mirror/lwip
1
0
Fork 0
mirror of https://github.com/lwip-tcpip/lwip.git synced 2024-07-02 19:08:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

SLIP netif: add support for multiple input strategies (threaded, polling, RX from ISR)

Browse Source
This commit is contained in:
goldsimon 2011-10-21 22:25:44 +02:00
parent 2750d61e92
commit 3d1a306518
2 changed files with 303 additions and 131 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
src/include/netif/slipif.h
View File

@ -34,14 +34,44 @@
#ifndef __NETIF_SLIPIF_H__
#define __NETIF_SLIPIF_H__
#include "lwip/opt.h"
#include "lwip/netif.h"
/** Set this to 1 to start a thread that blocks reading on the serial line
* (using sio_read()).
*/
#ifndef SLIP_USE_RX_THREAD
#define SLIP_USE_RX_THREAD !NO_SYS
#endif
/** Set this to 1 to enable functions to pass in RX bytes from ISR context.
* If enabled, slipif_received_byte[s]() process incoming bytes and put assembled
* packets on a queue, which is fed into lwIP from slipif_poll().
* If disabled, slipif_poll() polls the serila line (using sio_tryread()).
*/
#ifndef SLIP_RX_FROM_ISR
#define SLIP_RX_FROM_ISR 0
#endif
/** Set this to 1 (default for SLIP_RX_FROM_ISR) to queue incoming packets
* received by slipif_received_byte[s]() as long as PBUF_POOL pbufs are available.
* If disabled, packets will be dropped if more than one packet is received.
*/
#ifndef SLIP_RX_QUEUE
#define SLIP_RX_QUEUE SLIP_RX_FROM_ISR
#endif
#ifdef __cplusplus
extern "C" {
#endif
err_t slipif_init(struct netif * netif);
void slipif_poll(struct netif *netif);
#if SLIP_RX_FROM_ISR
void slipif_process_rxqueue(struct netif *netif);
void slipif_received_byte(struct netif *netif, u8_t data);
void slipif_received_bytes(struct netif *netif, u8_t *data, u8_t len);
#endif /* SLIP_RX_FROM_ISR */
#ifdef __cplusplus
}

404
src/netif/slipif.c
View File

@ -35,6 +35,19 @@
* This file is built upon the file: src/arch/rtxc/netif/sioslip.c
*
* Author: Magnus Ivarsson <magnus.ivarsson(at)volvo.com>
* Simon Goldschmidt
*
* Usage: This netif can be used in three ways:
* 1) For NO_SYS==0, an RX thread can be used which blocks on sio_read()
* until data is received.
* 2) In your main loop, call slipif_poll() to check for new RX bytes,
* completed packets are fed into netif->input().
* 3) Call slipif_received_byte[s]() from your serial RX ISR and
* slipif_process_rxqueue() from your main loop. ISR level decodes
* packets and puts completed packets on a queue which is fed into
* the stack from the main loop (needs SYS_LIGHTWEIGHT_PROT for
* pbuf_alloc to work on ISR level!).
*
*/
/*
@ -52,19 +65,25 @@
#include "lwip/stats.h"
#include "lwip/snmp.h"
#include "lwip/sio.h"
#if !NO_SYS
#include "lwip/sys.h"
#define SLIP_END 0xC0 /* 0300: start and end of every packet */
#define SLIP_ESC 0xDB /* 0333: escape start (one byte escaped data follows) */
#define SLIP_ESC_END 0xDC /* 0334: following escape: original byte is 0xC0 (END) */
#define SLIP_ESC_ESC 0xDD /* 0335: following escape: original byte is 0xDB (ESC) */
/** Maximum packet size that is received by this netif */
#ifndef SLIP_MAX_SIZE
#define SLIP_MAX_SIZE 1500
#endif
#define SLIP_BLOCK 1
#define SLIP_DONTBLOCK 0
#define SLIP_END 0300 /* 0xC0 */
#define SLIP_ESC 0333 /* 0xDB */
#define SLIP_ESC_END 0334 /* 0xDC */
#define SLIP_ESC_ESC 0335 /* 0xDD */
#define SLIP_MAX_SIZE 1500
/** Define this to the interface speed for SNMP
* (sio_fd is the sio_fd_t returned by sio_open).
