[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_seq_t *_hard_ack,
				 rxrpc_seq_t *_top)
{
	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);
	*_hard_ack = hard_ack;
	*_top = 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;
	rxrpc_seq_t hard_ack, top;
	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, &hard_ack, &top);
		call->ackr_reason = 0;

		spin_unlock_bh(&call->lock);


		pkt->whdr.flags |= RXRPC_SLOW_START_OK;

		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 (type == RXRPC_PACKET_TYPE_ACK &&
	    call->state < RXRPC_CALL_COMPLETE) {
		if (ret < 0) {
			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);
		} else {
			spin_lock_bh(&call->lock);
			if (after(hard_ack, call->ackr_consumed))
				call->ackr_consumed = hard_ack;
			if (after(top, call->ackr_seen))
				call->ackr_seen = top;
			spin_unlock_bh(&call->lock);
		}
	}

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,
			   bool retrans)
{
	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;
	bool lost = false;
	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.  Also request
	 * ACKs if a DATA packet appears to have been lost.
	 */
	if (retrans ||
	    call->cong_mode == RXRPC_CALL_SLOW_START ||
	    (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) {
			ret = 0;
			lost = true;
			goto done;
		}
	}

	_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,
			    retrans, lost);
	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,
805b21b9 DH	39	struct rxrpc_pkt_buffer *pkt,
	40	rxrpc_seq_t *_hard_ack,
	41	rxrpc_seq_t *_top)
8d94aa38	42	{
f3639df2	43	rxrpc_serial_t serial;
248f219c DH	44	rxrpc_seq_t hard_ack, top, seq;
248f219c DH	45	int ix;
8d94aa38 DH	46	u32 mtu, jmax;
	47	u8 *ackp = pkt->acks;
	48
248f219c	49	/* Barrier against rxrpc_input_data(). */
f3639df2	50	serial = call->ackr_serial;
248f219c DH	51	hard_ack = READ_ONCE(call->rx_hard_ack);
248f219c DH	52	top = smp_load_acquire(&call->rx_top);
805b21b9 DH	53	*_hard_ack = hard_ack;
805b21b9 DH	54	*_top = top;
248f219c	55
8d94aa38	56	pkt->ack.bufferSpace = htons(8);
248f219c DH	57	pkt->ack.maxSkew = htons(call->ackr_skew);
248f219c DH	58	pkt->ack.firstPacket = htonl(hard_ack + 1);
8d94aa38	59	pkt->ack.previousPacket = htonl(call->ackr_prev_seq);
f3639df2	60	pkt->ack.serial = htonl(serial);
248f219c DH	61	pkt->ack.reason = call->ackr_reason;
	62	pkt->ack.nAcks = top - hard_ack;
	63
8e83134d DH	64	if (pkt->ack.reason == RXRPC_ACK_PING)
	65	pkt->whdr.flags \|= RXRPC_REQUEST_ACK;
	66
248f219c DH	67	if (after(top, hard_ack)) {
	68	seq = hard_ack + 1;
	69	do {
	70	ix = seq & RXRPC_RXTX_BUFF_MASK;
	71	if (call->rxtx_buffer[ix])
	72	*ackp++ = RXRPC_ACK_TYPE_ACK;
	73	else
	74	*ackp++ = RXRPC_ACK_TYPE_NACK;
	75	seq++;
	76	} while (before_eq(seq, top));
	77	}
8d94aa38	78
248f219c DH	79	mtu = call->conn->params.peer->if_mtu;
248f219c DH	80	mtu -= call->conn->params.peer->hdrsize;
75e42126	81	jmax = (call->nr_jumbo_bad > 3) ? 1 : rxrpc_rx_jumbo_max;
8d94aa38 DH	82	pkt->ackinfo.rxMTU = htonl(rxrpc_rx_mtu);
8d94aa38 DH	83	pkt->ackinfo.maxMTU = htonl(mtu);
75e42126	84	pkt->ackinfo.rwind = htonl(call->rx_winsize);
8d94aa38 DH	85	pkt->ackinfo.jumbo_max = htonl(jmax);
	86
	87	*ackp++ = 0;
	88	*ackp++ = 0;
	89	*ackp++ = 0;
248f219c	90	return top - hard_ack + 3;
8d94aa38 DH	91	}
	92
	93	/*
248f219c	94	* Send an ACK or ABORT call packet.
