {
+		/*
+		 * The scsi_buf for a regular I/O will hold the FCP cmnd,
+		 * the FCP rsp, a BDE for each, and a BDE for up to
+		 * cfg_sg_seg_cnt data segments.
+		 */
+		phba->cfg_sg_dma_buf_size = sizeof(struct fcp_cmnd) +
+			sizeof(struct fcp_rsp) +
+			((phba->cfg_sg_seg_cnt + 2) * sizeof(struct ulp_bde64));
+
+		/* Total BDEs in BPL for scsi_sg_list */
+		phba->cfg_total_seg_cnt = phba->cfg_sg_seg_cnt + 2;
 	}
 
-	/* Also reinitialize the host templates with new values. */
-	lpfc_vport_template.sg_tablesize = phba->cfg_sg_seg_cnt;
-	lpfc_template.sg_tablesize = phba->cfg_sg_seg_cnt;
+	lpfc_printf_log(phba, KERN_INFO, LOG_INIT | LOG_FCP,
+			"9088 sg_tablesize:%d dmabuf_size:%d total_bde:%d\n",
+			phba->cfg_sg_seg_cnt, phba->cfg_sg_dma_buf_size,
+			phba->cfg_total_seg_cnt);
 
 	phba->max_vpi = LPFC_MAX_VPI;
 	/* This will be set to correct value after config_port mbox */
@@ -4789,13 +4865,13 @@ lpfc_sli_driver_resource_unset(struct lpfc_hba *phba)
 static int
 lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
 {
+	struct lpfc_vector_map_info *cpup;
 	struct lpfc_sli *psli;
 	LPFC_MBOXQ_t *mboxq;
-	int rc, i, hbq_count, buf_size, dma_buf_size, max_buf_size;
+	int rc, i, hbq_count, max_buf_size;
 	uint8_t pn_page[LPFC_MAX_SUPPORTED_PAGES] = {0};
 	struct lpfc_mqe *mqe;
-	int longs, sli_family;
-	int sges_per_segment;
+	int longs;
 
 	/* Before proceed, wait for POST done and device ready */
 	rc = lpfc_sli4_post_status_check(phba);
@@ -4863,11 +4939,6 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
 	phba->fc_map[1] = LPFC_FCOE_FCF_MAP1;
 	phba->fc_map[2] = LPFC_FCOE_FCF_MAP2;
 
-	/* With BlockGuard we can have multiple SGEs per Data Segemnt */
-	sges_per_segment = 1;
-	if (phba->cfg_enable_bg)
-		sges_per_segment = 2;
-
 	/*
 	 * For SLI4, instead of using ring 0 (LPFC_FCP_RING) for FCP commands
 	 * we will associate a new ring, for each FCP fastpath EQ/CQ/WQ tuple.
@@ -4878,43 +4949,71 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
 			sizeof(struct lpfc_sli_ring), GFP_KERNEL);
 	if (!phba->sli.ring)
 		return -ENOMEM;
+
 	/*
-	 * Since the sg_tablesize is module parameter, the sg_dma_buf_size
+	 * It doesn't matter what family our adapter is in, we are
+	 * limited to 2 Pages, 512 SGEs, for our SGL.
+	 * There are going to be 2 reserved SGEs: 1 FCP cmnd + 1 FCP rsp
+	 */
+	max_buf_size = (2 * SLI4_PAGE_SIZE);
+	if (phba->cfg_sg_seg_cnt > LPFC_MAX_SGL_SEG_CNT - 2)
+		phba->cfg_sg_seg_cnt = LPFC_MAX_SGL_SEG_CNT - 2;
+
+	/*
+	 * Since lpfc_sg_seg_cnt is module parameter, the sg_dma_buf_size
 	 * used to create the sg_dma_buf_pool must be dynamically calculated.
-	 * 2 segments are added since the IOCB needs a command and response bde.
-	 * To insure that the scsi sgl does not cross a 4k page boundary only
-	 * sgl sizes of must be a power of 2.
 	 */
-	buf_size = (sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp) +
-		    (((phba->cfg_sg_seg_cnt * sges_per_segment) + 2) *
-		    sizeof(struct sli4_sge)));
-
-	sli_family = bf_get(lpfc_sli_intf_sli_family, &phba->sli4_hba.sli_intf);
-	max_buf_size = LPFC_SLI4_MAX_BUF_SIZE;
-	switch (sli_family) {
-	case LPFC_SLI_INTF_FAMILY_BE2:
-	case LPFC_SLI_INTF_FAMILY_BE3:
-		/* There is a single hint for BE - 2 pages per BPL. */
-		if (bf_get(lpfc_sli_intf_sli_hint1, &phba->sli4_hba.sli_intf) ==
-		    LPFC_SLI_INTF_SLI_HINT1_1)
-			max_buf_size = LPFC_SLI4_FL1_MAX_BUF_SIZE;
-		break;
-	case LPFC_SLI_INTF_FAMILY_LNCR_A0:
-	case LPFC_SLI_INTF_FAMILY_LNCR_B0:
-	default:
-		break;
+
+	if (phba->cfg_enable_bg) {
+		/*
+		 * The scsi_buf for a T10-DIF I/O will hold the FCP cmnd,
+		 * the FCP rsp, and a SGE for each. Sice we have no control
+		 * over how many protection data segments the SCSI Layer
+		 * will hand us (ie: there could be one for every block
+		 * in the IO), we just allocate enough SGEs to accomidate
+		 * our max amount and we need to limit lpfc_sg_seg_cnt to
+		 * minimize the risk of running out.
+		 */
+		phba->cfg_sg_dma_buf_size = sizeof(struct fcp_cmnd) +
+			sizeof(struct fcp_rsp) + max_buf_size;
+
+		/* Total SGEs for scsi_sg_list and scsi_sg_prot_list */
+		phba->cfg_total_seg_cnt = LPFC_MAX_SGL_SEG_CNT;
+
+		if (phba->cfg_sg_seg_cnt > LPFC_MAX_SG_SLI4_SEG_CNT_DIF)
+			phba->cfg_sg_seg_cnt = LPFC_MAX_SG_SLI4_SEG_CNT_DIF;
+	} else {
+		/*
+		 * The scsi_buf for a regular I/O will hold the FCP cmnd,
+		 * the FCP rsp, a SGE for each, and a SGE for up to
+		 * cfg_sg_seg_cnt data segments.
+		 */
+		phba->cfg_sg_dma_buf_size = sizeof(struct fcp_cmnd) +
+			sizeof(struct fcp_rsp) +
+			((phba->cfg_sg_seg_cnt + 2) * sizeof(struct sli4_sge));
+
+		/* Total SGEs for scsi_sg_list */
+		phba->cfg_total_seg_cnt = phba->cfg_sg_seg_cnt + 2;
+		/*
+		 * NOTE: if (phba->cfg_sg_seg_cnt + 2) <= 256 we only need
+		 * to post 1 page for the SGL.
+		 */
 	}
 
-	for (dma_buf_size = LPFC_SLI4_MIN_BUF_SIZE;
-	     dma_buf_size < max_buf_size && buf_size > dma_buf_size;
-	     dma_buf_size = dma_buf_size << 1)
-		;
-	if (dma_buf_size == max_buf_size)
-		phba->cfg_sg_seg_cnt = (dma_buf_size -
-			sizeof(struct fcp_cmnd) - sizeof(struct fcp_rsp) -
-			(2 * sizeof(struct sli4_sge))) /
-				sizeof(struct sli4_sge);
-	phba->cfg_sg_dma_buf_size = dma_buf_size;
+	/* Initialize the host templates with the updated values. */
+	lpfc_vport_template.sg_tablesize = phba->cfg_sg_seg_cnt;
+	lpfc_template.sg_tablesize = phba->cfg_sg_seg_cnt;
+
+	if (phba->cfg_sg_dma_buf_size  <= LPFC_MIN_SG_SLI4_BUF_SZ)
+		phba->cfg_sg_dma_buf_size = LPFC_MIN_SG_SLI4_BUF_SZ;
+	else
+		phba->cfg_sg_dma_buf_size =
+			SLI4_PAGE_ALIGN(phba->cfg_sg_dma_buf_size);
+
+	lpfc_printf_log(phba, KERN_INFO, LOG_INIT | LOG_FCP,
+			"9087 sg_tablesize:%d dmabuf_size:%d total_sge:%d\n",
+			phba->cfg_sg_seg_cnt, phba->cfg_sg_dma_buf_size,
+			phba->cfg_total_seg_cnt);
 
 	/* Initialize buffer queue management fields */
 	hbq_count = lpfc_sli_hbq_count();
@@ -5104,6 +5203,26 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
 		goto out_free_fcp_eq_hdl;
 	}
 
+	phba->sli4_hba.cpu_map = kzalloc((sizeof(struct lpfc_vector_map_info) *
+					 phba->sli4_hba.num_present_cpu),
+					 GFP_KERNEL);
+	if (!phba->sli4_hba.cpu_map) {
+		lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+				"3327 Failed allocate memory for msi-x "
+				"interrupt vector mapping\n");
+		rc = -ENOMEM;
+		goto out_free_msix;
+	}
+	/* Initialize io channels for round robin */
+	cpup = phba->sli4_hba.cpu_map;
+	rc = 0;
+	for (i = 0; i < phba->sli4_hba.num_present_cpu; i++) {
+		cpup->channel_id = rc;
+		rc++;
+		if (rc >= phba->cfg_fcp_io_channel)
+			rc = 0;
+	}
+
 	/*
 	 * Enable sr-iov virtual functions if supported and configured
 	 * through the module parameter.
@@ -5123,6 +5242,8 @@ lpfc_sli4_driver_resource_setup(struct lpfc_hba *phba)
 
 	return 0;
 
+out_free_msix:
+	kfree(phba->sli4_hba.msix_entries);
 out_free_fcp_eq_hdl:
 	kfree(phba->sli4_hba.fcp_eq_hdl);
 out_free_fcf_rr_bmask:
@@ -5152,6 +5273,11 @@ lpfc_sli4_driver_resource_unset(struct lpfc_hba *phba)
 {
 	struct lpfc_fcf_conn_entry *conn_entry, *next_conn_entry;
 
+	/* Free memory allocated for msi-x interrupt vector to CPU mapping */
+	kfree(phba->sli4_hba.cpu_map);
+	phba->sli4_hba.num_present_cpu = 0;
+	phba->sli4_hba.num_online_cpu = 0;
+
 	/* Free memory allocated for msi-x interrupt vector entries */
 	kfree(phba->sli4_hba.msix_entries);
 
@@ -5260,8 +5386,10 @@ lpfc_setup_driver_resource_phase1(struct lpfc_hba *phba)
 	init_waitqueue_head(&phba->work_waitq);
 
 	/* Initialize the scsi buffer list used by driver for scsi IO */
-	spin_lock_init(&phba->scsi_buf_list_lock);
-	INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list);
+	spin_lock_init(&phba->scsi_buf_list_get_lock);
+	INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list_get);
+	spin_lock_init(&phba->scsi_buf_list_put_lock);
+	INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list_put);
 
 	/* Initialize the fabric iocb list */
 	INIT_LIST_HEAD(&phba->fabric_iocb_list);
@@ -6696,6 +6824,7 @@ lpfc_sli4_queue_verify(struct lpfc_hba *phba)
 	int cfg_fcp_io_channel;
 	uint32_t cpu;
 	uint32_t i = 0;
+	uint32_t j = 0;
 
 
 	/*
@@ -6706,15 +6835,21 @@ lpfc_sli4_queue_verify(struct lpfc_hba *phba)
 	/* Sanity check on HBA EQ parameters */
 	cfg_fcp_io_channel = phba->cfg_fcp_io_channel;
 
-	/* It doesn't make sense to have more io channels then CPUs */
-	for_each_online_cpu(cpu) {
-		i++;
+	/* It doesn't make sense to have more io channels then online CPUs */
+	for_each_present_cpu(cpu) {
+		if (cpu_online(cpu))
+			i++;
+		j++;
 	}
+	phba->sli4_hba.num_online_cpu = i;
+	phba->sli4_hba.num_present_cpu = j;
+
 	if (i < cfg_fcp_io_channel) {
 		lpfc_printf_log(phba,
 				KERN_ERR, LOG_INIT,
 				"3188 Reducing IO channels to match number of "
-				"CPUs: from %d to %d\n", cfg_fcp_io_channel, i);
+				"online CPUs: from %d to %d\n",
+				cfg_fcp_io_channel, i);
 		cfg_fcp_io_channel = i;
 	}
 
@@ -7743,8 +7878,13 @@ lpfc_pci_function_reset(struct lpfc_hba *phba)
 
 out:
 	/* Catch the not-ready port failure after a port reset. */
-	if (num_resets >= MAX_IF_TYPE_2_RESETS)
+	if (num_resets >= MAX_IF_TYPE_2_RESETS) {
+		lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+				"3317 HBA not functional: IP Reset Failed "
+				"after (%d) retries, try: "
+				"echo fw_reset > board_mode\n", num_resets);
 		rc = -ENODEV;
+	}
 
 	return rc;
 }
@@ -8209,6 +8349,269 @@ lpfc_sli_disable_intr(struct lpfc_hba *phba)
 }
 
