IPv6 enhancements.
diff --git a/libiptc/libip4tc.c b/libiptc/libip4tc.c
new file mode 100644
index 0000000..10ea1be
--- /dev/null
+++ b/libiptc/libip4tc.c
@@ -0,0 +1,444 @@
+/* Library which manipulates firewall rules. Version 0.1. */
+
+/* Architecture of firewall rules is as follows:
+ *
+ * Chains go INPUT, FORWARD, OUTPUT then user chains.
+ * Each user chain starts with an ERROR node.
+ * Every chain ends with an unconditional jump: a RETURN for user chains,
+ * and a POLICY for built-ins.
+ */
+
+/* (C)1999 Paul ``Rusty'' Russell - Placed under the GNU GPL (See
+ COPYING for details). */
+
+#include <assert.h>
+#include <string.h>
+#include <errno.h>
+#include <stdlib.h>
+#include <stdio.h>
+
+#ifdef DEBUG_CONNTRACK
+#define inline
+#endif
+
+#if !defined(__GLIBC__) || (__GLIBC__ < 2)
+typedef unsigned int socklen_t;
+#endif
+
+#include "libiptc/libiptc.h"
+
+#define IP_VERSION 4
+#define IP_OFFSET 0x1FFF
+
+#define HOOK_PRE_ROUTING NF_IP_PRE_ROUTING
+#define HOOK_LOCAL_IN NF_IP_LOCAL_IN
+#define HOOK_FORWARD NF_IP_FORWARD
+#define HOOK_LOCAL_OUT NF_IP_LOCAL_OUT
+#define HOOK_POST_ROUTING NF_IP_POST_ROUTING
+
+#define STRUCT_ENTRY_TARGET struct ipt_entry_target
+#define STRUCT_ENTRY struct ipt_entry
+#define STRUCT_ENTRY_MATCH struct ipt_entry_match
+#define STRUCT_GETINFO struct ipt_getinfo
+#define STRUCT_GET_ENTRIES struct ipt_get_entries
+#define STRUCT_COUNTERS struct ipt_counters
+#define STRUCT_COUNTERS_INFO struct ipt_counters_info
+#define STRUCT_STANDARD_TARGET struct ipt_standard_target
+#define STRUCT_REPLACE struct ipt_replace
+
+#define STRUCT_TC_HANDLE struct iptc_handle
+#define TC_HANDLE_T iptc_handle_t
+
+#define ENTRY_ITERATE IPT_ENTRY_ITERATE
+#define TABLE_MAXNAMELEN IPT_TABLE_MAXNAMELEN
+#define FUNCTION_MAXNAMELEN IPT_FUNCTION_MAXNAMELEN
+
+#define GET_TARGET ipt_get_target
+
+#define ERROR_TARGET IPT_ERROR_TARGET
+#define NUMHOOKS NF_IP_NUMHOOKS
+
+#define IPT_CHAINLABEL ipt_chainlabel
+
+#define TC_DUMP_ENTRIES dump_entries
+#define TC_IS_CHAIN iptc_is_chain
+#define TC_NEXT_CHAIN iptc_next_chain
+#define TC_NUM_RULES iptc_num_rules
+#define TC_GET_RULE iptc_get_rule
+#define TC_GET_TARGET iptc_get_target
+#define TC_BUILTIN iptc_builtin
+#define TC_GET_POLICY iptc_get_policy
+#define TC_INSERT_ENTRY iptc_insert_entry
+#define TC_REPLACE_ENTRY iptc_replace_entry
+#define TC_APPEND_ENTRY iptc_append_entry
+#define TC_DELETE_ENTRY iptc_delete_entry
+#define TC_DELETE_NUM_ENTRY iptc_delete_num_entry
+#define TC_CHECK_PACKET iptc_check_packet
+#define TC_FLUSH_ENTRIES iptc_flush_entries
+#define TC_ZERO_ENTRIES iptc_zero_entries
+#define TC_CREATE_CHAIN iptc_create_chain
+#define TC_GET_REFERENCES iptc_get_references
