diff mbox series

[v2,4/6] mpi3mr: Improve Shutdown times when firmware has faulted

Message ID 20231214205900.270488-5-ranjan.kumar@broadcom.com (mailing list archive)
State Changes Requested
Headers show
Series mpi3mr: Support PCIe Error Recovery | expand

Commit Message

Ranjan Kumar Dec. 14, 2023, 8:58 p.m. UTC 
The driver monitors the controller state periodically while
waiting for the shutdown notification MPI request to complete.
If the firmware is faulty, the driver resets the controller and
re-issues the shutdown notification. The driver will make three
attempts to complete the shutdown process and will not retry the
notification request if the controller reset is unsuccessful.

Signed-off-by: Prayas Patel <prayas.patel@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
---
 drivers/scsi/mpi3mr/mpi3mr.h    |  1 +
 drivers/scsi/mpi3mr/mpi3mr_fw.c | 36 +++++++++++++++++++++++++--------
 2 files changed, 29 insertions(+), 8 deletions(-)
diff mbox series

Patch

diff --git a/drivers/scsi/mpi3mr/mpi3mr.h b/drivers/scsi/mpi3mr/mpi3mr.h
index 1ac7f88dc1cd..de953eb055d0 100644
--- a/drivers/scsi/mpi3mr/mpi3mr.h
+++ b/drivers/scsi/mpi3mr/mpi3mr.h
@@ -159,6 +159,7 @@  extern atomic64_t event_counter;
 /* Controller Reset related definitions */
 #define MPI3MR_HOSTDIAG_UNLOCK_RETRY_COUNT	5
 #define MPI3MR_MAX_RESET_RETRY_COUNT		3
+#define MPI3MR_MAX_SHUTDOWN_RETRY_COUNT		2
 
 /* ResponseCode definitions */
 #define MPI3MR_RI_MASK_RESPCODE		(0x000000FF)
diff --git a/drivers/scsi/mpi3mr/mpi3mr_fw.c b/drivers/scsi/mpi3mr/mpi3mr_fw.c
index c50fc27231cd..491ef854fdba 100644
--- a/drivers/scsi/mpi3mr/mpi3mr_fw.c
+++ b/drivers/scsi/mpi3mr/mpi3mr_fw.c
@@ -4566,9 +4566,10 @@  void mpi3mr_free_mem(struct mpi3mr_ioc *mrioc)
  */
 static void mpi3mr_issue_ioc_shutdown(struct mpi3mr_ioc *mrioc)
 {
-	u32 ioc_config, ioc_status;
-	u8 retval = 1;
+	u32 ioc_config, ioc_status, shutdown_action;
+	u8 retval = 1, retry = 0;
 	u32 timeout = MPI3MR_DEFAULT_SHUTDOWN_TIME * 10;
+	u32 timeout_remaining = 0;
 
 	ioc_info(mrioc, "Issuing shutdown Notification\n");
 	if (mrioc->unrecoverable) {
@@ -4583,15 +4584,16 @@  static void mpi3mr_issue_ioc_shutdown(struct mpi3mr_ioc *mrioc)
 		return;
 	}
 
+	shutdown_action = MPI3_SYSIF_IOC_CONFIG_SHUTDOWN_NORMAL |
+	    MPI3_SYSIF_IOC_CONFIG_DEVICE_SHUTDOWN_SEND_REQ;
 	ioc_config = readl(&mrioc->sysif_regs->ioc_configuration);
-	ioc_config |= MPI3_SYSIF_IOC_CONFIG_SHUTDOWN_NORMAL;
-	ioc_config |= MPI3_SYSIF_IOC_CONFIG_DEVICE_SHUTDOWN_SEND_REQ;
+	ioc_config |= shutdown_action;
 
 	writel(ioc_config, &mrioc->sysif_regs->ioc_configuration);
 
 	if (mrioc->facts.shutdown_timeout)
 		timeout = mrioc->facts.shutdown_timeout * 10;
-
+	timeout_remaining = timeout;
 	do {
 		ioc_status = readl(&mrioc->sysif_regs->ioc_status);
 		if ((ioc_status & MPI3_SYSIF_IOC_STATUS_SHUTDOWN_MASK)
@@ -4599,8 +4601,26 @@  static void mpi3mr_issue_ioc_shutdown(struct mpi3mr_ioc *mrioc)
 			retval = 0;
 			break;
 		}
+		if (mrioc->unrecoverable)
+			break;
+		if (ioc_status & MPI3_SYSIF_IOC_STATUS_FAULT) {
+			mpi3mr_print_fault_info(mrioc);
+			if (retry >= MPI3MR_MAX_SHUTDOWN_RETRY_COUNT)
+				break;
+			if (mpi3mr_issue_reset(mrioc,
+			    MPI3_SYSIF_HOST_DIAG_RESET_ACTION_SOFT_RESET,
+			    MPI3MR_RESET_FROM_CTLR_CLEANUP))
+				break;
+			ioc_config =
+			    readl(&mrioc->sysif_regs->ioc_configuration);
+			ioc_config |= shutdown_action;
+			writel(ioc_config,
+			    &mrioc->sysif_regs->ioc_configuration);
+			timeout_remaining = timeout;
+			retry++;
+		}
 		msleep(100);
-	} while (--timeout);
+	} while (--timeout_remaining);
 
 	ioc_status = readl(&mrioc->sysif_regs->ioc_status);
 	ioc_config = readl(&mrioc->sysif_regs->ioc_configuration);
@@ -4609,11 +4629,11 @@  static void mpi3mr_issue_ioc_shutdown(struct mpi3mr_ioc *mrioc)
 		if ((ioc_status & MPI3_SYSIF_IOC_STATUS_SHUTDOWN_MASK)
 		    == MPI3_SYSIF_IOC_STATUS_SHUTDOWN_IN_PROGRESS)
 			ioc_warn(mrioc,
-			    "shutdown still in progress after timeout\n");
+			    "shutdown still in progress\n");
 	}
 
 	ioc_info(mrioc,
-	    "Base IOC Sts/Config after %s shutdown is (0x%x)/(0x%x)\n",
+	    "ioc_status/ioc_config after %s shutdown is (0x%x)/(0x%x)\n",
 	    (!retval) ? "successful" : "failed", ioc_status,
 	    ioc_config);
 }