diff --git a/drivers/scsi/hisi_sas/hisi_sas_main.c b/drivers/scsi/hisi_sas/hisi_sas_main.c
index 3596414d970b..7a484ad0f9ab 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_main.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_main.c
@@ -935,8 +935,28 @@ static void hisi_sas_phyup_work_common(struct work_struct *work,
 		container_of(work, typeof(*phy), works[event]);
 	struct hisi_hba *hisi_hba = phy->hisi_hba;
 	struct asd_sas_phy *sas_phy = &phy->sas_phy;
+	struct asd_sas_port *sas_port = sas_phy->port;
+	struct hisi_sas_port *port = phy->port;
+	struct device *dev = hisi_hba->dev;
+	struct domain_device *port_dev;
 	int phy_no = sas_phy->id;
 
+	if (!test_bit(HISI_SAS_RESETTING_BIT, &hisi_hba->flags) &&
+	    sas_port && port && (port->id != phy->port_id)) {
+		dev_info(dev, "phy%d's hw port id changed from %d to %llu\n",
+				phy_no, port->id, phy->port_id);
+		port_dev = sas_port->port_dev;
+		if (port_dev && !dev_is_expander(port_dev->dev_type)) {
+			/*
+			 * Set the device state to gone to block
+			 * sending IO to the device.
+			 */
+			set_bit(SAS_DEV_GONE, &port_dev->state);
+			hisi_sas_notify_phy_event(phy, HISI_PHYE_LINK_RESET);
+			return;
+		}
+	}
+
 	phy->wait_phyup_cnt = 0;
 	if (phy->identify.target_port_protocols == SAS_PROTOCOL_SSP)
 		hisi_hba->hw->sl_notify_ssp(hisi_hba, phy_no);
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
index 04ee02797ca3..be2c0b628595 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v2_hw.c
@@ -2501,6 +2501,7 @@ static void prep_ata_v2_hw(struct hisi_hba *hisi_hba,
 	struct hisi_sas_port *port = to_hisi_sas_port(sas_port);
 	struct sas_ata_task *ata_task = &task->ata_task;
 	struct sas_tmf_task *tmf = slot->tmf;
+	int phy_id;
 	u8 *buf_cmd;
 	int has_data = 0, hdr_tag = 0;
 	u32 dw0, dw1 = 0, dw2 = 0;
@@ -2508,10 +2509,14 @@ static void prep_ata_v2_hw(struct hisi_hba *hisi_hba,
 	/* create header */
 	/* dw0 */
 	dw0 = port->id << CMD_HDR_PORT_OFF;
-	if (parent_dev && dev_is_expander(parent_dev->dev_type))
+	if (parent_dev && dev_is_expander(parent_dev->dev_type)) {
 		dw0 |= 3 << CMD_HDR_CMD_OFF;
-	else
+	} else {
+		phy_id = device->phy->identify.phy_identifier;
+		dw0 |= (1U << phy_id) << CMD_HDR_PHY_ID_OFF;
+		dw0 |= CMD_HDR_FORCE_PHY_MSK;
 		dw0 |= 4 << CMD_HDR_CMD_OFF;
+	}
 
 	if (tmf && ata_task->force_phy) {
 		dw0 |= CMD_HDR_FORCE_PHY_MSK;
diff --git a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
index 095bbf80c34e..6a0656f3b596 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v3_hw.c
@@ -359,6 +359,10 @@
 #define CMD_HDR_RESP_REPORT_MSK		(0x1 << CMD_HDR_RESP_REPORT_OFF)
 #define CMD_HDR_TLR_CTRL_OFF		6
 #define CMD_HDR_TLR_CTRL_MSK		(0x3 << CMD_HDR_TLR_CTRL_OFF)
+#define CMD_HDR_PHY_ID_OFF		8
+#define CMD_HDR_PHY_ID_MSK		(0x1ff << CMD_HDR_PHY_ID_OFF)
+#define CMD_HDR_FORCE_PHY_OFF		17
+#define CMD_HDR_FORCE_PHY_MSK		(0x1U << CMD_HDR_FORCE_PHY_OFF)
 #define CMD_HDR_PORT_OFF		18
 #define CMD_HDR_PORT_MSK		(0xf << CMD_HDR_PORT_OFF)
 #define CMD_HDR_PRIORITY_OFF		27
@@ -1429,15 +1433,21 @@ static void prep_ata_v3_hw(struct hisi_hba *hisi_hba,
 	struct hisi_sas_cmd_hdr *hdr = slot->cmd_hdr;
 	struct asd_sas_port *sas_port = device->port;
 	struct hisi_sas_port *port = to_hisi_sas_port(sas_port);
+	int phy_id;
 	u8 *buf_cmd;
 	int has_data = 0, hdr_tag = 0;
 	u32 dw1 = 0, dw2 = 0;
 
