RDMA/hns: Add SCC context clr support for hip08

This patch adds SCC context clear support for DCQCN in kernel space
driver.

Signed-off-by: Yangyang Li <liyangyang20@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

4 files changed, 85 insertions, 1 deletions

diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index 7c93be4379e6..bbcc56f8fc76 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -202,6 +202,7 @@ enum {
 	HNS_ROCE_CAP_FLAG_SRQ			= BIT(5),
 	HNS_ROCE_CAP_FLAG_MW			= BIT(7),
 	HNS_ROCE_CAP_FLAG_FRMR                  = BIT(8),
+	HNS_ROCE_CAP_FLAG_QP_FLOW_CTRL		= BIT(9),
 	HNS_ROCE_CAP_FLAG_ATOMIC		= BIT(10),
 };
 
@@ -483,6 +484,7 @@ struct hns_roce_qp_table {
 	struct hns_roce_hem_table	irrl_table;
 	struct hns_roce_hem_table	trrl_table;
 	struct hns_roce_hem_table	sccc_table;
+	struct mutex			scc_mutex;
 };
 
 struct hns_roce_cq_table {
@@ -867,6 +869,8 @@ struct hns_roce_hw {
 			 int attr_mask, enum ib_qp_state cur_state,
 			 enum ib_qp_state new_state);
 	int (*destroy_qp)(struct ib_qp *ibqp);
+	int (*qp_flow_control_init)(struct hns_roce_dev *hr_dev,
+			 struct hns_roce_qp *hr_qp);
 	int (*post_send)(struct ib_qp *ibqp, const struct ib_send_wr *wr,
 			 const struct ib_send_wr **bad_wr);
 	int (*post_recv)(struct ib_qp *qp, const struct ib_recv_wr *recv_wr,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index 549a60113766..5e57f5f93d49 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -1436,7 +1436,9 @@ static int hns_roce_v2_profile(struct hns_roce_dev *hr_dev)
 
 	if (hr_dev->pci_dev->revision == 0x21) {
 		caps->flags |= HNS_ROCE_CAP_FLAG_ATOMIC |
-			       HNS_ROCE_CAP_FLAG_SRQ;
+			       HNS_ROCE_CAP_FLAG_SRQ |
+			       HNS_ROCE_CAP_FLAG_QP_FLOW_CTRL;
+
 		caps->sccc_entry_sz	= HNS_ROCE_V2_SCCC_ENTRY_SZ;
 		caps->sccc_ba_pg_sz	= 0;
 		caps->sccc_buf_pg_sz    = 0;
@@ -4277,6 +4279,60 @@ static int hns_roce_v2_destroy_qp(struct ib_qp *ibqp)
 	return 0;
 }
 
