[mirror_ubuntu-jammy-kernel.git] / net / rxrpc / output.c

/* RxRPC packet transmission
 *
 * 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.
 */

#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt

#include <linux/net.h>
#include <linux/gfp.h>
#include <linux/skbuff.h>
#include <linux/export.h>
#include <net/sock.h>
#include <net/af_rxrpc.h>
#include "ar-internal.h"

struct rxrpc_pkt_buffer {
	struct rxrpc_wire_header whdr;
	union {
		struct {
			struct rxrpc_ackpacket ack;
			u8 acks[255];
			u8 pad[3];
		};
		__be32 abort_code;
	};
	struct rxrpc_ackinfo ackinfo;
};

/*
 * Fill out an ACK packet.
 */
static size_t rxrpc_fill_out_ack(struct rxrpc_call *call,
				 struct rxrpc_pkt_buffer *pkt)
{
	rxrpc_serial_t serial;
	rxrpc_seq_t hard_ack, top, seq;
	int ix;
	u32 mtu, jmax;
	u8 *ackp = pkt->acks;

	/* Barrier against rxrpc_input_data(). */
	serial = call->ackr_serial;
	hard_ack = READ_ONCE(call->rx_hard_ack);
	top = smp_load_acquire(&call->rx_top);

	pkt->ack.bufferSpace	= htons(8);
	pkt->ack.maxSkew	= htons(call->ackr_skew);
	pkt->ack.firstPacket	= htonl(hard_ack + 1);
	pkt->ack.previousPacket	= htonl(call->ackr_prev_seq);
	pkt->ack.serial		= htonl(serial);
	pkt->ack.reason		= call->ackr_reason;
	pkt->ack.nAcks		= top - hard_ack;

	if (pkt->ack.reason == RXRPC_ACK_PING)
		pkt->whdr.flags |= RXRPC_REQUEST_ACK;

	if (after(top, hard_ack)) {
		seq = hard_ack + 1;
		do {
			ix = seq & RXRPC_RXTX_BUFF_MASK;
			if (call->rxtx_buffer[ix])
				*ackp++ = RXRPC_ACK_TYPE_ACK;
			else
				*ackp++ = RXRPC_ACK_TYPE_NACK;
			seq++;
		} while (before_eq(seq, top));
	}

	mtu = call->conn->params.peer->if_mtu;
	mtu -= call->conn->params.peer->hdrsize;
	jmax = (call->nr_jumbo_bad > 3) ? 1 : rxrpc_rx_jumbo_max;
	pkt->ackinfo.rxMTU	= htonl(rxrpc_rx_mtu);
	pkt->ackinfo.maxMTU	= htonl(mtu);
	pkt->ackinfo.rwind	= htonl(call->rx_winsize);
	pkt->ackinfo.jumbo_max	= htonl(jmax);

	*ackp++ = 0;
	*ackp++ = 0;
	*ackp++ = 0;
	return top - hard_ack + 3;
}

/*
 * Send an ACK or ABORT call packet.
 */
int rxrpc_send_call_packet(struct rxrpc_call *call, u8 type)
{
	struct rxrpc_connection *conn = NULL;
	struct rxrpc_pkt_buffer *pkt;
	struct msghdr msg;
	struct kvec iov[2];
	rxrpc_serial_t serial;
	size_t len, n;
	bool ping = false;
	int ioc, ret;
	u32 abort_code;

	_enter("%u,%s", call->debug_id, rxrpc_pkts[type]);

	spin_lock_bh(&call->lock);
	if (call->conn)
		conn = rxrpc_get_connection_maybe(call->conn);
	spin_unlock_bh(&call->lock);
	if (!conn)
		return -ECONNRESET;

	pkt = kzalloc(sizeof(*pkt), GFP_KERNEL);
	if (!pkt) {
		rxrpc_put_connection(conn);
		return -ENOMEM;
	}

	msg.msg_name	= &call->peer->srx.transport;
	msg.msg_namelen	= call->peer->srx.transport_len;
	msg.msg_control	= NULL;
	msg.msg_controllen = 0;
	msg.msg_flags	= 0;