* The default value of zero means 'unknown'.
*/
#ifndef SLIP_SIO_SPEED
#define SLIP_SIO_SPEED(sio_fd) 0
#endif
enum slipif_recv_state {
SLIP_RECV_NORMAL,
@ -75,8 +94,11 @@ struct slipif_priv {
sio_fd_t sd;
/* q is the whole pbuf chain for a packet, p is the current pbuf in the chain */
struct pbuf *p, *q;
enum slipif_recv_state state;
u8_t state;
u16_t i, recved;
#if SLIP_RX_FROM_ISR
struct pbuf *rxpackets;
#endif
};
/**
@ -103,9 +125,11 @@ slipif_output(struct netif *netif, struct pbuf *p, ip_addr_t *ipaddr)
LWIP_UNUSED_ARG(ipaddr);
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_output(%"U16_F"): sending %"U16_F" bytes\n", (u16_t)netif->num, p->tot_len));
priv = netif->state;
/* Send pbuf out on the serial I/O device. */
/* Start with packet delimiter. */
sio_send(SLIP_END, priv->sd);
for (q = p; q != NULL; q = q->next) {
@ -113,59 +137,39 @@ slipif_output(struct netif *netif, struct pbuf *p, ip_addr_t *ipaddr)
c = ((u8_t *)q->payload)[i];
switch (c) {
case SLIP_END:
/* need to escape this byte (0xC0 -> 0xDB, 0xDC) */
sio_send(SLIP_ESC, priv->sd);
sio_send(SLIP_ESC_END, priv->sd);
break;
case SLIP_ESC:
/* need to escape this byte (0xDB -> 0xDB, 0xDD) */
sio_send(SLIP_ESC, priv->sd);
sio_send(SLIP_ESC_ESC, priv->sd);
break;
default:
/* normal byte - no need for escaping */
sio_send(c, priv->sd);
break;
}
}
}
/* End with packet delimiter. */
sio_send(SLIP_END, priv->sd);
return ERR_OK;
}
/**
* Static function for easy use of blockig or non-blocking
* sio_read
*
* @param fd serial device handle
* @param data pointer to data buffer for receiving
* @param len maximum length (in bytes) of data to receive
* @param block if 1, call sio_read; if 0, call sio_tryread
* @return return value of sio_read of sio_tryread
*/
static u32_t
slip_sio_read(sio_fd_t fd, u8_t* data, u32_t len, u8_t block)
{
if (block) {
return sio_read(fd, data, len);
} else {
return sio_tryread(fd, data, len);
}
}
/**
* Handle the incoming SLIP stream character by character
*
* Poll the serial layer by calling sio_read() or sio_tryread().
*
* @param netif the lwip network interface structure for this slipif
* @param block if 1, block until data is received; if 0, return when all data
* from the buffer is received (multiple calls to this function will
* @param c received character (multiple calls to this function will
* return a complete packet, NULL is returned before - used for polling)
* @return The IP packet when SLIP_END is received
*/
static struct pbuf *
slipif_input(struct netif *netif, u8_t block)
static struct pbuf*
slipif_rxbyte(struct netif *netif, u8_t c)
{
struct slipif_priv *priv;
u8_t c;
struct pbuf *t;
LWIP_ASSERT("netif != NULL", (netif != NULL));
@ -173,89 +177,105 @@ slipif_input(struct netif *netif, u8_t block)
priv = netif->state;
while (slip_sio_read(priv->sd, &c, 1, block) > 0) {
switch (priv->state) {
case SLIP_RECV_NORMAL:
switch (c) {
case SLIP_END:
if (priv->recved > 0) {
/* Received whole packet. */
/* Trim the pbuf to the size of the received packet. */
pbuf_realloc(priv->q, priv->recved);
switch (priv->state) {
case SLIP_RECV_NORMAL:
switch (c) {
case SLIP_END:
if (priv->recved > 0) {
/* Received whole packet. */
/* Trim the pbuf to the size of the received packet. */
pbuf_realloc(priv->q, priv->recved);
LINK_STATS_INC(link.recv);
LINK_STATS_INC(link.recv);
LWIP_DEBUGF(SLIP_DEBUG, ("slipif: Got packet\n"));
t = priv->q;
priv->p = priv->q = NULL;
priv->i = priv->recved = 0;
return t;
}
continue;
