[SCSI] esas2r: ATTO Technology ExpressSAS 6G SAS/SATA RAID Adapter Driver

This is a new driver for ATTO Technology's ExpressSAS series of hardware RAID
adapters.  It supports the following adapters:

    - ExpressSAS R60F
    - ExpressSAS R680
    - ExpressSAS R608
    - ExpressSAS R644

Signed-off-by: Bradley Grove <bgrove@attotech.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

1 files changed, 1189 insertions, 0 deletions

diff --git a/drivers/scsi/esas2r/esas2r_disc.c b/drivers/scsi/esas2r/esas2r_disc.c
new file mode 100644
index 000000000000..dec6c334ce3e
--- /dev/null
+++ b/drivers/scsi/esas2r/esas2r_disc.c
@@ -0,0 +1,1189 @@
+/*
+ *  linux/drivers/scsi/esas2r/esas2r_disc.c
+ *      esas2r device discovery routines
+ *
+ *  Copyright (c) 2001-2013 ATTO Technology, Inc.
+ *  (mailto:linuxdrivers@attotech.com)
+ */
+/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/
+/*
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; version 2 of the License.
+ *
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *  GNU General Public License for more details.
+ *
+ *  NO WARRANTY
+ *  THE PROGRAM IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OR
+ *  CONDITIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED INCLUDING, WITHOUT
+ *  LIMITATION, ANY WARRANTIES OR CONDITIONS OF TITLE, NON-INFRINGEMENT,
+ *  MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Each Recipient is
+ *  solely responsible for determining the appropriateness of using and
+ *  distributing the Program and assumes all risks associated with its
+ *  exercise of rights under this Agreement, including but not limited to
+ *  the risks and costs of program errors, damage to or loss of data,
+ *  programs or equipment, and unavailability or interruption of operations.
+ *
+ *  DISCLAIMER OF LIABILITY
+ *  NEITHER RECIPIENT NOR ANY CONTRIBUTORS SHALL HAVE ANY LIABILITY FOR ANY
+ *  DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ *  DAMAGES (INCLUDING WITHOUT LIMITATION LOST PROFITS), HOWEVER CAUSED AND
+ *  ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR
+ *  TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+ *  USE OR DISTRIBUTION OF THE PROGRAM OR THE EXERCISE OF ANY RIGHTS GRANTED
+ *  HEREUNDER, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES
+ *
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
+ */
+/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/
+
+#include "esas2r.h"
+
+/* Miscellaneous internal discovery routines */
+static void esas2r_disc_abort(struct esas2r_adapter *a,
+			      struct esas2r_request *rq);
+static bool esas2r_disc_continue(struct esas2r_adapter *a,
+				 struct esas2r_request *rq);
+static void esas2r_disc_fix_curr_requests(struct esas2r_adapter *a);
+static u32 esas2r_disc_get_phys_addr(struct esas2r_sg_context *sgc, u64 *addr);
+static bool esas2r_disc_start_request(struct esas2r_adapter *a,
+				      struct esas2r_request *rq);
+
+/* Internal discovery routines that process the states */
+static bool esas2r_disc_block_dev_scan(struct esas2r_adapter *a,
+				       struct esas2r_request *rq);
+static void esas2r_disc_block_dev_scan_cb(struct esas2r_adapter *a,
+					  struct esas2r_request *rq);
+static bool esas2r_disc_dev_add(struct esas2r_adapter *a,
+				struct esas2r_request *rq);
+static bool esas2r_disc_dev_remove(struct esas2r_adapter *a,
+				   struct esas2r_request *rq);
+static bool esas2r_disc_part_info(struct esas2r_adapter *a,
+				  struct esas2r_request *rq);
+static void esas2r_disc_part_info_cb(struct esas2r_adapter *a,
+				     struct esas2r_request *rq);
+static bool esas2r_disc_passthru_dev_info(struct esas2r_adapter *a,
+					  struct esas2r_request *rq);
+static void esas2r_disc_passthru_dev_info_cb(struct esas2r_adapter *a,
+					     struct esas2r_request *rq);
+static bool esas2r_disc_passthru_dev_addr(struct esas2r_adapter *a,
+					  struct esas2r_request *rq);
+static void esas2r_disc_passthru_dev_addr_cb(struct esas2r_adapter *a,
+					     struct esas2r_request *rq);
+static bool esas2r_disc_raid_grp_info(struct esas2r_adapter *a,
+				      struct esas2r_request *rq);
+static void esas2r_disc_raid_grp_info_cb(struct esas2r_adapter *a,
+					 struct esas2r_request *rq);
+
+void esas2r_disc_initialize(struct esas2r_adapter *a)
+{
+	struct esas2r_sas_nvram *nvr = a->nvram;
+
+	esas2r_trace_enter();
+
+	esas2r_lock_clear_flags(&a->flags, AF_DISC_IN_PROG);
+	esas2r_lock_clear_flags(&a->flags2, AF2_DEV_SCAN);
+	esas2r_lock_clear_flags(&a->flags2, AF2_DEV_CNT_OK);
+
+	a->disc_start_time = jiffies_to_msecs(jiffies);
+	a->disc_wait_time = nvr->dev_wait_time * 1000;
+	a->disc_wait_cnt = nvr->dev_wait_count;
+
+	if (a->disc_wait_cnt > ESAS2R_MAX_TARGETS)
+		a->disc_wait_cnt = ESAS2R_MAX_TARGETS;
+
+	/*
+	 * If we are doing chip reset or power management processing, always
+	 * wait for devices.  use the NVRAM device count if it is greater than
+	 * previously discovered devices.
