[PATCH v2 for-next 0/4] RDMA/hns: Extend some capabilities for HIP09

[PATCH v2 for-next 0/4] RDMA/hns: Extend some capabilities for HIP09

[Weihang Li]

HIP09 supports larger entries/contexts to improve the performance of bus or
exchange more information with the hardware. The capbilities changes from
HIP08 to HIP09 is as follows:
- CEQE: 4B -> 64B
- AEQE: 16B -> 64B
- CQE: 32B -> 64B
- QPC: 256B -> 512B

Previous discussions can be found at:
v1: https://patchwork.kernel.org/cover/11718143/

Changes since v1:
- Fix comments from Lang Cheng about redundant comments and type of
  reserved fields in structure of eqe.
- Rename some variables.

Wenpeng Liang (3):
  RDMA/hns: Add support for EQE in size of 64 Bytes
  RDMA/hns: Add support for CQE in size of 64 Bytes
  RDMA/hns: Add support for QPC in size of 512 Bytes

Xi Wang (1):
  RDMA/hns: Export hardware capability flags to userspace

 drivers/infiniband/hw/hns/hns_roce_cq.c     | 19 ++++++-
 drivers/infiniband/hw/hns/hns_roce_device.h | 25 ++++++---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 17 +++----
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  4 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c  | 79 ++++++++++++++++++++---------
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h  | 18 ++++---
 drivers/infiniband/hw/hns/hns_roce_main.c   |  5 +-
 include/uapi/rdma/hns-abi.h                 |  4 ++
 8 files changed, 121 insertions(+), 50 deletions(-)

-- 
2.8.1

[PATCH v2 for-next 1/4] RDMA/hns: Export hardware capability flags to userspace

[Weihang Li]

From: Xi Wang <wangxi11@huawei.com>

The libhns in userspace for HIP09 will use the hardware's capability to
enable some features. So export the hardware capablility flags to userspace
by reusing the reserved fields in structure
"hns_roce_ib_alloc_ucontext_resp".

Signed-off-by: Xi Wang <wangxi11@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
---
 drivers/infiniband/hw/hns/hns_roce_main.c | 1 +
 include/uapi/rdma/hns-abi.h               | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 5907cfd..98945df 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -313,6 +313,7 @@ static int hns_roce_alloc_ucontext(struct ib_ucontext *uctx,
 		return -EAGAIN;
 
 	resp.qp_tab_size = hr_dev->caps.num_qps;
+	resp.cap_flags = (u32)hr_dev->caps.flags;
 
 	ret = hns_roce_uar_alloc(hr_dev, &context->uar);
 	if (ret)
diff --git a/include/uapi/rdma/hns-abi.h b/include/uapi/rdma/hns-abi.h
index eb76b38..5c38758 100644
--- a/include/uapi/rdma/hns-abi.h
+++ b/include/uapi/rdma/hns-abi.h
@@ -73,7 +73,7 @@ struct hns_roce_ib_create_qp_resp {
 
 struct hns_roce_ib_alloc_ucontext_resp {
 	__u32	qp_tab_size;
-	__u32	reserved;
+	__u32	cap_flags;
 };
 
 struct hns_roce_ib_alloc_pd_resp {
-- 
2.8.1

[PATCH v2 for-next 2/4] RDMA/hns: Add support for EQE in size of 64 Bytes

[Weihang Li]

From: Wenpeng Liang <liangwenpeng@huawei.com>

The new version of RoCEE supports using CEQE in size of 4B or 64B, AEQE in
size of 16B or 64B. The performance of bus can be improved by using larger
size of EQE.

Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
---
 drivers/infiniband/hw/hns/hns_roce_device.h | 14 ++++++++----
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  | 10 ++++-----
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c  | 33 ++++++++++++++++++++++-------
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h  |  7 ++++--
 4 files changed, 44 insertions(+), 20 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index da9888d..d871451 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -37,8 +37,8 @@
 
 #define DRV_NAME "hns_roce"
 
-/* hip08 is a pci device */
 #define PCI_REVISION_ID_HIP08			0x21
+#define PCI_REVISION_ID_HIP09			0x30
 
 #define HNS_ROCE_HW_VER1	('h' << 24 | 'i' << 16 | '0' << 8 | '6')
 
@@ -78,8 +78,10 @@
 #define HNS_ROCE_CEQ				0
 #define HNS_ROCE_AEQ				1
 
-#define HNS_ROCE_CEQ_ENTRY_SIZE			0x4
-#define HNS_ROCE_AEQ_ENTRY_SIZE			0x10
+#define HNS_ROCE_CEQE_SIZE 0x4
+#define HNS_ROCE_AEQE_SIZE 0x10
+
+#define HNS_ROCE_V3_EQE_SIZE 0x40
 
 #define HNS_ROCE_SL_SHIFT			28
 #define HNS_ROCE_TCLASS_SHIFT			20
@@ -680,7 +682,8 @@ enum {
 };
 
 struct hns_roce_ceqe {
-	__le32			comp;
+	__le32	comp;
+	__le32	rsv[15];
 };
 
 struct hns_roce_aeqe {
@@ -717,6 +720,7 @@ struct hns_roce_aeqe {
 			u8	rsv0;
 		} __packed cmd;
 	 } event;
+	__le32 rsv[12];
 };
 
 struct hns_roce_eq {
@@ -811,6 +815,8 @@ struct hns_roce_caps {
 	u32		pbl_hop_num;
 	int		aeqe_depth;
 	int		ceqe_depth;
+	u32		aeqe_size;
+	u32		ceqe_size;
 	enum ib_mtu	max_mtu;
 	u32		qpc_bt_num;
 	u32		qpc_timer_bt_num;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index 07b4c85..ae9a0ef 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -3775,8 +3775,7 @@ static void hns_roce_v1_db_overflow_handle(struct hns_roce_dev *hr_dev,
 
