1
linux/net/netfilter/nfnetlink.c

346 lines
8.1 KiB
C
Raw Normal View History

/* Netfilter messages via netlink socket. Allows for user space
* protocol helpers and general trouble making from userspace.
*
* (C) 2001 by Jay Schulist <jschlst@samba.org>,
* (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
* (C) 2005 by Pablo Neira Ayuso <pablo@eurodev.net>
*
* Initial netfilter messages via netlink development funded and
* generally made possible by Network Robots, Inc. (www.networkrobots.com)
*
* Further development of this code funded by Astaro AG (http://www.astaro.com)
*
* This software may be used and distributed according to the terms
* of the GNU General Public License, incorporated herein by reference.
*/
#include <linux/module.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/kernel.h>
#include <linux/major.h>
#include <linux/timer.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/fcntl.h>
#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <asm/system.h>
#include <net/sock.h>
#include <net/netlink.h>
#include <linux/init.h>
#include <linux/spinlock.h>
#include <linux/netfilter.h>
#include <linux/netlink.h>
#include <linux/netfilter/nfnetlink.h>
MODULE_LICENSE("GPL");
MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
static char __initdata nfversion[] = "0.30";
#if 0
#define DEBUGP(format, args...) \
printk(KERN_DEBUG "%s(%d):%s(): " format, __FILE__, \
__LINE__, __FUNCTION__, ## args)
#else
#define DEBUGP(format, args...)
#endif
static struct sock *nfnl = NULL;
static struct nfnetlink_subsystem *subsys_table[NFNL_SUBSYS_COUNT];
static DEFINE_MUTEX(nfnl_mutex);
static void nfnl_lock(void)
{
mutex_lock(&nfnl_mutex);
}
static int nfnl_trylock(void)
{
return !mutex_trylock(&nfnl_mutex);
}
static void __nfnl_unlock(void)
{
mutex_unlock(&nfnl_mutex);
}
static void nfnl_unlock(void)
{
mutex_unlock(&nfnl_mutex);
if (nfnl->sk_receive_queue.qlen)
nfnl->sk_data_ready(nfnl, 0);
}
int nfnetlink_subsys_register(struct nfnetlink_subsystem *n)
{
DEBUGP("registering subsystem ID %u\n", n->subsys_id);
nfnl_lock();
if (subsys_table[n->subsys_id]) {
nfnl_unlock();
return -EBUSY;
}
subsys_table[n->subsys_id] = n;
nfnl_unlock();
return 0;
}
int nfnetlink_subsys_unregister(struct nfnetlink_subsystem *n)
{
DEBUGP("unregistering subsystem ID %u\n", n->subsys_id);
nfnl_lock();
subsys_table[n->subsys_id] = NULL;
nfnl_unlock();
return 0;
}
static inline struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
{
u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
if (subsys_id >= NFNL_SUBSYS_COUNT
|| subsys_table[subsys_id] == NULL)
return NULL;
return subsys_table[subsys_id];
}
static inline struct nfnl_callback *
nfnetlink_find_client(u_int16_t type, struct nfnetlink_subsystem *ss)
{
u_int8_t cb_id = NFNL_MSG_TYPE(type);
if (cb_id >= ss->cb_count) {
DEBUGP("msgtype %u >= %u, returning\n", type, ss->cb_count);
return NULL;
}
return &ss->cb[cb_id];
}
void __nfa_fill(struct sk_buff *skb, int attrtype, int attrlen,
const void *data)
{
struct nfattr *nfa;
int size = NFA_LENGTH(attrlen);
nfa = (struct nfattr *)skb_put(skb, NFA_ALIGN(size));
nfa->nfa_type = attrtype;
nfa->nfa_len = size;
memcpy(NFA_DATA(nfa), data, attrlen);
memset(NFA_DATA(nfa) + attrlen, 0, NFA_ALIGN(size) - size);