+	 */
+
+	esas2r_hdebug("starting discovery...");
+
+	a->general_req.interrupt_cx = NULL;
+
+	if (a->flags & (AF_CHPRST_DETECTED | AF_POWER_MGT)) {
+		if (a->prev_dev_cnt == 0) {
+			/* Don't bother waiting if there is nothing to wait
+			 * for.
+			 */
+			a->disc_wait_time = 0;
+		} else {
+			/*
+			 * Set the device wait count to what was previously
+			 * found.  We don't care if the user only configured
+			 * a time because we know the exact count to wait for.
+			 * There is no need to honor the user's wishes to
+			 * always wait the full time.
+			 */
+			a->disc_wait_cnt = a->prev_dev_cnt;
+
+			/*
+			 * bump the minimum wait time to 15 seconds since the
+			 * default is 3 (system boot or the boot driver usually
+			 * buys us more time).
+			 */
+			if (a->disc_wait_time < 15000)
+				a->disc_wait_time = 15000;
+		}
+	}
+
+	esas2r_trace("disc wait count: %d", a->disc_wait_cnt);
+	esas2r_trace("disc wait time: %d", a->disc_wait_time);
+
+	if (a->disc_wait_time == 0)
+		esas2r_disc_check_complete(a);
+
+	esas2r_trace_exit();
+}
+
+void esas2r_disc_start_waiting(struct esas2r_adapter *a)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	if (a->disc_ctx.disc_evt)
+		esas2r_disc_start_port(a);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+}
+
+void esas2r_disc_check_for_work(struct esas2r_adapter *a)
+{
+	struct esas2r_request *rq = &a->general_req;
+
+	/* service any pending interrupts first */
+
+	esas2r_polled_interrupt(a);
+
+	/*
+	 * now, interrupt processing may have queued up a discovery event.  go
+	 * see if we have one to start.  we couldn't start it in the ISR since
+	 * polled discovery would cause a deadlock.
+	 */
+
+	esas2r_disc_start_waiting(a);
+
+	if (rq->interrupt_cx == NULL)
+		return;
+
+	if (rq->req_stat == RS_STARTED
+	    && rq->timeout <= RQ_MAX_TIMEOUT) {
+		/* wait for the current discovery request to complete. */
+		esas2r_wait_request(a, rq);
+
+		if (rq->req_stat == RS_TIMEOUT) {
+			esas2r_disc_abort(a, rq);
+			esas2r_local_reset_adapter(a);
+			return;
+		}
+	}
+
+	if (rq->req_stat == RS_PENDING
+	    || rq->req_stat == RS_STARTED)
+		return;
+
+	esas2r_disc_continue(a, rq);
+}
+
+void esas2r_disc_check_complete(struct esas2r_adapter *a)
+{
+	unsigned long flags;
+
+	esas2r_trace_enter();
+
+	/* check to see if we should be waiting for devices */
+	if (a->disc_wait_time) {
+		u32 currtime = jiffies_to_msecs(jiffies);
+		u32 time = currtime - a->disc_start_time;
+
+		/*
+		 * Wait until the device wait time is exhausted or the device
+		 * wait count is satisfied.
+		 */
+		if (time < a->disc_wait_time
+		    && (esas2r_targ_db_get_tgt_cnt(a) < a->disc_wait_cnt
+			|| a->disc_wait_cnt == 0)) {
+			/* After three seconds of waiting, schedule a scan. */
+			if (time >= 3000
+			    && !(esas2r_lock_set_flags(&a->flags2,
+						       AF2_DEV_SCAN) &
+				 ilog2(AF2_DEV_SCAN))) {
+				spin_lock_irqsave(&a->mem_lock, flags);
+				esas2r_disc_queue_event(a, DCDE_DEV_SCAN);
+				spin_unlock_irqrestore(&a->mem_lock, flags);
+			}
+
+			esas2r_trace_exit();
+			return;
+		}
+
+		/*
+		 * We are done waiting...we think.  Adjust the wait time to
+		 * consume events after the count is met.