 static struct hns_roce_aeqe *get_aeqe_v1(struct hns_roce_eq *eq, u32 entry)
 {
-	unsigned long off = (entry & (eq->entries - 1)) *
-			     HNS_ROCE_AEQ_ENTRY_SIZE;
+	unsigned long off = (entry & (eq->entries - 1)) * HNS_ROCE_AEQE_SIZE;
 
 	return (struct hns_roce_aeqe *)((u8 *)
 		(eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) +
@@ -3881,8 +3880,7 @@ static int hns_roce_v1_aeq_int(struct hns_roce_dev *hr_dev,
 
 static struct hns_roce_ceqe *get_ceqe_v1(struct hns_roce_eq *eq, u32 entry)
 {
-	unsigned long off = (entry & (eq->entries - 1)) *
-			     HNS_ROCE_CEQ_ENTRY_SIZE;
+	unsigned long off = (entry & (eq->entries - 1)) * HNS_ROCE_CEQE_SIZE;
 
 	return (struct hns_roce_ceqe *)((u8 *)
 			(eq->buf_list[off / HNS_ROCE_BA_SIZE].buf) +
@@ -4253,7 +4251,7 @@ static int hns_roce_v1_init_eq_table(struct hns_roce_dev *hr_dev)
 				       CEQ_REG_OFFSET * i;
 			eq->entries = hr_dev->caps.ceqe_depth;
 			eq->log_entries = ilog2(eq->entries);
-			eq->eqe_size = HNS_ROCE_CEQ_ENTRY_SIZE;
+			eq->eqe_size = HNS_ROCE_CEQE_SIZE;
 		} else {
 			/* AEQ */
 			eq_table->eqc_base[i] = hr_dev->reg_base +
@@ -4263,7 +4261,7 @@ static int hns_roce_v1_init_eq_table(struct hns_roce_dev *hr_dev)
 				       ROCEE_CAEP_AEQE_CONS_IDX_REG;
 			eq->entries = hr_dev->caps.aeqe_depth;
 			eq->log_entries = ilog2(eq->entries);
-			eq->eqe_size = HNS_ROCE_AEQ_ENTRY_SIZE;
+			eq->eqe_size = HNS_ROCE_AEQE_SIZE;
 		}
 	}
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index d296859..ea007e8 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -1740,6 +1740,8 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 	caps->gid_table_len[0]	= HNS_ROCE_V2_GID_INDEX_NUM;
 	caps->ceqe_depth	= HNS_ROCE_V2_COMP_EQE_NUM;
 	caps->aeqe_depth	= HNS_ROCE_V2_ASYNC_EQE_NUM;
+	caps->aeqe_size		= HNS_ROCE_AEQE_SIZE;
+	caps->ceqe_size		= HNS_ROCE_CEQE_SIZE;
 	caps->local_ca_ack_delay = 0;
 	caps->max_mtu = IB_MTU_4096;
 
@@ -1765,6 +1767,11 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 	caps->sccc_ba_pg_sz	  = 0;
 	caps->sccc_buf_pg_sz	  = 0;
 	caps->sccc_hop_num	  = HNS_ROCE_SCCC_HOP_NUM;
+
+	if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) {
+		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
+		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
+	}
 }
 
 static void calc_pg_sz(int obj_num, int obj_size, int hop_num, int ctx_bt_num,
@@ -1959,6 +1966,8 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 	caps->cqc_timer_entry_sz = HNS_ROCE_V2_CQC_TIMER_ENTRY_SZ;
 	caps->mtt_entry_sz = HNS_ROCE_V2_MTT_ENTRY_SZ;
 	caps->num_mtt_segs = HNS_ROCE_V2_MAX_MTT_SEGS;
+	caps->ceqe_size = HNS_ROCE_CEQE_SIZE;
+	caps->aeqe_size = HNS_ROCE_AEQE_SIZE;
 	caps->mtt_ba_pg_sz = 0;
 	caps->num_cqe_segs = HNS_ROCE_V2_MAX_CQE_SEGS;
 	caps->num_srqwqe_segs = HNS_ROCE_V2_MAX_SRQWQE_SEGS;
@@ -1982,6 +1991,11 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 					  V2_QUERY_PF_CAPS_D_RQWQE_HOP_NUM_M,
 					  V2_QUERY_PF_CAPS_D_RQWQE_HOP_NUM_S);
 
+	if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) {
+		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
+		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
+	}
+
 	calc_pg_sz(caps->num_qps, caps->qpc_entry_sz, caps->qpc_hop_num,
 		   caps->qpc_bt_num, &caps->qpc_buf_pg_sz, &caps->qpc_ba_pg_sz,
 		   HEM_TYPE_QPC);
@@ -5234,7 +5248,7 @@ static struct hns_roce_aeqe *next_aeqe_sw_v2(struct hns_roce_eq *eq)
 
 	aeqe = hns_roce_buf_offset(eq->mtr.kmem,
 				   (eq->cons_index & (eq->entries - 1)) *
-				   HNS_ROCE_AEQ_ENTRY_SIZE);
+				   eq->eqe_size);
 
 	return (roce_get_bit(aeqe->asyn, HNS_ROCE_V2_AEQ_AEQE_OWNER_S) ^
 		!!(eq->cons_index & eq->entries)) ? aeqe : NULL;
@@ -5334,7 +5348,8 @@ static struct hns_roce_ceqe *next_ceqe_sw_v2(struct hns_roce_eq *eq)
 
 	ceqe = hns_roce_buf_offset(eq->mtr.kmem,
 				   (eq->cons_index & (eq->entries - 1)) *
-				   HNS_ROCE_CEQ_ENTRY_SIZE);
+				   eq->eqe_size);
+
 	return (!!(roce_get_bit(ceqe->comp, HNS_ROCE_V2_CEQ_CEQE_OWNER_S))) ^
 		(!!(eq->cons_index & eq->entries)) ? ceqe : NULL;
 }
@@ -5610,14 +5625,16 @@ static int config_eqc(struct hns_roce_dev *hr_dev, struct hns_roce_eq *eq,
 	roce_set_field(eqc->byte_36, HNS_ROCE_EQC_CONS_INDX_M,
 		       HNS_ROCE_EQC_CONS_INDX_S, HNS_ROCE_EQ_INIT_CONS_IDX);
 
