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

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (45 commits)
  [SCSI] Fix block queue and elevator memory leak in scsi_alloc_sdev
  [SCSI] scsi_dh_alua: Fix the time inteval for alua rtpg commands
  [SCSI] scsi_transport_iscsi: Fix documentation os parameter
  [SCSI] mv_sas: OCZ RevoDrive3 & zDrive R4 support
  [SCSI] libfc: improve flogi retries to avoid lport stuck
  [SCSI] libfc: avoid exchanges collision during lport reset
  [SCSI] libfc: fix checking FC_TYPE_BLS
  [SCSI] edd: Treat "XPRS" host bus type the same as "PCI"
  [SCSI] isci: overriding max_concurr_spinup oem parameter by max(oem, user)
  [SCSI] isci: revert bcn filtering
  [SCSI] isci: Fix hard reset timeout conditions.
  [SCSI] isci: No need to manage the pending reset bit on pending requests.
  [SCSI] isci: Remove redundant isci_request.ttype field.
  [SCSI] isci: Fix task management for SMP, SATA and on dev remove.
  [SCSI] isci: No task_done callbacks in error handler paths.
  [SCSI] isci: Handle task request timeouts correctly.
  [SCSI] isci: Fix tag leak in tasks and terminated requests.
  [SCSI] isci: Immediately fail I/O to removed devices.
  [SCSI] isci: Lookup device references through requests in completions.
  [SCSI] ipr: add definitions for additional adapter
  ...

44 files changed, 1754 insertions, 1076 deletions

diff --git a/drivers/firmware/edd.c b/drivers/firmware/edd.c
index f1b7f659d3c..e2295766580 100644
--- a/drivers/firmware/edd.c
+++ b/drivers/firmware/edd.c
@@ -151,7 +151,8 @@ edd_show_host_bus(struct edd_device *edev, char *buf)
 		p += scnprintf(p, left, "\tbase_address: %x\n",
 			     info->params.interface_path.isa.base_address);
 	} else if (!strncmp(info->params.host_bus_type, "PCIX", 4) ||
-		   !strncmp(info->params.host_bus_type, "PCI", 3)) {
+		   !strncmp(info->params.host_bus_type, "PCI", 3) ||
+		   !strncmp(info->params.host_bus_type, "XPRS", 4)) {
 		p += scnprintf(p, left,
 			     "\t%02x:%02x.%d  channel: %u\n",
 			     info->params.interface_path.pci.bus,
@@ -159,7 +160,6 @@ edd_show_host_bus(struct edd_device *edev, char *buf)
 			     info->params.interface_path.pci.function,
 			     info->params.interface_path.pci.channel);
 	} else if (!strncmp(info->params.host_bus_type, "IBND", 4) ||
-		   !strncmp(info->params.host_bus_type, "XPRS", 4) ||
 		   !strncmp(info->params.host_bus_type, "HTPT", 4)) {
 		p += scnprintf(p, left,
 			     "\tTBD: %llx\n",
@@ -668,7 +668,7 @@ edd_get_pci_dev(struct edd_device *edev)
 {
 	struct edd_info *info = edd_dev_get_info(edev);
 
-	if (edd_dev_is_type(edev, "PCI")) {
+	if (edd_dev_is_type(edev, "PCI") || edd_dev_is_type(edev, "XPRS")) {
 		return pci_get_bus_and_slot(info->params.interface_path.pci.bus,
 				     PCI_DEVFN(info->params.interface_path.pci.slot,
 					       info->params.interface_path.pci.
diff --git a/drivers/scsi/bnx2fc/bnx2fc.h b/drivers/scsi/bnx2fc/bnx2fc.h
index 63de1c7cd0c..049ea907e04 100644
--- a/drivers/scsi/bnx2fc/bnx2fc.h
+++ b/drivers/scsi/bnx2fc/bnx2fc.h
@@ -62,7 +62,7 @@
 #include "bnx2fc_constants.h"
 
 #define BNX2FC_NAME		"bnx2fc"
-#define BNX2FC_VERSION		"1.0.8"
+#define BNX2FC_VERSION		"1.0.9"
 
 #define PFX			"bnx2fc: "
 
@@ -145,6 +145,9 @@
 #define REC_RETRY_COUNT			1
 #define BNX2FC_NUM_ERR_BITS		63
 
