[PATCH 0/5] support item NSH matching

[PATCH 0/5] support item NSH matching

[Haifei Luo]

NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
NSH fields matching is not supported.  
Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.
Add support for matching item NSH. The CLI is: nsh
Add support for HCA attribute query of NSH.
Enhance the validation for the case matching item NSH is supported. 
Add NSH support in net/mlx5. 


Haifei Luo (5):
  app/testpmd: support for VXLAN-GPE's next protocol
  common/mlx5: extend HCA attribute query for NSH
  net/mlx5: enhance the validation for item VXLAN-GPE
  app/testpmd: support for NSH flow item
  net/mlx5: add support for item NSH

 app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
 drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
 drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
 drivers/common/mlx5/mlx5_prm.h       |  4 ++-
 drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
 drivers/net/mlx5/mlx5_flow.h         |  6 ++++
 drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
 7 files changed, 96 insertions(+), 9 deletions(-)

-- 
2.34.1

[PATCH 1/5] app/testpmd: support for VXLAN-GPE's next protocol

[Haifei Luo]

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Example:
flow create 0 transfer group 1 pattern eth / ipv6 / udp dst is 4790
/ vxlan-gpe protocol is 0x04 / eth / ipv4  / tcp / end
actions port_id id 1 / end

Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com>
Signed-off-by: Haifei Luo <haifeil@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 94827bcc4a..d25f941259 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -385,6 +385,7 @@ enum index {
 	ITEM_GENEVE_OPTLEN,
 	ITEM_VXLAN_GPE,
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_ARP_ETH_IPV4,
 	ITEM_ARP_ETH_IPV4_SHA,
 	ITEM_ARP_ETH_IPV4_SPA,
@@ -1758,6 +1759,7 @@ static const enum index item_geneve[] = {
 
 static const enum index item_vxlan_gpe[] = {
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_NEXT,
 	ZERO,
 };
@@ -4804,6 +4806,14 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
 					     hdr.vni)),
 	},
+	[ITEM_VXLAN_GPE_PROTO] = {
+		.name = "protocol",
+		.help = "VXLAN-GPE next protocol",
+		.next = NEXT(item_vxlan_gpe, NEXT_ENTRY(COMMON_UNSIGNED),
+			     item_param),
+		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
+					     protocol)),
+	},
 	[ITEM_ARP_ETH_IPV4] = {
 		.name = "arp_eth_ipv4",
 		.help = "match ARP header for Ethernet/IPv4",
-- 
2.34.1

[PATCH 2/5] common/mlx5: extend HCA attribute query for NSH

[Haifei Luo]

Add NSH supporting field in two places:
1. New HCA capability indicating NSH is supported
2. New field in "mlx5_ifc_per_protocol_networking_offload_caps_bits"
   structure

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
---
 drivers/common/mlx5/mlx5_devx_cmds.c | 3 +++
 drivers/common/mlx5/mlx5_devx_cmds.h | 1 +
 drivers/common/mlx5/mlx5_prm.h       | 4 +++-
 3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/common/mlx5/mlx5_devx_cmds.c b/drivers/common/mlx5/mlx5_devx_cmds.c
index 66a77159a0..830199212e 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.c
+++ b/drivers/common/mlx5/mlx5_devx_cmds.c
@@ -1313,6 +1313,9 @@ mlx5_devx_cmd_query_hca_attr(void *ctx,
 	attr->tunnel_stateless_gtp = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, tunnel_stateless_gtp);
+	attr->tunnel_stateless_vxlan_gpe_nsh = MLX5_GET
+					(per_protocol_networking_offload_caps,
+					 hcattr, tunnel_stateless_vxlan_gpe_nsh);
 	attr->rss_ind_tbl_cap = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, rss_ind_tbl_cap);
diff --git a/drivers/common/mlx5/mlx5_devx_cmds.h b/drivers/common/mlx5/mlx5_devx_cmds.h
index e071cd841f..11772431ae 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.h
+++ b/drivers/common/mlx5/mlx5_devx_cmds.h
@@ -196,6 +196,7 @@ struct mlx5_hca_attr {
 	uint32_t tunnel_stateless_geneve_rx:1;
 	uint32_t geneve_max_opt_len:1; /* 0x0: 14DW, 0x1: 63DW */
 	uint32_t tunnel_stateless_gtp:1;
+	uint32_t tunnel_stateless_vxlan_gpe_nsh:1;
 	uint32_t max_lso_cap;
 	uint32_t scatter_fcs:1;
 	uint32_t lro_cap:1;
diff --git a/drivers/common/mlx5/mlx5_prm.h b/drivers/common/mlx5/mlx5_prm.h
index 51f426c614..f005877dd7 100644
--- a/drivers/common/mlx5/mlx5_prm.h
+++ b/drivers/common/mlx5/mlx5_prm.h
@@ -1964,7 +1964,9 @@ struct mlx5_ifc_per_protocol_networking_offload_caps_bits {
 	u8 swp_lso[0x1];
 	u8 reserved_at_23[0x8];
 	u8 tunnel_stateless_gtp[0x1];
-	u8 reserved_at_25[0x4];
+	u8 reserved_at_25[0x2];
+	u8 tunnel_stateless_vxlan_gpe_nsh[0x1];
+	u8 reserved_at_28[0x1];
 	u8 max_vxlan_udp_ports[0x8];
 	u8 reserved_at_38[0x6];
 	u8 max_geneve_opt_len[0x1];
-- 
2.34.1

[PATCH 3/5] net/mlx5: enhance the validation for item VXLAN-GPE

[Haifei Luo]

Enhance the validation so that configuring vxlan-gpe's next protocol as NSH
is supported.

1. The spec's protocol can have value and nic_mask's protocol is 0xff.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index e91eb636d0..7de6640ecd 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3198,6 +3198,11 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		uint8_t vni[4];
 	} id = { .vlan_id = 0, };
 
+	struct rte_flow_item_vxlan_gpe nic_mask = {
+		.vni = "\xff\xff\xff",
+		.protocol = 0xff,
+	};
+
 	if (!priv->sh->config.l3_vxlan_en)
 		return rte_flow_error_set(error, ENOTSUP,
 					  RTE_FLOW_ERROR_TYPE_ITEM, item,
@@ -3221,18 +3226,12 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		mask = &rte_flow_item_vxlan_gpe_mask;
 	ret = mlx5_flow_item_acceptable
 		(item, (const uint8_t *)mask,
-		 (const uint8_t *)&rte_flow_item_vxlan_gpe_mask,
+		 (const uint8_t *)&nic_mask,
 		 sizeof(struct rte_flow_item_vxlan_gpe),
 		 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 	if (ret < 0)
 		return ret;
 	if (spec) {
-		if (spec->hdr.proto)
-			return rte_flow_error_set(error, ENOTSUP,
-						  RTE_FLOW_ERROR_TYPE_ITEM,
-						  item,
-						  "VxLAN-GPE protocol"
-						  " not supported");
 		memcpy(&id.vni[1], spec->hdr.vni, 3);
 		memcpy(&id.vni[1], mask->hdr.vni, 3);
 	}
-- 
2.34.1

[PATCH 4/5] app/testpmd: support for NSH flow item

[Haifei Luo]

Add support for item NSH. The CLI is: nsh
Example:
flow create 0 transfer group 1
     pattern eth / ipv6 / udp dst is 4790 /
             vxlan-gpe / nsh / eth / ipv4 / tcp / end
     actions port_id id 1 / end

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index d25f941259..bf93d649e1 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -525,6 +525,7 @@ enum index {
 	ITEM_IB_BTH_PSN,
 	ITEM_IPV6_PUSH_REMOVE_EXT,
 	ITEM_IPV6_PUSH_REMOVE_EXT_TYPE,
+	ITEM_NSH,
 