	pkt->whdr.epoch		= htonl(conn->proto.epoch);
	pkt->whdr.cid		= htonl(call->cid);
	pkt->whdr.callNumber	= htonl(call->call_id);
	pkt->whdr.seq		= 0;
	pkt->whdr.type		= type;
	pkt->whdr.flags		= conn->out_clientflag;
	pkt->whdr.userStatus	= 0;
	pkt->whdr.securityIndex	= call->security_ix;
	pkt->whdr._rsvd		= 0;
	pkt->whdr.serviceId	= htons(call->service_id);

	iov[0].iov_base	= pkt;
	iov[0].iov_len	= sizeof(pkt->whdr);
	len = sizeof(pkt->whdr);

	switch (type) {
	case RXRPC_PACKET_TYPE_ACK:
		spin_lock_bh(&call->lock);
		if (!call->ackr_reason) {
			spin_unlock_bh(&call->lock);
			ret = 0;
			goto out;
		}
		ping = (call->ackr_reason == RXRPC_ACK_PING);
		n = rxrpc_fill_out_ack(call, pkt);
		call->ackr_reason = 0;

		spin_unlock_bh(&call->lock);


		iov[0].iov_len += sizeof(pkt->ack) + n;
		iov[1].iov_base = &pkt->ackinfo;
		iov[1].iov_len	= sizeof(pkt->ackinfo);
		len += sizeof(pkt->ack) + n + sizeof(pkt->ackinfo);
		ioc = 2;
		break;

	case RXRPC_PACKET_TYPE_ABORT:
		abort_code = call->abort_code;
		pkt->abort_code = htonl(abort_code);
		iov[0].iov_len += sizeof(pkt->abort_code);
		len += sizeof(pkt->abort_code);
		ioc = 1;
		break;

	default:
		BUG();
		ret = -ENOANO;
		goto out;
	}

	serial = atomic_inc_return(&conn->serial);
	pkt->whdr.serial = htonl(serial);
	switch (type) {
	case RXRPC_PACKET_TYPE_ACK:
		trace_rxrpc_tx_ack(call, serial,
				   ntohl(pkt->ack.firstPacket),
				   ntohl(pkt->ack.serial),
				   pkt->ack.reason, pkt->ack.nAcks);
		break;
	}

	if (ping) {
		call->ackr_ping = serial;
		smp_wmb();
		/* We need to stick a time in before we send the packet in case
		 * the reply gets back before kernel_sendmsg() completes - but
		 * asking UDP to send the packet can take a relatively long
		 * time, so we update the time after, on the assumption that
		 * the packet transmission is more likely to happen towards the
		 * end of the kernel_sendmsg() call.
		 */
		call->ackr_ping_time = ktime_get_real();
		set_bit(RXRPC_CALL_PINGING, &call->flags);
		trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_ping, serial);
	}
	ret = kernel_sendmsg(conn->params.local->socket,
			     &msg, iov, ioc, len);
	if (ping)
		call->ackr_ping_time = ktime_get_real();

	if (ret < 0 && call->state < RXRPC_CALL_COMPLETE) {
		switch (type) {
		case RXRPC_PACKET_TYPE_ACK:
			clear_bit(RXRPC_CALL_PINGING, &call->flags);
			rxrpc_propose_ACK(call, pkt->ack.reason,
					  ntohs(pkt->ack.maxSkew),
					  ntohl(pkt->ack.serial),
					  true, true,
					  rxrpc_propose_ack_retry_tx);
			break;
		case RXRPC_PACKET_TYPE_ABORT:
			break;
		}
	}

out:
	rxrpc_put_connection(conn);
	kfree(pkt);
	return ret;
}

/*
 * send a packet through the transport endpoint
 */
int rxrpc_send_data_packet(struct rxrpc_call *call, struct sk_buff *skb)
{
	struct rxrpc_connection *conn = call->conn;
	struct rxrpc_wire_header whdr;
	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
	struct msghdr msg;
	struct kvec iov[2];
	rxrpc_serial_t serial;
	size_t len;
	int ret, opt;

	_enter(",{%d}", skb->len);