+#define BNX2FC_RELOGIN_WAIT_TIME	200
+#define BNX2FC_RELOGIN_WAIT_CNT		10
+
 /* bnx2fc driver uses only one instance of fcoe_percpu_s */
 extern struct fcoe_percpu_s bnx2fc_global;
 
diff --git a/drivers/scsi/bnx2fc/bnx2fc_els.c b/drivers/scsi/bnx2fc/bnx2fc_els.c
index fd382fe33f6..ce0ce3e32f3 100644
--- a/drivers/scsi/bnx2fc/bnx2fc_els.c
+++ b/drivers/scsi/bnx2fc/bnx2fc_els.c
@@ -268,17 +268,6 @@ void bnx2fc_srr_compl(struct bnx2fc_els_cb_arg *cb_arg)
 
 	orig_io_req = cb_arg->aborted_io_req;
 	srr_req = cb_arg->io_req;
-	if (test_bit(BNX2FC_FLAG_IO_COMPL, &orig_io_req->req_flags)) {
-		BNX2FC_IO_DBG(srr_req, "srr_compl: xid - 0x%x completed",
-			orig_io_req->xid);
-		goto srr_compl_done;
-	}
-	if (test_bit(BNX2FC_FLAG_ISSUE_ABTS, &orig_io_req->req_flags)) {
-		BNX2FC_IO_DBG(srr_req, "rec abts in prog "
-		       "orig_io - 0x%x\n",
-			orig_io_req->xid);
-		goto srr_compl_done;
-	}
 	if (test_and_clear_bit(BNX2FC_FLAG_ELS_TIMEOUT, &srr_req->req_flags)) {
 		/* SRR timedout */
 		BNX2FC_IO_DBG(srr_req, "srr timed out, abort "
@@ -290,6 +279,12 @@ void bnx2fc_srr_compl(struct bnx2fc_els_cb_arg *cb_arg)
 				"failed. issue cleanup\n");
 			bnx2fc_initiate_cleanup(srr_req);
 		}
+		if (test_bit(BNX2FC_FLAG_IO_COMPL, &orig_io_req->req_flags) ||
+		    test_bit(BNX2FC_FLAG_ISSUE_ABTS, &orig_io_req->req_flags)) {
+			BNX2FC_IO_DBG(srr_req, "srr_compl:xid 0x%x flags = %lx",
+				      orig_io_req->xid, orig_io_req->req_flags);
+			goto srr_compl_done;
+		}
 		orig_io_req->srr_retry++;
 		if (orig_io_req->srr_retry <= SRR_RETRY_COUNT) {
 			struct bnx2fc_rport *tgt = orig_io_req->tgt;
@@ -311,6 +306,12 @@ void bnx2fc_srr_compl(struct bnx2fc_els_cb_arg *cb_arg)
 		}
 		goto srr_compl_done;
 	}
+	if (test_bit(BNX2FC_FLAG_IO_COMPL, &orig_io_req->req_flags) ||
+	    test_bit(BNX2FC_FLAG_ISSUE_ABTS, &orig_io_req->req_flags)) {
+		BNX2FC_IO_DBG(srr_req, "srr_compl:xid - 0x%x flags = %lx",
+			      orig_io_req->xid, orig_io_req->req_flags);
+		goto srr_compl_done;
+	}
 	mp_req = &(srr_req->mp_req);
 	fc_hdr = &(mp_req->resp_fc_hdr);
 	resp_len = mp_req->resp_len;
diff --git a/drivers/scsi/bnx2fc/bnx2fc_fcoe.c b/drivers/scsi/bnx2fc/bnx2fc_fcoe.c
index 85bcc4b5596..8c6156a10d9 100644
--- a/drivers/scsi/bnx2fc/bnx2fc_fcoe.c
+++ b/drivers/scsi/bnx2fc/bnx2fc_fcoe.c
@@ -22,7 +22,7 @@ DEFINE_PER_CPU(struct bnx2fc_percpu_s, bnx2fc_percpu);
 