-	/* set nex_eqe_ba[43:12] */
-	roce_set_field(eqc->nxt_eqe_ba0, HNS_ROCE_EQC_NXT_EQE_BA_L_M,
+	roce_set_field(eqc->byte_40, HNS_ROCE_EQC_NXT_EQE_BA_L_M,
 		       HNS_ROCE_EQC_NXT_EQE_BA_L_S, eqe_ba[1] >> 12);
 
-	/* set nex_eqe_ba[63:44] */
-	roce_set_field(eqc->nxt_eqe_ba1, HNS_ROCE_EQC_NXT_EQE_BA_H_M,
+	roce_set_field(eqc->byte_44, HNS_ROCE_EQC_NXT_EQE_BA_H_M,
 		       HNS_ROCE_EQC_NXT_EQE_BA_H_S, eqe_ba[1] >> 44);
 
+	roce_set_field(eqc->byte_44, HNS_ROCE_EQC_EQE_SIZE_M,
+		       HNS_ROCE_EQC_EQE_SIZE_S,
+		       eq->eqe_size == HNS_ROCE_V3_EQE_SIZE ? 1 : 0);
+
 	return 0;
 }
 
@@ -5808,7 +5825,7 @@ static int hns_roce_v2_init_eq_table(struct hns_roce_dev *hr_dev)
 			eq_cmd = HNS_ROCE_CMD_CREATE_CEQC;
 			eq->type_flag = HNS_ROCE_CEQ;
 			eq->entries = hr_dev->caps.ceqe_depth;
-			eq->eqe_size = HNS_ROCE_CEQ_ENTRY_SIZE;
+			eq->eqe_size = hr_dev->caps.ceqe_size;
 			eq->irq = hr_dev->irq[i + other_num + aeq_num];
 			eq->eq_max_cnt = HNS_ROCE_CEQ_DEFAULT_BURST_NUM;
 			eq->eq_period = HNS_ROCE_CEQ_DEFAULT_INTERVAL;
@@ -5817,7 +5834,7 @@ static int hns_roce_v2_init_eq_table(struct hns_roce_dev *hr_dev)
 			eq_cmd = HNS_ROCE_CMD_CREATE_AEQC;
 			eq->type_flag = HNS_ROCE_AEQ;
 			eq->entries = hr_dev->caps.aeqe_depth;
-			eq->eqe_size = HNS_ROCE_AEQ_ENTRY_SIZE;
+			eq->eqe_size = hr_dev->caps.aeqe_size;
 			eq->irq = hr_dev->irq[i - comp_num + other_num];
 			eq->eq_max_cnt = HNS_ROCE_AEQ_DEFAULT_BURST_NUM;
 			eq->eq_period = HNS_ROCE_AEQ_DEFAULT_INTERVAL;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
index 1fb1c58..9ed3339 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
@@ -1779,8 +1779,8 @@ struct hns_roce_eq_context {
 	__le32	byte_28;
 	__le32	byte_32;
 	__le32	byte_36;
-	__le32	nxt_eqe_ba0;
-	__le32	nxt_eqe_ba1;
+	__le32	byte_40;
+	__le32	byte_44;
 	__le32	rsv[5];
 };
 
@@ -1922,6 +1922,9 @@ struct hns_roce_eq_context {
 #define HNS_ROCE_EQC_NXT_EQE_BA_H_S 0
 #define HNS_ROCE_EQC_NXT_EQE_BA_H_M GENMASK(19, 0)
 
+#define HNS_ROCE_EQC_EQE_SIZE_S 20
+#define HNS_ROCE_EQC_EQE_SIZE_M GENMASK(21, 20)
+
 #define HNS_ROCE_V2_CEQE_COMP_CQN_S 0
 #define HNS_ROCE_V2_CEQE_COMP_CQN_M GENMASK(23, 0)
 
-- 
2.8.1

[PATCH v2 for-next 3/4] RDMA/hns: Add support for CQE in size of 64 Bytes

[Weihang Li]

From: Wenpeng Liang <liangwenpeng@huawei.com>

The new version of RoCEE supports using CQE in size of 32B or 64B. The
performance of bus can be improved by using larger size of CQE.

Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
---
 drivers/infiniband/hw/hns/hns_roce_cq.c     | 19 ++++++++++++++++++-
 drivers/infiniband/hw/hns/hns_roce_device.h |  6 +++++-
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  |  5 ++---
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  2 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c  | 20 +++++++++++++-------
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h  |  7 +++++--
 drivers/infiniband/hw/hns/hns_roce_main.c   |  2 ++
 include/uapi/rdma/hns-abi.h                 |  4 ++++
 8 files changed, 50 insertions(+), 15 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_cq.c b/drivers/infiniband/hw/hns/hns_roce_cq.c
index e87d616..9a2f745 100644
--- a/drivers/infiniband/hw/hns/hns_roce_cq.c
+++ b/drivers/infiniband/hw/hns/hns_roce_cq.c
@@ -150,7 +150,7 @@ static int alloc_cq_buf(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq,
 	int err;
 
 	buf_attr.page_shift = hr_dev->caps.cqe_buf_pg_sz + HNS_HW_PAGE_SHIFT;
-	buf_attr.region[0].size = hr_cq->cq_depth * hr_dev->caps.cq_entry_sz;
+	buf_attr.region[0].size = hr_cq->cq_depth * hr_cq->cqe_size;
 	buf_attr.region[0].hopnum = hr_dev->caps.cqe_hop_num;
 	buf_attr.region_count = 1;
 	buf_attr.fixed_page = true;
@@ -224,6 +224,21 @@ static void free_cq_db(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq,
 	}
 }
 