+		 */
+		if (!(esas2r_lock_set_flags(&a->flags2, AF2_DEV_CNT_OK)
+		      & ilog2(AF2_DEV_CNT_OK)))
+			a->disc_wait_time = time + 3000;
+
+		/* If we haven't done a full scan yet, do it now. */
+		if (!(esas2r_lock_set_flags(&a->flags2,
+					    AF2_DEV_SCAN) &
+		      ilog2(AF2_DEV_SCAN))) {
+			spin_lock_irqsave(&a->mem_lock, flags);
+			esas2r_disc_queue_event(a, DCDE_DEV_SCAN);
+			spin_unlock_irqrestore(&a->mem_lock, flags);
+
+			esas2r_trace_exit();
+			return;
+		}
+
+		/*
+		 * Now, if there is still time left to consume events, continue
+		 * waiting.
+		 */
+		if (time < a->disc_wait_time) {
+			esas2r_trace_exit();
+			return;
+		}
+	} else {
+		if (!(esas2r_lock_set_flags(&a->flags2,
+					    AF2_DEV_SCAN) &
+		      ilog2(AF2_DEV_SCAN))) {
+			spin_lock_irqsave(&a->mem_lock, flags);
+			esas2r_disc_queue_event(a, DCDE_DEV_SCAN);
+			spin_unlock_irqrestore(&a->mem_lock, flags);
+		}
+	}
+
+	/* We want to stop waiting for devices. */
+	a->disc_wait_time = 0;
+
+	if ((a->flags & AF_DISC_POLLED)
+	    && (a->flags & AF_DISC_IN_PROG)) {
+		/*
+		 * Polled discovery is still pending so continue the active
+		 * discovery until it is done.  At that point, we will stop
+		 * polled discovery and transition to interrupt driven
+		 * discovery.
+		 */
+	} else {
+		/*
+		 * Done waiting for devices.  Note that we get here immediately
+		 * after deferred waiting completes because that is interrupt
+		 * driven; i.e. There is no transition.
+		 */
+		esas2r_disc_fix_curr_requests(a);
+		esas2r_lock_clear_flags(&a->flags, AF_DISC_PENDING);
+
+		/*
+		 * We have deferred target state changes until now because we
+		 * don't want to report any removals (due to the first arrival)
+		 * until the device wait time expires.
+		 */
+		esas2r_lock_set_flags(&a->flags, AF_PORT_CHANGE);
+	}
+
+	esas2r_trace_exit();
+}
+
+void esas2r_disc_queue_event(struct esas2r_adapter *a, u8 disc_evt)
+{
+	struct esas2r_disc_context *dc = &a->disc_ctx;
+
+	esas2r_trace_enter();
+
+	esas2r_trace("disc_event: %d", disc_evt);
+
+	/* Initialize the discovery context */
+	dc->disc_evt |= disc_evt;
+
+	/*
+	 * Don't start discovery before or during polled discovery.  if we did,
+	 * we would have a deadlock if we are in the ISR already.
+	 */
+	if (!(a->flags & (AF_CHPRST_PENDING | AF_DISC_POLLED)))
+		esas2r_disc_start_port(a);
+
+	esas2r_trace_exit();
+}
+
+bool esas2r_disc_start_port(struct esas2r_adapter *a)
+{
+	struct esas2r_request *rq = &a->general_req;
+	struct esas2r_disc_context *dc = &a->disc_ctx;
+	bool ret;
+
+	esas2r_trace_enter();
+
+	if (a->flags & AF_DISC_IN_PROG) {
+		esas2r_trace_exit();
+
+		return false;
+	}
+
+	/* If there is a discovery waiting, process it. */
+	if (dc->disc_evt) {
+		if ((a->flags & AF_DISC_POLLED)
+		    && a->disc_wait_time == 0) {
+			/*
+			 * We are doing polled discovery, but we no longer want
+			 * to wait for devices.  Stop polled discovery and
+			 * transition to interrupt driven discovery.