 	/* Validate/create actions. */
 	ACTIONS,
@@ -1562,6 +1563,7 @@ static const enum index next_item[] = {
 	ITEM_AGGR_AFFINITY,
 	ITEM_TX_QUEUE,
 	ITEM_IB_BTH,
+	ITEM_NSH,
 	END_SET,
 	ZERO,
 };
@@ -2081,6 +2083,11 @@ static const enum index item_ib_bth[] = {
 	ZERO,
 };
 
+static const enum index item_nsh[] = {
+	ITEM_NEXT,
+	ZERO,
+};
+
 static const enum index next_action[] = {
 	ACTION_END,
 	ACTION_VOID,
@@ -5837,6 +5844,15 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_ib_bth,
 						 hdr.psn)),
 	},
+	[ITEM_NSH] = {
+		.name = "nsh",
+		.help = "match NSH header",
+		.priv = PRIV_ITEM(NSH,
+				  sizeof(struct rte_flow_item_nsh)),
+		.next = NEXT(item_nsh),
+		.call = parse_vc,
+	},
+
 	/* Validate/create actions. */
 	[ACTIONS] = {
 		.name = "actions",
-- 
2.34.1

[PATCH 5/5] net/mlx5: add support for item NSH

[Haifei Luo]

1. Add validation for item NSH.
   It will fail if HCA cap for NSH is false.
2. Add item_flags for NSH.
3. For vxlan-gpe if next header is NSH, set next_protocol as NSH.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c    | 39 +++++++++++++++++++++++++++++++++
 drivers/net/mlx5/mlx5_flow.h    |  6 +++++
 drivers/net/mlx5/mlx5_flow_dv.c | 13 ++++++++++-
 3 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 7de6640ecd..0e241acd62 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3905,6 +3905,45 @@ mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 					 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 }
 
+/**
+ * Validate the NSH item.
+ *
+ * @param[in] dev
+ *   Pointer to Ethernet device on which flow rule is being created on.
+ * @param[out] error
+ *   Pointer to error structure.
+ *
+ * @return
+ *   0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+int
+mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+			    const struct rte_flow_item *item,
+			    struct rte_flow_error *error)
+{
+	struct mlx5_priv *priv = dev->data->dev_private;
+
+	if (item->mask) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "NSH fields matching is not supported");
+	}
+
+	if (!priv->sh->config.dv_flow_en) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+					  NULL, "NSH support requires DV flow interface");
+	}
+
+	if (!priv->sh->cdev->config.hca_attr.tunnel_stateless_vxlan_gpe_nsh) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "Current FW does not support matching on NSH");
+	}
+
+	return 0;
+}
+
 static int
 flow_null_validate(struct rte_eth_dev *dev __rte_unused,
 		   const struct rte_flow_attr *attr __rte_unused,
diff --git a/drivers/net/mlx5/mlx5_flow.h b/drivers/net/mlx5/mlx5_flow.h
index 3a97975d69..ccb416e497 100644
--- a/drivers/net/mlx5/mlx5_flow.h
+++ b/drivers/net/mlx5/mlx5_flow.h
@@ -233,6 +233,9 @@ enum mlx5_feature_name {
 /* IB BTH ITEM. */
 #define MLX5_FLOW_ITEM_IB_BTH (1ull << 51)
 
+/* NSH ITEM */
+#define MLX5_FLOW_ITEM_NSH (1ull << 53)
+
 /* Outer Masks. */
 #define MLX5_FLOW_LAYER_OUTER_L3 \
 	(MLX5_FLOW_LAYER_OUTER_L3_IPV4 | MLX5_FLOW_LAYER_OUTER_L3_IPV6)
@@ -2453,6 +2456,9 @@ int mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 				  uint16_t ether_type,
 				  const struct rte_flow_item_ecpri *acc_mask,
 				  struct rte_flow_error *error);
+int mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+				const struct rte_flow_item *item,
+				struct rte_flow_error *error);
 int mlx5_flow_create_mtr_tbls(struct rte_eth_dev *dev,
 			      struct mlx5_flow_meter_info *fm,
 			      uint32_t mtr_idx,
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index a8dd9920e6..4a46793758 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -7815,6 +7815,12 @@ flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 
 			last_item = MLX5_FLOW_ITEM_IB_BTH;
 			break;
+		case RTE_FLOW_ITEM_TYPE_NSH:
+			ret = mlx5_flow_validate_item_nsh(dev, items, error);
+			if (ret < 0)
+				return ret;
+			last_item = MLX5_FLOW_ITEM_NSH;
+			break;
 		default:
 			return rte_flow_error_set(error, ENOTSUP,
 						  RTE_FLOW_ERROR_TYPE_ITEM,
@@ -9720,7 +9726,9 @@ flow_dv_translate_item_vxlan_gpe(void *key, const struct rte_flow_item *item,
 	v_protocol = vxlan_v->hdr.protocol;
 	if (!m_protocol) {
 		/* Force next protocol to ensure next headers parsing. */
-		if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
+		if (pattern_flags & MLX5_FLOW_ITEM_NSH)
+			v_protocol = RTE_VXLAN_GPE_TYPE_NSH;
+		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
 			v_protocol = RTE_VXLAN_GPE_TYPE_ETH;
 		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L3_IPV4)
 			v_protocol = RTE_VXLAN_GPE_TYPE_IPV4;
@@ -13910,6 +13918,9 @@ flow_dv_translate_items(struct rte_eth_dev *dev,
 		flow_dv_translate_item_ib_bth(key, items, tunnel, key_type);
 		last_item = MLX5_FLOW_ITEM_IB_BTH;
 		break;
+	case RTE_FLOW_ITEM_TYPE_NSH:
+		last_item = MLX5_FLOW_ITEM_NSH;
+		break;
 	default:
 		break;
 	}
-- 
2.34.1

[PATCH v2 0/5] support item NSH matching

[Haifei Luo]

NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
NSH fields matching is not supported.  

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Add support for matching item NSH. The CLI is: nsh
Add support for HCA attribute query of NSH.

Enhance the validation for the case matching item NSH is supported. 
Add NSH support in net/mlx5.

V2: Add Ack info in commit message.

Haifei Luo (5):
  app/testpmd: support for VXLAN-GPE's next protocol
  common/mlx5: extend HCA attribute query for NSH
  net/mlx5: enhance the validation for item VXLAN-GPE
  app/testpmd: support for NSH flow item
  net/mlx5: add support for item NSH

 app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
 drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
 drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
 drivers/common/mlx5/mlx5_prm.h       |  4 ++-
 drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
 drivers/net/mlx5/mlx5_flow.h         |  6 ++++
 drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
 7 files changed, 96 insertions(+), 9 deletions(-)

-- 
2.27.0

[PATCH v2 1/5] app/testpmd: support for VXLAN-GPE's next protocol

[Haifei Luo]

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Example:
flow create 0 transfer group 1 pattern eth / ipv6 / udp dst is 4790
/ vxlan-gpe protocol is 0x04 / eth / ipv4  / tcp / end
actions port_id id 1 / end

Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com>
Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 94827bcc4a..d25f941259 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -385,6 +385,7 @@ enum index {
 	ITEM_GENEVE_OPTLEN,
 	ITEM_VXLAN_GPE,
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_ARP_ETH_IPV4,
 	ITEM_ARP_ETH_IPV4_SHA,
 	ITEM_ARP_ETH_IPV4_SPA,
@@ -1758,6 +1759,7 @@ static const enum index item_geneve[] = {
 
 static const enum index item_vxlan_gpe[] = {
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_NEXT,
 	ZERO,
 };
@@ -4804,6 +4806,14 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
 					     hdr.vni)),
 	},
+	[ITEM_VXLAN_GPE_PROTO] = {
+		.name = "protocol",
+		.help = "VXLAN-GPE next protocol",
+		.next = NEXT(item_vxlan_gpe, NEXT_ENTRY(COMMON_UNSIGNED),
+			     item_param),
+		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
+					     protocol)),
+	},
 	[ITEM_ARP_ETH_IPV4] = {
 		.name = "arp_eth_ipv4",
 		.help = "match ARP header for Ethernet/IPv4",
-- 
2.27.0

[PATCH v2 2/5] common/mlx5: extend HCA attribute query for NSH

[Haifei Luo]

Add NSH supporting field in two places:
1. New HCA capability indicating NSH is supported
2. New field in "mlx5_ifc_per_protocol_networking_offload_caps_bits"
   structure

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
---
 drivers/common/mlx5/mlx5_devx_cmds.c | 3 +++
 drivers/common/mlx5/mlx5_devx_cmds.h | 1 +
 drivers/common/mlx5/mlx5_prm.h       | 4 +++-
 3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/common/mlx5/mlx5_devx_cmds.c b/drivers/common/mlx5/mlx5_devx_cmds.c
index 66a77159a0..830199212e 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.c
+++ b/drivers/common/mlx5/mlx5_devx_cmds.c
@@ -1313,6 +1313,9 @@ mlx5_devx_cmd_query_hca_attr(void *ctx,
 	attr->tunnel_stateless_gtp = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, tunnel_stateless_gtp);
+	attr->tunnel_stateless_vxlan_gpe_nsh = MLX5_GET
+					(per_protocol_networking_offload_caps,
+					 hcattr, tunnel_stateless_vxlan_gpe_nsh);
 	attr->rss_ind_tbl_cap = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, rss_ind_tbl_cap);
diff --git a/drivers/common/mlx5/mlx5_devx_cmds.h b/drivers/common/mlx5/mlx5_devx_cmds.h
index e071cd841f..11772431ae 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.h
+++ b/drivers/common/mlx5/mlx5_devx_cmds.h
@@ -196,6 +196,7 @@ struct mlx5_hca_attr {
 	uint32_t tunnel_stateless_geneve_rx:1;
 	uint32_t geneve_max_opt_len:1; /* 0x0: 14DW, 0x1: 63DW */
 	uint32_t tunnel_stateless_gtp:1;
+	uint32_t tunnel_stateless_vxlan_gpe_nsh:1;
 	uint32_t max_lso_cap;
 	uint32_t scatter_fcs:1;
 	uint32_t lro_cap:1;
diff --git a/drivers/common/mlx5/mlx5_prm.h b/drivers/common/mlx5/mlx5_prm.h
index 51f426c614..f005877dd7 100644
--- a/drivers/common/mlx5/mlx5_prm.h
+++ b/drivers/common/mlx5/mlx5_prm.h
@@ -1964,7 +1964,9 @@ struct mlx5_ifc_per_protocol_networking_offload_caps_bits {
 	u8 swp_lso[0x1];
 	u8 reserved_at_23[0x8];
 	u8 tunnel_stateless_gtp[0x1];
-	u8 reserved_at_25[0x4];
+	u8 reserved_at_25[0x2];
+	u8 tunnel_stateless_vxlan_gpe_nsh[0x1];
+	u8 reserved_at_28[0x1];
 	u8 max_vxlan_udp_ports[0x8];
 	u8 reserved_at_38[0x6];
 	u8 max_geneve_opt_len[0x1];
-- 
2.27.0

[PATCH v2 3/5] net/mlx5: enhance the validation for item VXLAN-GPE

[Haifei Luo]

Enhance the validation so that configuring vxlan-gpe's next protocol as NSH
is supported.

1. The spec's protocol can have value and nic_mask's protocol is 0xff.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index f7f8f54eb4..202e878ddf 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3198,6 +3198,11 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		uint8_t vni[4];
 	} id = { .vlan_id = 0, };
 
+	struct rte_flow_item_vxlan_gpe nic_mask = {
+		.vni = "\xff\xff\xff",
+		.protocol = 0xff,
+	};
+
 	if (!priv->sh->config.l3_vxlan_en)
 		return rte_flow_error_set(error, ENOTSUP,
 					  RTE_FLOW_ERROR_TYPE_ITEM, item,
@@ -3221,18 +3226,12 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		mask = &rte_flow_item_vxlan_gpe_mask;
 	ret = mlx5_flow_item_acceptable
 		(item, (const uint8_t *)mask,
-		 (const uint8_t *)&rte_flow_item_vxlan_gpe_mask,
+		 (const uint8_t *)&nic_mask,
 		 sizeof(struct rte_flow_item_vxlan_gpe),
 		 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 	if (ret < 0)
 		return ret;
 	if (spec) {
-		if (spec->hdr.proto)
-			return rte_flow_error_set(error, ENOTSUP,
-						  RTE_FLOW_ERROR_TYPE_ITEM,
-						  item,
-						  "VxLAN-GPE protocol"
-						  " not supported");
 		memcpy(&id.vni[1], spec->hdr.vni, 3);
 		memcpy(&id.vni[1], mask->hdr.vni, 3);
 	}
-- 
2.27.0

[PATCH v2 4/5] app/testpmd: support for NSH flow item

[Haifei Luo]

Add support for item NSH. The CLI is: nsh
Example:
flow create 0 transfer group 1
     pattern eth / ipv6 / udp dst is 4790 /
             vxlan-gpe / nsh / eth / ipv4 / tcp / end
     actions port_id id 1 / end

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index d25f941259..bf93d649e1 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -525,6 +525,7 @@ enum index {
 	ITEM_IB_BTH_PSN,
 	ITEM_IPV6_PUSH_REMOVE_EXT,
 	ITEM_IPV6_PUSH_REMOVE_EXT_TYPE,
+	ITEM_NSH,
 
 	/* Validate/create actions. */
 	ACTIONS,
@@ -1562,6 +1563,7 @@ static const enum index next_item[] = {
 	ITEM_AGGR_AFFINITY,
 	ITEM_TX_QUEUE,
 	ITEM_IB_BTH,
+	ITEM_NSH,
 	END_SET,
 	ZERO,
 };
@@ -2081,6 +2083,11 @@ static const enum index item_ib_bth[] = {
 	ZERO,
 };
 
+static const enum index item_nsh[] = {
+	ITEM_NEXT,
+	ZERO,
+};
+
 static const enum index next_action[] = {
 	ACTION_END,
 	ACTION_VOID,
@@ -5837,6 +5844,15 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_ib_bth,
 						 hdr.psn)),
 	},
+	[ITEM_NSH] = {
+		.name = "nsh",
+		.help = "match NSH header",
+		.priv = PRIV_ITEM(NSH,
+				  sizeof(struct rte_flow_item_nsh)),
+		.next = NEXT(item_nsh),
+		.call = parse_vc,
+	},
+
 	/* Validate/create actions. */
 	[ACTIONS] = {
 		.name = "actions",
-- 
2.27.0

[PATCH v2 5/5] net/mlx5: add support for item NSH

[Haifei Luo]

1. Add validation for item NSH.
   It will fail if HCA cap for NSH is false.
2. Add item_flags for NSH.
3. For vxlan-gpe if next header is NSH, set next_protocol as NSH.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c    | 39 +++++++++++++++++++++++++++++++++
 drivers/net/mlx5/mlx5_flow.h    |  6 +++++
 drivers/net/mlx5/mlx5_flow_dv.c | 13 ++++++++++-
 3 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 202e878ddf..8ad85e6027 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3905,6 +3905,45 @@ mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 					 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 }
 
