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: (60 commits)
  [SCSI] lpfc 8.3.24: Extend BSG infrastructure and add link diagnostics
  [SCSI] lpfc 8.3.24: Add resource extent support
  [SCSI] lpfc 8.3.24: Add request-firmware support
  [SCSI] lpfc 8.3.24: Add SR-IOV control
  [SCSI] lpfc 8.3.24: Extended hardware support and support dump images
  [SCSI] lpfc 8.3.24: Miscellaneous Fixes and Corrections
  [SCSI] libsas: Add option for SATA soft reset
  [SCSI] libsas: check dev->gone before submitting sata i/o
  [SCSI] libsas: fix/amend device gone notification in sas_deform_port()
  [SCSI] MAINTAINERS update for SCSI (new email address)
  [SCSI] Fix Ultrastor asm snippet
  [SCSI] osst: fix warning
  [SCSI] osst: wrong index used in inner loop
  [SCSI] aic94xx: world-writable sysfs update_bios file
  [SCSI] MAINTAINERS: Add drivers/target/ entry
  [SCSI] target: Convert TASK_ATTR to scsi_tcq.h definitions
  [SCSI] target: Convert REPORT_LUNs to use int_to_scsilun
  [SCSI] target: Fix task->task_execute_queue=1 clear bug + LUN_RESET OOPs
  [SCSI] target: Fix bug with task_sg chained transport_free_dev_tasks release
  [SCSI] target: Fix interrupt context bug with stats_lock and core_tmr_alloc_req
  ...

Documentation/scsi/ChangeLog.megaraid_sas

@@ -1,3 +1,17 @@
+Release Date    : Wed. May 11, 2011 17:00:00 PST 2010 -
+			(emaild-id:megaraidlinux@lsi.com)
+			Adam Radford
+Current Version : 00.00.05.38-rc1
+Old Version     : 00.00.05.34-rc1
+    1. Remove MSI-X black list, use MFI_REG_STATE.ready.msiEnable.
+    2. Remove un-used function megasas_return_cmd_for_smid().
+    3. Check MFI_REG_STATE.fault.resetAdapter in megasas_reset_fusion().
+    4. Disable interrupts/free_irq() in megasas_shutdown().
+    5. Fix bug where AENs could be lost in probe() and resume().
+    6. Convert 6,10,12 byte CDB's to 16 byte CDB for large LBA's for FastPath
+       IO.
+    7. Add 1078 OCR support.
+-------------------------------------------------------------------------------
 Release Date    : Thu. Feb 24, 2011 17:00:00 PST 2010 -
 			(emaild-id:megaraidlinux@lsi.com)
 			Adam Radford

MAINTAINERS

@@ -5515,7 +5515,7 @@ F:	drivers/scsi/sg.c
 F:	include/scsi/sg.h
 
 SCSI SUBSYSTEM
-M:	"James E.J. Bottomley" <James.Bottomley@suse.de>
+M:	"James E.J. Bottomley" <JBottomley@parallels.com>
 L:	linux-scsi@vger.kernel.org
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6.git
 T:	git git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-rc-fixes-2.6.git
@@ -6084,6 +6084,17 @@ F:	Documentation/filesystems/sysv-fs.txt
 F:	fs/sysv/
 F:	include/linux/sysv_fs.h
 
+TARGET SUBSYSTEM
+M:	Nicholas A. Bellinger <nab@linux-iscsi.org>
+L:	linux-scsi@vger.kernel.org
+L:	http://groups.google.com/group/linux-iscsi-target-dev
+W:	http://www.linux-iscsi.org
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/nab/lio-core-2.6.git master
+S:	Supported
+F:	drivers/target/
+F:	include/target/
+F:	Documentation/target/
+
 TASKSTATS STATISTICS INTERFACE
 M:	Balbir Singh <balbir@linux.vnet.ibm.com>
 S:	Maintained

drivers/scsi/aic94xx/aic94xx_init.c

@@ -486,7 +486,7 @@ static ssize_t asd_show_update_bios(struct device *dev,
 			flash_error_table[i].reason);
 }
 
