nuttx/net/tcp/tcp.h

2340 lines
76 KiB
C
Raw Normal View History

/****************************************************************************
* net/tcp/tcp.h
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership. The
* ASF licenses this file to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance with the
* License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations
* under the License.
*
****************************************************************************/
#ifndef __NET_TCP_TCP_H
#define __NET_TCP_TCP_H
/****************************************************************************
* Included Files
****************************************************************************/
#include <nuttx/config.h>
#include <sys/types.h>
#include <nuttx/clock.h>
#include <nuttx/queue.h>
#include <nuttx/semaphore.h>
#include <nuttx/mm/iob.h>
#include <nuttx/net/ip.h>
#include <nuttx/net/net.h>
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
#include <nuttx/net/tcp.h>
#include <nuttx/wqueue.h>
#ifdef CONFIG_NET_TCP
/****************************************************************************
* Pre-processor Definitions
****************************************************************************/
/* This is a helper pointer for accessing the contents of the tcp header */
#define TCPIPv4BUF ((FAR struct tcp_hdr_s *)IPBUF(IPv4_HDRLEN))
#define TCPIPv6BUF ((FAR struct tcp_hdr_s *)IPBUF(IPv6_HDRLEN))
#ifndef CONFIG_NET_TCP_NO_STACK
#define NET_TCP_HAVE_STACK 1
/* Allocate a new TCP data callback */
/* These macros allocate and free callback structures used for receiving
* notifications of TCP data-related events.
*/
#define tcp_callback_alloc(conn) \
devif_callback_alloc((conn)->dev, &(conn)->sconn.list, &(conn)->sconn.list_tail)
#define tcp_callback_free(conn,cb) \
devif_conn_callback_free((conn)->dev, (cb), &(conn)->sconn.list, &(conn)->sconn.list_tail)
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
/* TCP write buffer access macros */
# define TCP_WBSEQNO(wrb) ((wrb)->wb_seqno)
# define TCP_WBPKTLEN(wrb) ((wrb)->wb_iob->io_pktlen)
# define TCP_WBSENT(wrb) ((wrb)->wb_sent)
# define TCP_WBNRTX(wrb) ((wrb)->wb_nrtx)
#if defined(CONFIG_NET_TCP_FAST_RETRANSMIT) && !defined(CONFIG_NET_TCP_CC_NEWRENO)
# define TCP_WBNACK(wrb) ((wrb)->wb_nack)
#endif
# define TCP_WBIOB(wrb) ((wrb)->wb_iob)
# define TCP_WBCOPYOUT(wrb,dest,n) (iob_copyout(dest,(wrb)->wb_iob,(n),0))
tcp_send_buffered.c: improve tcp write buffering * Send data chunk-by-chunk Note: A stream socket doesn't have atomicity requirement. * Increase the chance to use full-sized segments Benchmark numbers in my environment: * Over ESP32 wifi * The peer is NetBSD, which has traditional delayed ack TCP * iperf uses 16384 bytes buffer --- without this patch, CONFIG_IOB_NBUFFERS=36 CONFIG_IOB_BUFSIZE=196 does not work. see https://github.com/apache/incubator-nuttx/pull/2772#discussion_r592820639 --- without this patch, CONFIG_IOB_NBUFFERS=128 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 4.11 Mbits/sec 3- 6 sec, 4.63 Mbits/sec 6- 9 sec, 4.89 Mbits/sec 9- 12 sec, 4.63 Mbits/sec 12- 15 sec, 4.85 Mbits/sec 15- 18 sec, 4.85 Mbits/sec 18- 21 sec, 5.02 Mbits/sec 21- 24 sec, 3.67 Mbits/sec 24- 27 sec, 4.94 Mbits/sec 27- 30 sec, 4.81 Mbits/sec 0- 30 sec, 4.64 Mbits/sec nsh> ``` --- with this patch, CONFIG_IOB_NBUFFERS=36 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 5.33 Mbits/sec 3- 6 sec, 5.59 Mbits/sec 6- 9 sec, 5.55 Mbits/sec 9- 12 sec, 5.59 Mbits/sec 12- 15 sec, 5.59 Mbits/sec 15- 18 sec, 5.72 Mbits/sec 18- 21 sec, 5.68 Mbits/sec 21- 24 sec, 5.29 Mbits/sec 24- 27 sec, 4.67 Mbits/sec 27- 30 sec, 4.50 Mbits/sec 0- 30 sec, 5.35 Mbits/sec nsh> ``` --- with this patch, CONFIG_IOB_NBUFFERS=128 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 5.51 Mbits/sec 3- 6 sec, 4.67 Mbits/sec 6- 9 sec, 4.54 Mbits/sec 9- 12 sec, 5.42 Mbits/sec 12- 15 sec, 5.37 Mbits/sec 15- 18 sec, 5.11 Mbits/sec 18- 21 sec, 5.07 Mbits/sec 21- 24 sec, 5.29 Mbits/sec 24- 27 sec, 5.77 Mbits/sec 27- 30 sec, 4.63 Mbits/sec 0- 30 sec, 5.14 Mbits/sec nsh> ```
2021-03-15 08:19:42 +01:00
# define TCP_WBCOPYIN(wrb,src,n,off) \
(iob_copyin((wrb)->wb_iob,src,(n),(off),true))
tcp_send_buffered.c: improve tcp write buffering * Send data chunk-by-chunk Note: A stream socket doesn't have atomicity requirement. * Increase the chance to use full-sized segments Benchmark numbers in my environment: * Over ESP32 wifi * The peer is NetBSD, which has traditional delayed ack TCP * iperf uses 16384 bytes buffer --- without this patch, CONFIG_IOB_NBUFFERS=36 CONFIG_IOB_BUFSIZE=196 does not work. see https://github.com/apache/incubator-nuttx/pull/2772#discussion_r592820639 --- without this patch, CONFIG_IOB_NBUFFERS=128 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 4.11 Mbits/sec 3- 6 sec, 4.63 Mbits/sec 6- 9 sec, 4.89 Mbits/sec 9- 12 sec, 4.63 Mbits/sec 12- 15 sec, 4.85 Mbits/sec 15- 18 sec, 4.85 Mbits/sec 18- 21 sec, 5.02 Mbits/sec 21- 24 sec, 3.67 Mbits/sec 24- 27 sec, 4.94 Mbits/sec 27- 30 sec, 4.81 Mbits/sec 0- 30 sec, 4.64 Mbits/sec nsh> ``` --- with this patch, CONFIG_IOB_NBUFFERS=36 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 5.33 Mbits/sec 3- 6 sec, 5.59 Mbits/sec 6- 9 sec, 5.55 Mbits/sec 9- 12 sec, 5.59 Mbits/sec 12- 15 sec, 5.59 Mbits/sec 15- 18 sec, 5.72 Mbits/sec 18- 21 sec, 5.68 Mbits/sec 21- 24 sec, 5.29 Mbits/sec 24- 27 sec, 4.67 Mbits/sec 27- 30 sec, 4.50 Mbits/sec 0- 30 sec, 5.35 Mbits/sec nsh> ``` --- with this patch, CONFIG_IOB_NBUFFERS=128 CONFIG_IOB_BUFSIZE=196 ``` nsh> iperf -c 192.168.8.1 IP: 192.168.8.103 mode=tcp-client sip=192.168.8.103:5001,dip=192.168.8.1:5001, interval=3, time=30 Interval Bandwidth 0- 3 sec, 5.51 Mbits/sec 3- 6 sec, 4.67 Mbits/sec 6- 9 sec, 4.54 Mbits/sec 9- 12 sec, 5.42 Mbits/sec 12- 15 sec, 5.37 Mbits/sec 15- 18 sec, 5.11 Mbits/sec 18- 21 sec, 5.07 Mbits/sec 21- 24 sec, 5.29 Mbits/sec 24- 27 sec, 5.77 Mbits/sec 27- 30 sec, 4.63 Mbits/sec 0- 30 sec, 5.14 Mbits/sec nsh> ```
2021-03-15 08:19:42 +01:00
# define TCP_WBTRYCOPYIN(wrb,src,n,off) \
(iob_trycopyin((wrb)->wb_iob,src,(n),(off),true))
# define TCP_WBTRIM(wrb,n) \
do { (wrb)->wb_iob = iob_trimhead((wrb)->wb_iob,(n)); } while (0)
#ifdef CONFIG_DEBUG_FEATURES
# define TCP_WBDUMP(msg,wrb,len,offset) \
tcp_wrbuffer_dump(msg,wrb,len,offset)
# else
# define TCP_WBDUMP(msg,wrb,len,offset)
# endif
#endif
/* 32-bit modular arithmetics for tcp sequence numbers */
#define TCP_SEQ_LT(a, b) ((int32_t)((a) - (b)) < 0)
#define TCP_SEQ_GT(a, b) TCP_SEQ_LT(b, a)
#define TCP_SEQ_LTE(a, b) (!TCP_SEQ_GT(a, b))
#define TCP_SEQ_GTE(a, b) (!TCP_SEQ_LT(a, b))
#define TCP_SEQ_ADD(a, b) ((uint32_t)((a) + (b)))
#define TCP_SEQ_SUB(a, b) ((uint32_t)((a) - (b)))
/* The TCP options flags */
#define TCP_WSCALE 0x01U /* Window Scale option enabled */
#define TCP_SACK 0x02U /* Selective ACKs enabled */
#define TCP_CLOSE_ARRANGED 0x04U /* Connection is arranged to be freed */
#ifdef CONFIG_NET_TCP_CC_NEWRENO
/* The TCP flags for congestion control */
#define TCP_INFR 0x08U /* The flag in Fast Recovery */
#define TCP_INFT 0x10U /* The flag in Fast Transmitted */
#endif
/* The Max Range count of TCP Selective ACKs */
#define TCP_SACK_RANGES_MAX 4
/* After receiving 3 duplicate ACKs, TCP performs a retransmission
* (RFC 5681 (3.2))
*/
#define TCP_FAST_RETRANSMISSION_THRESH 3
#define TCP_RTO_MAX 240 /* 120s,The unit is half a second */
#define TCP_RTO_MIN 1 /* 0.5s */
/****************************************************************************
* Public Type Definitions
****************************************************************************/
struct file; /* Forward reference */
struct sockaddr; /* Forward reference */
struct socket; /* Forward reference */
struct pollfd; /* Forward reference */
/* Representation of a TCP connection.