+static void set_cqe_size(struct hns_roce_cq *hr_cq, struct ib_udata *udata,
+			 struct hns_roce_ib_create_cq *ucmd)
+{
+	struct hns_roce_dev *hr_dev = to_hr_dev(hr_cq->ib_cq.device);
+
+	if (udata) {
+		if (udata->inlen >= offsetofend(typeof(*ucmd), cqe_size))
+			hr_cq->cqe_size = ucmd->cqe_size;
+		else
+			hr_cq->cqe_size = HNS_ROCE_V2_CQE_SIZE;
+	} else {
+		hr_cq->cqe_size = hr_dev->caps.cqe_sz;
+	}
+}
+
 int hns_roce_create_cq(struct ib_cq *ib_cq, const struct ib_cq_init_attr *attr,
 		       struct ib_udata *udata)
 {
@@ -266,6 +281,8 @@ int hns_roce_create_cq(struct ib_cq *ib_cq, const struct ib_cq_init_attr *attr,
 		}
 	}
 
+	set_cqe_size(hr_cq, udata, &ucmd);
+
 	ret = alloc_cq_buf(hr_dev, hr_cq, udata, ucmd.buf_addr);
 	if (ret) {
 		ibdev_err(ibdev, "Failed to alloc CQ buf, err %d\n", ret);
diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index d871451..f4ffc87 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -83,6 +83,9 @@
 
 #define HNS_ROCE_V3_EQE_SIZE 0x40
 
+#define HNS_ROCE_V2_CQE_SIZE 32
+#define HNS_ROCE_V3_CQE_SIZE 64
+
 #define HNS_ROCE_SL_SHIFT			28
 #define HNS_ROCE_TCLASS_SHIFT			20
 #define HNS_ROCE_FLOW_LABEL_MASK		0xfffff
@@ -471,6 +474,7 @@ struct hns_roce_cq {
 	void __iomem			*cq_db_l;
 	u16				*tptr_addr;
 	int				arm_sn;
+	int				cqe_size;
 	unsigned long			cqn;
 	u32				vector;
 	atomic_t			refcount;
@@ -797,7 +801,7 @@ struct hns_roce_caps {
 	int		num_pds;
 	int		reserved_pds;
 	u32		mtt_entry_sz;
-	u32		cq_entry_sz;
+	u32		cqe_sz;
 	u32		page_size_cap;
 	u32		reserved_lkey;
 	int		mtpt_entry_sz;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index ae9a0ef..e053421 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1476,7 +1476,7 @@ static int hns_roce_v1_profile(struct hns_roce_dev *hr_dev)
 	caps->cqc_entry_sz	= HNS_ROCE_V1_CQC_ENTRY_SIZE;
 	caps->mtpt_entry_sz	= HNS_ROCE_V1_MTPT_ENTRY_SIZE;
 	caps->mtt_entry_sz	= HNS_ROCE_V1_MTT_ENTRY_SIZE;
-	caps->cq_entry_sz	= HNS_ROCE_V1_CQE_ENTRY_SIZE;
+	caps->cqe_sz		= HNS_ROCE_V1_CQE_SIZE;
 	caps->page_size_cap	= HNS_ROCE_V1_PAGE_SIZE_SUPPORT;
 	caps->reserved_lkey	= 0;
 	caps->reserved_pds	= 0;
@@ -1897,8 +1897,7 @@ static int hns_roce_v1_write_mtpt(struct hns_roce_dev *hr_dev, void *mb_buf,
 
 static void *get_cqe(struct hns_roce_cq *hr_cq, int n)
 {
-	return hns_roce_buf_offset(hr_cq->mtr.kmem,
-				   n * HNS_ROCE_V1_CQE_ENTRY_SIZE);
+	return hns_roce_buf_offset(hr_cq->mtr.kmem, n * HNS_ROCE_V1_CQE_SIZE);
 }
 
 static void *get_sw_cqe(struct hns_roce_cq *hr_cq, int n)
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 52307b2..5996892 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -74,7 +74,7 @@
 #define HNS_ROCE_V1_MTPT_ENTRY_SIZE			64
 #define HNS_ROCE_V1_MTT_ENTRY_SIZE			64
 
-#define HNS_ROCE_V1_CQE_ENTRY_SIZE			32
+#define HNS_ROCE_V1_CQE_SIZE				32
 #define HNS_ROCE_V1_PAGE_SIZE_SUPPORT			0xFFFFF000
 
 #define HNS_ROCE_V1_TABLE_CHUNK_SIZE			(1 << 17)
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index ea007e8..ea20d82 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -1691,7 +1691,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 	caps->mtpt_entry_sz	= HNS_ROCE_V2_MTPT_ENTRY_SZ;
 	caps->mtt_entry_sz	= HNS_ROCE_V2_MTT_ENTRY_SZ;
 	caps->idx_entry_sz	= HNS_ROCE_V2_IDX_ENTRY_SZ;
-	caps->cq_entry_sz	= HNS_ROCE_V2_CQE_ENTRY_SIZE;
+	caps->cqe_sz		= HNS_ROCE_V2_CQE_SIZE;
 	caps->page_size_cap	= HNS_ROCE_V2_PAGE_SIZE_SUPPORTED;
 	caps->reserved_lkey	= 0;
 	caps->reserved_pds	= 0;
@@ -1771,6 +1771,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 	if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) {
 		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
+		caps->cqe_sz = HNS_ROCE_V3_CQE_SIZE;
 	}
 }
 