 #define DRV_MODULE_NAME		"bnx2fc"
 #define DRV_MODULE_VERSION	BNX2FC_VERSION
-#define DRV_MODULE_RELDATE	"Oct 02, 2011"
+#define DRV_MODULE_RELDATE	"Oct 21, 2011"
 
 
 static char version[] __devinitdata =
diff --git a/drivers/scsi/bnx2fc/bnx2fc_io.c b/drivers/scsi/bnx2fc/bnx2fc_io.c
index 0c64d184d73..84a78af83f9 100644
--- a/drivers/scsi/bnx2fc/bnx2fc_io.c
+++ b/drivers/scsi/bnx2fc/bnx2fc_io.c
@@ -1103,7 +1103,10 @@ int bnx2fc_eh_abort(struct scsi_cmnd *sc_cmd)
 	struct fc_rport_libfc_priv *rp = rport->dd_data;
 	struct bnx2fc_cmd *io_req;
 	struct fc_lport *lport;
+	struct fc_rport_priv *rdata;
 	struct bnx2fc_rport *tgt;
+	int logo_issued;
+	int wait_cnt = 0;
 	int rc = FAILED;
 
 
@@ -1192,8 +1195,40 @@ int bnx2fc_eh_abort(struct scsi_cmnd *sc_cmd)
 	} else {
 		printk(KERN_ERR PFX "eh_abort: io_req (xid = 0x%x) "
 				"already in abts processing\n", io_req->xid);
+		if (cancel_delayed_work(&io_req->timeout_work))
+			kref_put(&io_req->refcount,
+				 bnx2fc_cmd_release); /* drop timer hold */
+		bnx2fc_initiate_cleanup(io_req);
+
+		spin_unlock_bh(&tgt->tgt_lock);
+
+		wait_for_completion(&io_req->tm_done);
+
+		spin_lock_bh(&tgt->tgt_lock);
+		io_req->wait_for_comp = 0;
+		rdata = io_req->tgt->rdata;
+		logo_issued = test_and_set_bit(BNX2FC_FLAG_EXPL_LOGO,
+					       &tgt->flags);
 		kref_put(&io_req->refcount, bnx2fc_cmd_release);
 		spin_unlock_bh(&tgt->tgt_lock);
+
+		if (!logo_issued) {
+			BNX2FC_IO_DBG(io_req, "Expl logo - tgt flags = 0x%lx\n",
+				      tgt->flags);
+			mutex_lock(&lport->disc.disc_mutex);
+			lport->tt.rport_logoff(rdata);
+			mutex_unlock(&lport->disc.disc_mutex);
+			do {
+				msleep(BNX2FC_RELOGIN_WAIT_TIME);
+				/*
+				 * If session not recovered, let SCSI-ml
+				 * escalate error recovery.
+				 */
+				if (wait_cnt++ > BNX2FC_RELOGIN_WAIT_CNT)
+					return FAILED;
+			} while (!test_bit(BNX2FC_FLAG_SESSION_READY,
+					   &tgt->flags));
+		}
 		return SUCCESS;
 	}
 	if (rc == FAILED) {
@@ -1275,6 +1310,8 @@ void bnx2fc_process_cleanup_compl(struct bnx2fc_cmd *io_req,
 		   io_req->refcount.refcount.counter, io_req->cmd_type);
 	bnx2fc_scsi_done(io_req, DID_ERROR);
 	kref_put(&io_req->refcount, bnx2fc_cmd_release);
+	if (io_req->wait_for_comp)
+		complete(&io_req->tm_done);
 }
 
 void bnx2fc_process_abts_compl(struct bnx2fc_cmd *io_req,
diff --git a/drivers/scsi/device_handler/scsi_dh.c b/drivers/scsi/device_handler/scsi_dh.c
index 7c05fd9dccf..339ea23a867 100644
--- a/drivers/scsi/device_handler/scsi_dh.c
+++ b/drivers/scsi/device_handler/scsi_dh.c
@@ -441,7 +441,15 @@ int scsi_dh_activate(struct request_queue *q, activate_complete fn, void *data)
 