*
* The tcp_conn_s structure is used for identifying a connection. All
* but one field in the structure are to be considered read-only by an
* application. The only exception is the 'private' fields whose purpose
* is to let the application store application-specific state (e.g.,
* file pointers) for the connection.
*/
struct net_driver_s; /* Forward reference */
struct devif_callback_s; /* Forward reference */
struct tcp_backlog_s; /* Forward reference */
struct tcp_hdr_s; /* Forward reference */
/* This is a container that holds the poll-related information */
struct tcp_poll_s
{
FAR struct tcp_conn_s *conn; /* Needed to handle loss of connection */
FAR struct pollfd *fds; /* Needed to handle poll events */
FAR struct devif_callback_s *cb; /* Needed to teardown the poll */
};
/* Out-of-order segments */
struct tcp_ofoseg_s
{
uint32_t left; /* Left edge of segment */
uint32_t right; /* Right edge of segment */
FAR struct iob_s *data; /* Out-of-order buffering */
};
/* SACK ranges to include in ACK packets. */
struct tcp_sack_s
{
uint32_t left; /* Left edge of the SACK */
uint32_t right; /* Right edge of the SACK */
};
struct tcp_conn_s
{
/* Common prologue of all connection structures. */
/* TCP callbacks:
*
* Data transfer events are retained in 'list'. Event handlers in 'list'
* are called for events specified in the flags set within struct
* devif_callback_s
*
* When an callback is executed from 'list', the input flags are normally
* returned, however, the implementation may set one of the following:
*
* TCP_CLOSE - Gracefully close the current connection
* TCP_ABORT - Abort (reset) the current connection on an error that
* prevents TCP_CLOSE from working.
*
* And/Or set/clear the following:
*
* TCP_NEWDATA - May be cleared to indicate that the data was consumed
* and that no further process of the new data should be
* attempted.
* TCP_SNDACK - If TCP_NEWDATA is cleared, then TCP_SNDACK may be set to
* indicate that an ACK should be included in the response.
* (In TCP_NEWDATA is cleared bu TCP_SNDACK is not set,
* then dev->d_len should also be cleared).
*/
struct socket_conn_s sconn;
/* TCP-specific content follows */
union ip_binding_u u; /* IP address binding */
uint8_t rcvseq[4]; /* The sequence number that we expect to
* receive next */
uint8_t sndseq[4]; /* The sequence number that was last sent by us */
#if !defined(CONFIG_NET_TCP_WRITE_BUFFERS) || \
defined(CONFIG_NET_SENDFILE)
uint32_t rexmit_seq; /* The sequence number to be retrasmitted */
#endif
uint8_t crefs; /* Reference counts on this instance */
#if defined(CONFIG_NET_IPv4) && defined(CONFIG_NET_IPv6)
2016-02-26 13:45:37 +01:00
uint8_t domain; /* IP domain: PF_INET or PF_INET6 */
#endif
uint8_t sa; /* Retransmission time-out calculation state
* variable */
uint8_t sv; /* Retransmission time-out calculation state
* variable */
uint8_t rto; /* Retransmission time-out */
uint8_t tcpstateflags; /* TCP state and flags */
struct work_s work; /* TCP timer handle */
bool timeout; /* Trigger from timer expiry */
uint8_t timer; /* The retransmission timer (units: half-seconds) */
uint8_t nrtx; /* The number of retransmissions for the last
* segment sent */
#ifdef CONFIG_NET_TCP_DELAYED_ACK
uint8_t rx_unackseg; /* Number of un-ACKed received segments */
uint8_t rx_acktimer; /* Time since last ACK sent (units: half-seconds) */
#endif
uint16_t lport; /* The local TCP port, in network byte order */
uint16_t rport; /* The remoteTCP port, in network byte order */
uint16_t mss; /* Current maximum segment size for the
* connection */
#ifdef CONFIG_NET_TCPPROTO_OPTIONS
uint16_t user_mss; /* Configured maximum segment size for the
* connection */
#endif
uint32_t rcv_adv; /* The right edge of the recv window advertized */
#ifdef CONFIG_NET_TCP_CC_NEWRENO
uint32_t last_ackno; /* The ack number at the last receive ack */
uint32_t dupacks; /* The number of duplicate ack */
uint32_t fr_recover; /* The snd_seq at the retransmissions */
uint32_t cwnd; /* The Congestion window */
uint32_t max_cwnd; /* The Congestion window maximum value */
uint32_t ssthresh; /* The Slow start threshold */
#endif
#ifdef CONFIG_NET_TCP_WINDOW_SCALE
uint32_t snd_wnd; /* Sequence and acknowledgement numbers of last
* window update */
uint8_t snd_scale; /* Sender window scale factor */
uint8_t rcv_scale; /* Receiver windows scale factor */
#else
uint16_t snd_wnd; /* Sequence and acknowledgement numbers of last
* window update */
#endif
uint32_t snd_wl1;
uint32_t snd_wl2;
#if CONFIG_NET_RECV_BUFSIZE > 0
int32_t rcv_bufs; /* Maximum amount of bytes queued in recv */
#endif
#if CONFIG_NET_SEND_BUFSIZE > 0
int32_t snd_bufs; /* Maximum amount of bytes queued in send */
sem_t snd_sem; /* Semaphore signals send completion */
#endif
#if defined(CONFIG_NET_TCP_WRITE_BUFFERS) || \
defined(CONFIG_NET_TCP_WINDOW_SCALE)
uint32_t tx_unacked; /* Number bytes sent but not yet ACKed */
#else
uint16_t tx_unacked; /* Number bytes sent but not yet ACKed */
#endif
uint16_t flags; /* Flags of TCP-specific options */
#ifdef CONFIG_NET_SOLINGER
sclock_t ltimeout; /* Linger timeout expiration */
#endif
#ifdef CONFIG_NETDEV_RSS
int rcvcpu; /* Currect cpu id */
#endif
/* If the TCP socket is bound to a local address, then this is
* a reference to the device that routes traffic on the corresponding
* network.
*/
FAR struct net_driver_s *dev;
/* Read-ahead buffering.
*
* readahead - An IOB chain where the TCP/IP read-ahead data is retained.
*/
FAR struct iob_s *readahead; /* Read-ahead buffering */
#ifdef CONFIG_NET_TCP_OUT_OF_ORDER
/* Number of out-of-order segments */
uint8_t nofosegs;
/* This defines a out of order segment block. */
struct tcp_ofoseg_s ofosegs[TCP_SACK_RANGES_MAX];
#endif
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
/* Write buffering
*
* write_q - The queue of unsent I/O buffers. The head of this
* list may be partially sent. FIFO ordering.
* unacked_q - A queue of completely sent, but unacked I/O buffer
* chains. Sequence number ordering.
*/
sq_queue_t write_q; /* Write buffering for segments */
sq_queue_t unacked_q; /* Write buffering for un-ACKed segments */
uint16_t expired; /* Number segments retransmitted but not yet ACKed,
* it can only be updated at TCP_ESTABLISHED state */
uint32_t sent; /* The number of bytes sent (ACKed and un-ACKed) */
uint32_t isn; /* Initial sequence number */
uint32_t sndseq_max; /* The sequence number of next not-retransmitted
* segment (next greater sndseq) */
#endif
#ifdef CONFIG_NET_TCPBACKLOG
/* Listen backlog support
*
* blparent - The backlog parent. If this connection is backlogged,
* this field will be non-null and will refer to the TCP connection
* structure in which this connection is backlogged.
* backlog - The pending connection backlog. If this connection is
* configured as a listener with backlog, then this refers to the
* struct tcp_backlog_s tear-off structure that manages that backlog.
