scsi: smartpqi: make ioaccel references consistent
authorKevin Barnett <kevin.barnett@hpe.com>
Wed, 3 May 2017 23:55:25 +0000 (18:55 -0500)
committerMartin K. Petersen <martin.petersen@oracle.com>
Tue, 13 Jun 2017 00:48:04 +0000 (20:48 -0400)
 - make all references to RAID bypass consistent throughout driver.

Reviewed-by: Scott Benesh <scott.benesh@microsemi.com>
Signed-off-by: Kevin Barnett <kevin.barnett@microsemi.com>
Signed-off-by: Don Brace <don.brace@microsemi.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
drivers/scsi/smartpqi/smartpqi.h
drivers/scsi/smartpqi/smartpqi_init.c

index 2ed15cfd01d9601b5dafc92dfc0029467b4e254f..e9f113a0f4537f9d5f9c81f43744dab760d72315 100644 (file)
@@ -805,12 +805,11 @@ struct pqi_scsi_dev {
        u8      bay;
        u8      box[8];
        u16     phys_connector[8];
-       int     offload_configured;     /* I/O accel RAID offload configured */
-       int     offload_enabled;        /* I/O accel RAID offload enabled */
-       int     offload_enabled_pending;
-       int     offload_to_mirror;      /* Send next I/O accelerator RAID */
-                                       /* offload request to mirror drive. */
-       struct raid_map *raid_map;      /* I/O accelerator RAID map */
+       bool    raid_bypass_configured; /* RAID bypass configured */
+       bool    raid_bypass_enabled;    /* RAID bypass enabled */
+       int     offload_to_mirror;      /* Send next RAID bypass request */
+                                       /* to mirror drive. */
+       struct raid_map *raid_map;      /* RAID bypass map */
 
        struct pqi_sas_port *sas_port;
        struct scsi_device *sdev;
@@ -827,7 +826,7 @@ struct pqi_scsi_dev {
 #define SCSI_VPD_SUPPORTED_PAGES       0x0     /* standard page */
 #define SCSI_VPD_DEVICE_ID             0x83    /* standard page */
 #define CISS_VPD_LV_DEVICE_GEOMETRY    0xc1    /* vendor-specific page */
-#define CISS_VPD_LV_OFFLOAD_STATUS     0xc2    /* vendor-specific page */
+#define CISS_VPD_LV_BYPASS_STATUS      0xc2    /* vendor-specific page */
 #define CISS_VPD_LV_STATUS             0xc3    /* vendor-specific page */
 
 #define VPD_PAGE       (1 << 8)
index 4e016dfff9728eea21f1a8c2be7acebb472625b1..5f1c607905bf1405dff3aa93f9face40dd6aa474 100644 (file)
@@ -1112,35 +1112,33 @@ error:
        return rc;
 }
 
-static void pqi_get_offload_status(struct pqi_ctrl_info *ctrl_info,
+static void pqi_get_raid_bypass_status(struct pqi_ctrl_info *ctrl_info,
        struct pqi_scsi_dev *device)
 {
        int rc;
        u8 *buffer;
-       u8 offload_status;
+       u8 bypass_status;
 
        buffer = kmalloc(64, GFP_KERNEL);
        if (!buffer)
                return;
 
        rc = pqi_scsi_inquiry(ctrl_info, device->scsi3addr,
-               VPD_PAGE | CISS_VPD_LV_OFFLOAD_STATUS, buffer, 64);
+               VPD_PAGE | CISS_VPD_LV_BYPASS_STATUS, buffer, 64);
        if (rc)
                goto out;
 
-#define OFFLOAD_STATUS_BYTE    4
-#define OFFLOAD_CONFIGURED_BIT 0x1
-#define OFFLOAD_ENABLED_BIT    0x2
+#define RAID_BYPASS_STATUS     4
+#define RAID_BYPASS_CONFIGURED 0x1
+#define RAID_BYPASS_ENABLED    0x2
 