+			 */
+
+			esas2r_trace_exit();
+
+			return false;
+		}
+	} else {
+		/* Discovery is complete. */
+
+		esas2r_hdebug("disc done");
+
+		esas2r_lock_set_flags(&a->flags, AF_PORT_CHANGE);
+
+		esas2r_trace_exit();
+
+		return false;
+	}
+
+	/* Handle the discovery context */
+	esas2r_trace("disc_evt: %d", dc->disc_evt);
+	esas2r_lock_set_flags(&a->flags, AF_DISC_IN_PROG);
+	dc->flags = 0;
+
+	if (a->flags & AF_DISC_POLLED)
+		dc->flags |= DCF_POLLED;
+
+	rq->interrupt_cx = dc;
+	rq->req_stat = RS_SUCCESS;
+
+	/* Decode the event code */
+	if (dc->disc_evt & DCDE_DEV_SCAN) {
+		dc->disc_evt &= ~DCDE_DEV_SCAN;
+
+		dc->flags |= DCF_DEV_SCAN;
+		dc->state = DCS_BLOCK_DEV_SCAN;
+	} else if (dc->disc_evt & DCDE_DEV_CHANGE) {
+		dc->disc_evt &= ~DCDE_DEV_CHANGE;
+
+		dc->flags |= DCF_DEV_CHANGE;
+		dc->state = DCS_DEV_RMV;
+	}
+
+	/* Continue interrupt driven discovery */
+	if (!(a->flags & AF_DISC_POLLED))
+		ret = esas2r_disc_continue(a, rq);
+	else
+		ret = true;
+
+	esas2r_trace_exit();
+
+	return ret;
+}
+
+static bool esas2r_disc_continue(struct esas2r_adapter *a,
+				 struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+
+	/* Device discovery/removal */
+	while (dc->flags & (DCF_DEV_CHANGE | DCF_DEV_SCAN)) {
+		rslt = false;
+
+		switch (dc->state) {
+		case DCS_DEV_RMV:
+
+			rslt = esas2r_disc_dev_remove(a, rq);
+			break;
+
+		case DCS_DEV_ADD:
+
+			rslt = esas2r_disc_dev_add(a, rq);
+			break;
+
+		case DCS_BLOCK_DEV_SCAN:
+
+			rslt = esas2r_disc_block_dev_scan(a, rq);
+			break;
+
+		case DCS_RAID_GRP_INFO:
+
+			rslt = esas2r_disc_raid_grp_info(a, rq);
+			break;
+
+		case DCS_PART_INFO:
+
+			rslt = esas2r_disc_part_info(a, rq);
+			break;
+
+		case DCS_PT_DEV_INFO:
+
+			rslt = esas2r_disc_passthru_dev_info(a, rq);
+			break;
+		case DCS_PT_DEV_ADDR:
+
+			rslt = esas2r_disc_passthru_dev_addr(a, rq);
+			break;
+		case DCS_DISC_DONE:
+
+			dc->flags &= ~(DCF_DEV_CHANGE | DCF_DEV_SCAN);
+			break;
+
+		default:
+
+			esas2r_bugon();
+			dc->state = DCS_DISC_DONE;
+			break;
+		}
+
+		if (rslt)
+			return true;
+	}
+
+	/* Discovery is done...for now. */
+	rq->interrupt_cx = NULL;
+
+	if (!(a->flags & AF_DISC_PENDING))
+		esas2r_disc_fix_curr_requests(a);
+
+	esas2r_lock_clear_flags(&a->flags, AF_DISC_IN_PROG);
+
+	/* Start the next discovery. */
+	return esas2r_disc_start_port(a);
+}
+
+static bool esas2r_disc_start_request(struct esas2r_adapter *a,
+				      struct esas2r_request *rq)
+{
+	unsigned long flags;
+
+	/* Set the timeout to a minimum value. */
+	if (rq->timeout < ESAS2R_DEFAULT_TMO)
+		rq->timeout = ESAS2R_DEFAULT_TMO;
+
+	/*
+	 * Override the request type to distinguish discovery requests.  If we
+	 * end up deferring the request, esas2r_disc_local_start_request()
+	 * will be called to restart it.
+	 */
+	rq->req_type = RT_DISC_REQ;
+
+	spin_lock_irqsave(&a->queue_lock, flags);
+
+	if (!(a->flags & (AF_CHPRST_PENDING | AF_FLASHING)))
+		esas2r_disc_local_start_request(a, rq);
+	else
+		list_add_tail(&rq->req_list, &a->defer_list);
+
+	spin_unlock_irqrestore(&a->queue_lock, flags);
+
+	return true;
+}
+
+void esas2r_disc_local_start_request(struct esas2r_adapter *a,
+				     struct esas2r_request *rq)
+{
+	esas2r_trace_enter();
+
+	list_add_tail(&rq->req_list, &a->active_list);
+
+	esas2r_start_vda_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return;
+}
+
+static void esas2r_disc_abort(struct esas2r_adapter *a,
+			      struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+
+	esas2r_trace_enter();
+
+	/* abort the current discovery */
+
+	dc->state = DCS_DISC_DONE;
+
+	esas2r_trace_exit();
+}
+
+static bool esas2r_disc_block_dev_scan(struct esas2r_adapter *a,
+				       struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+
+	esas2r_trace_enter();
