[linux-2.6-block.git] / net / rxrpc / peer_event.c

/* Peer event handling, typically ICMP messages.
 *
 * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
 * Written by David Howells (dhowells@redhat.com)
 *
 * 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 <linux/module.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <linux/errqueue.h>
#include <linux/udp.h>
#include <linux/in.h>
#include <linux/in6.h>
#include <linux/icmp.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include <net/ip.h>
#include "ar-internal.h"

static void rxrpc_store_error(struct rxrpc_peer *, struct sock_exterr_skb *);

/*
 * Find the peer associated with an ICMP packet.
 */
static struct rxrpc_peer *rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local *local,
						     const struct sk_buff *skb)
{
	struct sock_exterr_skb *serr = SKB_EXT_ERR(skb);
	struct sockaddr_rxrpc srx;

	_enter("");

	memset(&srx, 0, sizeof(srx));
	srx.transport_type = local->srx.transport_type;
	srx.transport.family = local->srx.transport.family;

	/* Can we see an ICMP4 packet on an ICMP6 listening socket?  and vice
	 * versa?
	 */
	switch (srx.transport.family) {
	case AF_INET:
		srx.transport.sin.sin_port = serr->port;
		srx.transport_len = sizeof(struct sockaddr_in);
		switch (serr->ee.ee_origin) {
		case SO_EE_ORIGIN_ICMP:
			_net("Rx ICMP");
			memcpy(&srx.transport.sin.sin_addr,
			       skb_network_header(skb) + serr->addr_offset,
			       sizeof(struct in_addr));
			break;
		case SO_EE_ORIGIN_ICMP6:
			_net("Rx ICMP6 on v4 sock");
			memcpy(&srx.transport.sin.sin_addr,
			       skb_network_header(skb) + serr->addr_offset + 12,
			       sizeof(struct in_addr));
			break;
		default:
			memcpy(&srx.transport.sin.sin_addr, &ip_hdr(skb)->saddr,
			       sizeof(struct in_addr));
			break;
		}
		break;

	default:
		BUG();
	}

	return rxrpc_lookup_peer_rcu(local, &srx);
}

/*
 * Handle an MTU/fragmentation problem.
 */
static void rxrpc_adjust_mtu(struct rxrpc_peer *peer, struct sock_exterr_skb *serr)
{
	u32 mtu = serr->ee.ee_info;

	_net("Rx ICMP Fragmentation Needed (%d)", mtu);

	/* wind down the local interface MTU */
	if (mtu > 0 && peer->if_mtu == 65535 && mtu < peer->if_mtu) {
		peer->if_mtu = mtu;
		_net("I/F MTU %u", mtu);
	}

	if (mtu == 0) {
		/* they didn't give us a size, estimate one */
		mtu = peer->if_mtu;
		if (mtu > 1500) {
			mtu >>= 1;
			if (mtu < 1500)
				mtu = 1500;
		} else {
			mtu -= 100;
			if (mtu < peer->hdrsize)
				mtu = peer->hdrsize + 4;
		}
	}

	if (mtu < peer->mtu) {
		spin_lock_bh(&peer->lock);
		peer->mtu = mtu;
		peer->maxdata = peer->mtu - peer->hdrsize;
		spin_unlock_bh(&peer->lock);
		_net("Net MTU %u (maxdata %u)",
		     peer->mtu, peer->maxdata);
	}
}

/*
 * Handle an error received on the local endpoint.
 */
void rxrpc_error_report(struct sock *sk)
{
	struct sock_exterr_skb *serr;
	struct rxrpc_local *local = sk->sk_user_data;
	struct rxrpc_peer *peer;
	struct sk_buff *skb;

