ConnectX-3 RSS Implementation

index bfdd4e117813..e084eb88d44b 100644
@@ -81,6 +81,8 @@ static const char mlx4_ib_version[] =
 	DRV_VERSION "\n";
 
 static void do_slave_init(struct mlx4_ib_dev *ibdev, int slave, int do_init);
+static enum rdma_link_layer mlx4_ib_port_link_layer(struct ib_device *device,
+						    u8 port_num);
 
 static struct workqueue_struct *wq;
 

@@ -552,6 +554,11 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 	props->timestamp_mask = 0xFFFFFFFFFFFFULL;
 	props->max_ah = INT_MAX;
 
+	if ((dev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_RSS) &&
+	    (mlx4_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET ||
+	     mlx4_ib_port_link_layer(ibdev, 2) == IB_LINK_LAYER_ETHERNET))
+		props->max_wq_type_rq = MLX4_IB_MAX_CTX_WQS;
+
 	if (!mlx4_is_slave(dev->dev))
 		err = mlx4_get_internal_clock_params(dev->dev, &clock_params);
 

@@ -1087,6 +1094,21 @@ static struct ib_ucontext *mlx4_ib_alloc_ucontext(struct ib_device *ibdev,
 		return ERR_PTR(-EFAULT);
 	}
 
+	context->wq_use_cnt = MLX4_IB_MAX_CTX_WQS; /* WQs pool is empty */
+	context->base_wqn = 0xFFFFFFFF; /* invalid wqn */
+
+	if ((rdma_port_get_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) ||
+	    (rdma_port_get_link_layer(ibdev, 2) == IB_LINK_LAYER_ETHERNET)) {
+		if (!mlx4_qp_reserve_range(dev->dev, MLX4_IB_MAX_CTX_WQS,
+					   MLX4_IB_MAX_CTX_WQS,
+					   &context->base_wqn, 0,
+					   MLX4_RES_USAGE_USER_VERBS)) {
+			context->wq_use_cnt = 0;
+		} else {
+			pr_warn("Failed to reserve WQNs for user context\n");
+		}
+	}
+
 	return &context->ibucontext;
 }
 

@@ -1094,6 +1116,10 @@ static int mlx4_ib_dealloc_ucontext(struct ib_ucontext *ibcontext)
 {
 	struct mlx4_ib_ucontext *context = to_mucontext(ibcontext);
 
+	if (context->base_wqn != 0xFFFFFFFF)
+		mlx4_qp_release_range(to_mdev(ibcontext->device)->dev,
+				      context->base_wqn, MLX4_IB_MAX_CTX_WQS);
+
 	mlx4_uar_free(to_mdev(ibcontext->device)->dev, &context->uar);
 	kfree(context);
 

@@ -2720,6 +2746,20 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	ibdev->ib_dev.get_dev_fw_str    = get_fw_ver_str;
 	ibdev->ib_dev.disassociate_ucontext = mlx4_ib_disassociate_ucontext;
 
+	if ((dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_RSS) &&
+	    ((mlx4_ib_port_link_layer(&ibdev->ib_dev, 1) ==
+	    IB_LINK_LAYER_ETHERNET) ||
+	    (mlx4_ib_port_link_layer(&ibdev->ib_dev, 2) ==
+	    IB_LINK_LAYER_ETHERNET))) {
+		ibdev->ib_dev.create_wq		= mlx4_ib_create_wq;
+		ibdev->ib_dev.modify_wq		= mlx4_ib_modify_wq;
+		ibdev->ib_dev.destroy_wq	= mlx4_ib_destroy_wq;
+		ibdev->ib_dev.uverbs_ex_cmd_mask |=
+			(1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) |
+			(1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) |
+			(1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ);
+	}
+
 	if (!mlx4_is_slave(ibdev->dev)) {
 		ibdev->ib_dev.alloc_fmr		= mlx4_ib_fmr_alloc;
 		ibdev->ib_dev.map_phys_fmr	= mlx4_ib_map_phys_fmr;

index a6337f3161cf..6d8ef6506f09 100644
@@ -82,12 +82,16 @@ struct mlx4_ib_vma_private_data {
 	struct vm_area_struct *vma;
 };
 