+/**
+ * Validate the NSH item.
+ *
+ * @param[in] dev
+ *   Pointer to Ethernet device on which flow rule is being created on.
+ * @param[out] error
+ *   Pointer to error structure.
+ *
+ * @return
+ *   0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+int
+mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+			    const struct rte_flow_item *item,
+			    struct rte_flow_error *error)
+{
+	struct mlx5_priv *priv = dev->data->dev_private;
+
+	if (item->mask) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "NSH fields matching is not supported");
+	}
+
+	if (!priv->sh->config.dv_flow_en) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+					  NULL, "NSH support requires DV flow interface");
+	}
+
+	if (!priv->sh->cdev->config.hca_attr.tunnel_stateless_vxlan_gpe_nsh) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "Current FW does not support matching on NSH");
+	}
+
+	return 0;
+}
+
 static int
 flow_null_validate(struct rte_eth_dev *dev __rte_unused,
 		   const struct rte_flow_attr *attr __rte_unused,
diff --git a/drivers/net/mlx5/mlx5_flow.h b/drivers/net/mlx5/mlx5_flow.h
index 3a97975d69..ccb416e497 100644
--- a/drivers/net/mlx5/mlx5_flow.h
+++ b/drivers/net/mlx5/mlx5_flow.h
@@ -233,6 +233,9 @@ enum mlx5_feature_name {
 /* IB BTH ITEM. */
 #define MLX5_FLOW_ITEM_IB_BTH (1ull << 51)
 
+/* NSH ITEM */
+#define MLX5_FLOW_ITEM_NSH (1ull << 53)
+
 /* Outer Masks. */
 #define MLX5_FLOW_LAYER_OUTER_L3 \
 	(MLX5_FLOW_LAYER_OUTER_L3_IPV4 | MLX5_FLOW_LAYER_OUTER_L3_IPV6)
@@ -2453,6 +2456,9 @@ int mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 				  uint16_t ether_type,
 				  const struct rte_flow_item_ecpri *acc_mask,
 				  struct rte_flow_error *error);
+int mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+				const struct rte_flow_item *item,
+				struct rte_flow_error *error);
 int mlx5_flow_create_mtr_tbls(struct rte_eth_dev *dev,
 			      struct mlx5_flow_meter_info *fm,
 			      uint32_t mtr_idx,
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 3f4325c5c8..5cd04b1d93 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -7815,6 +7815,12 @@ flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 
 			last_item = MLX5_FLOW_ITEM_IB_BTH;
 			break;
+		case RTE_FLOW_ITEM_TYPE_NSH:
+			ret = mlx5_flow_validate_item_nsh(dev, items, error);
+			if (ret < 0)
+				return ret;
+			last_item = MLX5_FLOW_ITEM_NSH;
+			break;
 		default:
 			return rte_flow_error_set(error, ENOTSUP,
 						  RTE_FLOW_ERROR_TYPE_ITEM,
@@ -9720,7 +9726,9 @@ flow_dv_translate_item_vxlan_gpe(void *key, const struct rte_flow_item *item,
 	v_protocol = vxlan_v->hdr.protocol;
 	if (!m_protocol) {
 		/* Force next protocol to ensure next headers parsing. */
-		if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
+		if (pattern_flags & MLX5_FLOW_ITEM_NSH)
+			v_protocol = RTE_VXLAN_GPE_TYPE_NSH;
+		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
 			v_protocol = RTE_VXLAN_GPE_TYPE_ETH;
 		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L3_IPV4)
 			v_protocol = RTE_VXLAN_GPE_TYPE_IPV4;
@@ -13910,6 +13918,9 @@ flow_dv_translate_items(struct rte_eth_dev *dev,
 		flow_dv_translate_item_ib_bth(key, items, tunnel, key_type);
 		last_item = MLX5_FLOW_ITEM_IB_BTH;
 		break;
+	case RTE_FLOW_ITEM_TYPE_NSH:
+		last_item = MLX5_FLOW_ITEM_NSH;
+		break;
 	default:
 		break;
 	}
-- 
2.27.0

[PATCH v3 0/5] support item NSH matching

[Haifei Luo]

NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
NSH fields matching is not supported.  

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Add support for matching item NSH. The CLI is: nsh
Add support for HCA attribute query of NSH.

Enhance the validation for the case matching item NSH is supported. 
Add NSH support in net/mlx5.

V2: Add Ack info in commit message.
V3: Add more Ack in commi message.

Haifei Luo (5):
  app/testpmd: support for VXLAN-GPE's next protocol
  common/mlx5: extend HCA attribute query for NSH
  net/mlx5: enhance the validation for item VXLAN-GPE
  app/testpmd: support for NSH flow item
  net/mlx5: add support for item NSH

 app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
 drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
 drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
 drivers/common/mlx5/mlx5_prm.h       |  4 ++-
 drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
 drivers/net/mlx5/mlx5_flow.h         |  6 ++++
 drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
 7 files changed, 96 insertions(+), 9 deletions(-)

-- 
2.27.0

[PATCH v3 1/5] app/testpmd: support for VXLAN-GPE's next protocol

[Haifei Luo]

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Example:
flow create 0 transfer group 1 pattern eth / ipv6 / udp dst is 4790
/ vxlan-gpe protocol is 0x04 / eth / ipv4  / tcp / end
actions port_id id 1 / end

Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com>
Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 94827bcc4a..d25f941259 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -385,6 +385,7 @@ enum index {
 	ITEM_GENEVE_OPTLEN,
 	ITEM_VXLAN_GPE,
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_ARP_ETH_IPV4,
 	ITEM_ARP_ETH_IPV4_SHA,
 	ITEM_ARP_ETH_IPV4_SPA,
@@ -1758,6 +1759,7 @@ static const enum index item_geneve[] = {
 
 static const enum index item_vxlan_gpe[] = {
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_NEXT,
 	ZERO,
 };
@@ -4804,6 +4806,14 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
 					     hdr.vni)),
 	},
+	[ITEM_VXLAN_GPE_PROTO] = {
+		.name = "protocol",
+		.help = "VXLAN-GPE next protocol",
+		.next = NEXT(item_vxlan_gpe, NEXT_ENTRY(COMMON_UNSIGNED),
+			     item_param),
+		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
+					     protocol)),
+	},
 	[ITEM_ARP_ETH_IPV4] = {
 		.name = "arp_eth_ipv4",
 		.help = "match ARP header for Ethernet/IPv4",
-- 
2.27.0

[PATCH v3 2/5] common/mlx5: extend HCA attribute query for NSH

[Haifei Luo]

Add NSH supporting field in two places:
1. New HCA capability indicating NSH is supported
2. New field in "mlx5_ifc_per_protocol_networking_offload_caps_bits"
   structure