}
void nfattr_parse(struct nfattr *tb[], int maxattr, struct nfattr *nfa, int len)
{
memset(tb, 0, sizeof(struct nfattr *) * maxattr);
while (NFA_OK(nfa, len)) {
unsigned flavor = NFA_TYPE(nfa);
if (flavor && flavor <= maxattr)
tb[flavor-1] = nfa;
nfa = NFA_NEXT(nfa, len);
}
}
/**
* nfnetlink_check_attributes - check and parse nfnetlink attributes
*
* subsys: nfnl subsystem for which this message is to be parsed
* nlmsghdr: netlink message to be checked/parsed
* cda: array of pointers, needs to be at least subsys->attr_count big
*
*/
static int
nfnetlink_check_attributes(struct nfnetlink_subsystem *subsys,
struct nlmsghdr *nlh, struct nfattr *cda[])
{
int min_len;
u_int16_t attr_count;
u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
if (unlikely(cb_id >= subsys->cb_count)) {
DEBUGP("msgtype %u >= %u, returning\n",
cb_id, subsys->cb_count);
return -EINVAL;
}
min_len = NLMSG_SPACE(sizeof(struct nfgenmsg));
if (unlikely(nlh->nlmsg_len < min_len))
return -EINVAL;
attr_count = subsys->cb[cb_id].attr_count;
memset(cda, 0, sizeof(struct nfattr *) * attr_count);
/* check attribute lengths. */
if (likely(nlh->nlmsg_len > min_len)) {
struct nfattr *attr = NFM_NFA(NLMSG_DATA(nlh));
int attrlen = nlh->nlmsg_len - NLMSG_ALIGN(min_len);
while (NFA_OK(attr, attrlen)) {
unsigned flavor = NFA_TYPE(attr);
if (flavor) {
if (flavor > attr_count)
return -EINVAL;
cda[flavor - 1] = attr;
}
attr = NFA_NEXT(attr, attrlen);
}
}
/* implicit: if nlmsg_len == min_len, we return 0, and an empty
* (zeroed) cda[] array. The message is valid, but empty. */
return 0;
}
int nfnetlink_has_listeners(unsigned int group)
{
return netlink_has_listeners(nfnl, group);
}
EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
int nfnetlink_send(struct sk_buff *skb, u32 pid, unsigned group, int echo)
{
int err = 0;
NETLINK_CB(skb).dst_group = group;
if (echo)
atomic_inc(&skb->users);
[NET]: Handle disabled preemption in gfp_any() ctnetlink uses netlink_unicast from an atomic_notifier_chain (which is called within a RCU read side critical section) without holding further locks. netlink_unicast calls netlink_trim with the result of gfp_any() for the gfp flags, which are passed down to pskb_expand_header. gfp_any() only checks for softirq context and returns GFP_KERNEL, resulting in this warning: BUG: sleeping function called from invalid context at mm/slab.c:3032 in_atomic():1, irqs_disabled():0 no locks held by rmmod/7010. Call Trace: [<ffffffff8109467f>] debug_show_held_locks+0x9/0xb [<ffffffff8100b0b4>] __might_sleep+0xd9/0xdb [<ffffffff810b5082>] __kmalloc+0x68/0x110 [<ffffffff811ba8f2>] pskb_expand_head+0x4d/0x13b [<ffffffff81053147>] netlink_broadcast+0xa5/0x2e0 [<ffffffff881cd1d7>] :nfnetlink:nfnetlink_send+0x83/0x8a [<ffffffff8834f6a6>] :nf_conntrack_netlink:ctnetlink_conntrack_event+0x94c/0x96a [<ffffffff810624d6>] notifier_call_chain+0x29/0x3e [<ffffffff8106251d>] atomic_notifier_call_chain+0x32/0x60 [<ffffffff881d266d>] :nf_conntrack:destroy_conntrack+0xa5/0x1d3 [<ffffffff881d194e>] :nf_conntrack:nf_ct_cleanup+0x8c/0x12c [<ffffffff881d4614>] :nf_conntrack:kill_l3proto+0x0/0x13 [<ffffffff881d482a>] :nf_conntrack:nf_conntrack_l3proto_unregister+0x90/0x94 [<ffffffff883551b3>] :nf_conntrack_ipv4:nf_conntrack_l3proto_ipv4_fini+0x2b/0x5d [<ffffffff8109d44f>] sys_delete_module+0x1b5/0x1e6 [<ffffffff8105f245>] trace_hardirqs_on_thunk+0x35/0x37 [<ffffffff8105911e>] system_call+0x7e/0x83 Since netlink_unicast is supposed to be callable from within RCU read side critical sections, make gfp_any() check for in_atomic() instead of in_softirq(). Additionally nfnetlink_send needs to use gfp_any() as well for the call to netlink_broadcast). Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-02-27 10:56:42 -07:00
netlink_broadcast(nfnl, skb, pid, group, gfp_any());
if (echo)
err = netlink_unicast(nfnl, skb, pid, MSG_DONTWAIT);
return err;
}
int nfnetlink_unicast(struct sk_buff *skb, u_int32_t pid, int flags)
{
return netlink_unicast(nfnl, skb, pid, flags);
}
/* Process one complete nfnetlink message. */
static int nfnetlink_rcv_msg(struct sk_buff *skb,
struct nlmsghdr *nlh, int *errp)
{
struct nfnl_callback *nc;
struct nfnetlink_subsystem *ss;
int type, err = 0;
DEBUGP("entered; subsys=%u, msgtype=%u\n",
NFNL_SUBSYS_ID(nlh->nlmsg_type),
NFNL_MSG_TYPE(nlh->nlmsg_type));
if (security_netlink_recv(skb, CAP_NET_ADMIN)) {
DEBUGP("missing CAP_NET_ADMIN\n");
*errp = -EPERM;
return -1;
}
/* Only requests are handled by kernel now. */
if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
DEBUGP("received non-request message\n");
return 0;
}
/* All the messages must at least contain nfgenmsg */
if (nlh->nlmsg_len < NLMSG_SPACE(sizeof(struct nfgenmsg))) {
DEBUGP("received message was too short\n");
return 0;
}
type = nlh->nlmsg_type;
ss = nfnetlink_get_subsys(type);
if (!ss) {
#ifdef CONFIG_KMOD
/* don't call nfnl_unlock, since it would reenter
* with further packet processing */
__nfnl_unlock();
request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
nfnl_lock();
ss = nfnetlink_get_subsys(type);
if (!ss)
#endif
goto err_inval;
}
nc = nfnetlink_find_client(type, ss);
if (!nc) {
DEBUGP("unable to find client for type %d\n", type);
goto err_inval;
}
{
u_int16_t attr_count =
ss->cb[NFNL_MSG_TYPE(nlh->nlmsg_type)].attr_count;
struct nfattr *cda[attr_count];
memset(cda, 0, sizeof(struct nfattr *) * attr_count);
err = nfnetlink_check_attributes(ss, nlh, cda);
if (err < 0)
goto err_inval;
DEBUGP("calling handler\n");
err = nc->call(nfnl, skb, nlh, cda, errp);
*errp = err;
return err;
}
err_inval:
DEBUGP("returning -EINVAL\n");
*errp = -EINVAL;
return -1;
}
static void nfnetlink_rcv(struct sock *sk, int len)
{
unsigned int qlen = 0;
do {
if (nfnl_trylock())
return;
netlink_run_queue(sk, &qlen, nfnetlink_rcv_msg);
__nfnl_unlock();
} while (qlen);
}
static void __exit nfnetlink_exit(void)
{
printk("Removing netfilter NETLINK layer.\n");
sock_release(nfnl->sk_socket);
return;
}
static int __init nfnetlink_init(void)
{
printk("Netfilter messages via NETLINK v%s.\n", nfversion);
nfnl = netlink_kernel_create(NETLINK_NETFILTER, NFNLGRP_MAX,
nfnetlink_rcv, THIS_MODULE);
if (!nfnl) {
printk(KERN_ERR "cannot initialize nfnetlink!\n");
return -1;
}
return 0;
}
module_init(nfnetlink_init);
module_exit(nfnetlink_exit);
EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
EXPORT_SYMBOL_GPL(nfnetlink_send);
EXPORT_SYMBOL_GPL(nfnetlink_unicast);
EXPORT_SYMBOL_GPL(nfattr_parse);
EXPORT_SYMBOL_GPL(__nfa_fill);