@@ -1863,7 +1864,7 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 	caps->max_sq_desc_sz	     = resp_a->max_sq_desc_sz;
 	caps->max_rq_desc_sz	     = resp_a->max_rq_desc_sz;
 	caps->max_srq_desc_sz	     = resp_a->max_srq_desc_sz;
-	caps->cq_entry_sz	     = resp_a->cq_entry_sz;
+	caps->cqe_sz		     = HNS_ROCE_V2_CQE_SIZE;
 
 	caps->mtpt_entry_sz	     = resp_b->mtpt_entry_sz;
 	caps->irrl_entry_sz	     = resp_b->irrl_entry_sz;
@@ -1994,6 +1995,7 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 	if (hr_dev->pci_dev->revision >= PCI_REVISION_ID_HIP09) {
 		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
+		caps->cqe_sz = HNS_ROCE_V3_CQE_SIZE;
 	}
 
 	calc_pg_sz(caps->num_qps, caps->qpc_entry_sz, caps->qpc_hop_num,
@@ -2772,8 +2774,7 @@ static int hns_roce_v2_mw_write_mtpt(void *mb_buf, struct hns_roce_mw *mw)
 
 static void *get_cqe_v2(struct hns_roce_cq *hr_cq, int n)
 {
-	return hns_roce_buf_offset(hr_cq->mtr.kmem,
-				   n * HNS_ROCE_V2_CQE_ENTRY_SIZE);
+	return hns_roce_buf_offset(hr_cq->mtr.kmem, n * hr_cq->cqe_size);
 }
 
 static void *get_sw_cqe_v2(struct hns_roce_cq *hr_cq, int n)
@@ -2873,6 +2874,10 @@ static void hns_roce_v2_write_cqc(struct hns_roce_dev *hr_dev,
 	roce_set_field(cq_context->byte_8_cqn, V2_CQC_BYTE_8_CQN_M,
 		       V2_CQC_BYTE_8_CQN_S, hr_cq->cqn);
 
+	roce_set_field(cq_context->byte_8_cqn, V2_CQC_BYTE_8_CQE_SIZE_M,
+		       V2_CQC_BYTE_8_CQE_SIZE_S, hr_cq->cqe_size ==
+		       HNS_ROCE_V3_CQE_SIZE ? 1 : 0);
+
 	cq_context->cqe_cur_blk_addr = cpu_to_le32(to_hr_hw_page_addr(mtts[0]));
 
 	roce_set_field(cq_context->byte_16_hop_addr,
@@ -3040,7 +3045,8 @@ static int hns_roce_v2_sw_poll_cq(struct hns_roce_cq *hr_cq, int num_entries,
 }
 
 static void get_cqe_status(struct hns_roce_dev *hr_dev, struct hns_roce_qp *qp,
-			   struct hns_roce_v2_cqe *cqe, struct ib_wc *wc)
+			   struct hns_roce_cq *cq, struct hns_roce_v2_cqe *cqe,
+			   struct ib_wc *wc)
 {
 	static const struct {
 		u32 cqe_status;
@@ -3081,7 +3087,7 @@ static void get_cqe_status(struct hns_roce_dev *hr_dev, struct hns_roce_qp *qp,
 
 	ibdev_err(&hr_dev->ib_dev, "error cqe status 0x%x:\n", cqe_status);
 	print_hex_dump(KERN_ERR, "", DUMP_PREFIX_NONE, 16, 4, cqe,
-		       sizeof(*cqe), false);
+		       cq->cqe_size, false);
 
 	/*
 	 * For hns ROCEE, GENERAL_ERR is an error type that is not defined in
@@ -3178,7 +3184,7 @@ static int hns_roce_v2_poll_one(struct hns_roce_cq *hr_cq,
 		++wq->tail;
 	}
 
-	get_cqe_status(hr_dev, *cur_qp, cqe, wc);
+	get_cqe_status(hr_dev, *cur_qp, hr_cq, cqe, wc);
 	if (unlikely(wc->status != IB_WC_SUCCESS))
 		return 0;
 
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
index 9ed3339..a78e5b3 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
@@ -86,7 +86,6 @@
 #define HNS_ROCE_V2_MTPT_ENTRY_SZ		64
 #define HNS_ROCE_V2_MTT_ENTRY_SZ		64
 #define HNS_ROCE_V2_IDX_ENTRY_SZ		4
-#define HNS_ROCE_V2_CQE_ENTRY_SIZE		32
 #define HNS_ROCE_V2_SCCC_ENTRY_SZ		32
 #define HNS_ROCE_V2_QPC_TIMER_ENTRY_SZ		PAGE_SIZE
 #define HNS_ROCE_V2_CQC_TIMER_ENTRY_SZ		PAGE_SIZE
@@ -311,6 +310,9 @@ struct hns_roce_v2_cq_context {
 #define	V2_CQC_BYTE_8_CQN_S 0
 #define V2_CQC_BYTE_8_CQN_M GENMASK(23, 0)
 
+#define V2_CQC_BYTE_8_CQE_SIZE_S 27
+#define V2_CQC_BYTE_8_CQE_SIZE_M GENMASK(28, 27)
+
 #define	V2_CQC_BYTE_16_CQE_CUR_BLK_ADDR_S 0
 #define V2_CQC_BYTE_16_CQE_CUR_BLK_ADDR_M GENMASK(19, 0)
 
@@ -898,6 +900,7 @@ struct hns_roce_v2_cqe {
 	u8	smac[4];
 	__le32	byte_28;
 	__le32	byte_32;
+	__le32	rsv[8];
 };
 