*/
FAR struct tcp_conn_s *blparent;
FAR struct tcp_backlog_s *backlog;
#endif
#ifdef CONFIG_NET_TCP_KEEPALIVE
/* There fields manage TCP/IP keep-alive. All times are in units of the
* system clock tick.
*/
uint32_t keeptimer; /* KeepAlive timer (dsec) */
uint32_t keepidle; /* Elapsed idle time before first probe sent (dsec) */
uint32_t keepintvl; /* Interval between probes (dsec) */
bool keepalive; /* True: KeepAlive enabled; false: disabled */
uint8_t keepcnt; /* Number of retries before the socket is closed */
uint8_t keepretries; /* Number of retries attempted */
#endif
#if defined(CONFIG_NET_SENDFILE) && defined(CONFIG_NET_TCP_WRITE_BUFFERS)
bool sendfile; /* True if sendfile operation is in progress */
#endif
bool zero_probe; /* TCP zero window probe timer */
/* connevents is a list of callbacks for each socket the uses this
* connection (there can be more that one in the event that the the socket
* was dup'ed). It is used with the network monitor to handle
* asynchronous loss-of-connection events.
*/
FAR struct devif_callback_s *connevents;
FAR struct devif_callback_s *connevents_tail;
/* Reference to TCP shutdown/close callback instance */
FAR struct devif_callback_s *shdcb;
FAR struct devif_callback_s *clscb;
net/tcp: use independent work to free the conn instance I noticed that the conn instance will leak during stress test, The close work queued from tcp_close_eventhandler() will be canceled by tcp_timer() immediately: Breakpoint 1, tcp_close_eventhandler (dev=0x565cd338 <up_irq_restore+108>, pvpriv=0x5655e6ff <getpid+12>, flags=0) at tcp/tcp_close.c:71 (gdb) bt | #0 tcp_close_eventhandler (dev=0x565cd338 <up_irq_restore+108>, pvpriv=0x5655e6ff <getpid+12>, flags=0) at tcp/tcp_close.c:71 | #1 0x5658bf1e in devif_conn_event (dev=0x5660bd80 <g_sim_dev>, flags=512, list=0x5660d558 <g_cbprealloc+312>) at devif/devif_callback.c:508 | #2 0x5658a219 in tcp_callback (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>, flags=512) at tcp/tcp_callback.c:167 | #3 0x56589253 in tcp_timer (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:378 | #4 0x5658dd47 in tcp_poll (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_devpoll.c:95 | #5 0x5658b95f in devif_poll_tcp_connections (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:601 | #6 0x5658b9ea in devif_poll (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:722 | #7 0x56577230 in netdriver_txavail_work (arg=0x5660bd80 <g_sim_dev>) at sim/up_netdriver.c:308 | #8 0x5655999e in work_thread (argc=2, argv=0xf3db5dd0) at wqueue/kwork_thread.c:178 | #9 0x5655983f in nxtask_start () at task/task_start.c:129 (gdb) c Continuing. Breakpoint 2, tcp_update_timer (conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:178 (gdb) bt | #0 tcp_update_timer (conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:178 | #1 0x5658952a in tcp_timer (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_timer.c:708 | #2 0x5658dd47 in tcp_poll (dev=0x5660bd80 <g_sim_dev>, conn=0x5660c4a0 <g_tcp_connections>) at tcp/tcp_devpoll.c:95 | #3 0x5658b95f in devif_poll_tcp_connections (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:601 | #4 0x5658b9ea in devif_poll (dev=0x5660bd80 <g_sim_dev>, callback=0x565770f2 <netdriver_txpoll>) at devif/devif_poll.c:722 | #5 0x56577230 in netdriver_txavail_work (arg=0x5660bd80 <g_sim_dev>) at sim/up_netdriver.c:308 | #6 0x5655999e in work_thread (argc=2, argv=0xf3db5dd0) at wqueue/kwork_thread.c:178 | #7 0x5655983f in nxtask_start () at task/task_start.c:129 Since a separate work will add 24 bytes to each conn instance, but in order to support the feature of asynchronous close(), I can not find a better way than adding a separate work, for resource constraints, I recommend the developers to enable CONFIG_NET_ALLOC_CONNS, which will reduce the ram usage. Signed-off-by: chao an <anchao@xiaomi.com>
2022-08-30 05:04:31 +02:00
struct work_s clswork;
#if defined(CONFIG_NET_TCP_WRITE_BUFFERS)
/* Callback instance for TCP send() */
FAR struct devif_callback_s *sndcb;
#endif
/* accept() is called when the TCP logic has created a connection
*
* accept_private: This is private data that will be available to the
* accept() handler when it is invoked with a point to this structure
* as an argument.
* accept: This is the pointer to the accept handler.
*/
FAR void *accept_private;
CODE int (*accept)(FAR struct tcp_conn_s *listener,
FAR struct tcp_conn_s *conn);
/* The following is a list of poll structures of threads waiting for
* socket events.
*/
struct tcp_poll_s pollinfo[CONFIG_NET_TCP_NPOLLWAITERS];
};
/* This structure supports TCP write buffering */
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
struct tcp_wrbuffer_s
{
sq_entry_t wb_node; /* Supports a singly linked list */
uint32_t wb_seqno; /* Sequence number of the write segment */
uint16_t wb_sent; /* Number of bytes sent from the I/O buffer chain */
uint8_t wb_nrtx; /* The number of retransmissions for the last
* segment sent */
#if defined(CONFIG_NET_TCP_FAST_RETRANSMIT) && !defined(CONFIG_NET_TCP_CC_NEWRENO)
uint8_t wb_nack; /* The number of ack count */
#endif
struct iob_s *wb_iob; /* Head of the I/O buffer chain */
};
#endif
/* Support for listen backlog:
*
* struct tcp_blcontainer_s describes one backlogged connection
* struct tcp_backlog_s is a "tear-off" describing all backlog for a
* listener connection
*/
#ifdef CONFIG_NET_TCPBACKLOG
struct tcp_blcontainer_s
{
sq_entry_t bc_node; /* Implements a singly linked list */
FAR struct tcp_conn_s *bc_conn; /* Holds reference to the new connection structure */
};
struct tcp_backlog_s
{
sq_queue_t bl_free; /* Implements a singly-linked list of free containers */
sq_queue_t bl_pending; /* Implements a singly-linked list of pending connections */
};
#endif
/****************************************************************************
* Public Data
****************************************************************************/
#ifdef __cplusplus
extern "C"
{
#endif
/****************************************************************************
* Public Function Prototypes
****************************************************************************/
/****************************************************************************
* Name: tcp_initialize
*
* Description:
* Initialize the TCP/IP connection structures. Called only once and only
* from the network layer at start-up.
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
void tcp_initialize(void);
/****************************************************************************
* Name: tcp_alloc
*
* Description:
* Find a free TCP/IP connection structure and allocate it
* for use. This is normally something done by the implementation of the
* socket() API but is also called from the event processing logic when a
* TCP packet is received while "listening"
*
****************************************************************************/
FAR struct tcp_conn_s *tcp_alloc(uint8_t domain);
/****************************************************************************
* Name: tcp_free_rx_buffers
*
* Description:
* Free rx buffer of a connection
*
****************************************************************************/
void tcp_free_rx_buffers(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_free
*
* Description:
* Free a connection structure that is no longer in use. This should be
* done by the implementation of close()
*
****************************************************************************/
void tcp_free(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_active
*
* Description:
* Find a connection structure that is the appropriate
* connection to be used with the provided TCP/IP header
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
FAR struct tcp_conn_s *tcp_active(FAR struct net_driver_s *dev,
FAR struct tcp_hdr_s *tcp);
/****************************************************************************
* Name: tcp_nextconn
*
* Description:
* Traverse the list of active TCP connections
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
FAR struct tcp_conn_s *tcp_nextconn(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_local_ipv4_device
*
* Description:
* Select the network driver to use with the IPv4 TCP transaction based
* on the locally bound IPv4 address
*
* Input Parameters:
* conn - TCP connection structure. The locally bound address, laddr,
* should be set to a non-zero value in this structure.
*
* Returned Value:
* Zero (OK) is returned on success. A negated errno value is returned
2015-09-03 04:03:28 +02:00
* on failure. -ENETUNREACH is the only expected error value.
*
****************************************************************************/
#ifdef CONFIG_NET_IPv4
int tcp_local_ipv4_device(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_remote_ipv4_device
*
* Description:
* Select the network driver to use with the IPv4 TCP transaction based
* on the remotely connected IPv4 address
*
* Input Parameters:
2020-02-14 15:50:45 +01:00
* conn - TCP connection structure. The remotely connected address, raddr,
* should be set to a non-zero value in this structure.
*
* Returned Value:
* Zero (OK) is returned on success. A negated errno value is returned
2015-09-03 04:03:28 +02:00
* on failure. -ENETUNREACH is the only expected error value.