 	spin_lock_irqsave(q->queue_lock, flags);
 	sdev = q->queuedata;
-	if (sdev && sdev->scsi_dh_data)
+	if (!sdev) {
+		spin_unlock_irqrestore(q->queue_lock, flags);
+		err = SCSI_DH_NOSYS;
+		if (fn)
+			fn(data, err);
+		return err;
+	}
+
+	if (sdev->scsi_dh_data)
 		scsi_dh = sdev->scsi_dh_data->scsi_dh;
 	dev = get_device(&sdev->sdev_gendev);
 	if (!scsi_dh || !dev ||
diff --git a/drivers/scsi/device_handler/scsi_dh_alua.c b/drivers/scsi/device_handler/scsi_dh_alua.c
index 627f4b5e517..fe4df2da309 100644
--- a/drivers/scsi/device_handler/scsi_dh_alua.c
+++ b/drivers/scsi/device_handler/scsi_dh_alua.c
@@ -507,7 +507,7 @@ static int alua_rtpg(struct scsi_device *sdev, struct alua_dh_data *h)
 	int len, k, off, valid_states = 0;
 	unsigned char *ucp;
 	unsigned err;
-	unsigned long expiry, interval = 1;
+	unsigned long expiry, interval = 1000;
 
 	expiry = round_jiffies_up(jiffies + ALUA_FAILOVER_TIMEOUT);
  retry:
@@ -734,6 +734,7 @@ static int alua_bus_attach(struct scsi_device *sdev)
 	spin_lock_irqsave(sdev->request_queue->queue_lock, flags);
 	sdev->scsi_dh_data = scsi_dh_data;
 	spin_unlock_irqrestore(sdev->request_queue->queue_lock, flags);
+	sdev_printk(KERN_NOTICE, sdev, "%s: Attached\n", ALUA_DH_NAME);
 
 	return 0;
 
diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
index 61384ee4049..cefbe44bb84 100644
--- a/drivers/scsi/fcoe/fcoe.c
+++ b/drivers/scsi/fcoe/fcoe.c
@@ -2347,14 +2347,11 @@ static void fcoe_flogi_resp(struct fc_seq *seq, struct fc_frame *fp, void *arg)
 		goto done;
 
 	mac = fr_cb(fp)->granted_mac;
-	if (is_zero_ether_addr(mac)) {
-		/* pre-FIP */
-		if (fcoe_ctlr_recv_flogi(fip, lport, fp)) {
-			fc_frame_free(fp);
-			return;
-		}
-	}
-	fcoe_update_src_mac(lport, mac);
+	/* pre-FIP */
+	if (is_zero_ether_addr(mac))
+		fcoe_ctlr_recv_flogi(fip, lport, fp);
+	if (!is_zero_ether_addr(mac))
+		fcoe_update_src_mac(lport, mac);
 done:
 	fc_lport_flogi_resp(seq, fp, lport);
 }
diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
index 4f7a5829ea4..351dc0b86fa 100644
--- a/drivers/scsi/hosts.c
+++ b/drivers/scsi/hosts.c
@@ -286,6 +286,7 @@ static void scsi_host_dev_release(struct device *dev)
 {
 	struct Scsi_Host *shost = dev_to_shost(dev);
 	struct device *parent = dev->parent;
+	struct request_queue *q;
 
 	scsi_proc_hostdir_rm(shost->hostt);
 
@@ -293,9 +294,11 @@ static void scsi_host_dev_release(struct device *dev)
 		kthread_stop(shost->ehandler);
 	if (shost->work_q)
 		destroy_workqueue(shost->work_q);
-	if (shost->uspace_req_q) {
-		kfree(shost->uspace_req_q->queuedata);
-		scsi_free_queue(shost->uspace_req_q);
+	q = shost->uspace_req_q;
+	if (q) {
+		kfree(q->queuedata);
+		q->queuedata = NULL;
+		scsi_free_queue(q);
 	}
 
 	scsi_destroy_command_freelist(shost);
diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index bbdc9f960a6..e76107b2ade 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -48,6 +48,7 @@
 #include <linux/bitmap.h>
 #include <linux/atomic.h>
 #include <linux/kthread.h>
+#include <linux/jiffies.h>
 #include "hpsa_cmd.h"
 #include "hpsa.h"
 
@@ -127,6 +128,10 @@ static struct board_type products[] = {
 
 static int number_of_controllers;
 
+static struct list_head hpsa_ctlr_list = LIST_HEAD_INIT(hpsa_ctlr_list);
+static spinlock_t lockup_detector_lock;
+static struct task_struct *hpsa_lockup_detector;
+
 static irqreturn_t do_hpsa_intr_intx(int irq, void *dev_id);
 static irqreturn_t do_hpsa_intr_msi(int irq, void *dev_id);
 static int hpsa_ioctl(struct scsi_device *dev, int cmd, void *arg);
@@ -484,6 +489,7 @@ static struct scsi_host_template hpsa_driver_template = {
 #endif
 	.sdev_attrs = hpsa_sdev_attrs,
 	.shost_attrs = hpsa_shost_attrs,
+	.max_sectors = 8192,
 };
 
