netsched: Allow meta match on vlan tag on receive
[safe/jmp/linux-2.6] / net / sched / act_ipt.c
index 39a22a3..082c520 100644 (file)
  * Copyright:  Jamal Hadi Salim (2002-4)
  */
 
-#include <asm/uaccess.h>
-#include <asm/system.h>
-#include <asm/bitops.h>
-#include <linux/config.h>
 #include <linux/types.h>
 #include <linux/kernel.h>
-#include <linux/sched.h>
 #include <linux/string.h>
-#include <linux/mm.h>
-#include <linux/socket.h>
-#include <linux/sockios.h>
-#include <linux/in.h>
 #include <linux/errno.h>
-#include <linux/interrupt.h>
-#include <linux/netdevice.h>
 #include <linux/skbuff.h>
 #include <linux/rtnetlink.h>
 #include <linux/module.h>
 #include <linux/init.h>
-#include <linux/proc_fs.h>
-#include <linux/kmod.h>
-#include <net/sock.h>
+#include <net/netlink.h>
 #include <net/pkt_sched.h>
 #include <linux/tc_act/tc_ipt.h>
 #include <net/tc_act/tc_ipt.h>
 
 #include <linux/netfilter_ipv4/ip_tables.h>
 
-/* use generic hash table */
-#define MY_TAB_SIZE     16
-#define MY_TAB_MASK     15
 
-static u32 idx_gen;
-static struct tcf_ipt *tcf_ipt_ht[MY_TAB_SIZE];
-/* ipt hash table lock */
+#define IPT_TAB_MASK     15
+static struct tcf_common *tcf_ipt_ht[IPT_TAB_MASK + 1];
+static u32 ipt_idx_gen;
 static DEFINE_RWLOCK(ipt_lock);
 
-/* ovewrride the defaults */
-#define tcf_st         tcf_ipt
-#define tcf_t_lock     ipt_lock
-#define tcf_ht         tcf_ipt_ht
-
-#define CONFIG_NET_ACT_INIT
-#include <net/pkt_act.h>
+static struct tcf_hashinfo ipt_hash_info = {
+       .htab   =       tcf_ipt_ht,
+       .hmask  =       IPT_TAB_MASK,
+       .lock   =       &ipt_lock,
+};
 
-static int
-ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
+static int ipt_init_target(struct ipt_entry_target *t, char *table, unsigned int hook)
 {
-       struct ipt_target *target;
+       struct xt_tgchk_param par;
+       struct xt_target *target;
        int ret = 0;
 
-       target = xt_find_target(AF_INET, t->u.user.name, t->u.user.revision);
+       target = xt_request_find_target(AF_INET, t->u.user.name,
+                                       t->u.user.revision);
        if (!target)
                return -ENOENT;
 
-       DPRINTK("ipt_init_target: found %s\n", target->name);
        t->u.kernel.target = target;
-
-       if (t->u.kernel.target->checkentry
-           && !t->u.kernel.target->checkentry(table, NULL, t->data,
-                                              t->u.target_size - sizeof(*t),
-                                              hook)) {
-               DPRINTK("ipt_init_target: check failed for `%s'.\n",
-                       t->u.kernel.target->name);
+       par.table     = table;
+       par.entryinfo = NULL;
+       par.target    = target;
+       par.targinfo  = t->data;
+       par.hook_mask = hook;
+       par.family    = NFPROTO_IPV4;
+
+       ret = xt_check_target(&par, t->u.target_size - sizeof(*t), 0, false);
+       if (ret < 0) {
                module_put(t->u.kernel.target->me);
-               ret = -EINVAL;
+               return ret;
        }
-
-       return ret;
+       return 0;
 }
 
-static void
-ipt_destroy_target(struct ipt_entry_target *t)
+static void ipt_destroy_target(struct ipt_entry_target *t)
 {
-       if (t->u.kernel.target->destroy)
-               t->u.kernel.target->destroy(t->data,
-                                           t->u.target_size - sizeof(*t));
-        module_put(t->u.kernel.target->me);
+       struct xt_tgdtor_param par = {
+               .target   = t->u.kernel.target,
+               .targinfo = t->data,
+       };
+       if (par.target->destroy != NULL)
+               par.target->destroy(&par);
+       module_put(par.target->me);
 }
 
