[PATCH net-next v4 0/3] Fix traceroute in the presence of SRv6

[PATCH net-next v4 0/3] Fix traceroute in the presence of SRv6

[Andrew Lunn]

When using SRv6 the destination IP address in the IPv6 header is not
always the true destination, it can be a router along the path that
SRv6 is using.

When ICMP reports an error, e.g, time exceeded, which is what
traceroute uses, it included the packet which invoked the error into
the ICMP message body. Upon receiving such an ICMP packet, the
invoking packet is examined and an attempt is made to find the socket
which sent the packet, so the error can be reported. Lookup is
performed using the source and destination address. If the
intermediary router IP address from the IP header is used, the lookup
fails. It is necessary to dig into the header and find the true
destination address in the Segment Router header, SRH.

v2:
Play games with the skb->network_header rather than clone the skb
v3:
Move helpers into seg6.c
v4:
Move short helper into header file.
Rework getting SRH destination address

Patch 1 exports a helper which can find the SRH in a packet
Patch 2 does the actual examination of the invoking packet
Patch 3 makes use of the results when trying to find the socket.

Andrew Lunn (3):
  seg6: export get_srh() for ICMP handling
  icmp: ICMPV6: Examine invoking packet for Segment Route Headers.
  udp6: Use Segment Routing Header for dest address if present

 include/linux/ipv6.h  |  2 ++
 include/net/seg6.h    |  1 +
 net/ipv6/icmp.c       | 36 +++++++++++++++++++++++++++++++++++-
 net/ipv6/seg6.c       | 29 +++++++++++++++++++++++++++++
 net/ipv6/seg6_local.c | 33 ++-------------------------------
 net/ipv6/udp.c        |  7 +++++++
 6 files changed, 76 insertions(+), 32 deletions(-)

-- 
2.33.1

[PATCH net-next v4 1/3] seg6: export get_srh() for ICMP handling

[Andrew Lunn]

An ICMP error message can contain in its message body part of an IPv6
packet which invoked the error. Such a packet might contain a segment
router header. Export get_srh() so the ICMP code can make use of it.

Since his changes the scope of the function from local to global, add
the seg6_ prefix to keep the namespace clean. And move it into seg6.c
so it is always available, not just when IPV6_SEG6_LWTUNNEL is
enabled.

Signed-off-by: Andrew Lunn <andrew@lunn.ch>
---
 include/net/seg6.h    |  1 +
 net/ipv6/seg6.c       | 29 +++++++++++++++++++++++++++++
 net/ipv6/seg6_local.c | 33 ++-------------------------------
 3 files changed, 32 insertions(+), 31 deletions(-)

diff --git a/include/net/seg6.h b/include/net/seg6.h
index 9d19c15e8545..a6f25983670a 100644
--- a/include/net/seg6.h
+++ b/include/net/seg6.h
@@ -58,6 +58,7 @@ extern int seg6_local_init(void);
 extern void seg6_local_exit(void);
 
 extern bool seg6_validate_srh(struct ipv6_sr_hdr *srh, int len, bool reduced);
+extern struct ipv6_sr_hdr *seg6_get_srh(struct sk_buff *skb, int flags);
 extern int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
 			     int proto);
 extern int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh);
diff --git a/net/ipv6/seg6.c b/net/ipv6/seg6.c
index a8b5784afb1a..5bc9bf892199 100644
--- a/net/ipv6/seg6.c
+++ b/net/ipv6/seg6.c
@@ -75,6 +75,35 @@ bool seg6_validate_srh(struct ipv6_sr_hdr *srh, int len, bool reduced)
 	return true;
 }
 
+struct ipv6_sr_hdr *seg6_get_srh(struct sk_buff *skb, int flags)
+{
+	struct ipv6_sr_hdr *srh;
+	int len, srhoff = 0;
+
+	if (ipv6_find_hdr(skb, &srhoff, IPPROTO_ROUTING, NULL, &flags) < 0)
+		return NULL;
+
+	if (!pskb_may_pull(skb, srhoff + sizeof(*srh)))
+		return NULL;
+
+	srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
+
+	len = (srh->hdrlen + 1) << 3;
+
+	if (!pskb_may_pull(skb, srhoff + len))
+		return NULL;
+
+	/* note that pskb_may_pull may change pointers in header;
+	 * for this reason it is necessary to reload them when needed.
+	 */
+	srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
+
+	if (!seg6_validate_srh(srh, len, true))
+		return NULL;
+
+	return srh;
+}
+
 static struct genl_family seg6_genl_family;
 
