RDMA/hns: Add QP operations support for hip08 SoC

This patch implements QP operations for hip08 RoCE driver and
fixes some checkpatch warning about print message in QP function.
The QP operations includes create QP, query QP, modify QP and
destroy QP.
Signed-off-by: Lijun Ou <oulijun@huawei.com>
Signed-off-by: Shaobo Xu <xushaobo2@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>

drivers/infiniband/hw/hns/hns_roce_cmd.h

@@ -48,6 +48,10 @@ enum {
 	HNS_ROCE_CMD_DESTROY_QPC_BT1	= 0x9,
 	HNS_ROCE_CMD_DESTROY_QPC_BT2	= 0xa,
 
+	/* QPC operation */
+	HNS_ROCE_CMD_MODIFY_QPC		= 0x41,
+	HNS_ROCE_CMD_QUERY_QPC		= 0x42,
+
 	/* CQC BT commands */
 	HNS_ROCE_CMD_WRITE_CQC_BT0	= 0x10,
 	HNS_ROCE_CMD_WRITE_CQC_BT1	= 0x11,

drivers/infiniband/hw/hns/hns_roce_device.h

@@ -304,6 +304,12 @@ struct hns_roce_wq {
 	void __iomem	*db_reg_l;
 };
 
+struct hns_roce_sge {
+	int		sge_cnt;  /* SGE num */
+	int		offset;
+	int		sge_shift;/* SGE size */
+};
+
 struct hns_roce_buf_list {
 	void		*buf;
 	dma_addr_t	map;
@@ -455,6 +461,9 @@ struct hns_roce_qp {
 
 	atomic_t		refcount;
 	struct completion	free;
+
+	struct hns_roce_sge	sge;
+	u32			next_sge;
 };
 
 struct hns_roce_sqp {
@@ -509,6 +518,7 @@ struct hns_roce_caps {
 	int		num_cqs;
 	int		max_cqes;
 	int		min_cqes;
+	u32		min_wqes;
 	int		reserved_cqs;
 	int		num_aeq_vectors;	/* 1 */
 	int		num_comp_vectors;	/* 32 ceq */
@@ -788,6 +798,7 @@ int hns_roce_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		       int attr_mask, struct ib_udata *udata);
 void *get_recv_wqe(struct hns_roce_qp *hr_qp, int n);
 void *get_send_wqe(struct hns_roce_qp *hr_qp, int n);
+void *get_send_extend_sge(struct hns_roce_qp *hr_qp, int n);
 bool hns_roce_wq_overflow(struct hns_roce_wq *hr_wq, int nreq,
 			  struct ib_cq *ib_cq);
 enum hns_roce_qp_state to_hns_roce_state(enum ib_qp_state state);

drivers/infiniband/hw/hns/hns_roce_hw_v1.c

@@ -1475,6 +1475,7 @@ int hns_roce_v1_profile(struct hns_roce_dev *hr_dev)
 
 	caps->num_qps		= HNS_ROCE_V1_MAX_QP_NUM;
 	caps->max_wqes		= HNS_ROCE_V1_MAX_WQE_NUM;
+	caps->min_wqes		= HNS_ROCE_MIN_WQE_NUM;
 	caps->num_cqs		= HNS_ROCE_V1_MAX_CQ_NUM;
 	caps->min_cqes		= HNS_ROCE_MIN_CQE_NUM;
 	caps->max_cqes		= HNS_ROCE_V1_MAX_CQE_NUM;

drivers/infiniband/hw/hns/hns_roce_qp.c

@@ -286,20 +286,27 @@ static int hns_roce_set_rq_size(struct hns_roce_dev *hr_dev,
 			return -EINVAL;
 		}
 