+#define TC_DELETE_CHAIN iptc_delete_chain
+#define TC_RENAME_CHAIN iptc_rename_chain
+#define TC_SET_POLICY iptc_set_policy
+#define TC_GET_RAW_SOCKET iptc_get_raw_socket
+#define TC_INIT iptc_init
+#define TC_COMMIT iptc_commit
+#define TC_STRERROR iptc_strerror
+
+#define TC_AF AF_INET
+#define TC_IPPROTO IPPROTO_IP
+
+#define SO_SET_REPLACE IPT_SO_SET_REPLACE
+#define SO_SET_ADD_COUNTERS IPT_SO_SET_ADD_COUNTERS
+#define SO_GET_INFO IPT_SO_GET_INFO
+#define SO_GET_ENTRIES IPT_SO_GET_ENTRIES
+#define SO_GET_VERSION IPT_SO_GET_VERSION
+
+#define STANDARD_TARGET IPT_STANDARD_TARGET
+#define LABEL_RETURN IPTC_LABEL_RETURN
+#define LABEL_ACCEPT IPTC_LABEL_ACCEPT
+#define LABEL_DROP IPTC_LABEL_DROP
+
+#define ALIGN IPT_ALIGN
+#define RETURN IPT_RETURN
+
+#include "libiptc.c"
+
+#define IP_PARTS_NATIVE(n) \
+(unsigned int)((n)>>24)&0xFF, \
+(unsigned int)((n)>>16)&0xFF, \
+(unsigned int)((n)>>8)&0xFF, \
+(unsigned int)((n)&0xFF)
+
+#define IP_PARTS(n) IP_PARTS_NATIVE(ntohl(n))
+
+int
+dump_entry(STRUCT_ENTRY *e, const TC_HANDLE_T handle)
+{
+ size_t i;
+ STRUCT_ENTRY_TARGET *t;
+
+ printf("Entry %u (%lu):\n", entry2index(handle, e),
+ entry2offset(handle, e));
+ printf("SRC IP: %u.%u.%u.%u/%u.%u.%u.%u\n",
+ IP_PARTS(e->ip.src.s_addr),IP_PARTS(e->ip.smsk.s_addr));
+ printf("DST IP: %u.%u.%u.%u/%u.%u.%u.%u\n",
+ IP_PARTS(e->ip.dst.s_addr),IP_PARTS(e->ip.dmsk.s_addr));
+ printf("Interface: `%s'/", e->ip.iniface);
+ for (i = 0; i < IFNAMSIZ; i++)
+ printf("%c", e->ip.iniface_mask[i] ? 'X' : '.');
+ printf("to `%s'/", e->ip.outiface);
+ for (i = 0; i < IFNAMSIZ; i++)
+ printf("%c", e->ip.outiface_mask[i] ? 'X' : '.');
+ printf("\nProtocol: %u\n", e->ip.proto);
+ printf("Flags: %02X\n", e->ip.flags);
+ printf("Invflags: %02X\n", e->ip.invflags);
+ printf("Counters: %llu packets, %llu bytes\n",
+ e->counters.pcnt, e->counters.bcnt);
+ printf("Cache: %08X ", e->nfcache);
+ if (e->nfcache & NFC_ALTERED) printf("ALTERED ");
+ if (e->nfcache & NFC_UNKNOWN) printf("UNKNOWN ");
+ if (e->nfcache & NFC_IP_SRC) printf("IP_SRC ");
+ if (e->nfcache & NFC_IP_DST) printf("IP_DST ");
+ if (e->nfcache & NFC_IP_IF_IN) printf("IP_IF_IN ");
+ if (e->nfcache & NFC_IP_IF_OUT) printf("IP_IF_OUT ");
+ if (e->nfcache & NFC_IP_TOS) printf("IP_TOS ");
+ if (e->nfcache & NFC_IP_PROTO) printf("IP_PROTO ");
+ if (e->nfcache & NFC_IP_OPTIONS) printf("IP_OPTIONS ");
+ if (e->nfcache & NFC_IP_TCPFLAGS) printf("IP_TCPFLAGS ");
+ if (e->nfcache & NFC_IP_SRC_PT) printf("IP_SRC_PT ");
+ if (e->nfcache & NFC_IP_DST_PT) printf("IP_DST_PT ");
+ if (e->nfcache & NFC_IP_PROTO_UNKNOWN) printf("IP_PROTO_UNKNOWN ");
+ printf("\n");
+
+ IPT_MATCH_ITERATE(e, print_match);
+
+ t = GET_TARGET(e);