 #define	V2_CQE_BYTE_4_OPCODE_S 0
@@ -1573,7 +1576,7 @@ struct hns_roce_query_pf_caps_a {
 	u8 max_sq_desc_sz;
 	u8 max_rq_desc_sz;
 	u8 max_srq_desc_sz;
-	u8 cq_entry_sz;
+	u8 cqe_sz;
 };
 
 struct hns_roce_query_pf_caps_b {
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 98945df..2871d7e 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -324,6 +324,8 @@ static int hns_roce_alloc_ucontext(struct ib_ucontext *uctx,
 		mutex_init(&context->page_mutex);
 	}
 
+	resp.cqe_size = hr_dev->caps.cqe_sz;
+
 	ret = ib_copy_to_udata(udata, &resp, sizeof(resp));
 	if (ret)
 		goto error_fail_copy_to_udata;
diff --git a/include/uapi/rdma/hns-abi.h b/include/uapi/rdma/hns-abi.h
index 5c38758..99a11b6 100644
--- a/include/uapi/rdma/hns-abi.h
+++ b/include/uapi/rdma/hns-abi.h
@@ -39,6 +39,8 @@
 struct hns_roce_ib_create_cq {
 	__aligned_u64 buf_addr;
 	__aligned_u64 db_addr;
+	__u32 cqe_size;
+	__u32 reserved;
 };
 
 struct hns_roce_ib_create_cq_resp {
@@ -74,6 +76,8 @@ struct hns_roce_ib_create_qp_resp {
 struct hns_roce_ib_alloc_ucontext_resp {
 	__u32	qp_tab_size;
 	__u32	cap_flags;
+	__u32	cqe_size;
+	__u32	reserved;
 };
 
 struct hns_roce_ib_alloc_pd_resp {
-- 
2.8.1

[PATCH v2 for-next 4/4] RDMA/hns: Add support for QPC in size of 512 Bytes

[Weihang Li]

From: Wenpeng Liang <liangwenpeng@huawei.com>

The new version of RoCEE supports using QPC in size of 256B or 512B, so
that HIP09 can supports new congestion control algorithms by using QPC in
larger size.

Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Weihang Li <liweihang@huawei.com>
---
 drivers/infiniband/hw/hns/hns_roce_device.h |  5 ++++-
 drivers/infiniband/hw/hns/hns_roce_hw_v1.c  |  2 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v1.h  |  2 +-
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c  | 26 +++++++++++++++++---------
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h  |  4 ++--
 drivers/infiniband/hw/hns/hns_roce_main.c   |  2 +-
 6 files changed, 26 insertions(+), 15 deletions(-)

diff --git a/drivers/infiniband/hw/hns/hns_roce_device.h b/drivers/infiniband/hw/hns/hns_roce_device.h
index f4ffc87..0fcf650 100644
--- a/drivers/infiniband/hw/hns/hns_roce_device.h
+++ b/drivers/infiniband/hw/hns/hns_roce_device.h
@@ -86,6 +86,9 @@
 #define HNS_ROCE_V2_CQE_SIZE 32
 #define HNS_ROCE_V3_CQE_SIZE 64
 
+#define HNS_ROCE_V2_QPC_SZ 256
+#define HNS_ROCE_V3_QPC_SZ 512
+
 #define HNS_ROCE_SL_SHIFT			28
 #define HNS_ROCE_TCLASS_SHIFT			20
 #define HNS_ROCE_FLOW_LABEL_MASK		0xfffff
@@ -805,7 +808,7 @@ struct hns_roce_caps {
 	u32		page_size_cap;
 	u32		reserved_lkey;
 	int		mtpt_entry_sz;
-	int		qpc_entry_sz;
+	int		qpc_sz;
 	int		irrl_entry_sz;
 	int		trrl_entry_sz;
 	int		cqc_entry_sz;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
index e053421..cda0145 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.c
@@ -1471,7 +1471,7 @@ static int hns_roce_v1_profile(struct hns_roce_dev *hr_dev)
 	caps->max_qp_dest_rdma	= HNS_ROCE_V1_MAX_QP_DEST_RDMA;
 	caps->max_sq_desc_sz	= HNS_ROCE_V1_MAX_SQ_DESC_SZ;
 	caps->max_rq_desc_sz	= HNS_ROCE_V1_MAX_RQ_DESC_SZ;
-	caps->qpc_entry_sz	= HNS_ROCE_V1_QPC_ENTRY_SIZE;
+	caps->qpc_sz		= HNS_ROCE_V1_QPC_SIZE;
 	caps->irrl_entry_sz	= HNS_ROCE_V1_IRRL_ENTRY_SIZE;
 	caps->cqc_entry_sz	= HNS_ROCE_V1_CQC_ENTRY_SIZE;
 	caps->mtpt_entry_sz	= HNS_ROCE_V1_MTPT_ENTRY_SIZE;
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
index 5996892..ffd0156 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v1.h
@@ -68,7 +68,7 @@
 #define HNS_ROCE_V1_COMP_EQE_NUM			0x8000
 #define HNS_ROCE_V1_ASYNC_EQE_NUM			0x400
 