-		/* In v1 engine, parameter verification procession */
-		max_cnt = cap->max_recv_wr > HNS_ROCE_MIN_WQE_NUM ?
-			  cap->max_recv_wr : HNS_ROCE_MIN_WQE_NUM;
+		if (hr_dev->caps.min_wqes)
+			max_cnt = max(cap->max_recv_wr, hr_dev->caps.min_wqes);
+		else
+			max_cnt = cap->max_recv_wr;
+
 		hr_qp->rq.wqe_cnt = roundup_pow_of_two(max_cnt);
 
 		if ((u32)hr_qp->rq.wqe_cnt > hr_dev->caps.max_wqes) {
-			dev_err(dev, "hns_roce_set_rq_size rq.wqe_cnt too large\n");
+			dev_err(dev, "while setting rq size, rq.wqe_cnt too large\n");
 			return -EINVAL;
 		}
 
 		max_cnt = max(1U, cap->max_recv_sge);
 		hr_qp->rq.max_gs = roundup_pow_of_two(max_cnt);
-		/* WQE is fixed for 64B */
-		hr_qp->rq.wqe_shift = ilog2(hr_dev->caps.max_rq_desc_sz);
+		if (hr_dev->caps.max_rq_sg <= 2)
+			hr_qp->rq.wqe_shift =
+					ilog2(hr_dev->caps.max_rq_desc_sz);
+		else
+			hr_qp->rq.wqe_shift =
+					ilog2(hr_dev->caps.max_rq_desc_sz
+					      * hr_qp->rq.max_gs);
 	}
 
 	cap->max_recv_wr = hr_qp->rq.max_post = hr_qp->rq.wqe_cnt;
@@ -309,11 +316,13 @@ static int hns_roce_set_rq_size(struct hns_roce_dev *hr_dev,
 }
 
 static int hns_roce_set_user_sq_size(struct hns_roce_dev *hr_dev,
+				     struct ib_qp_cap *cap,
 				     struct hns_roce_qp *hr_qp,
 				     struct hns_roce_ib_create_qp *ucmd)
 {
 	u32 roundup_sq_stride = roundup_pow_of_two(hr_dev->caps.max_sq_desc_sz);
 	u8 max_sq_stride = ilog2(roundup_sq_stride);
+	u32 max_cnt;
 
 	/* Sanity check SQ size before proceeding */
 	if ((u32)(1 << ucmd->log_sq_bb_count) > hr_dev->caps.max_wqes ||
@@ -323,18 +332,61 @@ static int hns_roce_set_user_sq_size(struct hns_roce_dev *hr_dev,
 		return -EINVAL;
 	}
 
+	if (cap->max_send_sge > hr_dev->caps.max_sq_sg) {
+		dev_err(hr_dev->dev, "SQ sge error! max_send_sge=%d\n",
+			cap->max_send_sge);
+		return -EINVAL;
+	}
+
 	hr_qp->sq.wqe_cnt = 1 << ucmd->log_sq_bb_count;
 	hr_qp->sq.wqe_shift = ucmd->log_sq_stride;
 
+	max_cnt = max(1U, cap->max_send_sge);
+	if (hr_dev->caps.max_sq_sg <= 2)
+		hr_qp->sq.max_gs = roundup_pow_of_two(max_cnt);
+	else
+		hr_qp->sq.max_gs = max_cnt;
+
+	if (hr_qp->sq.max_gs > 2)
+		hr_qp->sge.sge_cnt = roundup_pow_of_two(hr_qp->sq.wqe_cnt *
+							(hr_qp->sq.max_gs - 2));
+	hr_qp->sge.sge_shift = 4;
+
 	/* Get buf size, SQ and RQ  are aligned to page_szie */
-	hr_qp->buff_size = HNS_ROCE_ALOGN_UP((hr_qp->rq.wqe_cnt <<
+	if (hr_dev->caps.max_sq_sg <= 2) {
+		hr_qp->buff_size = HNS_ROCE_ALOGN_UP((hr_qp->rq.wqe_cnt <<
 					     hr_qp->rq.wqe_shift), PAGE_SIZE) +
-			   HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
+				   HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
 					     hr_qp->sq.wqe_shift), PAGE_SIZE);
 
