path: root/net/ipv4/udp.c

/*
 * INET		An implementation of the TCP/IP protocol suite for the LINUX
 *		operating system.  INET is implemented using the  BSD Socket
 *		interface as the means of communication with the user level.
 *
 *		The User Datagram Protocol (UDP).
 *
 * Version:	$Id: udp.c,v 1.102 2002/02/01 22:01:04 davem Exp $
 *
 * Authors:	Ross Biro
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
 *		Alan Cox, <Alan.Cox@linux.org>
 *		Hirokazu Takahashi, <taka@valinux.co.jp>
 *
 * Fixes:
 *		Alan Cox	:	verify_area() calls
 *		Alan Cox	: 	stopped close while in use off icmp
 *					messages. Not a fix but a botch that
 *					for udp at least is 'valid'.
 *		Alan Cox	:	Fixed icmp handling properly
 *		Alan Cox	: 	Correct error for oversized datagrams
 *		Alan Cox	:	Tidied select() semantics.
 *		Alan Cox	:	udp_err() fixed properly, also now
 *					select and read wake correctly on errors
 *		Alan Cox	:	udp_send verify_area moved to avoid mem leak
 *		Alan Cox	:	UDP can count its memory
 *		Alan Cox	:	send to an unknown connection causes
 *					an ECONNREFUSED off the icmp, but
 *					does NOT close.
 *		Alan Cox	:	Switched to new sk_buff handlers. No more backlog!
 *		Alan Cox	:	Using generic datagram code. Even smaller and the PEEK
 *					bug no longer crashes it.
 *		Fred Van Kempen	: 	Net2e support for sk->broadcast.
 *		Alan Cox	:	Uses skb_free_datagram
 *		Alan Cox	:	Added get/set sockopt support.
 *		Alan Cox	:	Broadcasting without option set returns EACCES.
 *		Alan Cox	:	No wakeup calls. Instead we now use the callbacks.
 *		Alan Cox	:	Use ip_tos and ip_ttl
 *		Alan Cox	:	SNMP Mibs
 *		Alan Cox	:	MSG_DONTROUTE, and 0.0.0.0 support.
 *		Matt Dillon	:	UDP length checks.
 *		Alan Cox	:	Smarter af_inet used properly.
 *		Alan Cox	:	Use new kernel side addressing.
 *		Alan Cox	:	Incorrect return on truncated datagram receive.
 *	Arnt Gulbrandsen 	:	New udp_send and stuff
 *		Alan Cox	:	Cache last socket
 *		Alan Cox	:	Route cache
 *		Jon Peatfield	:	Minor efficiency fix to sendto().
 *		Mike Shaver	:	RFC1122 checks.
 *		Alan Cox	:	Nonblocking error fix.
 *	Willy Konynenberg	:	Transparent proxying support.
 *		Mike McLagan	:	Routing by source
 *		David S. Miller	:	New socket lookup architecture.
 *					Last socket cache retained as it
 *					does have a high hit rate.
 *		Olaf Kirch	:	Don't linearise iovec on sendmsg.
 *		Andi Kleen	:	Some cleanups, cache destination entry
 *					for connect.
 *	Vitaly E. Lavrov	:	Transparent proxy revived after year coma.
 *		Melvin Smith	:	Check msg_name not msg_namelen in sendto(),
 *					return ENOTCONN for unconnected sockets (POSIX)
 *		Janos Farkas	:	don't deliver multi/broadcasts to a different
 *					bound-to-device socket
 *	Hirokazu Takahashi	:	HW checksumming for outgoing UDP
 *					datagrams.
 *	Hirokazu Takahashi	:	sendfile() on UDP works now.
 *		Arnaldo C. Melo :	convert /proc/net/udp to seq_file
 *	YOSHIFUJI Hideaki @USAGI and:	Support IPV6_V6ONLY socket option, which
 *	Alexey Kuznetsov:		allow both IPv4 and IPv6 sockets to bind
 *					a single port at the same time.
 *	Derek Atkins <derek@ihtfp.com>: Add Encapulation Support
 *	James Chapman		:	Add L2TP encapsulation type.
 *
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 */

#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/ioctls.h>
#include <linux/bootmem.h>
#include <linux/types.h>
#include <linux/fcntl.h>
#include <linux/module.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/igmp.h>
#include <linux/in.h>
#include <linux/errno.h>
#include <linux/timer.h>
#include <linux/mm.h>
#include <linux/inet.h>
#include <linux/netdevice.h>
#include <net/tcp_states.h>
#include <linux/skbuff.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <net/net_namespace.h>
#include <net/icmp.h>
#include <net/route.h>
#include <net/checksum.h>
#include <net/xfrm.h>
#include "udp_impl.h"

/*
 *	Snmp MIB for the UDP layer
 */

DEFINE_SNMP_STAT(struct udp_mib, udp_statistics) __read_mostly;
EXPORT_SYMBOL(udp_statistics);