*
****************************************************************************/
#ifdef CONFIG_NET_IPv4
int tcp_remote_ipv4_device(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_local_ipv6_device
*
* Description:
* Select the network driver to use with the IPv6 TCP transaction based
* on the locally bound IPv6 address
*
* Input Parameters:
* conn - TCP connection structure. The locally bound address, laddr,
* should be set to a non-zero value in this structure.
*
* Returned Value:
* Zero (OK) is returned on success. A negated errno value is returned
* on failure. -EHOSTUNREACH is the only expected error value.
*
****************************************************************************/
#ifdef CONFIG_NET_IPv6
int tcp_local_ipv6_device(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_remote_ipv6_device
*
* Description:
* Select the network driver to use with the IPv6 TCP transaction based
* on the remotely connected IPv6 address
*
* Input Parameters:
* conn - TCP connection structure. The remotely connected address, raddr,
* should be set to a non-zero value in this structure.
*
* Returned Value:
* Zero (OK) is returned on success. A negated errno value is returned
* on failure. -EHOSTUNREACH is the only expected error value.
*
****************************************************************************/
#ifdef CONFIG_NET_IPv6
int tcp_remote_ipv6_device(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_alloc_accept
*
* Description:
* Called when driver processing matches the incoming packet with a
* connection in LISTEN. In that case, this function will create a new
* connection and initialize it to send a SYNACK in return.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
FAR struct tcp_conn_s *tcp_alloc_accept(FAR struct net_driver_s *dev,
FAR struct tcp_hdr_s *tcp,
FAR struct tcp_conn_s *listener);
/****************************************************************************
* Name: tcp_selectport
*
* Description:
* If the port number is zero; select an unused port for the connection.
* If the port number is non-zero, verify that no other connection has
* been created with this port number.
*
* Returned Value:
* Selected or verified port number in network order on success, a negated
* errno on failure.
*
* Assumptions:
* Interrupts are disabled
*
****************************************************************************/
int tcp_selectport(uint8_t domain,
FAR const union ip_addr_u *ipaddr,
uint16_t portno);
/****************************************************************************
* Name: tcp_bind
*
* Description:
* This function implements the lower level parts of the standard TCP
* bind() operation.
*
* Returned Value:
* 0 on success or -EADDRINUSE on failure
*
* Assumptions:
* This function is called from normal user level code.
*
****************************************************************************/
int tcp_bind(FAR struct tcp_conn_s *conn, FAR const struct sockaddr *addr);
/****************************************************************************
* Name: tcp_connect
*
* Description:
* This function implements the lower level parts of the standard
* TCP connect() operation: It connects to a remote host using TCP.
*
* This function is used to start a new connection to the specified
* port on the specified host. It uses the connection structure that was
* allocated by a preceding socket() call. It sets the connection to
* the SYN_SENT state and sets the retransmission timer to 0. This will
* cause a TCP SYN segment to be sent out the next time this connection
* is periodically processed, which usually is done within 0.5 seconds
* after the call to tcp_connect().
*
* Assumptions:
* This function is called from normal user level code.
*
****************************************************************************/
int tcp_connect(FAR struct tcp_conn_s *conn,
FAR const struct sockaddr *addr);
2014-06-25 02:12:49 +02:00
/****************************************************************************
* Name: psock_tcp_connect
*
* Description:
* Perform a TCP connection
*
* Input Parameters:
* psock - A reference to the structure of the socket to be connected
* addr - The address of the remote server to connect to
*
* Returned Value:
* None
*
* Assumptions:
* The network is locked
*
****************************************************************************/
int psock_tcp_connect(FAR struct socket *psock,
FAR const struct sockaddr *addr);
/****************************************************************************
* Name: tcp_start_monitor
*
* Description:
* Set up to receive TCP connection state changes for a given socket
*
* Input Parameters:
* psock - The socket of interest
*
* Returned Value:
* On success, tcp_start_monitor returns OK; On any failure,
* tcp_start_monitor will return a negated errno value. The only failure
* that can occur is if the socket has already been closed and, in this
* case, -ENOTCONN is returned.
*
* Assumptions:
* The caller holds the network lock (if not, it will be locked momentarily
* by this function).
*
****************************************************************************/
int tcp_start_monitor(FAR struct socket *psock);
/****************************************************************************
* Name: tcp_stop_monitor
*
* Description:
* Stop monitoring TCP connection changes for a sockets associated with
* a given TCP connection structure.
*
* Input Parameters:
* conn - The TCP connection of interest
* flags Set of disconnection events
*
* Returned Value:
* None
*
* Assumptions:
* The caller holds the network lock (if not, it will be locked momentarily
* by this function).
*
****************************************************************************/
void tcp_stop_monitor(FAR struct tcp_conn_s *conn, uint16_t flags);
/****************************************************************************
* Name: tcp_lost_connection
*
* Description:
* Called when a loss-of-connection event has been detected by network
* event handling logic. Perform operations like tcp_stop_monitor but
* (1) explicitly mark this socket and (2) disable further callbacks
* the event handler.
*
* Input Parameters:
* conn - The TCP connection of interest
* cb - devif callback structure
* flags - Set of connection events
*
* Returned Value:
* None
*
* Assumptions:
* The caller holds the network lock (if not, it will be locked momentarily
* by this function).
*
****************************************************************************/
void tcp_lost_connection(FAR struct tcp_conn_s *conn,
FAR struct devif_callback_s *cb, uint16_t flags);
/****************************************************************************
* Name: tcp_close
*
* Description:
* Break any current TCP connection
*
* Input Parameters:
* psock - An instance of the internal socket structure.
*
* Assumptions:
* Called from normal user-level logic
*
****************************************************************************/
int tcp_close(FAR struct socket *psock);
/****************************************************************************
* Name: tcp_shutdown
*
* Description:
* Gracefully shutdown a TCP connection by sending a SYN
*
* Input Parameters:
* psock - An instance of the internal socket structure.
* how - Specifies the type of shutdown.
*
* Assumptions:
* Called from normal user-level logic
*
****************************************************************************/
int tcp_shutdown(FAR struct socket *psock, int how);
/****************************************************************************
* Name: tcp_ipv4_select
*
* Description:
* Configure to send or receive an TCP IPv4 packet
*
****************************************************************************/
#ifdef CONFIG_NET_IPv4
void tcp_ipv4_select(FAR struct net_driver_s *dev);
#endif
/****************************************************************************
* Name: tcp_ipv6_select
*
* Description:
* Configure to send or receive an TCP IPv6 packet
*
****************************************************************************/
#ifdef CONFIG_NET_IPv6
void tcp_ipv6_select(FAR struct net_driver_s *dev);
#endif
/****************************************************************************
* Name: tcp_ip_select
*
* Description:
* Configure to send or receive an TCP IPv[4|6] packet for connection
*
****************************************************************************/
void tcp_ip_select(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_setsequence
*
* Description:
* Set the TCP/IP sequence number
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
void tcp_setsequence(FAR uint8_t *seqno, uint32_t value);
/****************************************************************************
* Name: tcp_getsequence
*
* Description:
* Get the TCP/IP sequence number
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
uint32_t tcp_getsequence(FAR uint8_t *seqno);
/****************************************************************************
* Name: tcp_addsequence
*
* Description:
* Add the length to get the next TCP sequence number.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
uint32_t tcp_addsequence(FAR uint8_t *seqno, uint16_t len);
/****************************************************************************
* Name: tcp_initsequence
*
* Description:
* Set the (initial) the TCP/IP sequence number when a TCP connection is
* established.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
void tcp_initsequence(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_nextsequence
*
* Description:
* Increment the TCP/IP sequence number
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
void tcp_nextsequence(void);
/****************************************************************************
* Name: tcp_poll
*
* Description:
* Poll a TCP connection structure for availability of TX data
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP "connection" to poll for TX data
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
2014-06-28 00:48:12 +02:00
void tcp_poll(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn);
2014-06-25 02:12:49 +02:00
/****************************************************************************
* Name: tcp_timer
*
* Description:
* Handle a TCP timer expiration for the provided TCP connection
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP "connection" to poll for TX data
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
void tcp_timer(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_update_timer
*
* Description:
* Update the TCP timer for the provided TCP connection,
* The timeout is accurate
*
* Input Parameters:
* conn - The TCP "connection" to poll for TX data
*
* Returned Value:
* None
*
* Assumptions:
* conn is not NULL.
* The connection (conn) is bound to the polling device (dev).
*
****************************************************************************/
void tcp_update_timer(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_update_retrantimer
*
* Description:
* Update the retransmit TCP timer for the provided TCP connection,
* The timeout is accurate
*
* Input Parameters:
* conn - The TCP "connection" to poll for TX data
* timeout - Time for the next timeout
*
* Returned Value:
* None
*
* Assumptions:
* conn is not NULL.
* The connection (conn) is bound to the polling device (dev).
*
****************************************************************************/
void tcp_update_retrantimer(FAR struct tcp_conn_s *conn, int timeout);
/****************************************************************************
* Name: tcp_update_keeptimer
*
* Description:
* Update the keeplive TCP timer for the provided TCP connection,
* The timeout is accurate
*
* Input Parameters:
* conn - The TCP "connection" to poll for TX data
* timeout - Time for the next timeout
*
* Returned Value:
* None
*
* Assumptions:
* conn is not NULL.