	/* Each transmission of a Tx packet needs a new serial number */
	serial = atomic_inc_return(&conn->serial);

	whdr.epoch	= htonl(conn->proto.epoch);
	whdr.cid	= htonl(call->cid);
	whdr.callNumber	= htonl(call->call_id);
	whdr.seq	= htonl(sp->hdr.seq);
	whdr.serial	= htonl(serial);
	whdr.type	= RXRPC_PACKET_TYPE_DATA;
	whdr.flags	= sp->hdr.flags;
	whdr.userStatus	= 0;
	whdr.securityIndex = call->security_ix;
	whdr._rsvd	= htons(sp->hdr._rsvd);
	whdr.serviceId	= htons(call->service_id);

	iov[0].iov_base = &whdr;
	iov[0].iov_len = sizeof(whdr);
	iov[1].iov_base = skb->head;
	iov[1].iov_len = skb->len;
	len = iov[0].iov_len + iov[1].iov_len;

	msg.msg_name = &call->peer->srx.transport;
	msg.msg_namelen = call->peer->srx.transport_len;
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_flags = 0;

	/* If our RTT cache needs working on, request an ACK. */
	if ((call->peer->rtt_usage < 3 && sp->hdr.seq & 1) ||
	    ktime_before(ktime_add_ms(call->peer->rtt_last_req, 1000),
			 ktime_get_real()))
		whdr.flags |= RXRPC_REQUEST_ACK;

	if (IS_ENABLED(CONFIG_AF_RXRPC_INJECT_LOSS)) {
		static int lose;
		if ((lose++ & 7) == 7) {
			trace_rxrpc_tx_data(call, sp->hdr.seq, serial,
					    whdr.flags, true);
			rxrpc_lose_skb(skb, rxrpc_skb_tx_lost);
			_leave(" = 0 [lose]");
			return 0;
		}
	}

	_proto("Tx DATA %%%u { #%u }", serial, sp->hdr.seq);

	/* send the packet with the don't fragment bit set if we currently
	 * think it's small enough */
	if (iov[1].iov_len >= call->peer->maxdata)
		goto send_fragmentable;

	down_read(&conn->params.local->defrag_sem);
	/* send the packet by UDP
	 * - returns -EMSGSIZE if UDP would have to fragment the packet
	 *   to go out of the interface
	 *   - in which case, we'll have processed the ICMP error
	 *     message and update the peer record
	 */
	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);

	up_read(&conn->params.local->defrag_sem);
	if (ret == -EMSGSIZE)
		goto send_fragmentable;

done:
	trace_rxrpc_tx_data(call, sp->hdr.seq, serial, whdr.flags, false);
	if (ret >= 0) {
		ktime_t now = ktime_get_real();
		skb->tstamp = now;
		smp_wmb();
		sp->hdr.serial = serial;
		if (whdr.flags & RXRPC_REQUEST_ACK) {
			call->peer->rtt_last_req = now;
			trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_data, serial);
		}
	}
	_leave(" = %d [%u]", ret, call->peer->maxdata);
	return ret;

send_fragmentable:
	/* attempt to send this message with fragmentation enabled */
	_debug("send fragment");

	down_write(&conn->params.local->defrag_sem);

	switch (conn->params.local->srx.transport.family) {
	case AF_INET:
		opt = IP_PMTUDISC_DONT;
		ret = kernel_setsockopt(conn->params.local->socket,
					SOL_IP, IP_MTU_DISCOVER,
					(char *)&opt, sizeof(opt));
		if (ret == 0) {
			ret = kernel_sendmsg(conn->params.local->socket, &msg,
					     iov, 2, len);

			opt = IP_PMTUDISC_DO;
			kernel_setsockopt(conn->params.local->socket, SOL_IP,
					  IP_MTU_DISCOVER,
					  (char *)&opt, sizeof(opt));
		}
		break;

#ifdef CONFIG_AF_RXRPC_IPV6
	case AF_INET6:
		opt = IPV6_PMTUDISC_DONT;
		ret = kernel_setsockopt(conn->params.local->socket,
					SOL_IPV6, IPV6_MTU_DISCOVER,
					(char *)&opt, sizeof(opt));
		if (ret == 0) {
			ret = kernel_sendmsg(conn->params.local->socket, &msg,
					     iov, 1, iov[0].iov_len);

			opt = IPV6_PMTUDISC_DO;
			kernel_setsockopt(conn->params.local->socket,
					  SOL_IPV6, IPV6_MTU_DISCOVER,
					  (char *)&opt, sizeof(opt));
		}
		break;
#endif
	}

	up_write(&conn->params.local->defrag_sem);
	goto done;
}

/*
 * reject packets through the local endpoint
 */
void rxrpc_reject_packets(struct rxrpc_local *local)
{
	struct sockaddr_rxrpc srx;
	struct rxrpc_skb_priv *sp;
	struct rxrpc_wire_header whdr;
	struct sk_buff *skb;
	struct msghdr msg;
	struct kvec iov[2];
	size_t size;
	__be32 code;