 
@@ -566,16 +572,16 @@ static int hpsa_find_target_lun(struct ctlr_info *h,
 	 * assumes h->devlock is held
 	 */
 	int i, found = 0;
-	DECLARE_BITMAP(lun_taken, HPSA_MAX_SCSI_DEVS_PER_HBA);
+	DECLARE_BITMAP(lun_taken, HPSA_MAX_DEVICES);
 
-	memset(&lun_taken[0], 0, HPSA_MAX_SCSI_DEVS_PER_HBA >> 3);
+	memset(&lun_taken[0], 0, HPSA_MAX_DEVICES >> 3);
 
 	for (i = 0; i < h->ndevices; i++) {
 		if (h->dev[i]->bus == bus && h->dev[i]->target != -1)
 			set_bit(h->dev[i]->target, lun_taken);
 	}
 
-	for (i = 0; i < HPSA_MAX_SCSI_DEVS_PER_HBA; i++) {
+	for (i = 0; i < HPSA_MAX_DEVICES; i++) {
 		if (!test_bit(i, lun_taken)) {
 			/* *bus = 1; */
 			*target = i;
@@ -598,7 +604,7 @@ static int hpsa_scsi_add_entry(struct ctlr_info *h, int hostno,
 	unsigned char addr1[8], addr2[8];
 	struct hpsa_scsi_dev_t *sd;
 
-	if (n >= HPSA_MAX_SCSI_DEVS_PER_HBA) {
+	if (n >= HPSA_MAX_DEVICES) {
 		dev_err(&h->pdev->dev, "too many devices, some will be "
 			"inaccessible.\n");
 		return -1;
@@ -673,7 +679,7 @@ static void hpsa_scsi_replace_entry(struct ctlr_info *h, int hostno,
 	struct hpsa_scsi_dev_t *removed[], int *nremoved)
 {
 	/* assumes h->devlock is held */
-	BUG_ON(entry < 0 || entry >= HPSA_MAX_SCSI_DEVS_PER_HBA);
+	BUG_ON(entry < 0 || entry >= HPSA_MAX_DEVICES);
 	removed[*nremoved] = h->dev[entry];
 	(*nremoved)++;
 
@@ -702,7 +708,7 @@ static void hpsa_scsi_remove_entry(struct ctlr_info *h, int hostno, int entry,
 	int i;
 	struct hpsa_scsi_dev_t *sd;
 
-	BUG_ON(entry < 0 || entry >= HPSA_MAX_SCSI_DEVS_PER_HBA);
+	BUG_ON(entry < 0 || entry >= HPSA_MAX_DEVICES);
 
 	sd = h->dev[entry];
 	removed[*nremoved] = h->dev[entry];
@@ -814,10 +820,8 @@ static void adjust_hpsa_scsi_table(struct ctlr_info *h, int hostno,
 	int nadded, nremoved;
 	struct Scsi_Host *sh = NULL;
 
-	added = kzalloc(sizeof(*added) * HPSA_MAX_SCSI_DEVS_PER_HBA,
-		GFP_KERNEL);
-	removed = kzalloc(sizeof(*removed) * HPSA_MAX_SCSI_DEVS_PER_HBA,
-		GFP_KERNEL);
+	added = kzalloc(sizeof(*added) * HPSA_MAX_DEVICES, GFP_KERNEL);
+	removed = kzalloc(sizeof(*removed) * HPSA_MAX_DEVICES, GFP_KERNEL);
 
 	if (!added || !removed) {
 		dev_warn(&h->pdev->dev, "out of memory in "
@@ -1338,6 +1342,22 @@ static inline void hpsa_scsi_do_simple_cmd_core(struct ctlr_info *h,
 	wait_for_completion(&wait);
 }
 