* The connection (conn) is bound to the polling device (dev).
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_KEEPALIVE
void tcp_update_keeptimer(FAR struct tcp_conn_s *conn, int timeout);
#endif
/****************************************************************************
* Name: tcp_stop_timer
*
* Description:
* Stop TCP timer for the provided TCP connection
* When the connection is closed
*
* Input Parameters:
* conn - The TCP "connection" to poll for TX data
*
* Returned Value:
* None
*
* Assumptions:
* conn is not NULL.
*
****************************************************************************/
void tcp_stop_timer(FAR struct tcp_conn_s *conn);
2014-06-25 02:12:49 +02:00
/****************************************************************************
* Name: tcp_findlistener
*
* Description:
* Return the connection listener for connections on this port (if any)
*
* Assumptions:
* The network is locked
*
****************************************************************************/
#if defined(CONFIG_NET_IPv4) && defined(CONFIG_NET_IPv6)
FAR struct tcp_conn_s *tcp_findlistener(FAR union ip_binding_u *uaddr,
uint16_t portno,
uint8_t domain);
#else
FAR struct tcp_conn_s *tcp_findlistener(FAR union ip_binding_u *uaddr,
uint16_t portno);
#endif
/****************************************************************************
* Name: tcp_unlisten
*
* Description:
* Stop listening to the port bound to the specified TCP connection
*
* Assumptions:
* Called from normal user code.
*
****************************************************************************/
int tcp_unlisten(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_listen
*
* Description:
* Start listening to the port bound to the specified TCP connection
*
* Assumptions:
* Called from normal user code.
*
****************************************************************************/
int tcp_listen(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_islistener
*
* Description:
* Return true is there is a listener for the specified port
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
#if defined(CONFIG_NET_IPv4) && defined(CONFIG_NET_IPv6)
bool tcp_islistener(FAR union ip_binding_u *uaddr, uint16_t portno,
uint8_t domain);
#else
bool tcp_islistener(FAR union ip_binding_u *uaddr, uint16_t portno);
#endif
/****************************************************************************
* Name: tcp_accept_connection
*
* Description:
* Accept the new connection for the specified listening port.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
2014-06-28 00:48:12 +02:00
int tcp_accept_connection(FAR struct net_driver_s *dev,
2014-06-25 02:12:49 +02:00
FAR struct tcp_conn_s *conn, uint16_t portno);
/****************************************************************************
* Name: tcp_send
*
* Description:
* Setup to send a TCP packet
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP connection structure holding connection information
* flags - flags to apply to the TCP header
* len - length of the message
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
2014-06-28 00:48:12 +02:00
void tcp_send(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn,
2014-06-25 02:12:49 +02:00
uint16_t flags, uint16_t len);
/****************************************************************************
* Name: tcp_sendfile
*
* Description:
* The tcp_sendfile() call may be used only when the INET socket is in a
* connected state (so that the intended recipient is known).
*
* Input Parameters:
* psock An instance of the internal socket structure.
* buf Data to send
* len Length of data to send
* flags Send flags
*
* Returned Value:
* On success, returns the number of characters sent. On error,
* a negated errno value is returned. See sendfile() for a list
* appropriate error return values.
*
****************************************************************************/
#ifdef CONFIG_NET_SENDFILE
ssize_t tcp_sendfile(FAR struct socket *psock, FAR struct file *infile,
FAR off_t *offset, size_t count);
#endif
/****************************************************************************
* Name: tcp_reset
*
* Description:
* Send a TCP reset (no-data) message
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP connection structure holding connection information
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
void tcp_reset(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_rx_mss
*
* Description:
* Return the MSS to advertize to the peer.
*
* Input Parameters:
* dev - The device driver structure
*
* Returned Value:
* The MSS value.
*
****************************************************************************/
uint16_t tcp_rx_mss(FAR struct net_driver_s *dev);
/****************************************************************************
* Name: tcp_synack
*
* Description:
* Send the SYN or SYNACK response.
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP connection structure holding connection information
* ack - The ACK response to send
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
void tcp_synack(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn,
uint8_t ack);
2014-06-25 02:12:49 +02:00
/****************************************************************************
* Name: tcp_appsend
*
* Description:
* Handle application or TCP protocol response. If this function is called
* with dev->d_sndlen > 0, then this is an application attempting to send
* packet.
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP connection structure holding connection information
* result - App result event sent
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
2014-06-28 00:48:12 +02:00
void tcp_appsend(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn,
2014-06-25 02:12:49 +02:00
uint16_t result);
/****************************************************************************
* Name: tcp_rexmit
*
* Description:
* Handle application retransmission
*
* Input Parameters:
* dev - The device driver structure to use in the send operation
* conn - The TCP connection structure holding connection information
* result - App result event sent
*
* Returned Value:
* None
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-28 00:48:12 +02:00
void tcp_rexmit(FAR struct net_driver_s *dev, FAR struct tcp_conn_s *conn,
2014-06-25 02:12:49 +02:00
uint16_t result);
/****************************************************************************
* Name: tcp_send_txnotify
*
* Description:
* Notify the appropriate device driver that we are have data ready to
* be send (TCP)
*
* Input Parameters:
* psock - Socket state structure
* conn - The TCP connection structure
*
* Returned Value:
* None
*
****************************************************************************/
void tcp_send_txnotify(FAR struct socket *psock,
FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_ipv4_input
*
* Description:
* Handle incoming TCP input with IPv4 header
*
* Input Parameters:
* dev - The device driver structure containing the received TCP packet.
*
* Returned Value:
* None
*
* Assumptions:
* Called from the Ethernet driver with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
#ifdef CONFIG_NET_IPv4
void tcp_ipv4_input(FAR struct net_driver_s *dev);
#endif
/****************************************************************************
* Name: tcp_ipv6_input
*
* Description:
* Handle incoming TCP input with IPv4 header
*
* Input Parameters:
* dev - The device driver structure containing the received TCP packet.
* iplen - The size of the IPv6 header. This may be larger than
* IPv6_HDRLEN the IPv6 header if IPv6 extension headers are
* present.
*
* Returned Value:
* None
*
* Assumptions:
* Called from the Ethernet driver with the network stack locked
*
****************************************************************************/
#ifdef CONFIG_NET_IPv6
void tcp_ipv6_input(FAR struct net_driver_s *dev, unsigned int iplen);
#endif
2014-06-25 02:12:49 +02:00
/****************************************************************************
* Name: tcp_callback
*
* Description:
* Inform the application holding the TCP socket of a change in state.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
2014-06-28 00:48:12 +02:00
uint16_t tcp_callback(FAR struct net_driver_s *dev,
2014-06-25 02:12:49 +02:00
FAR struct tcp_conn_s *conn, uint16_t flags);
/****************************************************************************
* Name: tcp_datahandler
*
* Description:
* Handle data that is not accepted by the application. This may be called
* either (1) from the data receive logic if it cannot buffer the data, or
* (2) from the TCP event logic is there is no listener in place ready to
* receive the data.
*
* Input Parameters:
* conn - A pointer to the TCP connection structure
* buffer - A pointer to the buffer to be copied to the read-ahead
* buffers
* buflen - The number of bytes to copy to the read-ahead buffer.
*
* Returned Value:
* The number of bytes actually buffered is returned. This will be either
* zero or equal to buflen; partial packets are not buffered.
*
* Assumptions:
* - The caller has checked that TCP_NEWDATA is set in flags and that is no
* other handler available to process the incoming data.
* - Called from network stack logic with the network stack locked
*
****************************************************************************/
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
uint16_t tcp_datahandler(FAR struct net_driver_s *dev,
FAR struct tcp_conn_s *conn,
uint16_t offset);
/****************************************************************************
* Name: tcp_backlogcreate
*
* Description:
* Called from the listen() logic to setup the backlog as specified in the
* the listen arguments.
*
* Assumptions:
* Called from network socket logic. The network may or may not be locked.
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
int tcp_backlogcreate(FAR struct tcp_conn_s *conn, int nblg);
#else
# define tcp_backlogcreate(c,n) (-ENOSYS)
#endif
/****************************************************************************
* Name: tcp_backlogdestroy
*
* Description:
* (1) Called from tcp_free() whenever a connection is freed.
* (2) Called from tcp_backlogcreate() to destroy any old backlog
*
* NOTE: This function may re-enter tcp_free when a connection that
* is freed that has pending connections.
*
* Assumptions:
* Called from network socket logic with the network stack locked
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
int tcp_backlogdestroy(FAR struct tcp_conn_s *conn);
#else
# define tcp_backlogdestroy(conn) (-ENOSYS)
#endif
/****************************************************************************
* Name: tcp_backlogadd
*
* Description:
* Called tcp_listen when a new connection is made with a listener socket
* but when there is no accept() in place to receive the connection. This
* function adds the new connection to the backlog.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
int tcp_backlogadd(FAR struct tcp_conn_s *conn,
FAR struct tcp_conn_s *blconn);
#else
# define tcp_backlogadd(conn,blconn) (-ENOSYS)
#endif
/****************************************************************************
* Name: tcp_backlogpending
*
* Description:
* Called from poll(). Before waiting for a new connection, poll will
* call this API to see if there are pending connections in the backlog.
