西厢1客户端

xt_ZHANG.c

/*
 *      "ZHANG" target extension for Xtables
 *      copied from "DELUDE" target
 *      Copyright © Jan Engelhardt <jengelh [at] medozas de>, 2007 - 2008
 *      Klzgrad <klzgrad@gmail.com>, 2010
 *
 *      Based upon linux-2.6.18.5/net/ipv4/netfilter/ipt_REJECT.c:
 *      (C) 1999-2001 Paul `Rusty' Russell
 *      (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
 *
 *      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
 *      published by the Free Software Foundation.
 */
#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/ip.h>
#include <linux/tcp.h>
#include <linux/netfilter/x_tables.h>
#ifdef CONFIG_BRIDGE_NETFILTER
#       include <linux/netfilter_bridge.h>
#endif
#include <net/tcp.h>
#include "compat_xtables.h"
#define PFX KBUILD_MODNAME ": "

static void zhang_send_reset(struct sk_buff *oldskb, unsigned int hook)
{
        struct tcphdr _otcph, *tcph, *tcph2;
        const struct tcphdr *oth;
        const struct iphdr *oiph;
        struct sk_buff *skb, *skb2;
        struct iphdr *iph;
        unsigned int addr_type;

        oiph = ip_hdr(oldskb);

        /* do not deal with fragment */
        if (oiph->frag_off & htons(IP_OFFSET))
                return;

        oth = skb_header_pointer(oldskb, ip_hdrlen(oldskb),
                                 sizeof(_otcph), &_otcph);
        if (oth == NULL)
                return;

        /* syn[ack] only */
        if (!oth->syn || !oth->ack || oth->rst || oth->fin)
                return;

        /* check checksum */
        if (nf_ip_checksum(oldskb, hook, ip_hdrlen(oldskb), IPPROTO_TCP))
                return;

        skb = alloc_skb(sizeof(struct iphdr) + sizeof(struct tcphdr) +
                         LL_MAX_HEADER, GFP_ATOMIC);
        if (skb == NULL)
                return;

        /* construct from scratch */
        skb_reserve(skb, LL_MAX_HEADER);
        skb_reset_network_header(skb);
        iph = (struct iphdr *)skb_put(skb, sizeof(struct iphdr));
        iph->version  = 4;
        iph->ihl      = sizeof(struct iphdr) / 4;
        iph->tos      = 0;
        iph->id       = 0;
        iph->frag_off = htons(IP_DF);
        iph->protocol = IPPROTO_TCP;
        iph->check    = 0;
        iph->saddr    = oiph->daddr;
        iph->daddr    = oiph->saddr;

        tcph = (struct tcphdr *)skb_put(skb, sizeof(struct tcphdr));
        memset(tcph, 0, sizeof(*tcph));
        tcph->source = oth->dest;
        tcph->dest   = oth->source;
        tcph->doff   = sizeof(struct tcphdr) / 4;
        tcph->window = 0xffffU;

        /*
         * essential part 1
         * inject an FIN with bad sequence number, obfuscating the handshake.
         * it will be dropped by rfc-compliant endpoint, 
         * meanwhile thwarting eavesdroppers on the same direction (c -> s).
         */
        tcph->fin     = true;
        tcph->seq     = oth->ack_seq;
        tcph->ack_seq = oth->seq;

#if LINUX_VERSION_CODE <= KERNEL_VERSION(2, 6, 20)
        tcph->check = tcp_v4_check(tcph, sizeof(struct tcphdr), iph->saddr,
                      iph->daddr, csum_partial((char *)tcph,
                      sizeof(struct tcphdr), 0));
#else
        tcph->check = tcp_v4_check(sizeof(struct tcphdr), iph->saddr,
                      iph->daddr, csum_partial((char *)tcph,
                      sizeof(struct tcphdr), 0));
#endif

        addr_type = RTN_UNSPEC;
#ifdef CONFIG_BRIDGE_NETFILTER
        if (hook != NF_INET_FORWARD || (skb->nf_bridge != NULL &&
            skb->nf_bridge->mask & BRNF_BRIDGED))
#else
        if (hook != NF_INET_FORWARD)
#endif
                addr_type = RTN_LOCAL;