 static const struct nla_policy seg6_genl_policy[SEG6_ATTR_MAX + 1] = {
diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c
index 2dc40b3f373e..ef88489c71f5 100644
--- a/net/ipv6/seg6_local.c
+++ b/net/ipv6/seg6_local.c
@@ -150,40 +150,11 @@ static struct seg6_local_lwt *seg6_local_lwtunnel(struct lwtunnel_state *lwt)
 	return (struct seg6_local_lwt *)lwt->data;
 }
 
-static struct ipv6_sr_hdr *get_srh(struct sk_buff *skb, int flags)
-{
-	struct ipv6_sr_hdr *srh;
-	int len, srhoff = 0;
-
-	if (ipv6_find_hdr(skb, &srhoff, IPPROTO_ROUTING, NULL, &flags) < 0)
-		return NULL;
-
-	if (!pskb_may_pull(skb, srhoff + sizeof(*srh)))
-		return NULL;
-
-	srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
-
-	len = (srh->hdrlen + 1) << 3;
-
-	if (!pskb_may_pull(skb, srhoff + len))
-		return NULL;
-
-	/* note that pskb_may_pull may change pointers in header;
-	 * for this reason it is necessary to reload them when needed.
-	 */
-	srh = (struct ipv6_sr_hdr *)(skb->data + srhoff);
-
-	if (!seg6_validate_srh(srh, len, true))
-		return NULL;
-
-	return srh;
-}
-
 static struct ipv6_sr_hdr *get_and_validate_srh(struct sk_buff *skb)
 {
 	struct ipv6_sr_hdr *srh;
 
-	srh = get_srh(skb, IP6_FH_F_SKIP_RH);
+	srh = seg6_get_srh(skb, IP6_FH_F_SKIP_RH);
 	if (!srh)
 		return NULL;
 
@@ -200,7 +171,7 @@ static bool decap_and_validate(struct sk_buff *skb, int proto)
 	struct ipv6_sr_hdr *srh;
 	unsigned int off = 0;
 
-	srh = get_srh(skb, 0);
+	srh = seg6_get_srh(skb, 0);
 	if (srh && srh->segments_left > 0)
 		return false;
 
-- 
2.33.1

[PATCH net-next v4 1/3] seg6: export get_srh() for ICMP handling

[Andrew Lunn]

From 387a5e9d6b2749d0457ccc760a5b785c2df8f799 Mon Sep 17 00:00:00 2001
From: Andrew Lunn <andrew@lunn.ch>
Date: Sat, 20 Nov 2021 12:51:07 -0600
Subject: [PATCH net-next v4 2/3] icmp: ICMPV6: Examine invoking packet for
 Segment Route Headers.

RFC8754 says:

ICMP error packets generated within the SR domain are sent to source
nodes within the SR domain.  The invoking packet in the ICMP error
message may contain an SRH.  Since the destination address of a packet
with an SRH changes as each segment is processed, it may not be the
destination used by the socket or application that generated the
invoking packet.

For the source of an invoking packet to process the ICMP error
message, the ultimate destination address of the IPv6 header may be
required.  The following logic is used to determine the destination
address for use by protocol-error handlers.

*  Walk all extension headers of the invoking IPv6 packet to the
   routing extension header preceding the upper-layer header.

   -  If routing header is type 4 Segment Routing Header (SRH)

      o  The SID at Segment List[0] may be used as the destination
         address of the invoking packet.

Mangle the skb so the network header points to the invoking packet
inside the ICMP packet. The seg6 helpers can then be used on the skb
to find any segment routing headers. If found, mark this fact in the
IPv6 control block of the skb, and store the offset into the packet of
the SRH. Then restore the skb back to its old state.