 	hdr->dw0 = cpu_to_le32(port->id << CMD_HDR_PORT_OFF);
-	if (parent_dev && dev_is_expander(parent_dev->dev_type))
+	if (parent_dev && dev_is_expander(parent_dev->dev_type)) {
 		hdr->dw0 |= cpu_to_le32(3 << CMD_HDR_CMD_OFF);
-	else
+	} else {
+		phy_id = device->phy->identify.phy_identifier;
+		hdr->dw0 |= cpu_to_le32((1U << phy_id)
+				<< CMD_HDR_PHY_ID_OFF);
+		hdr->dw0 |= CMD_HDR_FORCE_PHY_MSK;
 		hdr->dw0 |= cpu_to_le32(4U << CMD_HDR_CMD_OFF);
+	}
 
 	switch (task->data_dir) {
 	case DMA_TO_DEVICE:
diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
index 4e1e214fc5a2..3288a7da73dc 100644
--- a/drivers/ufs/core/ufshcd.c
+++ b/drivers/ufs/core/ufshcd.c
@@ -3176,16 +3176,10 @@ static int ufshcd_wait_for_dev_cmd(struct ufs_hba *hba,
 	int err;
 
 retry:
-	time_left = wait_for_completion_timeout(hba->dev_cmd.complete,
+	time_left = wait_for_completion_timeout(&hba->dev_cmd.complete,
 						time_left);
 
 	if (likely(time_left)) {
-		/*
-		 * The completion handler called complete() and the caller of
-		 * this function still owns the @lrbp tag so the code below does
-		 * not trigger any race conditions.
-		 */
-		hba->dev_cmd.complete = NULL;
 		err = ufshcd_get_tr_ocs(lrbp, NULL);
 		if (!err)
 			err = ufshcd_dev_cmd_completion(hba, lrbp);
@@ -3199,7 +3193,6 @@ static int ufshcd_wait_for_dev_cmd(struct ufs_hba *hba,
 			/* successfully cleared the command, retry if needed */
 			if (ufshcd_clear_cmd(hba, lrbp->task_tag) == 0)
 				err = -EAGAIN;
-			hba->dev_cmd.complete = NULL;
 			return err;
 		}
 
@@ -3215,11 +3208,9 @@ static int ufshcd_wait_for_dev_cmd(struct ufs_hba *hba,
 			spin_lock_irqsave(&hba->outstanding_lock, flags);
 			pending = test_bit(lrbp->task_tag,
 					   &hba->outstanding_reqs);
-			if (pending) {
-				hba->dev_cmd.complete = NULL;
+			if (pending)
 				__clear_bit(lrbp->task_tag,
 					    &hba->outstanding_reqs);
-			}
 			spin_unlock_irqrestore(&hba->outstanding_lock, flags);
 
 			if (!pending) {
@@ -3237,8 +3228,6 @@ static int ufshcd_wait_for_dev_cmd(struct ufs_hba *hba,
 			spin_lock_irqsave(&hba->outstanding_lock, flags);
 			pending = test_bit(lrbp->task_tag,
 					   &hba->outstanding_reqs);
-			if (pending)
-				hba->dev_cmd.complete = NULL;
 			spin_unlock_irqrestore(&hba->outstanding_lock, flags);
 
 			if (!pending) {
@@ -3272,13 +3261,9 @@ static void ufshcd_dev_man_unlock(struct ufs_hba *hba)
 static int ufshcd_issue_dev_cmd(struct ufs_hba *hba, struct ufshcd_lrb *lrbp,
 			  const u32 tag, int timeout)
 {
-	DECLARE_COMPLETION_ONSTACK(wait);
 	int err;
 
-	hba->dev_cmd.complete = &wait;
-
 	ufshcd_add_query_upiu_trace(hba, UFS_QUERY_SEND, lrbp->ucd_req_ptr);
-
 	ufshcd_send_command(hba, tag, hba->dev_cmd_queue);
 	err = ufshcd_wait_for_dev_cmd(hba, lrbp, timeout);
 
@@ -5585,12 +5570,12 @@ void ufshcd_compl_one_cqe(struct ufs_hba *hba, int task_tag,
 		ufshcd_release_scsi_cmd(hba, lrbp);
 		/* Do not touch lrbp after scsi done */
 		scsi_done(cmd);
-	} else if (hba->dev_cmd.complete) {
+	} else {
 		if (cqe) {
 			ocs = le32_to_cpu(cqe->status) & MASK_OCS;
 			lrbp->utr_descriptor_ptr->header.ocs = ocs;
 		}
-		complete(hba->dev_cmd.complete);
+		complete(&hba->dev_cmd.complete);
 	}
 }
 
@@ -10475,6 +10460,8 @@ int ufshcd_init(struct ufs_hba *hba, void __iomem *mmio_base, unsigned int irq)
 	 */
 	spin_lock_init(&hba->clk_gating.lock);
 
+	init_completion(&hba->dev_cmd.complete);
+
 	err = ufshcd_hba_init(hba);
 	if (err)
 		goto out_error;
diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
index e3909cc691b2..f56050ce9445 100644
--- a/include/ufs/ufshcd.h
+++ b/include/ufs/ufshcd.h
@@ -246,7 +246,7 @@ struct ufs_query {
 struct ufs_dev_cmd {
 	enum dev_cmd_type type;
 	struct mutex lock;
-	struct completion *complete;
+	struct completion complete;
 	struct ufs_query query;
 };