	_enter("%p{%d}", sk, local->debug_id);

	skb = sock_dequeue_err_skb(sk);
	if (!skb) {
		_leave("UDP socket errqueue empty");
		return;
	}
	serr = SKB_EXT_ERR(skb);
	if (!skb->len && serr->ee.ee_origin == SO_EE_ORIGIN_TIMESTAMPING) {
		_leave("UDP empty message");
		kfree_skb(skb);
		return;
	}

	rxrpc_new_skb(skb);

	rcu_read_lock();
	peer = rxrpc_lookup_peer_icmp_rcu(local, skb);
	if (peer && !rxrpc_get_peer_maybe(peer))
		peer = NULL;
	if (!peer) {
		rcu_read_unlock();
		rxrpc_free_skb(skb);
		_leave(" [no peer]");
		return;
	}

	if ((serr->ee.ee_origin == SO_EE_ORIGIN_ICMP &&
	     serr->ee.ee_type == ICMP_DEST_UNREACH &&
	     serr->ee.ee_code == ICMP_FRAG_NEEDED)) {
		rxrpc_adjust_mtu(peer, serr);
		rcu_read_unlock();
		rxrpc_free_skb(skb);
		rxrpc_put_peer(peer);
		_leave(" [MTU update]");
		return;
	}

	rxrpc_store_error(peer, serr);
	rcu_read_unlock();
	rxrpc_free_skb(skb);

	/* The ref we obtained is passed off to the work item */
	rxrpc_queue_work(&peer->error_distributor);
	_leave("");
}

/*
 * Map an error report to error codes on the peer record.
 */
static void rxrpc_store_error(struct rxrpc_peer *peer,
			      struct sock_exterr_skb *serr)
{
	struct sock_extended_err *ee;
	int err;

	_enter("");

	ee = &serr->ee;

	_net("Rx Error o=%d t=%d c=%d e=%d",
	     ee->ee_origin, ee->ee_type, ee->ee_code, ee->ee_errno);

	err = ee->ee_errno;

	switch (ee->ee_origin) {
	case SO_EE_ORIGIN_ICMP:
		switch (ee->ee_type) {
		case ICMP_DEST_UNREACH:
			switch (ee->ee_code) {
			case ICMP_NET_UNREACH:
				_net("Rx Received ICMP Network Unreachable");
				break;
			case ICMP_HOST_UNREACH:
				_net("Rx Received ICMP Host Unreachable");
				break;
			case ICMP_PORT_UNREACH:
				_net("Rx Received ICMP Port Unreachable");
				break;
			case ICMP_NET_UNKNOWN:
				_net("Rx Received ICMP Unknown Network");
				break;
			case ICMP_HOST_UNKNOWN:
				_net("Rx Received ICMP Unknown Host");
				break;
			default:
				_net("Rx Received ICMP DestUnreach code=%u",
				     ee->ee_code);
				break;
			}
			break;

		case ICMP_TIME_EXCEEDED:
			_net("Rx Received ICMP TTL Exceeded");
			break;

		default:
			_proto("Rx Received ICMP error { type=%u code=%u }",
			       ee->ee_type, ee->ee_code);
			break;
		}
		break;

	case SO_EE_ORIGIN_NONE:
	case SO_EE_ORIGIN_LOCAL:
		_proto("Rx Received local error { error=%d }", err);
		err += RXRPC_LOCAL_ERROR_OFFSET;
		break;

	case SO_EE_ORIGIN_ICMP6:
	default:
		_proto("Rx Received error report { orig=%u }", ee->ee_origin);
		break;
	}

	peer->error_report = err;
}

/*
 * Distribute an error that occurred on a peer
 */
void rxrpc_peer_error_distributor(struct work_struct *work)
{
	struct rxrpc_peer *peer =
		container_of(work, struct rxrpc_peer, error_distributor);
	struct rxrpc_call *call;
	int error_report;

	_enter("");

	error_report = READ_ONCE(peer->error_report);