Signed-off-by: Andrew Lunn <andrew@lunn.ch>
---
 include/linux/ipv6.h |  2 ++
 include/net/seg6.h   |  1 +
 net/ipv6/icmp.c      |  6 +++++-
 net/ipv6/seg6.c      | 30 ++++++++++++++++++++++++++++++
 4 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index 20c1f968da7c..a59d25f19385 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -133,6 +133,7 @@ struct inet6_skb_parm {
 	__u16			dsthao;
 #endif
 	__u16			frag_max_size;
+	__u16			srhoff;
 
 #define IP6SKB_XFRM_TRANSFORMED	1
 #define IP6SKB_FORWARDED	2
@@ -142,6 +143,7 @@ struct inet6_skb_parm {
 #define IP6SKB_HOPBYHOP        32
 #define IP6SKB_L3SLAVE         64
 #define IP6SKB_JUMBOGRAM      128
+#define IP6SKB_SEG6	      256
 };
 
 #if defined(CONFIG_NET_L3_MASTER_DEV)
diff --git a/include/net/seg6.h b/include/net/seg6.h
index a6f25983670a..02b0cd305787 100644
--- a/include/net/seg6.h
+++ b/include/net/seg6.h
@@ -59,6 +59,7 @@ extern void seg6_local_exit(void);
 
 extern bool seg6_validate_srh(struct ipv6_sr_hdr *srh, int len, bool reduced);
 extern struct ipv6_sr_hdr *seg6_get_srh(struct sk_buff *skb, int flags);
+extern void seg6_icmp_srh(struct sk_buff *skb, struct inet6_skb_parm *opt);
 extern int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
 			     int proto);
 extern int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh);
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index a7c31ab67c5d..96c5cc0f30ce 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -57,6 +57,7 @@
 #include <net/protocol.h>
 #include <net/raw.h>
 #include <net/rawv6.h>
+#include <net/seg6.h>
 #include <net/transp_v6.h>
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
@@ -820,6 +821,7 @@ static void icmpv6_echo_reply(struct sk_buff *skb)
 
 void icmpv6_notify(struct sk_buff *skb, u8 type, u8 code, __be32 info)
 {
+	struct inet6_skb_parm *opt = IP6CB(skb);
 	const struct inet6_protocol *ipprot;
 	int inner_offset;
 	__be16 frag_off;
@@ -829,6 +831,8 @@ void icmpv6_notify(struct sk_buff *skb, u8 type, u8 code, __be32 info)
 	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
 		goto out;
 
+	seg6_icmp_srh(skb, opt);
+
 	nexthdr = ((struct ipv6hdr *)skb->data)->nexthdr;
 	if (ipv6_ext_hdr(nexthdr)) {
 		/* now skip over extension headers */
@@ -853,7 +857,7 @@ void icmpv6_notify(struct sk_buff *skb, u8 type, u8 code, __be32 info)
 
 	ipprot = rcu_dereference(inet6_protos[nexthdr]);
 	if (ipprot && ipprot->err_handler)
-		ipprot->err_handler(skb, NULL, type, code, inner_offset, info);
+		ipprot->err_handler(skb, opt, type, code, inner_offset, info);
 
 	raw6_icmp_error(skb, nexthdr, type, code, inner_offset, info);
 	return;
diff --git a/net/ipv6/seg6.c b/net/ipv6/seg6.c
index 5bc9bf892199..73aaabf0e966 100644
--- a/net/ipv6/seg6.c
+++ b/net/ipv6/seg6.c
@@ -104,6 +104,36 @@ struct ipv6_sr_hdr *seg6_get_srh(struct sk_buff *skb, int flags)
 	return srh;
 }
 
+/* Determine if an ICMP invoking packet contains a segment routing
+ * header.  If it does, extract the offset to the true destination
+ * address, which is in the first segment address.
+ */
+void seg6_icmp_srh(struct sk_buff *skb, struct inet6_skb_parm *opt)
+{
+	__u16 network_header = skb->network_header;
+	struct ipv6_sr_hdr *srh;
+
+	/* Update network header to point to the invoking packet
+	 * inside the ICMP packet, so we can use the seg6_get_srh()
+	 * helper.
+	 */
+	skb_reset_network_header(skb);
+
+	srh = seg6_get_srh(skb, 0);
+	if (!srh)
+		goto out;
+
+	if (srh->type != IPV6_SRCRT_TYPE_4)
+		goto out;
+
+	opt->flags |= IP6SKB_SEG6;
+	opt->srhoff = (unsigned char *)srh - skb->data;
+
+out:
+	/* Restore the network header back to the ICMP packet */
+	skb->network_header = network_header;
+}
+
 static struct genl_family seg6_genl_family;
 
 static const struct nla_policy seg6_genl_policy[SEG6_ATTR_MAX + 1] = {
-- 
2.33.1

