Harald Welte | f6ebe77 | 2005-08-09 20:21:49 -0700 | [diff] [blame] | 1 | /* netfilter.c: look after the filters for various protocols. |
| 2 | * Heavily influenced by the old firewall.c by David Bonn and Alan Cox. |
| 3 | * |
| 4 | * Thanks to Rob `CmdrTaco' Malda for not influencing this code in any |
| 5 | * way. |
| 6 | * |
| 7 | * Rusty Russell (C)2000 -- This code is GPL. |
| 8 | * |
| 9 | * February 2000: Modified by James Morris to have 1 queue per protocol. |
| 10 | * 15-Mar-2000: Added NF_REPEAT --RR. |
| 11 | * 08-May-2003: Internal logging interface added by Jozsef Kadlecsik. |
| 12 | */ |
| 13 | #include <linux/config.h> |
| 14 | #include <linux/kernel.h> |
| 15 | #include <linux/netfilter.h> |
| 16 | #include <net/protocol.h> |
| 17 | #include <linux/init.h> |
| 18 | #include <linux/skbuff.h> |
| 19 | #include <linux/wait.h> |
| 20 | #include <linux/module.h> |
| 21 | #include <linux/interrupt.h> |
| 22 | #include <linux/if.h> |
| 23 | #include <linux/netdevice.h> |
| 24 | #include <linux/inetdevice.h> |
| 25 | #include <linux/proc_fs.h> |
| 26 | #include <net/sock.h> |
| 27 | |
| 28 | #include "nf_internals.h" |
| 29 | |
Patrick McHardy | bce8032 | 2006-04-06 14:18:09 -0700 | [diff] [blame] | 30 | static DEFINE_SPINLOCK(afinfo_lock); |
| 31 | |
| 32 | struct nf_afinfo *nf_afinfo[NPROTO]; |
| 33 | EXPORT_SYMBOL(nf_afinfo); |
| 34 | |
| 35 | int nf_register_afinfo(struct nf_afinfo *afinfo) |
| 36 | { |
| 37 | spin_lock(&afinfo_lock); |
| 38 | rcu_assign_pointer(nf_afinfo[afinfo->family], afinfo); |
| 39 | spin_unlock(&afinfo_lock); |
| 40 | return 0; |
| 41 | } |
| 42 | EXPORT_SYMBOL_GPL(nf_register_afinfo); |
| 43 | |
| 44 | void nf_unregister_afinfo(struct nf_afinfo *afinfo) |
| 45 | { |
| 46 | spin_lock(&afinfo_lock); |
| 47 | rcu_assign_pointer(nf_afinfo[afinfo->family], NULL); |
| 48 | spin_unlock(&afinfo_lock); |
| 49 | synchronize_rcu(); |
| 50 | } |
| 51 | EXPORT_SYMBOL_GPL(nf_unregister_afinfo); |
| 52 | |
Harald Welte | f6ebe77 | 2005-08-09 20:21:49 -0700 | [diff] [blame] | 53 | /* In this code, we can be waiting indefinitely for userspace to |
| 54 | * service a packet if a hook returns NF_QUEUE. We could keep a count |
| 55 | * of skbuffs queued for userspace, and not deregister a hook unless |
| 56 | * this is zero, but that sucks. Now, we simply check when the |
| 57 | * packets come back: if the hook is gone, the packet is discarded. */ |
| 58 | struct list_head nf_hooks[NPROTO][NF_MAX_HOOKS]; |
| 59 | EXPORT_SYMBOL(nf_hooks); |
| 60 | static DEFINE_SPINLOCK(nf_hook_lock); |
| 61 | |
| 62 | int nf_register_hook(struct nf_hook_ops *reg) |
| 63 | { |
| 64 | struct list_head *i; |
| 65 | |
| 66 | spin_lock_bh(&nf_hook_lock); |
| 67 | list_for_each(i, &nf_hooks[reg->pf][reg->hooknum]) { |
| 68 | if (reg->priority < ((struct nf_hook_ops *)i)->priority) |
| 69 | break; |
| 70 | } |
| 71 | list_add_rcu(®->list, i->prev); |
| 72 | spin_unlock_bh(&nf_hook_lock); |
| 73 | |
| 74 | synchronize_net(); |
| 75 | return 0; |
| 76 | } |
| 77 | EXPORT_SYMBOL(nf_register_hook); |
| 78 | |
| 79 | void nf_unregister_hook(struct nf_hook_ops *reg) |