-static int
-tcf_ipt_release(struct tcf_ipt *p, int bind)
+static int tcf_ipt_release(struct tcf_ipt *ipt, int bind)
 {
        int ret = 0;
-       if (p) {
+       if (ipt) {
                if (bind)
-                       p->bindcnt--;
-               p->refcnt--;
-               if (p->bindcnt <= 0 && p->refcnt <= 0) {
-                       ipt_destroy_target(p->t);
-                       kfree(p->tname);
-                       kfree(p->t);
-                       tcf_hash_destroy(p);
+                       ipt->tcf_bindcnt--;
+               ipt->tcf_refcnt--;
+               if (ipt->tcf_bindcnt <= 0 && ipt->tcf_refcnt <= 0) {
+                       ipt_destroy_target(ipt->tcfi_t);
+                       kfree(ipt->tcfi_tname);
+                       kfree(ipt->tcfi_t);
+                       tcf_hash_destroy(&ipt->common, &ipt_hash_info);
                        ret = ACT_P_DELETED;
                }
        }
        return ret;
 }
 
-static int
-tcf_ipt_init(struct rtattr *rta, struct rtattr *est, struct tc_action *a,
-             int ovr, int bind)
+static const struct nla_policy ipt_policy[TCA_IPT_MAX + 1] = {
+       [TCA_IPT_TABLE] = { .type = NLA_STRING, .len = IFNAMSIZ },
+       [TCA_IPT_HOOK]  = { .type = NLA_U32 },
+       [TCA_IPT_INDEX] = { .type = NLA_U32 },
+       [TCA_IPT_TARG]  = { .len = sizeof(struct ipt_entry_target) },
+};
+
+static int tcf_ipt_init(struct nlattr *nla, struct nlattr *est,
+                       struct tc_action *a, int ovr, int bind)
 {
-       struct rtattr *tb[TCA_IPT_MAX];
-       struct tcf_ipt *p;
+       struct nlattr *tb[TCA_IPT_MAX + 1];
+       struct tcf_ipt *ipt;
+       struct tcf_common *pc;
        struct ipt_entry_target *td, *t;
        char *tname;
        int ret = 0, err;
        u32 hook = 0;
        u32 index = 0;
 
-       if (rta == NULL || rtattr_parse_nested(tb, TCA_IPT_MAX, rta) < 0)
+       if (nla == NULL)
                return -EINVAL;
 
-       if (tb[TCA_IPT_HOOK-1] == NULL ||
-           RTA_PAYLOAD(tb[TCA_IPT_HOOK-1]) < sizeof(u32))
+       err = nla_parse_nested(tb, TCA_IPT_MAX, nla, ipt_policy);
+       if (err < 0)
+               return err;
+
+       if (tb[TCA_IPT_HOOK] == NULL)
                return -EINVAL;
-       if (tb[TCA_IPT_TARG-1] == NULL ||
-           RTA_PAYLOAD(tb[TCA_IPT_TARG-1]) < sizeof(*t))
+       if (tb[TCA_IPT_TARG] == NULL)
                return -EINVAL;
-       td = (struct ipt_entry_target *)RTA_DATA(tb[TCA_IPT_TARG-1]);
-       if (RTA_PAYLOAD(tb[TCA_IPT_TARG-1]) < td->u.target_size)
+
+       td = (struct ipt_entry_target *)nla_data(tb[TCA_IPT_TARG]);
+       if (nla_len(tb[TCA_IPT_TARG]) < td->u.target_size)
                return -EINVAL;
 
-       if (tb[TCA_IPT_INDEX-1] != NULL &&
-           RTA_PAYLOAD(tb[TCA_IPT_INDEX-1]) >= sizeof(u32))
-               index = *(u32 *)RTA_DATA(tb[TCA_IPT_INDEX-1]);
+       if (tb[TCA_IPT_INDEX] != NULL)
+               index = nla_get_u32(tb[TCA_IPT_INDEX]);
 
-       p = tcf_hash_check(index, a, ovr, bind);
-       if (p == NULL) {
-               p = tcf_hash_create(index, est, a, sizeof(*p), ovr, bind);
-               if (p == NULL)
-                       return -ENOMEM;
+       pc = tcf_hash_check(index, a, bind, &ipt_hash_info);
+       if (!pc) {
+               pc = tcf_hash_create(index, est, a, sizeof(*ipt), bind,
+                                    &ipt_idx_gen, &ipt_hash_info);
+               if (IS_ERR(pc))
+                   return PTR_ERR(pc);
                ret = ACT_P_CREATED;
        } else {
                if (!ovr) {
-                       tcf_ipt_release(p, bind);
+                       tcf_ipt_release(to_ipt(pc), bind);
                        return -EEXIST;
                }
        }
+       ipt = to_ipt(pc);
 
