diff mbox series

[v2] net/mlx5: fix vni matching with non-std port at ConnectX-5

Message ID 20210802122048.555306-1-rongweil@nvidia.com (mailing list archive)
State Accepted, archived
Delegated to: Thomas Monjalon
Headers show
Series [v2] net/mlx5: fix vni matching with non-std port at ConnectX-5 | expand

Checks

Context Check Description
ci/iol-x86_64-compile-testing success Testing PASS
ci/iol-x86_64-unit-testing success Testing PASS
ci/iol-aarch64-unit-testing success Testing PASS
ci/iol-aarch64-compile-testing success Testing PASS
ci/iol-abi-testing success Testing PASS
ci/iol-x86_64-compile-testing-testing success Testing PASS
ci/iol-mellanox-Performance success Performance Testing PASS
ci/iol-spell-check-testing warning Testing issues
ci/iol-testing success Testing PASS
ci/intel-Testing success Testing PASS
ci/Intel-compilation success Compilation OK
ci/iol-intel-Performance success Performance Testing PASS
ci/iol-intel-Functional success Functional Testing PASS
ci/github-robot success github build: passed
ci/checkpatch success coding style OK

Commit Message

Rongwei Liu Aug. 2, 2021, 12:20 p.m. UTC
In the recent update, the misc5 matcher was introduced to
match VxLAN header extra fields. However, ConnectX-5
doesn't support misc5 for the UDP ports different from
VXLAN's standard one (4789).

Need to fall back to the previous approach and use legacy
misc matcher if non-standard UDP port is recognized
in VxLAN flow.

Fixes: 630a587bfb37 ("net/mlx5: support matching on VXLAN reserved field")
Cc: stable@dpdk.org

Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
---
 doc/guides/nics/mlx5.rst           |  1 +
 drivers/net/mlx5/mlx5_flow.c       | 21 +++++++++++++++------
 drivers/net/mlx5/mlx5_flow.h       |  1 +
 drivers/net/mlx5/mlx5_flow_dv.c    | 21 +++++++++++++++++----
 drivers/net/mlx5/mlx5_flow_verbs.c | 16 +++++++++++++---
 5 files changed, 47 insertions(+), 13 deletions(-)

Comments

Slava Ovsiienko Aug. 3, 2021, 6:43 a.m. UTC | #1
> -----Original Message-----
> From: Rongwei Liu <rongweil@nvidia.com>
> Sent: Monday, August 2, 2021 15:21
> To: Matan Azrad <matan@nvidia.com>; Slava Ovsiienko
> <viacheslavo@nvidia.com>; Ori Kam <orika@nvidia.com>; NBU-Contact-
> Thomas Monjalon <thomas@monjalon.net>; Shahaf Shuler
> <shahafs@nvidia.com>; Raslan Darawsheh <rasland@nvidia.com>
> Cc: dev@dpdk.org; stable@dpdk.org
> Subject: [PATCH v2] net/mlx5: fix vni matching with non-std port at
> ConnectX-5
> 
> In the recent update, the misc5 matcher was introduced to match VxLAN
> header extra fields. However, ConnectX-5 doesn't support misc5 for the UDP
> ports different from VXLAN's standard one (4789).
> 
> Need to fall back to the previous approach and use legacy misc matcher if
> non-standard UDP port is recognized in VxLAN flow.
> 
> Fixes: 630a587bfb37 ("net/mlx5: support matching on VXLAN reserved field")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Slava Ovsiienko Aug. 3, 2021, 6:51 a.m. UTC | #2
> -----Original Message-----
> From: Rongwei Liu <rongweil@nvidia.com>
> Sent: Monday, August 2, 2021 15:21
> To: Matan Azrad <matan@nvidia.com>; Slava Ovsiienko
> <viacheslavo@nvidia.com>; Ori Kam <orika@nvidia.com>; NBU-Contact-
> Thomas Monjalon <thomas@monjalon.net>; Shahaf Shuler
> <shahafs@nvidia.com>; Raslan Darawsheh <rasland@nvidia.com>
> Cc: dev@dpdk.org; stable@dpdk.org
> Subject: [PATCH v2] net/mlx5: fix vni matching with non-std port at
> ConnectX-5
> 
> In the recent update, the misc5 matcher was introduced to match VxLAN
> header extra fields. However, ConnectX-5 doesn't support misc5 for the UDP
> ports different from VXLAN's standard one (4789).
> 
> Need to fall back to the previous approach and use legacy misc matcher if
> non-standard UDP port is recognized in VxLAN flow.
> 
> Fixes: 630a587bfb37 ("net/mlx5: support matching on VXLAN reserved field")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Thomas Monjalon Aug. 3, 2021, 12:45 p.m. UTC | #3
> > In the recent update, the misc5 matcher was introduced to match VxLAN
> > header extra fields. However, ConnectX-5 doesn't support misc5 for the UDP
> > ports different from VXLAN's standard one (4789).
> > 
> > Need to fall back to the previous approach and use legacy misc matcher if
> > non-standard UDP port is recognized in VxLAN flow.
> > 
> > Fixes: 630a587bfb37 ("net/mlx5: support matching on VXLAN reserved field")
> > Cc: stable@dpdk.org
> > 
> > Signed-off-by: Rongwei Liu <rongweil@nvidia.com>
> Acked-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>