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/common/mlx5/mlx5_devx_cmds.c | 3 +++
 drivers/common/mlx5/mlx5_devx_cmds.h | 1 +
 drivers/common/mlx5/mlx5_prm.h       | 4 +++-
 3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/common/mlx5/mlx5_devx_cmds.c b/drivers/common/mlx5/mlx5_devx_cmds.c
index 66a77159a0..830199212e 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.c
+++ b/drivers/common/mlx5/mlx5_devx_cmds.c
@@ -1313,6 +1313,9 @@ mlx5_devx_cmd_query_hca_attr(void *ctx,
 	attr->tunnel_stateless_gtp = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, tunnel_stateless_gtp);
+	attr->tunnel_stateless_vxlan_gpe_nsh = MLX5_GET
+					(per_protocol_networking_offload_caps,
+					 hcattr, tunnel_stateless_vxlan_gpe_nsh);
 	attr->rss_ind_tbl_cap = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, rss_ind_tbl_cap);
diff --git a/drivers/common/mlx5/mlx5_devx_cmds.h b/drivers/common/mlx5/mlx5_devx_cmds.h
index e071cd841f..11772431ae 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.h
+++ b/drivers/common/mlx5/mlx5_devx_cmds.h
@@ -196,6 +196,7 @@ struct mlx5_hca_attr {
 	uint32_t tunnel_stateless_geneve_rx:1;
 	uint32_t geneve_max_opt_len:1; /* 0x0: 14DW, 0x1: 63DW */
 	uint32_t tunnel_stateless_gtp:1;
+	uint32_t tunnel_stateless_vxlan_gpe_nsh:1;
 	uint32_t max_lso_cap;
 	uint32_t scatter_fcs:1;
 	uint32_t lro_cap:1;
diff --git a/drivers/common/mlx5/mlx5_prm.h b/drivers/common/mlx5/mlx5_prm.h
index 51f426c614..f005877dd7 100644
--- a/drivers/common/mlx5/mlx5_prm.h
+++ b/drivers/common/mlx5/mlx5_prm.h
@@ -1964,7 +1964,9 @@ struct mlx5_ifc_per_protocol_networking_offload_caps_bits {
 	u8 swp_lso[0x1];
 	u8 reserved_at_23[0x8];
 	u8 tunnel_stateless_gtp[0x1];
-	u8 reserved_at_25[0x4];
+	u8 reserved_at_25[0x2];
+	u8 tunnel_stateless_vxlan_gpe_nsh[0x1];
+	u8 reserved_at_28[0x1];
 	u8 max_vxlan_udp_ports[0x8];
 	u8 reserved_at_38[0x6];
 	u8 max_geneve_opt_len[0x1];
-- 
2.27.0

[PATCH v3 3/5] net/mlx5: enhance the validation for item VXLAN-GPE

[Haifei Luo]

Enhance the validation so that configuring vxlan-gpe's next protocol as NSH
is supported.

1. The spec's protocol can have value and nic_mask's protocol is 0xff.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index f7f8f54eb4..202e878ddf 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3198,6 +3198,11 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		uint8_t vni[4];
 	} id = { .vlan_id = 0, };
 
+	struct rte_flow_item_vxlan_gpe nic_mask = {
+		.vni = "\xff\xff\xff",
+		.protocol = 0xff,
+	};
+
 	if (!priv->sh->config.l3_vxlan_en)
 		return rte_flow_error_set(error, ENOTSUP,
 					  RTE_FLOW_ERROR_TYPE_ITEM, item,
@@ -3221,18 +3226,12 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		mask = &rte_flow_item_vxlan_gpe_mask;
 	ret = mlx5_flow_item_acceptable
 		(item, (const uint8_t *)mask,
-		 (const uint8_t *)&rte_flow_item_vxlan_gpe_mask,
+		 (const uint8_t *)&nic_mask,
 		 sizeof(struct rte_flow_item_vxlan_gpe),
 		 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 	if (ret < 0)
 		return ret;
 	if (spec) {
-		if (spec->hdr.proto)
-			return rte_flow_error_set(error, ENOTSUP,
-						  RTE_FLOW_ERROR_TYPE_ITEM,
-						  item,
-						  "VxLAN-GPE protocol"
-						  " not supported");
 		memcpy(&id.vni[1], spec->hdr.vni, 3);
 		memcpy(&id.vni[1], mask->hdr.vni, 3);
 	}
-- 
2.27.0

[PATCH v3 4/5] app/testpmd: support for NSH flow item

[Haifei Luo]

Add support for item NSH. The CLI is: nsh
Example:
flow create 0 transfer group 1
     pattern eth / ipv6 / udp dst is 4790 /
             vxlan-gpe / nsh / eth / ipv4 / tcp / end
     actions port_id id 1 / end

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index d25f941259..bf93d649e1 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -525,6 +525,7 @@ enum index {
 	ITEM_IB_BTH_PSN,
 	ITEM_IPV6_PUSH_REMOVE_EXT,
 	ITEM_IPV6_PUSH_REMOVE_EXT_TYPE,
+	ITEM_NSH,
 
 	/* Validate/create actions. */
 	ACTIONS,
@@ -1562,6 +1563,7 @@ static const enum index next_item[] = {
 	ITEM_AGGR_AFFINITY,
 	ITEM_TX_QUEUE,
 	ITEM_IB_BTH,
+	ITEM_NSH,
 	END_SET,
 	ZERO,
 };
@@ -2081,6 +2083,11 @@ static const enum index item_ib_bth[] = {
 	ZERO,
 };
 
+static const enum index item_nsh[] = {
+	ITEM_NEXT,
+	ZERO,
+};
+
 static const enum index next_action[] = {
 	ACTION_END,
 	ACTION_VOID,
@@ -5837,6 +5844,15 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_ib_bth,
 						 hdr.psn)),
 	},
+	[ITEM_NSH] = {
+		.name = "nsh",
+		.help = "match NSH header",
+		.priv = PRIV_ITEM(NSH,
+				  sizeof(struct rte_flow_item_nsh)),
+		.next = NEXT(item_nsh),
+		.call = parse_vc,
+	},
+
 	/* Validate/create actions. */
 	[ACTIONS] = {
 		.name = "actions",
-- 
2.27.0

[PATCH v3 5/5] net/mlx5: add support for item NSH

[Haifei Luo]

1. Add validation for item NSH.
   It will fail if HCA cap for NSH is false.
2. Add item_flags for NSH.
3. For vxlan-gpe if next header is NSH, set next_protocol as NSH.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c    | 39 +++++++++++++++++++++++++++++++++
 drivers/net/mlx5/mlx5_flow.h    |  6 +++++
 drivers/net/mlx5/mlx5_flow_dv.c | 13 ++++++++++-
 3 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 202e878ddf..8ad85e6027 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3905,6 +3905,45 @@ mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 					 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 }
 
+/**
+ * Validate the NSH item.
+ *
+ * @param[in] dev
+ *   Pointer to Ethernet device on which flow rule is being created on.
+ * @param[out] error
+ *   Pointer to error structure.
+ *
+ * @return
+ *   0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+int
+mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+			    const struct rte_flow_item *item,
+			    struct rte_flow_error *error)
+{
+	struct mlx5_priv *priv = dev->data->dev_private;
+
+	if (item->mask) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "NSH fields matching is not supported");
+	}
+
+	if (!priv->sh->config.dv_flow_en) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+					  NULL, "NSH support requires DV flow interface");
+	}
+
+	if (!priv->sh->cdev->config.hca_attr.tunnel_stateless_vxlan_gpe_nsh) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "Current FW does not support matching on NSH");
+	}
+
+	return 0;
+}
+
 static int
 flow_null_validate(struct rte_eth_dev *dev __rte_unused,
 		   const struct rte_flow_attr *attr __rte_unused,
diff --git a/drivers/net/mlx5/mlx5_flow.h b/drivers/net/mlx5/mlx5_flow.h
index 3a97975d69..ccb416e497 100644
--- a/drivers/net/mlx5/mlx5_flow.h
+++ b/drivers/net/mlx5/mlx5_flow.h
@@ -233,6 +233,9 @@ enum mlx5_feature_name {
 /* IB BTH ITEM. */
 #define MLX5_FLOW_ITEM_IB_BTH (1ull << 51)
 