	_enter("%d", local->debug_id);

	iov[0].iov_base = &whdr;
	iov[0].iov_len = sizeof(whdr);
	iov[1].iov_base = &code;
	iov[1].iov_len = sizeof(code);
	size = sizeof(whdr) + sizeof(code);

	msg.msg_name = &srx.transport;
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_flags = 0;

	memset(&whdr, 0, sizeof(whdr));
	whdr.type = RXRPC_PACKET_TYPE_ABORT;

	while ((skb = skb_dequeue(&local->reject_queue))) {
		rxrpc_see_skb(skb, rxrpc_skb_rx_seen);
		sp = rxrpc_skb(skb);

		if (rxrpc_extract_addr_from_skb(&srx, skb) == 0) {
			msg.msg_namelen = srx.transport_len;

			code = htonl(skb->priority);

			whdr.epoch	= htonl(sp->hdr.epoch);
			whdr.cid	= htonl(sp->hdr.cid);
			whdr.callNumber	= htonl(sp->hdr.callNumber);
			whdr.serviceId	= htons(sp->hdr.serviceId);
			whdr.flags	= sp->hdr.flags;
			whdr.flags	^= RXRPC_CLIENT_INITIATED;
			whdr.flags	&= RXRPC_CLIENT_INITIATED;

			kernel_sendmsg(local->socket, &msg, iov, 2, size);
		}

		rxrpc_free_skb(skb, rxrpc_skb_rx_freed);
	}

	_leave("");
}
Commit	Line	Data
17926a79 DH	1	/* RxRPC packet transmission
	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
9b6d5398 JP	12	#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
9b6d5398 JP	13
17926a79	14	#include <linux/net.h>
5a0e3ad6	15	#include <linux/gfp.h>
17926a79	16	#include <linux/skbuff.h>
bc3b2d7f	17	#include <linux/export.h>
17926a79 DH	18	#include <net/sock.h>
	19	#include <net/af_rxrpc.h>
	20	#include "ar-internal.h"
	21
8d94aa38 DH	22	struct rxrpc_pkt_buffer {
	23	struct rxrpc_wire_header whdr;
	24	union {
	25	struct {
	26	struct rxrpc_ackpacket ack;
	27	u8 acks[255];
	28	u8 pad[3];
	29	};
	30	__be32 abort_code;
	31	};
	32	struct rxrpc_ackinfo ackinfo;
	33	};
	34
	35	/*
	36	* Fill out an ACK packet.
	37	*/
	38	static size_t rxrpc_fill_out_ack(struct rxrpc_call *call,
	39	struct rxrpc_pkt_buffer *pkt)
	40	{
f3639df2	41	rxrpc_serial_t serial;
248f219c DH	42	rxrpc_seq_t hard_ack, top, seq;
248f219c DH	43	int ix;
8d94aa38 DH	44	u32 mtu, jmax;
	45	u8 *ackp = pkt->acks;
	46
248f219c	47	/* Barrier against rxrpc_input_data(). */
f3639df2	48	serial = call->ackr_serial;
248f219c DH	49	hard_ack = READ_ONCE(call->rx_hard_ack);
	50	top = smp_load_acquire(&call->rx_top);
	51
8d94aa38	52	pkt->ack.bufferSpace = htons(8);
248f219c DH	53	pkt->ack.maxSkew = htons(call->ackr_skew);
248f219c DH	54	pkt->ack.firstPacket = htonl(hard_ack + 1);
8d94aa38	55	pkt->ack.previousPacket = htonl(call->ackr_prev_seq);
f3639df2	56	pkt->ack.serial = htonl(serial);
248f219c DH	57	pkt->ack.reason = call->ackr_reason;
	58	pkt->ack.nAcks = top - hard_ack;
	59
8e83134d DH	60	if (pkt->ack.reason == RXRPC_ACK_PING)
	61	pkt->whdr.flags \|= RXRPC_REQUEST_ACK;
	62
248f219c DH	63	if (after(top, hard_ack)) {
	64	seq = hard_ack + 1;
	65	do {
	66	ix = seq & RXRPC_RXTX_BUFF_MASK;
	67	if (call->rxtx_buffer[ix])