8d94aa38 DH	95	*/
	96	int rxrpc_send_call_packet(struct rxrpc_call *call, u8 type)
	97	{
	98	struct rxrpc_connection *conn = NULL;
	99	struct rxrpc_pkt_buffer *pkt;
	100	struct msghdr msg;
	101	struct kvec iov[2];
	102	rxrpc_serial_t serial;
805b21b9	103	rxrpc_seq_t hard_ack, top;
8d94aa38	104	size_t len, n;
8e83134d	105	bool ping = false;
8d94aa38 DH	106	int ioc, ret;
	107	u32 abort_code;
	108
	109	_enter("%u,%s", call->debug_id, rxrpc_pkts[type]);
	110
	111	spin_lock_bh(&call->lock);
	112	if (call->conn)
	113	conn = rxrpc_get_connection_maybe(call->conn);
	114	spin_unlock_bh(&call->lock);
	115	if (!conn)
	116	return -ECONNRESET;
	117
	118	pkt = kzalloc(sizeof(*pkt), GFP_KERNEL);
	119	if (!pkt) {
	120	rxrpc_put_connection(conn);
	121	return -ENOMEM;
	122	}
	123
8d94aa38 DH	124	msg.msg_name = &call->peer->srx.transport;
	125	msg.msg_namelen = call->peer->srx.transport_len;
	126	msg.msg_control = NULL;
	127	msg.msg_controllen = 0;
	128	msg.msg_flags = 0;
	129
	130	pkt->whdr.epoch = htonl(conn->proto.epoch);
	131	pkt->whdr.cid = htonl(call->cid);
	132	pkt->whdr.callNumber = htonl(call->call_id);
	133	pkt->whdr.seq = 0;
8d94aa38 DH	134	pkt->whdr.type = type;
	135	pkt->whdr.flags = conn->out_clientflag;
	136	pkt->whdr.userStatus = 0;
	137	pkt->whdr.securityIndex = call->security_ix;
	138	pkt->whdr._rsvd = 0;
	139	pkt->whdr.serviceId = htons(call->service_id);
	140
	141	iov[0].iov_base = pkt;
	142	iov[0].iov_len = sizeof(pkt->whdr);
	143	len = sizeof(pkt->whdr);
	144
	145	switch (type) {
	146	case RXRPC_PACKET_TYPE_ACK:
	147	spin_lock_bh(&call->lock);
27d0fc43 DH	148	if (!call->ackr_reason) {
	149	spin_unlock_bh(&call->lock);
	150	ret = 0;
	151	goto out;
	152	}
8e83134d	153	ping = (call->ackr_reason == RXRPC_ACK_PING);
805b21b9	154	n = rxrpc_fill_out_ack(call, pkt, &hard_ack, &top);
8d94aa38 DH	155	call->ackr_reason = 0;
	156
	157	spin_unlock_bh(&call->lock);
	158
8d94aa38	159
57494343 DH	160	pkt->whdr.flags \|= RXRPC_SLOW_START_OK;
57494343 DH	161
8d94aa38 DH	162	iov[0].iov_len += sizeof(pkt->ack) + n;
	163	iov[1].iov_base = &pkt->ackinfo;
	164	iov[1].iov_len = sizeof(pkt->ackinfo);
	165	len += sizeof(pkt->ack) + n + sizeof(pkt->ackinfo);
	166	ioc = 2;
	167	break;
	168
	169	case RXRPC_PACKET_TYPE_ABORT:
	170	abort_code = call->abort_code;
	171	pkt->abort_code = htonl(abort_code);
8d94aa38 DH	172	iov[0].iov_len += sizeof(pkt->abort_code);
	173	len += sizeof(pkt->abort_code);
	174	ioc = 1;
	175	break;
	176
	177	default:
	178	BUG();
	179	ret = -ENOANO;
	180	goto out;
	181	}
	182
b86e218e DH	183	serial = atomic_inc_return(&conn->serial);
	184	pkt->whdr.serial = htonl(serial);
	185	switch (type) {
	186	case RXRPC_PACKET_TYPE_ACK:
be832aec	187	trace_rxrpc_tx_ack(call, serial,
b86e218e DH	188	ntohl(pkt->ack.firstPacket),
	189	ntohl(pkt->ack.serial),
	190	pkt->ack.reason, pkt->ack.nAcks);
	191	break;
	192	}
	193
8e83134d DH	194	if (ping) {
	195	call->ackr_ping = serial;
	196	smp_wmb();
	197	/* We need to stick a time in before we send the packet in case
	198	* the reply gets back before kernel_sendmsg() completes - but
	199	* asking UDP to send the packet can take a relatively long
	200	* time, so we update the time after, on the assumption that
	201	* the packet transmission is more likely to happen towards the
	202	* end of the kernel_sendmsg() call.