-	hr_qp->sq.offset = 0;
-	hr_qp->rq.offset = HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
+		hr_qp->sq.offset = 0;
+		hr_qp->rq.offset = HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
 					     hr_qp->sq.wqe_shift), PAGE_SIZE);
+	} else {
+		hr_qp->buff_size = HNS_ROCE_ALOGN_UP((hr_qp->rq.wqe_cnt <<
+					     hr_qp->rq.wqe_shift), PAGE_SIZE) +
+				   HNS_ROCE_ALOGN_UP((hr_qp->sge.sge_cnt <<
+					     hr_qp->sge.sge_shift), PAGE_SIZE) +
+				   HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
+					     hr_qp->sq.wqe_shift), PAGE_SIZE);
+
+		hr_qp->sq.offset = 0;
+		if (hr_qp->sge.sge_cnt) {
+			hr_qp->sge.offset = HNS_ROCE_ALOGN_UP(
+							(hr_qp->sq.wqe_cnt <<
+							hr_qp->sq.wqe_shift),
+							PAGE_SIZE);
+			hr_qp->rq.offset = hr_qp->sge.offset +
+					HNS_ROCE_ALOGN_UP((hr_qp->sge.sge_cnt <<
+						hr_qp->sge.sge_shift),
+						PAGE_SIZE);
+		} else {
+			hr_qp->rq.offset = HNS_ROCE_ALOGN_UP(
+							(hr_qp->sq.wqe_cnt <<
+							hr_qp->sq.wqe_shift),
+							PAGE_SIZE);
+		}
+	}
 
 	return 0;
 }
@@ -345,11 +397,12 @@ static int hns_roce_set_kernel_sq_size(struct hns_roce_dev *hr_dev,
 {
 	struct device *dev = hr_dev->dev;
 	u32 max_cnt;
+	int size;
 
 	if (cap->max_send_wr  > hr_dev->caps.max_wqes  ||
 	    cap->max_send_sge > hr_dev->caps.max_sq_sg ||
 	    cap->max_inline_data > hr_dev->caps.max_sq_inline) {
-		dev_err(dev, "hns_roce_set_kernel_sq_size error1\n");
+		dev_err(dev, "SQ WR or sge or inline data error!\n");
 		return -EINVAL;
 	}
 
@@ -357,27 +410,45 @@ static int hns_roce_set_kernel_sq_size(struct hns_roce_dev *hr_dev,
 	hr_qp->sq_max_wqes_per_wr = 1;
 	hr_qp->sq_spare_wqes = 0;
 
-	/* In v1 engine, parameter verification procession */
-	max_cnt = cap->max_send_wr > HNS_ROCE_MIN_WQE_NUM ?
-		  cap->max_send_wr : HNS_ROCE_MIN_WQE_NUM;
+	if (hr_dev->caps.min_wqes)
+		max_cnt = max(cap->max_send_wr, hr_dev->caps.min_wqes);
+	else
+		max_cnt = cap->max_send_wr;
+
 	hr_qp->sq.wqe_cnt = roundup_pow_of_two(max_cnt);
 	if ((u32)hr_qp->sq.wqe_cnt > hr_dev->caps.max_wqes) {
-		dev_err(dev, "hns_roce_set_kernel_sq_size sq.wqe_cnt too large\n");
+		dev_err(dev, "while setting kernel sq size, sq.wqe_cnt too large\n");
 		return -EINVAL;
 	}
 
 	/* Get data_seg numbers */
 	max_cnt = max(1U, cap->max_send_sge);
-	hr_qp->sq.max_gs = roundup_pow_of_two(max_cnt);
+	if (hr_dev->caps.max_sq_sg <= 2)
+		hr_qp->sq.max_gs = roundup_pow_of_two(max_cnt);
+	else
+		hr_qp->sq.max_gs = max_cnt;
 
