1
0
mirror of synced 2024-11-16 21:43:50 +00:00
MassiveKnob/Arduino/MassiveKnob/min.c
2021-02-28 11:55:23 +01:00

642 lines
24 KiB
C

// Copyright (c) 2014-2017 JK Energy Ltd.
//
// Use authorized under the MIT license.
#include "min.h"
#define TRANSPORT_FIFO_SIZE_FRAMES_MASK ((uint8_t)((1U << TRANSPORT_FIFO_SIZE_FRAMES_BITS) - 1U))
#define TRANSPORT_FIFO_SIZE_FRAME_DATA_MASK ((uint16_t)((1U << TRANSPORT_FIFO_SIZE_FRAME_DATA_BITS) - 1U))
// Number of bytes needed for a frame with a given payload length, excluding stuff bytes
// 3 header bytes, ID/control byte, length byte, seq byte, 4 byte CRC, EOF byte
#define ON_WIRE_SIZE(p) ((p) + 11U)
// Special protocol bytes
enum {
HEADER_BYTE = 0xaaU,
STUFF_BYTE = 0x55U,
EOF_BYTE = 0x55U,
};
// Receiving state machine
enum {
SEARCHING_FOR_SOF,
RECEIVING_ID_CONTROL,
RECEIVING_SEQ,
RECEIVING_LENGTH,
RECEIVING_PAYLOAD,
RECEIVING_CHECKSUM_3,
RECEIVING_CHECKSUM_2,
RECEIVING_CHECKSUM_1,
RECEIVING_CHECKSUM_0,
RECEIVING_EOF,
};
#ifdef TRANSPORT_PROTOCOL
#ifndef TRANSPORT_ACK_RETRANSMIT_TIMEOUT_MS
#define TRANSPORT_ACK_RETRANSMIT_TIMEOUT_MS (25U)
#endif
#ifndef TRANSPORT_FRAME_RETRANSMIT_TIMEOUT_MS
#define TRANSPORT_FRAME_RETRANSMIT_TIMEOUT_MS (50U) // Should be long enough for a whole window to be transmitted plus an ACK / NACK to get back
#endif
#ifndef TRANSPORT_MAX_WINDOW_SIZE
#define TRANSPORT_MAX_WINDOW_SIZE (16U)
#endif
#ifndef TRANSPORT_IDLE_TIMEOUT_MS
#define TRANSPORT_IDLE_TIMEOUT_MS (1000U)
#endif
enum {
// Top bit must be set: these are for the transport protocol to use
// 0x7f and 0x7e are reserved MIN identifiers.
ACK = 0xffU,
RESET = 0xfeU,
};
// Where the payload data of the frame FIFO is stored
uint8_t payloads_ring_buffer[TRANSPORT_FIFO_MAX_FRAME_DATA];
static uint32_t now;
static void send_reset(struct min_context *self);
#endif
static void crc32_init_context(struct crc32_context *context)
{
context->crc = 0xffffffffU;
}
static void crc32_step(struct crc32_context *context, uint8_t byte)
{
context->crc ^= byte;
for(uint32_t j = 0; j < 8; j++) {
uint32_t mask = (uint32_t) -(context->crc & 1U);
context->crc = (context->crc >> 1) ^ (0xedb88320U & mask);
}
}
static uint32_t crc32_finalize(struct crc32_context *context)
{
return ~context->crc;
}
static void stuffed_tx_byte(struct min_context *self, uint8_t byte)
{
// Transmit the byte
min_tx_byte(self->port, byte);
crc32_step(&self->tx_checksum, byte);
// See if an additional stuff byte is needed
if(byte == HEADER_BYTE) {
if(--self->tx_header_byte_countdown == 0) {
min_tx_byte(self->port, STUFF_BYTE); // Stuff byte
self->tx_header_byte_countdown = 2U;
}
}
else {
self->tx_header_byte_countdown = 2U;
}
}
static void on_wire_bytes(struct min_context *self, uint8_t id_control, uint8_t seq, uint8_t *payload_base, uint16_t payload_offset, uint16_t payload_mask, uint8_t payload_len)
{
uint8_t n, i;
uint32_t checksum;
self->tx_header_byte_countdown = 2U;
crc32_init_context(&self->tx_checksum);
min_tx_start(self->port);
// Header is 3 bytes; because unstuffed will reset receiver immediately
min_tx_byte(self->port, HEADER_BYTE);
min_tx_byte(self->port, HEADER_BYTE);
min_tx_byte(self->port, HEADER_BYTE);
stuffed_tx_byte(self, id_control);
if(id_control & 0x80U) {
// Send the sequence number if it is a transport frame
stuffed_tx_byte(self, seq);
}
stuffed_tx_byte(self, payload_len);
for(i = 0, n = payload_len; n > 0; n--, i++) {
stuffed_tx_byte(self, payload_base[payload_offset]);
payload_offset++;
payload_offset &= payload_mask;
}
checksum = crc32_finalize(&self->tx_checksum);
// Network order is big-endian. A decent C compiler will spot that this
// is extracting bytes and will use efficient instructions.