| 80 | { |
| 81 | spin_lock_bh(&nf_hook_lock); |
| 82 | list_del_rcu(®->list); |
| 83 | spin_unlock_bh(&nf_hook_lock); |
| 84 | |
| 85 | synchronize_net(); |
| 86 | } |
| 87 | EXPORT_SYMBOL(nf_unregister_hook); |
| 88 | |
Patrick McHardy | 972d1cb | 2006-04-06 14:09:12 -0700 | [diff] [blame] | 89 | int nf_register_hooks(struct nf_hook_ops *reg, unsigned int n) |
| 90 | { |
| 91 | unsigned int i; |
| 92 | int err = 0; |
| 93 | |
| 94 | for (i = 0; i < n; i++) { |
| 95 | err = nf_register_hook(®[i]); |
| 96 | if (err) |
| 97 | goto err; |
| 98 | } |
| 99 | return err; |
| 100 | |
| 101 | err: |
| 102 | if (i > 0) |
| 103 | nf_unregister_hooks(reg, i); |
| 104 | return err; |
| 105 | } |
| 106 | EXPORT_SYMBOL(nf_register_hooks); |
| 107 | |
| 108 | void nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n) |
| 109 | { |
| 110 | unsigned int i; |
| 111 | |
| 112 | for (i = 0; i < n; i++) |
| 113 | nf_unregister_hook(®[i]); |
| 114 | } |
| 115 | EXPORT_SYMBOL(nf_unregister_hooks); |
| 116 | |
Harald Welte | f6ebe77 | 2005-08-09 20:21:49 -0700 | [diff] [blame] | 117 | unsigned int nf_iterate(struct list_head *head, |
| 118 | struct sk_buff **skb, |
| 119 | int hook, |
| 120 | const struct net_device *indev, |
| 121 | const struct net_device *outdev, |
| 122 | struct list_head **i, |
| 123 | int (*okfn)(struct sk_buff *), |
| 124 | int hook_thresh) |
| 125 | { |
| 126 | unsigned int verdict; |
| 127 | |
| 128 | /* |
| 129 | * The caller must not block between calls to this |
| 130 | * function because of risk of continuing from deleted element. |
| 131 | */ |
| 132 | list_for_each_continue_rcu(*i, head) { |
| 133 | struct nf_hook_ops *elem = (struct nf_hook_ops *)*i; |
| 134 | |
| 135 | if (hook_thresh > elem->priority) |
| 136 | continue; |
| 137 | |
| 138 | /* Optimization: we don't need to hold module |
| 139 | reference here, since function can't sleep. --RR */ |
| 140 | verdict = elem->hook(hook, skb, indev, outdev, okfn); |
| 141 | if (verdict != NF_ACCEPT) { |
| 142 | #ifdef CONFIG_NETFILTER_DEBUG |
| 143 | if (unlikely((verdict & NF_VERDICT_MASK) |
| 144 | > NF_MAX_VERDICT)) { |
| 145 | NFDEBUG("Evil return from %p(%u).\n", |
| 146 | elem->hook, hook); |
| 147 | continue; |
| 148 | } |
| 149 | #endif |
| 150 | if (verdict != NF_REPEAT) |
| 151 | return verdict; |
| 152 | *i = (*i)->prev; |
| 153 | } |
| 154 | } |
| 155 | return NF_ACCEPT; |
| 156 | } |
| 157 | |
| 158 | |
| 159 | /* Returns 1 if okfn() needs to be executed by the caller, |
| 160 | * -EPERM for NF_DROP, 0 otherwise. */ |
| 161 | int nf_hook_slow(int pf, unsigned int hook, struct sk_buff **pskb, |
| 162 | struct net_device *indev, |
| 163 | struct net_device *outdev, |
| 164 | int (*okfn)(struct sk_buff *), |
| 165 | int hook_thresh) |
| 166 | { |
| 167 | struct list_head *elem; |
| 168 | unsigned int verdict; |
| 169 | int ret = 0; |
| 170 | |
| 171 | /* We may already have this, but read-locks nest anyway */ |
| 172 | rcu_read_lock(); |
| 173 | |
| 174 | elem = &nf_hooks[pf][hook]; |
| 175 | next_hook: |
| 176 | verdict = nf_iterate(&nf_hooks[pf][hook], pskb, hook, indev, |