+/* NSH ITEM */
+#define MLX5_FLOW_ITEM_NSH (1ull << 53)
+
 /* Outer Masks. */
 #define MLX5_FLOW_LAYER_OUTER_L3 \
 	(MLX5_FLOW_LAYER_OUTER_L3_IPV4 | MLX5_FLOW_LAYER_OUTER_L3_IPV6)
@@ -2453,6 +2456,9 @@ int mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 				  uint16_t ether_type,
 				  const struct rte_flow_item_ecpri *acc_mask,
 				  struct rte_flow_error *error);
+int mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+				const struct rte_flow_item *item,
+				struct rte_flow_error *error);
 int mlx5_flow_create_mtr_tbls(struct rte_eth_dev *dev,
 			      struct mlx5_flow_meter_info *fm,
 			      uint32_t mtr_idx,
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 3f4325c5c8..5cd04b1d93 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -7815,6 +7815,12 @@ flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 
 			last_item = MLX5_FLOW_ITEM_IB_BTH;
 			break;
+		case RTE_FLOW_ITEM_TYPE_NSH:
+			ret = mlx5_flow_validate_item_nsh(dev, items, error);
+			if (ret < 0)
+				return ret;
+			last_item = MLX5_FLOW_ITEM_NSH;
+			break;
 		default:
 			return rte_flow_error_set(error, ENOTSUP,
 						  RTE_FLOW_ERROR_TYPE_ITEM,
@@ -9720,7 +9726,9 @@ flow_dv_translate_item_vxlan_gpe(void *key, const struct rte_flow_item *item,
 	v_protocol = vxlan_v->hdr.protocol;
 	if (!m_protocol) {
 		/* Force next protocol to ensure next headers parsing. */
-		if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
+		if (pattern_flags & MLX5_FLOW_ITEM_NSH)
+			v_protocol = RTE_VXLAN_GPE_TYPE_NSH;
+		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
 			v_protocol = RTE_VXLAN_GPE_TYPE_ETH;
 		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L3_IPV4)
 			v_protocol = RTE_VXLAN_GPE_TYPE_IPV4;
@@ -13910,6 +13918,9 @@ flow_dv_translate_items(struct rte_eth_dev *dev,
 		flow_dv_translate_item_ib_bth(key, items, tunnel, key_type);
 		last_item = MLX5_FLOW_ITEM_IB_BTH;
 		break;
+	case RTE_FLOW_ITEM_TYPE_NSH:
+		last_item = MLX5_FLOW_ITEM_NSH;
+		break;
 	default:
 		break;
 	}
-- 
2.27.0

[PATCH v4 0/5] support item NSH matching

[Haifei Luo]

NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
NSH fields matching is not supported.  

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Add support for matching item NSH. The CLI is: nsh
Add support for HCA attribute query of NSH.

Enhance the validation for the case matching item NSH is supported. 
Add NSH support in net/mlx5.

V2: Add Ack info in commit message.
V3: Add more Ack in commi message.
V4: Add nsh in feature ini file.

Haifei Luo (5):
  app/testpmd: support for VXLAN-GPE's next protocol
  common/mlx5: extend HCA attribute query for NSH
  net/mlx5: enhance the validation for item VXLAN-GPE
  app/testpmd: support for NSH flow item
  net/mlx5: add support for item NSH

 app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
 doc/guides/nics/features/mlx5.ini    |  1 +
 drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
 drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
 drivers/common/mlx5/mlx5_prm.h       |  4 ++-
 drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
 drivers/net/mlx5/mlx5_flow.h         |  6 ++++
 drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
 8 files changed, 97 insertions(+), 9 deletions(-)

-- 
2.27.0

[PATCH v4 1/5] app/testpmd: support for VXLAN-GPE's next protocol

[Haifei Luo]

Add support for configuring VXLAN-GPE's next protocol.
The CLI is: vxlan-gpe protocol is <value>.

Example:
flow create 0 transfer group 1 pattern eth / ipv6 / udp dst is 4790
/ vxlan-gpe protocol is 0x04 / eth / ipv4  / tcp / end
actions port_id id 1 / end

Signed-off-by: Jiawei Wang <jiaweiw@nvidia.com>
Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 app/test-pmd/cmdline_flow.c | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 95c0a19beb..8a4e06c50a 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -386,6 +386,7 @@ enum index {
 	ITEM_GENEVE_OPTLEN,
 	ITEM_VXLAN_GPE,
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_ARP_ETH_IPV4,
 	ITEM_ARP_ETH_IPV4_SHA,
 	ITEM_ARP_ETH_IPV4_SPA,
@@ -1760,6 +1761,7 @@ static const enum index item_geneve[] = {
 
 static const enum index item_vxlan_gpe[] = {
 	ITEM_VXLAN_GPE_VNI,
+	ITEM_VXLAN_GPE_PROTO,
 	ITEM_NEXT,
 	ZERO,
 };
@@ -4813,6 +4815,14 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
 					     hdr.vni)),
 	},
+	[ITEM_VXLAN_GPE_PROTO] = {
+		.name = "protocol",
+		.help = "VXLAN-GPE next protocol",
+		.next = NEXT(item_vxlan_gpe, NEXT_ENTRY(COMMON_UNSIGNED),
+			     item_param),
+		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_vxlan_gpe,
+					     protocol)),
+	},
 	[ITEM_ARP_ETH_IPV4] = {
 		.name = "arp_eth_ipv4",
 		.help = "match ARP header for Ethernet/IPv4",
-- 
2.27.0

[PATCH v4 2/5] common/mlx5: extend HCA attribute query for NSH

[Haifei Luo]

Add NSH supporting field in two places:
1. New HCA capability indicating NSH is supported
2. New field in "mlx5_ifc_per_protocol_networking_offload_caps_bits"
   structure

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/common/mlx5/mlx5_devx_cmds.c | 3 +++
 drivers/common/mlx5/mlx5_devx_cmds.h | 1 +
 drivers/common/mlx5/mlx5_prm.h       | 4 +++-
 3 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/drivers/common/mlx5/mlx5_devx_cmds.c b/drivers/common/mlx5/mlx5_devx_cmds.c
index 66a77159a0..830199212e 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.c
+++ b/drivers/common/mlx5/mlx5_devx_cmds.c
@@ -1313,6 +1313,9 @@ mlx5_devx_cmd_query_hca_attr(void *ctx,
 	attr->tunnel_stateless_gtp = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, tunnel_stateless_gtp);
+	attr->tunnel_stateless_vxlan_gpe_nsh = MLX5_GET
+					(per_protocol_networking_offload_caps,
+					 hcattr, tunnel_stateless_vxlan_gpe_nsh);
 	attr->rss_ind_tbl_cap = MLX5_GET
 					(per_protocol_networking_offload_caps,
 					 hcattr, rss_ind_tbl_cap);
diff --git a/drivers/common/mlx5/mlx5_devx_cmds.h b/drivers/common/mlx5/mlx5_devx_cmds.h
index e071cd841f..11772431ae 100644
--- a/drivers/common/mlx5/mlx5_devx_cmds.h
+++ b/drivers/common/mlx5/mlx5_devx_cmds.h
@@ -196,6 +196,7 @@ struct mlx5_hca_attr {
 	uint32_t tunnel_stateless_geneve_rx:1;
 	uint32_t geneve_max_opt_len:1; /* 0x0: 14DW, 0x1: 63DW */
 	uint32_t tunnel_stateless_gtp:1;
+	uint32_t tunnel_stateless_vxlan_gpe_nsh:1;
 	uint32_t max_lso_cap;
 	uint32_t scatter_fcs:1;
 	uint32_t lro_cap:1;
diff --git a/drivers/common/mlx5/mlx5_prm.h b/drivers/common/mlx5/mlx5_prm.h
index 51f426c614..f005877dd7 100644
--- a/drivers/common/mlx5/mlx5_prm.h
+++ b/drivers/common/mlx5/mlx5_prm.h
@@ -1964,7 +1964,9 @@ struct mlx5_ifc_per_protocol_networking_offload_caps_bits {
 	u8 swp_lso[0x1];
 	u8 reserved_at_23[0x8];
 	u8 tunnel_stateless_gtp[0x1];
-	u8 reserved_at_25[0x4];
+	u8 reserved_at_25[0x2];
+	u8 tunnel_stateless_vxlan_gpe_nsh[0x1];
+	u8 reserved_at_28[0x1];
 	u8 max_vxlan_udp_ports[0x8];
 	u8 reserved_at_38[0x6];
 	u8 max_geneve_opt_len[0x1];
-- 
2.27.0