+ printf("Target name: `%s' [%u]\n", t->u.user.name, t->u.target_size);
+ if (strcmp(t->u.user.name, STANDARD_TARGET) == 0) {
+ int pos = *(int *)t->data;
+ if (pos < 0)
+ printf("verdict=%s\n",
+ pos == -NF_ACCEPT-1 ? "NF_ACCEPT"
+ : pos == -NF_DROP-1 ? "NF_DROP"
+ : pos == -NF_QUEUE-1 ? "NF_QUEUE"
+ : pos == RETURN ? "RETURN"
+ : "UNKNOWN");
+ else
+ printf("verdict=%u\n", pos);
+ } else if (strcmp(t->u.user.name, IPT_ERROR_TARGET) == 0)
+ printf("error=`%s'\n", t->data);
+
+ printf("\n");
+ return 0;
+}
+
+static int
+is_same(const STRUCT_ENTRY *a, const STRUCT_ENTRY *b, unsigned char *matchmask)
+{
+ unsigned int i;
+ STRUCT_ENTRY_TARGET *ta, *tb;
+ unsigned char *mptr;
+
+ /* Always compare head structures: ignore mask here. */
+ if (a->ip.src.s_addr != b->ip.src.s_addr
+ || a->ip.dst.s_addr != b->ip.dst.s_addr
+ || a->ip.smsk.s_addr != b->ip.smsk.s_addr
+ || a->ip.smsk.s_addr != b->ip.smsk.s_addr
+ || a->ip.proto != b->ip.proto
+ || a->ip.flags != b->ip.flags
+ || a->ip.invflags != b->ip.invflags)
+ return 0;
+
+ for (i = 0; i < IFNAMSIZ; i++) {
+ if (a->ip.iniface_mask[i] != b->ip.iniface_mask[i])
+ return 0;
+ if ((a->ip.iniface[i] & a->ip.iniface_mask[i])
+ != (b->ip.iniface[i] & b->ip.iniface_mask[i]))
+ return 0;
+ if (a->ip.outiface_mask[i] != b->ip.outiface_mask[i])
+ return 0;
+ if ((a->ip.outiface[i] & a->ip.outiface_mask[i])
+ != (b->ip.outiface[i] & b->ip.outiface_mask[i]))
+ return 0;
+ }
+
+ if (a->nfcache != b->nfcache
+ || a->target_offset != b->target_offset
+ || a->next_offset != b->next_offset)
+ return 0;
+
+ mptr = matchmask + sizeof(STRUCT_ENTRY);
+ if (IPT_MATCH_ITERATE(a, match_different, a->elems, b->elems, &mptr))
+ return 0;
+
+ ta = GET_TARGET((STRUCT_ENTRY *)a);
+ tb = GET_TARGET((STRUCT_ENTRY *)b);
+ if (ta->u.target_size != tb->u.target_size)
+ return 0;
+ if (strcmp(ta->u.user.name, tb->u.user.name) != 0)
+ return 0;
+
+ mptr += sizeof(*ta);
+ if (target_different(ta->data, tb->data,
+ ta->u.target_size - sizeof(*ta), mptr))
+ return 0;
+
+ return 1;
+}
+
+/***************************** DEBUGGING ********************************/
+static inline int
+unconditional(const struct ipt_ip *ip)
+{
+ unsigned int i;
+
+ for (i = 0; i < sizeof(*ip)/sizeof(u_int32_t); i++)
+ if (((u_int32_t *)ip)[i])
+ return 0;
+
+ return 1;
+}
+
+static inline int
+check_match(const STRUCT_ENTRY_MATCH *m, unsigned int *off)
+{
+ assert(m->u.match_size >= sizeof(STRUCT_ENTRY_MATCH));
+ assert(ALIGN(m->u.match_size) == m->u.match_size);
+
+ (*off) += m->u.match_size;
+ return 0;
+}
+
+static inline int
+check_entry(const STRUCT_ENTRY *e, unsigned int *i, unsigned int *off,
+ unsigned int user_offset, int *was_return,