case SLIP_ESC:
priv->state = SLIP_RECV_ESCAPE;
continue;
LWIP_DEBUGF(SLIP_DEBUG, ("slipif: Got packet (%"U16_F" bytes)\n", priv->recved));
t = priv->q;
priv->p = priv->q = NULL;
priv->i = priv->recved = 0;
return t;
}
return NULL;
case SLIP_ESC:
priv->state = SLIP_RECV_ESCAPE;
return NULL;
} /* end switch (c) */
break;
case SLIP_RECV_ESCAPE:
/* un-escape END or ESC bytes, leave other bytes
(although that would be a protocol error) */
switch (c) {
case SLIP_ESC_END:
c = SLIP_END;
break;
case SLIP_ESC_ESC:
c = SLIP_ESC;
break;
case SLIP_RECV_ESCAPE:
switch (c) {
case SLIP_ESC_END:
c = SLIP_END;
break;
case SLIP_ESC_ESC:
c = SLIP_ESC;
break;
}
priv->state = SLIP_RECV_NORMAL;
/* FALLTHROUGH */
}
priv->state = SLIP_RECV_NORMAL;
break;
} /* end switch (priv->state) */
/* byte received, packet not yet completely received */
if (priv->p == NULL) {
/* allocate a new pbuf */
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_input: alloc\n"));
priv->p = pbuf_alloc(PBUF_LINK, (PBUF_POOL_BUFSIZE - PBUF_LINK_HLEN), PBUF_POOL);
/* byte received, packet not yet completely received */
if (priv->p == NULL) {
/* allocate a new pbuf */
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_input: alloc\n"));
priv->p = pbuf_alloc(PBUF_LINK, (PBUF_POOL_BUFSIZE - PBUF_LINK_HLEN), PBUF_POOL);
if (priv->p == NULL) {
LINK_STATS_INC(link.drop);
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_input: no new pbuf! (DROP)\n"));
/* don't process any further since we got no pbuf to receive to */
break;
}
if (priv->q != NULL) {
/* 'chain' the pbuf to the existing chain */
pbuf_cat(priv->q, priv->p);
} else {
/* p is the first pbuf in the chain */
priv->q = priv->p;
}
LINK_STATS_INC(link.drop);
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_input: no new pbuf! (DROP)\n"));
/* don't process any further since we got no pbuf to receive to */
return NULL;
}
/* this automatically drops bytes if > SLIP_MAX_SIZE */
if ((priv->p != NULL) && (priv->recved <= SLIP_MAX_SIZE)) {
((u8_t *)priv->p->payload)[priv->i] = c;
priv->recved++;
priv->i++;
if (priv->i >= priv->p->len) {
/* on to the next pbuf */
priv->i = 0;
if (priv->p->next != NULL && priv->p->next->len > 0) {
/* p is a chain, on to the next in the chain */
priv->p = priv->p->next;
} else {
/* p is a single pbuf, set it to NULL so next time a new
* pbuf is allocated */
priv->p = NULL;
}
}
if (priv->q != NULL) {
/* 'chain' the pbuf to the existing chain */
pbuf_cat(priv->q, priv->p);
} else {
/* p is the first pbuf in the chain */
priv->q = priv->p;
}
}
/* this automatically drops bytes if > SLIP_MAX_SIZE */
if ((priv->p != NULL) && (priv->recved <= SLIP_MAX_SIZE)) {
((u8_t *)priv->p->payload)[priv->i] = c;
priv->recved++;
priv->i++;
if (priv->i >= priv->p->len) {
/* on to the next pbuf */
priv->i = 0;
if (priv->p->next != NULL && priv->p->next->len > 0) {
/* p is a chain, on to the next in the chain */
priv->p = priv->p->next;
} else {
/* p is a single pbuf, set it to NULL so next time a new
* pbuf is allocated */
priv->p = NULL;
}
}
}
return NULL;
}
#if !NO_SYS
/** Like slipif_rxbyte, but passes completed packets to netif->input
*
* @param netif The lwip network interface structure for this slipif
* @param data received character
*/
static void
slipif_rxbyte_input(struct netif *netif, u8_t c)
{
struct pbuf *p;
p = slipif_rxbyte(netif, c);
if (p != NULL) {
if (netif->input(p, netif) != ERR_OK) {
pbuf_free(p);
}
}
}
#if SLIP_USE_RX_THREAD
/**
* The SLIP input thread.