+static int hns_roce_v2_qp_flow_control_init(struct hns_roce_dev *hr_dev,
+						struct hns_roce_qp *hr_qp)
+{
+	struct hns_roce_sccc_clr_done *rst, *resp;
+	struct hns_roce_sccc_clr *clr;
+	struct hns_roce_cmq_desc desc;
+	int ret, i;
+
+	mutex_lock(&hr_dev->qp_table.scc_mutex);
+
+	/* set scc ctx clear done flag */
+	hns_roce_cmq_setup_basic_desc(&desc, HNS_ROCE_OPC_RESET_SCCC, false);
+	rst = (struct hns_roce_sccc_clr_done *)desc.data;
+	ret =  hns_roce_cmq_send(hr_dev, &desc, 1);
+	if (ret) {
+		dev_err(hr_dev->dev, "Reset SCC ctx  failed(%d)\n", ret);
+		goto out;
+	}
+
+	/* clear scc context */
+	hns_roce_cmq_setup_basic_desc(&desc, HNS_ROCE_OPC_CLR_SCCC, false);
+	clr = (struct hns_roce_sccc_clr *)desc.data;
+	clr->qpn = cpu_to_le32(hr_qp->qpn);
+	ret =  hns_roce_cmq_send(hr_dev, &desc, 1);
+	if (ret) {
+		dev_err(hr_dev->dev, "Clear SCC ctx failed(%d)\n", ret);
+		goto out;
+	}
+
+	/* query scc context clear is done or not */
+	resp = (struct hns_roce_sccc_clr_done *)desc.data;
+	for (i = 0; i <= HNS_ROCE_CMQ_SCC_CLR_DONE_CNT; i++) {
+		hns_roce_cmq_setup_basic_desc(&desc,
+					      HNS_ROCE_OPC_QUERY_SCCC, true);
+		ret = hns_roce_cmq_send(hr_dev, &desc, 1);
+		if (ret) {
+			dev_err(hr_dev->dev, "Query clr cmq failed(%d)\n", ret);
+			goto out;
+		}
+
+		if (resp->clr_done)
+			goto out;
+
+		msleep(20);
+	}
+
+	dev_err(hr_dev->dev, "Query SCC clr done flag overtime.\n");
+	ret = -ETIMEDOUT;
+
+out:
+	mutex_unlock(&hr_dev->qp_table.scc_mutex);
+	return ret;
+}
+
 static int hns_roce_v2_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(cq->device);
@@ -5835,6 +5891,7 @@ static const struct hns_roce_hw hns_roce_hw_v2 = {
 	.modify_qp = hns_roce_v2_modify_qp,
 	.query_qp = hns_roce_v2_query_qp,
 	.destroy_qp = hns_roce_v2_destroy_qp,
+	.qp_flow_control_init = hns_roce_v2_qp_flow_control_init,
 	.modify_cq = hns_roce_v2_modify_cq,
 	.post_send = hns_roce_v2_post_send,
 	.post_recv = hns_roce_v2_post_recv,
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
index 007caa21195b..4da247d5ea18 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
@@ -123,6 +123,8 @@
 #define HNS_ROCE_CMQ_EN_B		16
 #define HNS_ROCE_CMQ_ENABLE		BIT(HNS_ROCE_CMQ_EN_B)
 
+#define HNS_ROCE_CMQ_SCC_CLR_DONE_CNT		5
+
 #define check_whether_last_step(hop_num, step_idx) \
 	((step_idx == 0 && hop_num == HNS_ROCE_HOP_NUM_0) || \
 	(step_idx == 1 && hop_num == 1) || \
@@ -232,6 +234,9 @@ enum hns_roce_opcode_type {
 	HNS_ROCE_OPC_POST_MB				= 0x8504,
 	HNS_ROCE_OPC_QUERY_MB_ST			= 0x8505,
 	HNS_ROCE_OPC_CFG_BT_ATTR			= 0x8506,
+	HNS_ROCE_OPC_CLR_SCCC				= 0x8509,
+	HNS_ROCE_OPC_QUERY_SCCC				= 0x850a,
+	HNS_ROCE_OPC_RESET_SCCC				= 0x850b,
 	HNS_SWITCH_PARAMETER_CFG			= 0x1033,
 };
 
@@ -1757,4 +1762,14 @@ struct hns_roce_wqe_atomic_seg {
 	__le64          cmp_data;
 };
 
+struct hns_roce_sccc_clr {
+	__le32 qpn;
+	__le32 rsv[5];
+};
+
+struct hns_roce_sccc_clr_done {
+	__le32 clr_done;
+	__le32 rsv[5];
+};
+
 #endif
diff --git a/drivers/infiniband/hw/hns/hns_roce_qp.c b/drivers/infiniband/hw/hns/hns_roce_qp.c
index dc9a30e3b7ce..73066bf38e47 100644
--- a/drivers/infiniband/hw/hns/hns_roce_qp.c
+++ b/drivers/infiniband/hw/hns/hns_roce_qp.c
@@ -811,6 +811,13 @@ static int hns_roce_create_qp_common(struct hns_roce_dev *hr_dev,
 		if (ret)
 			goto err_qp;
 	}
+
+	if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_QP_FLOW_CTRL) {
+		ret = hr_dev->hw->qp_flow_control_init(hr_dev, hr_qp);
+		if (ret)
+			goto err_qp;
+	}
+
 	hr_qp->event = hns_roce_ib_qp_event;
 
 	return 0;
@@ -1152,6 +1159,7 @@ int hns_roce_init_qp_table(struct hns_roce_dev *hr_dev)
 	int reserved_from_bot;
 	int ret;
 
+	mutex_init(&qp_table->scc_mutex);
 	spin_lock_init(&qp_table->lock);
 	INIT_RADIX_TREE(&hr_dev->qp_table_tree, GFP_ATOMIC);