-#define HNS_ROCE_V1_QPC_ENTRY_SIZE			256
+#define HNS_ROCE_V1_QPC_SIZE				256
 #define HNS_ROCE_V1_IRRL_ENTRY_SIZE			8
 #define HNS_ROCE_V1_CQC_ENTRY_SIZE			64
 #define HNS_ROCE_V1_MTPT_ENTRY_SIZE			64
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
index ea20d82..2ced155 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.c
@@ -1683,7 +1683,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 	caps->max_sq_desc_sz	= HNS_ROCE_V2_MAX_SQ_DESC_SZ;
 	caps->max_rq_desc_sz	= HNS_ROCE_V2_MAX_RQ_DESC_SZ;
 	caps->max_srq_desc_sz	= HNS_ROCE_V2_MAX_SRQ_DESC_SZ;
-	caps->qpc_entry_sz	= HNS_ROCE_V2_QPC_ENTRY_SZ;
+	caps->qpc_sz		= HNS_ROCE_V2_QPC_SZ;
 	caps->irrl_entry_sz	= HNS_ROCE_V2_IRRL_ENTRY_SZ;
 	caps->trrl_entry_sz	= HNS_ROCE_V2_EXT_ATOMIC_TRRL_ENTRY_SZ;
 	caps->cqc_entry_sz	= HNS_ROCE_V2_CQC_ENTRY_SZ;
@@ -1772,6 +1772,7 @@ static void set_default_caps(struct hns_roce_dev *hr_dev)
 		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->cqe_sz = HNS_ROCE_V3_CQE_SIZE;
+		caps->qpc_sz = HNS_ROCE_V3_QPC_SZ;
 	}
 }
 
@@ -1874,7 +1875,7 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 	caps->idx_entry_sz	     = resp_b->idx_entry_sz;
 	caps->sccc_entry_sz	     = resp_b->scc_ctx_entry_sz;
 	caps->max_mtu		     = resp_b->max_mtu;
-	caps->qpc_entry_sz	     = le16_to_cpu(resp_b->qpc_entry_sz);
+	caps->qpc_sz		     = HNS_ROCE_V2_QPC_SZ;
 	caps->min_cqes		     = resp_b->min_cqes;
 	caps->min_wqes		     = resp_b->min_wqes;
 	caps->page_size_cap	     = le32_to_cpu(resp_b->page_size_cap);
@@ -1996,9 +1997,10 @@ static int hns_roce_query_pf_caps(struct hns_roce_dev *hr_dev)
 		caps->ceqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->aeqe_size = HNS_ROCE_V3_EQE_SIZE;
 		caps->cqe_sz = HNS_ROCE_V3_CQE_SIZE;
+		caps->qpc_sz = HNS_ROCE_V3_QPC_SZ;
 	}
 
-	calc_pg_sz(caps->num_qps, caps->qpc_entry_sz, caps->qpc_hop_num,
+	calc_pg_sz(caps->num_qps, caps->qpc_sz, caps->qpc_hop_num,
 		   caps->qpc_bt_num, &caps->qpc_buf_pg_sz, &caps->qpc_ba_pg_sz,
 		   HEM_TYPE_QPC);
 	calc_pg_sz(caps->num_mtpts, caps->mtpt_entry_sz, caps->mpt_hop_num,
@@ -3535,16 +3537,21 @@ static int hns_roce_v2_clear_hem(struct hns_roce_dev *hr_dev,
 
 static int hns_roce_v2_qp_modify(struct hns_roce_dev *hr_dev,
 				 struct hns_roce_v2_qp_context *context,
+				 struct hns_roce_v2_qp_context *qpc_mask,
 				 struct hns_roce_qp *hr_qp)
 {
 	struct hns_roce_cmd_mailbox *mailbox;
+	int qpc_size;
 	int ret;
 
 	mailbox = hns_roce_alloc_cmd_mailbox(hr_dev);
 	if (IS_ERR(mailbox))
 		return PTR_ERR(mailbox);
 
-	memcpy(mailbox->buf, context, sizeof(*context) * 2);
+	/* The qpc size of HIP08 is only 256B, which is half of HIP09 */
+	qpc_size = hr_dev->caps.qpc_sz;
+	memcpy(mailbox->buf, context, qpc_size);
+	memcpy(mailbox->buf + qpc_size, qpc_mask, qpc_size);
 
 	ret = hns_roce_cmd_mbox(hr_dev, mailbox->dma, 0, hr_qp->qpn, 0,
 				HNS_ROCE_CMD_MODIFY_QPC,
@@ -4330,7 +4337,7 @@ static int hns_roce_v2_set_abs_fields(struct ib_qp *ibqp,
 	}
 
 	if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) {
-		memset(qpc_mask, 0, sizeof(*qpc_mask));
+		memset(qpc_mask, 0, hr_dev->caps.qpc_sz);
 		modify_qp_reset_to_init(ibqp, attr, attr_mask, context,
 					qpc_mask);
 	} else if (cur_state == IB_QPS_INIT && new_state == IB_QPS_INIT) {
@@ -4553,8 +4560,9 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 	 * we should set all bits of the relevant fields in context mask to
 	 * 0 at the same time, else set them to 0x1.
 	 */
-	memset(context, 0, sizeof(*context));
-	memset(qpc_mask, 0xff, sizeof(*qpc_mask));
+	memset(context, 0, hr_dev->caps.qpc_sz);
+	memset(qpc_mask, 0xff, hr_dev->caps.qpc_sz);
+
 	ret = hns_roce_v2_set_abs_fields(ibqp, attr, attr_mask, cur_state,
 					 new_state, context, qpc_mask);
 	if (ret)
@@ -4604,7 +4612,7 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 		       V2_QPC_BYTE_60_QP_ST_S, 0);
 
 	/* SW pass context to HW */
-	ret = hns_roce_v2_qp_modify(hr_dev, ctx, hr_qp);
+	ret = hns_roce_v2_qp_modify(hr_dev, context, qpc_mask, hr_qp);
 	if (ret) {
 		ibdev_err(ibdev, "failed to modify QP, ret = %d\n", ret);
 		goto out;
@@ -4667,7 +4675,7 @@ static int hns_roce_v2_query_qpc(struct hns_roce_dev *hr_dev,
 	if (ret)
 		goto out;
 