+
+	esas2r_rq_init_request(rq, a);
+
+	esas2r_build_mgt_req(a,
+			     rq,
+			     VDAMGT_DEV_SCAN,
+			     0,
+			     0,
+			     0,
+			     NULL);
+
+	rq->comp_cb = esas2r_disc_block_dev_scan_cb;
+
+	rq->timeout = 30000;
+	rq->interrupt_cx = dc;
+
+	rslt = esas2r_disc_start_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return rslt;
+}
+
+static void esas2r_disc_block_dev_scan_cb(struct esas2r_adapter *a,
+					  struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	unsigned long flags;
+
+	esas2r_trace_enter();
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	if (rq->req_stat == RS_SUCCESS)
+		dc->scan_gen = rq->func_rsp.mgt_rsp.scan_generation;
+
+	dc->state = DCS_RAID_GRP_INFO;
+	dc->raid_grp_ix = 0;
+
+	esas2r_rq_destroy_request(rq, a);
+
+	/* continue discovery if it's interrupt driven */
+
+	if (!(dc->flags & DCF_POLLED))
+		esas2r_disc_continue(a, rq);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+
+	esas2r_trace_exit();
+}
+
+static bool esas2r_disc_raid_grp_info(struct esas2r_adapter *a,
+				      struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+	struct atto_vda_grp_info *grpinfo;
+
+	esas2r_trace_enter();
+
+	esas2r_trace("raid_group_idx: %d", dc->raid_grp_ix);
+
+	if (dc->raid_grp_ix >= VDA_MAX_RAID_GROUPS) {
+		dc->state = DCS_DISC_DONE;
+
+		esas2r_trace_exit();
+
+		return false;
+	}
+
+	esas2r_rq_init_request(rq, a);
+
+	grpinfo = &rq->vda_rsp_data->mgt_data.data.grp_info;
+
+	memset(grpinfo, 0, sizeof(struct atto_vda_grp_info));
+
+	esas2r_build_mgt_req(a,
+			     rq,
+			     VDAMGT_GRP_INFO,
+			     dc->scan_gen,
+			     0,
+			     sizeof(struct atto_vda_grp_info),
+			     NULL);
+
+	grpinfo->grp_index = dc->raid_grp_ix;
+
+	rq->comp_cb = esas2r_disc_raid_grp_info_cb;
+
+	rq->interrupt_cx = dc;
+
+	rslt = esas2r_disc_start_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return rslt;
+}
+
+static void esas2r_disc_raid_grp_info_cb(struct esas2r_adapter *a,
+					 struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	unsigned long flags;
+	struct atto_vda_grp_info *grpinfo;
+
+	esas2r_trace_enter();
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	if (rq->req_stat == RS_SCAN_GEN) {
+		dc->scan_gen = rq->func_rsp.mgt_rsp.scan_generation;
+		dc->raid_grp_ix = 0;
+		goto done;
+	}
+
+	if (rq->req_stat == RS_SUCCESS) {
+		grpinfo = &rq->vda_rsp_data->mgt_data.data.grp_info;
+
+		if (grpinfo->status != VDA_GRP_STAT_ONLINE
+		    && grpinfo->status != VDA_GRP_STAT_DEGRADED) {
+			/* go to the next group. */
+
+			dc->raid_grp_ix++;
+		} else {
+			memcpy(&dc->raid_grp_name[0],
+			       &grpinfo->grp_name[0],
+			       sizeof(grpinfo->grp_name));
+
+			dc->interleave = le32_to_cpu(grpinfo->interleave);
+			dc->block_size = le32_to_cpu(grpinfo->block_size);
+
+			dc->state = DCS_PART_INFO;
+			dc->part_num = 0;
+		}
+	} else {
+		if (!(rq->req_stat == RS_GRP_INVALID)) {
+			esas2r_log(ESAS2R_LOG_WARN,
+				   "A request for RAID group info failed - "
+				   "returned with %x",
+				   rq->req_stat);
+		}
+
+		dc->dev_ix = 0;
+		dc->state = DCS_PT_DEV_INFO;
+	}
+
+done:
+
+	esas2r_rq_destroy_request(rq, a);
+
+	/* continue discovery if it's interrupt driven */
+
+	if (!(dc->flags & DCF_POLLED))
+		esas2r_disc_continue(a, rq);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+
+	esas2r_trace_exit();
+}
+
+static bool esas2r_disc_part_info(struct esas2r_adapter *a,
+				  struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+	struct atto_vdapart_info *partinfo;
+
+	esas2r_trace_enter();
+
+	esas2r_trace("part_num: %d", dc->part_num);
+
+	if (dc->part_num >= VDA_MAX_PARTITIONS) {
+		dc->state = DCS_RAID_GRP_INFO;
+		dc->raid_grp_ix++;
+
+		esas2r_trace_exit();
+
+		return false;
+	}
+
+	esas2r_rq_init_request(rq, a);
+