-	/* Get buf size, SQ and RQ  are aligned to page_szie */
-	hr_qp->buff_size = HNS_ROCE_ALOGN_UP((hr_qp->rq.wqe_cnt <<
-					     hr_qp->rq.wqe_shift), PAGE_SIZE) +
-			   HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
-					     hr_qp->sq.wqe_shift), PAGE_SIZE);
+	if (hr_qp->sq.max_gs > 2) {
+		hr_qp->sge.sge_cnt = roundup_pow_of_two(hr_qp->sq.wqe_cnt *
+				     (hr_qp->sq.max_gs - 2));
+		hr_qp->sge.sge_shift = 4;
+	}
+
+	/* Get buf size, SQ and RQ are aligned to PAGE_SIZE */
 	hr_qp->sq.offset = 0;
-	hr_qp->rq.offset = HNS_ROCE_ALOGN_UP((hr_qp->sq.wqe_cnt <<
-					      hr_qp->sq.wqe_shift), PAGE_SIZE);
+	size = HNS_ROCE_ALOGN_UP(hr_qp->sq.wqe_cnt << hr_qp->sq.wqe_shift,
+				 PAGE_SIZE);
+
+	if (hr_dev->caps.max_sq_sg > 2 && hr_qp->sge.sge_cnt) {
+		hr_qp->sge.offset = size;
+		size += HNS_ROCE_ALOGN_UP(hr_qp->sge.sge_cnt <<
+					  hr_qp->sge.sge_shift, PAGE_SIZE);
+	}
+
+	hr_qp->rq.offset = size;
+	size += HNS_ROCE_ALOGN_UP((hr_qp->rq.wqe_cnt << hr_qp->rq.wqe_shift),
+				  PAGE_SIZE);
+	hr_qp->buff_size = size;
 
 	/* Get wr and sge number which send */
 	cap->max_send_wr = hr_qp->sq.max_post = hr_qp->sq.wqe_cnt;
@@ -425,7 +496,8 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev,
 			goto err_out;
 		}
 
-		ret = hns_roce_set_user_sq_size(hr_dev, hr_qp, &ucmd);
+		ret = hns_roce_set_user_sq_size(hr_dev, &init_attr->cap, hr_qp,
+						&ucmd);
 		if (ret) {
 			dev_err(dev, "hns_roce_set_user_sq_size error for create qp\n");
 			goto err_out;
@@ -528,7 +600,9 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev,
 		}
 	}
 
-	if ((init_attr->qp_type) == IB_QPT_GSI) {
+	if (init_attr->qp_type == IB_QPT_GSI &&
+	    hr_dev->hw_rev == HNS_ROCE_HW_VER1) {
+		/* In v1 engine, GSI QP context in RoCE engine's register */
 		ret = hns_roce_gsi_qp_alloc(hr_dev, qpn, hr_qp);
 		if (ret) {
 			dev_err(dev, "hns_roce_qp_alloc failed!\n");
@@ -700,7 +774,10 @@ int hns_roce_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		p = attr_mask & IB_QP_PORT ? (attr->port_num - 1) : hr_qp->port;
 		active_mtu = iboe_get_mtu(hr_dev->iboe.netdevs[p]->mtu);
 
-		if (attr->path_mtu > IB_MTU_2048 ||
+		if ((hr_dev->caps.max_mtu == IB_MTU_4096 &&
+		    attr->path_mtu > IB_MTU_4096) ||
+		    (hr_dev->caps.max_mtu == IB_MTU_2048 &&
+		    attr->path_mtu > IB_MTU_2048) ||
 		    attr->path_mtu < IB_MTU_256 ||
 		    attr->path_mtu > active_mtu) {
 			dev_err(dev, "attr path_mtu(%d)invalid while modify qp",
@@ -724,9 +801,7 @@ int hns_roce_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 	}
 
 	if (cur_state == new_state && cur_state == IB_QPS_RESET) {
-		ret = -EPERM;
-		dev_err(dev, "cur_state=%d new_state=%d\n", cur_state,
-			new_state);
+		ret = 0;
 		goto out;
 	}
 
@@ -804,6 +879,13 @@ void *get_send_wqe(struct hns_roce_qp *hr_qp, int n)
 }
 EXPORT_SYMBOL_GPL(get_send_wqe);
 
+void *get_send_extend_sge(struct hns_roce_qp *hr_qp, int n)
+{
+	return hns_roce_buf_offset(&hr_qp->hr_buf, hr_qp->sge.offset +
+					(n << hr_qp->sge.sge_shift));
+}
+EXPORT_SYMBOL_GPL(get_send_extend_sge);
+
 bool hns_roce_wq_overflow(struct hns_roce_wq *hr_wq, int nreq,
 			  struct ib_cq *ib_cq)
 {