+#define MLX4_IB_MAX_CTX_WQS 128
+
 struct mlx4_ib_ucontext {
 	struct ib_ucontext	ibucontext;
 	struct mlx4_uar		uar;
 	struct list_head	db_page_list;
 	struct mutex		db_page_mutex;
 	struct mlx4_ib_vma_private_data hw_bar_info[HW_BAR_COUNT];
+	int			base_wqn;
+	int			wq_use_cnt;
 };
 
 struct mlx4_ib_pd {

@@ -290,7 +294,10 @@ struct mlx4_roce_smac_vlan_info {
 };
 
 struct mlx4_ib_qp {
-	struct ib_qp		ibqp;
+	union {
+		struct ib_qp	ibqp;
+		struct ib_wq	ibwq;
+	};
 	struct mlx4_qp		mqp;
 	struct mlx4_buf		buf;
 

@@ -329,6 +336,8 @@ struct mlx4_ib_qp {
 	struct list_head	cq_recv_list;
 	struct list_head	cq_send_list;
 	struct counter_index	*counter_index;
+	/* Number of RSS QP parents that uses this WQ */
+	u32			rss_usecnt;
 };
 
 struct mlx4_ib_srq {

@@ -893,4 +902,11 @@ void mlx4_sched_ib_sl2vl_update_work(struct mlx4_ib_dev *ibdev,
 
 void mlx4_ib_sl2vl_update(struct mlx4_ib_dev *mdev, int port);
 
+struct ib_wq *mlx4_ib_create_wq(struct ib_pd *pd,
+				struct ib_wq_init_attr *init_attr,
+				struct ib_udata *udata);
+int mlx4_ib_destroy_wq(struct ib_wq *wq);
+int mlx4_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr,
+		      u32 wq_attr_mask, struct ib_udata *udata);
+
 #endif /* MLX4_IB_H */

657 lines not on this page · raw message

index bf3bdba2f326..12b23b80dae5 100644
@@ -105,4 +105,16 @@ struct mlx4_ib_create_qp {
 	__u8	reserved;
 };
 
+struct mlx4_ib_create_wq {
+	__u64	buf_addr;
+	__u64	db_addr;
+	__u32   comp_mask;
+	__u32   reserved;
+};
+
+struct mlx4_ib_modify_wq {
+	__u32	comp_mask;
+	__u32	reserved;
+};
+
 #endif /* MLX4_ABI_USER_H */

index e084eb88d44b..97b82b0c680c 100644
@@ -2754,10 +2754,16 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 		ibdev->ib_dev.create_wq		= mlx4_ib_create_wq;
 		ibdev->ib_dev.modify_wq		= mlx4_ib_modify_wq;
 		ibdev->ib_dev.destroy_wq	= mlx4_ib_destroy_wq;
+		ibdev->ib_dev.create_rwq_ind_table  =
+			mlx4_ib_create_rwq_ind_table;
+		ibdev->ib_dev.destroy_rwq_ind_table =
+			mlx4_ib_destroy_rwq_ind_table;
 		ibdev->ib_dev.uverbs_ex_cmd_mask |=
-			(1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) |
-			(1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) |
-			(1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ);
+			(1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ)	  |
+			(1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ)	  |
+			(1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ)	  |
+			(1ull << IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL) |
+			(1ull << IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL);
 	}
 
 	if (!mlx4_is_slave(ibdev->dev)) {

index 6d8ef6506f09..352b8bc975d6 100644
@@ -909,4 +909,10 @@ int mlx4_ib_destroy_wq(struct ib_wq *wq);
 int mlx4_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr,
 		      u32 wq_attr_mask, struct ib_udata *udata);
 
+struct ib_rwq_ind_table
+*mlx4_ib_create_rwq_ind_table(struct ib_device *device,
+			      struct ib_rwq_ind_table_init_attr *init_attr,
+			      struct ib_udata *udata);
+int mlx4_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table);
+
 #endif /* MLX4_IB_H */

72 lines not on this page · raw message

index 12b23b80dae5..2371d41156b1 100644
@@ -117,4 +117,8 @@ struct mlx4_ib_modify_wq {
 	__u32	reserved;
 };
 