-static DEVICE_ATTR(update_bios, S_IRUGO|S_IWUGO,
+static DEVICE_ATTR(update_bios, S_IRUGO|S_IWUSR,
 	asd_show_update_bios, asd_store_update_bios);
 
 static int asd_create_dev_attrs(struct asd_ha_struct *asd_ha)

drivers/scsi/bfa/bfa_ioc.c

@@ -56,6 +56,8 @@ BFA_TRC_FILE(CNA, IOC);
 #define bfa_ioc_map_port(__ioc) ((__ioc)->ioc_hwif->ioc_map_port(__ioc))
 #define bfa_ioc_notify_fail(__ioc)              \
 			((__ioc)->ioc_hwif->ioc_notify_fail(__ioc))
+#define bfa_ioc_sync_start(__ioc)               \
+			((__ioc)->ioc_hwif->ioc_sync_start(__ioc))
 #define bfa_ioc_sync_join(__ioc)                \
 			((__ioc)->ioc_hwif->ioc_sync_join(__ioc))
 #define bfa_ioc_sync_leave(__ioc)               \
@@ -647,7 +649,7 @@ bfa_iocpf_sm_fwcheck(struct bfa_iocpf_s *iocpf, enum iocpf_event event)
 	switch (event) {
 	case IOCPF_E_SEMLOCKED:
 		if (bfa_ioc_firmware_lock(ioc)) {
-			if (bfa_ioc_sync_complete(ioc)) {
+			if (bfa_ioc_sync_start(ioc)) {
 				iocpf->retry_count = 0;
 				bfa_ioc_sync_join(ioc);
 				bfa_fsm_set_state(iocpf, bfa_iocpf_sm_hwinit);

drivers/scsi/bfa/bfa_ioc.h

@@ -263,6 +263,7 @@ struct bfa_ioc_hwif_s {
 					bfa_boolean_t msix);
 	void		(*ioc_notify_fail)	(struct bfa_ioc_s *ioc);
 	void		(*ioc_ownership_reset)	(struct bfa_ioc_s *ioc);
+	bfa_boolean_t   (*ioc_sync_start)       (struct bfa_ioc_s *ioc);
 	void		(*ioc_sync_join)	(struct bfa_ioc_s *ioc);
 	void		(*ioc_sync_leave)	(struct bfa_ioc_s *ioc);
 	void		(*ioc_sync_ack)		(struct bfa_ioc_s *ioc);

drivers/scsi/bfa/bfa_ioc_cb.c

@@ -32,6 +32,7 @@ static void bfa_ioc_cb_map_port(struct bfa_ioc_s *ioc);
 static void bfa_ioc_cb_isr_mode_set(struct bfa_ioc_s *ioc, bfa_boolean_t msix);
 static void bfa_ioc_cb_notify_fail(struct bfa_ioc_s *ioc);
 static void bfa_ioc_cb_ownership_reset(struct bfa_ioc_s *ioc);
+static bfa_boolean_t bfa_ioc_cb_sync_start(struct bfa_ioc_s *ioc);
 static void bfa_ioc_cb_sync_join(struct bfa_ioc_s *ioc);
 static void bfa_ioc_cb_sync_leave(struct bfa_ioc_s *ioc);
 static void bfa_ioc_cb_sync_ack(struct bfa_ioc_s *ioc);
@@ -53,6 +54,7 @@ bfa_ioc_set_cb_hwif(struct bfa_ioc_s *ioc)
 	hwif_cb.ioc_isr_mode_set = bfa_ioc_cb_isr_mode_set;
 	hwif_cb.ioc_notify_fail = bfa_ioc_cb_notify_fail;
 	hwif_cb.ioc_ownership_reset = bfa_ioc_cb_ownership_reset;
+	hwif_cb.ioc_sync_start = bfa_ioc_cb_sync_start;
 	hwif_cb.ioc_sync_join = bfa_ioc_cb_sync_join;
 	hwif_cb.ioc_sync_leave = bfa_ioc_cb_sync_leave;
 	hwif_cb.ioc_sync_ack = bfa_ioc_cb_sync_ack;
@@ -194,6 +196,15 @@ bfa_ioc_cb_isr_mode_set(struct bfa_ioc_s *ioc, bfa_boolean_t msix)
 {
 }
 