[PATCH v4 3/5] net/mlx5: enhance the validation for item VXLAN-GPE

[Haifei Luo]

Enhance the validation so that configuring vxlan-gpe's next protocol as NSH
is supported.

1. The spec's protocol can have value and nic_mask's protocol is 0xff.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index f7f8f54eb4..202e878ddf 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3198,6 +3198,11 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		uint8_t vni[4];
 	} id = { .vlan_id = 0, };
 
+	struct rte_flow_item_vxlan_gpe nic_mask = {
+		.vni = "\xff\xff\xff",
+		.protocol = 0xff,
+	};
+
 	if (!priv->sh->config.l3_vxlan_en)
 		return rte_flow_error_set(error, ENOTSUP,
 					  RTE_FLOW_ERROR_TYPE_ITEM, item,
@@ -3221,18 +3226,12 @@ mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 		mask = &rte_flow_item_vxlan_gpe_mask;
 	ret = mlx5_flow_item_acceptable
 		(item, (const uint8_t *)mask,
-		 (const uint8_t *)&rte_flow_item_vxlan_gpe_mask,
+		 (const uint8_t *)&nic_mask,
 		 sizeof(struct rte_flow_item_vxlan_gpe),
 		 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 	if (ret < 0)
 		return ret;
 	if (spec) {
-		if (spec->hdr.proto)
-			return rte_flow_error_set(error, ENOTSUP,
-						  RTE_FLOW_ERROR_TYPE_ITEM,
-						  item,
-						  "VxLAN-GPE protocol"
-						  " not supported");
 		memcpy(&id.vni[1], spec->hdr.vni, 3);
 		memcpy(&id.vni[1], mask->hdr.vni, 3);
 	}
-- 
2.27.0

[PATCH v4 4/5] app/testpmd: support for NSH flow item

[Haifei Luo]

Add support for item NSH. The CLI is: nsh
Example:
flow create 0 transfer group 1
     pattern eth / ipv6 / udp dst is 4790 /
             vxlan-gpe / nsh / eth / ipv4 / tcp / end
     actions port_id id 1 / end

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 app/test-pmd/cmdline_flow.c       | 16 ++++++++++++++++
 doc/guides/nics/features/mlx5.ini |  1 +
 2 files changed, 17 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 8a4e06c50a..84de3a2506 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -526,6 +526,7 @@ enum index {
 	ITEM_IB_BTH_PSN,
 	ITEM_IPV6_PUSH_REMOVE_EXT,
 	ITEM_IPV6_PUSH_REMOVE_EXT_TYPE,
+	ITEM_NSH,
 
 	/* Validate/create actions. */
 	ACTIONS,
@@ -1563,6 +1564,7 @@ static const enum index next_item[] = {
 	ITEM_AGGR_AFFINITY,
 	ITEM_TX_QUEUE,
 	ITEM_IB_BTH,
+	ITEM_NSH,
 	END_SET,
 	ZERO,
 };
@@ -2083,6 +2085,11 @@ static const enum index item_ib_bth[] = {
 	ZERO,
 };
 
+static const enum index item_nsh[] = {
+	ITEM_NEXT,
+	ZERO,
+};
+
 static const enum index next_action[] = {
 	ACTION_END,
 	ACTION_VOID,
@@ -5846,6 +5853,15 @@ static const struct token token_list[] = {
 		.args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_ib_bth,
 						 hdr.psn)),
 	},
+	[ITEM_NSH] = {
+		.name = "nsh",
+		.help = "match NSH header",
+		.priv = PRIV_ITEM(NSH,
+				  sizeof(struct rte_flow_item_nsh)),
+		.next = NEXT(item_nsh),
+		.call = parse_vc,
+	},
+
 	/* Validate/create actions. */
 	[ACTIONS] = {
 		.name = "actions",
diff --git a/doc/guides/nics/features/mlx5.ini b/doc/guides/nics/features/mlx5.ini
index c0e0b779cf..fc67415c6c 100644
--- a/doc/guides/nics/features/mlx5.ini
+++ b/doc/guides/nics/features/mlx5.ini
@@ -82,6 +82,7 @@ mark                 = Y
 meta                 = Y
 meter_color          = Y
 mpls                 = Y
+nsh                  = Y
 nvgre                = Y
 port_id              = Y
 port_representor     = Y
-- 
2.27.0

[PATCH v4 5/5] net/mlx5: add support for item NSH

[Haifei Luo]

1. Add validation for item NSH.
   It will fail if HCA cap for NSH is false.
2. Add item_flags for NSH.
3. For vxlan-gpe if next header is NSH, set next_protocol as NSH.

Signed-off-by: Haifei Luo <haifeil@nvidia.com>
Acked-by: Dariusz Sosnowski <dsosnowski@nvidia.com>
Acked-by: Suanming Mou <suanmingm@nvidia.com>
---
 drivers/net/mlx5/mlx5_flow.c    | 39 +++++++++++++++++++++++++++++++++
 drivers/net/mlx5/mlx5_flow.h    |  6 +++++
 drivers/net/mlx5/mlx5_flow_dv.c | 13 ++++++++++-
 3 files changed, 57 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 202e878ddf..8ad85e6027 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -3905,6 +3905,45 @@ mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 					 MLX5_ITEM_RANGE_NOT_ACCEPTED, error);
 }
 
+/**
+ * Validate the NSH item.
+ *
+ * @param[in] dev
+ *   Pointer to Ethernet device on which flow rule is being created on.
+ * @param[out] error
+ *   Pointer to error structure.
+ *
+ * @return
+ *   0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+int
+mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+			    const struct rte_flow_item *item,
+			    struct rte_flow_error *error)
+{
+	struct mlx5_priv *priv = dev->data->dev_private;
+
+	if (item->mask) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "NSH fields matching is not supported");
+	}
+
+	if (!priv->sh->config.dv_flow_en) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+					  NULL, "NSH support requires DV flow interface");
+	}
+
+	if (!priv->sh->cdev->config.hca_attr.tunnel_stateless_vxlan_gpe_nsh) {
+		return rte_flow_error_set(error, ENOTSUP,
+					  RTE_FLOW_ERROR_TYPE_ITEM, item,
+					  "Current FW does not support matching on NSH");
+	}
+
+	return 0;
+}
+
 static int
 flow_null_validate(struct rte_eth_dev *dev __rte_unused,
 		   const struct rte_flow_attr *attr __rte_unused,
diff --git a/drivers/net/mlx5/mlx5_flow.h b/drivers/net/mlx5/mlx5_flow.h
index 3a97975d69..ccb416e497 100644
--- a/drivers/net/mlx5/mlx5_flow.h
+++ b/drivers/net/mlx5/mlx5_flow.h
@@ -233,6 +233,9 @@ enum mlx5_feature_name {
 /* IB BTH ITEM. */
 #define MLX5_FLOW_ITEM_IB_BTH (1ull << 51)
 
