[PATCH] RP filter support for IPv6, kernel 2.6.15

From: Denis Semmau
Date: Fri Jan 13 2006 - 13:33:36 EST


I've made a patch in order to enable reverse path filtering for IPv6.
The rp_filter sysctl has been added, and some structures have changed in order to support this new feature.
The main part has been added into net/ipv6/ip6_output.c, with the same name of functions than for IPv4 (the new function is called rt6_validate_source).


Signed-off-by: Denis Semmau <denis.semmau@xxxxxxxxxx>


diff -Naur linux-2.6.15.old/include/linux/ipv6.h linux-2.6.15.new/include/linux/ipv6.h
--- linux-2.6.15.old/include/linux/ipv6.h 2006-01-03 04:21:10.000000000 +0100
+++ linux-2.6.15.new/include/linux/ipv6.h 2006-01-11 20:12:03.000000000 +0100
@@ -127,6 +127,7 @@
*/
struct ipv6_devconf {
__s32 forwarding;
+ __s32 rp_filter;
__s32 hop_limit;
__s32 mtu6;
__s32 accept_ra;
@@ -151,6 +152,7 @@
/* index values for the variables in ipv6_devconf */
enum {
DEVCONF_FORWARDING = 0,
+ DEVCONF_RPFILTER = 0,
DEVCONF_HOPLIMIT,
DEVCONF_MTU6,
DEVCONF_ACCEPT_RA,
diff -Naur linux-2.6.15.old/include/linux/sysctl.h linux-2.6.15.new/include/linux/sysctl.h
--- linux-2.6.15.old/include/linux/sysctl.h 2006-01-03 04:21:10.000000000 +0100
+++ linux-2.6.15.new/include/linux/sysctl.h 2006-01-11 20:11:51.000000000 +0100
@@ -524,6 +524,7 @@
NET_IPV6_MAX_DESYNC_FACTOR=15,
NET_IPV6_MAX_ADDRESSES=16,
NET_IPV6_FORCE_MLD_VERSION=17,
+ NET_IPV6_RPFILTER=18,
__NET_IPV6_MAX
};

diff -Naur linux-2.6.15.old/net/ipv6/addrconf.c linux-2.6.15.new/net/ipv6/addrconf.c
--- linux-2.6.15.old/net/ipv6/addrconf.c 2006-01-03 04:21:10.000000000 +0100
+++ linux-2.6.15.new/net/ipv6/addrconf.c 2006-01-11 20:12:30.000000000 +0100
@@ -150,6 +150,7 @@

struct ipv6_devconf ipv6_devconf = {
.forwarding = 0,
+ .rp_filter = 0,
.hop_limit = IPV6_DEFAULT_HOPLIMIT,
.mtu6 = IPV6_MIN_MTU,
.accept_ra = 1,
@@ -172,6 +173,7 @@

static struct ipv6_devconf ipv6_devconf_dflt = {
.forwarding = 0,
+ .rp_filter = 0,
.hop_limit = IPV6_DEFAULT_HOPLIMIT,
.mtu6 = IPV6_MIN_MTU,
.accept_ra = 1,
@@ -3111,6 +3113,7 @@
{
memset(array, 0, bytes);
array[DEVCONF_FORWARDING] = cnf->forwarding;
+ array[DEVCONF_RPFILTER] = cnf->rp_filter;
array[DEVCONF_HOPLIMIT] = cnf->hop_limit;
array[DEVCONF_MTU6] = cnf->mtu6;
array[DEVCONF_ACCEPT_RA] = cnf->accept_ra;
@@ -3586,6 +3589,14 @@
.proc_handler = &proc_dointvec,
},
{
+ .ctl_name = NET_IPV6_RPFILTER,
+ .procname = "rp_filter",
+ .data = &ipv6_devconf.rp_filter,
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .proc_handler = &proc_dointvec,
+ },
+ {
.ctl_name = 0, /* sentinel */
}
},
diff -Naur linux-2.6.15.old/net/ipv6/ip6_output.c linux-2.6.15.new/net/ipv6/ip6_output.c
--- linux-2.6.15.old/net/ipv6/ip6_output.c 2006-01-03 04:21:10.000000000 +0100
+++ linux-2.6.15.new/net/ipv6/ip6_output.c 2006-01-13 01:24:22.000000000 +0100
@@ -306,11 +306,24 @@
return 0;
}

+static int rt6_validate_source( struct sk_buff *skb) {
+ struct rt6_info *rt;
+ rt=rt6_lookup(&skb->nh.ipv6h->saddr,NULL,0,0);
+ if ( rt!=NULL ) {
+ if (rt->rt6i_idev->dev == skb->dev )
+ return 0;
+ }
+ return -1;
+}
+
+
static inline int ip6_forward_finish(struct sk_buff *skb)
{
return dst_output(skb);
}

+
+
int ip6_forward(struct sk_buff *skb)
{
struct dst_entry *dst = skb->dst;
@@ -320,6 +333,26 @@
if (ipv6_devconf.forwarding == 0)
goto error;

+ /*RP_FILTER*/
+ struct inet6_dev *idev = NULL;
+ idev = in6_dev_get(skb->dev);
+ if (!idev) {
+ printk(KERN_WARNING "idev error for RP_Filter\n");
+ goto error;
+ }
+
+ if (ipv6_devconf.rp_filter & idev->cnf.rp_filter ) {
+ if (rt6_validate_source(skb)<0) {
+ printk(KERN_WARNING "RP_FILTER-- Packet refused from %x:%x:%x:%x:%x:%x:%x:%x to %x:%x:%x:%x:%x:%x:%x:%x from %s\n",NIP6(skb->nh.ipv6h->saddr),NIP6(skb->nh.ipv6h->daddr),skb->dev->name);
+ goto error;
+ }
+
+ }
+
+ /*RP_FILTER END*/
+
+
+
if (!xfrm6_policy_check(NULL, XFRM_POLICY_FWD, skb)) {
IP6_INC_STATS(IPSTATS_MIB_INDISCARDS);
goto drop;

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/