+	partinfo = &rq->vda_rsp_data->mgt_data.data.part_info;
+
+	memset(partinfo, 0, sizeof(struct atto_vdapart_info));
+
+	esas2r_build_mgt_req(a,
+			     rq,
+			     VDAMGT_PART_INFO,
+			     dc->scan_gen,
+			     0,
+			     sizeof(struct atto_vdapart_info),
+			     NULL);
+
+	partinfo->part_no = dc->part_num;
+
+	memcpy(&partinfo->grp_name[0],
+	       &dc->raid_grp_name[0],
+	       sizeof(partinfo->grp_name));
+
+	rq->comp_cb = esas2r_disc_part_info_cb;
+
+	rq->interrupt_cx = dc;
+
+	rslt = esas2r_disc_start_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return rslt;
+}
+
+static void esas2r_disc_part_info_cb(struct esas2r_adapter *a,
+				     struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	unsigned long flags;
+	struct atto_vdapart_info *partinfo;
+
+	esas2r_trace_enter();
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	if (rq->req_stat == RS_SCAN_GEN) {
+		dc->scan_gen = rq->func_rsp.mgt_rsp.scan_generation;
+		dc->raid_grp_ix = 0;
+		dc->state = DCS_RAID_GRP_INFO;
+	} else if (rq->req_stat == RS_SUCCESS) {
+		partinfo = &rq->vda_rsp_data->mgt_data.data.part_info;
+
+		dc->part_num = partinfo->part_no;
+
+		dc->curr_virt_id = le16_to_cpu(partinfo->target_id);
+
+		esas2r_targ_db_add_raid(a, dc);
+
+		dc->part_num++;
+	} else {
+		if (!(rq->req_stat == RS_PART_LAST)) {
+			esas2r_log(ESAS2R_LOG_WARN,
+				   "A request for RAID group partition info "
+				   "failed - status:%d", rq->req_stat);
+		}
+
+		dc->state = DCS_RAID_GRP_INFO;
+		dc->raid_grp_ix++;
+	}
+
+	esas2r_rq_destroy_request(rq, a);
+
+	/* continue discovery if it's interrupt driven */
+
+	if (!(dc->flags & DCF_POLLED))
+		esas2r_disc_continue(a, rq);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+
+	esas2r_trace_exit();
+}
+
+static bool esas2r_disc_passthru_dev_info(struct esas2r_adapter *a,
+					  struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+	struct atto_vda_devinfo *devinfo;
+
+	esas2r_trace_enter();
+
+	esas2r_trace("dev_ix: %d", dc->dev_ix);
+
+	esas2r_rq_init_request(rq, a);
+
+	devinfo = &rq->vda_rsp_data->mgt_data.data.dev_info;
+
+	memset(devinfo, 0, sizeof(struct atto_vda_devinfo));
+
+	esas2r_build_mgt_req(a,
+			     rq,
+			     VDAMGT_DEV_PT_INFO,
+			     dc->scan_gen,
+			     dc->dev_ix,
+			     sizeof(struct atto_vda_devinfo),
+			     NULL);
+
+	rq->comp_cb = esas2r_disc_passthru_dev_info_cb;
+
+	rq->interrupt_cx = dc;
+
+	rslt = esas2r_disc_start_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return rslt;
+}
+
+static void esas2r_disc_passthru_dev_info_cb(struct esas2r_adapter *a,
+					     struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	unsigned long flags;
+	struct atto_vda_devinfo *devinfo;
+
+	esas2r_trace_enter();
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	if (rq->req_stat == RS_SCAN_GEN) {
+		dc->scan_gen = rq->func_rsp.mgt_rsp.scan_generation;
+		dc->dev_ix = 0;
+		dc->state = DCS_PT_DEV_INFO;
+	} else if (rq->req_stat == RS_SUCCESS) {
+		devinfo = &rq->vda_rsp_data->mgt_data.data.dev_info;
+
+		dc->dev_ix = le16_to_cpu(rq->func_rsp.mgt_rsp.dev_index);
+
+		dc->curr_virt_id = le16_to_cpu(devinfo->target_id);
+
+		if (le16_to_cpu(devinfo->features) & VDADEVFEAT_PHYS_ID) {
+			dc->curr_phys_id =
+				le16_to_cpu(devinfo->phys_target_id);
+			dc->dev_addr_type = ATTO_GDA_AT_PORT;
+			dc->state = DCS_PT_DEV_ADDR;
+
+			esas2r_trace("curr_virt_id: %d", dc->curr_virt_id);
+			esas2r_trace("curr_phys_id: %d", dc->curr_phys_id);
+		} else {
+			dc->dev_ix++;
+		}
+	} else {
+		if (!(rq->req_stat == RS_DEV_INVALID)) {
+			esas2r_log(ESAS2R_LOG_WARN,
+				   "A request for device information failed - "
+				   "status:%d", rq->req_stat);
+		}
+
+		dc->state = DCS_DISC_DONE;
+	}
+
+	esas2r_rq_destroy_request(rq, a);
+
+	/* continue discovery if it's interrupt driven */
