Merge branch 'erspan-version-2'

William Tu says:

====================
ERSPAN version 2 (type III) support

ERSPAN has two versions, v1 (type II) and v2 (type III).  This patch
series add support for erspan v2 based on existing erspan v1
implementation.  The first patch refactors the existing erspan v1's
header structure, making it extensible to put additional v2's header.
The second and third patch introduces erspan v2's implementation to
ipv4 and ipv6 erspan, for both native mode and collect metadata mode.
Finally, test cases are added under the samples/bpf.

Note:
ERSPAN version 2 has many features and this patch does not implement
all.  One major use case of version 2 over version 1 is its timestamp
and direction.  So the traffic collector is able to distinguish the
mirrorred traffic better.  Other features such as SGT (security group
tag), FT (frame type) for carrying non-ethernet packet, and optional
subheader are not implemented yet.

Example commandline for ERSPAN version 2:
ip link add dev ip6erspan11 type ip6erspan seq key 102 \
	local fc00:100::2 remote fc00:100::1 \
	erspan_ver 2 erspan_dir 1 erspan_hwid 17

The corresponding iproute2 patch:
https://marc.info/?l=linux-netdev&m=151321141525106&w=2

William Tu (4):
  net: erspan: refactor existing erspan code
  net: erspan: introduce erspan v2 for ip_gre
  ip6_gre: add erspan v2 support
  samples/bpf: add erspan v2 sample code

 include/net/erspan.h           | 152 ++++++++++++++++++++++++++++++++++++++---
 include/net/ip6_tunnel.h       |   3 +
 include/net/ip_tunnels.h       |   5 +-
 include/uapi/linux/if_ether.h  |   1 +
 include/uapi/linux/if_tunnel.h |   3 +
 net/ipv4/ip_gre.c              | 124 +++++++++++++++++++++++++++------
 net/ipv6/ip6_gre.c             | 139 +++++++++++++++++++++++++++++++------
 net/openvswitch/flow_netlink.c |   8 +--
 samples/bpf/tcbpf2_kern.c      |  77 ++++++++++++++++++---
 samples/bpf/test_tunnel_bpf.sh |  38 ++++++++---
 10 files changed, 472 insertions(+), 78 deletions(-)

--
A simple script to test it:

set -ex
function cleanup() {
	set +ex
	ip netns del ns0
	ip link del ip6erspan11
	ip link del veth1
}

function main() {
	trap cleanup 0 2 3 9

	ip netns add ns0
	ip link add veth0 type veth peer name veth1
	ip link set veth0 netns ns0

	# non-namespace
	ip addr add dev veth1 fc00:100::2/96

	if [ "$1" == "v1" ]; then
		echo "create IP6 ERSPAN v1 tunnel"
		ip link add dev ip6erspan11 type ip6erspan seq key 102 \
			local fc00:100::2 remote fc00:100::1 \
			erspan 123 erspan_ver 1
	else
		echo "create IP6 ERSPAN v2 tunnel"
		ip link add dev ip6erspan11 type ip6erspan seq key 102 \
			local fc00:100::2 remote fc00:100::1 \
			erspan_ver 2 erspan_dir 1 erspan_hwid 17
	fi
	ip addr add dev ip6erspan11 fc00:200::2/96
	ip addr add dev ip6erspan11 10.10.200.2/24

	# namespace: ns0
	ip netns exec ns0 ip addr add fc00:100::1/96 dev veth0

	if [ "$1" == "v1" ]; then
		ip netns exec ns0 \
		ip link add dev ip6erspan00 type ip6erspan seq key 102 \
			local fc00:100::1 remote fc00:100::2 \
			erspan 123 erspan_ver 1
	else
		ip netns exec ns0 \
		ip link add dev ip6erspan00 type ip6erspan seq key 102 \
			local fc00:100::1 remote fc00:100::2 \
			erspan_ver 2 erspan_dir 1 erspan_hwid 7
	fi

	ip netns exec ns0 ip addr add dev ip6erspan00 fc00:200::1/96
	ip netns exec ns0 ip addr add dev ip6erspan00 10.10.200.1/24