 /**
+ * lpfc_find_next_cpu - Find next available CPU that matches the phys_id
+ * @phba: pointer to lpfc hba data structure.
+ *
+ * Find next available CPU to use for IRQ to CPU affinity.
+ */
+static int
+lpfc_find_next_cpu(struct lpfc_hba *phba, uint32_t phys_id)
+{
+	struct lpfc_vector_map_info *cpup;
+	int cpu;
+
+	cpup = phba->sli4_hba.cpu_map;
+	for (cpu = 0; cpu < phba->sli4_hba.num_present_cpu; cpu++) {
+		/* CPU must be online */
+		if (cpu_online(cpu)) {
+			if ((cpup->irq == LPFC_VECTOR_MAP_EMPTY) &&
+			    (lpfc_used_cpu[cpu] == LPFC_VECTOR_MAP_EMPTY) &&
+			    (cpup->phys_id == phys_id)) {
+				return cpu;
+			}
+		}
+		cpup++;
+	}
+
+	/*
+	 * If we get here, we have used ALL CPUs for the specific
+	 * phys_id. Now we need to clear out lpfc_used_cpu and start
+	 * reusing CPUs.
+	 */
+
+	for (cpu = 0; cpu < phba->sli4_hba.num_present_cpu; cpu++) {
+		if (lpfc_used_cpu[cpu] == phys_id)
+			lpfc_used_cpu[cpu] = LPFC_VECTOR_MAP_EMPTY;
+	}
+
+	cpup = phba->sli4_hba.cpu_map;
+	for (cpu = 0; cpu < phba->sli4_hba.num_present_cpu; cpu++) {
+		/* CPU must be online */
+		if (cpu_online(cpu)) {
+			if ((cpup->irq == LPFC_VECTOR_MAP_EMPTY) &&
+			    (cpup->phys_id == phys_id)) {
+				return cpu;
+			}
+		}
+		cpup++;
+	}
+	return LPFC_VECTOR_MAP_EMPTY;
+}
+
+/**
+ * lpfc_sli4_set_affinity - Set affinity for HBA IRQ vectors
+ * @phba:	pointer to lpfc hba data structure.
+ * @vectors:	number of HBA vectors
+ *
+ * Affinitize MSIX IRQ vectors to CPUs. Try to equally spread vector
+ * affinization across multple physical CPUs (numa nodes).
+ * In addition, this routine will assign an IO channel for each CPU
+ * to use when issuing I/Os.
+ */
+static int
+lpfc_sli4_set_affinity(struct lpfc_hba *phba, int vectors)
+{
+	int i, idx, saved_chann, used_chann, cpu, phys_id;
+	int max_phys_id, num_io_channel, first_cpu;
+	struct lpfc_vector_map_info *cpup;
+#ifdef CONFIG_X86
+	struct cpuinfo_x86 *cpuinfo;
+#endif
+	struct cpumask *mask;
+	uint8_t chann[LPFC_FCP_IO_CHAN_MAX+1];
+
+	/* If there is no mapping, just return */
+	if (!phba->cfg_fcp_cpu_map)
+		return 1;
+
+	/* Init cpu_map array */
+	memset(phba->sli4_hba.cpu_map, 0xff,
+	       (sizeof(struct lpfc_vector_map_info) *
+		phba->sli4_hba.num_present_cpu));
+
+	max_phys_id = 0;
+	phys_id = 0;
+	num_io_channel = 0;
+	first_cpu = LPFC_VECTOR_MAP_EMPTY;
+
+	/* Update CPU map with physical id and core id of each CPU */
+	cpup = phba->sli4_hba.cpu_map;
+	for (cpu = 0; cpu < phba->sli4_hba.num_present_cpu; cpu++) {
+#ifdef CONFIG_X86
+		cpuinfo = &cpu_data(cpu);
+		cpup->phys_id = cpuinfo->phys_proc_id;
+		cpup->core_id = cpuinfo->cpu_core_id;
+#else
+		/* No distinction between CPUs for other platforms */
+		cpup->phys_id = 0;
+		cpup->core_id = 0;
+#endif
+
+		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
+				"3328 CPU physid %d coreid %d\n",
+				cpup->phys_id, cpup->core_id);
+
+		if (cpup->phys_id > max_phys_id)
+			max_phys_id = cpup->phys_id;
+		cpup++;
+	}
+
+	/* Now associate the HBA vectors with specific CPUs */
+	for (idx = 0; idx < vectors; idx++) {
+		cpup = phba->sli4_hba.cpu_map;
+		cpu = lpfc_find_next_cpu(phba, phys_id);
+		if (cpu == LPFC_VECTOR_MAP_EMPTY) {
+
+			/* Try for all phys_id's */
+			for (i = 1; i < max_phys_id; i++) {
+				phys_id++;
+				if (phys_id > max_phys_id)
+					phys_id = 0;
+				cpu = lpfc_find_next_cpu(phba, phys_id);
+				if (cpu == LPFC_VECTOR_MAP_EMPTY)
+					continue;
+				goto found;
+			}
+
+			lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+					"3329 Cannot set affinity:"
+					"Error mapping vector %d (%d)\n",
+					idx, vectors);
+			return 0;
+		}
+found:
+		cpup += cpu;
+		if (phba->cfg_fcp_cpu_map == LPFC_DRIVER_CPU_MAP)
+			lpfc_used_cpu[cpu] = phys_id;
+
+		/* Associate vector with selected CPU */
+		cpup->irq = phba->sli4_hba.msix_entries[idx].vector;
+
+		/* Associate IO channel with selected CPU */
+		cpup->channel_id = idx;
+		num_io_channel++;
+
+		if (first_cpu == LPFC_VECTOR_MAP_EMPTY)
+			first_cpu = cpu;
+
+		/* Now affinitize to the selected CPU */
+		mask = &cpup->maskbits;
+		cpumask_clear(mask);
+		cpumask_set_cpu(cpu, mask);
+		i = irq_set_affinity_hint(phba->sli4_hba.msix_entries[idx].
+					  vector, mask);
+
+		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
+				"3330 Set Affinity: CPU %d channel %d "
+				"irq %d (%x)\n",
+				cpu, cpup->channel_id,
+				phba->sli4_hba.msix_entries[idx].vector, i);
+
+		/* Spread vector mapping across multple physical CPU nodes */
+		phys_id++;
+		if (phys_id > max_phys_id)
+			phys_id = 0;
+	}
+
+	/*
+	 * Finally fill in the IO channel for any remaining CPUs.
+	 * At this point, all IO channels have been assigned to a specific
+	 * MSIx vector, mapped to a specific CPU.
+	 * Base the remaining IO channel assigned, to IO channels already
+	 * assigned to other CPUs on the same phys_id.
+	 */
+	for (i = 0; i <= max_phys_id; i++) {
+		/*
+		 * If there are no io channels already mapped to
+		 * this phys_id, just round robin thru the io_channels.
+		 * Setup chann[] for round robin.
+		 */
+		for (idx = 0; idx < phba->cfg_fcp_io_channel; idx++)
+			chann[idx] = idx;
+
+		saved_chann = 0;
+		used_chann = 0;
+
+		/*
+		 * First build a list of IO channels already assigned
+		 * to this phys_id before reassigning the same IO
+		 * channels to the remaining CPUs.
+		 */
+		cpup = phba->sli4_hba.cpu_map;
+		cpu = first_cpu;
+		cpup += cpu;
+		for (idx = 0; idx < phba->sli4_hba.num_present_cpu;
+		     idx++) {
+			if (cpup->phys_id == i) {
+				/*
+				 * Save any IO channels that are
+				 * already mapped to this phys_id.
+				 */
+				if (cpup->irq != LPFC_VECTOR_MAP_EMPTY) {
+					chann[saved_chann] =
+						cpup->channel_id;
+					saved_chann++;
+					goto out;
+				}
+
+				/* See if we are using round-robin */
+				if (saved_chann == 0)
+					saved_chann =
+						phba->cfg_fcp_io_channel;
+
+				/* Associate next IO channel with CPU */
+				cpup->channel_id = chann[used_chann];
+				num_io_channel++;
+				used_chann++;
+				if (used_chann == saved_chann)
+					used_chann = 0;
+
+				lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
+						"3331 Set IO_CHANN "
+						"CPU %d channel %d\n",
+						idx, cpup->channel_id);
+			}
+out:
+			cpu++;
+			if (cpu >= phba->sli4_hba.num_present_cpu) {
+				cpup = phba->sli4_hba.cpu_map;
+				cpu = 0;
+			} else {
+				cpup++;
+			}
+		}
+	}
+
+	if (phba->sli4_hba.num_online_cpu != phba->sli4_hba.num_present_cpu) {
+		cpup = phba->sli4_hba.cpu_map;
+		for (idx = 0; idx < phba->sli4_hba.num_present_cpu; idx++) {
+			if (cpup->channel_id == LPFC_VECTOR_MAP_EMPTY) {
+				cpup->channel_id = 0;
+				num_io_channel++;
+
+				lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
+						"3332 Assign IO_CHANN "
+						"CPU %d channel %d\n",
+						idx, cpup->channel_id);
+			}
+			cpup++;
+		}
+	}
+
+	/* Sanity check */
+	if (num_io_channel != phba->sli4_hba.num_present_cpu)
+		lpfc_printf_log(phba, KERN_ERR, LOG_INIT,
+				"3333 Set affinity mismatch:"
+				"%d chann != %d cpus: %d vactors\n",
+				num_io_channel, phba->sli4_hba.num_present_cpu,
+				vectors);
+
+	phba->cfg_fcp_io_sched = LPFC_FCP_SCHED_BY_CPU;
+	return 1;
+}
+
+
+/**
  * lpfc_sli4_enable_msix - Enable MSI-X interrupt mode to SLI-4 device
  * @phba: pointer to lpfc hba data structure.
  *
@@ -8259,9 +8662,7 @@ enable_msix_vectors:
 				phba->sli4_hba.msix_entries[index].vector,
 				phba->sli4_hba.msix_entries[index].entry);
 
-	/*
-	 * Assign MSI-X vectors to interrupt handlers
-	 */
+	/* Assign MSI-X vectors to interrupt handlers */
 	for (index = 0; index < vectors; index++) {
 		memset(&phba->sli4_hba.handler_name[index], 0, 16);
 		sprintf((char *)&phba->sli4_hba.handler_name[index],
@@ -8289,6 +8690,8 @@ enable_msix_vectors:
 				phba->cfg_fcp_io_channel, vectors);
 		phba->cfg_fcp_io_channel = vectors;
 	}
+
+	lpfc_sli4_set_affinity(phba, vectors);
 	return rc;
 
 cfg_fail_out:
@@ -9213,15 +9616,15 @@ lpfc_sli_prep_dev_for_reset(struct lpfc_hba *phba)
 	/* Block all SCSI devices' I/Os on the host */
 	lpfc_scsi_dev_block(phba);
 
+	/* Flush all driver's outstanding SCSI I/Os as we are to reset */
+	lpfc_sli_flush_fcp_rings(phba);
+
 	/* stop all timers */
 	lpfc_stop_hba_timers(phba);
 
 	/* Disable interrupt and pci device */
 	lpfc_sli_disable_intr(phba);
 	pci_disable_device(phba->pcidev);
-
-	/* Flush all driver's outstanding SCSI I/Os as we are to reset */
-	lpfc_sli_flush_fcp_rings(phba);
 }
 
 /**
@@ -9966,6 +10369,9 @@ lpfc_sli4_prep_dev_for_reset(struct lpfc_hba *phba)
 	/* Block all SCSI devices' I/Os on the host */
 	lpfc_scsi_dev_block(phba);
 
+	/* Flush all driver's outstanding SCSI I/Os as we are to reset */
+	lpfc_sli_flush_fcp_rings(phba);
+
 	/* stop all timers */
 	lpfc_stop_hba_timers(phba);
 
@@ -9973,9 +10379,6 @@ lpfc_sli4_prep_dev_for_reset(struct lpfc_hba *phba)
 	lpfc_sli4_disable_intr(phba);
 	lpfc_sli4_queue_destroy(phba);
 	pci_disable_device(phba->pcidev);
-
-	/* Flush all driver's outstanding SCSI I/Os as we are to reset */
-	lpfc_sli_flush_fcp_rings(phba);
 }
 
 /**
@@ -10535,6 +10938,7 @@ static struct miscdevice lpfc_mgmt_dev = {
 static int __init
 lpfc_init(void)
 {
+	int cpu;
 	int error = 0;
 
 	printk(LPFC_MODULE_DESC "\n");
@@ -10561,6 +10965,11 @@ lpfc_init(void)
 			return -ENOMEM;
 		}
 	}
+
+	/* Initialize in case vector mapping is needed */
+	for (cpu = 0; cpu < LPFC_MAX_CPU; cpu++)
+		lpfc_used_cpu[cpu] = LPFC_VECTOR_MAP_EMPTY;
+
 	error = pci_register_driver(&lpfc_driver);
 	if (error) {
 		fc_release_transport(lpfc_transport_template);
diff --git a/drivers/scsi/lpfc/lpfc_logmsg.h b/drivers/scsi/lpfc/lpfc_logmsg.h
index baf53e6c2bd1..2a4e5d21eab2 100644
--- a/drivers/scsi/lpfc/lpfc_logmsg.h
+++ b/drivers/scsi/lpfc/lpfc_logmsg.h
@@ -37,6 +37,7 @@
 #define LOG_EVENT	0x00010000	/* CT,TEMP,DUMP, logging */
 #define LOG_FIP		0x00020000	/* FIP events */
 #define LOG_FCP_UNDER	0x00040000	/* FCP underruns errors */
+#define LOG_SCSI_CMD	0x00080000	/* ALL SCSI commands */
 #define LOG_ALL_MSG	0xffffffff	/* LOG all messages */
 
 #define lpfc_printf_vlog(vport, level, mask, fmt, arg...) \
diff --git a/drivers/scsi/lpfc/lpfc_mbox.c b/drivers/scsi/lpfc/lpfc_mbox.c
index a7a9fa468308..41363db7d426 100644
--- a/drivers/scsi/lpfc/lpfc_mbox.c
+++ b/drivers/scsi/lpfc/lpfc_mbox.c
@@ -2149,18 +2149,21 @@ lpfc_reg_vfi(struct lpfcMboxq *mbox, struct lpfc_vport *vport, dma_addr_t phys)
 