	_debug("ISSUE ERROR %d", error_report);

	spin_lock_bh(&peer->lock);

	while (!hlist_empty(&peer->error_targets)) {
		call = hlist_entry(peer->error_targets.first,
				   struct rxrpc_call, error_link);
		hlist_del_init(&call->error_link);

		write_lock(&call->state_lock);
		if (call->state != RXRPC_CALL_COMPLETE &&
		    call->state < RXRPC_CALL_NETWORK_ERROR) {
			call->error_report = error_report;
			call->state = RXRPC_CALL_NETWORK_ERROR;
			set_bit(RXRPC_CALL_EV_RCVD_ERROR, &call->events);
			rxrpc_queue_call(call);
		}
		write_unlock(&call->state_lock);
	}

	spin_unlock_bh(&peer->lock);

	rxrpc_put_peer(peer);
	_leave("");
}
Commit	Line	Data
f66d7490	1	/* Peer event handling, typically ICMP messages.
17926a79 DH	2	*
	3	* Copyright (C) 2007 Red Hat, Inc. All Rights Reserved.
	4	* Written by David Howells (dhowells@redhat.com)
	5	*
	6	* This program is free software; you can redistribute it and/or
	7	* modify it under the terms of the GNU General Public License
	8	* as published by the Free Software Foundation; either version
	9	* 2 of the License, or (at your option) any later version.
	10	*/
	11
	12	#include <linux/module.h>
	13	#include <linux/net.h>
	14	#include <linux/skbuff.h>
	15	#include <linux/errqueue.h>
	16	#include <linux/udp.h>
	17	#include <linux/in.h>
	18	#include <linux/in6.h>
	19	#include <linux/icmp.h>
	20	#include <net/sock.h>
	21	#include <net/af_rxrpc.h>
	22	#include <net/ip.h>
	23	#include "ar-internal.h"
	24
f66d7490 DH	25	static void rxrpc_store_error(struct rxrpc_peer , struct sock_exterr_skb );
f66d7490 DH	26
be6e6707 DH	27	/*
	28	* Find the peer associated with an ICMP packet.
	29	*/
	30	static struct rxrpc_peer rxrpc_lookup_peer_icmp_rcu(struct rxrpc_local local,
	31	const struct sk_buff *skb)
	32	{
	33	struct sock_exterr_skb *serr = SKB_EXT_ERR(skb);
	34	struct sockaddr_rxrpc srx;
	35
	36	_enter("");
	37
	38	memset(&srx, 0, sizeof(srx));
	39	srx.transport_type = local->srx.transport_type;
	40	srx.transport.family = local->srx.transport.family;
	41
	42	/* Can we see an ICMP4 packet on an ICMP6 listening socket? and vice
	43	* versa?
	44	*/
	45	switch (srx.transport.family) {
	46	case AF_INET:
	47	srx.transport.sin.sin_port = serr->port;
	48	srx.transport_len = sizeof(struct sockaddr_in);
	49	switch (serr->ee.ee_origin) {
	50	case SO_EE_ORIGIN_ICMP:
	51	_net("Rx ICMP");
	52	memcpy(&srx.transport.sin.sin_addr,
	53	skb_network_header(skb) + serr->addr_offset,
	54	sizeof(struct in_addr));
	55	break;
	56	case SO_EE_ORIGIN_ICMP6:
	57	_net("Rx ICMP6 on v4 sock");
	58	memcpy(&srx.transport.sin.sin_addr,
	59	skb_network_header(skb) + serr->addr_offset + 12,
	60	sizeof(struct in_addr));
	61	break;
	62	default:
	63	memcpy(&srx.transport.sin.sin_addr, &ip_hdr(skb)->saddr,
	64	sizeof(struct in_addr));
	65	break;
	66	}
	67	break;
	68
	69	default:
	70	BUG();
	71	}
	72
	73	return rxrpc_lookup_peer_rcu(local, &srx);
	74	}
	75
1a70c05b DH	76	/*