new title: net/mlx5: fix VXLAN VNI matching on ConnectX-5

Applied, thanks.
diff mbox series

Patch

diff --git a/doc/guides/nics/mlx5.rst b/doc/guides/nics/mlx5.rst
index 3e9c736cae..a4d0f054d9 100644
--- a/doc/guides/nics/mlx5.rst
+++ b/doc/guides/nics/mlx5.rst
@@ -202,6 +202,7 @@  Limitations
 
   Last reserved 8-bits matching is only supported When using DV flow
   engine (``dv_flow_en`` = 1).
+  For ConnectX-5, the UDP destination port must be the standard one (4789).
   Group zero's behavior may differ which depends on FW.
   Matching value equals 0 (value & mask) is not supported.
 
diff --git a/drivers/net/mlx5/mlx5_flow.c b/drivers/net/mlx5/mlx5_flow.c
index 0689e6d45d..e63a297e2a 100644
--- a/drivers/net/mlx5/mlx5_flow.c
+++ b/drivers/net/mlx5/mlx5_flow.c
@@ -2432,6 +2432,8 @@  mlx5_flow_validate_item_tcp(const struct rte_flow_item *item,
  *
  * @param[in] dev
  *   Pointer to the Ethernet device structure.
+ * @param[in] udp_dport
+ *   UDP destination port
  * @param[in] item
  *   Item specification.
  * @param[in] item_flags
@@ -2446,6 +2448,7 @@  mlx5_flow_validate_item_tcp(const struct rte_flow_item *item,
  */
 int
 mlx5_flow_validate_item_vxlan(struct rte_eth_dev *dev,
+			      uint16_t udp_dport,
 			      const struct rte_flow_item *item,
 			      uint64_t item_flags,
 			      const struct rte_flow_attr *attr,
@@ -2481,12 +2484,18 @@  mlx5_flow_validate_item_vxlan(struct rte_eth_dev *dev,
 					  "no outer UDP layer found");
 	if (!mask)
 		mask = &rte_flow_item_vxlan_mask;
-	/* FDB domain & NIC domain non-zero group */
-	if ((attr->transfer || attr->group) && priv->sh->misc5_cap)
-		valid_mask = &nic_mask;
-	/* Group zero in NIC domain */
-	if (!attr->group && !attr->transfer && priv->sh->tunnel_header_0_1)
-		valid_mask = &nic_mask;
+
+	if (priv->sh->steering_format_version !=
+	    MLX5_STEERING_LOGIC_FORMAT_CONNECTX_5 ||
+	    !udp_dport || udp_dport == MLX5_UDP_PORT_VXLAN) {
+		/* FDB domain & NIC domain non-zero group */
+		if ((attr->transfer || attr->group) && priv->sh->misc5_cap)
+			valid_mask = &nic_mask;
+		/* Group zero in NIC domain */
+		if (!attr->group && !attr->transfer &&
+		    priv->sh->tunnel_header_0_1)
+			valid_mask = &nic_mask;
+	}
 	ret = mlx5_flow_item_acceptable
 		(item, (const uint8_t *)mask,
 		 (const uint8_t *)valid_mask,
diff --git a/drivers/net/mlx5/mlx5_flow.h b/drivers/net/mlx5/mlx5_flow.h
index 3724293d26..22fa007b42 100644
--- a/drivers/net/mlx5/mlx5_flow.h
+++ b/drivers/net/mlx5/mlx5_flow.h
@@ -1522,6 +1522,7 @@  int mlx5_flow_validate_item_vlan(const struct rte_flow_item *item,
 				 struct rte_eth_dev *dev,
 				 struct rte_flow_error *error);
 int mlx5_flow_validate_item_vxlan(struct rte_eth_dev *dev,
+				  uint16_t udp_dport,
 				  const struct rte_flow_item *item,
 				  uint64_t item_flags,
 				  const struct rte_flow_attr *attr,
diff --git a/drivers/net/mlx5/mlx5_flow_dv.c b/drivers/net/mlx5/mlx5_flow_dv.c
index 059bd25efc..ae0975e8b3 100644
--- a/drivers/net/mlx5/mlx5_flow_dv.c
+++ b/drivers/net/mlx5/mlx5_flow_dv.c
@@ -6926,6 +6926,7 @@  flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 	const struct rte_flow_item *rule_items = items;
 	const struct rte_flow_item *port_id_item = NULL;
 	bool def_policy = false;
+	uint16_t udp_dport = 0;
 