 	/* Only FC supports upd bit */
 	if ((phba->sli4_hba.lnk_info.lnk_tp == LPFC_LNK_TYPE_FC) &&
-	    (vport->fc_flag & FC_VFI_REGISTERED)) {
+	    (vport->fc_flag & FC_VFI_REGISTERED) &&
+	    (!phba->fc_topology_changed)) {
 		bf_set(lpfc_reg_vfi_vp, reg_vfi, 0);
 		bf_set(lpfc_reg_vfi_upd, reg_vfi, 1);
 	}
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_MBOX,
 			"3134 Register VFI, mydid:x%x, fcfi:%d, "
-			" vfi:%d, vpi:%d, fc_pname:%x%x\n",
+			" vfi:%d, vpi:%d, fc_pname:%x%x fc_flag:x%x"
+			" port_state:x%x topology chg:%d\n",
 			vport->fc_myDID,
 			phba->fcf.fcfi,
 			phba->sli4_hba.vfi_ids[vport->vfi],
 			phba->vpi_ids[vport->vpi],
-			reg_vfi->wwn[0], reg_vfi->wwn[1]);
+			reg_vfi->wwn[0], reg_vfi->wwn[1], vport->fc_flag,
+			vport->port_state, phba->fc_topology_changed);
 }
 
 /**
diff --git a/drivers/scsi/lpfc/lpfc_mem.c b/drivers/scsi/lpfc/lpfc_mem.c
index cd86069a0ba8..812d0cd7c86d 100644
--- a/drivers/scsi/lpfc/lpfc_mem.c
+++ b/drivers/scsi/lpfc/lpfc_mem.c
@@ -64,18 +64,26 @@ lpfc_mem_alloc(struct lpfc_hba *phba, int align)
 	struct lpfc_dma_pool *pool = &phba->lpfc_mbuf_safety_pool;
 	int i;
 
-	if (phba->sli_rev == LPFC_SLI_REV4)
+	if (phba->sli_rev == LPFC_SLI_REV4) {
+		/* Calculate alignment */
+		if (phba->cfg_sg_dma_buf_size < SLI4_PAGE_SIZE)
+			i = phba->cfg_sg_dma_buf_size;
+		else
+			i = SLI4_PAGE_SIZE;
+
 		phba->lpfc_scsi_dma_buf_pool =
 			pci_pool_create("lpfc_scsi_dma_buf_pool",
 				phba->pcidev,
 				phba->cfg_sg_dma_buf_size,
-				phba->cfg_sg_dma_buf_size,
+				i,
 				0);
-	else
+	} else {
 		phba->lpfc_scsi_dma_buf_pool =
 			pci_pool_create("lpfc_scsi_dma_buf_pool",
 				phba->pcidev, phba->cfg_sg_dma_buf_size,
 				align, 0);
+	}
+
 	if (!phba->lpfc_scsi_dma_buf_pool)
 		goto fail;
 
diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index 82f4d3542289..31e9b92f5a9b 100644
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -332,9 +332,11 @@ lpfc_rcv_plogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 
 	/* PLOGI chkparm OK */
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
-			 "0114 PLOGI chkparm OK Data: x%x x%x x%x x%x\n",
+			 "0114 PLOGI chkparm OK Data: x%x x%x x%x "
+			 "x%x x%x x%x\n",
 			 ndlp->nlp_DID, ndlp->nlp_state, ndlp->nlp_flag,
-			 ndlp->nlp_rpi);
+			 ndlp->nlp_rpi, vport->port_state,
+			 vport->fc_flag);
 
 	if (vport->cfg_fcp_class == 2 && sp->cls2.classValid)
 		ndlp->nlp_fcp_info |= CLASS2;
@@ -574,7 +576,7 @@ out:
 	lpfc_els_rsp_reject(vport, stat.un.lsRjtError, cmdiocb, ndlp, NULL);
 
 	/* 1 sec timeout */
-	mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ);
+	mod_timer(&ndlp->nlp_delayfunc, jiffies + msecs_to_jiffies(1000));
 
 	spin_lock_irq(shost->host_lock);
 	ndlp->nlp_flag |= NLP_DELAY_TMO;
@@ -631,7 +633,8 @@ lpfc_rcv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 			 * If there are other active VLinks present,
 			 * re-instantiate the Vlink using FDISC.
 			 */
-			mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ);
+			mod_timer(&ndlp->nlp_delayfunc,
+				  jiffies + msecs_to_jiffies(1000));
 			spin_lock_irq(shost->host_lock);
 			ndlp->nlp_flag |= NLP_DELAY_TMO;
 			spin_unlock_irq(shost->host_lock);
@@ -648,7 +651,8 @@ lpfc_rcv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		!(ndlp->nlp_type & NLP_FCP_INITIATOR))) ||
 		(ndlp->nlp_state == NLP_STE_ADISC_ISSUE)) {
 		/* Only try to re-login if this is NOT a Fabric Node */
-		mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ * 1);
+		mod_timer(&ndlp->nlp_delayfunc,
+			  jiffies + msecs_to_jiffies(1000 * 1));
 		spin_lock_irq(shost->host_lock);
 		ndlp->nlp_flag |= NLP_DELAY_TMO;
 		spin_unlock_irq(shost->host_lock);
@@ -969,7 +973,7 @@ lpfc_rcv_els_plogi_issue(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 	}
 
 	/* Put ndlp in npr state set plogi timer for 1 sec */
-	mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ * 1);
+	mod_timer(&ndlp->nlp_delayfunc, jiffies + msecs_to_jiffies(1000 * 1));
 	spin_lock_irq(shost->host_lock);
 	ndlp->nlp_flag |= NLP_DELAY_TMO;
 	spin_unlock_irq(shost->host_lock);
@@ -1303,7 +1307,8 @@ lpfc_cmpl_adisc_adisc_issue(struct lpfc_vport *vport,
 	if ((irsp->ulpStatus) ||
 	    (!lpfc_check_adisc(vport, ndlp, &ap->nodeName, &ap->portName))) {
 		/* 1 sec timeout */
-		mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ);
+		mod_timer(&ndlp->nlp_delayfunc,
+			  jiffies + msecs_to_jiffies(1000));
 		spin_lock_irq(shost->host_lock);
 		ndlp->nlp_flag |= NLP_DELAY_TMO;
 		spin_unlock_irq(shost->host_lock);
@@ -1509,7 +1514,8 @@ lpfc_cmpl_reglogin_reglogin_issue(struct lpfc_vport *vport,
 		}
 
 		/* Put ndlp in npr state set plogi timer for 1 sec */
-		mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ * 1);
+		mod_timer(&ndlp->nlp_delayfunc,
+			  jiffies + msecs_to_jiffies(1000 * 1));
 		spin_lock_irq(shost->host_lock);
 		ndlp->nlp_flag |= NLP_DELAY_TMO;
 		spin_unlock_irq(shost->host_lock);
@@ -2145,7 +2151,8 @@ lpfc_rcv_prlo_npr_node(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 	lpfc_els_rsp_acc(vport, ELS_CMD_ACC, cmdiocb, ndlp, NULL);
 
 	if ((ndlp->nlp_flag & NLP_DELAY_TMO) == 0) {
-		mod_timer(&ndlp->nlp_delayfunc, jiffies + HZ * 1);
+		mod_timer(&ndlp->nlp_delayfunc,
+			  jiffies + msecs_to_jiffies(1000 * 1));
 		spin_lock_irq(shost->host_lock);
 		ndlp->nlp_flag |= NLP_DELAY_TMO;
 		ndlp->nlp_flag &= ~NLP_NPR_ADISC;
diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index 74b8710e1e90..8523b278ec9d 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -24,6 +24,8 @@
 #include <linux/export.h>
 #include <linux/delay.h>
 #include <asm/unaligned.h>
+#include <linux/crc-t10dif.h>
+#include <net/checksum.h>
 
 #include <scsi/scsi.h>
 #include <scsi/scsi_device.h>
@@ -48,7 +50,7 @@
 #define LPFC_RESET_WAIT  2
 #define LPFC_ABORT_WAIT  2
 
-int _dump_buf_done;
+int _dump_buf_done = 1;
 
 static char *dif_op_str[] = {
 	"PROT_NORMAL",
@@ -66,6 +68,10 @@ struct scsi_dif_tuple {
 	__be32 ref_tag;         /* Target LBA or indirect LBA */
 };
 
+#if !defined(SCSI_PROT_GUARD_CHECK) || !defined(SCSI_PROT_REF_CHECK)
+#define scsi_prot_flagged(sc, flg)	sc
+#endif
+
 static void
 lpfc_release_scsi_buf_s4(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb);
 static void
@@ -534,7 +540,16 @@ lpfc_new_scsi_buf_s3(struct lpfc_vport *vport, int num_to_alloc)
 	dma_addr_t pdma_phys_fcp_rsp;
 	dma_addr_t pdma_phys_bpl;
 	uint16_t iotag;
-	int bcnt;
+	int bcnt, bpl_size;
+
+	bpl_size = phba->cfg_sg_dma_buf_size -
+		(sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
+
+	lpfc_printf_vlog(vport, KERN_INFO, LOG_FCP,
+			 "9067 ALLOC %d scsi_bufs: %d (%d + %d + %d)\n",
+			 num_to_alloc, phba->cfg_sg_dma_buf_size,
+			 (int)sizeof(struct fcp_cmnd),
+			 (int)sizeof(struct fcp_rsp), bpl_size);
 
 	for (bcnt = 0; bcnt < num_to_alloc; bcnt++) {
 		psb = kzalloc(sizeof(struct lpfc_scsi_buf), GFP_KERNEL);
@@ -759,7 +774,7 @@ lpfc_sli4_post_scsi_sgl_list(struct lpfc_hba *phba,
 			     struct list_head *post_sblist, int sb_count)
 {
 	struct lpfc_scsi_buf *psb, *psb_next;
-	int status;
+	int status, sgl_size;
 	int post_cnt = 0, block_cnt = 0, num_posting = 0, num_posted = 0;
 	dma_addr_t pdma_phys_bpl1;
 	int last_xritag = NO_XRI;
@@ -771,6 +786,9 @@ lpfc_sli4_post_scsi_sgl_list(struct lpfc_hba *phba,
 	if (sb_count <= 0)
 		return -EINVAL;
 
+	sgl_size = phba->cfg_sg_dma_buf_size -
+		(sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
+
 	list_for_each_entry_safe(psb, psb_next, post_sblist, list) {
 		list_del_init(&psb->list);
 		block_cnt++;
@@ -803,7 +821,7 @@ lpfc_sli4_post_scsi_sgl_list(struct lpfc_hba *phba,
 				post_cnt = block_cnt;
 			} else if (block_cnt == 1) {
 				/* last single sgl with non-contiguous xri */
-				if (phba->cfg_sg_dma_buf_size > SGL_PAGE_SIZE)
+				if (sgl_size > SGL_PAGE_SIZE)
 					pdma_phys_bpl1 = psb->dma_phys_bpl +
 								SGL_PAGE_SIZE;
 				else
@@ -885,9 +903,12 @@ lpfc_sli4_repost_scsi_sgl_list(struct lpfc_hba *phba)
 	int num_posted, rc = 0;
 
 	/* get all SCSI buffers need to repost to a local list */
-	spin_lock_irq(&phba->scsi_buf_list_lock);
-	list_splice_init(&phba->lpfc_scsi_buf_list, &post_sblist);
-	spin_unlock_irq(&phba->scsi_buf_list_lock);
+	spin_lock_irq(&phba->scsi_buf_list_get_lock);
+	spin_lock_irq(&phba->scsi_buf_list_put_lock);
+	list_splice_init(&phba->lpfc_scsi_buf_list_get, &post_sblist);
+	list_splice(&phba->lpfc_scsi_buf_list_put, &post_sblist);
+	spin_unlock_irq(&phba->scsi_buf_list_put_lock);
+	spin_unlock_irq(&phba->scsi_buf_list_get_lock);
 
 	/* post the list of scsi buffer sgls to port if available */
 	if (!list_empty(&post_sblist)) {
@@ -923,13 +944,22 @@ lpfc_new_scsi_buf_s4(struct lpfc_vport *vport, int num_to_alloc)
 	IOCB_t *iocb;
 	dma_addr_t pdma_phys_fcp_cmd;
 	dma_addr_t pdma_phys_fcp_rsp;
-	dma_addr_t pdma_phys_bpl, pdma_phys_bpl1;
+	dma_addr_t pdma_phys_bpl;
 	uint16_t iotag, lxri = 0;
-	int bcnt, num_posted;
+	int bcnt, num_posted, sgl_size;
 	LIST_HEAD(prep_sblist);
 	LIST_HEAD(post_sblist);
 	LIST_HEAD(scsi_sblist);
 
+	sgl_size = phba->cfg_sg_dma_buf_size -
+		(sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
+
+	lpfc_printf_vlog(vport, KERN_INFO, LOG_FCP,
+			 "9068 ALLOC %d scsi_bufs: %d (%d + %d + %d)\n",
+			 num_to_alloc, phba->cfg_sg_dma_buf_size, sgl_size,
+			 (int)sizeof(struct fcp_cmnd),
+			 (int)sizeof(struct fcp_rsp));
+
 	for (bcnt = 0; bcnt < num_to_alloc; bcnt++) {
 		psb = kzalloc(sizeof(struct lpfc_scsi_buf), GFP_KERNEL);
 		if (!psb)
@@ -948,6 +978,15 @@ lpfc_new_scsi_buf_s4(struct lpfc_vport *vport, int num_to_alloc)
 		}
 		memset(psb->data, 0, phba->cfg_sg_dma_buf_size);
 
+		/* Page alignment is CRITICAL, double check to be sure */
+		if (((unsigned long)(psb->data) &
+		    (unsigned long)(SLI4_PAGE_SIZE - 1)) != 0) {
+			pci_pool_free(phba->lpfc_scsi_dma_buf_pool,
+				      psb->data, psb->dma_handle);
+			kfree(psb);
+			break;
+		}
+
 		/* Allocate iotag for psb->cur_iocbq. */
 		iotag = lpfc_sli_next_iotag(phba, &psb->cur_iocbq);
 		if (iotag == 0) {
@@ -968,17 +1007,14 @@ lpfc_new_scsi_buf_s4(struct lpfc_vport *vport, int num_to_alloc)
 		psb->cur_iocbq.sli4_xritag = phba->sli4_hba.xri_ids[lxri];
 		psb->cur_iocbq.iocb_flag |= LPFC_IO_FCP;
 		psb->fcp_bpl = psb->data;
-		psb->fcp_cmnd = (psb->data + phba->cfg_sg_dma_buf_size)
-			- (sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
+		psb->fcp_cmnd = (psb->data + sgl_size);
 		psb->fcp_rsp = (struct fcp_rsp *)((uint8_t *)psb->fcp_cmnd +
 					sizeof(struct fcp_cmnd));
 
