tcp: add tracepoint for checksum errors

Add a tracepoint for capturing TCP segments with
a bad checksum. This makes it easy to identify
sources of bad frames in the fleet (e.g. machines
with faulty NICs).

It should also help tools like IOvisor's tcpdrop.py
which are used today to get detailed information
about such packets.

We don't have a socket in many cases so we must
open code the address extraction based just on
the skb.

v2: add missing export for ipv6=m

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 83 insertions, 0 deletions

diff --git a/include/trace/events/tcp.h b/include/trace/events/tcp.h
index ba94857eea11..521059d8dc0a 100644
--- a/include/trace/events/tcp.h
+++ b/include/trace/events/tcp.h
@@ -295,6 +295,82 @@ TRACE_EVENT(tcp_probe,
 		  __entry->srtt, __entry->rcv_wnd, __entry->sock_cookie)
 );
 
+#define TP_STORE_ADDR_PORTS_SKB_V4(__entry, skb)			\
+	do {								\
+		const struct tcphdr *th = (const struct tcphdr *)skb->data; \
+		struct sockaddr_in *v4 = (void *)__entry->saddr;	\
+									\
+		v4->sin_family = AF_INET;				\
+		v4->sin_port = th->source;				\
+		v4->sin_addr.s_addr = ip_hdr(skb)->saddr;		\
+		v4 = (void *)__entry->daddr;				\
+		v4->sin_family = AF_INET;				\
+		v4->sin_port = th->dest;				\
+		v4->sin_addr.s_addr = ip_hdr(skb)->daddr;		\
+	} while (0)
+
+#if IS_ENABLED(CONFIG_IPV6)
+
+#define TP_STORE_ADDR_PORTS_SKB(__entry, skb)				\
+	do {								\
+		const struct iphdr *iph = ip_hdr(skb);			\
+									\
+		if (iph->version == 6) {				\
+			const struct tcphdr *th = (const struct tcphdr *)skb->data; \
+			struct sockaddr_in6 *v6 = (void *)__entry->saddr; \
+									\
+			v6->sin6_family = AF_INET6;			\
+			v6->sin6_port = th->source;			\
+			v6->sin6_addr = ipv6_hdr(skb)->saddr;		\
+			v6 = (void *)__entry->daddr;			\
+			v6->sin6_family = AF_INET6;			\
+			v6->sin6_port = th->dest;			\
+			v6->sin6_addr = ipv6_hdr(skb)->daddr;		\
+		} else							\
+			TP_STORE_ADDR_PORTS_SKB_V4(__entry, skb);	\
+	} while (0)
+
+#else
+
+#define TP_STORE_ADDR_PORTS_SKB(__entry, skb)		\
+	TP_STORE_ADDR_PORTS_SKB_V4(__entry, skb)
+
+#endif
+
+/*
+ * tcp event with only skb
+ */
+DECLARE_EVENT_CLASS(tcp_event_skb,
+
+	TP_PROTO(const struct sk_buff *skb),
+
+	TP_ARGS(skb),
+
+	TP_STRUCT__entry(
+		__field(const void *, skbaddr)
+		__array(__u8, saddr, sizeof(struct sockaddr_in6))
+		__array(__u8, daddr, sizeof(struct sockaddr_in6))
+	),
+
+	TP_fast_assign(
+		__entry->skbaddr = skb;
+
+		memset(__entry->saddr, 0, sizeof(struct sockaddr_in6));
+		memset(__entry->daddr, 0, sizeof(struct sockaddr_in6));
+
+		TP_STORE_ADDR_PORTS_SKB(__entry, skb);
+	),
+
+	TP_printk("src=%pISpc dest=%pISpc", __entry->saddr, __entry->daddr)
+);
+
+DEFINE_EVENT(tcp_event_skb, tcp_bad_csum,
+
+	TP_PROTO(const struct sk_buff *skb),
+
+	TP_ARGS(skb)
+);
+
 #endif /* _TRACE_TCP_H */
 
 /* This part must be outside protection */
diff --git a/net/core/net-traces.c b/net/core/net-traces.c
index 283ddb2dbc7d..c40cd8dd75c7 100644
--- a/net/core/net-traces.c
+++ b/net/core/net-traces.c
@@ -60,3 +60,4 @@ EXPORT_TRACEPOINT_SYMBOL_GPL(kfree_skb);
 EXPORT_TRACEPOINT_SYMBOL_GPL(napi_poll);
 
 EXPORT_TRACEPOINT_SYMBOL_GPL(tcp_send_reset);
+EXPORT_TRACEPOINT_SYMBOL_GPL(tcp_bad_csum);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 4cf4dd532d1c..cd52ce0a2a85 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -5885,6 +5885,7 @@ step5:
 	return;
 
 csum_error:
+	trace_tcp_bad_csum(skb);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 312184cead57..4f5b68a90be9 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1731,6 +1731,7 @@ discard:
 	return 0;
 
 csum_err:
+	trace_tcp_bad_csum(skb);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
 	goto discard;
@@ -1801,6 +1802,7 @@ bool tcp_add_backlog(struct sock *sk, struct sk_buff *skb)
 
 	if (unlikely(tcp_checksum_complete(skb))) {
 		bh_unlock_sock(sk);
+		trace_tcp_bad_csum(skb);
 		__TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
 		__TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
 		return true;
@@ -2098,6 +2100,7 @@ no_tcp_socket:
 
 	if (tcp_checksum_complete(skb)) {
 csum_error:
+		trace_tcp_bad_csum(skb);
 		__TCP_INC_STATS(net, TCP_MIB_CSUMERRORS);
 bad_packet:
 		__TCP_INC_STATS(net, TCP_MIB_INERRS);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 5f47c0b6e3de..4435fa342e7a 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1538,6 +1538,7 @@ discard:
 	kfree_skb(skb);
 	return 0;
 csum_err:
+	trace_tcp_bad_csum(skb);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_CSUMERRORS);
 	TCP_INC_STATS(sock_net(sk), TCP_MIB_INERRS);
 	goto discard;
@@ -1754,6 +1755,7 @@ no_tcp_socket:
 
 	if (tcp_checksum_complete(skb)) {
 csum_error:
+		trace_tcp_bad_csum(skb);
 		__TCP_INC_STATS(net, TCP_MIB_CSUMERRORS);
 bad_packet:
 		__TCP_INC_STATS(net, TCP_MIB_INERRS);