+
+	if (!(dc->flags & DCF_POLLED))
+		esas2r_disc_continue(a, rq);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+
+	esas2r_trace_exit();
+}
+
+static bool esas2r_disc_passthru_dev_addr(struct esas2r_adapter *a,
+					  struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	bool rslt;
+	struct atto_ioctl *hi;
+	struct esas2r_sg_context sgc;
+
+	esas2r_trace_enter();
+
+	esas2r_rq_init_request(rq, a);
+
+	/* format the request. */
+
+	sgc.cur_offset = NULL;
+	sgc.get_phys_addr = (PGETPHYSADDR)esas2r_disc_get_phys_addr;
+	sgc.length = offsetof(struct atto_ioctl, data)
+		     + sizeof(struct atto_hba_get_device_address);
+
+	esas2r_sgc_init(&sgc, a, rq, rq->vrq->ioctl.sge);
+
+	esas2r_build_ioctl_req(a, rq, sgc.length, VDA_IOCTL_HBA);
+
+	if (!esas2r_build_sg_list(a, rq, &sgc)) {
+		esas2r_rq_destroy_request(rq, a);
+
+		esas2r_trace_exit();
+
+		return false;
+	}
+
+	rq->comp_cb = esas2r_disc_passthru_dev_addr_cb;
+
+	rq->interrupt_cx = dc;
+
+	/* format the IOCTL data. */
+
+	hi = (struct atto_ioctl *)a->disc_buffer;
+
+	memset(a->disc_buffer, 0, ESAS2R_DISC_BUF_LEN);
+
+	hi->version = ATTO_VER_GET_DEV_ADDR0;
+	hi->function = ATTO_FUNC_GET_DEV_ADDR;
+	hi->flags = HBAF_TUNNEL;
+
+	hi->data.get_dev_addr.target_id = le32_to_cpu(dc->curr_phys_id);
+	hi->data.get_dev_addr.addr_type = dc->dev_addr_type;
+
+	/* start it up. */
+
+	rslt = esas2r_disc_start_request(a, rq);
+
+	esas2r_trace_exit();
+
+	return rslt;
+}
+
+static void esas2r_disc_passthru_dev_addr_cb(struct esas2r_adapter *a,
+					     struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	struct esas2r_target *t = NULL;
+	unsigned long flags;
+	struct atto_ioctl *hi;
+	u16 addrlen;
+
+	esas2r_trace_enter();
+
+	spin_lock_irqsave(&a->mem_lock, flags);
+
+	hi = (struct atto_ioctl *)a->disc_buffer;
+
+	if (rq->req_stat == RS_SUCCESS
+	    && hi->status == ATTO_STS_SUCCESS) {
+		addrlen = le16_to_cpu(hi->data.get_dev_addr.addr_len);
+
+		if (dc->dev_addr_type == ATTO_GDA_AT_PORT) {
+			if (addrlen == sizeof(u64))
+				memcpy(&dc->sas_addr,
+				       &hi->data.get_dev_addr.address[0],
+				       addrlen);
+			else
+				memset(&dc->sas_addr, 0, sizeof(dc->sas_addr));
+
+			/* Get the unique identifier. */
+			dc->dev_addr_type = ATTO_GDA_AT_UNIQUE;
+
+			goto next_dev_addr;
+		} else {
+			/* Add the pass through target. */
+			if (HIBYTE(addrlen) == 0) {
+				t = esas2r_targ_db_add_pthru(a,
+							     dc,
+							     &hi->data.
+							     get_dev_addr.
+							     address[0],
+							     (u8)hi->data.
+							     get_dev_addr.
+							     addr_len);
+
+				if (t)
+					memcpy(&t->sas_addr, &dc->sas_addr,
+					       sizeof(t->sas_addr));
+			} else {
+				/* getting the back end data failed */
+
+				esas2r_log(ESAS2R_LOG_WARN,
+					   "an error occurred retrieving the "
+					   "back end data (%s:%d)",
+					   __func__,
+					   __LINE__);
+			}
+		}
+	} else {
+		/* getting the back end data failed */
+
+		esas2r_log(ESAS2R_LOG_WARN,
+			   "an error occurred retrieving the back end data - "
+			   "rq->req_stat:%d hi->status:%d",
+			   rq->req_stat, hi->status);
+	}
+
+	/* proceed to the next device. */
+
+	if (dc->flags & DCF_DEV_SCAN) {
+		dc->dev_ix++;
+		dc->state = DCS_PT_DEV_INFO;
+	} else if (dc->flags & DCF_DEV_CHANGE) {
+		dc->curr_targ++;
+		dc->state = DCS_DEV_ADD;
+	} else {
+		esas2r_bugon();
+	}
+
+next_dev_addr:
+	esas2r_rq_destroy_request(rq, a);
+
+	/* continue discovery if it's interrupt driven */
+
+	if (!(dc->flags & DCF_POLLED))
+		esas2r_disc_continue(a, rq);
+
+	spin_unlock_irqrestore(&a->mem_lock, flags);
+
+	esas2r_trace_exit();
+}
+
+static u32 esas2r_disc_get_phys_addr(struct esas2r_sg_context *sgc, u64 *addr)
+{
+	struct esas2r_adapter *a = sgc->adapter;
+