 		/* Initialize local short-hand pointers. */
 		sgl = (struct sli4_sge *)psb->fcp_bpl;
 		pdma_phys_bpl = psb->dma_handle;
-		pdma_phys_fcp_cmd =
-			(psb->dma_handle + phba->cfg_sg_dma_buf_size)
-			 - (sizeof(struct fcp_cmnd) + sizeof(struct fcp_rsp));
+		pdma_phys_fcp_cmd = (psb->dma_handle + sgl_size);
 		pdma_phys_fcp_rsp = pdma_phys_fcp_cmd + sizeof(struct fcp_cmnd);
 
 		/*
@@ -1020,17 +1056,13 @@ lpfc_new_scsi_buf_s4(struct lpfc_vport *vport, int num_to_alloc)
 		iocb->ulpLe = 1;
 		iocb->ulpClass = CLASS3;
 		psb->cur_iocbq.context1 = psb;
-		if (phba->cfg_sg_dma_buf_size > SGL_PAGE_SIZE)
-			pdma_phys_bpl1 = pdma_phys_bpl + SGL_PAGE_SIZE;
-		else
-			pdma_phys_bpl1 = 0;
 		psb->dma_phys_bpl = pdma_phys_bpl;
 
 		/* add the scsi buffer to a post list */
 		list_add_tail(&psb->list, &post_sblist);
-		spin_lock_irq(&phba->scsi_buf_list_lock);
+		spin_lock_irq(&phba->scsi_buf_list_get_lock);
 		phba->sli4_hba.scsi_xri_cnt++;
-		spin_unlock_irq(&phba->scsi_buf_list_lock);
+		spin_unlock_irq(&phba->scsi_buf_list_get_lock);
 	}
 	lpfc_printf_log(phba, KERN_INFO, LOG_BG,
 			"3021 Allocate %d out of %d requested new SCSI "
@@ -1079,17 +1111,23 @@ static struct lpfc_scsi_buf*
 lpfc_get_scsi_buf_s3(struct lpfc_hba *phba, struct lpfc_nodelist *ndlp)
 {
 	struct  lpfc_scsi_buf * lpfc_cmd = NULL;
-	struct list_head *scsi_buf_list = &phba->lpfc_scsi_buf_list;
-	unsigned long iflag = 0;
-
-	spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag);
-	list_remove_head(scsi_buf_list, lpfc_cmd, struct lpfc_scsi_buf, list);
-	if (lpfc_cmd) {
-		lpfc_cmd->seg_cnt = 0;
-		lpfc_cmd->nonsg_phys = 0;
-		lpfc_cmd->prot_seg_cnt = 0;
+	struct list_head *scsi_buf_list_get = &phba->lpfc_scsi_buf_list_get;
+	unsigned long gflag = 0;
+	unsigned long pflag = 0;
+
+	spin_lock_irqsave(&phba->scsi_buf_list_get_lock, gflag);
+	list_remove_head(scsi_buf_list_get, lpfc_cmd, struct lpfc_scsi_buf,
+			 list);
+	if (!lpfc_cmd) {
+		spin_lock_irqsave(&phba->scsi_buf_list_put_lock, pflag);
+		list_splice(&phba->lpfc_scsi_buf_list_put,
+			    &phba->lpfc_scsi_buf_list_get);
+		INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list_put);
+		list_remove_head(scsi_buf_list_get, lpfc_cmd,
+				 struct lpfc_scsi_buf, list);
+		spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, pflag);
 	}
-	spin_unlock_irqrestore(&phba->scsi_buf_list_lock, iflag);
+	spin_unlock_irqrestore(&phba->scsi_buf_list_get_lock, gflag);
 	return  lpfc_cmd;
 }
 /**
@@ -1107,28 +1145,39 @@ static struct lpfc_scsi_buf*
 lpfc_get_scsi_buf_s4(struct lpfc_hba *phba, struct lpfc_nodelist *ndlp)
 {
 	struct lpfc_scsi_buf *lpfc_cmd ;
-	unsigned long iflag = 0;
+	unsigned long gflag = 0;
+	unsigned long pflag = 0;
 	int found = 0;
 
-	spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag);
-	list_for_each_entry(lpfc_cmd, &phba->lpfc_scsi_buf_list,
-							list) {
+	spin_lock_irqsave(&phba->scsi_buf_list_get_lock, gflag);
+	list_for_each_entry(lpfc_cmd, &phba->lpfc_scsi_buf_list_get, list) {
 		if (lpfc_test_rrq_active(phba, ndlp,
 					 lpfc_cmd->cur_iocbq.sli4_lxritag))
 			continue;
 		list_del(&lpfc_cmd->list);
 		found = 1;
-		lpfc_cmd->seg_cnt = 0;
-		lpfc_cmd->nonsg_phys = 0;
-		lpfc_cmd->prot_seg_cnt = 0;
 		break;
 	}
-	spin_unlock_irqrestore(&phba->scsi_buf_list_lock,
-						 iflag);
+	if (!found) {
+		spin_lock_irqsave(&phba->scsi_buf_list_put_lock, pflag);
+		list_splice(&phba->lpfc_scsi_buf_list_put,
+			    &phba->lpfc_scsi_buf_list_get);
+		INIT_LIST_HEAD(&phba->lpfc_scsi_buf_list_put);
+		spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, pflag);
+		list_for_each_entry(lpfc_cmd, &phba->lpfc_scsi_buf_list_get,
+				    list) {
+			if (lpfc_test_rrq_active(
+				phba, ndlp, lpfc_cmd->cur_iocbq.sli4_lxritag))
+				continue;
+			list_del(&lpfc_cmd->list);
+			found = 1;
+			break;
+		}
+	}
+	spin_unlock_irqrestore(&phba->scsi_buf_list_get_lock, gflag);
 	if (!found)
 		return NULL;
-	else
-		return  lpfc_cmd;
+	return  lpfc_cmd;
 }
 /**
  * lpfc_get_scsi_buf - Get a scsi buffer from lpfc_scsi_buf_list of the HBA
@@ -1160,10 +1209,15 @@ lpfc_release_scsi_buf_s3(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb)
 {
 	unsigned long iflag = 0;
 
-	spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag);
+	psb->seg_cnt = 0;
+	psb->nonsg_phys = 0;
+	psb->prot_seg_cnt = 0;
+
+	spin_lock_irqsave(&phba->scsi_buf_list_put_lock, iflag);
 	psb->pCmd = NULL;
-	list_add_tail(&psb->list, &phba->lpfc_scsi_buf_list);
-	spin_unlock_irqrestore(&phba->scsi_buf_list_lock, iflag);
+	psb->cur_iocbq.iocb_flag = LPFC_IO_FCP;
+	list_add_tail(&psb->list, &phba->lpfc_scsi_buf_list_put);
+	spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, iflag);
 }
 
 /**
@@ -1181,6 +1235,10 @@ lpfc_release_scsi_buf_s4(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb)
 {
 	unsigned long iflag = 0;
 
+	psb->seg_cnt = 0;
+	psb->nonsg_phys = 0;
+	psb->prot_seg_cnt = 0;
+
 	if (psb->exch_busy) {
 		spin_lock_irqsave(&phba->sli4_hba.abts_scsi_buf_list_lock,
 					iflag);
@@ -1190,11 +1248,11 @@ lpfc_release_scsi_buf_s4(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb)
 		spin_unlock_irqrestore(&phba->sli4_hba.abts_scsi_buf_list_lock,
 					iflag);
 	} else {
-
-		spin_lock_irqsave(&phba->scsi_buf_list_lock, iflag);
 		psb->pCmd = NULL;
-		list_add_tail(&psb->list, &phba->lpfc_scsi_buf_list);
-		spin_unlock_irqrestore(&phba->scsi_buf_list_lock, iflag);
+		psb->cur_iocbq.iocb_flag = LPFC_IO_FCP;
+		spin_lock_irqsave(&phba->scsi_buf_list_put_lock, iflag);
+		list_add_tail(&psb->list, &phba->lpfc_scsi_buf_list_put);
+		spin_unlock_irqrestore(&phba->scsi_buf_list_put_lock, iflag);
 	}
 }
 
@@ -1268,6 +1326,7 @@ lpfc_scsi_prep_dma_buf_s3(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
 			       "dma_map_sg.  Config %d, seg_cnt %d\n",
 			       __func__, phba->cfg_sg_seg_cnt,
 			       lpfc_cmd->seg_cnt);
+			lpfc_cmd->seg_cnt = 0;
 			scsi_dma_unmap(scsi_cmnd);
 			return 1;
 		}
@@ -2013,9 +2072,21 @@ lpfc_bg_setup_bpl(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 	bf_set(pde6_type, pde6, LPFC_PDE6_DESCRIPTOR);
 	bf_set(pde6_optx, pde6, txop);
 	bf_set(pde6_oprx, pde6, rxop);
+
+	/*
+	 * We only need to check the data on READs, for WRITEs
+	 * protection data is automatically generated, not checked.
+	 */
 	if (datadir == DMA_FROM_DEVICE) {
-		bf_set(pde6_ce, pde6, checking);
-		bf_set(pde6_re, pde6, checking);
+		if (scsi_prot_flagged(sc, SCSI_PROT_GUARD_CHECK))
+			bf_set(pde6_ce, pde6, checking);
+		else
+			bf_set(pde6_ce, pde6, 0);
+
+		if (scsi_prot_flagged(sc, SCSI_PROT_REF_CHECK))
+			bf_set(pde6_re, pde6, checking);
+		else
+			bf_set(pde6_re, pde6, 0);
 	}
 	bf_set(pde6_ai, pde6, 1);
 	bf_set(pde6_ae, pde6, 0);
@@ -2145,6 +2216,10 @@ lpfc_bg_setup_bpl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 
 	split_offset = 0;
 	do {
+		/* Check to see if we ran out of space */
+		if (num_bde >= (phba->cfg_total_seg_cnt - 2))
+			return num_bde + 3;
+
 		/* setup PDE5 with what we have */
 		pde5 = (struct lpfc_pde5 *) bpl;
 		memset(pde5, 0, sizeof(struct lpfc_pde5));
@@ -2164,8 +2239,17 @@ lpfc_bg_setup_bpl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 		bf_set(pde6_type, pde6, LPFC_PDE6_DESCRIPTOR);
 		bf_set(pde6_optx, pde6, txop);
 		bf_set(pde6_oprx, pde6, rxop);
-		bf_set(pde6_ce, pde6, checking);
-		bf_set(pde6_re, pde6, checking);
+
+		if (scsi_prot_flagged(sc, SCSI_PROT_GUARD_CHECK))
+			bf_set(pde6_ce, pde6, checking);
+		else
+			bf_set(pde6_ce, pde6, 0);
+
+		if (scsi_prot_flagged(sc, SCSI_PROT_REF_CHECK))
+			bf_set(pde6_re, pde6, checking);
+		else
+			bf_set(pde6_re, pde6, 0);
+
 		bf_set(pde6_ai, pde6, 1);
 		bf_set(pde6_ae, pde6, 0);
 		bf_set(pde6_apptagval, pde6, 0);
@@ -2213,6 +2297,10 @@ lpfc_bg_setup_bpl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 		pgdone = 0;
 		subtotal = 0; /* total bytes processed for current prot grp */
 		while (!pgdone) {
+			/* Check to see if we ran out of space */
+			if (num_bde >= phba->cfg_total_seg_cnt)
+				return num_bde + 1;
+
 			if (!sgde) {
 				lpfc_printf_log(phba, KERN_ERR, LOG_BG,
 					"9065 BLKGRD:%s Invalid data segment\n",
@@ -2324,7 +2412,6 @@ lpfc_bg_setup_sgl(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 	struct sli4_sge_diseed *diseed = NULL;
 	dma_addr_t physaddr;
 	int i = 0, num_sge = 0, status;
-	int datadir = sc->sc_data_direction;
 	uint32_t reftag;
 	unsigned blksize;
 	uint8_t txop, rxop;
@@ -2362,13 +2449,26 @@ lpfc_bg_setup_sgl(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 	diseed->ref_tag = cpu_to_le32(reftag);
 	diseed->ref_tag_tran = diseed->ref_tag;
 
+	/*
+	 * We only need to check the data on READs, for WRITEs
+	 * protection data is automatically generated, not checked.
+	 */
+	if (sc->sc_data_direction == DMA_FROM_DEVICE) {
+		if (scsi_prot_flagged(sc, SCSI_PROT_GUARD_CHECK))
+			bf_set(lpfc_sli4_sge_dif_ce, diseed, checking);
+		else
+			bf_set(lpfc_sli4_sge_dif_ce, diseed, 0);
+
+		if (scsi_prot_flagged(sc, SCSI_PROT_REF_CHECK))
+			bf_set(lpfc_sli4_sge_dif_re, diseed, checking);
+		else
+			bf_set(lpfc_sli4_sge_dif_re, diseed, 0);
+	}
+
 	/* setup DISEED with the rest of the info */
 	bf_set(lpfc_sli4_sge_dif_optx, diseed, txop);
 	bf_set(lpfc_sli4_sge_dif_oprx, diseed, rxop);
-	if (datadir == DMA_FROM_DEVICE) {
-		bf_set(lpfc_sli4_sge_dif_ce, diseed, checking);
-		bf_set(lpfc_sli4_sge_dif_re, diseed, checking);
-	}
+
 	bf_set(lpfc_sli4_sge_dif_ai, diseed, 1);
 	bf_set(lpfc_sli4_sge_dif_me, diseed, 0);
 