*
@ -266,20 +286,17 @@ slipif_input(struct netif *netif, u8_t block)
static void
slipif_loop_thread(void *nf)
{
struct pbuf *p;
u8_t c;
struct netif *netif = (struct netif *)nf;
struct slipif_priv *priv = (struct slipif_priv *)netif->state;
while (1) {
p = slipif_input(netif, SLIP_BLOCK);
if (p != NULL) {
if (netif->input(p, netif) != ERR_OK) {
pbuf_free(p);
p = NULL;
}
if (sio_read(priv->sd, &c, 1) > 0) {
slipif_rxbyte_input(netif, c);
}
}
}
#endif /* !NO_SYS */
#endif /* SLIP_USE_RX_THREAD */
/**
* SLIP netif initialization
@ -293,17 +310,20 @@ slipif_loop_thread(void *nf)
* ERR_IF is serial line couldn't be opened
*
* @note netif->num must contain the number of the serial port to open
* (0 by default)
* (0 by default). If netif->state is != NULL, it is interpreted as an
* u8_t pointer pointing to the serial port number instead of netif->num.
*
*/
err_t
slipif_init(struct netif *netif)
{
struct slipif_priv *priv;
u8_t sio_num;
LWIP_DEBUGF(SLIP_DEBUG, ("slipif_init: netif->num=%"U16_F"\n", (u16_t)netif->num));
/* Allocate private data */
priv = mem_malloc(sizeof(struct slipif_priv));
priv = (struct slipif_priv *)mem_malloc(sizeof(struct slipif_priv));
if (!priv) {
return ERR_MEM;
}
@ -311,11 +331,20 @@ slipif_init(struct netif *netif)
netif->name[0] = 's';
netif->name[1] = 'l';
netif->output = slipif_output;
#if LWIP_IPV6
netif->output_ip6 = slipif_output;
#endif /* LWIP_IPV6 */
netif->mtu = SLIP_MAX_SIZE;
netif->flags |= NETIF_FLAG_POINTTOPOINT;
/* Try to open the serial port (netif->num contains the port number). */
priv->sd = sio_open(netif->num);
/* netif->state or netif->num contain the port number */
if (netif->state != NULL) {
sio_num = *(u8_t*)netif->state;
} else {
sio_num = netif->num;
}
/* Try to open the serial port. */
priv->sd = sio_open(sio_num);
if (!priv->sd) {
/* Opening the serial port failed. */
mem_free(priv);
@ -328,20 +357,20 @@ slipif_init(struct netif *netif)
priv->state = SLIP_RECV_NORMAL;
priv->i = 0;
priv->recved = 0;
#if SLIP_RX_FROM_ISR
priv->rxpackets = NULL;
#endif
netif->state = priv;
/* initialize the snmp variables and counters inside the struct netif
* ifSpeed: no assumption can be made without knowing more about the
* serial line!