+/* NSH ITEM */
+#define MLX5_FLOW_ITEM_NSH (1ull << 53)
+
 /* Outer Masks. */
 #define MLX5_FLOW_LAYER_OUTER_L3 \
 	(MLX5_FLOW_LAYER_OUTER_L3_IPV4 | MLX5_FLOW_LAYER_OUTER_L3_IPV6)
@@ -2453,6 +2456,9 @@ int mlx5_flow_validate_item_ecpri(const struct rte_flow_item *item,
 				  uint16_t ether_type,
 				  const struct rte_flow_item_ecpri *acc_mask,
 				  struct rte_flow_error *error);
+int mlx5_flow_validate_item_nsh(struct rte_eth_dev *dev,
+				const struct rte_flow_item *item,
+				struct rte_flow_error *error);
 int mlx5_flow_create_mtr_tbls(struct rte_eth_dev *dev,
 			      struct mlx5_flow_meter_info *fm,
 			      uint32_t mtr_idx,
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 3f4325c5c8..5cd04b1d93 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -7815,6 +7815,12 @@ flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 
 			last_item = MLX5_FLOW_ITEM_IB_BTH;
 			break;
+		case RTE_FLOW_ITEM_TYPE_NSH:
+			ret = mlx5_flow_validate_item_nsh(dev, items, error);
+			if (ret < 0)
+				return ret;
+			last_item = MLX5_FLOW_ITEM_NSH;
+			break;
 		default:
 			return rte_flow_error_set(error, ENOTSUP,
 						  RTE_FLOW_ERROR_TYPE_ITEM,
@@ -9720,7 +9726,9 @@ flow_dv_translate_item_vxlan_gpe(void *key, const struct rte_flow_item *item,
 	v_protocol = vxlan_v->hdr.protocol;
 	if (!m_protocol) {
 		/* Force next protocol to ensure next headers parsing. */
-		if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
+		if (pattern_flags & MLX5_FLOW_ITEM_NSH)
+			v_protocol = RTE_VXLAN_GPE_TYPE_NSH;
+		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L2)
 			v_protocol = RTE_VXLAN_GPE_TYPE_ETH;
 		else if (pattern_flags & MLX5_FLOW_LAYER_INNER_L3_IPV4)
 			v_protocol = RTE_VXLAN_GPE_TYPE_IPV4;
@@ -13910,6 +13918,9 @@ flow_dv_translate_items(struct rte_eth_dev *dev,
 		flow_dv_translate_item_ib_bth(key, items, tunnel, key_type);
 		last_item = MLX5_FLOW_ITEM_IB_BTH;
 		break;
+	case RTE_FLOW_ITEM_TYPE_NSH:
+		last_item = MLX5_FLOW_ITEM_NSH;
+		break;
 	default:
 		break;
 	}
-- 
2.27.0

RE: [PATCH v2 0/5] support item NSH matching

[Raslan Darawsheh]

Hi,

> -----Original Message-----
> From: Haifei Luo <haifeil@nvidia.com>
> Sent: Monday, September 25, 2023 5:09 AM
> To: dev@dpdk.org
> Cc: Ori Kam <orika@nvidia.com>; Slava Ovsiienko <viacheslavo@nvidia.com>;
> Raslan Darawsheh <rasland@nvidia.com>; Xueming(Steven) Li
> <xuemingl@nvidia.com>; Haifei Luo <haifeil@nvidia.com>
> Subject: [PATCH v2 0/5] support item NSH matching
> 
> NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
> NSH fields matching is not supported.
> 
> Add support for configuring VXLAN-GPE's next protocol.
> The CLI is: vxlan-gpe protocol is <value>.
> 
> Add support for matching item NSH. The CLI is: nsh Add support for HCA
> attribute query of NSH.
> 
> Enhance the validation for the case matching item NSH is supported.
> Add NSH support in net/mlx5.
> 
> V2: Add Ack info in commit message.
> 
> Haifei Luo (5):
>   app/testpmd: support for VXLAN-GPE's next protocol
>   common/mlx5: extend HCA attribute query for NSH
>   net/mlx5: enhance the validation for item VXLAN-GPE
>   app/testpmd: support for NSH flow item
>   net/mlx5: add support for item NSH
> 
>  app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
>  drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
> drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
>  drivers/common/mlx5/mlx5_prm.h       |  4 ++-
>  drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
>  drivers/net/mlx5/mlx5_flow.h         |  6 ++++
>  drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
>  7 files changed, 96 insertions(+), 9 deletions(-)
> 
> --
> 2.27.0

Series applied to next-net-mlx,

Kindest regards,
Raslan Darawsheh

RE: [PATCH v4 0/5] support item NSH matching

[Raslan Darawsheh]

Hi,

> -----Original Message-----
> From: Haifei Luo <haifeil@nvidia.com>
> Sent: Sunday, October 8, 2023 6:10 AM
> To: dev@dpdk.org
> Cc: Ori Kam <orika@nvidia.com>; Slava Ovsiienko <viacheslavo@nvidia.com>;
> Raslan Darawsheh <rasland@nvidia.com>; Xueming(Steven) Li
> <xuemingl@nvidia.com>; Haifei Luo <haifeil@nvidia.com>
> Subject: [PATCH v4 0/5] support item NSH matching
> 
> NSH can be matched using the existed item: RTE_FLOW_ITEM_TYPE_NSH.
> NSH fields matching is not supported.
> 
> Add support for configuring VXLAN-GPE's next protocol.
> The CLI is: vxlan-gpe protocol is <value>.
> 
> Add support for matching item NSH. The CLI is: nsh Add support for HCA
> attribute query of NSH.
> 
> Enhance the validation for the case matching item NSH is supported.
> Add NSH support in net/mlx5.
> 
> V2: Add Ack info in commit message.
> V3: Add more Ack in commi message.
> V4: Add nsh in feature ini file.
> 
> Haifei Luo (5):
>   app/testpmd: support for VXLAN-GPE's next protocol
>   common/mlx5: extend HCA attribute query for NSH
>   net/mlx5: enhance the validation for item VXLAN-GPE
>   app/testpmd: support for NSH flow item
>   net/mlx5: add support for item NSH
> 
>  app/test-pmd/cmdline_flow.c          | 26 ++++++++++++++
>  doc/guides/nics/features/mlx5.ini    |  1 +
>  drivers/common/mlx5/mlx5_devx_cmds.c |  3 ++
> drivers/common/mlx5/mlx5_devx_cmds.h |  1 +
>  drivers/common/mlx5/mlx5_prm.h       |  4 ++-
>  drivers/net/mlx5/mlx5_flow.c         | 52 ++++++++++++++++++++++++----
>  drivers/net/mlx5/mlx5_flow.h         |  6 ++++
>  drivers/net/mlx5/mlx5_flow_dv.c      | 13 ++++++-
>  8 files changed, 97 insertions(+), 9 deletions(-)
> 
> --
> 2.27.0

Series applied to next-net-mlx,
Replied originally to V2
Now to the correct version !

Kindest regards,
Raslan Darawsheh

Re: [PATCH v4 0/5] support item NSH matching

[Thomas Monjalon]

09/10/2023 13:15, Raslan Darawsheh:
> Series applied to next-net-mlx,
> Replied originally to V2
> Now to the correct version !


A release note should be added:

* **Updated NVIDIA mlx5 net driver.**

  * Added support for Network Service Header (NSH) flow matching.