[SCSI] be2iscsi: Fix Unrecoverable Error Detection

Driver periodically checks adapter state,is up fine or not.
Based on the value updates the internal structures of driver.

Signed-off-by: John Soni Jose <sony.john-n@emulex.com>
Signed-off-by: Jayamohan Kallickal <jayamohan.kallickal@emulex.com>
Reviewed-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

5 files changed, 178 insertions, 2 deletions

diff --git a/drivers/scsi/be2iscsi/be_cmds.c b/drivers/scsi/be2iscsi/be_cmds.c
index 352fc53e91b..5c87768c109 100644
--- a/drivers/scsi/be2iscsi/be_cmds.c
+++ b/drivers/scsi/be2iscsi/be_cmds.c
@@ -157,6 +157,9 @@ int beiscsi_mccq_compl(struct beiscsi_hba *phba,
 	struct be_cmd_req_hdr *ioctl_hdr;
 	struct be_queue_info *mccq = &phba->ctrl.mcc_obj.q;
 
+	if (beiscsi_error(phba))
+		return -EIO;
+
 	/* wait for the mccq completion */
 	rc = wait_event_interruptible_timeout(
 				phba->ctrl.mcc_wait[tag],
@@ -423,7 +426,7 @@ static int be_mcc_wait_compl(struct beiscsi_hba *phba)
 {
 	int i, status;
 	for (i = 0; i < mcc_timeout; i++) {
-		if (phba->fw_timeout)
+		if (beiscsi_error(phba))
 			return -EIO;
 
 		status = beiscsi_process_mcc(phba);
@@ -439,6 +442,7 @@ static int be_mcc_wait_compl(struct beiscsi_hba *phba)
 			    BEISCSI_LOG_CONFIG | BEISCSI_LOG_MBOX,
 			    "BC_%d : FW Timed Out\n");
 		phba->fw_timeout = true;
+		beiscsi_ue_detect(phba);
 		return -EBUSY;
 	}
 	return 0;
@@ -479,7 +483,8 @@ static int be_mbox_db_ready_wait(struct be_ctrl_info *ctrl)
 	u32 ready;
 
 	do {
-		if (phba->fw_timeout)
+
+		if (beiscsi_error(phba))
 			return -EIO;
 
 		ready = ioread32(db) & MPU_MAILBOX_DB_RDY_MASK;
@@ -491,6 +496,7 @@ static int be_mbox_db_ready_wait(struct be_ctrl_info *ctrl)
 				    BEISCSI_LOG_CONFIG | BEISCSI_LOG_MBOX,
 				    "BC_%d : FW Timed Out\n");
 			phba->fw_timeout = true;
+			beiscsi_ue_detect(phba);
 			return -EBUSY;
 		}
 
diff --git a/drivers/scsi/be2iscsi/be_main.c b/drivers/scsi/be2iscsi/be_main.c
index 73a29b2c18b..48d37dded8f 100644
--- a/drivers/scsi/be2iscsi/be_main.c
+++ b/drivers/scsi/be2iscsi/be_main.c
@@ -4717,6 +4717,8 @@ static void beiscsi_quiesce(struct beiscsi_hba *phba)
 			    phba->ctrl.mbox_mem_alloced.size,
 			    phba->ctrl.mbox_mem_alloced.va,
 			    phba->ctrl.mbox_mem_alloced.dma);
+
+	cancel_delayed_work_sync(&phba->beiscsi_hw_check_task);
 }
 
 static void beiscsi_remove(struct pci_dev *pcidev)
@@ -4769,6 +4771,25 @@ static void beiscsi_msix_enable(struct beiscsi_hba *phba)
 	return;
 }
 