-	memcpy(hr_context, mailbox->buf, sizeof(*hr_context));
+	memcpy(hr_context, mailbox->buf, hr_dev->caps.qpc_sz);
 
 out:
 	hns_roce_free_cmd_mailbox(hr_dev, mailbox);
diff --git a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
index a78e5b3..1595ab6 100644
--- a/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
+++ b/drivers/infiniband/hw/hns/hns_roce_hw_v2.h
@@ -77,7 +77,6 @@
 #define HNS_ROCE_V2_MAX_SQ_DESC_SZ		64
 #define HNS_ROCE_V2_MAX_RQ_DESC_SZ		16
 #define HNS_ROCE_V2_MAX_SRQ_DESC_SZ		64
-#define HNS_ROCE_V2_QPC_ENTRY_SZ		256
 #define HNS_ROCE_V2_IRRL_ENTRY_SZ		64
 #define HNS_ROCE_V2_TRRL_ENTRY_SZ		48
 #define HNS_ROCE_V2_EXT_ATOMIC_TRRL_ENTRY_SZ	100
@@ -516,6 +515,7 @@ struct hns_roce_v2_qp_context {
 	__le32	byte_248_ack_psn;
 	__le32	byte_252_err_txcqn;
 	__le32	byte_256_sqflush_rqcqe;
+	__le32	ext[64];
 };
 
 #define	V2_QPC_BYTE_4_TST_S 0
@@ -1588,7 +1588,7 @@ struct hns_roce_query_pf_caps_b {
 	u8 idx_entry_sz;
 	u8 scc_ctx_entry_sz;
 	u8 max_mtu;
-	__le16 qpc_entry_sz;
+	__le16 qpc_sz;
 	__le16 qpc_timer_entry_sz;
 	__le16 cqc_timer_entry_sz;
 	u8 min_cqes;
diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
index 2871d7e..b1b0f18 100644
--- a/drivers/infiniband/hw/hns/hns_roce_main.c
+++ b/drivers/infiniband/hw/hns/hns_roce_main.c
@@ -590,7 +590,7 @@ static int hns_roce_init_hem(struct hns_roce_dev *hr_dev)
 	}
 
 	ret = hns_roce_init_hem_table(hr_dev, &hr_dev->qp_table.qp_table,
-				      HEM_TYPE_QPC, hr_dev->caps.qpc_entry_sz,
+				      HEM_TYPE_QPC, hr_dev->caps.qpc_sz,
 				      hr_dev->caps.num_qps, 1);
 	if (ret) {
 		dev_err(dev, "Failed to init QP context memory, aborting.\n");
-- 
2.8.1

Re: [PATCH v2 for-next 1/4] RDMA/hns: Export hardware capability flags to userspace

[Jason Gunthorpe]

On Thu, Aug 20, 2020 at 09:17:46PM +0800, Weihang Li wrote:
> From: Xi Wang <wangxi11@huawei.com>
> 
> The libhns in userspace for HIP09 will use the hardware's capability to
> enable some features. So export the hardware capablility flags to userspace
> by reusing the reserved fields in structure
> "hns_roce_ib_alloc_ucontext_resp".
> 
> Signed-off-by: Xi Wang <wangxi11@huawei.com>
> Signed-off-by: Weihang Li <liweihang@huawei.com>
>  drivers/infiniband/hw/hns/hns_roce_main.c | 1 +
>  include/uapi/rdma/hns-abi.h               | 2 +-
>  2 files changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
> index 5907cfd..98945df 100644
> +++ b/drivers/infiniband/hw/hns/hns_roce_main.c
> @@ -313,6 +313,7 @@ static int hns_roce_alloc_ucontext(struct ib_ucontext *uctx,
>  		return -EAGAIN;
>  
>  	resp.qp_tab_size = hr_dev->caps.num_qps;
> +	resp.cap_flags = (u32)hr_dev->caps.flags;

This makes all the HNS_ROCE_CAP_FLAG_* values uapi, they need to be
moved to the uapi header too.

Or rethink this to only expose what you want to be uapi.

Jason

Re: [PATCH v2 for-next 1/4] RDMA/hns: Export hardware capability flags to userspace

[liweihang]

On 2020/8/27 20:40, Jason Gunthorpe wrote:
> On Thu, Aug 20, 2020 at 09:17:46PM +0800, Weihang Li wrote:
>> From: Xi Wang <wangxi11@huawei.com>
>>
>> The libhns in userspace for HIP09 will use the hardware's capability to
>> enable some features. So export the hardware capablility flags to userspace
>> by reusing the reserved fields in structure
>> "hns_roce_ib_alloc_ucontext_resp".
>>
>> Signed-off-by: Xi Wang <wangxi11@huawei.com>
>> Signed-off-by: Weihang Li <liweihang@huawei.com>
>>  drivers/infiniband/hw/hns/hns_roce_main.c | 1 +
>>  include/uapi/rdma/hns-abi.h               | 2 +-
>>  2 files changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/infiniband/hw/hns/hns_roce_main.c b/drivers/infiniband/hw/hns/hns_roce_main.c
>> index 5907cfd..98945df 100644
>> +++ b/drivers/infiniband/hw/hns/hns_roce_main.c
>> @@ -313,6 +313,7 @@ static int hns_roce_alloc_ucontext(struct ib_ucontext *uctx,
>>  		return -EAGAIN;
>>  
>>  	resp.qp_tab_size = hr_dev->caps.num_qps;
>> +	resp.cap_flags = (u32)hr_dev->caps.flags;
> 
> This makes all the HNS_ROCE_CAP_FLAG_* values uapi, they need to be
> moved to the uapi header too.
> 
> Or rethink this to only expose what you want to be uapi.
> 
> Jason
> 

Hi Jason,

Thanks for your comments, we will only expose the needed flag to be
uapi.

Considering that this cap_flags field is not necessary for this series,
I will drop #1 and send a new version. And the modification about
cap_flags will be sent later.

Weihang