	203	*/
	204	call->ackr_ping_time = ktime_get_real();
	205	set_bit(RXRPC_CALL_PINGING, &call->flags);
	206	trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_ping, serial);
	207	}
8d94aa38 DH	208	ret = kernel_sendmsg(conn->params.local->socket,
8d94aa38 DH	209	&msg, iov, ioc, len);
8e83134d DH	210	if (ping)
8e83134d DH	211	call->ackr_ping_time = ktime_get_real();
8d94aa38	212
805b21b9 DH	213	if (type == RXRPC_PACKET_TYPE_ACK &&
	214	call->state < RXRPC_CALL_COMPLETE) {
	215	if (ret < 0) {
8e83134d	216	clear_bit(RXRPC_CALL_PINGING, &call->flags);
248f219c DH	217	rxrpc_propose_ACK(call, pkt->ack.reason,
	218	ntohs(pkt->ack.maxSkew),
	219	ntohl(pkt->ack.serial),
9c7ad434 DH	220	true, true,
9c7ad434 DH	221	rxrpc_propose_ack_retry_tx);
805b21b9 DH	222	} else {
	223	spin_lock_bh(&call->lock);
	224	if (after(hard_ack, call->ackr_consumed))
	225	call->ackr_consumed = hard_ack;
	226	if (after(top, call->ackr_seen))
	227	call->ackr_seen = top;
	228	spin_unlock_bh(&call->lock);
248f219c DH	229	}
	230	}
	231
8d94aa38 DH	232	out:
	233	rxrpc_put_connection(conn);
	234	kfree(pkt);
	235	return ret;
	236	}
	237
17926a79 DH	238	/*
	239	* send a packet through the transport endpoint
	240	*/
a1767077 DH	241	int rxrpc_send_data_packet(struct rxrpc_call call, struct sk_buff skb,
a1767077 DH	242	bool retrans)
17926a79	243	{
5a924b89 DH	244	struct rxrpc_connection *conn = call->conn;
	245	struct rxrpc_wire_header whdr;
	246	struct rxrpc_skb_priv *sp = rxrpc_skb(skb);
17926a79	247	struct msghdr msg;
5a924b89 DH	248	struct kvec iov[2];
	249	rxrpc_serial_t serial;
	250	size_t len;
a1767077	251	bool lost = false;
17926a79 DH	252	int ret, opt;
	253
	254	_enter(",{%d}", skb->len);
	255
5a924b89 DH	256	/* Each transmission of a Tx packet needs a new serial number */
5a924b89 DH	257	serial = atomic_inc_return(&conn->serial);
17926a79	258
5a924b89 DH	259	whdr.epoch = htonl(conn->proto.epoch);
	260	whdr.cid = htonl(call->cid);
	261	whdr.callNumber = htonl(call->call_id);
	262	whdr.seq = htonl(sp->hdr.seq);
	263	whdr.serial = htonl(serial);
	264	whdr.type = RXRPC_PACKET_TYPE_DATA;
	265	whdr.flags = sp->hdr.flags;
	266	whdr.userStatus = 0;
	267	whdr.securityIndex = call->security_ix;
	268	whdr._rsvd = htons(sp->hdr._rsvd);
	269	whdr.serviceId = htons(call->service_id);
	270
	271	iov[0].iov_base = &whdr;
	272	iov[0].iov_len = sizeof(whdr);
	273	iov[1].iov_base = skb->head;
	274	iov[1].iov_len = skb->len;
	275	len = iov[0].iov_len + iov[1].iov_len;
	276
	277	msg.msg_name = &call->peer->srx.transport;
	278	msg.msg_namelen = call->peer->srx.transport_len;
17926a79 DH	279	msg.msg_control = NULL;
	280	msg.msg_controllen = 0;
	281	msg.msg_flags = 0;
	282
57494343 DH	283	/* If our RTT cache needs working on, request an ACK. Also request
	284	* ACKs if a DATA packet appears to have been lost.