+static void hpsa_scsi_do_simple_cmd_core_if_no_lockup(struct ctlr_info *h,
+	struct CommandList *c)
+{
+	unsigned long flags;
+
+	/* If controller lockup detected, fake a hardware error. */
+	spin_lock_irqsave(&h->lock, flags);
+	if (unlikely(h->lockup_detected)) {
+		spin_unlock_irqrestore(&h->lock, flags);
+		c->err_info->CommandStatus = CMD_HARDWARE_ERR;
+	} else {
+		spin_unlock_irqrestore(&h->lock, flags);
+		hpsa_scsi_do_simple_cmd_core(h, c);
+	}
+}
+
 static void hpsa_scsi_do_simple_cmd_with_retry(struct ctlr_info *h,
 	struct CommandList *c, int data_direction)
 {
@@ -1735,7 +1755,6 @@ static int add_msa2xxx_enclosure_device(struct ctlr_info *h,
 	if (is_scsi_rev_5(h))
 		return 0; /* p1210m doesn't need to do this. */
 
-#define MAX_MSA2XXX_ENCLOSURES 32
 	if (*nmsa2xxx_enclosures >= MAX_MSA2XXX_ENCLOSURES) {
 		dev_warn(&h->pdev->dev, "Maximum number of MSA2XXX "
 			"enclosures exceeded.  Check your hardware "
@@ -1846,8 +1865,7 @@ static void hpsa_update_scsi_devices(struct ctlr_info *h, int hostno)
 	int raid_ctlr_position;
 	DECLARE_BITMAP(lunzerobits, HPSA_MAX_TARGETS_PER_CTLR);
 
-	currentsd = kzalloc(sizeof(*currentsd) * HPSA_MAX_SCSI_DEVS_PER_HBA,
-		GFP_KERNEL);
+	currentsd = kzalloc(sizeof(*currentsd) * HPSA_MAX_DEVICES, GFP_KERNEL);
 	physdev_list = kzalloc(reportlunsize, GFP_KERNEL);
 	logdev_list = kzalloc(reportlunsize, GFP_KERNEL);
 	tmpdevice = kzalloc(sizeof(*tmpdevice), GFP_KERNEL);
@@ -1870,6 +1888,13 @@ static void hpsa_update_scsi_devices(struct ctlr_info *h, int hostno)
 
 	/* Allocate the per device structures */
 	for (i = 0; i < ndevs_to_allocate; i++) {
+		if (i >= HPSA_MAX_DEVICES) {
+			dev_warn(&h->pdev->dev, "maximum devices (%d) exceeded."
+				"  %d devices ignored.\n", HPSA_MAX_DEVICES,
+				ndevs_to_allocate - HPSA_MAX_DEVICES);
+			break;
+		}
+
 		currentsd[i] = kzalloc(sizeof(*currentsd[i]), GFP_KERNEL);
 		if (!currentsd[i]) {
 			dev_warn(&h->pdev->dev, "out of memory at %s:%d\n",
@@ -1956,7 +1981,7 @@ static void hpsa_update_scsi_devices(struct ctlr_info *h, int hostno)
 		default:
 			break;
 		}
-		if (ncurrent >= HPSA_MAX_SCSI_DEVS_PER_HBA)
+		if (ncurrent >= HPSA_MAX_DEVICES)
 			break;
 	}
 	adjust_hpsa_scsi_table(h, hostno, currentsd, ncurrent);
@@ -2048,8 +2073,14 @@ static int hpsa_scsi_queue_command_lck(struct scsi_cmnd *cmd,
 	}
 	memcpy(scsi3addr, dev->scsi3addr, sizeof(scsi3addr));
 