	77	* Handle an MTU/fragmentation problem.
	78	*/
	79	static void rxrpc_adjust_mtu(struct rxrpc_peer peer, struct sock_exterr_skb serr)
	80	{
	81	u32 mtu = serr->ee.ee_info;
	82
	83	_net("Rx ICMP Fragmentation Needed (%d)", mtu);
	84
	85	/* wind down the local interface MTU */
	86	if (mtu > 0 && peer->if_mtu == 65535 && mtu < peer->if_mtu) {
	87	peer->if_mtu = mtu;
	88	_net("I/F MTU %u", mtu);
	89	}
	90
	91	if (mtu == 0) {
	92	/* they didn't give us a size, estimate one */
	93	mtu = peer->if_mtu;
	94	if (mtu > 1500) {
	95	mtu >>= 1;
	96	if (mtu < 1500)
	97	mtu = 1500;
	98	} else {
	99	mtu -= 100;
	100	if (mtu < peer->hdrsize)
	101	mtu = peer->hdrsize + 4;
	102	}
	103	}
	104
	105	if (mtu < peer->mtu) {
	106	spin_lock_bh(&peer->lock);
	107	peer->mtu = mtu;
	108	peer->maxdata = peer->mtu - peer->hdrsize;
	109	spin_unlock_bh(&peer->lock);
	110	_net("Net MTU %u (maxdata %u)",
	111	peer->mtu, peer->maxdata);
	112	}
	113	}
	114
17926a79	115	/*
f66d7490	116	* Handle an error received on the local endpoint.
17926a79	117	*/
abe89ef0	118	void rxrpc_error_report(struct sock *sk)
17926a79 DH	119	{
17926a79 DH	120	struct sock_exterr_skb *serr;
17926a79 DH	121	struct rxrpc_local *local = sk->sk_user_data;
	122	struct rxrpc_peer *peer;
	123	struct sk_buff *skb;
17926a79 DH	124
	125	_enter("%p{%d}", sk, local->debug_id);
	126
364a9e93	127	skb = sock_dequeue_err_skb(sk);
17926a79 DH	128	if (!skb) {
	129	_leave("UDP socket errqueue empty");
	130	return;
	131	}
c247f053 WB	132	serr = SKB_EXT_ERR(skb);
c247f053 WB	133	if (!skb->len && serr->ee.ee_origin == SO_EE_ORIGIN_TIMESTAMPING) {
49ca0d8b WB	134	_leave("UDP empty message");
	135	kfree_skb(skb);
	136	return;
	137	}
17926a79 DH	138
	139	rxrpc_new_skb(skb);
	140
be6e6707 DH	141	rcu_read_lock();
	142	peer = rxrpc_lookup_peer_icmp_rcu(local, skb);
	143	if (peer && !rxrpc_get_peer_maybe(peer))
	144	peer = NULL;
	145	if (!peer) {
	146	rcu_read_unlock();
17926a79 DH	147	rxrpc_free_skb(skb);
	148	_leave(" [no peer]");
	149	return;
	150	}
	151
1a70c05b DH	152	if ((serr->ee.ee_origin == SO_EE_ORIGIN_ICMP &&
	153	serr->ee.ee_type == ICMP_DEST_UNREACH &&
	154	serr->ee.ee_code == ICMP_FRAG_NEEDED)) {
	155	rxrpc_adjust_mtu(peer, serr);
f66d7490	156	rcu_read_unlock();
1a70c05b	157	rxrpc_free_skb(skb);
f66d7490 DH	158	rxrpc_put_peer(peer);
	159	_leave(" [MTU update]");
	160	return;
17926a79 DH	161	}
17926a79 DH	162
f66d7490	163	rxrpc_store_error(peer, serr);
be6e6707	164	rcu_read_unlock();
f66d7490	165	rxrpc_free_skb(skb);
17926a79	166
f66d7490 DH	167	/* The ref we obtained is passed off to the work item */
f66d7490 DH	168	rxrpc_queue_work(&peer->error_distributor);
17926a79 DH	169	_leave("");
	170	}
	171
	172	/*
f66d7490	173	* Map an error report to error codes on the peer record.
17926a79	174	*/
f66d7490 DH	175	static void rxrpc_store_error(struct rxrpc_peer *peer,
f66d7490 DH	176	struct sock_exterr_skb *serr)
17926a79 DH	177	{
17926a79 DH	178	struct sock_extended_err *ee;