	ip link set dev veth1 up
	ip link set dev ip6erspan11 up
	ip netns exec ns0 ip link set dev ip6erspan00 up
	ip netns exec ns0 ip link set dev veth0 up
}

main $1

ping6 -c 1 fc00:100::1 || true

ping -c 3 10.10.200.1
exit 0
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 152 insertions, 12 deletions

diff --git a/include/net/erspan.h b/include/net/erspan.h
index 6e758d08c9ee..acdf6843095d 100644
--- a/include/net/erspan.h
+++ b/include/net/erspan.h
@@ -15,7 +15,7 @@
  *  s, Recur, Flags, Version fields only S (bit 03) is set to 1. The
  *  other fields are set to zero, so only a sequence number follows.
  *
- *  ERSPAN Type II header (8 octets [42:49])
+ *  ERSPAN Version 1 (Type II) header (8 octets [42:49])
  *  0                   1                   2                   3
  *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
  * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
@@ -24,11 +24,29 @@
  * |      Reserved         |                  Index                |
  * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
  *
+ *
+ *  ERSPAN Version 2 (Type III) header (12 octets [42:49])
+ *  0                   1                   2                   3
+ *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |  Ver  |          VLAN         | COS |BSO|T|     Session ID    |
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |                          Timestamp                            |
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |             SGT               |P|    FT   |   Hw ID   |D|Gra|O|
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *
+ *      Platform Specific SubHeader (8 octets, optional)
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |  Platf ID |               Platform Specific Info              |
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ * |                  Platform Specific Info                       |
+ * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
+ *
  * GRE proto ERSPAN type II = 0x88BE, type III = 0x22EB
  */
 
-#define ERSPAN_VERSION	0x1
-
+#define ERSPAN_VERSION	0x1	/* ERSPAN type II */
 #define VER_MASK	0xf000
 #define VLAN_MASK	0x0fff
 #define COS_MASK	0xe000
@@ -37,6 +55,28 @@
 #define ID_MASK		0x03ff
 #define INDEX_MASK	0xfffff
 
+#define ERSPAN_VERSION2	0x2	/* ERSPAN type III*/
+#define BSO_MASK	EN_MASK
+#define SGT_MASK	0xffff0000
+#define P_MASK		0x8000
+#define FT_MASK		0x7c00
+#define HWID_MASK	0x03f0
+#define DIR_MASK	0x0008
+#define GRA_MASK	0x0006
+#define O_MASK		0x0001
+
+/* ERSPAN version 2 metadata header */
+struct erspan_md2 {
+	__be32 timestamp;
+	__be16 sgt;	/* security group tag */
+	__be16 flags;
+#define P_OFFSET	15
+#define FT_OFFSET	10
+#define HWID_OFFSET	4
+#define DIR_OFFSET	3
+#define GRA_OFFSET	1
+};
+
 enum erspan_encap_type {
 	ERSPAN_ENCAP_NOVLAN = 0x0,	/* originally without VLAN tag */
 	ERSPAN_ENCAP_ISL = 0x1,		/* originally ISL encapsulated */
@@ -44,20 +84,32 @@ enum erspan_encap_type {
 	ERSPAN_ENCAP_INFRAME = 0x3,	/* VLAN tag perserved in frame */
 };
 
+#define ERSPAN_V1_MDSIZE	4
+#define ERSPAN_V2_MDSIZE	8
 struct erspan_metadata {
-	__be32 index;   /* type II */
+	union {
+		__be32 index;		/* Version 1 (type II)*/
+		struct erspan_md2 md2;	/* Version 2 (type III) */
+	} u;
+	int version;
 };
 
-struct erspanhdr {
+struct erspan_base_hdr {
 	__be16 ver_vlan;
 #define VER_OFFSET  12
 	__be16 session_id;
 #define COS_OFFSET  13
 #define EN_OFFSET   11
+#define BSO_OFFSET  EN_OFFSET
 #define T_OFFSET    10
-	struct erspan_metadata md;
 };
 