-	/* Need a lock as this is being allocated from the pool */
 	spin_lock_irqsave(&h->lock, flags);
+	if (unlikely(h->lockup_detected)) {
+		spin_unlock_irqrestore(&h->lock, flags);
+		cmd->result = DID_ERROR << 16;
+		done(cmd);
+		return 0;
+	}
+	/* Need a lock as this is being allocated from the pool */
 	c = cmd_alloc(h);
 	spin_unlock_irqrestore(&h->lock, flags);
 	if (c == NULL) {			/* trouble... */
@@ -2601,7 +2632,7 @@ static int hpsa_passthru_ioctl(struct ctlr_info *h, void __user *argp)
 		c->SG[0].Len = iocommand.buf_size;
 		c->SG[0].Ext = 0; /* we are not chaining*/
 	}
-	hpsa_scsi_do_simple_cmd_core(h, c);
+	hpsa_scsi_do_simple_cmd_core_if_no_lockup(h, c);
 	if (iocommand.buf_size > 0)
 		hpsa_pci_unmap(h->pdev, c, 1, PCI_DMA_BIDIRECTIONAL);
 	check_ioctl_unit_attention(h, c);
@@ -2724,7 +2755,7 @@ static int hpsa_big_passthru_ioctl(struct ctlr_info *h, void __user *argp)
 			c->SG[i].Ext = 0;
 		}
 	}
-	hpsa_scsi_do_simple_cmd_core(h, c);
+	hpsa_scsi_do_simple_cmd_core_if_no_lockup(h, c);
 	if (sg_used)
 		hpsa_pci_unmap(h->pdev, c, sg_used, PCI_DMA_BIDIRECTIONAL);
 	check_ioctl_unit_attention(h, c);
@@ -2872,6 +2903,8 @@ static void fill_cmd(struct CommandList *c, u8 cmd, struct ctlr_info *h,
 			c->Request.Timeout = 0;
 			c->Request.CDB[0] = BMIC_WRITE;
 			c->Request.CDB[6] = BMIC_CACHE_FLUSH;
+			c->Request.CDB[7] = (size >> 8) & 0xFF;
+			c->Request.CDB[8] = size & 0xFF;
 			break;
 		case TEST_UNIT_READY:
 			c->Request.CDBLen = 6;
@@ -3091,6 +3124,7 @@ static irqreturn_t hpsa_intx_discard_completions(int irq, void *dev_id)
 	if (interrupt_not_for_us(h))
 		return IRQ_NONE;
 	spin_lock_irqsave(&h->lock, flags);
+	h->last_intr_timestamp = get_jiffies_64();
 	while (interrupt_pending(h)) {
 		raw_tag = get_next_completion(h);
 		while (raw_tag != FIFO_EMPTY)
@@ -3110,6 +3144,7 @@ static irqreturn_t hpsa_msix_discard_completions(int irq, void *dev_id)
 		return IRQ_NONE;
 
 	spin_lock_irqsave(&h->lock, flags);
+	h->last_intr_timestamp = get_jiffies_64();
 	raw_tag = get_next_completion(h);
 	while (raw_tag != FIFO_EMPTY)
 		raw_tag = next_command(h);
@@ -3126,6 +3161,7 @@ static irqreturn_t do_hpsa_intr_intx(int irq, void *dev_id)
 	if (interrupt_not_for_us(h))
 		return IRQ_NONE;
 	spin_lock_irqsave(&h->lock, flags);
+	h->last_intr_timestamp = get_jiffies_64();
 	while (interrupt_pending(h)) {
 		raw_tag = get_next_completion(h);
 		while (raw_tag != FIFO_EMPTY) {
@@ -3146,6 +3182,7 @@ static irqreturn_t do_hpsa_intr_msi(int irq, void *dev_id)
 	u32 raw_tag;
 
 	spin_lock_irqsave(&h->lock, flags);
+	h->last_intr_timestamp = get_jiffies_64();
 	raw_tag = get_next_completion(h);
 	while (raw_tag != FIFO_EMPTY) {
 		if (hpsa_tag_contains_index(raw_tag))
@@ -4090,6 +4127,149 @@ static void hpsa_undo_allocations_after_kdump_soft_reset(struct ctlr_info *h)
 	kfree(h);
 }
 
