net: ipv6: sysctl to specify IPv6 ND traffic class

Add a per-device sysctl to specify the default traffic class to use for
kernel originated IPv6 Neighbour Discovery packets.

Currently this includes:

  - Router Solicitation (ICMPv6 type 133)
    ndisc_send_rs() -> ndisc_send_skb() -> ip6_nd_hdr()

  - Neighbour Solicitation (ICMPv6 type 135)
    ndisc_send_ns() -> ndisc_send_skb() -> ip6_nd_hdr()

  - Neighbour Advertisement (ICMPv6 type 136)
    ndisc_send_na() -> ndisc_send_skb() -> ip6_nd_hdr()

  - Redirect (ICMPv6 type 137)
    ndisc_send_redirect() -> ndisc_send_skb() -> ip6_nd_hdr()

and if the kernel ever gets around to generating RA's,
it would presumably also include:

  - Router Advertisement (ICMPv6 type 134)
    (radvd daemon could pick up on the kernel setting and use it)

Interface drivers may examine the Traffic Class value and translate
the DiffServ Code Point into a link-layer appropriate traffic
prioritization scheme.  An example of mapping IETF DSCP values to
IEEE 802.11 User Priority values can be found here:

    https://tools.ietf.org/html/draft-ietf-tsvwg-ieee-802-11

The expected primary use case is to properly prioritize ND over wifi.

Testing:
  jzem22:~# cat /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  0
  jzem22:~# echo -1 > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  -bash: echo: write error: Invalid argument
  jzem22:~# echo 256 > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  -bash: echo: write error: Invalid argument
  jzem22:~# echo 0 > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  jzem22:~# echo 255 > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  jzem22:~# cat /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  255
  jzem22:~# echo 34 > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  jzem22:~# cat /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  34

  jzem22:~# echo $[0xDC] > /proc/sys/net/ipv6/conf/eth0/ndisc_tclass
  jzem22:~# tcpdump -v -i eth0 icmp6 and src host jzem22.pgc and dst host fe80::1
  tcpdump: listening on eth0, link-type EN10MB (Ethernet), capture size 262144 bytes
  IP6 (class 0xdc, hlim 255, next-header ICMPv6 (58) payload length: 24)
  jzem22.pgc > fe80::1: [icmp6 sum ok] ICMP6, neighbor advertisement,
  length 24, tgt is jzem22.pgc, Flags [solicited]

(based on original change written by Erik Kline, with minor changes)

v2: fix 'suspicious rcu_dereference_check() usage'
    by explicitly grabbing the rcu_read_lock.

Cc: Lorenzo Colitti <lorenzo@google.com>
Signed-off-by: Erik Kline <ek@google.com>
Signed-off-by: Maciej Żenczykowski <maze@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 30 insertions, 1 deletions

diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index 54410a1d4065..d8676dda7fa6 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -1732,6 +1732,15 @@ ndisc_notify - BOOLEAN
 	1 - Generate unsolicited neighbour advertisements when device is brought
 	    up or hardware address changes.
 
+ndisc_tclass - INTEGER
+	The IPv6 Traffic Class to use by default when sending IPv6 Neighbor
+	Discovery (Router Solicitation, Router Advertisement, Neighbor
+	Solicitation, Neighbor Advertisement, Redirect) messages.
+	These 8 bits can be interpreted as 6 high order bits holding the DSCP
+	value and 2 low order bits representing ECN (which you probably want
+	to leave cleared).
+	0 - (default)
+
 mldv1_unsolicited_report_interval - INTEGER
 	The interval in milliseconds in which the next unsolicited
 	MLDv1 report retransmit will take place.
diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h
index ea04ca024f0d..cb18c6290ca8 100644
--- a/include/linux/ipv6.h
+++ b/include/linux/ipv6.h
@@ -73,6 +73,7 @@ struct ipv6_devconf {
 	__u32		enhanced_dad;
 	__u32		addr_gen_mode;
 	__s32		disable_policy;
+	__s32           ndisc_tclass;
 
 	struct ctl_table_header *sysctl_header;
 };
diff --git a/include/uapi/linux/ipv6.h b/include/uapi/linux/ipv6.h
index b22a9c4e1b12..9c0f4a92bcff 100644
--- a/include/uapi/linux/ipv6.h
+++ b/include/uapi/linux/ipv6.h
@@ -186,6 +186,7 @@ enum {
 	DEVCONF_ADDR_GEN_MODE,
 	DEVCONF_DISABLE_POLICY,
 	DEVCONF_ACCEPT_RA_RT_INFO_MIN_PLEN,
+	DEVCONF_NDISC_TCLASS,
 	DEVCONF_MAX
 };
 
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 6233e06fa35c..a6dffd65eb9d 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -5059,6 +5059,7 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
 	array[DEVCONF_ENHANCED_DAD] = cnf->enhanced_dad;
 	array[DEVCONF_ADDR_GEN_MODE] = cnf->addr_gen_mode;
 	array[DEVCONF_DISABLE_POLICY] = cnf->disable_policy;
+	array[DEVCONF_NDISC_TCLASS] = cnf->ndisc_tclass;
 }
 
 static inline size_t inet6_ifla6_size(void)
@@ -5986,6 +5987,7 @@ int addrconf_sysctl_disable_policy(struct ctl_table *ctl, int write,
 }
 
 static int minus_one = -1;
+static const int zero = 0;
 static const int one = 1;
 static const int two_five_five = 255;
 
@@ -6357,6 +6359,15 @@ static const struct ctl_table addrconf_sysctl[] = {
 		.proc_handler   = addrconf_sysctl_disable_policy,
 	},
 	{
+		.procname	= "ndisc_tclass",
+		.data		= &ipv6_devconf.ndisc_tclass,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= (void *)&zero,
+		.extra2		= (void *)&two_five_five,
+	},
+	{
 		/* sentinel */
 	}
 };
diff --git a/net/ipv6/ndisc.c b/net/ipv6/ndisc.c
index f9c3ffe04382..b3cea200c85e 100644
--- a/net/ipv6/ndisc.c
+++ b/net/ipv6/ndisc.c
@@ -427,12 +427,19 @@ static void ip6_nd_hdr(struct sk_buff *skb,
 		       int hop_limit, int len)
 {
 	struct ipv6hdr *hdr;
+	struct inet6_dev *idev;
+	unsigned tclass;
+
+	rcu_read_lock();
+	idev = __in6_dev_get(skb->dev);
+	tclass = idev ? idev->cnf.ndisc_tclass : 0;
+	rcu_read_unlock();
 
 	skb_push(skb, sizeof(*hdr));
 	skb_reset_network_header(skb);
 	hdr = ipv6_hdr(skb);
 
-	ip6_flow_hdr(hdr, 0, 0);
+	ip6_flow_hdr(hdr, tclass, 0);
 
 	hdr->payload_len = htons(len);
 	hdr->nexthdr = IPPROTO_ICMPV6;