 	if (items == NULL)
 		return -1;
@@ -7100,6 +7101,14 @@  flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 			ret = mlx5_flow_validate_item_udp(items, item_flags,
 							  next_protocol,
 							  error);
+			const struct rte_flow_item_udp *spec = items->spec;
+			const struct rte_flow_item_udp *mask = items->mask;
+			if (!mask)
+				mask = &rte_flow_item_udp_mask;
+			if (spec != NULL)
+				udp_dport = rte_be_to_cpu_16
+						(spec->hdr.dst_port &
+						 mask->hdr.dst_port);
 			if (ret < 0)
 				return ret;
 			last_item = tunnel ? MLX5_FLOW_LAYER_INNER_L4_UDP :
@@ -7129,9 +7138,9 @@  flow_dv_validate(struct rte_eth_dev *dev, const struct rte_flow_attr *attr,
 			last_item = MLX5_FLOW_LAYER_GRE_KEY;
 			break;
 		case RTE_FLOW_ITEM_TYPE_VXLAN:
-			ret = mlx5_flow_validate_item_vxlan(dev, items,
-							    item_flags, attr,
-							    error);
+			ret = mlx5_flow_validate_item_vxlan(dev, udp_dport,
+							    items, item_flags,
+							    attr, error);
 			if (ret < 0)
 				return ret;
 			last_item = MLX5_FLOW_LAYER_VXLAN;
@@ -8927,6 +8936,7 @@  flow_dv_translate_item_vxlan(struct rte_eth_dev *dev,
 		MLX5_SET(fte_match_set_lyr_2_4, headers_m, udp_dport, 0xFFFF);
 		MLX5_SET(fte_match_set_lyr_2_4, headers_v, udp_dport, dport);
 	}
+	dport = MLX5_GET16(fte_match_set_lyr_2_4, headers_v, udp_dport);
 	if (!vxlan_v)
 		return;
 	if (!vxlan_m) {
@@ -8936,7 +8946,10 @@  flow_dv_translate_item_vxlan(struct rte_eth_dev *dev,
 		else
 			vxlan_m = &nic_mask;
 	}
-	if ((!attr->group && !attr->transfer && !priv->sh->tunnel_header_0_1) ||
+	if ((priv->sh->steering_format_version ==
+	    MLX5_STEERING_LOGIC_FORMAT_CONNECTX_5 &&
+	    dport != MLX5_UDP_PORT_VXLAN) ||
+	    (!attr->group && !attr->transfer && !priv->sh->tunnel_header_0_1) ||
 	    ((attr->group || attr->transfer) && !priv->sh->misc5_cap)) {
 		void *misc_m;
 		void *misc_v;
diff --git a/drivers/net/mlx5/mlx5_flow_verbs.c b/drivers/net/mlx5/mlx5_flow_verbs.c
index a36b8adf6b..b93fd4d2c9 100644
--- a/drivers/net/mlx5/mlx5_flow_verbs.c
+++ b/drivers/net/mlx5/mlx5_flow_verbs.c
@@ -1257,6 +1257,7 @@  flow_verbs_validate(struct rte_eth_dev *dev,
 	uint8_t next_protocol = 0xff;
 	uint16_t ether_type = 0;
 	bool is_empty_vlan = false;
+	uint16_t udp_dport = 0;
 
 	if (items == NULL)
 		return -1;
@@ -1364,6 +1365,15 @@  flow_verbs_validate(struct rte_eth_dev *dev,
 			ret = mlx5_flow_validate_item_udp(items, item_flags,
 							  next_protocol,
 							  error);
+			const struct rte_flow_item_udp *spec = items->spec;
+			const struct rte_flow_item_udp *mask = items->mask;
+			if (!mask)
+				mask = &rte_flow_item_udp_mask;
+			if (spec != NULL)
+				udp_dport = rte_be_to_cpu_16
+						(spec->hdr.dst_port &
+						 mask->hdr.dst_port);
+
 			if (ret < 0)
 				return ret;
 			last_item = tunnel ? MLX5_FLOW_LAYER_INNER_L4_UDP :
@@ -1381,9 +1391,9 @@  flow_verbs_validate(struct rte_eth_dev *dev,
 					     MLX5_FLOW_LAYER_OUTER_L4_TCP;
 			break;
 		case RTE_FLOW_ITEM_TYPE_VXLAN:
-			ret = mlx5_flow_validate_item_vxlan(dev, items,
-							    item_flags, attr,
-							    error);
+			ret = mlx5_flow_validate_item_vxlan(dev, udp_dport,
+							    items, item_flags,
+							    attr, error);
 			if (ret < 0)
 				return ret;
 			last_item = MLX5_FLOW_LAYER_VXLAN;