Re: [PATCH net-next v4 1/3] seg6: export get_srh() for ICMP handling

[Willem de Bruijn]

On Wed, Dec 8, 2021 at 12:38 PM Andrew Lunn <andrew@lunn.ch> wrote:
>
> From 387a5e9d6b2749d0457ccc760a5b785c2df8f799 Mon Sep 17 00:00:00 2001
> From: Andrew Lunn <andrew@lunn.ch>
> Date: Sat, 20 Nov 2021 12:51:07 -0600
> Subject: [PATCH net-next v4 2/3] icmp: ICMPV6: Examine invoking packet for
>  Segment Route Headers.

Something went wrong when sending this patch series.

The above header is part of the commit message, and this patch is
marked as 1/3. See also in
https://patchwork.kernel.org/project/netdevbpf

Otherwise the code looks good to me.

Re: [PATCH net-next v4 1/3] seg6: export get_srh() for ICMP handling

[Andrew Lunn]

On Wed, Dec 08, 2021 at 04:24:56PM -0500, Willem de Bruijn wrote:
> On Wed, Dec 8, 2021 at 12:38 PM Andrew Lunn <andrew@lunn.ch> wrote:
> >
> > From 387a5e9d6b2749d0457ccc760a5b785c2df8f799 Mon Sep 17 00:00:00 2001
> > From: Andrew Lunn <andrew@lunn.ch>
> > Date: Sat, 20 Nov 2021 12:51:07 -0600
> > Subject: [PATCH net-next v4 2/3] icmp: ICMPV6: Examine invoking packet for
> >  Segment Route Headers.
> 
> Something went wrong when sending this patch series.
> 
> The above header is part of the commit message, and this patch is
> marked as 1/3. See also in
> https://patchwork.kernel.org/project/netdevbpf

Now that is odd. I just did the usual git format-patch ; git
send-email.

I will try out v5 sending it just to myself first.

Thanks
	Andrew

[PATCH net-next v4 3/3] udp6: Use Segment Routing Header for dest address if present

[Andrew Lunn]

When finding the socket to report an error on, if the invoking packet
is using Segment Routing, the IPv6 destination address is that of an
intermediate router, not the end destination. Extract the ultimate
destination address from the segment address.

This change allows traceroute to function in the presence of Segment
Routing.

Signed-off-by: Andrew Lunn <andrew@lunn.ch>
---
 include/net/seg6.h | 19 +++++++++++++++++++
 net/ipv6/udp.c     |  3 ++-
 2 files changed, 21 insertions(+), 1 deletion(-)

diff --git a/include/net/seg6.h b/include/net/seg6.h
index 02b0cd305787..af668f17b398 100644
--- a/include/net/seg6.h
+++ b/include/net/seg6.h
@@ -65,4 +65,23 @@ extern int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
 extern int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh);
 extern int seg6_lookup_nexthop(struct sk_buff *skb, struct in6_addr *nhaddr,
 			       u32 tbl_id);
+
+/* If the packet which invoked an ICMP error contains an SRH return
+ * the true destination address from within the SRH, otherwise use the
+ * destination address in the IP header.
+ */
+static inline const struct in6_addr *seg6_get_daddr(struct sk_buff *skb,
+						    struct inet6_skb_parm *opt)
+{
+	struct ipv6_sr_hdr *srh;
+
+	if (opt->flags & IP6SKB_SEG6) {
+		srh = (struct ipv6_sr_hdr *)(skb->data + opt->srhoff);
+		return  &srh->segments[0];
+	}
+
+	return NULL;
+}
+
+
 #endif
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 6a0e569f0bb8..a14375e3f923 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -40,6 +40,7 @@
 #include <net/transp_v6.h>
 #include <net/ip6_route.h>
 #include <net/raw.h>
+#include <net/seg6.h>
 #include <net/tcp_states.h>
 #include <net/ip6_checksum.h>
 #include <net/ip6_tunnel.h>
@@ -561,7 +562,7 @@ int __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 	struct ipv6_pinfo *np;
 	const struct ipv6hdr *hdr = (const struct ipv6hdr *)skb->data;
 	const struct in6_addr *saddr = &hdr->saddr;
-	const struct in6_addr *daddr = &hdr->daddr;
+	const struct in6_addr *daddr = seg6_get_daddr(skb, opt) ? : &hdr->daddr;
 	struct udphdr *uh = (struct udphdr *)(skb->data+offset);
 	bool tunnel = false;
 	struct sock *sk;
-- 
2.33.1