-       offload_status = buffer[OFFLOAD_STATUS_BYTE];
-       device->offload_configured =
-               !!(offload_status & OFFLOAD_CONFIGURED_BIT);
-       if (device->offload_configured) {
-               device->offload_enabled_pending =
-                       !!(offload_status & OFFLOAD_ENABLED_BIT);
-               if (pqi_get_raid_map(ctrl_info, device))
-                       device->offload_enabled_pending = false;
-       }
+       bypass_status = buffer[RAID_BYPASS_STATUS];
+       device->raid_bypass_configured =
+               (bypass_status & RAID_BYPASS_CONFIGURED) != 0;
+       if (device->raid_bypass_configured &&
+               (bypass_status & RAID_BYPASS_ENABLED) &&
+               pqi_get_raid_map(ctrl_info, device) == 0)
+               device->raid_bypass_enabled = true;
 
 out:
        kfree(buffer);
@@ -1214,7 +1212,7 @@ static int pqi_get_device_info(struct pqi_ctrl_info *ctrl_info,
                        device->volume_offline = false;
                } else {
                        pqi_get_raid_level(ctrl_info, device);
-                       pqi_get_offload_status(ctrl_info, device);
+                       pqi_get_raid_bypass_status(ctrl_info, device);
                        pqi_get_volume_status(ctrl_info, device);
                }
        }
@@ -1492,9 +1490,8 @@ static void pqi_dev_info(struct pqi_ctrl_info *ctrl_info,
                        count += snprintf(buffer + count,
                                PQI_DEV_INFO_BUFFER_LENGTH - count,
                                "SSDSmartPathCap%c En%c %-12s",
-                               device->offload_configured ? '+' : '-',
-                               (device->offload_enabled ||
-                               device->offload_enabled_pending) ? '+' : '-',
+                               device->raid_bypass_configured ? '+' : '-',
+                               device->raid_bypass_enabled ? '+' : '-',
                                pqi_raid_level_to_string(device->raid_level));
        } else {
                count += snprintf(buffer + count,
@@ -1546,13 +1543,13 @@ static void pqi_scsi_update_device(struct pqi_scsi_dev *existing_device,
                sizeof(existing_device->box));
        memcpy(existing_device->phys_connector, new_device->phys_connector,
                sizeof(existing_device->phys_connector));
-       existing_device->offload_configured = new_device->offload_configured;
-       existing_device->offload_enabled = false;
-       existing_device->offload_enabled_pending =
-               new_device->offload_enabled_pending;
        existing_device->offload_to_mirror = 0;
        kfree(existing_device->raid_map);
        existing_device->raid_map = new_device->raid_map;
+       existing_device->raid_bypass_configured =
+               new_device->raid_bypass_configured;
+       existing_device->raid_bypass_enabled =
+               new_device->raid_bypass_enabled;
 
        /* To prevent this from being freed later. */
        new_device->raid_map = NULL;
@@ -1670,11 +1667,6 @@ static void pqi_update_device_list(struct pqi_ctrl_info *ctrl_info,
                device->keep_device = true;
        }
 
-       list_for_each_entry(device, &ctrl_info->scsi_device_list,
-               scsi_device_list_entry)
-               device->offload_enabled =
-                       device->offload_enabled_pending;
-
        spin_unlock_irqrestore(&ctrl_info->scsi_device_list_lock, flags);
 
        /* Remove all devices that have gone away. */
@@ -2044,7 +2036,7 @@ static inline void pqi_set_encryption_info(
 }
 
 /*
- * Attempt to perform offload RAID mapping for a logical volume I/O.
+ * Attempt to perform RAID bypass mapping for a logical volume I/O.
  */
 
 #define PQI_RAID_BYPASS_INELIGIBLE     1
@@ -2448,7 +2440,7 @@ static inline void pqi_aio_path_disabled(struct pqi_io_request *io_request)
        struct pqi_scsi_dev *device;
 
        device = io_request->scmd->device->hostdata;
-       device->offload_enabled = false;
+       device->raid_bypass_enabled = false;
        device->aio_enabled = false;
 }
 
@@ -5002,7 +4994,7 @@ static int pqi_scsi_queue_command(struct Scsi_Host *shost,
 
        if (pqi_is_logical_device(device)) {
                raid_bypassed = false;
-               if (device->offload_enabled &&
+               if (device->raid_bypass_enabled &&
                                !blk_rq_is_passthrough(scmd->request)) {
                        rc = pqi_raid_bypass_submit_scsi_cmd(ctrl_info, device,
                                scmd, queue_group);
@@ -5753,7 +5745,7 @@ static ssize_t pqi_ssd_smart_path_enabled_show(struct device *dev,
        spin_lock_irqsave(&ctrl_info->scsi_device_list_lock, flags);
 
        device = sdev->hostdata;
-       buffer[0] = device->offload_enabled ? '1' : '0';
+       buffer[0] = device->raid_bypass_enabled ? '1' : '0';
        buffer[1] = '\n';
        buffer[2] = '\0';