+struct mlx4_ib_create_rwq_ind_tbl_resp {
+	__u32	response_length;
+	__u32	reserved;
+};
 #endif /* MLX4_ABI_USER_H */

index 352b8bc975d6..6a71e159d1b2 100644
@@ -46,6 +46,7 @@
 
 #include <linux/mlx4/device.h>
 #include <linux/mlx4/doorbell.h>
+#include <linux/mlx4/qp.h>
 
 #define MLX4_IB_DRV_NAME	"mlx4_ib"
 

@@ -293,6 +294,12 @@ struct mlx4_roce_smac_vlan_info {
 	int update_vid;
 };
 
+struct mlx4_ib_rss {
+	unsigned int		base_qpn_tbl_sz;
+	u8			flags;
+	u8			rss_key[MLX4_EN_RSS_KEY_SIZE];
+};
+
 struct mlx4_ib_qp {
 	union {
 		struct ib_qp	ibqp;

@@ -338,6 +345,7 @@ struct mlx4_ib_qp {
 	struct counter_index	*counter_index;
 	/* Number of RSS QP parents that uses this WQ */
 	u32			rss_usecnt;
+	struct mlx4_ib_rss	*rss_ctx;
 };
 
 struct mlx4_ib_srq {

567 lines not on this page · raw message

index 2371d41156b1..61c868c35578 100644
@@ -95,6 +95,16 @@ struct mlx4_ib_create_srq_resp {
 	__u32	reserved;
 };
 
+struct mlx4_ib_create_qp_rss {
+	__u64   rx_hash_fields_mask;
+	__u8    rx_hash_function;
+	__u8    rx_key_len;
+	__u8    reserved[6];
+	__u8    rx_hash_key[40];
+	__u32   comp_mask;
+	__u32   reserved1;
+};
+
 struct mlx4_ib_create_qp {
 	__u64	buf_addr;
 	__u64	db_addr;

@@ -121,4 +131,27 @@ struct mlx4_ib_create_rwq_ind_tbl_resp {
 	__u32	response_length;
 	__u32	reserved;
 };
+
+/* RX Hash function flags */
+enum mlx4_ib_rx_hash_function_flags {
+	MLX4_IB_RX_HASH_FUNC_TOEPLITZ	= 1 << 0,
+};
+
+/*
+ * RX Hash flags, these flags allows to set which incoming packet's field should
+ * participates in RX Hash. Each flag represent certain packet's field,
+ * when the flag is set the field that is represented by the flag will
+ * participate in RX Hash calculation.
+ */
+enum mlx4_ib_rx_hash_fields {
+	MLX4_IB_RX_HASH_SRC_IPV4	= 1 << 0,
+	MLX4_IB_RX_HASH_DST_IPV4	= 1 << 1,
+	MLX4_IB_RX_HASH_SRC_IPV6	= 1 << 2,
+	MLX4_IB_RX_HASH_DST_IPV6	= 1 << 3,
+	MLX4_IB_RX_HASH_SRC_PORT_TCP	= 1 << 4,
+	MLX4_IB_RX_HASH_DST_PORT_TCP	= 1 << 5,
+	MLX4_IB_RX_HASH_SRC_PORT_UDP	= 1 << 6,
+	MLX4_IB_RX_HASH_DST_PORT_UDP	= 1 << 7
+};
+
 #endif /* MLX4_ABI_USER_H */

index 97b82b0c680c..6c11aed181a5 100644
@@ -556,8 +556,13 @@ static int mlx4_ib_query_device(struct ib_device *ibdev,
 
 	if ((dev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_RSS) &&
 	    (mlx4_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET ||
-	     mlx4_ib_port_link_layer(ibdev, 2) == IB_LINK_LAYER_ETHERNET))
+	     mlx4_ib_port_link_layer(ibdev, 2) == IB_LINK_LAYER_ETHERNET)) {
+		props->rss_caps.max_rwq_indirection_tables = 128;
+		props->rss_caps.max_rwq_indirection_table_size =
+			dev->dev->caps.max_rss_tbl_sz;
+		props->rss_caps.supported_qpts = 1 << IB_QPT_RAW_PACKET;
 		props->max_wq_type_rq = MLX4_IB_MAX_CTX_WQS;
+	}
 
 	if (!mlx4_is_slave(dev->dev))
 		err = mlx4_get_internal_clock_params(dev->dev, &clock_params);