Re: [PATCH net-next v4 3/3] udp6: Use Segment Routing Header for dest address if present

[Willem de Bruijn]

On Wed, Dec 8, 2021 at 12:38 PM Andrew Lunn <andrew@lunn.ch> wrote:
>
> When finding the socket to report an error on, if the invoking packet
> is using Segment Routing, the IPv6 destination address is that of an
> intermediate router, not the end destination. Extract the ultimate
> destination address from the segment address.
>
> This change allows traceroute to function in the presence of Segment
> Routing.
>
> Signed-off-by: Andrew Lunn <andrew@lunn.ch>
> ---
>  include/net/seg6.h | 19 +++++++++++++++++++
>  net/ipv6/udp.c     |  3 ++-
>  2 files changed, 21 insertions(+), 1 deletion(-)
>
> diff --git a/include/net/seg6.h b/include/net/seg6.h
> index 02b0cd305787..af668f17b398 100644
> --- a/include/net/seg6.h
> +++ b/include/net/seg6.h
> @@ -65,4 +65,23 @@ extern int seg6_do_srh_encap(struct sk_buff *skb, struct ipv6_sr_hdr *osrh,
>  extern int seg6_do_srh_inline(struct sk_buff *skb, struct ipv6_sr_hdr *osrh);
>  extern int seg6_lookup_nexthop(struct sk_buff *skb, struct in6_addr *nhaddr,
>                                u32 tbl_id);
> +
> +/* If the packet which invoked an ICMP error contains an SRH return
> + * the true destination address from within the SRH, otherwise use the
> + * destination address in the IP header.
> + */

nit: the above describes the behavior of the caller, not the function.
The function returns NULL if no SRH is found.

> +static inline const struct in6_addr *seg6_get_daddr(struct sk_buff *skb,
> +                                                   struct inet6_skb_parm *opt)
> +{
> +       struct ipv6_sr_hdr *srh;
> +
> +       if (opt->flags & IP6SKB_SEG6) {
> +               srh = (struct ipv6_sr_hdr *)(skb->data + opt->srhoff);
> +               return  &srh->segments[0];
> +       }
> +
> +       return NULL;
> +}
> +
> +
>  #endif
> diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> index 6a0e569f0bb8..a14375e3f923 100644
> --- a/net/ipv6/udp.c
> +++ b/net/ipv6/udp.c
> @@ -40,6 +40,7 @@
>  #include <net/transp_v6.h>
>  #include <net/ip6_route.h>
>  #include <net/raw.h>
> +#include <net/seg6.h>
>  #include <net/tcp_states.h>
>  #include <net/ip6_checksum.h>
>  #include <net/ip6_tunnel.h>
> @@ -561,7 +562,7 @@ int __udp6_lib_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
>         struct ipv6_pinfo *np;
>         const struct ipv6hdr *hdr = (const struct ipv6hdr *)skb->data;
>         const struct in6_addr *saddr = &hdr->saddr;
> -       const struct in6_addr *daddr = &hdr->daddr;
> +       const struct in6_addr *daddr = seg6_get_daddr(skb, opt) ? : &hdr->daddr;
>         struct udphdr *uh = (struct udphdr *)(skb->data+offset);
>         bool tunnel = false;
>         struct sock *sk;
> --
> 2.33.1
>

Re: [PATCH net-next v4 3/3] udp6: Use Segment Routing Header for dest address if present

[Andrew Lunn]

> > +/* If the packet which invoked an ICMP error contains an SRH return
> > + * the true destination address from within the SRH, otherwise use the
> > + * destination address in the IP header.
> > + */
> 
> nit: the above describes the behavior of the caller, not the function.
> The function returns NULL if no SRH is found.

Yes, i will fix that.

Thanks
	Andrew