c9d10c49	179	int err;
17926a79 DH	180
	181	_enter("");
	182
17926a79 DH	183	ee = &serr->ee;
	184
	185	_net("Rx Error o=%d t=%d c=%d e=%d",
	186	ee->ee_origin, ee->ee_type, ee->ee_code, ee->ee_errno);
	187
	188	err = ee->ee_errno;
	189
	190	switch (ee->ee_origin) {
	191	case SO_EE_ORIGIN_ICMP:
17926a79 DH	192	switch (ee->ee_type) {
	193	case ICMP_DEST_UNREACH:
	194	switch (ee->ee_code) {
	195	case ICMP_NET_UNREACH:
	196	_net("Rx Received ICMP Network Unreachable");
17926a79 DH	197	break;
	198	case ICMP_HOST_UNREACH:
	199	_net("Rx Received ICMP Host Unreachable");
17926a79 DH	200	break;
	201	case ICMP_PORT_UNREACH:
	202	_net("Rx Received ICMP Port Unreachable");
17926a79 DH	203	break;
	204	case ICMP_NET_UNKNOWN:
	205	_net("Rx Received ICMP Unknown Network");
17926a79 DH	206	break;
	207	case ICMP_HOST_UNKNOWN:
	208	_net("Rx Received ICMP Unknown Host");
17926a79 DH	209	break;
	210	default:
	211	_net("Rx Received ICMP DestUnreach code=%u",
	212	ee->ee_code);
	213	break;
	214	}
	215	break;
	216
	217	case ICMP_TIME_EXCEEDED:
	218	_net("Rx Received ICMP TTL Exceeded");
	219	break;
	220
	221	default:
	222	_proto("Rx Received ICMP error { type=%u code=%u }",
	223	ee->ee_type, ee->ee_code);
	224	break;
	225	}
	226	break;
	227
f66d7490	228	case SO_EE_ORIGIN_NONE:
17926a79	229	case SO_EE_ORIGIN_LOCAL:
fe77d5fc	230	_proto("Rx Received local error { error=%d }", err);
f66d7490	231	err += RXRPC_LOCAL_ERROR_OFFSET;
17926a79 DH	232	break;
17926a79 DH	233
17926a79 DH	234	case SO_EE_ORIGIN_ICMP6:
17926a79 DH	235	default:
fe77d5fc	236	_proto("Rx Received error report { orig=%u }", ee->ee_origin);
17926a79 DH	237	break;
	238	}
	239
f66d7490 DH	240	peer->error_report = err;
	241	}
	242
	243	/*
	244	* Distribute an error that occurred on a peer
	245	*/
	246	void rxrpc_peer_error_distributor(struct work_struct *work)
	247	{
	248	struct rxrpc_peer *peer =
	249	container_of(work, struct rxrpc_peer, error_distributor);
	250	struct rxrpc_call *call;
	251	int error_report;
	252
	253	_enter("");
17926a79	254
f66d7490	255	error_report = READ_ONCE(peer->error_report);
17926a79	256
f66d7490	257	_debug("ISSUE ERROR %d", error_report);
17926a79	258
f66d7490	259	spin_lock_bh(&peer->lock);
17926a79	260
f66d7490 DH	261	while (!hlist_empty(&peer->error_targets)) {
	262	call = hlist_entry(peer->error_targets.first,
	263	struct rxrpc_call, error_link);
	264	hlist_del_init(&call->error_link);
	265
	266	write_lock(&call->state_lock);
	267	if (call->state != RXRPC_CALL_COMPLETE &&
	268	call->state < RXRPC_CALL_NETWORK_ERROR) {
	269	call->error_report = error_report;
	270	call->state = RXRPC_CALL_NETWORK_ERROR;
	271	set_bit(RXRPC_CALL_EV_RCVD_ERROR, &call->events);
	272	rxrpc_queue_call(call);
	273	}
	274	write_unlock(&call->state_lock);
17926a79 DH	275	}
17926a79 DH	276
f66d7490	277	spin_unlock_bh(&peer->lock);
17926a79	278
f66d7490	279	rxrpc_put_peer(peer);
17926a79 DH	280	_leave("");
17926a79 DH	281	}