+static inline int erspan_hdr_len(int version)
+{
+	return sizeof(struct erspan_base_hdr) +
+	       (version == 1 ? ERSPAN_V1_MDSIZE : ERSPAN_V2_MDSIZE);
+}
+
 static inline u8 tos_to_cos(u8 tos)
 {
 	u8 dscp, cos;
@@ -73,7 +125,8 @@ static inline void erspan_build_header(struct sk_buff *skb,
 {
 	struct ethhdr *eth = eth_hdr(skb);
 	enum erspan_encap_type enc_type;
-	struct erspanhdr *ershdr;
+	struct erspan_base_hdr *ershdr;
+	struct erspan_metadata *ersmd;
 	struct qtag_prefix {
 		__be16 eth_type;
 		__be16 tci;
@@ -96,17 +149,94 @@ static inline void erspan_build_header(struct sk_buff *skb,
 		enc_type = ERSPAN_ENCAP_INFRAME;
 	}
 
-	skb_push(skb, sizeof(*ershdr));
-	ershdr = (struct erspanhdr *)skb->data;
-	memset(ershdr, 0, sizeof(*ershdr));
+	skb_push(skb, sizeof(*ershdr) + ERSPAN_V1_MDSIZE);
+	ershdr = (struct erspan_base_hdr *)skb->data;
+	memset(ershdr, 0, sizeof(*ershdr) + ERSPAN_V1_MDSIZE);
 
+	/* Build base header */
 	ershdr->ver_vlan = htons((vlan_tci & VLAN_MASK) |
 				 (ERSPAN_VERSION << VER_OFFSET));
 	ershdr->session_id = htons((u16)(ntohl(id) & ID_MASK) |
 			   ((tos_to_cos(tos) << COS_OFFSET) & COS_MASK) |
 			   (enc_type << EN_OFFSET & EN_MASK) |
 			   ((truncate << T_OFFSET) & T_MASK));
-	ershdr->md.index = htonl(index & INDEX_MASK);
+
+	/* Build metadata */
+	ersmd = (struct erspan_metadata *)(ershdr + 1);
+	ersmd->u.index = htonl(index & INDEX_MASK);
+}
+
+/* ERSPAN GRA: timestamp granularity
+ *   00b --> granularity = 100 microseconds
+ *   01b --> granularity = 100 nanoseconds
+ *   10b --> granularity = IEEE 1588
+ * Here we only support 100 microseconds.
+ */
+static inline __be32 erspan_get_timestamp(void)
+{
+	u64 h_usecs;
+	ktime_t kt;
+
+	kt = ktime_get_real();
+	h_usecs = ktime_divns(kt, 100 * NSEC_PER_USEC);
+
+	/* ERSPAN base header only has 32-bit,
+	 * so it wraps around 4 days.
+	 */
+	return htonl((u32)h_usecs);
+}
+
+static inline void erspan_build_header_v2(struct sk_buff *skb,
+					  __be32 id, u8 direction, u16 hwid,
+					  bool truncate, bool is_ipv4)
+{
+	struct ethhdr *eth = eth_hdr(skb);
+	struct erspan_base_hdr *ershdr;
+	struct erspan_metadata *md;
+	struct qtag_prefix {
+		__be16 eth_type;
+		__be16 tci;
+	} *qp;
+	u16 vlan_tci = 0;
+	u16 session_id;
+	u8 gra = 0; /* 100 usec */
+	u8 bso = 0; /* Bad/Short/Oversized */
+	u8 sgt = 0;
+	u8 tos;
+
+	tos = is_ipv4 ? ip_hdr(skb)->tos :
+			(ipv6_hdr(skb)->priority << 4) +
+			(ipv6_hdr(skb)->flow_lbl[0] >> 4);
+
+	/* Unlike v1, v2 does not have En field,
+	 * so only extract vlan tci field.
+	 */
+	if (eth->h_proto == htons(ETH_P_8021Q)) {
+		qp = (struct qtag_prefix *)(skb->data + 2 * ETH_ALEN);
+		vlan_tci = ntohs(qp->tci);
+	}
+
+	skb_push(skb, sizeof(*ershdr) + ERSPAN_V2_MDSIZE);
+	ershdr = (struct erspan_base_hdr *)skb->data;
+	memset(ershdr, 0, sizeof(*ershdr) + ERSPAN_V2_MDSIZE);
+
+	/* Build base header */
+	ershdr->ver_vlan = htons((vlan_tci & VLAN_MASK) |
+				 (ERSPAN_VERSION2 << VER_OFFSET));
+	session_id = (u16)(ntohl(id) & ID_MASK) |
+		     ((tos_to_cos(tos) << COS_OFFSET) & COS_MASK) |
+		     (bso << BSO_OFFSET & BSO_MASK) |
+		     ((truncate << T_OFFSET) & T_MASK);
+	ershdr->session_id = htons(session_id);
+
+	/* Build metadata */
+	md = (struct erspan_metadata *)(ershdr + 1);
+	md->u.md2.timestamp = erspan_get_timestamp();
+	md->u.md2.sgt = htons(sgt);
+	md->u.md2.flags = htons(((1 << P_OFFSET) & P_MASK) |
+				((hwid << HWID_OFFSET) & HWID_MASK) |
+				((direction << DIR_OFFSET) & DIR_MASK) |
+				((gra << GRA_OFFSET) & GRA_MASK));
 }
 