+/*
+ * beiscsi_hw_health_check()- Check adapter health
+ * @work: work item to check HW health
+ *
+ * Check if adapter in an unrecoverable state or not.
+ **/
+static void
+beiscsi_hw_health_check(struct work_struct *work)
+{
+	struct beiscsi_hba *phba =
+		container_of(work, struct beiscsi_hba,
+			     beiscsi_hw_check_task.work);
+
+	beiscsi_ue_detect(phba);
+
+	schedule_delayed_work(&phba->beiscsi_hw_check_task,
+			      msecs_to_jiffies(1000));
+}
+
 static int __devinit beiscsi_dev_probe(struct pci_dev *pcidev,
 				const struct pci_device_id *id)
 {
@@ -4892,6 +4913,8 @@ static int __devinit beiscsi_dev_probe(struct pci_dev *pcidev,
 		goto free_twq;
 	}
 
+	INIT_DELAYED_WORK(&phba->beiscsi_hw_check_task,
+			  beiscsi_hw_health_check);
 
 	phwi_ctrlr = phba->phwi_ctrlr;
 	phwi_context = phwi_ctrlr->phwi_ctxt;
@@ -4941,6 +4964,9 @@ static int __devinit beiscsi_dev_probe(struct pci_dev *pcidev,
 			    "iSCSI boot info.\n");
 
 	beiscsi_create_def_ifaces(phba);
+	schedule_delayed_work(&phba->beiscsi_hw_check_task,
+			      msecs_to_jiffies(1000));
+
 	beiscsi_log(phba, KERN_INFO, BEISCSI_LOG_INIT,
 		    "\n\n\n BM_%d : SUCCESS - DRIVER LOADED\n\n\n");
 	return 0;
diff --git a/drivers/scsi/be2iscsi/be_main.h b/drivers/scsi/be2iscsi/be_main.h
index 033c053d947..5b27275cc81 100644
--- a/drivers/scsi/be2iscsi/be_main.h
+++ b/drivers/scsi/be2iscsi/be_main.h
@@ -750,6 +750,11 @@ free_mgmt_sgl_handle(struct beiscsi_hba *phba, struct sgl_handle *psgl_handle);
 
 void beiscsi_process_all_cqs(struct work_struct *work);
 