+/*
+ * Synchronized IOC failure processing routines
+ */
+static bfa_boolean_t
+bfa_ioc_cb_sync_start(struct bfa_ioc_s *ioc)
+{
+	return bfa_ioc_cb_sync_complete(ioc);
+}
+
 /*
  * Cleanup hw semaphore and usecnt registers
  */

drivers/scsi/bfa/bfa_ioc_ct.c

@@ -41,6 +41,7 @@ static void bfa_ioc_ct_map_port(struct bfa_ioc_s *ioc);
 static void bfa_ioc_ct_isr_mode_set(struct bfa_ioc_s *ioc, bfa_boolean_t msix);
 static void bfa_ioc_ct_notify_fail(struct bfa_ioc_s *ioc);
 static void bfa_ioc_ct_ownership_reset(struct bfa_ioc_s *ioc);
+static bfa_boolean_t bfa_ioc_ct_sync_start(struct bfa_ioc_s *ioc);
 static void bfa_ioc_ct_sync_join(struct bfa_ioc_s *ioc);
 static void bfa_ioc_ct_sync_leave(struct bfa_ioc_s *ioc);
 static void bfa_ioc_ct_sync_ack(struct bfa_ioc_s *ioc);
@@ -62,6 +63,7 @@ bfa_ioc_set_ct_hwif(struct bfa_ioc_s *ioc)
 	hwif_ct.ioc_isr_mode_set = bfa_ioc_ct_isr_mode_set;
 	hwif_ct.ioc_notify_fail = bfa_ioc_ct_notify_fail;
 	hwif_ct.ioc_ownership_reset = bfa_ioc_ct_ownership_reset;
+	hwif_ct.ioc_sync_start = bfa_ioc_ct_sync_start;
 	hwif_ct.ioc_sync_join = bfa_ioc_ct_sync_join;
 	hwif_ct.ioc_sync_leave = bfa_ioc_ct_sync_leave;
 	hwif_ct.ioc_sync_ack = bfa_ioc_ct_sync_ack;
@@ -351,6 +353,30 @@ bfa_ioc_ct_ownership_reset(struct bfa_ioc_s *ioc)
 	writel(1, ioc->ioc_regs.ioc_sem_reg);
 }
 
+static bfa_boolean_t
+bfa_ioc_ct_sync_start(struct bfa_ioc_s *ioc)
+{
+	uint32_t r32 = readl(ioc->ioc_regs.ioc_fail_sync);
+	uint32_t sync_reqd = bfa_ioc_ct_get_sync_reqd(r32);
+
+	/*
+	 * Driver load time.  If the sync required bit for this PCI fn
+	 * is set, it is due to an unclean exit by the driver for this
+	 * PCI fn in the previous incarnation. Whoever comes here first
+	 * should clean it up, no matter which PCI fn.
+	 */
+
+	if (sync_reqd & bfa_ioc_ct_sync_pos(ioc)) {
+		writel(0, ioc->ioc_regs.ioc_fail_sync);
+		writel(1, ioc->ioc_regs.ioc_usage_reg);
+		writel(BFI_IOC_UNINIT, ioc->ioc_regs.ioc_fwstate);
+		writel(BFI_IOC_UNINIT, ioc->ioc_regs.alt_ioc_fwstate);
+		return BFA_TRUE;
+	}
+
+	return bfa_ioc_ct_sync_complete(ioc);
+}
+
 /*
  * Synchronized IOC failure processing routines
  */

drivers/scsi/bnx2i/bnx2i.h

@@ -66,11 +66,11 @@
 #define BD_SPLIT_SIZE			32768
 
 /* min, max & default values for SQ/RQ/CQ size, configurable via' modparam */
-#define BNX2I_SQ_WQES_MIN 		16
-#define BNX2I_570X_SQ_WQES_MAX 		128
-#define BNX2I_5770X_SQ_WQES_MAX 	512
-#define BNX2I_570X_SQ_WQES_DEFAULT 	128
-#define BNX2I_5770X_SQ_WQES_DEFAULT 	256
+#define BNX2I_SQ_WQES_MIN		16
+#define BNX2I_570X_SQ_WQES_MAX		128
+#define BNX2I_5770X_SQ_WQES_MAX		512
+#define BNX2I_570X_SQ_WQES_DEFAULT	128
+#define BNX2I_5770X_SQ_WQES_DEFAULT	128
 