| 177 | outdev, &elem, okfn, hook_thresh); |
| 178 | if (verdict == NF_ACCEPT || verdict == NF_STOP) { |
| 179 | ret = 1; |
| 180 | goto unlock; |
| 181 | } else if (verdict == NF_DROP) { |
| 182 | kfree_skb(*pskb); |
| 183 | ret = -EPERM; |
| 184 | } else if ((verdict & NF_VERDICT_MASK) == NF_QUEUE) { |
| 185 | NFDEBUG("nf_hook: Verdict = QUEUE.\n"); |
| 186 | if (!nf_queue(pskb, elem, pf, hook, indev, outdev, okfn, |
| 187 | verdict >> NF_VERDICT_BITS)) |
| 188 | goto next_hook; |
| 189 | } |
| 190 | unlock: |
| 191 | rcu_read_unlock(); |
| 192 | return ret; |
| 193 | } |
| 194 | EXPORT_SYMBOL(nf_hook_slow); |
| 195 | |
| 196 | |
| 197 | int skb_make_writable(struct sk_buff **pskb, unsigned int writable_len) |
| 198 | { |
| 199 | struct sk_buff *nskb; |
| 200 | |
| 201 | if (writable_len > (*pskb)->len) |
| 202 | return 0; |
| 203 | |
| 204 | /* Not exclusive use of packet? Must copy. */ |
| 205 | if (skb_shared(*pskb) || skb_cloned(*pskb)) |
| 206 | goto copy_skb; |
| 207 | |
| 208 | return pskb_may_pull(*pskb, writable_len); |
| 209 | |
| 210 | copy_skb: |
| 211 | nskb = skb_copy(*pskb, GFP_ATOMIC); |
| 212 | if (!nskb) |
| 213 | return 0; |
| 214 | BUG_ON(skb_is_nonlinear(nskb)); |
| 215 | |
| 216 | /* Rest of kernel will get very unhappy if we pass it a |
| 217 | suddenly-orphaned skbuff */ |
| 218 | if ((*pskb)->sk) |
| 219 | skb_set_owner_w(nskb, (*pskb)->sk); |
| 220 | kfree_skb(*pskb); |
| 221 | *pskb = nskb; |
| 222 | return 1; |
| 223 | } |
| 224 | EXPORT_SYMBOL(skb_make_writable); |
| 225 | |
| 226 | |
| 227 | /* This does not belong here, but locally generated errors need it if connection |
| 228 | tracking in use: without this, connection may not be in hash table, and hence |
| 229 | manufactured ICMP or RST packets will not be associated with it. */ |
| 230 | void (*ip_ct_attach)(struct sk_buff *, struct sk_buff *); |
| 231 | EXPORT_SYMBOL(ip_ct_attach); |
| 232 | |
| 233 | void nf_ct_attach(struct sk_buff *new, struct sk_buff *skb) |
| 234 | { |
| 235 | void (*attach)(struct sk_buff *, struct sk_buff *); |
| 236 | |
| 237 | if (skb->nfct && (attach = ip_ct_attach) != NULL) { |
| 238 | mb(); /* Just to be sure: must be read before executing this */ |
| 239 | attach(new, skb); |
| 240 | } |
| 241 | } |
| 242 | EXPORT_SYMBOL(nf_ct_attach); |
| 243 | |
| 244 | #ifdef CONFIG_PROC_FS |
| 245 | struct proc_dir_entry *proc_net_netfilter; |
| 246 | EXPORT_SYMBOL(proc_net_netfilter); |
| 247 | #endif |
| 248 | |
| 249 | void __init netfilter_init(void) |
| 250 | { |
| 251 | int i, h; |
| 252 | for (i = 0; i < NPROTO; i++) { |
| 253 | for (h = 0; h < NF_MAX_HOOKS; h++) |
| 254 | INIT_LIST_HEAD(&nf_hooks[i][h]); |
| 255 | } |
| 256 | |
| 257 | #ifdef CONFIG_PROC_FS |
| 258 | proc_net_netfilter = proc_mkdir("netfilter", proc_net); |
| 259 | if (!proc_net_netfilter) |
| 260 | panic("cannot create netfilter proc entry"); |
| 261 | #endif |
| 262 | |
| 263 | if (netfilter_queue_init() < 0) |
| 264 | panic("cannot initialize nf_queue"); |
| 265 | if (netfilter_log_init() < 0) |
| 266 | panic("cannot initialize nf_log"); |
| 267 | } |