+static inline bool beiscsi_error(struct beiscsi_hba *phba)
+{
+	return phba->ue_detected || phba->fw_timeout;
+}
+
 struct pdu_nop_out {
 	u32 dw[12];
 };
diff --git a/drivers/scsi/be2iscsi/be_mgmt.c b/drivers/scsi/be2iscsi/be_mgmt.c
index 688bf64741e..a6c2fe4b4d6 100644
--- a/drivers/scsi/be2iscsi/be_mgmt.c
+++ b/drivers/scsi/be2iscsi/be_mgmt.c
@@ -22,6 +22,138 @@
 #include <scsi/scsi_bsg_iscsi.h>
 #include "be_mgmt.h"
 #include "be_iscsi.h"
+#include "be_main.h"
+
+/* UE Status Low CSR */
+static const char * const desc_ue_status_low[] = {
+	"CEV",
+	"CTX",
+	"DBUF",
+	"ERX",
+	"Host",
+	"MPU",
+	"NDMA",
+	"PTC ",
+	"RDMA ",
+	"RXF ",
+	"RXIPS ",
+	"RXULP0 ",
+	"RXULP1 ",
+	"RXULP2 ",
+	"TIM ",
+	"TPOST ",
+	"TPRE ",
+	"TXIPS ",
+	"TXULP0 ",
+	"TXULP1 ",
+	"UC ",
+	"WDMA ",
+	"TXULP2 ",
+	"HOST1 ",
+	"P0_OB_LINK ",
+	"P1_OB_LINK ",
+	"HOST_GPIO ",
+	"MBOX ",
+	"AXGMAC0",
+	"AXGMAC1",
+	"JTAG",
+	"MPU_INTPEND"
+};
+
+/* UE Status High CSR */
+static const char * const desc_ue_status_hi[] = {
+	"LPCMEMHOST",
+	"MGMT_MAC",
+	"PCS0ONLINE",
+	"MPU_IRAM",
+	"PCS1ONLINE",
+	"PCTL0",
+	"PCTL1",
+	"PMEM",
+	"RR",
+	"TXPB",
+	"RXPP",
+	"XAUI",
+	"TXP",
+	"ARM",
+	"IPC",
+	"HOST2",
+	"HOST3",
+	"HOST4",
+	"HOST5",
+	"HOST6",
+	"HOST7",
+	"HOST8",
+	"HOST9",
+	"NETC",
+	"Unknown",
+	"Unknown",
+	"Unknown",
+	"Unknown",
+	"Unknown",
+	"Unknown",
+	"Unknown",
+	"Unknown"
+};
+
+/*
+ * beiscsi_ue_detec()- Detect Unrecoverable Error on adapter
+ * @phba: Driver priv structure
+ *
+ * Read registers linked to UE and check for the UE status
+ **/
+void beiscsi_ue_detect(struct beiscsi_hba *phba)
+{
+	uint32_t ue_hi = 0, ue_lo = 0;
+	uint32_t ue_mask_hi = 0, ue_mask_lo = 0;
+	uint8_t i = 0;
+
+	if (phba->ue_detected)
+		return;
+
+	pci_read_config_dword(phba->pcidev,
+			      PCICFG_UE_STATUS_LOW, &ue_lo);
+	pci_read_config_dword(phba->pcidev,
+			      PCICFG_UE_STATUS_MASK_LOW,
+			      &ue_mask_lo);
+	pci_read_config_dword(phba->pcidev,
+			      PCICFG_UE_STATUS_HIGH,
+			      &ue_hi);
+	pci_read_config_dword(phba->pcidev,
+			      PCICFG_UE_STATUS_MASK_HI,
+			      &ue_mask_hi);
+
+	ue_lo = (ue_lo & ~ue_mask_lo);
+	ue_hi = (ue_hi & ~ue_mask_hi);
+
+
+	if (ue_lo || ue_hi) {
+		phba->ue_detected = true;
+		beiscsi_log(phba, KERN_ERR,
+			    BEISCSI_LOG_CONFIG | BEISCSI_LOG_MBOX,
+			    "BG_%d : Error detected on the adapter\n");
+	}
+
+	if (ue_lo) {
+		for (i = 0; ue_lo; ue_lo >>= 1, i++) {
+			if (ue_lo & 1)
+				beiscsi_log(phba, KERN_ERR,
+					    BEISCSI_LOG_CONFIG,
+					    "BG_%d : UE_LOW %s bit set\n",
+					    desc_ue_status_low[i]);
+		}
+	}
+
+	if (ue_hi) {
+		for (i = 0; ue_hi; ue_hi >>= 1, i++) {
+			if (ue_hi & 1)
+				beiscsi_log(phba, KERN_ERR,
+					    BEISCSI_LOG_CONFIG,
+					    "BG_%d : UE_HIGH %s bit set\n",
+					    desc_ue_status_hi[i]);
+		}
+	}
+}
 
 /**
  * mgmt_reopen_session()- Reopen a session based on reopen_type
diff --git a/drivers/scsi/be2iscsi/be_mgmt.h b/drivers/scsi/be2iscsi/be_mgmt.h
index 88a8ed21d7f..2e4968add79 100644
--- a/drivers/scsi/be2iscsi/be_mgmt.h
+++ b/drivers/scsi/be2iscsi/be_mgmt.h
@@ -30,6 +30,12 @@
 #define IP_V6_LEN	16
 #define IP_V4_LEN	4
 
+/* UE Status and Mask register */
+#define PCICFG_UE_STATUS_LOW            0xA0
+#define PCICFG_UE_STATUS_HIGH           0xA4
+#define PCICFG_UE_STATUS_MASK_LOW       0xA8
+#define PCICFG_UE_STATUS_MASK_HI        0xAC
+
 /**
  * Pseudo amap definition in which each bit of the actual structure is defined
  * as a byte: used to calculate offset/shift/mask of each field
@@ -314,5 +320,6 @@ void beiscsi_offload_cxn_v0(struct beiscsi_offload_params *params,
 
 void beiscsi_offload_cxn_v2(struct beiscsi_offload_params *params,
 			     struct wrb_handle *pwrb_handle);
+void beiscsi_ue_detect(struct beiscsi_hba *phba);
 
 #endif