+ TC_HANDLE_T h)
+{
+ unsigned int toff;
+ STRUCT_STANDARD_TARGET *t;
+
+ assert(e->target_offset >= sizeof(STRUCT_ENTRY));
+ assert(e->next_offset >= e->target_offset
+ + sizeof(STRUCT_ENTRY_TARGET));
+ toff = sizeof(STRUCT_ENTRY);
+ IPT_MATCH_ITERATE(e, check_match, &toff);
+
+ assert(toff == e->target_offset);
+
+ t = (STRUCT_STANDARD_TARGET *)
+ GET_TARGET((STRUCT_ENTRY *)e);
+ /* next_offset will have to be multiple of entry alignment. */
+ assert(e->next_offset == ALIGN(e->next_offset));
+ assert(e->target_offset == ALIGN(e->target_offset));
+ assert(t->target.u.target_size == ALIGN(t->target.u.target_size));
+ assert(!TC_IS_CHAIN(t->target.u.user.name, h));
+
+ if (strcmp(t->target.u.user.name, STANDARD_TARGET) == 0) {
+ assert(t->target.u.target_size
+ == ALIGN(sizeof(STRUCT_STANDARD_TARGET)));
+
+ assert(t->verdict == -NF_DROP-1
+ || t->verdict == -NF_ACCEPT-1
+ || t->verdict == RETURN
+ || t->verdict < (int)h->entries.size);
+
+ if (t->verdict >= 0) {
+ STRUCT_ENTRY *te = get_entry(h, t->verdict);
+ int idx;
+
+ idx = entry2index(h, te);
+ assert(strcmp(GET_TARGET(te)->u.user.name,
+ IPT_ERROR_TARGET)
+ != 0);
+ assert(te != e);
+
+ /* Prior node must be error node, or this node. */
+ assert(t->verdict == entry2offset(h, e)+e->next_offset
+ || strcmp(GET_TARGET(index2entry(h, idx-1))
+ ->u.user.name, IPT_ERROR_TARGET)
+ == 0);
+ }
+
+ if (t->verdict == RETURN
+ && unconditional(&e->ip)
+ && e->target_offset == sizeof(*e))
+ *was_return = 1;
+ else
+ *was_return = 0;
+ } else if (strcmp(t->target.u.user.name, IPT_ERROR_TARGET) == 0) {
+ assert(t->target.u.target_size
+ == ALIGN(sizeof(struct ipt_error_target)));
+
+ /* If this is in user area, previous must have been return */
+ if (*off > user_offset)
+ assert(*was_return);
+
+ *was_return = 0;
+ }
+ else *was_return = 0;
+
+ if (*off == user_offset)
+ assert(strcmp(t->target.u.user.name, IPT_ERROR_TARGET) == 0);
+
+ (*off) += e->next_offset;
+ (*i)++;
+ return 0;
+}
+
+#ifndef NDEBUG
+/* Do every conceivable sanity check on the handle */
+static void
+do_check(TC_HANDLE_T h, unsigned int line)
+{
+ unsigned int i, n;
+ unsigned int user_offset; /* Offset of first user chain */
+ int was_return;
+
+ assert(h->changed == 0 || h->changed == 1);
+ if (strcmp(h->info.name, "filter") == 0) {
+ assert(h->info.valid_hooks
+ == (1 << NF_IP_LOCAL_IN
+ | 1 << NF_IP_FORWARD
+ | 1 << NF_IP_LOCAL_OUT));
+
+ /* Hooks should be first three */
+ assert(h->info.hook_entry[NF_IP_LOCAL_IN] == 0);
+
+ n = get_chain_end(h, 0);
+ n += get_entry(h, n)->next_offset;
+ assert(h->info.hook_entry[NF_IP_FORWARD] == n);
+
+ n = get_chain_end(h, n);
+ n += get_entry(h, n)->next_offset;
+ assert(h->info.hook_entry[NF_IP_LOCAL_OUT] == n);