	68	*ackp++ = RXRPC_ACK_TYPE_ACK;
	69	else
	70	*ackp++ = RXRPC_ACK_TYPE_NACK;
	71	seq++;
	72	} while (before_eq(seq, top));
	73	}
8d94aa38	74
248f219c DH	75	mtu = call->conn->params.peer->if_mtu;
248f219c DH	76	mtu -= call->conn->params.peer->hdrsize;
75e42126	77	jmax = (call->nr_jumbo_bad > 3) ? 1 : rxrpc_rx_jumbo_max;
8d94aa38 DH	78	pkt->ackinfo.rxMTU = htonl(rxrpc_rx_mtu);
8d94aa38 DH	79	pkt->ackinfo.maxMTU = htonl(mtu);
75e42126	80	pkt->ackinfo.rwind = htonl(call->rx_winsize);
8d94aa38 DH	81	pkt->ackinfo.jumbo_max = htonl(jmax);
	82
	83	*ackp++ = 0;
	84	*ackp++ = 0;
	85	*ackp++ = 0;
248f219c	86	return top - hard_ack + 3;
8d94aa38 DH	87	}
	88
	89	/*
248f219c	90	* Send an ACK or ABORT call packet.
8d94aa38 DH	91	*/
	92	int rxrpc_send_call_packet(struct rxrpc_call *call, u8 type)
	93	{
	94	struct rxrpc_connection *conn = NULL;
	95	struct rxrpc_pkt_buffer *pkt;
	96	struct msghdr msg;
	97	struct kvec iov[2];
	98	rxrpc_serial_t serial;
	99	size_t len, n;
8e83134d	100	bool ping = false;
8d94aa38 DH	101	int ioc, ret;
	102	u32 abort_code;
	103
	104	_enter("%u,%s", call->debug_id, rxrpc_pkts[type]);
	105
	106	spin_lock_bh(&call->lock);
	107	if (call->conn)
	108	conn = rxrpc_get_connection_maybe(call->conn);
	109	spin_unlock_bh(&call->lock);
	110	if (!conn)
	111	return -ECONNRESET;
	112
	113	pkt = kzalloc(sizeof(*pkt), GFP_KERNEL);
	114	if (!pkt) {
	115	rxrpc_put_connection(conn);
	116	return -ENOMEM;
	117	}
	118
8d94aa38 DH	119	msg.msg_name = &call->peer->srx.transport;
	120	msg.msg_namelen = call->peer->srx.transport_len;
	121	msg.msg_control = NULL;
	122	msg.msg_controllen = 0;
	123	msg.msg_flags = 0;
	124
	125	pkt->whdr.epoch = htonl(conn->proto.epoch);
	126	pkt->whdr.cid = htonl(call->cid);
	127	pkt->whdr.callNumber = htonl(call->call_id);
	128	pkt->whdr.seq = 0;
8d94aa38 DH	129	pkt->whdr.type = type;
	130	pkt->whdr.flags = conn->out_clientflag;
	131	pkt->whdr.userStatus = 0;
	132	pkt->whdr.securityIndex = call->security_ix;
	133	pkt->whdr._rsvd = 0;
	134	pkt->whdr.serviceId = htons(call->service_id);
	135
	136	iov[0].iov_base = pkt;
	137	iov[0].iov_len = sizeof(pkt->whdr);
	138	len = sizeof(pkt->whdr);
	139
	140	switch (type) {
	141	case RXRPC_PACKET_TYPE_ACK:
	142	spin_lock_bh(&call->lock);
27d0fc43 DH	143	if (!call->ackr_reason) {
	144	spin_unlock_bh(&call->lock);
	145	ret = 0;
	146	goto out;
	147	}
8e83134d	148	ping = (call->ackr_reason == RXRPC_ACK_PING);
8d94aa38 DH	149	n = rxrpc_fill_out_ack(call, pkt);
	150	call->ackr_reason = 0;
	151
	152	spin_unlock_bh(&call->lock);
	153
8d94aa38 DH	154
	155	iov[0].iov_len += sizeof(pkt->ack) + n;
	156	iov[1].iov_base = &pkt->ackinfo;
	157	iov[1].iov_len = sizeof(pkt->ackinfo);
	158	len += sizeof(pkt->ack) + n + sizeof(pkt->ackinfo);
	159	ioc = 2;
	160	break;
	161
	162	case RXRPC_PACKET_TYPE_ABORT:
	163	abort_code = call->abort_code;