*/
NETIF_INIT_SNMP(netif, snmp_ifType_slip, 0);
/* initialize the snmp variables and counters inside the struct netif */
NETIF_INIT_SNMP(netif, snmp_ifType_slip, SLIP_SIO_SPEED(priv->sd));
#if !NO_SYS
#if SLIP_USE_RX_THREAD
/* Create a thread to poll the serial line. */
sys_thread_new(SLIPIF_THREAD_NAME, slipif_loop_thread, netif,
SLIPIF_THREAD_STACKSIZE, SLIPIF_THREAD_PRIO);
#endif
#endif /* SLIP_USE_RX_THREAD */
return ERR_OK;
}
@ -353,19 +382,132 @@ slipif_init(struct netif *netif)
void
slipif_poll(struct netif *netif)
{
struct pbuf *p;
u8_t c;
struct slipif_priv *priv;
LWIP_ASSERT("netif != NULL", (netif != NULL));
LWIP_ASSERT("netif->state != NULL", (netif->state != NULL));
priv = netif->state;
priv = (struct slipif_priv *)netif->state;
while ((p = slipif_input(netif, SLIP_DONTBLOCK)) != NULL) {
if (netif->input(p, netif) != ERR_OK) {
pbuf_free(p);
}
while (sio_tryread(priv->sd, &c, 1) > 0) {
slipif_rxbyte_input(netif, c);
}
}
#if SLIP_RX_FROM_ISR
/**
* Feeds the IP layer with incoming packets that were receive
*
* @param netif The lwip network interface structure for this slipif
*/
void
slipif_process_rxqueue(struct netif *netif)
{
struct slipif_priv *priv;
SYS_ARCH_DECL_PROTECT(old_level);
LWIP_ASSERT("netif != NULL", (netif != NULL));
LWIP_ASSERT("netif->state != NULL", (netif->state != NULL));
priv = (struct slipif_priv *)netif->state;
SYS_ARCH_PROTECT(old_level);
while (priv->rxpackets != NULL) {
struct pbuf *p = priv->rxpackets;
#if SLIP_RX_QUEUE
/* dequeue packet */
struct pbuf *q = p;
while ((q->len != q->tot_len) && (q->next != NULL)) {
q = q->next;
}
priv->rxpackets = q->next;
q->next = NULL;
#else /* SLIP_RX_QUEUE */
priv->rxpackets = NULL;
#endif /* SLIP_RX_QUEUE */
SYS_ARCH_UNPROTECT(old_level);
if (netif->input(p, netif) != ERR_OK) {
pbuf_free(p);
}
SYS_ARCH_PROTECT(old_level);
}
}
/** Like slipif_rxbyte, but queues completed packets.
*
* @param netif The lwip network interface structure for this slipif
* @param data Received serial byte
*/
static void
slipif_rxbyte_enqueue(struct netif *netif, u8_t data)
{
struct pbuf *p;
struct slipif_priv *priv = (struct slipif_priv *)netif->state;
SYS_ARCH_DECL_PROTECT(old_level);
p = slipif_rxbyte(netif, data);
if (p != NULL) {
SYS_ARCH_PROTECT(old_level);
if (priv->rxpackets != NULL) {
#if SLIP_RX_QUEUE
/* queue multiple pbufs */
struct pbuf *q = p;
while(q->next != NULL) {
q = q->next;
}
q->next = p;
} else {
#else /* SLIP_RX_QUEUE */
pbuf_free(priv->rxpackets);
}
{
#endif /* SLIP_RX_QUEUE */
priv->rxpackets = p;
}
SYS_ARCH_UNPROTECT(old_level);
}
}
/**
* Process a received byte, completed packets are put on a queue that is
* fed into IP through slipif_process_rxqueue().
*
* This function can be called from ISR if SYS_LIGHTWEIGHT_PROT is enabled.
*
* @param netif The lwip network interface structure for this slipif
* @param data received character
*/
void
slipif_received_byte(struct netif *netif, u8_t data)
{
LWIP_ASSERT("netif != NULL", (netif != NULL));
LWIP_ASSERT("netif->state != NULL", (netif->state != NULL));
slipif_rxbyte_enqueue(netif, data);
}
/**
* Process multiple received byte, completed packets are put on a queue that is
* fed into IP through slipif_process_rxqueue().
*
* This function can be called from ISR if SYS_LIGHTWEIGHT_PROT is enabled.
*
* @param netif The lwip network interface structure for this slipif
* @param data received character
* @param len Number of received characters
*/
void
slipif_received_bytes(struct netif *netif, u8_t *data, u8_t len)
{
u8_t i;
u8_t *rxdata = data;
LWIP_ASSERT("netif != NULL", (netif != NULL));
LWIP_ASSERT("netif->state != NULL", (netif->state != NULL));
for (i = 0; i < len; i++, rxdata++) {
slipif_rxbyte_enqueue(netif, *rxdata);
}
}
#endif /* SLIP_RX_FROM_ISR */
#endif /* LWIP_HAVE_SLIPIF */