+
+ user_offset = h->info.hook_entry[NF_IP_LOCAL_OUT];
+ } else if (strcmp(h->info.name, "nat") == 0) {
+ assert(h->info.valid_hooks
+ == (1 << NF_IP_PRE_ROUTING
+ | 1 << NF_IP_POST_ROUTING
+ | 1 << NF_IP_LOCAL_OUT));
+
+ assert(h->info.hook_entry[NF_IP_PRE_ROUTING] == 0);
+
+ n = get_chain_end(h, 0);
+ n += get_entry(h, n)->next_offset;
+ assert(h->info.hook_entry[NF_IP_POST_ROUTING] == n);
+
+ n = get_chain_end(h, n);
+ n += get_entry(h, n)->next_offset;
+ assert(h->info.hook_entry[NF_IP_LOCAL_OUT] == n);
+
+ user_offset = h->info.hook_entry[NF_IP_LOCAL_OUT];
+ } else if (strcmp(h->info.name, "mangle") == 0) {
+ assert(h->info.valid_hooks
+ == (1 << NF_IP_PRE_ROUTING
+ | 1 << NF_IP_LOCAL_OUT));
+
+ /* Hooks should be first three */
+ assert(h->info.hook_entry[NF_IP_PRE_ROUTING] == 0);
+
+ n = get_chain_end(h, 0);
+ n += get_entry(h, n)->next_offset;
+ assert(h->info.hook_entry[NF_IP_LOCAL_OUT] == n);
+
+ user_offset = h->info.hook_entry[NF_IP_LOCAL_OUT];
+ } else
+ abort();
+
+ /* User chain == end of last builtin + policy entry */
+ user_offset = get_chain_end(h, user_offset);
+ user_offset += get_entry(h, user_offset)->next_offset;
+
+ /* Overflows should be end of entry chains, and unconditional
+ policy nodes. */
+ for (i = 0; i < NUMHOOKS; i++) {
+ STRUCT_ENTRY *e;
+ STRUCT_STANDARD_TARGET *t;
+
+ if (!(h->info.valid_hooks & (1 << i)))
+ continue;
+ assert(h->info.underflow[i]
+ == get_chain_end(h, h->info.hook_entry[i]));
+
+ e = get_entry(h, get_chain_end(h, h->info.hook_entry[i]));
+ assert(unconditional(&e->ip));
+ assert(e->target_offset == sizeof(*e));
+ assert(e->next_offset == sizeof(*e) + sizeof(*t));
+ t = (STRUCT_STANDARD_TARGET *)GET_TARGET(e);
+
+ assert(strcmp(t->target.u.user.name, STANDARD_TARGET)==0);
+ assert(t->verdict == -NF_DROP-1 || t->verdict == -NF_ACCEPT-1);
+
+ /* Hooks and underflows must be valid entries */
+ entry2index(h, get_entry(h, h->info.hook_entry[i]));
+ entry2index(h, get_entry(h, h->info.underflow[i]));
+ }
+
+ assert(h->info.size
+ >= h->info.num_entries * (sizeof(STRUCT_ENTRY)
+ +sizeof(STRUCT_STANDARD_TARGET)));
+
+ assert(h->entries.size
+ >= (h->new_number
+ * (sizeof(STRUCT_ENTRY)
+ + sizeof(STRUCT_STANDARD_TARGET))));
+ assert(strcmp(h->info.name, h->entries.name) == 0);
+
+ i = 0; n = 0;
+ was_return = 0;
+ /* Check all the entries. */
+ ENTRY_ITERATE(h->entries.entries, h->entries.size,
+ check_entry, &i, &n, user_offset, &was_return, h);
+
+ assert(i == h->new_number);
+ assert(n == h->entries.size);
+
+ /* Final entry must be error node */
+ assert(strcmp(GET_TARGET(index2entry(h, h->new_number-1))
+ ->u.user.name,
+ IPT_ERROR_TARGET) == 0);
+}
+#endif /*NDEBUG*/