 #endif
diff --git a/include/net/ip6_tunnel.h b/include/net/ip6_tunnel.h
index 109a5a8877ef..236e40ba06bf 100644
--- a/include/net/ip6_tunnel.h
+++ b/include/net/ip6_tunnel.h
@@ -37,6 +37,9 @@ struct __ip6_tnl_parm {
 
 	__u32			fwmark;
 	__u32			index;	/* ERSPAN type II index */
+	__u8			erspan_ver;	/* ERSPAN version */
+	__u8			dir;	/* direction */
+	__u16			hwid;	/* hwid */
 };
 
 /* IPv6 tunnel */
diff --git a/include/net/ip_tunnels.h b/include/net/ip_tunnels.h
index 24628f6b09bf..1f16773cfd76 100644
--- a/include/net/ip_tunnels.h
+++ b/include/net/ip_tunnels.h
@@ -116,8 +116,11 @@ struct ip_tunnel {
 	u32		o_seqno;	/* The last output seqno */
 	int		tun_hlen;	/* Precalculated header length */
 
-	/* This field used only by ERSPAN */
+	/* These four fields used only by ERSPAN */
 	u32		index;		/* ERSPAN type II index */
+	u8		erspan_ver;	/* ERSPAN version */
+	u8		dir;		/* ERSPAN direction */
+	u16		hwid;		/* ERSPAN hardware ID */
 
 	struct dst_cache dst_cache;
 
diff --git a/include/uapi/linux/if_ether.h b/include/uapi/linux/if_ether.h
index 3ee3bf7c8526..87b7529fcdfe 100644
--- a/include/uapi/linux/if_ether.h
+++ b/include/uapi/linux/if_ether.h
@@ -47,6 +47,7 @@
 #define ETH_P_PUP	0x0200		/* Xerox PUP packet		*/
 #define ETH_P_PUPAT	0x0201		/* Xerox PUP Addr Trans packet	*/
 #define ETH_P_TSN	0x22F0		/* TSN (IEEE 1722) packet	*/
+#define ETH_P_ERSPAN2	0x22EB		/* ERSPAN version 2 (type III)	*/
 #define ETH_P_IP	0x0800		/* Internet Protocol packet	*/
 #define ETH_P_X25	0x0805		/* CCITT X.25			*/
 #define ETH_P_ARP	0x0806		/* Address Resolution packet	*/
diff --git a/include/uapi/linux/if_tunnel.h b/include/uapi/linux/if_tunnel.h
index e68dadbd6d45..1b3d148c4560 100644
--- a/include/uapi/linux/if_tunnel.h
+++ b/include/uapi/linux/if_tunnel.h
@@ -137,6 +137,9 @@ enum {
 	IFLA_GRE_IGNORE_DF,
 	IFLA_GRE_FWMARK,
 	IFLA_GRE_ERSPAN_INDEX,
+	IFLA_GRE_ERSPAN_VER,
+	IFLA_GRE_ERSPAN_DIR,
+	IFLA_GRE_ERSPAN_HWID,
 	__IFLA_GRE_MAX,
 };