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
bool tcp_backlogpending(FAR struct tcp_conn_s *conn);
#else
# define tcp_backlogpending(c) (false)
#endif
/****************************************************************************
* Name: tcp_backlogavailable
*
* Description:
* Called from tcp_input(). Before alloc a new accept connection, tcp_input
* will call this API to see if there are free node in the backlog.
*
* Assumptions:
* Called from network socket logic with the network locked
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
bool tcp_backlogavailable(FAR struct tcp_conn_s *conn);
#else
# define tcp_backlogavailable(c) (true)
#endif
/****************************************************************************
* Name: tcp_backlogremove
*
* Description:
* Called from accept(). Before waiting for a new connection, accept will
* call this API to see if there are pending connections in the backlog.
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
FAR struct tcp_conn_s *tcp_backlogremove(FAR struct tcp_conn_s *conn);
#else
# define tcp_backlogremove(c) (NULL)
#endif
/****************************************************************************
* Name: tcp_backlogdelete
*
* Description:
* Called from tcp_free() when a connection is freed that this also
* retained in the pending connection list of a listener. We simply need
* to remove the defunct connection from the list.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
#ifdef CONFIG_NET_TCPBACKLOG
int tcp_backlogdelete(FAR struct tcp_conn_s *conn,
FAR struct tcp_conn_s *blconn);
#else
# define tcp_backlogdelete(c,b) (-ENOSYS)
2014-06-25 02:12:49 +02:00
#endif
/****************************************************************************
* Name: tcp_accept
*
* Description:
* This function implements accept() for TCP/IP sockets. See the
* description of accept() for further information.
*
* Input Parameters:
* psock The listening TCP socket structure
* addr Receives the address of the connecting client
* addrlen Input: allocated size of 'addr'
* Return: returned size of 'addr'
* newconn The new, accepted TCP connection structure
*
* Returned Value:
* Returns zero (OK) on success or a negated errno value on failure.
* See the description of accept of the possible errno values in the
* description of accept().
*
* Assumptions:
* Network is locked.
*
****************************************************************************/
int psock_tcp_accept(FAR struct socket *psock, FAR struct sockaddr *addr,
FAR socklen_t *addrlen, FAR void **newconn);
/****************************************************************************
* Name: psock_tcp_recvfrom
*
* Description:
* Perform the recvfrom operation for a TCP/IP SOCK_STREAM
*
* Input Parameters:
* psock Pointer to the socket structure for the SOCK_DRAM socket
* msg Receive info and buffer for receive data
* flags Receive flags
*
* Returned Value:
* On success, returns the number of characters received. On error,
* -errno is returned (see recvfrom for list of errnos).
*
* Assumptions:
*
****************************************************************************/
ssize_t psock_tcp_recvfrom(FAR struct socket *psock, FAR struct msghdr *msg,
int flags);
/****************************************************************************
* Name: psock_tcp_send
*
* Description:
2014-06-25 02:12:49 +02:00
* The psock_tcp_send() call may be used only when the TCP socket is in a
* connected state (so that the intended recipient is known).
*
* Input Parameters:
* psock An instance of the internal socket structure.
* buf Data to send
* len Length of data to send
* flags Send flags
*
* Returned Value:
* On success, returns the number of characters sent. On error,
* -1 is returned, and errno is set appropriately:
*
* EAGAIN or EWOULDBLOCK
* The socket is marked non-blocking and the requested operation
* would block.
* EBADF
* An invalid descriptor was specified.
* ECONNRESET
* Connection reset by peer.
* EDESTADDRREQ
* The socket is not connection-mode, and no peer address is set.
* EFAULT
* An invalid user space address was specified for a parameter.
* EINTR
* A signal occurred before any data was transmitted.
* EINVAL
* Invalid argument passed.
* EISCONN
* The connection-mode socket was connected already but a recipient
* was specified. (Now either this error is returned, or the recipient
* specification is ignored.)
* EMSGSIZE
* The socket type requires that message be sent atomically, and the
* size of the message to be sent made this impossible.
* ENOBUFS
* The output queue for a network interface was full. This generally
* indicates that the interface has stopped sending, but may be
* caused by transient congestion.
* ENOMEM
* No memory available.
* ENOTCONN
* The socket is not connected, and no target has been given.
* ENOTSOCK
* The argument s is not a socket.
* EPIPE
* The local end has been shut down on a connection oriented socket.
* In this case the process will also receive a SIGPIPE unless
* MSG_NOSIGNAL is set.
*
* Assumptions:
*
****************************************************************************/
2014-06-25 02:12:49 +02:00
ssize_t psock_tcp_send(FAR struct socket *psock, FAR const void *buf,
size_t len, int flags);
/****************************************************************************
* Name: tcp_setsockopt
*
* Description:
* tcp_setsockopt() sets the TCP-protocol option specified by the
* 'option' argument to the value pointed to by the 'value' argument for
* the socket specified by the 'psock' argument.
*
* See <netinet/tcp.h> for the a complete list of values of TCP protocol
* options.
*
* Input Parameters:
* psock Socket structure of socket to operate on
* option identifies the option to set
* value Points to the argument value
* value_len The length of the argument value
*
* Returned Value:
* Returns zero (OK) on success. On failure, it returns a negated errno
* value to indicate the nature of the error. See psock_setcockopt() for
* the list of possible error values.
*
****************************************************************************/
#ifdef CONFIG_NET_TCPPROTO_OPTIONS
int tcp_setsockopt(FAR struct socket *psock, int option,
FAR const void *value, socklen_t value_len);
#endif
/****************************************************************************
* Name: tcp_getsockopt
*
* Description:
* tcp_getsockopt() retrieves the value for the option specified by the
* 'option' argument for the socket specified by the 'psock' argument. If
* the size of the option value is greater than 'value_len', the value
* stored in the object pointed to by the 'value' argument will be silently
* truncated. Otherwise, the length pointed to by the 'value_len' argument
* will be modified to indicate the actual length of the 'value'.
*
* The 'level' argument specifies the protocol level of the option. To
* retrieve options at the socket level, specify the level argument as
* SOL_SOCKET; to retrieve options at the TCP-protocol level, the level
* argument is SOL_TCP.
*
* See <sys/socket.h> a complete list of values for the socket-level
* 'option' argument. Protocol-specific options are are protocol specific
* header files (such as netinet/tcp.h for the case of the TCP protocol).
*
* Input Parameters:
* psock Socket structure of the socket to query
* level Protocol level to set the option
* option identifies the option to get
* value Points to the argument value
* value_len The length of the argument value
*
* Returned Value:
* Returns zero (OK) on success. On failure, it returns a negated errno
* value to indicate the nature of the error. See psock_getsockopt() for
* the complete list of appropriate return error codes.
*
****************************************************************************/
#ifdef CONFIG_NET_TCPPROTO_OPTIONS
int tcp_getsockopt(FAR struct socket *psock, int option,
FAR void *value, FAR socklen_t *value_len);
#endif
/****************************************************************************
* Name: tcp_get_recvwindow
*
* Description:
* Calculate the TCP receive window for the specified device.
*
* Input Parameters:
* dev - The device whose TCP receive window will be updated.
* conn - The TCP connection structure holding connection information.
*
* Returned Value:
* The value of the TCP receive window to use.
*
****************************************************************************/
uint32_t tcp_get_recvwindow(FAR struct net_driver_s *dev,
FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_should_send_recvwindow
*
* Description:
* Determine if we should advertize the new recv window to the peer.
*
* Input Parameters:
* conn - The TCP connection structure holding connection information.
*
* Returned Value:
* If we should send an update.
*
****************************************************************************/
bool tcp_should_send_recvwindow(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: psock_tcp_cansend
*
* Description:
* psock_tcp_cansend() returns a value indicating if a write to the socket
* would block. No space in the buffer is actually reserved, so it is
* possible that the write may still block if the buffer is filled by
* another means.
*
* Input Parameters:
* conn The TCP connection of interest
*
* Returned Value:
* OK
* At least one byte of data could be successfully written.
* -EWOULDBLOCK
* There is no room in the output buffer.
* -EBADF
* An invalid descriptor was specified.
* -ENOTCONN
* The socket is not connected.
*
****************************************************************************/
int psock_tcp_cansend(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_wrbuffer_initialize
*
* Description:
* Initialize the list of free write buffers
*
* Assumptions:
* Called once early initialization.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
void tcp_wrbuffer_initialize(void);
#endif /* CONFIG_NET_TCP_WRITE_BUFFERS */
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
struct tcp_wrbuffer_s;
/****************************************************************************
* Name: tcp_wrbuffer_timedalloc
*
* Description:
* Allocate a TCP write buffer by taking a pre-allocated buffer from
* the free list. This function is called from TCP logic when a buffer
* of TCP data is about to sent
* This function is wrapped version of tcp_wrbuffer_alloc(),
* this wait will be terminated when the specified timeout expires.