	164	pkt->abort_code = htonl(abort_code);
8d94aa38 DH	165	iov[0].iov_len += sizeof(pkt->abort_code);
	166	len += sizeof(pkt->abort_code);
	167	ioc = 1;
	168	break;
	169
	170	default:
	171	BUG();
	172	ret = -ENOANO;
	173	goto out;
	174	}
	175
b86e218e DH	176	serial = atomic_inc_return(&conn->serial);
	177	pkt->whdr.serial = htonl(serial);
	178	switch (type) {
	179	case RXRPC_PACKET_TYPE_ACK:
be832aec	180	trace_rxrpc_tx_ack(call, serial,
b86e218e DH	181	ntohl(pkt->ack.firstPacket),
	182	ntohl(pkt->ack.serial),
	183	pkt->ack.reason, pkt->ack.nAcks);
	184	break;
	185	}
	186
8e83134d DH	187	if (ping) {
	188	call->ackr_ping = serial;
	189	smp_wmb();
	190	/* We need to stick a time in before we send the packet in case
	191	* the reply gets back before kernel_sendmsg() completes - but
	192	* asking UDP to send the packet can take a relatively long
	193	* time, so we update the time after, on the assumption that
	194	* the packet transmission is more likely to happen towards the
	195	* end of the kernel_sendmsg() call.
	196	*/
	197	call->ackr_ping_time = ktime_get_real();
	198	set_bit(RXRPC_CALL_PINGING, &call->flags);
	199	trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_ping, serial);
	200	}
8d94aa38 DH	201	ret = kernel_sendmsg(conn->params.local->socket,
8d94aa38 DH	202	&msg, iov, ioc, len);
8e83134d DH	203	if (ping)
8e83134d DH	204	call->ackr_ping_time = ktime_get_real();
8d94aa38	205
248f219c	206	if (ret < 0 && call->state < RXRPC_CALL_COMPLETE) {
2311e327	207	switch (type) {
248f219c	208	case RXRPC_PACKET_TYPE_ACK:
8e83134d	209	clear_bit(RXRPC_CALL_PINGING, &call->flags);
248f219c DH	210	rxrpc_propose_ACK(call, pkt->ack.reason,
	211	ntohs(pkt->ack.maxSkew),
	212	ntohl(pkt->ack.serial),
9c7ad434 DH	213	true, true,
9c7ad434 DH	214	rxrpc_propose_ack_retry_tx);
248f219c DH	215	break;
	216	case RXRPC_PACKET_TYPE_ABORT:
	217	break;
	218	}
	219	}
	220
8d94aa38 DH	221	out:
	222	rxrpc_put_connection(conn);
	223	kfree(pkt);
	224	return ret;
	225	}
	226
17926a79 DH	227	/*
	228	* send a packet through the transport endpoint
	229	*/
5a924b89	230	int rxrpc_send_data_packet(struct rxrpc_call call, struct sk_buff skb)
17926a79	231	{
5a924b89 DH	232	struct rxrpc_connection *conn = call->conn;
	233	struct rxrpc_wire_header whdr;
	234	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
17926a79	235	struct msghdr msg;
5a924b89 DH	236	struct kvec iov[2];
	237	rxrpc_serial_t serial;
	238	size_t len;
17926a79 DH	239	int ret, opt;
	240
	241	_enter(",{%d}", skb->len);
	242
5a924b89 DH	243	/* Each transmission of a Tx packet needs a new serial number */
5a924b89 DH	244	serial = atomic_inc_return(&conn->serial);
17926a79	245
5a924b89 DH	246	whdr.epoch = htonl(conn->proto.epoch);
	247	whdr.cid = htonl(call->cid);
	248	whdr.callNumber = htonl(call->call_id);
	249	whdr.seq = htonl(sp->hdr.seq);
