scsi: hisi_sas: Release resource directly in hisi_sas_abort_task() when NCQ error

driver inclusion
category: bugfix
bugzilla: https://gitee.com/openeuler/kernel/issues/I5SXSB
CVE: NA

------------------------------------------------

When the port is detached, EH will clear ATA_EH_RESET in ehc->i.action when
call ata_eh_reset(), and device reset won't be executed.

As the disk won't return other I/Os normally after NCQ Error. In addition,
the abort operation is added, then resource release is safe, so release NCQ
command lldd resource directly in hisi_sas_abort_task() when NCQ error
without soft reset to make sure read log command can be executed success
later. But Soft reset still need to be used in other scenario.
Signed-off-by: NXingui Yang <yangxingui@huawei.com>
Reviewed-by: Nkang fenglong <kangfenglong@huawei.com>
Signed-off-by: NYongqiang Liu <liuyongqiang13@huawei.com>

drivers/scsi/hisi_sas/hisi_sas.h

@@ -101,6 +101,7 @@ enum {
 enum dev_status {
 	HISI_SAS_DEV_INIT,
 	HISI_SAS_DEV_NORMAL,
+	HISI_SAS_DEV_NCQ_ERR,
 };
 
 enum {

drivers/scsi/hisi_sas/hisi_sas_main.c

@@ -1673,7 +1673,7 @@ static int hisi_sas_abort_task(struct sas_task *task)
 	task->task_state_flags |= SAS_TASK_STATE_ABORTED;
 	spin_unlock_irqrestore(&task->task_state_lock, flags);
 
-	if (task->lldd_task && task->task_proto & SAS_PROTOCOL_SSP) {
+	if (slot && task->task_proto & SAS_PROTOCOL_SSP) {
 		struct scsi_cmnd *cmnd = task->uldd_task;
 		u32 tag = slot->idx;
 		int rc2;
@@ -1706,8 +1706,7 @@ static int hisi_sas_abort_task(struct sas_task *task)
 	} else if (task->task_proto & SAS_PROTOCOL_SATA ||
 		task->task_proto & SAS_PROTOCOL_STP) {
 		if (task->dev->dev_type == SAS_SATA_DEV) {
-			struct ata_port *ap = device->sata_dev.ap;
-			struct ata_link *link = &ap->link;
+			struct ata_queued_cmd *qc = task->uldd_task;
 
 			rc = hisi_sas_internal_task_abort(hisi_hba, device,
 						HISI_SAS_INT_ABT_DEV, 0);
@@ -1717,14 +1716,19 @@ static int hisi_sas_abort_task(struct sas_task *task)
 			}
 			hisi_sas_dereg_device(hisi_hba, device);
 
-			if (link->eh_info.action & ATA_EH_RESET) {
-				slot->task = NULL;
+			/*
+			 * If an ATA internal command times out in ATA EH, it
+			 * need to execute soft reset, so check the scsicmd
+			 */
+			if ((sas_dev->dev_status == HISI_SAS_DEV_NCQ_ERR) &&
+			    qc && qc->scsicmd) {
+				hisi_sas_do_release_task(hisi_hba, task, slot);
 				rc = TMF_RESP_FUNC_COMPLETE;
 			} else {
 				rc = hisi_sas_softreset_ata_disk(device);
 			}
 		}
-	} else if (task->lldd_task && task->task_proto & SAS_PROTOCOL_SMP) {
+	} else if (slot && task->task_proto & SAS_PROTOCOL_SMP) {
 		/* SMP */
 		u32 tag = slot->idx;
 		struct hisi_sas_cq *cq = &hisi_hba->cq[slot->dlvry_queue];
@@ -1843,8 +1847,12 @@ static int hisi_sas_I_T_nexus_reset(struct domain_device *device)
 {
 	struct hisi_hba *hisi_hba = dev_to_hisi_hba(device);
 	struct device *dev = hisi_hba->dev;
+	struct hisi_sas_device *sas_dev = device->lldd_dev;
 	int rc;
 
+	if (sas_dev->dev_status == HISI_SAS_DEV_NCQ_ERR)
+		sas_dev->dev_status = HISI_SAS_DEV_NORMAL;
+
 	rc = hisi_sas_internal_task_abort(hisi_hba, device,
 					HISI_SAS_INT_ABT_DEV, 0);
 	if (rc < 0) {

drivers/scsi/hisi_sas/hisi_sas_v3_hw.c

@@ -398,6 +398,8 @@
 #define CMPLT_HDR_CMPLT_MSK		(0x3 << CMPLT_HDR_CMPLT_OFF)
 #define CMPLT_HDR_ERROR_PHASE_OFF   2
 #define CMPLT_HDR_ERROR_PHASE_MSK   (0xff << CMPLT_HDR_ERROR_PHASE_OFF)
+/* bit[9:2] Error Phase */
+#define ERR_PHASE_RESPONSE_FRAME_REV_STAGE	BIT(8)
 #define CMPLT_HDR_RSPNS_XFRD_OFF	10
 #define CMPLT_HDR_RSPNS_XFRD_MSK	(0x1 << CMPLT_HDR_RSPNS_XFRD_OFF)
 #define CMPLT_HDR_RSPNS_GOOD_OFF	11
@@ -423,6 +425,10 @@
 #define CMPLT_HDR_SATA_DISK_ERR_MSK	(0x1 << COMLT_HDR_SATA_DISK_ERR_OFF)
 #define CMPLT_HDR_IO_IN_TARGET_OFF	17
 #define CMPLT_HDR_IO_IN_TARGET_MSK	(0x1 << CMPLT_HDR_IO_IN_TARGET_OFF)
+/* bit[23:18] ERR_FIS_ATA_STATUS */
+#define FIS_ATA_STATUS_ERR		BIT(18)
+/* bit[31:24] ERR_FIS_TYPE */
+#define FIS_TYPE_SDB			BIT(31)
 
 /* ITCT header */
 /* qw0 */
@@ -2247,6 +2253,18 @@ static void hisi_sas_set_sense_data(struct sas_task *task,
 	}
 }
 
+static bool is_ncq_err(struct hisi_sas_complete_v3_hdr *complete_hdr)
+{
+	u32 dw0, dw3;
+
+	dw0 = le32_to_cpu(complete_hdr->dw0);
+	dw3 = le32_to_cpu(complete_hdr->dw3);
+
+	return (dw0 & ERR_PHASE_RESPONSE_FRAME_REV_STAGE) &&
+		(dw3 & FIS_TYPE_SDB) &&
+		(dw3 & FIS_ATA_STATUS_ERR);
+}
+
 static void
 slot_err_v3_hw(struct hisi_hba *hisi_hba, struct sas_task *task,
 	       struct hisi_sas_slot *slot)
@@ -2582,6 +2600,10 @@ static void cq_tasklet_v3_hw(unsigned long val)
 			dev_err(dev, "erroneous completion disk err dev id=%d sas_addr=0x%llx CQ hdr: 0x%x 0x%x 0x%x 0x%x\n",
 				device_id, itct->sas_addr, dw0, dw1,
 				complete_hdr->act, dw3);
+
+			if (is_ncq_err(complete_hdr))
+				sas_dev->dev_status = HISI_SAS_DEV_NCQ_ERR;
+
 			link->eh_info.err_mask |= AC_ERR_DEV;
 			link->eh_info.action |= ATA_EH_RESET;
 			ata_link_abort(link);