*
* Input Parameters:
* timeout - The relative time to wait until a timeout is declared.
*
* Assumptions:
* Called from user logic with the network locked.
*
****************************************************************************/
FAR struct tcp_wrbuffer_s *tcp_wrbuffer_timedalloc(unsigned int timeout);
/****************************************************************************
* Name: tcp_wrbuffer_alloc
*
* Description:
* Allocate a TCP write buffer by taking a pre-allocated buffer from
* the free list. This function is called from TCP logic when a buffer
* of TCP data is about to sent
*
* Input Parameters:
* None
*
* Assumptions:
* Called from user logic with the network locked.
*
****************************************************************************/
FAR struct tcp_wrbuffer_s *tcp_wrbuffer_alloc(void);
/****************************************************************************
* Name: tcp_wrbuffer_tryalloc
*
* Description:
* Try to allocate a TCP write buffer by taking a pre-allocated buffer from
* the free list. This function is called from TCP logic when a buffer
* of TCP data is about to be sent if the socket is non-blocking. Returns
* immediately if allocation fails.
*
* Input parameters:
* None
*
* Assumptions:
* Called from user logic with the network locked.
*
****************************************************************************/
FAR struct tcp_wrbuffer_s *tcp_wrbuffer_tryalloc(void);
#endif /* CONFIG_NET_TCP_WRITE_BUFFERS */
/****************************************************************************
* Name: tcp_wrbuffer_release
*
* Description:
* Release a TCP write buffer by returning the buffer to the free list.
* This function is called from user logic after it is consumed the
* buffered data.
*
* Assumptions:
* Called from network stack logic with the network stack locked
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
void tcp_wrbuffer_release(FAR struct tcp_wrbuffer_s *wrb);
#endif /* CONFIG_NET_TCP_WRITE_BUFFERS */
/****************************************************************************
* Name: tcp_wrbuffer_inqueue_size
*
* Description:
* Get the in-queued write buffer size from connection
*
* Input Parameters:
* conn - The TCP connection of interest
*
* Assumptions:
* Called from user logic with the network locked.
*
****************************************************************************/
#if CONFIG_NET_SEND_BUFSIZE > 0
uint32_t tcp_wrbuffer_inqueue_size(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_wrbuffer_test
*
* Description:
* Check if there is room in the write buffer. Does not reserve any space.
*
* Assumptions:
* None.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
int tcp_wrbuffer_test(void);
#endif /* CONFIG_NET_TCP_WRITE_BUFFERS */
/****************************************************************************
* Name: tcp_event_handler_dump
*
* Description:
* Dump the TCP event handler related variables
*
****************************************************************************/
#ifdef CONFIG_DEBUG_FEATURES
void tcp_event_handler_dump(FAR struct net_driver_s *dev,
FAR void *pvpriv,
uint16_t flags,
FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_wrbuffer_dump
*
* Description:
* Dump the contents of a write buffer.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_WRITE_BUFFERS
#ifdef CONFIG_DEBUG_FEATURES
void tcp_wrbuffer_dump(FAR const char *msg, FAR struct tcp_wrbuffer_s *wrb,
unsigned int len, unsigned int offset);
#else
# define tcp_wrbuffer_dump(msg,wrb)
#endif
#endif /* CONFIG_NET_TCP_WRITE_BUFFERS */
/****************************************************************************
* Name: tcp_pollsetup
*
* Description:
* Setup to monitor events on one TCP/IP socket
*
* Input Parameters:
* psock - The TCP/IP socket of interest
* fds - The structure describing the events to be monitored, OR NULL if
* this is a request to stop monitoring events.
*
* Returned Value:
* 0: Success; Negated errno on failure
*
****************************************************************************/
int tcp_pollsetup(FAR struct socket *psock, FAR struct pollfd *fds);
/****************************************************************************
* Name: tcp_pollteardown
*
* Description:
* Teardown monitoring of events on an TCP/IP socket
*
* Input Parameters:
* psock - The TCP/IP socket of interest
* fds - The structure describing the events to be monitored, OR NULL if
* this is a request to stop monitoring events.
*
* Returned Value:
* 0: Success; Negated errno on failure
*
****************************************************************************/
int tcp_pollteardown(FAR struct socket *psock, FAR struct pollfd *fds);
/****************************************************************************
* Name: tcp_readahead_notifier_setup
*
* Description:
* Set up to perform a callback to the worker function when an TCP data
* is added to the read-ahead buffer. The worker function will execute
* on the high priority worker thread.
*
* Input Parameters:
* worker - The worker function to execute on the high priority work
* queue when data is available in the TCP read-ahead buffer.
* conn - The TCP connection where read-ahead data is needed.
* arg - A user-defined argument that will be available to the worker
* function when it runs.
*
* Returned Value:
* > 0 - The signal notification is in place. The returned value is a
* key that may be used later in a call to
* tcp_notifier_teardown().
* == 0 - There is already buffered read-ahead data. No signal
* notification will be provided.
* < 0 - An unexpected error occurred and no signal will be sent. The
* returned value is a negated errno value that indicates the
* nature of the failure.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
int tcp_readahead_notifier_setup(worker_t worker,
FAR struct tcp_conn_s *conn,
FAR void *arg);
#endif
/****************************************************************************
* Name: tcp_writebuffer_notifier_setup
*
* Description:
* Set up to perform a callback to the worker function when an TCP write
* buffer is emptied. The worker function will execute on the high
* priority worker thread.
*
* Input Parameters:
* worker - The worker function to execute on the high priority work
* queue when all buffer TX data has been sent.
* conn - The TCP connection where buffer write data is pending.
* arg - A user-defined argument that will be available to the worker
* function when it runs.
*
* Returned Value:
* > 0 - The signal notification is in place. The returned value is a
* key that may be used later in a call to
* tcp_notifier_teardown().
* == 0 - There is already buffered read-ahead data. No signal
* notification will be provided.
* < 0 - An unexpected error occurred and no signal will be sent. The
* returned value is a negated errno value that indicates the
* nature of the failure.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
int tcp_writebuffer_notifier_setup(worker_t worker,
FAR struct tcp_conn_s *conn,
FAR void *arg);
#endif
/****************************************************************************
* Name: tcp_disconnect_notifier_setup
*
* Description:
* Set up to perform a callback to the worker function if the TCP
* connection is lost.
*
* Input Parameters:
* worker - The worker function to execute on the high priority work
* queue when data is available in the TCP read-ahead buffer.
* conn - The TCP connection where read-ahead data is needed.
* arg - A user-defined argument that will be available to the worker
* function when it runs.
*
* Returned Value:
* > 0 - The signal notification is in place. The returned value is a
* key that may be used later in a call to
* tcp_notifier_teardown().
* == 0 - No connection has been established.
* < 0 - An unexpected error occurred and no signal will be sent. The
* returned value is a negated errno value that indicates the
* nature of the failure.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
int tcp_disconnect_notifier_setup(worker_t worker,
FAR struct tcp_conn_s *conn,
FAR void *arg);
#endif
/****************************************************************************
* Name: tcp_notifier_teardown
*
* Description:
* Eliminate a TCP read-ahead notification previously setup by
* tcp_readahead_notifier_setup(). This function should only be called
* if the notification should be aborted prior to the notification. The
* notification will automatically be torn down after the signal is sent.
*
* Input Parameters:
* key - The key value returned from a previous call to
* tcp_readahead_notifier_setup().
*
* Returned Value:
* None.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
void tcp_notifier_teardown(int key);
#endif
/****************************************************************************
* Name: tcp_readahead_signal
*
* Description:
* Read-ahead data has been buffered. Signal all threads waiting for
* read-ahead data to become available.
*
* When read-ahead data becomes available, *all* of the workers waiting
* for read-ahead data will be executed. If there are multiple workers
* waiting for read-ahead data then only the first to execute will get the
* data. Others will need to call tcp_readahead_notifier_setup() once
* again.
*
* Input Parameters:
* conn - The TCP connection where read-ahead data was just buffered.
*
* Returned Value:
* None.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
void tcp_readahead_signal(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_writebuffer_signal
*
* Description:
* All buffer Tx data has been sent. Signal all threads waiting for the
* write buffers to become empty.
*
* When write buffer becomes empty, *all* of the workers waiting
* for that event data will be executed. If there are multiple workers
* waiting for read-ahead data then only the first to execute will get the
* data. Others will need to call tcp_writebuffer_notifier_setup() once
* again.
*
* Input Parameters:
* conn - The TCP connection where read-ahead data was just buffered.
*
* Returned Value:
* None.
*
****************************************************************************/
#if defined(CONFIG_NET_TCP_WRITE_BUFFERS) && defined(CONFIG_NET_TCP_NOTIFIER)
void tcp_writebuffer_signal(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_disconnect_signal
*
* Description:
* The TCP connection has been lost. Signal all threads monitoring TCP
* state events.