stuffed_tx_byte(self, (uint8_t)((checksum >> 24) & 0xffU));
stuffed_tx_byte(self, (uint8_t)((checksum >> 16) & 0xffU));
stuffed_tx_byte(self, (uint8_t)((checksum >> 8) & 0xffU));
stuffed_tx_byte(self, (uint8_t)((checksum >> 0) & 0xffU));
// Ensure end-of-frame doesn't contain 0xaa and confuse search for start-of-frame
min_tx_byte(self->port, EOF_BYTE);
min_tx_finished(self->port);
}
#ifdef TRANSPORT_PROTOCOL
// Pops frame from front of queue, reclaims its ring buffer space
static void transport_fifo_pop(struct min_context *self)
{
#ifdef ASSERTION_CHECKING
assert(self->transport_fifo.n_frames != 0);
#endif
struct transport_frame *frame = &self->transport_fifo.frames[self->transport_fifo.head_idx];
min_debug_print("Popping frame id=%d seq=%d\n", frame->min_id, frame->seq);
#ifdef ASSERTION_CHECKING
assert(self->transport_fifo.n_ring_buffer_bytes >= frame->payload_len);
#endif
self->transport_fifo.n_frames--;
self->transport_fifo.head_idx++;
self->transport_fifo.head_idx &= TRANSPORT_FIFO_SIZE_FRAMES_MASK;
self->transport_fifo.n_ring_buffer_bytes -= frame->payload_len;
}
// Claim a buffer slot from the FIFO. Returns 0 if there is no space.
static struct transport_frame *transport_fifo_push(struct min_context *self, uint16_t data_size)
{
// A frame is only queued if there aren't too many frames in the FIFO and there is space in the
// data ring buffer.
struct transport_frame *ret = 0;
if (self->transport_fifo.n_frames < TRANSPORT_FIFO_MAX_FRAMES) {
// Is there space in the ring buffer for the frame payload?
if(self->transport_fifo.n_ring_buffer_bytes <= TRANSPORT_FIFO_MAX_FRAME_DATA - data_size) {
self->transport_fifo.n_frames++;
if (self->transport_fifo.n_frames > self->transport_fifo.n_frames_max) {
// High-water mark of FIFO (for diagnostic purposes)
self->transport_fifo.n_frames_max = self->transport_fifo.n_frames;
}
// Create FIFO entry
ret = &(self->transport_fifo.frames[self->transport_fifo.tail_idx]);
ret->payload_offset = self->transport_fifo.ring_buffer_tail_offset;
// Claim ring buffer space
self->transport_fifo.n_ring_buffer_bytes += data_size;
if(self->transport_fifo.n_ring_buffer_bytes > self->transport_fifo.n_ring_buffer_bytes_max) {
// High-water mark of ring buffer usage (for diagnostic purposes)
self->transport_fifo.n_ring_buffer_bytes_max = self->transport_fifo.n_ring_buffer_bytes;
}
self->transport_fifo.ring_buffer_tail_offset += data_size;
self->transport_fifo.ring_buffer_tail_offset &= TRANSPORT_FIFO_SIZE_FRAME_DATA_MASK;
// Claim FIFO space
self->transport_fifo.tail_idx++;
self->transport_fifo.tail_idx &= TRANSPORT_FIFO_SIZE_FRAMES_MASK;
}
else {
min_debug_print("No FIFO payload space: data_size=%d, n_ring_buffer_bytes=%d\n", data_size, self->transport_fifo.n_ring_buffer_bytes);
}
}
else {
min_debug_print("No FIFO frame slots\n");
}
return ret;
}
// Return the nth frame in the FIFO
static struct transport_frame *transport_fifo_get(struct min_context *self, uint8_t n)
{
uint8_t idx = self->transport_fifo.head_idx;
return &self->transport_fifo.frames[(idx + n) & TRANSPORT_FIFO_SIZE_FRAMES_MASK];