@@ -2497,6 +2597,10 @@ lpfc_bg_setup_sgl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 
 	split_offset = 0;
 	do {
+		/* Check to see if we ran out of space */
+		if (num_sge >= (phba->cfg_total_seg_cnt - 2))
+			return num_sge + 3;
+
 		/* setup DISEED with what we have */
 		diseed = (struct sli4_sge_diseed *) sgl;
 		memset(diseed, 0, sizeof(struct sli4_sge_diseed));
@@ -2506,11 +2610,34 @@ lpfc_bg_setup_sgl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 		diseed->ref_tag = cpu_to_le32(reftag);
 		diseed->ref_tag_tran = diseed->ref_tag;
 
+		if (scsi_prot_flagged(sc, SCSI_PROT_GUARD_CHECK)) {
+			bf_set(lpfc_sli4_sge_dif_ce, diseed, checking);
+
+		} else {
+			bf_set(lpfc_sli4_sge_dif_ce, diseed, 0);
+			/*
+			 * When in this mode, the hardware will replace
+			 * the guard tag from the host with a
+			 * newly generated good CRC for the wire.
+			 * Switch to raw mode here to avoid this
+			 * behavior. What the host sends gets put on the wire.
+			 */
+			if (txop == BG_OP_IN_CRC_OUT_CRC) {
+				txop = BG_OP_RAW_MODE;
+				rxop = BG_OP_RAW_MODE;
+			}
+		}
+
+
+		if (scsi_prot_flagged(sc, SCSI_PROT_REF_CHECK))
+			bf_set(lpfc_sli4_sge_dif_re, diseed, checking);
+		else
+			bf_set(lpfc_sli4_sge_dif_re, diseed, 0);
+
 		/* setup DISEED with the rest of the info */
 		bf_set(lpfc_sli4_sge_dif_optx, diseed, txop);
 		bf_set(lpfc_sli4_sge_dif_oprx, diseed, rxop);
-		bf_set(lpfc_sli4_sge_dif_ce, diseed, checking);
-		bf_set(lpfc_sli4_sge_dif_re, diseed, checking);
+
 		bf_set(lpfc_sli4_sge_dif_ai, diseed, 1);
 		bf_set(lpfc_sli4_sge_dif_me, diseed, 0);
 
@@ -2556,6 +2683,10 @@ lpfc_bg_setup_sgl_prot(struct lpfc_hba *phba, struct scsi_cmnd *sc,
 		pgdone = 0;
 		subtotal = 0; /* total bytes processed for current prot grp */
 		while (!pgdone) {
+			/* Check to see if we ran out of space */
+			if (num_sge >= phba->cfg_total_seg_cnt)
+				return num_sge + 1;
+
 			if (!sgde) {
 				lpfc_printf_log(phba, KERN_ERR, LOG_BG,
 					"9086 BLKGRD:%s Invalid data segment\n",
@@ -2670,6 +2801,47 @@ lpfc_prot_group_type(struct lpfc_hba *phba, struct scsi_cmnd *sc)
 }
 
 /**
+ * lpfc_bg_scsi_adjust_dl - Adjust SCSI data length for BlockGuard
+ * @phba: The Hba for which this call is being executed.
+ * @lpfc_cmd: The scsi buffer which is going to be adjusted.
+ *
+ * Adjust the data length to account for how much data
+ * is actually on the wire.
+ *
+ * returns the adjusted data length
+ **/
+static int
+lpfc_bg_scsi_adjust_dl(struct lpfc_hba *phba,
+		       struct lpfc_scsi_buf *lpfc_cmd)
+{
+	struct scsi_cmnd *sc = lpfc_cmd->pCmd;
+	int fcpdl;
+
+	fcpdl = scsi_bufflen(sc);
+
+	/* Check if there is protection data on the wire */
+	if (sc->sc_data_direction == DMA_FROM_DEVICE) {
+		/* Read */
+		if (scsi_get_prot_op(sc) ==  SCSI_PROT_READ_INSERT)
+			return fcpdl;
+
+	} else {
+		/* Write */
+		if (scsi_get_prot_op(sc) ==  SCSI_PROT_WRITE_STRIP)
+			return fcpdl;
+	}
+
+	/*
+	 * If we are in DIF Type 1 mode every data block has a 8 byte
+	 * DIF (trailer) attached to it. Must ajust FCP data length.
+	 */
+	if (scsi_prot_flagged(sc, SCSI_PROT_TRANSFER_PI))
+		fcpdl += (fcpdl / lpfc_cmd_blksize(sc)) * 8;
+
+	return fcpdl;
+}
+
+/**
  * lpfc_bg_scsi_prep_dma_buf_s3 - DMA mapping for scsi buffer to SLI3 IF spec
  * @phba: The Hba for which this call is being executed.
  * @lpfc_cmd: The scsi buffer which is going to be prep'ed.
@@ -2689,8 +2861,7 @@ lpfc_bg_scsi_prep_dma_buf_s3(struct lpfc_hba *phba,
 	uint32_t num_bde = 0;
 	int datasegcnt, protsegcnt, datadir = scsi_cmnd->sc_data_direction;
 	int prot_group_type = 0;
-	int diflen, fcpdl;
-	unsigned blksize;
+	int fcpdl;
 
 	/*
 	 * Start the lpfc command prep by bumping the bpl beyond fcp_cmnd
@@ -2711,28 +2882,28 @@ lpfc_bg_scsi_prep_dma_buf_s3(struct lpfc_hba *phba,
 			return 1;
 
 		lpfc_cmd->seg_cnt = datasegcnt;
-		if (lpfc_cmd->seg_cnt > phba->cfg_sg_seg_cnt) {
-			lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-					"9067 BLKGRD: %s: Too many sg segments"
-					" from dma_map_sg.  Config %d, seg_cnt"
-					" %d\n",
-					__func__, phba->cfg_sg_seg_cnt,
-					lpfc_cmd->seg_cnt);
-			scsi_dma_unmap(scsi_cmnd);
-			return 1;
-		}
+
+		/* First check if data segment count from SCSI Layer is good */
+		if (lpfc_cmd->seg_cnt > phba->cfg_sg_seg_cnt)
+			goto err;
 
 		prot_group_type = lpfc_prot_group_type(phba, scsi_cmnd);
 
 		switch (prot_group_type) {
 		case LPFC_PG_TYPE_NO_DIF:
+
+			/* Here we need to add a PDE5 and PDE6 to the count */
+			if ((lpfc_cmd->seg_cnt + 2) > phba->cfg_total_seg_cnt)
+				goto err;
+
 			num_bde = lpfc_bg_setup_bpl(phba, scsi_cmnd, bpl,
 					datasegcnt);
 			/* we should have 2 or more entries in buffer list */
 			if (num_bde < 2)
 				goto err;
 			break;
-		case LPFC_PG_TYPE_DIF_BUF:{
+
+		case LPFC_PG_TYPE_DIF_BUF:
 			/*
 			 * This type indicates that protection buffers are
 			 * passed to the driver, so that needs to be prepared
@@ -2747,31 +2918,28 @@ lpfc_bg_scsi_prep_dma_buf_s3(struct lpfc_hba *phba,
 			}
 
 			lpfc_cmd->prot_seg_cnt = protsegcnt;
-			if (lpfc_cmd->prot_seg_cnt
-			    > phba->cfg_prot_sg_seg_cnt) {
-				lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-					"9068 BLKGRD: %s: Too many prot sg "
-					"segments from dma_map_sg.  Config %d,"
-						"prot_seg_cnt %d\n", __func__,
-						phba->cfg_prot_sg_seg_cnt,
-						lpfc_cmd->prot_seg_cnt);
-				dma_unmap_sg(&phba->pcidev->dev,
-					     scsi_prot_sglist(scsi_cmnd),
-					     scsi_prot_sg_count(scsi_cmnd),
-					     datadir);
-				scsi_dma_unmap(scsi_cmnd);
-				return 1;
-			}
+
+			/*
+			 * There is a minimun of 4 BPLs used for every
+			 * protection data segment.
+			 */
+			if ((lpfc_cmd->prot_seg_cnt * 4) >
+			    (phba->cfg_total_seg_cnt - 2))
+				goto err;
 
 			num_bde = lpfc_bg_setup_bpl_prot(phba, scsi_cmnd, bpl,
 					datasegcnt, protsegcnt);
 			/* we should have 3 or more entries in buffer list */
-			if (num_bde < 3)
+			if ((num_bde < 3) ||
+			    (num_bde > phba->cfg_total_seg_cnt))
 				goto err;
 			break;
-		}
+
 		case LPFC_PG_TYPE_INVALID:
 		default:
+			scsi_dma_unmap(scsi_cmnd);
+			lpfc_cmd->seg_cnt = 0;
+
 			lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
 					"9022 Unexpected protection group %i\n",
 					prot_group_type);
@@ -2790,18 +2958,7 @@ lpfc_bg_scsi_prep_dma_buf_s3(struct lpfc_hba *phba,
 	iocb_cmd->ulpBdeCount = 1;
 	iocb_cmd->ulpLe = 1;
 
-	fcpdl = scsi_bufflen(scsi_cmnd);
-
-	if (scsi_get_prot_type(scsi_cmnd) == SCSI_PROT_DIF_TYPE1) {
-		/*
-		 * We are in DIF Type 1 mode
-		 * Every data block has a 8 byte DIF (trailer)
-		 * attached to it.  Must ajust FCP data length
-		 */
-		blksize = lpfc_cmd_blksize(scsi_cmnd);
-		diflen = (fcpdl / blksize) * 8;
-		fcpdl += diflen;
-	}
+	fcpdl = lpfc_bg_scsi_adjust_dl(phba, lpfc_cmd);
 	fcp_cmnd->fcpDl = be32_to_cpu(fcpdl);
 
 	/*
@@ -2812,14 +2969,234 @@ lpfc_bg_scsi_prep_dma_buf_s3(struct lpfc_hba *phba,
 
 	return 0;
 err:
+	if (lpfc_cmd->seg_cnt)
+		scsi_dma_unmap(scsi_cmnd);
+	if (lpfc_cmd->prot_seg_cnt)
+		dma_unmap_sg(&phba->pcidev->dev, scsi_prot_sglist(scsi_cmnd),
+			     scsi_prot_sg_count(scsi_cmnd),
+			     scsi_cmnd->sc_data_direction);
+
 	lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
-			"9023 Could not setup all needed BDE's"
-			"prot_group_type=%d, num_bde=%d\n",
+			"9023 Cannot setup S/G List for HBA"
+			"IO segs %d/%d BPL %d SCSI %d: %d %d\n",
+			lpfc_cmd->seg_cnt, lpfc_cmd->prot_seg_cnt,
+			phba->cfg_total_seg_cnt, phba->cfg_sg_seg_cnt,
 			prot_group_type, num_bde);
+
+	lpfc_cmd->seg_cnt = 0;
+	lpfc_cmd->prot_seg_cnt = 0;
 	return 1;
 }
 