*
* Input Parameters:
* conn - The TCP connection where read-ahead data was just buffered.
*
* Returned Value:
* None.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_NOTIFIER
void tcp_disconnect_signal(FAR struct tcp_conn_s *conn);
#endif
/****************************************************************************
* Name: tcp_txdrain
*
* Description:
* Wait for all write buffers to be sent (or for a timeout to occur).
*
* Input Parameters:
* psock - An instance of the internal socket structure.
* timeout - The relative time when the timeout will occur
*
* Returned Value:
* Zero (OK) is returned on success; a negated errno value is returned
* on any failure.
*
****************************************************************************/
#if defined(CONFIG_NET_TCP_WRITE_BUFFERS) && defined(CONFIG_NET_TCP_NOTIFIER)
int tcp_txdrain(FAR struct socket *psock, unsigned int timeout);
#else
# define tcp_txdrain(conn, timeout) (0)
#endif
/****************************************************************************
* Name: tcp_ioctl
*
* Description:
* This function performs tcp specific ioctl() operations.
*
* Parameters:
* conn The TCP connection of interest
* cmd The ioctl command
* arg The argument of the ioctl cmd
*
****************************************************************************/
int tcp_ioctl(FAR struct tcp_conn_s *conn, int cmd, unsigned long arg);
/****************************************************************************
* Name: tcp_sendbuffer_notify
*
* Description:
* Notify the send buffer semaphore
*
* Input Parameters:
* conn - The TCP connection of interest
*
* Assumptions:
* Called from user logic with the network locked.
*
****************************************************************************/
#if CONFIG_NET_SEND_BUFSIZE > 0
void tcp_sendbuffer_notify(FAR struct tcp_conn_s *conn);
#endif /* CONFIG_NET_SEND_BUFSIZE */
net/l2/l3/l4: add support of iob offload 1. Add new config CONFIG_NET_LL_GUARDSIZE to isolation of l2 stack, which will benefit l3(IP) layer for multi-MAC(l2) implementation, especially in some NICs such as celluler net driver. new configuration options: CONFIG_NET_LL_GUARDSIZE CONFIG_NET_LL_GUARDSIZE will reserved l2 buffer header size of network buffer to isolate the L2/L3 (MAC/IP) data on network layer, which will be beneficial to L3 network layer protocol transparent transmission and forwarding ------------------------------------------------------------ Layout of frist iob entry: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- iob | Reserved | io_len | ------------------------------------------------- ------------------------------------------------------------- Layout of different NICs implementation: iob_data (aligned by CONFIG_IOB_ALIGNMENT) | | io_offset(CONFIG_NET_LL_GUARDSIZE) | | ------------------------------------------------- Ethernet | Reserved | ETH_HDRLEN | io_len | ---------------------------------|--------------- 8021Q | Reserved | ETH_8021Q_HDRLEN | io_len | ---------------------------------|--------------- ipforward | Reserved | io_len | ------------------------------------------------- -------------------------------------------------------------------- 2. Support iob offload to l2 driver to avoid unnecessary memory copy Support send/receive iob vectors directly between the NICs and l3/l4 stack to avoid unnecessary memory copies, especially on hardware that supports Scatter/gather, which can greatly improve performance. new interface to support iob offload: ------------------------------------------ | IOB version | original | |----------------------------------------| | devif_iob_poll() | devif_poll() | | ... | ... | ------------------------------------------ -------------------------------------------------------------------- 1> NIC hardware support Scatter/gather transfer TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | dev->d_iob: | --------------- --------------- io_data iob1 | | | iob3 | | | \ --------------- --------------- --------------- | --------------- | iob0 | | | | iob2 | | | | --------------- | --------------- | \ | / / \ | / / ---------------------------------------------- NICs io vector | | | | | | | | | | ---------------------------------------------- RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to each iobs) -------------------------------------------------------------------- 2> CONFIG_IOB_BUFSIZE is greater than MTU: TX: "(CONFIG_IOB_BUFSIZE) > (MAX_NETDEV_PKTSIZE + CONFIG_NET_GUARDSIZE + CONFIG_NET_LL_GUARDSIZE)" tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll("NIC"_txpoll) callback() // "NIC"_txpoll | "NIC"_send() (dev->d_iob->io_data[CONFIG_NET_LL_GUARDSIZE - NET_LL_HDRLEN(dev)]) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | pkt/ipv[4/6]_input()/... | | NICs io vector receive(iov_base to io_data) -------------------------------------------------------------------- 3> Compatible with all old flat buffer NICs TX: tcp_poll()/udp_poll()/pkt_poll()/...(l3|l4) / \ / \ devif_poll_[l3|l4]_connections() devif_iob_send() (nocopy:udp/icmp/...) / \ (copy:tcp) / \ devif_iob_poll(devif_poll_callback()) devif_poll_callback() /* new interface, gather iobs to flat buffer */ / \ / \ devif_poll("NIC"_txpoll) "NIC"_send()(dev->d_buf) RX: [tcp|udp|icmp|...]ipv[4|6]_data_handler()(iob_concat/append to readahead) | | [tcp|udp|icmp|...]_ipv[4|6]_in()/... | | netdev_input() /* new interface, Scatter/gather flat/iob buffer */ | | pkt/ipv[4|6]_input()/... | | NICs io vector receive(Orignal flat buffer) 3. Iperf passthrough on NuttX simulator: ------------------------------------------------- | Protocol | Server | Client | | |-----------------------------------------------| | TCP | 813 | 834 | Mbits/sec | | TCP(Offload) | 1720 | 1100 | Mbits/sec | | UDP | 22 | 757 | Mbits/sec | | UDP(Offload) | 25 | 1250 | Mbits/sec | ------------------------------------------------- Signed-off-by: chao an <anchao@xiaomi.com>
2022-11-23 12:39:40 +01:00
/****************************************************************************
* Name: tcpip_hdrsize
*
* Description:
* Get the total size of L3 and L4 TCP header
*
* Input Parameters:
* conn The connection structure associated with the socket
*
* Returned Value:
* the total size of L3 and L4 TCP header
*
****************************************************************************/
uint16_t tcpip_hdrsize(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_ofoseg_bufsize
*
* Description:
* Calculate the pending size of out-of-order buffer
*
* Input Parameters:
* conn - The TCP connection of interest
*
* Returned Value:
* Total size of out-of-order buffer
*
* Assumptions:
* This function must be called with the network locked.
*
****************************************************************************/
int tcp_ofoseg_bufsize(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_reorder_ofosegs
*
* Description:
* Sort out-of-order segments by left edge
*
* Input Parameters:
* nofosegs - Number of out-of-order semgnets
* ofosegs - Pointer to out-of-order segments
*
* Returned Value:
* True if re-order occurs
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
bool tcp_reorder_ofosegs(int nofosegs, FAR struct tcp_ofoseg_s *ofosegs);
/****************************************************************************
* Name: tcp_cc_init
*
* Description:
* Initialize the congestion control variables, cwnd, ssthresh and dupacks.
* The function is called on starting a new connection.
*
* Input Parameters:
* conn - The TCP connection of interest
*
* Returned Value:
* None
*
* Assumptions:
* The normal user level code is calling the connect/accept to start a new
* connection.
*
****************************************************************************/
#ifdef CONFIG_NET_TCP_CC_NEWRENO
void tcp_cc_init(FAR struct tcp_conn_s *conn);
/****************************************************************************
* Name: tcp_cc_update
*
* Description:
* Update the congestion control variables when recieve the SYNACK/ACK
* packet from the peer in the connection phase.
*
* Input Parameters:
* conn - The TCP connection of interest
* tcp - The TCP header.
*
* Returned Value:
* None
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
void tcp_cc_update(FAR struct tcp_conn_s *conn, FAR struct tcp_hdr_s *tcp);
/****************************************************************************
* Name: tcp_cc_recv_ack
*
* Description:
* Update congestion control variables
*
* Input Parameters:
* conn - The TCP connection of interest
* tcp - The TCP header.
*
* Returned Value:
* None
*
* Assumptions:
* The network is locked.
*
****************************************************************************/
void tcp_cc_recv_ack(FAR struct tcp_conn_s *conn, FAR struct tcp_hdr_s *tcp);
#endif
#ifdef __cplusplus
}
#endif
/****************************************************************************
* Name: tcp_set_zero_probe
*
* Description:
* Update the TCP probe timer for the provided TCP connection,
* The timeout is accurate
*
* Input Parameters:
* conn - The TCP "connection" to poll for TX data
* flags - Set of connection events
*
* Returned Value:
* None
*
* Assumptions:
* conn is not NULL.
*
****************************************************************************/
void tcp_set_zero_probe(FAR struct tcp_conn_s *conn, uint16_t flags);
#endif /* !CONFIG_NET_TCP_NO_STACK */
#endif /* CONFIG_NET_TCP */
#endif /* __NET_TCP_TCP_H */