        /* ip_route_me_harder expects skb->dst to be set */
        skb_dst_set(skb, dst_clone(skb_dst(oldskb)));

        if (ip_route_me_harder(&skb, addr_type))
                goto free_skb;
        else
                iph = ip_hdr(skb);

        iph->ttl       = dst_metric(skb_dst(skb), RTAX_HOPLIMIT);
        skb->ip_summed = CHECKSUM_NONE;

        /* "Never happens" */
        if (skb->len > dst_mtu(skb_dst(skb)))
                goto free_skb;

        skb2 = skb_copy(skb, GFP_ATOMIC);
        if (skb2 == NULL)
                goto free_skb;

        /*
         * essential part 2
         * inject an ACK with correct SEQ but bad ACK.
         * this causes an RST from server which should have no real impact on 
         * the original connection,
         * thus thwarts eavesdroppers on the other direction (s -> c).
         * 
         * RFC793: 
         *   2.  If the connection is in any non-synchronized state (LISTEN,
         *   SYN-SENT, SYN-RECEIVED), and the incoming segment acknowledges
         *   something not yet sent (the segment carries an unacceptable ACK),
         *   ..., a reset is sent.
         * 
         *   If the incoming segment has an ACK field, the reset takes its
         *   sequence number from the ACK field of the segment, otherwise the
         *   reset has sequence number zero and the ACK field is set to the sum
         *   of the sequence number and segment length of the incoming segment.
         *   The connection remains in the same state.
         * 
         * sometimes certain kind of rfc non-compliant tcp stacks or firewalls
         * may have unexpected response or no reply at all.
         *
         * seems that the seq is not nessesarily correct.
         */
        tcph2 = (struct tcphdr *)(skb_network_header(skb2) + ip_hdrlen(skb2));
        tcph2->fin     = false;
        tcph2->ack     = true;
        //tcph2->seq     = oth->ack_seq;
        //tcph2->ack_seq = oth->seq;

        /* checksum shortcut */
        //csum_replace4(&tcph2->check, tcph->seq, tcph2->seq);
        //csum_replace4(&tcph2->check, tcph->ack_seq, tcph2->ack_seq);
        csum_replace2(&tcph2->check, htons(((u_int8_t *)tcph)[13]), 
                        htons(((u_int8_t *)tcph2)[13]));

        ip_local_out(skb);
        ip_local_out(skb2);

        return;
free_skb:
        kfree_skb(skb);
}

static unsigned int
zhang_tg(struct sk_buff **pskb, const struct xt_action_param *par)
{
        /* WARNING: This code causes reentry within iptables.
           This means that the iptables jump stack is now crap.  We
           must return an absolute verdict. --RR */
        zhang_send_reset(*pskb, par->hooknum);
        return NF_ACCEPT;
}

static struct xt_target zhang_tg_reg __read_mostly = {
        .name     = "ZHANG",
        .revision = 0,
        .family   = NFPROTO_IPV4,
        .table    = "filter",
        .hooks    = (1 << NF_INET_LOCAL_IN),
        .proto    = IPPROTO_TCP,
        .target   = zhang_tg,
        .me       = THIS_MODULE,
};

static int __init zhang_tg_init(void)
{
        return xt_register_target(&zhang_tg_reg);
}

static void __exit zhang_tg_exit(void)
{
        xt_unregister_target(&zhang_tg_reg);
}

module_init(zhang_tg_init);
module_exit(zhang_tg_exit);
MODULE_DESCRIPTION("Xtables: Scholar Zhang");
MODULE_AUTHOR("Jan Engelhardt <jengelh@medozas.de>");
MODULE_AUTHOR("Klzgrad <klzgrad@gmail.com>");
MODULE_LICENSE("GPL");
MODULE_ALIAS("ipt_ZHANG");