 /*
+ * This function calcuates the T10 DIF guard tag
+ * on the specified data using a CRC algorithmn
+ * using crc_t10dif.
+ */
+uint16_t
+lpfc_bg_crc(uint8_t *data, int count)
+{
+	uint16_t crc = 0;
+	uint16_t x;
+
+	crc = crc_t10dif(data, count);
+	x = cpu_to_be16(crc);
+	return x;
+}
+
+/*
+ * This function calcuates the T10 DIF guard tag
+ * on the specified data using a CSUM algorithmn
+ * using ip_compute_csum.
+ */
+uint16_t
+lpfc_bg_csum(uint8_t *data, int count)
+{
+	uint16_t ret;
+
+	ret = ip_compute_csum(data, count);
+	return ret;
+}
+
+/*
+ * This function examines the protection data to try to determine
+ * what type of T10-DIF error occurred.
+ */
+void
+lpfc_calc_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
+{
+	struct scatterlist *sgpe; /* s/g prot entry */
+	struct scatterlist *sgde; /* s/g data entry */
+	struct scsi_cmnd *cmd = lpfc_cmd->pCmd;
+	struct scsi_dif_tuple *src = NULL;
+	uint8_t *data_src = NULL;
+	uint16_t guard_tag, guard_type;
+	uint16_t start_app_tag, app_tag;
+	uint32_t start_ref_tag, ref_tag;
+	int prot, protsegcnt;
+	int err_type, len, data_len;
+	int chk_ref, chk_app, chk_guard;
+	uint16_t sum;
+	unsigned blksize;
+
+	err_type = BGS_GUARD_ERR_MASK;
+	sum = 0;
+	guard_tag = 0;
+
+	/* First check to see if there is protection data to examine */
+	prot = scsi_get_prot_op(cmd);
+	if ((prot == SCSI_PROT_READ_STRIP) ||
+	    (prot == SCSI_PROT_WRITE_INSERT) ||
+	    (prot == SCSI_PROT_NORMAL))
+		goto out;
+
+	/* Currently the driver just supports ref_tag and guard_tag checking */
+	chk_ref = 1;
+	chk_app = 0;
+	chk_guard = 0;
+
+	/* Setup a ptr to the protection data provided by the SCSI host */
+	sgpe = scsi_prot_sglist(cmd);
+	protsegcnt = lpfc_cmd->prot_seg_cnt;
+
+	if (sgpe && protsegcnt) {
+
+		/*
+		 * We will only try to verify guard tag if the segment
+		 * data length is a multiple of the blksize.
+		 */
+		sgde = scsi_sglist(cmd);
+		blksize = lpfc_cmd_blksize(cmd);
+		data_src = (uint8_t *)sg_virt(sgde);
+		data_len = sgde->length;
+		if ((data_len & (blksize - 1)) == 0)
+			chk_guard = 1;
+		guard_type = scsi_host_get_guard(cmd->device->host);
+
+		start_ref_tag = (uint32_t)scsi_get_lba(cmd); /* Truncate LBA */
+		start_app_tag = src->app_tag;
+		src = (struct scsi_dif_tuple *)sg_virt(sgpe);
+		len = sgpe->length;
+		while (src && protsegcnt) {
+			while (len) {
+
+				/*
+				 * First check to see if a protection data
+				 * check is valid
+				 */
+				if ((src->ref_tag == 0xffffffff) ||
+				    (src->app_tag == 0xffff)) {
+					start_ref_tag++;
+					goto skipit;
+				}
+
+				/* App Tag checking */
+				app_tag = src->app_tag;
+				if (chk_app && (app_tag != start_app_tag)) {
+					err_type = BGS_APPTAG_ERR_MASK;
+					goto out;
+				}
+
+				/* Reference Tag checking */
+				ref_tag = be32_to_cpu(src->ref_tag);
+				if (chk_ref && (ref_tag != start_ref_tag)) {
+					err_type = BGS_REFTAG_ERR_MASK;
+					goto out;
+				}
+				start_ref_tag++;
+
+				/* Guard Tag checking */
+				if (chk_guard) {
+					guard_tag = src->guard_tag;
+					if (guard_type == SHOST_DIX_GUARD_IP)
+						sum = lpfc_bg_csum(data_src,
+								   blksize);
+					else
+						sum = lpfc_bg_crc(data_src,
+								  blksize);
+					if ((guard_tag != sum)) {
+						err_type = BGS_GUARD_ERR_MASK;
+						goto out;
+					}
+				}
+skipit:
+				len -= sizeof(struct scsi_dif_tuple);
+				if (len < 0)
+					len = 0;
+				src++;
+
+				data_src += blksize;
+				data_len -= blksize;
+
+				/*
+				 * Are we at the end of the Data segment?
+				 * The data segment is only used for Guard
+				 * tag checking.
+				 */
+				if (chk_guard && (data_len == 0)) {
+					chk_guard = 0;
+					sgde = sg_next(sgde);
+					if (!sgde)
+						goto out;
+
+					data_src = (uint8_t *)sg_virt(sgde);
+					data_len = sgde->length;
+					if ((data_len & (blksize - 1)) == 0)
+						chk_guard = 1;
+				}
+			}
+
+			/* Goto the next Protection data segment */
+			sgpe = sg_next(sgpe);
+			if (sgpe) {
+				src = (struct scsi_dif_tuple *)sg_virt(sgpe);
+				len = sgpe->length;
+			} else {
+				src = NULL;
+			}
+			protsegcnt--;
+		}
+	}
+out:
+	if (err_type == BGS_GUARD_ERR_MASK) {
+		scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
+					0x10, 0x1);
+		cmd->result = DRIVER_SENSE << 24
+			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
+		phba->bg_guard_err_cnt++;
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9069 BLKGRD: LBA %lx grd_tag error %x != %x\n",
+				(unsigned long)scsi_get_lba(cmd),
+				sum, guard_tag);
+
+	} else if (err_type == BGS_REFTAG_ERR_MASK) {
+		scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
+					0x10, 0x3);
+		cmd->result = DRIVER_SENSE << 24
+			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
+
+		phba->bg_reftag_err_cnt++;
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9066 BLKGRD: LBA %lx ref_tag error %x != %x\n",
+				(unsigned long)scsi_get_lba(cmd),
+				ref_tag, start_ref_tag);
+
+	} else if (err_type == BGS_APPTAG_ERR_MASK) {
+		scsi_build_sense_buffer(1, cmd->sense_buffer, ILLEGAL_REQUEST,
+					0x10, 0x2);
+		cmd->result = DRIVER_SENSE << 24
+			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
+
+		phba->bg_apptag_err_cnt++;
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9041 BLKGRD: LBA %lx app_tag error %x != %x\n",
+				(unsigned long)scsi_get_lba(cmd),
+				app_tag, start_app_tag);
+	}
+}
+
+
+/*
  * This function checks for BlockGuard errors detected by
  * the HBA.  In case of errors, the ASC/ASCQ fields in the
  * sense buffer will be set accordingly, paired with
@@ -2842,12 +3219,6 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 	uint32_t bgstat = bgf->bgstat;
 	uint64_t failing_sector = 0;
 
-	lpfc_printf_log(phba, KERN_ERR, LOG_BG, "9069 BLKGRD: BG ERROR in cmd"
-			" 0x%x lba 0x%llx blk cnt 0x%x "
-			"bgstat=0x%x bghm=0x%x\n",
-			cmd->cmnd[0], (unsigned long long)scsi_get_lba(cmd),
-			blk_rq_sectors(cmd->request), bgstat, bghm);
-
 	spin_lock(&_dump_buf_lock);
 	if (!_dump_buf_done) {
 		lpfc_printf_log(phba, KERN_ERR, LOG_BG,  "9070 BLKGRD: Saving"
@@ -2870,18 +3241,24 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 
 	if (lpfc_bgs_get_invalid_prof(bgstat)) {
 		cmd->result = ScsiResult(DID_ERROR, 0);
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG, "9072 BLKGRD: Invalid"
-			" BlockGuard profile. bgstat:0x%x\n",
-			bgstat);
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9072 BLKGRD: Invalid BG Profile in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
 		ret = (-1);
 		goto out;
 	}
 
 	if (lpfc_bgs_get_uninit_dif_block(bgstat)) {
 		cmd->result = ScsiResult(DID_ERROR, 0);
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG, "9073 BLKGRD: "
-				"Invalid BlockGuard DIF Block. bgstat:0x%x\n",
-				bgstat);
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9073 BLKGRD: Invalid BG PDIF Block in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
 		ret = (-1);
 		goto out;
 	}
@@ -2894,8 +3271,12 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 		cmd->result = DRIVER_SENSE << 24
 			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
 		phba->bg_guard_err_cnt++;
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-			"9055 BLKGRD: guard_tag error\n");
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9055 BLKGRD: Guard Tag error in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
 	}
 
 	if (lpfc_bgs_get_reftag_err(bgstat)) {
@@ -2907,8 +3288,12 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
 
 		phba->bg_reftag_err_cnt++;
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-			"9056 BLKGRD: ref_tag error\n");
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9056 BLKGRD: Ref Tag error in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
 	}
 
 	if (lpfc_bgs_get_apptag_err(bgstat)) {
@@ -2920,8 +3305,12 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 			| ScsiResult(DID_ABORT, SAM_STAT_CHECK_CONDITION);
 
 		phba->bg_apptag_err_cnt++;
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-			"9061 BLKGRD: app_tag error\n");
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9061 BLKGRD: App Tag error in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
 	}
 
 	if (lpfc_bgs_get_hi_water_mark_present(bgstat)) {
@@ -2960,11 +3349,16 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
 
 	if (!ret) {
 		/* No error was reported - problem in FW? */
-		cmd->result = ScsiResult(DID_ERROR, 0);
-		lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-			"9057 BLKGRD: Unknown error reported!\n");
+		lpfc_printf_log(phba, KERN_WARNING, LOG_FCP | LOG_BG,
+				"9057 BLKGRD: Unknown error in cmd"
+				" 0x%x lba 0x%llx blk cnt 0x%x "
+				"bgstat=x%x bghm=x%x\n", cmd->cmnd[0],
+				(unsigned long long)scsi_get_lba(cmd),
+				blk_rq_sectors(cmd->request), bgstat, bghm);
+
+		/* Calcuate what type of error it was */
+		lpfc_calc_bg_err(phba, lpfc_cmd);
 	}
-
 out:
 	return ret;
 }
@@ -3028,6 +3422,7 @@ lpfc_scsi_prep_dma_buf_s4(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
 				"dma_map_sg.  Config %d, seg_cnt %d\n",
 				__func__, phba->cfg_sg_seg_cnt,
 			       lpfc_cmd->seg_cnt);
+			lpfc_cmd->seg_cnt = 0;
 			scsi_dma_unmap(scsi_cmnd);
 			return 1;
 		}
@@ -3094,45 +3489,6 @@ lpfc_scsi_prep_dma_buf_s4(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd)
 }
 
 /**
- * lpfc_bg_scsi_adjust_dl - Adjust SCSI data length for BlockGuard
- * @phba: The Hba for which this call is being executed.
- * @lpfc_cmd: The scsi buffer which is going to be adjusted.
- *
- * Adjust the data length to account for how much data
- * is actually on the wire.
- *
- * returns the adjusted data length
- **/
-static int
-lpfc_bg_scsi_adjust_dl(struct lpfc_hba *phba,
-		struct lpfc_scsi_buf *lpfc_cmd)
-{
-	struct scsi_cmnd *sc = lpfc_cmd->pCmd;
-	int diflen, fcpdl;
-	unsigned blksize;
-
-	fcpdl = scsi_bufflen(sc);
-
-	/* Check if there is protection data on the wire */
-	if (sc->sc_data_direction == DMA_FROM_DEVICE) {
-		/* Read */
-		if (scsi_get_prot_op(sc) ==  SCSI_PROT_READ_INSERT)
-			return fcpdl;
-
-	} else {
-		/* Write */
-		if (scsi_get_prot_op(sc) ==  SCSI_PROT_WRITE_STRIP)
-			return fcpdl;
-	}
-
-	/* If protection data on the wire, adjust the count accordingly */
-	blksize = lpfc_cmd_blksize(sc);
-	diflen = (fcpdl / blksize) * 8;
-	fcpdl += diflen;
-	return fcpdl;
-}
-
-/**
  * lpfc_bg_scsi_prep_dma_buf_s4 - DMA mapping for scsi buffer to SLI4 IF spec
  * @phba: The Hba for which this call is being executed.
  * @lpfc_cmd: The scsi buffer which is going to be mapped.
@@ -3149,14 +3505,14 @@ lpfc_bg_scsi_prep_dma_buf_s4(struct lpfc_hba *phba,
 	struct fcp_cmnd *fcp_cmnd = lpfc_cmd->fcp_cmnd;
 	struct sli4_sge *sgl = (struct sli4_sge *)(lpfc_cmd->fcp_bpl);
 	IOCB_t *iocb_cmd = &lpfc_cmd->cur_iocbq.iocb;
-	uint32_t num_bde = 0;
+	uint32_t num_sge = 0;
 	int datasegcnt, protsegcnt, datadir = scsi_cmnd->sc_data_direction;
 	int prot_group_type = 0;
 	int fcpdl;
 
 	/*
 	 * Start the lpfc command prep by bumping the sgl beyond fcp_cmnd
-	 *  fcp_rsp regions to the first data bde entry
+	 *  fcp_rsp regions to the first data sge entry
 	 */
 	if (scsi_sg_count(scsi_cmnd)) {
 		/*
@@ -3179,28 +3535,28 @@ lpfc_bg_scsi_prep_dma_buf_s4(struct lpfc_hba *phba,
 
 		sgl += 1;
 		lpfc_cmd->seg_cnt = datasegcnt;
-		if (lpfc_cmd->seg_cnt > phba->cfg_sg_seg_cnt) {
-			lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-					"9087 BLKGRD: %s: Too many sg segments"
-					" from dma_map_sg.  Config %d, seg_cnt"
-					" %d\n",
-					__func__, phba->cfg_sg_seg_cnt,
-					lpfc_cmd->seg_cnt);
-			scsi_dma_unmap(scsi_cmnd);
-			return 1;
-		}
+
+		/* First check if data segment count from SCSI Layer is good */
+		if (lpfc_cmd->seg_cnt > phba->cfg_sg_seg_cnt)
+			goto err;
 
 		prot_group_type = lpfc_prot_group_type(phba, scsi_cmnd);
 
 		switch (prot_group_type) {
 		case LPFC_PG_TYPE_NO_DIF:
-			num_bde = lpfc_bg_setup_sgl(phba, scsi_cmnd, sgl,
+			/* Here we need to add a DISEED to the count */
+			if ((lpfc_cmd->seg_cnt + 1) > phba->cfg_total_seg_cnt)
+				goto err;
+
+			num_sge = lpfc_bg_setup_sgl(phba, scsi_cmnd, sgl,
 					datasegcnt);
+
 			/* we should have 2 or more entries in buffer list */
-			if (num_bde < 2)
+			if (num_sge < 2)
 				goto err;
 			break;
-		case LPFC_PG_TYPE_DIF_BUF:{
+
+		case LPFC_PG_TYPE_DIF_BUF:
 			/*
 			 * This type indicates that protection buffers are
 			 * passed to the driver, so that needs to be prepared
@@ -3215,31 +3571,28 @@ lpfc_bg_scsi_prep_dma_buf_s4(struct lpfc_hba *phba,
 			}
 
 			lpfc_cmd->prot_seg_cnt = protsegcnt;
-			if (lpfc_cmd->prot_seg_cnt
-			    > phba->cfg_prot_sg_seg_cnt) {
-				lpfc_printf_log(phba, KERN_ERR, LOG_BG,
-					"9088 BLKGRD: %s: Too many prot sg "
-					"segments from dma_map_sg.  Config %d,"
-						"prot_seg_cnt %d\n", __func__,
-						phba->cfg_prot_sg_seg_cnt,
-						lpfc_cmd->prot_seg_cnt);
-				dma_unmap_sg(&phba->pcidev->dev,
-					     scsi_prot_sglist(scsi_cmnd),
-					     scsi_prot_sg_count(scsi_cmnd),
-					     datadir);
-				scsi_dma_unmap(scsi_cmnd);
-				return 1;
-			}
+			/*
+			 * There is a minimun of 3 SGEs used for every
+			 * protection data segment.
+			 */
+			if ((lpfc_cmd->prot_seg_cnt * 3) >
+			    (phba->cfg_total_seg_cnt - 2))
+				goto err;
 