-       hook = *(u32 *)RTA_DATA(tb[TCA_IPT_HOOK-1]);
+       hook = nla_get_u32(tb[TCA_IPT_HOOK]);
 
        err = -ENOMEM;
        tname = kmalloc(IFNAMSIZ, GFP_KERNEL);
-       if (tname == NULL)
+       if (unlikely(!tname))
                goto err1;
-       if (tb[TCA_IPT_TABLE - 1] == NULL ||
-           rtattr_strlcpy(tname, tb[TCA_IPT_TABLE-1], IFNAMSIZ) >= IFNAMSIZ)
+       if (tb[TCA_IPT_TABLE] == NULL ||
+           nla_strlcpy(tname, tb[TCA_IPT_TABLE], IFNAMSIZ) >= IFNAMSIZ)
                strcpy(tname, "mangle");
 
-       t = kmalloc(td->u.target_size, GFP_KERNEL);
-       if (t == NULL)
+       t = kmemdup(td, td->u.target_size, GFP_KERNEL);
+       if (unlikely(!t))
                goto err2;
-       memcpy(t, td, td->u.target_size);
 
        if ((err = ipt_init_target(t, tname, hook)) < 0)
                goto err3;
 
-       spin_lock_bh(&p->lock);
+       spin_lock_bh(&ipt->tcf_lock);
        if (ret != ACT_P_CREATED) {
-               ipt_destroy_target(p->t);
-               kfree(p->tname);
-               kfree(p->t);
+               ipt_destroy_target(ipt->tcfi_t);
+               kfree(ipt->tcfi_tname);
+               kfree(ipt->tcfi_t);
        }
-       p->tname = tname;
-       p->t     = t;
-       p->hook  = hook;
-       spin_unlock_bh(&p->lock);
+       ipt->tcfi_tname = tname;
+       ipt->tcfi_t     = t;
+       ipt->tcfi_hook  = hook;
+       spin_unlock_bh(&ipt->tcf_lock);
        if (ret == ACT_P_CREATED)
-               tcf_hash_insert(p);
+               tcf_hash_insert(pc, &ipt_hash_info);
        return ret;
 
 err3:
@@ -189,50 +183,51 @@ err3:
 err2:
        kfree(tname);
 err1:
-       kfree(p);
+       kfree(pc);
        return err;
 }
 
-static int
-tcf_ipt_cleanup(struct tc_action *a, int bind)
+static int tcf_ipt_cleanup(struct tc_action *a, int bind)
 {
-       struct tcf_ipt *p = PRIV(a, ipt);
-       return tcf_ipt_release(p, bind);
+       struct tcf_ipt *ipt = a->priv;
+       return tcf_ipt_release(ipt, bind);
 }
 
-static int
-tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
+static int tcf_ipt(struct sk_buff *skb, struct tc_action *a,
+                  struct tcf_result *res)
 {
        int ret = 0, result = 0;
-       struct tcf_ipt *p = PRIV(a, ipt);
+       struct tcf_ipt *ipt = a->priv;
+       struct xt_target_param par;
 
        if (skb_cloned(skb)) {
                if (pskb_expand_head(skb, 0, 0, GFP_ATOMIC))
                        return TC_ACT_UNSPEC;
        }
 
-       spin_lock(&p->lock);
+       spin_lock(&ipt->tcf_lock);
 
-       p->tm.lastuse = jiffies;
-       p->bstats.bytes += skb->len;
-       p->bstats.packets++;
+       ipt->tcf_tm.lastuse = jiffies;
+       ipt->tcf_bstats.bytes += qdisc_pkt_len(skb);
+       ipt->tcf_bstats.packets++;
 
        /* yes, we have to worry about both in and out dev
         worry later - danger - this API seems to have changed
         from earlier kernels */
+       par.in       = skb->dev;
+       par.out      = NULL;
+       par.hooknum  = ipt->tcfi_hook;
+       par.target   = ipt->tcfi_t->u.kernel.target;
+       par.targinfo = ipt->tcfi_t->data;
+       ret = par.target->target(skb, &par);
 
-       /* iptables targets take a double skb pointer in case the skb
-        * needs to be replaced. We don't own the skb, so this must not
-        * happen. The pskb_expand_head above should make sure of this */
-       ret = p->t->u.kernel.target->target(&skb, skb->dev, NULL,
-                                           p->hook, p->t->data, NULL);
        switch (ret) {
        case NF_ACCEPT:
                result = TC_ACT_OK;
                break;
        case NF_DROP:
                result = TC_ACT_SHOT;
-               p->qstats.drops++;
+               ipt->tcf_qstats.drops++;
                break;
        case IPT_CONTINUE:
                result = TC_ACT_PIPE;
@@ -243,60 +238,53 @@ tcf_ipt(struct sk_buff *skb, struct tc_action *a, struct tcf_result *res)
                result = TC_POLICE_OK;
                break;
        }
-       spin_unlock(&p->lock);
+       spin_unlock(&ipt->tcf_lock);
        return result;
 