+	if (sgc->length > ESAS2R_DISC_BUF_LEN)
+		esas2r_bugon();
+
+	*addr = a->uncached_phys
+		+ (u64)((u8 *)a->disc_buffer - a->uncached);
+
+	return sgc->length;
+}
+
+static bool esas2r_disc_dev_remove(struct esas2r_adapter *a,
+				   struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	struct esas2r_target *t;
+	struct esas2r_target *t2;
+
+	esas2r_trace_enter();
+
+	/* process removals. */
+
+	for (t = a->targetdb; t < a->targetdb_end; t++) {
+		if (t->new_target_state != TS_NOT_PRESENT)
+			continue;
+
+		t->new_target_state = TS_INVALID;
+
+		/* remove the right target! */
+
+		t2 =
+			esas2r_targ_db_find_by_virt_id(a,
+						       esas2r_targ_get_id(t,
+									  a));
+
+		if (t2)
+			esas2r_targ_db_remove(a, t2);
+	}
+
+	/* removals complete.  process arrivals. */
+
+	dc->state = DCS_DEV_ADD;
+	dc->curr_targ = a->targetdb;
+
+	esas2r_trace_exit();
+
+	return false;
+}
+
+static bool esas2r_disc_dev_add(struct esas2r_adapter *a,
+				struct esas2r_request *rq)
+{
+	struct esas2r_disc_context *dc =
+		(struct esas2r_disc_context *)rq->interrupt_cx;
+	struct esas2r_target *t = dc->curr_targ;
+
+	if (t >= a->targetdb_end) {
+		/* done processing state changes. */
+
+		dc->state = DCS_DISC_DONE;
+	} else if (t->new_target_state == TS_PRESENT) {
+		struct atto_vda_ae_lu *luevt = &t->lu_event;
+
+		esas2r_trace_enter();
+
+		/* clear this now in case more events come in. */
+
+		t->new_target_state = TS_INVALID;
+
+		/* setup the discovery context for adding this device. */
+
+		dc->curr_virt_id = esas2r_targ_get_id(t, a);
+
+		if ((luevt->hdr.bylength >= offsetof(struct atto_vda_ae_lu, id)
+		     + sizeof(struct atto_vda_ae_lu_tgt_lun_raid))
+		    && !(luevt->dwevent & VDAAE_LU_PASSTHROUGH)) {
+			dc->block_size = luevt->id.tgtlun_raid.dwblock_size;
+			dc->interleave = luevt->id.tgtlun_raid.dwinterleave;
+		} else {
+			dc->block_size = 0;
+			dc->interleave = 0;
+		}
+
+		/* determine the device type being added. */
+
+		if (luevt->dwevent & VDAAE_LU_PASSTHROUGH) {
+			if (luevt->dwevent & VDAAE_LU_PHYS_ID) {
+				dc->state = DCS_PT_DEV_ADDR;
+				dc->dev_addr_type = ATTO_GDA_AT_PORT;
+				dc->curr_phys_id = luevt->wphys_target_id;
+			} else {
+				esas2r_log(ESAS2R_LOG_WARN,
+					   "luevt->dwevent does not have the "
+					   "VDAAE_LU_PHYS_ID bit set (%s:%d)",
+					   __func__, __LINE__);
+			}
+		} else {
+			dc->raid_grp_name[0] = 0;
+
+			esas2r_targ_db_add_raid(a, dc);
+		}
+
+		esas2r_trace("curr_virt_id: %d", dc->curr_virt_id);
+		esas2r_trace("curr_phys_id: %d", dc->curr_phys_id);
+		esas2r_trace("dwevent: %d", luevt->dwevent);
+
+		esas2r_trace_exit();
+	}
+
+	if (dc->state == DCS_DEV_ADD) {
+		/* go to the next device. */
+
+		dc->curr_targ++;
+	}
+
+	return false;
+}
+
+/*
+ * When discovery is done, find all requests on defer queue and
+ * test if they need to be modified. If a target is no longer present
+ * then complete the request with RS_SEL. Otherwise, update the
+ * target_id since after a hibernate it can be a different value.
+ * VDA does not make passthrough target IDs persistent.
+ */
+static void esas2r_disc_fix_curr_requests(struct esas2r_adapter *a)
+{
+	unsigned long flags;
+	struct esas2r_target *t;
+	struct esas2r_request *rq;
+	struct list_head *element;
+
+	/* update virt_targ_id in any outstanding esas2r_requests  */
+
+	spin_lock_irqsave(&a->queue_lock, flags);
+
+	list_for_each(element, &a->defer_list) {
+		rq = list_entry(element, struct esas2r_request, req_list);
+		if (rq->vrq->scsi.function == VDA_FUNC_SCSI) {
+			t = a->targetdb + rq->target_id;
+
+			if (t->target_state == TS_PRESENT)
+				rq->vrq->scsi.target_id = le16_to_cpu(
+					t->virt_targ_id);
+			else
+				rq->req_stat = RS_SEL;
+		}
+
+	}
+
+	spin_unlock_irqrestore(&a->queue_lock, flags);
+}