-			num_bde = lpfc_bg_setup_sgl_prot(phba, scsi_cmnd, sgl,
+			num_sge = lpfc_bg_setup_sgl_prot(phba, scsi_cmnd, sgl,
 					datasegcnt, protsegcnt);
+
 			/* we should have 3 or more entries in buffer list */
-			if (num_bde < 3)
+			if ((num_sge < 3) ||
+			    (num_sge > phba->cfg_total_seg_cnt))
 				goto err;
 			break;
-		}
+
 		case LPFC_PG_TYPE_INVALID:
 		default:
+			scsi_dma_unmap(scsi_cmnd);
+			lpfc_cmd->seg_cnt = 0;
+
 			lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
 					"9083 Unexpected protection group %i\n",
 					prot_group_type);
@@ -3263,7 +3616,6 @@ lpfc_bg_scsi_prep_dma_buf_s4(struct lpfc_hba *phba,
 	}
 
 	fcpdl = lpfc_bg_scsi_adjust_dl(phba, lpfc_cmd);
-
 	fcp_cmnd->fcpDl = be32_to_cpu(fcpdl);
 
 	/*
@@ -3274,10 +3626,22 @@ lpfc_bg_scsi_prep_dma_buf_s4(struct lpfc_hba *phba,
 
 	return 0;
 err:
+	if (lpfc_cmd->seg_cnt)
+		scsi_dma_unmap(scsi_cmnd);
+	if (lpfc_cmd->prot_seg_cnt)
+		dma_unmap_sg(&phba->pcidev->dev, scsi_prot_sglist(scsi_cmnd),
+			     scsi_prot_sg_count(scsi_cmnd),
+			     scsi_cmnd->sc_data_direction);
+
 	lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
-			"9084 Could not setup all needed BDE's"
-			"prot_group_type=%d, num_bde=%d\n",
-			prot_group_type, num_bde);
+			"9084 Cannot setup S/G List for HBA"
+			"IO segs %d/%d SGL %d SCSI %d: %d %d\n",
+			lpfc_cmd->seg_cnt, lpfc_cmd->prot_seg_cnt,
+			phba->cfg_total_seg_cnt, phba->cfg_sg_seg_cnt,
+			prot_group_type, num_sge);
+
+	lpfc_cmd->seg_cnt = 0;
+	lpfc_cmd->prot_seg_cnt = 0;
 	return 1;
 }
 
@@ -4357,7 +4721,8 @@ lpfc_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *cmnd)
 
 	if (scsi_get_prot_op(cmnd) != SCSI_PROT_NORMAL) {
 		if (vport->phba->cfg_enable_bg) {
-			lpfc_printf_vlog(vport, KERN_INFO, LOG_BG,
+			lpfc_printf_vlog(vport,
+					 KERN_INFO, LOG_SCSI_CMD,
 					 "9033 BLKGRD: rcvd %s cmd:x%x "
 					 "sector x%llx cnt %u pt %x\n",
 					 dif_op_str[scsi_get_prot_op(cmnd)],
@@ -4369,7 +4734,8 @@ lpfc_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *cmnd)
 		err = lpfc_bg_scsi_prep_dma_buf(phba, lpfc_cmd);
 	} else {
 		if (vport->phba->cfg_enable_bg) {
-			lpfc_printf_vlog(vport, KERN_INFO, LOG_BG,
+			lpfc_printf_vlog(vport,
+					 KERN_INFO, LOG_SCSI_CMD,
 					 "9038 BLKGRD: rcvd PROT_NORMAL cmd: "
 					 "x%x sector x%llx cnt %u pt %x\n",
 					 cmnd->cmnd[0],
@@ -4542,7 +4908,7 @@ lpfc_abort_handler(struct scsi_cmnd *cmnd)
 	/* Wait for abort to complete */
 	wait_event_timeout(waitq,
 			  (lpfc_cmd->pCmd != cmnd),
-			   (2*vport->cfg_devloss_tmo*HZ));
+			   msecs_to_jiffies(2*vport->cfg_devloss_tmo*1000));
 	lpfc_cmd->waitq = NULL;
 
 	if (lpfc_cmd->pCmd == cmnd) {
@@ -5012,16 +5378,24 @@ lpfc_host_reset_handler(struct scsi_cmnd *cmnd)
 	struct lpfc_hba *phba = vport->phba;
 	int rc, ret = SUCCESS;
 
+	lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
+			 "3172 SCSI layer issued Host Reset Data:\n");
+
 	lpfc_offline_prep(phba, LPFC_MBX_WAIT);
 	lpfc_offline(phba);
 	rc = lpfc_sli_brdrestart(phba);
 	if (rc)
 		ret = FAILED;
-	lpfc_online(phba);
+	rc = lpfc_online(phba);
+	if (rc)
+		ret = FAILED;
 	lpfc_unblock_mgmt_io(phba);
 
-	lpfc_printf_log(phba, KERN_ERR, LOG_FCP,
-			"3172 SCSI layer issued Host Reset Data: x%x\n", ret);
+	if (ret == FAILED) {
+		lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
+				 "3323 Failed host reset, bring it offline\n");
+		lpfc_sli4_offline_eratt(phba);
+	}
 	return ret;
 }
 
@@ -5088,11 +5462,11 @@ lpfc_slave_alloc(struct scsi_device *sdev)
 	}
 	num_allocated = lpfc_new_scsi_buf(vport, num_to_alloc);
 	if (num_to_alloc != num_allocated) {
-			lpfc_printf_vlog(vport, KERN_WARNING, LOG_FCP,
-				 "0708 Allocation request of %d "
-				 "command buffers did not succeed.  "
-				 "Allocated %d buffers.\n",
-				 num_to_alloc, num_allocated);
+			lpfc_printf_vlog(vport, KERN_ERR, LOG_FCP,
+					 "0708 Allocation request of %d "
+					 "command buffers did not succeed.  "
+					 "Allocated %d buffers.\n",
+					 num_to_alloc, num_allocated);
 	}
 	if (num_allocated > 0)
 		phba->total_scsi_bufs += num_allocated;
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 35dd17eb0f27..572579f87de4 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -667,7 +667,7 @@ lpfc_handle_rrq_active(struct lpfc_hba *phba)
 
 	spin_lock_irqsave(&phba->hbalock, iflags);
 	phba->hba_flag &= ~HBA_RRQ_ACTIVE;
-	next_time = jiffies + HZ * (phba->fc_ratov + 1);
+	next_time = jiffies + msecs_to_jiffies(1000 * (phba->fc_ratov + 1));
 	list_for_each_entry_safe(rrq, nextrrq,
 				 &phba->active_rrq_list, list) {
 		if (time_after(jiffies, rrq->rrq_stop_time))
@@ -782,7 +782,7 @@ lpfc_cleanup_wt_rrqs(struct lpfc_hba *phba)
 		return;
 	spin_lock_irqsave(&phba->hbalock, iflags);
 	phba->hba_flag &= ~HBA_RRQ_ACTIVE;
-	next_time = jiffies + HZ * (phba->fc_ratov * 2);
+	next_time = jiffies + msecs_to_jiffies(1000 * (phba->fc_ratov * 2));
 	list_splice_init(&phba->active_rrq_list, &rrq_list);
 	spin_unlock_irqrestore(&phba->hbalock, iflags);
 
@@ -878,7 +878,8 @@ lpfc_set_rrq_active(struct lpfc_hba *phba, struct lpfc_nodelist *ndlp,
 	else
 		rrq->send_rrq = 0;
 	rrq->xritag = xritag;
-	rrq->rrq_stop_time = jiffies + HZ * (phba->fc_ratov + 1);
+	rrq->rrq_stop_time = jiffies +
+				msecs_to_jiffies(1000 * (phba->fc_ratov + 1));
 	rrq->ndlp = ndlp;
 	rrq->nlp_DID = ndlp->nlp_DID;
 	rrq->vport = ndlp->vport;
@@ -926,8 +927,7 @@ __lpfc_sli_get_sglq(struct lpfc_hba *phba, struct lpfc_iocbq *piocbq)
 	} else  if ((piocbq->iocb.ulpCommand == CMD_GEN_REQUEST64_CR) &&
 			!(piocbq->iocb_flag & LPFC_IO_LIBDFC))
 		ndlp = piocbq->context_un.ndlp;
-	else  if ((piocbq->iocb.ulpCommand == CMD_ELS_REQUEST64_CR) &&
-			(piocbq->iocb_flag & LPFC_IO_LIBDFC))
+	else  if (piocbq->iocb_flag & LPFC_IO_LIBDFC)
 		ndlp = piocbq->context_un.ndlp;
 	else
 		ndlp = piocbq->context1;
@@ -1339,7 +1339,8 @@ lpfc_sli_ringtxcmpl_put(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 			BUG();
 		else
 			mod_timer(&piocb->vport->els_tmofunc,
-				  jiffies + HZ * (phba->fc_ratov << 1));
+				jiffies +
+				msecs_to_jiffies(1000 * (phba->fc_ratov << 1)));
 	}
 
 
@@ -2340,7 +2341,8 @@ lpfc_sli_handle_mb_event(struct lpfc_hba *phba)
 		/* Mailbox cmd <cmd> Cmpl <cmpl> */
 		lpfc_printf_log(phba, KERN_INFO, LOG_MBOX | LOG_SLI,
 				"(%d):0307 Mailbox cmd x%x (x%x/x%x) Cmpl x%p "
-				"Data: x%x x%x x%x x%x x%x x%x x%x x%x x%x\n",
+				"Data: x%x x%x x%x x%x x%x x%x x%x x%x x%x "
+				"x%x x%x x%x\n",
 				pmb->vport ? pmb->vport->vpi : 0,
 				pmbox->mbxCommand,
 				lpfc_sli_config_mbox_subsys_get(phba, pmb),
@@ -2354,7 +2356,10 @@ lpfc_sli_handle_mb_event(struct lpfc_hba *phba)
 				pmbox->un.varWords[4],
 				pmbox->un.varWords[5],
 				pmbox->un.varWords[6],
-				pmbox->un.varWords[7]);
+				pmbox->un.varWords[7],
+				pmbox->un.varWords[8],
+				pmbox->un.varWords[9],
+				pmbox->un.varWords[10]);
 
 		if (pmb->mbox_cmpl)
 			pmb->mbox_cmpl(phba,pmb);
@@ -2908,8 +2913,9 @@ void lpfc_poll_eratt(unsigned long ptr)
 		lpfc_worker_wake_up(phba);
 	else
 		/* Restart the timer for next eratt poll */
-		mod_timer(&phba->eratt_poll, jiffies +
-					HZ * LPFC_ERATT_POLL_INTERVAL);
+		mod_timer(&phba->eratt_poll,
+			  jiffies +
+			  msecs_to_jiffies(1000 * LPFC_ERATT_POLL_INTERVAL));
 	return;
 }
 
@@ -5511,6 +5517,7 @@ lpfc_sli4_dealloc_extent(struct lpfc_hba *phba, uint16_t type)
 			list_del_init(&rsrc_blk->list);
 			kfree(rsrc_blk);
 		}
+		phba->sli4_hba.max_cfg_param.vpi_used = 0;
 		break;
 	case LPFC_RSC_TYPE_FCOE_XRI:
 		kfree(phba->sli4_hba.xri_bmask);
@@ -5811,6 +5818,7 @@ lpfc_sli4_dealloc_resource_identifiers(struct lpfc_hba *phba)
 		lpfc_sli4_dealloc_extent(phba, LPFC_RSC_TYPE_FCOE_VFI);
 	} else {
 		kfree(phba->vpi_bmask);
+		phba->sli4_hba.max_cfg_param.vpi_used = 0;
 		kfree(phba->vpi_ids);
 		bf_set(lpfc_vpi_rsrc_rdy, &phba->sli4_hba.sli4_flags, 0);
 		kfree(phba->sli4_hba.xri_bmask);
@@ -5992,7 +6000,7 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
 	struct lpfc_sglq *sglq_entry = NULL;
 	struct lpfc_sglq *sglq_entry_next = NULL;
 	struct lpfc_sglq *sglq_entry_first = NULL;
-	int status, post_cnt = 0, num_posted = 0, block_cnt = 0;
+	int status, total_cnt, post_cnt = 0, num_posted = 0, block_cnt = 0;
 	int last_xritag = NO_XRI;
 	LIST_HEAD(prep_sgl_list);
 	LIST_HEAD(blck_sgl_list);
@@ -6004,6 +6012,7 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
 	list_splice_init(&phba->sli4_hba.lpfc_sgl_list, &allc_sgl_list);
 	spin_unlock_irq(&phba->hbalock);
 
+	total_cnt = phba->sli4_hba.els_xri_cnt;
 	list_for_each_entry_safe(sglq_entry, sglq_entry_next,
 				 &allc_sgl_list, list) {
 		list_del_init(&sglq_entry->list);
@@ -6055,9 +6064,7 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
 						sglq_entry->sli4_xritag);
 					list_add_tail(&sglq_entry->list,
 						      &free_sgl_list);
-					spin_lock_irq(&phba->hbalock);
-					phba->sli4_hba.els_xri_cnt--;
-					spin_unlock_irq(&phba->hbalock);
+					total_cnt--;
 				}
 			}
 		}
@@ -6085,9 +6092,7 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
 					(sglq_entry_first->sli4_xritag +
 					 post_cnt - 1));
 			list_splice_init(&blck_sgl_list, &free_sgl_list);
-			spin_lock_irq(&phba->hbalock);
-			phba->sli4_hba.els_xri_cnt -= post_cnt;
-			spin_unlock_irq(&phba->hbalock);
+			total_cnt -= post_cnt;
 		}
 
 		/* don't reset xirtag due to hole in xri block */
@@ -6097,6 +6102,8 @@ lpfc_sli4_repost_els_sgl_list(struct lpfc_hba *phba)
 		/* reset els sgl post count for next round of posting */
 		post_cnt = 0;
 	}
+	/* update the number of XRIs posted for ELS */
+	phba->sli4_hba.els_xri_cnt = total_cnt;
 
 	/* free the els sgls failed to post */
 	lpfc_free_sgl_list(phba, &free_sgl_list);
@@ -6446,16 +6453,17 @@ lpfc_sli4_hba_setup(struct lpfc_hba *phba)
 