	250	whdr.serial = htonl(serial);
	251	whdr.type = RXRPC_PACKET_TYPE_DATA;
	252	whdr.flags = sp->hdr.flags;
	253	whdr.userStatus = 0;
	254	whdr.securityIndex = call->security_ix;
	255	whdr._rsvd = htons(sp->hdr._rsvd);
	256	whdr.serviceId = htons(call->service_id);
	257
	258	iov[0].iov_base = &whdr;
	259	iov[0].iov_len = sizeof(whdr);
	260	iov[1].iov_base = skb->head;
	261	iov[1].iov_len = skb->len;
	262	len = iov[0].iov_len + iov[1].iov_len;
	263
	264	msg.msg_name = &call->peer->srx.transport;
	265	msg.msg_namelen = call->peer->srx.transport_len;
17926a79 DH	266	msg.msg_control = NULL;
	267	msg.msg_controllen = 0;
	268	msg.msg_flags = 0;
	269
0d4b103c DH	270	/* If our RTT cache needs working on, request an ACK. */
	271	if ((call->peer->rtt_usage < 3 && sp->hdr.seq & 1) \|\|
	272	ktime_before(ktime_add_ms(call->peer->rtt_last_req, 1000),
	273	ktime_get_real()))
	274	whdr.flags \|= RXRPC_REQUEST_ACK;
	275
8a681c36 DH	276	if (IS_ENABLED(CONFIG_AF_RXRPC_INJECT_LOSS)) {
	277	static int lose;
	278	if ((lose++ & 7) == 7) {
be832aec DH	279	trace_rxrpc_tx_data(call, sp->hdr.seq, serial,
be832aec DH	280	whdr.flags, true);
8a681c36 DH	281	rxrpc_lose_skb(skb, rxrpc_skb_tx_lost);
	282	_leave(" = 0 [lose]");
	283	return 0;
	284	}
	285	}
	286
5a924b89 DH	287	_proto("Tx DATA %%%u { #%u }", serial, sp->hdr.seq);
5a924b89 DH	288
17926a79 DH	289	/* send the packet with the don't fragment bit set if we currently
17926a79 DH	290	* think it's small enough */
5a924b89 DH	291	if (iov[1].iov_len >= call->peer->maxdata)
	292	goto send_fragmentable;
	293
	294	down_read(&conn->params.local->defrag_sem);
	295	/* send the packet by UDP
	296	* - returns -EMSGSIZE if UDP would have to fragment the packet
	297	* to go out of the interface
	298	* - in which case, we'll have processed the ICMP error
	299	* message and update the peer record
	300	*/
	301	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);
	302
	303	up_read(&conn->params.local->defrag_sem);
	304	if (ret == -EMSGSIZE)
	305	goto send_fragmentable;
	306
	307	done:
be832aec	308	trace_rxrpc_tx_data(call, sp->hdr.seq, serial, whdr.flags, false);
50235c4b	309	if (ret >= 0) {
0d4b103c DH	310	ktime_t now = ktime_get_real();
0d4b103c DH	311	skb->tstamp = now;
50235c4b	312	smp_wmb();
5a924b89	313	sp->hdr.serial = serial;
0d4b103c DH	314	if (whdr.flags & RXRPC_REQUEST_ACK) {
0d4b103c DH	315	call->peer->rtt_last_req = now;
50235c4b	316	trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_data, serial);
0d4b103c	317	}
17926a79	318	}
5a924b89 DH	319	_leave(" = %d [%u]", ret, call->peer->maxdata);
5a924b89 DH	320	return ret;
17926a79 DH	321
	322	send_fragmentable:
	323	/* attempt to send this message with fragmentation enabled */
	324	_debug("send fragment");
	325
985a5c82 DH	326	down_write(&conn->params.local->defrag_sem);
	327