 }
 
-static int
-tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
+static int tcf_ipt_dump(struct sk_buff *skb, struct tc_action *a, int bind, int ref)
 {
+       unsigned char *b = skb_tail_pointer(skb);
+       struct tcf_ipt *ipt = a->priv;
        struct ipt_entry_target *t;
        struct tcf_t tm;
        struct tc_cnt c;
-       unsigned char *b = skb->tail;
-       struct tcf_ipt *p = PRIV(a, ipt);
 
        /* for simple targets kernel size == user size
        ** user name = target name
        ** for foolproof you need to not assume this
        */
 
-       t = kmalloc(p->t->u.user.target_size, GFP_ATOMIC);
-       if (t == NULL)
-               goto rtattr_failure;
-
-       c.bindcnt = p->bindcnt - bind;
-       c.refcnt = p->refcnt - ref;
-       memcpy(t, p->t, p->t->u.user.target_size);
-       strcpy(t->u.user.name, p->t->u.kernel.target->name);
-
-       DPRINTK("\ttcf_ipt_dump tablename %s length %d\n", p->tname,
-               strlen(p->tname));
-       DPRINTK("\tdump target name %s size %d size user %d "
-               "data[0] %x data[1] %x\n", p->t->u.kernel.target->name,
-               p->t->u.target_size, p->t->u.user.target_size,
-               p->t->data[0], p->t->data[1]);
-       RTA_PUT(skb, TCA_IPT_TARG, p->t->u.user.target_size, t);
-       RTA_PUT(skb, TCA_IPT_INDEX, 4, &p->index);
-       RTA_PUT(skb, TCA_IPT_HOOK, 4, &p->hook);
-       RTA_PUT(skb, TCA_IPT_CNT, sizeof(struct tc_cnt), &c);
-       RTA_PUT(skb, TCA_IPT_TABLE, IFNAMSIZ, p->tname);
-       tm.install = jiffies_to_clock_t(jiffies - p->tm.install);
-       tm.lastuse = jiffies_to_clock_t(jiffies - p->tm.lastuse);
-       tm.expires = jiffies_to_clock_t(p->tm.expires);
-       RTA_PUT(skb, TCA_IPT_TM, sizeof (tm), &tm);
+       t = kmemdup(ipt->tcfi_t, ipt->tcfi_t->u.user.target_size, GFP_ATOMIC);
+       if (unlikely(!t))
+               goto nla_put_failure;
+
+       c.bindcnt = ipt->tcf_bindcnt - bind;
+       c.refcnt = ipt->tcf_refcnt - ref;
+       strcpy(t->u.user.name, ipt->tcfi_t->u.kernel.target->name);
+
+       NLA_PUT(skb, TCA_IPT_TARG, ipt->tcfi_t->u.user.target_size, t);
+       NLA_PUT_U32(skb, TCA_IPT_INDEX, ipt->tcf_index);
+       NLA_PUT_U32(skb, TCA_IPT_HOOK, ipt->tcfi_hook);
+       NLA_PUT(skb, TCA_IPT_CNT, sizeof(struct tc_cnt), &c);
+       NLA_PUT_STRING(skb, TCA_IPT_TABLE, ipt->tcfi_tname);
+       tm.install = jiffies_to_clock_t(jiffies - ipt->tcf_tm.install);
+       tm.lastuse = jiffies_to_clock_t(jiffies - ipt->tcf_tm.lastuse);
+       tm.expires = jiffies_to_clock_t(ipt->tcf_tm.expires);
+       NLA_PUT(skb, TCA_IPT_TM, sizeof (tm), &tm);
        kfree(t);
        return skb->len;
 
-      rtattr_failure:
-       skb_trim(skb, b - skb->data);
+nla_put_failure:
+       nlmsg_trim(skb, b);
        kfree(t);
        return -1;
 }
 
 static struct tc_action_ops act_ipt_ops = {
        .kind           =       "ipt",
+       .hinfo          =       &ipt_hash_info,
        .type           =       TCA_ACT_IPT,
        .capab          =       TCA_CAP_NONE,
        .owner          =       THIS_MODULE,
@@ -312,14 +300,12 @@ MODULE_AUTHOR("Jamal Hadi Salim(2002-4)");
 MODULE_DESCRIPTION("Iptables target actions");
 MODULE_LICENSE("GPL");
 
-static int __init
-ipt_init_module(void)
+static int __init ipt_init_module(void)
 {
        return tcf_register_action(&act_ipt_ops);
 }
 
-static void __exit
-ipt_cleanup_module(void)
+static void __exit ipt_cleanup_module(void)
 {
        tcf_unregister_action(&act_ipt_ops);
 }