 	/* Start the ELS watchdog timer */
 	mod_timer(&vport->els_tmofunc,
-		  jiffies + HZ * (phba->fc_ratov * 2));
+		  jiffies + msecs_to_jiffies(1000 * (phba->fc_ratov * 2)));
 
 	/* Start heart beat timer */
 	mod_timer(&phba->hb_tmofunc,
-		  jiffies + HZ * LPFC_HB_MBOX_INTERVAL);
+		  jiffies + msecs_to_jiffies(1000 * LPFC_HB_MBOX_INTERVAL));
 	phba->hb_outstanding = 0;
 	phba->last_completion_time = jiffies;
 
 	/* Start error attention (ERATT) polling timer */
-	mod_timer(&phba->eratt_poll, jiffies + HZ * LPFC_ERATT_POLL_INTERVAL);
+	mod_timer(&phba->eratt_poll,
+		  jiffies + msecs_to_jiffies(1000 * LPFC_ERATT_POLL_INTERVAL));
 
 	/* Enable PCIe device Advanced Error Reporting (AER) if configured */
 	if (phba->cfg_aer_support == 1 && !(phba->hba_flag & HBA_AER_ENABLED)) {
@@ -6822,8 +6830,9 @@ lpfc_sli_issue_mbox_s3(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmbox,
 			goto out_not_finished;
 		}
 		/* timeout active mbox command */
-		mod_timer(&psli->mbox_tmo, (jiffies +
-			       (HZ * lpfc_mbox_tmo_val(phba, pmbox))));
+		timeout = msecs_to_jiffies(lpfc_mbox_tmo_val(phba, pmbox) *
+					   1000);
+		mod_timer(&psli->mbox_tmo, jiffies + timeout);
 	}
 
 	/* Mailbox cmd <cmd> issue */
@@ -7496,7 +7505,7 @@ lpfc_sli4_post_async_mbox(struct lpfc_hba *phba)
 
 	/* Start timer for the mbox_tmo and log some mailbox post messages */
 	mod_timer(&psli->mbox_tmo, (jiffies +
-		  (HZ * lpfc_mbox_tmo_val(phba, mboxq))));
+		  msecs_to_jiffies(1000 * lpfc_mbox_tmo_val(phba, mboxq))));
 
 	lpfc_printf_log(phba, KERN_INFO, LOG_MBOX | LOG_SLI,
 			"(%d):0355 Mailbox cmd x%x (x%x/x%x) issue Data: "
@@ -7914,15 +7923,21 @@ lpfc_sli4_bpl2sgl(struct lpfc_hba *phba, struct lpfc_iocbq *piocbq,
 static inline uint32_t
 lpfc_sli4_scmd_to_wqidx_distr(struct lpfc_hba *phba)
 {
-	int i;
-
-	if (phba->cfg_fcp_io_sched == LPFC_FCP_SCHED_BY_CPU)
-		i = smp_processor_id();
-	else
-		i = atomic_add_return(1, &phba->fcp_qidx);
+	struct lpfc_vector_map_info *cpup;
+	int chann, cpu;
 
-	i = (i % phba->cfg_fcp_io_channel);
-	return i;
+	if (phba->cfg_fcp_io_sched == LPFC_FCP_SCHED_BY_CPU) {
+		cpu = smp_processor_id();
+		if (cpu < phba->sli4_hba.num_present_cpu) {
+			cpup = phba->sli4_hba.cpu_map;
+			cpup += cpu;
+			return cpup->channel_id;
+		}
+		chann = cpu;
+	}
+	chann = atomic_add_return(1, &phba->fcp_qidx);
+	chann = (chann % phba->cfg_fcp_io_channel);
+	return chann;
 }
 
 /**
@@ -8444,10 +8459,14 @@ __lpfc_sli_issue_iocb_s4(struct lpfc_hba *phba, uint32_t ring_number,
 
 	if ((piocb->iocb_flag & LPFC_IO_FCP) ||
 		(piocb->iocb_flag & LPFC_USE_FCPWQIDX)) {
+		if (unlikely(!phba->sli4_hba.fcp_wq))
+			return IOCB_ERROR;
 		if (lpfc_sli4_wq_put(phba->sli4_hba.fcp_wq[piocb->fcp_wqidx],
 				     &wqe))
 			return IOCB_ERROR;
 	} else {
+		if (unlikely(!phba->sli4_hba.els_wq))
+			return IOCB_ERROR;
 		if (lpfc_sli4_wq_put(phba->sli4_hba.els_wq, &wqe))
 			return IOCB_ERROR;
 	}
@@ -10003,7 +10022,7 @@ lpfc_sli_issue_iocb_wait(struct lpfc_hba *phba,
 	retval = lpfc_sli_issue_iocb(phba, ring_number, piocb,
 				     SLI_IOCB_RET_IOCB);
 	if (retval == IOCB_SUCCESS) {
-		timeout_req = timeout * HZ;
+		timeout_req = msecs_to_jiffies(timeout * 1000);
 		timeleft = wait_event_timeout(done_q,
 				lpfc_chk_iocb_flg(phba, piocb, LPFC_IO_WAKE),
 				timeout_req);
@@ -10108,7 +10127,7 @@ lpfc_sli_issue_mbox_wait(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmboxq,
 	if (retval == MBX_BUSY || retval == MBX_SUCCESS) {
 		wait_event_interruptible_timeout(done_q,
 				pmboxq->mbox_flag & LPFC_MBX_WAKE,
-				timeout * HZ);
+				msecs_to_jiffies(timeout * 1000));
 
 		spin_lock_irqsave(&phba->hbalock, flag);
 		pmboxq->context1 = NULL;
@@ -12899,8 +12918,9 @@ lpfc_wq_create(struct lpfc_hba *phba, struct lpfc_queue *wq,
 		}
 		wq->db_regaddr = bar_memmap_p + db_offset;
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
-				"3264 WQ[%d]: barset:x%x, offset:x%x\n",
-				wq->queue_id, pci_barset, db_offset);
+				"3264 WQ[%d]: barset:x%x, offset:x%x, "
+				"format:x%x\n", wq->queue_id, pci_barset,
+				db_offset, wq->db_format);
 	} else {
 		wq->db_format = LPFC_DB_LIST_FORMAT;
 		wq->db_regaddr = phba->sli4_hba.WQDBregaddr;
@@ -13120,8 +13140,9 @@ lpfc_rq_create(struct lpfc_hba *phba, struct lpfc_queue *hrq,
 		}
 		hrq->db_regaddr = bar_memmap_p + db_offset;
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
-				"3266 RQ[qid:%d]: barset:x%x, offset:x%x\n",
-				hrq->queue_id, pci_barset, db_offset);
+				"3266 RQ[qid:%d]: barset:x%x, offset:x%x, "
+				"format:x%x\n", hrq->queue_id, pci_barset,
+				db_offset, hrq->db_format);
 	} else {
 		hrq->db_format = LPFC_DB_RING_FORMAT;
 		hrq->db_regaddr = phba->sli4_hba.RQDBregaddr;
@@ -13971,13 +13992,14 @@ lpfc_fc_frame_check(struct lpfc_hba *phba, struct fc_frame_header *fc_hdr)
 	}
 
 	lpfc_printf_log(phba, KERN_INFO, LOG_ELS,
-			"2538 Received frame rctl:%s type:%s "
-			"Frame Data:%08x %08x %08x %08x %08x %08x\n",
-			rctl_names[fc_hdr->fh_r_ctl],
-			type_names[fc_hdr->fh_type],
+			"2538 Received frame rctl:%s (x%x), type:%s (x%x), "
+			"frame Data:%08x %08x %08x %08x %08x %08x %08x\n",
+			rctl_names[fc_hdr->fh_r_ctl], fc_hdr->fh_r_ctl,
+			type_names[fc_hdr->fh_type], fc_hdr->fh_type,
 			be32_to_cpu(header[0]), be32_to_cpu(header[1]),
 			be32_to_cpu(header[2]), be32_to_cpu(header[3]),
-			be32_to_cpu(header[4]), be32_to_cpu(header[5]));
+			be32_to_cpu(header[4]), be32_to_cpu(header[5]),
+			be32_to_cpu(header[6]));
 	return 0;
 drop:
 	lpfc_printf_log(phba, KERN_WARNING, LOG_ELS,
diff --git a/drivers/scsi/lpfc/lpfc_sli4.h b/drivers/scsi/lpfc/lpfc_sli4.h
index be02b59ea279..67af460184ba 100644
--- a/drivers/scsi/lpfc/lpfc_sli4.h
+++ b/drivers/scsi/lpfc/lpfc_sli4.h
@@ -346,11 +346,6 @@ struct lpfc_bmbx {
 #define SLI4_CT_VFI 2
 #define SLI4_CT_FCFI 3
 
-#define LPFC_SLI4_FL1_MAX_SEGMENT_SIZE	0x10000
-#define LPFC_SLI4_FL1_MAX_BUF_SIZE	0X2000
-#define LPFC_SLI4_MIN_BUF_SIZE		0x400
-#define LPFC_SLI4_MAX_BUF_SIZE		0x20000
-
 /*
  * SLI4 specific data structures
  */
@@ -440,6 +435,17 @@ struct lpfc_sli4_lnk_info {
 
 #define LPFC_SLI4_HANDLER_NAME_SZ	16
 
+/* Used for IRQ vector to CPU mapping */
+struct lpfc_vector_map_info {
+	uint16_t	phys_id;
+	uint16_t	core_id;
+	uint16_t	irq;
+	uint16_t	channel_id;
+	struct cpumask	maskbits;
+};
+#define LPFC_VECTOR_MAP_EMPTY	0xffff
+#define LPFC_MAX_CPU		256
+
 /* SLI4 HBA data structure entries */
 struct lpfc_sli4_hba {
 	void __iomem *conf_regs_memmap_p; /* Kernel memory mapped address for
@@ -573,6 +579,11 @@ struct lpfc_sli4_hba {
 	struct lpfc_iov iov;
 	spinlock_t abts_scsi_buf_list_lock; /* list of aborted SCSI IOs */
 	spinlock_t abts_sgl_list_lock; /* list of aborted els IOs */
+
+	/* CPU to vector mapping information */
+	struct lpfc_vector_map_info *cpu_map;
+	uint16_t num_online_cpu;
+	uint16_t num_present_cpu;
 };
 
 enum lpfc_sge_type {
diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h
index 664cd04f7cd8..a38dc3b16969 100644
--- a/drivers/scsi/lpfc/lpfc_version.h
+++ b/drivers/scsi/lpfc/lpfc_version.h
@@ -18,7 +18,7 @@
  * included with this package.                                     *
  *******************************************************************/
 
-#define LPFC_DRIVER_VERSION "8.3.38"
+#define LPFC_DRIVER_VERSION "8.3.39"
 #define LPFC_DRIVER_NAME		"lpfc"
 
 /* Used for SLI 2/3 */
diff --git a/drivers/scsi/lpfc/lpfc_vport.c b/drivers/scsi/lpfc/lpfc_vport.c
index 0fe188e66000..e28e431564b0 100644
--- a/drivers/scsi/lpfc/lpfc_vport.c
+++ b/drivers/scsi/lpfc/lpfc_vport.c
@@ -80,7 +80,7 @@ inline void lpfc_vport_set_state(struct lpfc_vport *vport,
 	}
 }
 
-static int
+int
 lpfc_alloc_vpi(struct lpfc_hba *phba)
 {
 	unsigned long vpi;
@@ -568,6 +568,7 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 	struct lpfc_vport *vport = *(struct lpfc_vport **)fc_vport->dd_data;
 	struct lpfc_hba   *phba = vport->phba;
 	long timeout;
+	bool ns_ndlp_referenced = false;
 
 	if (vport->port_type == LPFC_PHYSICAL_PORT) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_VPORT,
@@ -628,6 +629,18 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 
 	lpfc_debugfs_terminate(vport);
 
+	/*
+	 * The call to fc_remove_host might release the NameServer ndlp. Since
+	 * we might need to use the ndlp to send the DA_ID CT command,
+	 * increment the reference for the NameServer ndlp to prevent it from
+	 * being released.
+	 */
+	ndlp = lpfc_findnode_did(vport, NameServer_DID);
+	if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
+		lpfc_nlp_get(ndlp);
+		ns_ndlp_referenced = true;
+	}
+
 	/* Remove FC host and then SCSI host with the vport */
 	fc_remove_host(lpfc_shost_from_vport(vport));
 	scsi_remove_host(lpfc_shost_from_vport(vport));
@@ -734,6 +747,16 @@ lpfc_vport_delete(struct fc_vport *fc_vport)
 		lpfc_discovery_wait(vport);
 
 skip_logo:
+
+	/*
+	 * If the NameServer ndlp has been incremented to allow the DA_ID CT
+	 * command to be sent, decrement the ndlp now.
+	 */
+	if (ns_ndlp_referenced) {
+		ndlp = lpfc_findnode_did(vport, NameServer_DID);
+		lpfc_nlp_put(ndlp);
+	}
+
 	lpfc_cleanup(vport);
 	lpfc_sli_host_down(vport);
 
diff --git a/drivers/scsi/lpfc/lpfc_vport.h b/drivers/scsi/lpfc/lpfc_vport.h
index 90828340acea..6b2c94eb8134 100644
--- a/drivers/scsi/lpfc/lpfc_vport.h
+++ b/drivers/scsi/lpfc/lpfc_vport.h
@@ -90,6 +90,7 @@ int lpfc_vport_getinfo(struct Scsi_Host *, struct vport_info *);
 int lpfc_vport_tgt_remove(struct Scsi_Host *, uint, uint);
 struct lpfc_vport **lpfc_create_vport_work_array(struct lpfc_hba *);
 void lpfc_destroy_vport_work_array(struct lpfc_hba *, struct lpfc_vport **);
+int lpfc_alloc_vpi(struct lpfc_hba *phba);
 
 /*
  *  queuecommand  VPORT-specific return codes. Specified in  the host byte code.