	328	switch (conn->params.local->srx.transport.family) {
	329	case AF_INET:
	330	opt = IP_PMTUDISC_DONT;
	331	ret = kernel_setsockopt(conn->params.local->socket,
	332	SOL_IP, IP_MTU_DISCOVER,
	333	(char *)&opt, sizeof(opt));
	334	if (ret == 0) {
5a924b89 DH	335	ret = kernel_sendmsg(conn->params.local->socket, &msg,
5a924b89 DH	336	iov, 2, len);
985a5c82 DH	337
	338	opt = IP_PMTUDISC_DO;
	339	kernel_setsockopt(conn->params.local->socket, SOL_IP,
	340	IP_MTU_DISCOVER,
	341	(char *)&opt, sizeof(opt));
	342	}
	343	break;
75b54cb5	344
d1912747	345	#ifdef CONFIG_AF_RXRPC_IPV6
75b54cb5 DH	346	case AF_INET6:
	347	opt = IPV6_PMTUDISC_DONT;
	348	ret = kernel_setsockopt(conn->params.local->socket,
	349	SOL_IPV6, IPV6_MTU_DISCOVER,
	350	(char *)&opt, sizeof(opt));
	351	if (ret == 0) {
	352	ret = kernel_sendmsg(conn->params.local->socket, &msg,
	353	iov, 1, iov[0].iov_len);
	354
	355	opt = IPV6_PMTUDISC_DO;
	356	kernel_setsockopt(conn->params.local->socket,
	357	SOL_IPV6, IPV6_MTU_DISCOVER,
	358	(char *)&opt, sizeof(opt));
	359	}
	360	break;
d1912747	361	#endif
17926a79 DH	362	}
17926a79 DH	363
985a5c82	364	up_write(&conn->params.local->defrag_sem);
5a924b89	365	goto done;
17926a79	366	}
248f219c DH	367
	368	/*
	369	* reject packets through the local endpoint
	370	*/
	371	void rxrpc_reject_packets(struct rxrpc_local *local)
	372	{
1c2bc7b9	373	struct sockaddr_rxrpc srx;
248f219c DH	374	struct rxrpc_skb_priv *sp;
	375	struct rxrpc_wire_header whdr;
	376	struct sk_buff *skb;
	377	struct msghdr msg;
	378	struct kvec iov[2];
	379	size_t size;
	380	__be32 code;
	381
	382	_enter("%d", local->debug_id);
	383
	384	iov[0].iov_base = &whdr;
	385	iov[0].iov_len = sizeof(whdr);
	386	iov[1].iov_base = &code;
	387	iov[1].iov_len = sizeof(code);
	388	size = sizeof(whdr) + sizeof(code);
	389
1c2bc7b9	390	msg.msg_name = &srx.transport;
248f219c DH	391	msg.msg_control = NULL;
	392	msg.msg_controllen = 0;
	393	msg.msg_flags = 0;
	394
248f219c DH	395	memset(&whdr, 0, sizeof(whdr));
	396	whdr.type = RXRPC_PACKET_TYPE_ABORT;
	397
	398	while ((skb = skb_dequeue(&local->reject_queue))) {
71f3ca40	399	rxrpc_see_skb(skb, rxrpc_skb_rx_seen);
248f219c	400	sp = rxrpc_skb(skb);
1c2bc7b9 DH	401
	402	if (rxrpc_extract_addr_from_skb(&srx, skb) == 0) {
	403	msg.msg_namelen = srx.transport_len;
	404
248f219c DH	405	code = htonl(skb->priority);
	406
	407	whdr.epoch = htonl(sp->hdr.epoch);
	408	whdr.cid = htonl(sp->hdr.cid);
	409	whdr.callNumber = htonl(sp->hdr.callNumber);
	410	whdr.serviceId = htons(sp->hdr.serviceId);
	411	whdr.flags = sp->hdr.flags;
	412	whdr.flags ^= RXRPC_CLIENT_INITIATED;
	413	whdr.flags &= RXRPC_CLIENT_INITIATED;
	414
	415	kernel_sendmsg(local->socket, &msg, iov, 2, size);
248f219c DH	416	}
248f219c DH	417
71f3ca40	418	rxrpc_free_skb(skb, rxrpc_skb_rx_freed);
248f219c DH	419	}
	420
	421	_leave("");
	422	}