DEFINE_SNMP_STAT(struct udp_mib, udp_stats_in6) __read_mostly;
EXPORT_SYMBOL(udp_stats_in6);

struct hlist_head udp_hash[UDP_HTABLE_SIZE];
DEFINE_RWLOCK(udp_hash_lock);

int sysctl_udp_mem[3] __read_mostly;
int sysctl_udp_rmem_min __read_mostly;
int sysctl_udp_wmem_min __read_mostly;

EXPORT_SYMBOL(sysctl_udp_mem);
EXPORT_SYMBOL(sysctl_udp_rmem_min);
EXPORT_SYMBOL(sysctl_udp_wmem_min);

atomic_t udp_memory_allocated;
EXPORT_SYMBOL(udp_memory_allocated);

static inline int __udp_lib_lport_inuse(struct net *net, __u16 num,
					const struct hlist_head udptable[])
{
	struct sock *sk;
	struct hlist_node *node;

	sk_for_each(sk, node, &udptable[num & (UDP_HTABLE_SIZE - 1)])
		if (sk->sk_net == net && sk->sk_hash == num)
			return 1;
	return 0;
}

/**
 *  __udp_lib_get_port  -  UDP/-Lite port lookup for IPv4 and IPv6
 *
 *  @sk:          socket struct in question
 *  @snum:        port number to look up
 *  @udptable:    hash list table, must be of UDP_HTABLE_SIZE
 *  @saddr_comp:  AF-dependent comparison of bound local IP addresses
 */
int __udp_lib_get_port(struct sock *sk, unsigned short snum,
		       struct hlist_head udptable[],
		       int (*saddr_comp)(const struct sock *sk1,
					 const struct sock *sk2 )    )
{
	struct hlist_node *node;
	struct hlist_head *head;
	struct sock *sk2;
	int    error = 1;
	struct net *net = sk->sk_net;

	write_lock_bh(&udp_hash_lock);

	if (!snum) {
		int i, low, high, remaining;
		unsigned rover, best, best_size_so_far;

		inet_get_local_port_range(&low, &high);
		remaining = (high - low) + 1;

		best_size_so_far = UINT_MAX;
		best = rover = net_random() % remaining + low;

		/* 1st pass: look for empty (or shortest) hash chain */
		for (i = 0; i < UDP_HTABLE_SIZE; i++) {
			int size = 0;

			head = &udptable[rover & (UDP_HTABLE_SIZE - 1)];
			if (hlist_empty(head))
				goto gotit;

			sk_for_each(sk2, node, head) {
				if (++size >= best_size_so_far)
					goto next;
			}
			best_size_so_far = size;
			best = rover;
		next:
			/* fold back if end of range */
			if (++rover > high)
				rover = low + ((rover - low)
					       & (UDP_HTABLE_SIZE - 1));


		}

		/* 2nd pass: find hole in shortest hash chain */
		rover = best;
		for (i = 0; i < (1 << 16) / UDP_HTABLE_SIZE; i++) {
			if (! __udp_lib_lport_inuse(net, rover, udptable))
				goto gotit;
			rover += UDP_HTABLE_SIZE;
			if (rover > high)
				rover = low + ((rover - low)
					       & (UDP_HTABLE_SIZE - 1));
		}


		/* All ports in use! */
		goto fail;

gotit:
		snum = rover;
	} else {
		head = &udptable[snum & (UDP_HTABLE_SIZE - 1)];

		sk_for_each(sk2, node, head)
			if (sk2->sk_hash == snum                             &&
			    sk2 != sk                                        &&
			    sk2->sk_net == net				     &&
			    (!sk2->sk_reuse        || !sk->sk_reuse)         &&
			    (!sk2->sk_bound_dev_if || !sk->sk_bound_dev_if
			     || sk2->sk_bound_dev_if == sk->sk_bound_dev_if) &&
			    (*saddr_comp)(sk, sk2)                             )
				goto fail;
	}

	inet_sk(sk)->num = snum;
	sk->sk_hash = snum;
	if (sk_unhashed(sk)) {
		head = &udptable[snum & (UDP_HTABLE_SIZE - 1)];
		sk_add_node(sk, head);
		sock_prot_inuse_add(sk->sk_prot, 1);
	}
	error = 0;
fail:
	write_unlock_bh(&udp_hash_lock);
	return error;
}

int udp_get_port(struct sock *sk, unsigned short snum,
			int (*scmp)(const struct sock *, const struct sock *))
{
	return  __udp_lib_get_port(sk, snum, udp_hash, scmp);
}

/*
 *	IOCTL requests applicable to the UDP protocol
 */

int udp_ioctl(struct sock *sk, int cmd, unsigned long arg)
{
	switch (cmd) {
	case SIOCOUTQ:
	{
		int amount = atomic_read(