 #define BNX2I_570X_CQ_WQES_MAX 		128
 #define BNX2I_5770X_CQ_WQES_MAX 	512
@@ -115,6 +115,7 @@
 #define BNX2X_MAX_CQS			8
 
 #define CNIC_ARM_CQE			1
+#define CNIC_ARM_CQE_FP			2
 #define CNIC_DISARM_CQE			0
 
 #define REG_RD(__hba, offset)				\
@@ -666,7 +667,9 @@ enum {
  *                      after HBA reset is completed by bnx2i/cnic/bnx2
  *                      modules
  * @state:              tracks offload connection state machine
- * @teardown_mode:      indicates if conn teardown is abortive or orderly
+ * @timestamp:          tracks the start time when the ep begins to connect
+ * @num_active_cmds:    tracks the number of outstanding commands for this ep
+ * @ec_shift:           the amount of shift as part of the event coal calc
  * @qp:                 QP information
  * @ids:                contains chip allocated *context id* & driver assigned
  *                      *iscsi cid*
@@ -685,6 +688,7 @@ struct bnx2i_endpoint {
 	u32 state;
 	unsigned long timestamp;
 	int num_active_cmds;
+	u32 ec_shift;
 
 	struct qp_info qp;
 	struct ep_handles ids;

drivers/scsi/bnx2i/bnx2i_hwi.c

@@ -138,7 +138,6 @@ void bnx2i_arm_cq_event_coalescing(struct bnx2i_endpoint *ep, u8 action)
 	u16 next_index;
 	u32 num_active_cmds;
 
-
 	/* Coalesce CQ entries only on 10G devices */
 	if (!test_bit(BNX2I_NX2_DEV_57710, &ep->hba->cnic_dev_type))
 		return;
@@ -148,16 +147,19 @@ void bnx2i_arm_cq_event_coalescing(struct bnx2i_endpoint *ep, u8 action)
 	 * interrupts and other unwanted results
 	 */
 	cq_db = (struct bnx2i_5771x_cq_db *) ep->qp.cq_pgtbl_virt;
-	if (cq_db->sqn[0] && cq_db->sqn[0] != 0xFFFF)
-		return;
 
-	if (action == CNIC_ARM_CQE) {
+	if (action != CNIC_ARM_CQE_FP)
+		if (cq_db->sqn[0] && cq_db->sqn[0] != 0xFFFF)
+			return;
+
+	if (action == CNIC_ARM_CQE || action == CNIC_ARM_CQE_FP) {
 		num_active_cmds = ep->num_active_cmds;
 		if (num_active_cmds <= event_coal_min)
 			next_index = 1;
 		else
 			next_index = event_coal_min +
-				(num_active_cmds - event_coal_min) / event_coal_div;
+				     ((num_active_cmds - event_coal_min) >>
+				     ep->ec_shift);
 		if (!next_index)
 			next_index = 1;
 		cq_index = ep->qp.cqe_exp_seq_sn + next_index - 1;
@@ -1274,6 +1276,7 @@ int bnx2i_send_fw_iscsi_init_msg(struct bnx2i_hba *hba)
 	iscsi_init.dummy_buffer_addr_hi =
 		(u32) ((u64) hba->dummy_buf_dma >> 32);
 
+	hba->num_ccell = hba->max_sqes >> 1;
 	hba->ctx_ccell_tasks =
 			((hba->num_ccell & 0xFFFF) | (hba->max_sqes << 16));
 	iscsi_init.num_ccells_per_conn = hba->num_ccell;
@@ -1934,7 +1937,6 @@ static void bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 			qp->cq_cons_idx++;
 		}
 	}
-	bnx2i_arm_cq_event_coalescing(bnx2i_conn->ep, CNIC_ARM_CQE);
 }
 
 /**
@@ -1948,22 +1950,23 @@ static void bnx2i_process_new_cqes(struct bnx2i_conn *bnx2i_conn)
 static void bnx2i_fastpath_notification(struct bnx2i_hba *hba,
 					struct iscsi_kcqe *new_cqe_kcqe)
 {
-	struct bnx2i_conn *conn;
+	struct bnx2i_conn *bnx2i_conn;
 	u32 iscsi_cid;
 
 	iscsi_cid = new_cqe_kcqe->iscsi_conn_id;
-	conn = bnx2i_get_conn_from_id(hba, iscsi_cid);
+	bnx2i_conn = bnx2i_get_conn_from_id(hba, iscsi_cid);
 