+static void remove_ctlr_from_lockup_detector_list(struct ctlr_info *h)
+{
+	assert_spin_locked(&lockup_detector_lock);
+	if (!hpsa_lockup_detector)
+		return;
+	if (h->lockup_detected)
+		return; /* already stopped the lockup detector */
+	list_del(&h->lockup_list);
+}
+
+/* Called when controller lockup detected. */
+static void fail_all_cmds_on_list(struct ctlr_info *h, struct list_head *list)
+{
+	struct CommandList *c = NULL;
+
+	assert_spin_locked(&h->lock);
+	/* Mark all outstanding commands as failed and complete them. */
+	while (!list_empty(list)) {
+		c = list_entry(list->next, struct CommandList, list);
+		c->err_info->CommandStatus = CMD_HARDWARE_ERR;
+		finish_cmd(c, c->Header.Tag.lower);
+	}
+}
+
+static void controller_lockup_detected(struct ctlr_info *h)
+{
+	unsigned long flags;
+
+	assert_spin_locked(&lockup_detector_lock);
+	remove_ctlr_from_lockup_detector_list(h);
+	h->access.set_intr_mask(h, HPSA_INTR_OFF);
+	spin_lock_irqsave(&h->lock, flags);
+	h->lockup_detected = readl(h->vaddr + SA5_SCRATCHPAD_OFFSET);
+	spin_unlock_irqrestore(&h->lock, flags);
+	dev_warn(&h->pdev->dev, "Controller lockup detected: 0x%08x\n",
+			h->lockup_detected);
+	pci_disable_device(h->pdev);
+	spin_lock_irqsave(&h->lock, flags);
+	fail_all_cmds_on_list(h, &h->cmpQ);
+	fail_all_cmds_on_list(h, &h->reqQ);
+	spin_unlock_irqrestore(&h->lock, flags);
+}
+
+#define HEARTBEAT_SAMPLE_INTERVAL (10 * HZ)
+#define HEARTBEAT_CHECK_MINIMUM_INTERVAL (HEARTBEAT_SAMPLE_INTERVAL / 2)
+
+static void detect_controller_lockup(struct ctlr_info *h)
+{
+	u64 now;
+	u32 heartbeat;
+	unsigned long flags;
+
+	assert_spin_locked(&lockup_detector_lock);
+	now = get_jiffies_64();
+	/* If we've received an interrupt recently, we're ok. */
+	if (time_after64(h->last_intr_timestamp +
+				(HEARTBEAT_CHECK_MINIMUM_INTERVAL), now))
+		return;
+
+	/*
+	 * If we've already checked the heartbeat recently, we're ok.
+	 * This could happen if someone sends us a signal. We
+	 * otherwise don't care about signals in this thread.
+	 */
+	if (time_after64(h->last_heartbeat_timestamp +
+				(HEARTBEAT_CHECK_MINIMUM_INTERVAL), now))
+		return;
+
+	/* If heartbeat has not changed since we last looked, we're not ok. */
+	spin_lock_irqsave(&h->lock, flags);
+	heartbeat = readl(&h->cfgtable->HeartBeat);
+	spin_unlock_irqrestore(&h->lock, flags);
+	if (h->last_heartbeat == heartbeat) {
+		controller_lockup_detected(h);
+		return;
+	}
+
+	/* We're ok. */
+	h->last_heartbeat = heartbeat;
+	h->last_heartbeat_timestamp = now;
+}
+
+static int detect_controller_lockup_thread(void *notused)
+{
+	struct ctlr_info *h;
+	unsigned long flags;
+
+	while (1) {
+		struct list_head *this, *tmp;
+
+		schedule_timeout_interruptible(HEARTBEAT_SAMPLE_INTERVAL);
+		if (kthread_should_stop())
+			break;
+		spin_lock_irqsave(&lockup_detector_lock, flags);
+		list_for_each_safe(this, tmp, &hpsa_ctlr_list) {
+			h = list_entry(this, struct ctlr_info, lockup_list);
+			detect_controller_lockup(h);
+		}
+		spin_unlock_irqrestore(&lockup_detector_lock, flags);
+	}
+	return 0;
+}
+
+static void add_ctlr_to_lockup_detector_list(struct ctlr_info *h)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&lockup_detector_lock, flags);
+	list_add_tail(&h->lockup_list, &hpsa_ctlr_list);
+	spin_unlock_irqrestore(&lockup_detector_lock, flags);
+}
+
+static void start_controller_lockup_detector(struct ctlr_info *h)
+{
+	/* Start the lockup detector thread if not already started */
+	if (!hpsa_lockup_detector) {
+		spin_lock_init(&lockup_detector_lock);
+		hpsa_lockup_detector =
+			kthread_run(detect_controller_lockup_thread,
+						NULL, "hpsa");
+	}
+	if (!hpsa_lockup_detector) {
+		dev_warn(&h->pdev->dev,
+			"Could not start lockup detector thread\n");
+		return;
+	}
+	add_ctlr_to_lockup_detector_list(h);
+}
+
+static void stop_controller_lockup_detector(struct ctlr_info *h)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&lockup_detector_lock, flags);
+	remove_ctlr_from_lockup_detector_list(h);
+	/* If the list of ctlr's to monitor is empty, stop the thread */
+	if (list_empty(&hpsa_ctlr_list)) {
+		kthread_stop(hpsa_lockup_detector);
+		hpsa_lockup_detector = NULL;
+	}
+	spin_unlock