	285	*/
a1767077	286	if (retrans \|\|
b112a670	287	call->cong_mode == RXRPC_CALL_SLOW_START \|\|
57494343	288	(call->peer->rtt_usage < 3 && sp->hdr.seq & 1) \|\|
0d4b103c DH	289	ktime_before(ktime_add_ms(call->peer->rtt_last_req, 1000),
	290	ktime_get_real()))
	291	whdr.flags \|= RXRPC_REQUEST_ACK;
	292
8a681c36 DH	293	if (IS_ENABLED(CONFIG_AF_RXRPC_INJECT_LOSS)) {
	294	static int lose;
	295	if ((lose++ & 7) == 7) {
a1767077 DH	296	ret = 0;
	297	lost = true;
	298	goto done;
8a681c36 DH	299	}
	300	}
	301
5a924b89 DH	302	_proto("Tx DATA %%%u { #%u }", serial, sp->hdr.seq);
5a924b89 DH	303
17926a79 DH	304	/* send the packet with the don't fragment bit set if we currently
17926a79 DH	305	* think it's small enough */
5a924b89 DH	306	if (iov[1].iov_len >= call->peer->maxdata)
	307	goto send_fragmentable;
	308
	309	down_read(&conn->params.local->defrag_sem);
	310	/* send the packet by UDP
	311	* - returns -EMSGSIZE if UDP would have to fragment the packet
	312	* to go out of the interface
	313	* - in which case, we'll have processed the ICMP error
	314	* message and update the peer record
	315	*/
	316	ret = kernel_sendmsg(conn->params.local->socket, &msg, iov, 2, len);
	317
	318	up_read(&conn->params.local->defrag_sem);
	319	if (ret == -EMSGSIZE)
	320	goto send_fragmentable;
	321
	322	done:
a1767077 DH	323	trace_rxrpc_tx_data(call, sp->hdr.seq, serial, whdr.flags,
a1767077 DH	324	retrans, lost);
50235c4b	325	if (ret >= 0) {
0d4b103c DH	326	ktime_t now = ktime_get_real();
0d4b103c DH	327	skb->tstamp = now;
50235c4b	328	smp_wmb();
5a924b89	329	sp->hdr.serial = serial;
0d4b103c DH	330	if (whdr.flags & RXRPC_REQUEST_ACK) {
0d4b103c DH	331	call->peer->rtt_last_req = now;
50235c4b	332	trace_rxrpc_rtt_tx(call, rxrpc_rtt_tx_data, serial);
0d4b103c	333	}
17926a79	334	}
5a924b89 DH	335	_leave(" = %d [%u]", ret, call->peer->maxdata);
5a924b89 DH	336	return ret;
17926a79 DH	337
	338	send_fragmentable:
	339	/* attempt to send this message with fragmentation enabled */
	340	_debug("send fragment");
	341
985a5c82 DH	342	down_write(&conn->params.local->defrag_sem);
	343
	344	switch (conn->params.local->srx.transport.family) {
	345	case AF_INET:
	346	opt = IP_PMTUDISC_DONT;
	347	ret = kernel_setsockopt(conn->params.local->socket,
	348	SOL_IP, IP_MTU_DISCOVER,
	349	(char *)&opt, sizeof(opt));
	350	if (ret == 0) {
5a924b89 DH	351	ret = kernel_sendmsg(conn->params.local->socket, &msg,
5a924b89 DH	352	iov, 2, len);
985a5c82 DH	353
