[NETFILTER]: xt_iprange match, revision 1
authorJan Engelhardt <jengelh@computergmbh.de>
Tue, 15 Jan 2008 07:43:03 +0000 (23:43 -0800)
committerDavid S. Miller <davem@davemloft.net>
Mon, 28 Jan 2008 23:02:28 +0000 (15:02 -0800)
Adds IPv6 support to xt_iprange, making it possible to match on IPv6
address ranges with ip6tables.

Signed-off-by: Jan Engelhardt <jengelh@computergmbh.de>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/netfilter/xt_iprange.c

index c57a6cf..dbea0e0 100644 (file)
@@ -2,6 +2,7 @@
  *     xt_iprange - Netfilter module to match IP address ranges
  *
  *     (C) 2003 Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
+ *     (C) CC Computer Consultants GmbH, 2008
  *
  *     This program is free software; you can redistribute it and/or modify
  *     it under the terms of the GNU General Public License version 2 as
@@ -10,6 +11,7 @@
 #include <linux/module.h>
 #include <linux/skbuff.h>
 #include <linux/ip.h>
+#include <linux/ipv6.h>
 #include <linux/netfilter/x_tables.h>
 #include <linux/netfilter_ipv4/ipt_iprange.h>
 
@@ -51,26 +53,128 @@ iprange_mt_v0(const struct sk_buff *skb, const struct net_device *in,
        return true;
 }
 
-static struct xt_match iprange_mt_reg __read_mostly = {
-       .name           = "iprange",
-       .family         = AF_INET,
-       .match          = iprange_mt_v0,
-       .matchsize      = sizeof(struct ipt_iprange_info),
-       .me             = THIS_MODULE
+static bool
+iprange_mt4(const struct sk_buff *skb, const struct net_device *in,
+            const struct net_device *out, const struct xt_match *match,
+            const void *matchinfo, int offset, unsigned int protoff,
+            bool *hotdrop)
+{
+       const struct xt_iprange_mtinfo *info = matchinfo;
+       const struct iphdr *iph = ip_hdr(skb);
+       bool m;
+
+       if (info->flags & IPRANGE_SRC) {
+               m  = ntohl(iph->saddr) < ntohl(info->src_min.ip);
+               m |= ntohl(iph->saddr) > ntohl(info->src_max.ip);
+               m ^= info->flags & IPRANGE_SRC_INV;
+               if (m) {
+                       pr_debug("src IP " NIPQUAD_FMT " NOT in range %s"
+                                NIPQUAD_FMT "-" NIPQUAD_FMT "\n",
+                                NIPQUAD(iph->saddr),
+                                (info->flags & IPRANGE_SRC_INV) ? "(INV) " : "",
+                                NIPQUAD(info->src_max.ip),
+                                NIPQUAD(info->src_max.ip));
+                       return false;
+               }
+       }
+       if (info->flags & IPRANGE_DST) {
+               m  = ntohl(iph->daddr) < ntohl(info->dst_min.ip);
+               m |= ntohl(iph->daddr) > ntohl(info->dst_max.ip);
+               m ^= info->flags & IPRANGE_DST_INV;
+               if (m) {
+                       pr_debug("dst IP " NIPQUAD_FMT " NOT in range %s"
+                                NIPQUAD_FMT "-" NIPQUAD_FMT "\n",
+                                NIPQUAD(iph->daddr),
+                                (info->flags & IPRANGE_DST_INV) ? "(INV) " : "",
+                                NIPQUAD(info->dst_min.ip),
+                                NIPQUAD(info->dst_max.ip));
+                       return false;
+               }
+       }
+       return true;
+}
+
+static inline int
+iprange_ipv6_sub(const struct in6_addr *a, const struct in6_addr *b)
+{
+       unsigned int i;
+       int r;
+
+       for (i = 0; i < 4; ++i) {
+               r = a->s6_addr32[i] - b->s6_addr32[i];
+               if (r != 0)
+                       return r;
+       }
+
+       return 0;
+}
+
+static bool
+iprange_mt6(const struct sk_buff *skb, const struct net_device *in,
+            const struct net_device *out, const struct xt_match *match,
+            const void *matchinfo, int offset, unsigned int protoff,
+            bool *hotdrop)
+{
+       const struct xt_iprange_mtinfo *info = matchinfo;
+       const struct ipv6hdr *iph = ipv6_hdr(skb);
+       bool m;
+
+       if (info->flags & IPRANGE_SRC) {
+               m  = iprange_ipv6_sub(&iph->saddr, &info->src_min.in6) < 0;
+               m |= iprange_ipv6_sub(&iph->saddr, &info->src_max.in6) > 0;
+               m ^= info->flags & IPRANGE_SRC_INV;
+               if (m)
+                       return false;
+       }
+       if (info->flags & IPRANGE_DST) {
+               m  = iprange_ipv6_sub(&iph->daddr, &info->dst_min.in6) < 0;
+               m |= iprange_ipv6_sub(&iph->daddr, &info->dst_max.in6) > 0;
+               m ^= info->flags & IPRANGE_DST_INV;
+               if (m)
+                       return false;
+       }
+       return true;
+}
+
+static struct xt_match iprange_mt_reg[] __read_mostly = {
+       {
+               .name      = "iprange",
+               .revision  = 0,
+               .family    = AF_INET,
+               .match     = iprange_mt_v0,
+               .matchsize = sizeof(struct ipt_iprange_info),
+               .me        = THIS_MODULE,
+       },
+       {
+               .name      = "iprange",
+               .revision  = 1,
+               .family    = AF_INET6,
+               .match     = iprange_mt4,
+               .matchsize = sizeof(struct xt_iprange_mtinfo),
+               .me        = THIS_MODULE,
+       },
+       {
+               .name      = "iprange",
+               .revision  = 1,
+               .family    = AF_INET6,
+               .match     = iprange_mt6,
+               .matchsize = sizeof(struct xt_iprange_mtinfo),
+               .me        = THIS_MODULE,
+       },
 };
 
 static int __init iprange_mt_init(void)
 {
-       return xt_register_match(&iprange_mt_reg);
+       return xt_register_matches(iprange_mt_reg, ARRAY_SIZE(iprange_mt_reg));
 }
 
 static void __exit iprange_mt_exit(void)
 {
-       xt_unregister_match(&iprange_mt_reg);
+       xt_unregister_matches(iprange_mt_reg, ARRAY_SIZE(iprange_mt_reg));
 }
 
 module_init(iprange_mt_init);
 module_exit(iprange_mt_exit);
 MODULE_LICENSE("GPL");
-MODULE_AUTHOR("Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>");
+MODULE_AUTHOR("Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>, Jan Engelhardt <jengelh@computergmbh.de>");
 MODULE_DESCRIPTION("Xtables: arbitrary IPv4 range matching");