-	if (!conn) {
+	if (!bnx2i_conn) {
 		printk(KERN_ALERT "cid #%x not valid\n", iscsi_cid);
 		return;
 	}
-	if (!conn->ep) {
+	if (!bnx2i_conn->ep) {
 		printk(KERN_ALERT "cid #%x - ep not bound\n", iscsi_cid);
 		return;
 	}
-
-	bnx2i_process_new_cqes(conn);
+	bnx2i_process_new_cqes(bnx2i_conn);
+	bnx2i_arm_cq_event_coalescing(bnx2i_conn->ep, CNIC_ARM_CQE_FP);
+	bnx2i_process_new_cqes(bnx2i_conn);
 }
 
 

drivers/scsi/bnx2i/bnx2i_init.c

@@ -244,7 +244,7 @@ void bnx2i_stop(void *handle)
 	wait_event_interruptible_timeout(hba->eh_wait,
 					 (list_empty(&hba->ep_ofld_list) &&
 					 list_empty(&hba->ep_destroy_list)),
-					 10 * HZ);
+					 2 * HZ);
 	/* Wait for all endpoints to be torn down, Chip will be reset once
 	 *  control returns to network driver. So it is required to cleanup and
 	 * release all connection resources before returning from this routine.

drivers/scsi/bnx2i/bnx2i_iscsi.c

@@ -379,6 +379,7 @@ static struct iscsi_endpoint *bnx2i_alloc_ep(struct bnx2i_hba *hba)
 {
 	struct iscsi_endpoint *ep;
 	struct bnx2i_endpoint *bnx2i_ep;
+	u32 ec_div;
 
 	ep = iscsi_create_endpoint(sizeof(*bnx2i_ep));
 	if (!ep) {
@@ -393,6 +394,11 @@ static struct iscsi_endpoint *bnx2i_alloc_ep(struct bnx2i_hba *hba)
 	bnx2i_ep->ep_iscsi_cid = (u16) -1;
 	bnx2i_ep->hba = hba;
 	bnx2i_ep->hba_age = hba->age;
+
+	ec_div = event_coal_div;
+	while (ec_div >>= 1)
+		bnx2i_ep->ec_shift += 1;
+
 	hba->ofld_conns_active++;
 	init_waitqueue_head(&bnx2i_ep->ofld_wait);
 	return ep;
@@ -858,7 +864,7 @@ struct bnx2i_hba *bnx2i_alloc_hba(struct cnic_dev *cnic)
 	mutex_init(&hba->net_dev_lock);
 	init_waitqueue_head(&hba->eh_wait);
 	if (test_bit(BNX2I_NX2_DEV_57710, &hba->cnic_dev_type)) {
-		hba->hba_shutdown_tmo = 20 * HZ;
+		hba->hba_shutdown_tmo = 30 * HZ;
 		hba->conn_teardown_tmo = 20 * HZ;
 		hba->conn_ctx_destroy_tmo = 6 * HZ;
 	} else {	/* 5706/5708/5709 */
@@ -1208,6 +1214,9 @@ static int bnx2i_task_xmit(struct iscsi_task *task)
 	struct bnx2i_cmd *cmd = task->dd_data;
 	struct iscsi_cmd *hdr = (struct iscsi_cmd *) task->hdr;
 
+	if (bnx2i_conn->ep->num_active_cmds + 1 > hba->max_sqes)
+		return -ENOMEM;
+
 	/*
 	 * If there is no scsi_cmnd this must be a mgmt task
 	 */
@@ -2156,7 +2165,7 @@ static struct scsi_host_template bnx2i_host_template = {
 	.change_queue_depth	= iscsi_change_queue_depth,
 	.can_queue		= 1024,
 	.max_sectors		= 127,
-	.cmd_per_lun		= 32,
+	.cmd_per_lun		= 24,
 	.this_id		= -1,
 	.use_clustering		= ENABLE_CLUSTERING,
 	.sg_tablesize		= ISCSI_MAX_BDS_PER_CMD,