	354	opt = IP_PMTUDISC_DO;
	355	kernel_setsockopt(conn->params.local->socket, SOL_IP,
	356	IP_MTU_DISCOVER,
	357	(char *)&opt, sizeof(opt));
	358	}
	359	break;
75b54cb5	360
d1912747	361	#ifdef CONFIG_AF_RXRPC_IPV6
75b54cb5 DH	362	case AF_INET6:
	363	opt = IPV6_PMTUDISC_DONT;
	364	ret = kernel_setsockopt(conn->params.local->socket,
	365	SOL_IPV6, IPV6_MTU_DISCOVER,
	366	(char *)&opt, sizeof(opt));
	367	if (ret == 0) {
	368	ret = kernel_sendmsg(conn->params.local->socket, &msg,
	369	iov, 1, iov[0].iov_len);
	370
	371	opt = IPV6_PMTUDISC_DO;
	372	kernel_setsockopt(conn->params.local->socket,
	373	SOL_IPV6, IPV6_MTU_DISCOVER,
	374	(char *)&opt, sizeof(opt));
	375	}
	376	break;
d1912747	377	#endif
17926a79 DH	378	}
17926a79 DH	379
985a5c82	380	up_write(&conn->params.local->defrag_sem);
5a924b89	381	goto done;
17926a79	382	}
248f219c DH	383
	384	/*
	385	* reject packets through the local endpoint
	386	*/
	387	void rxrpc_reject_packets(struct rxrpc_local *local)
	388	{
1c2bc7b9	389	struct sockaddr_rxrpc srx;
248f219c DH	390	struct rxrpc_skb_priv *sp;
	391	struct rxrpc_wire_header whdr;
	392	struct sk_buff *skb;
	393	struct msghdr msg;
	394	struct kvec iov[2];
	395	size_t size;
	396	__be32 code;
	397
	398	_enter("%d", local->debug_id);
	399
	400	iov[0].iov_base = &whdr;
	401	iov[0].iov_len = sizeof(whdr);
	402	iov[1].iov_base = &code;
	403	iov[1].iov_len = sizeof(code);
	404	size = sizeof(whdr) + sizeof(code);
	405
1c2bc7b9	406	msg.msg_name = &srx.transport;
248f219c DH	407	msg.msg_control = NULL;
	408	msg.msg_controllen = 0;
	409	msg.msg_flags = 0;
	410
248f219c DH	411	memset(&whdr, 0, sizeof(whdr));
	412	whdr.type = RXRPC_PACKET_TYPE_ABORT;
	413
	414	while ((skb = skb_dequeue(&local->reject_queue))) {
71f3ca40	415	rxrpc_see_skb(skb, rxrpc_skb_rx_seen);
248f219c	416	sp = rxrpc_skb(skb);
1c2bc7b9 DH	417
	418	if (rxrpc_extract_addr_from_skb(&srx, skb) == 0) {
	419	msg.msg_namelen = srx.transport_len;
	420
248f219c DH	421	code = htonl(skb->priority);
	422
	423	whdr.epoch = htonl(sp->hdr.epoch);
	424	whdr.cid = htonl(sp->hdr.cid);
	425	whdr.callNumber = htonl(sp->hdr.callNumber);
	426	whdr.serviceId = htons(sp->hdr.serviceId);
	427	whdr.flags = sp->hdr.flags;
	428	whdr.flags ^= RXRPC_CLIENT_INITIATED;
	429	whdr.flags &= RXRPC_CLIENT_INITIATED;
	430
	431	kernel_sendmsg(local->socket, &msg, iov, 2, size);
248f219c DH	432	}
248f219c DH	433
71f3ca40	434	rxrpc_free_skb(skb, rxrpc_skb_rx_freed);
248f219c DH	435	}
	436
	437	_leave("");
	438	}