}
// Sends the given frame to the serial line
static void transport_fifo_send(struct min_context *self, struct transport_frame *frame)
{
min_debug_print("transport_fifo_send: min_id=%d, seq=%d, payload_len=%d\n", frame->min_id, frame->seq, frame->payload_len);
on_wire_bytes(self, frame->min_id | (uint8_t)0x80U, frame->seq, payloads_ring_buffer, frame->payload_offset, TRANSPORT_FIFO_SIZE_FRAME_DATA_MASK, frame->payload_len);
frame->last_sent_time_ms = now;
}
// We don't queue an ACK frame - we send it straight away (if there's space to do so)
static void send_ack(struct min_context *self)
{
// In the embedded end we don't reassemble out-of-order frames and so never ask for retransmits. Payload is
// always the same as the sequence number.
min_debug_print("send ACK: seq=%d\n", self->transport_fifo.rn);
if(ON_WIRE_SIZE(0) <= min_tx_space(self->port)) {
on_wire_bytes(self, ACK, self->transport_fifo.rn, &self->transport_fifo.rn, 0, 0xffU, 1U);
self->transport_fifo.last_sent_ack_time_ms = now;
}
}
// We don't queue an RESET frame - we send it straight away (if there's space to do so)
static void send_reset(struct min_context *self)
{
min_debug_print("send RESET\n");
if(ON_WIRE_SIZE(0) <= min_tx_space(self->port)) {
on_wire_bytes(self, RESET, 0, 0, 0, 0, 0);
}
}
static void transport_fifo_reset(struct min_context *self)
{
// Clear down the transmission FIFO queue
self->transport_fifo.n_frames = 0;
self->transport_fifo.head_idx = 0;
self->transport_fifo.tail_idx = 0;
self->transport_fifo.n_ring_buffer_bytes = 0;
self->transport_fifo.ring_buffer_tail_offset = 0;
self->transport_fifo.sn_max = 0;
self->transport_fifo.sn_min = 0;
self->transport_fifo.rn = 0;
// Reset the timers
self->transport_fifo.last_received_anything_ms = now;
self->transport_fifo.last_sent_ack_time_ms = now;
self->transport_fifo.last_received_frame_ms = 0;
}
void min_transport_reset(struct min_context *self, bool inform_other_side)
{
if (inform_other_side) {
// Tell the other end we have gone away
send_reset(self);
}
// Throw our frames away
transport_fifo_reset(self);
}
// Queues a MIN ID / payload frame into the outgoing FIFO
// API call.
// Returns true if the frame was queued OK.
bool min_queue_frame(struct min_context *self, uint8_t min_id, uint8_t *payload, uint8_t payload_len)
{
struct transport_frame *frame = transport_fifo_push(self, payload_len); // Claim a FIFO slot, reserve space for payload
// We are just queueing here: the poll() function puts the frame into the window and on to the wire
if(frame != 0) {
// Copy frame details into frame slot, copy payload into ring buffer
frame->min_id = min_id & (uint8_t)0x3fU;
frame->payload_len = payload_len;
uint16_t payload_offset = frame->payload_offset;
for(uint32_t i = 0; i < payload_len; i++) {
payloads_ring_buffer[payload_offset] = payload[i];
payload_offset++;
payload_offset &= TRANSPORT_FIFO_SIZE_FRAME_DATA_MASK;
}
min_debug_print("Queued ID=%d, len=%d\n", min_id, payload_len);
return true;
}
else {
self->transport_fifo.dropped_frames++;
return false;
}
}
bool min_queue_has_space_for_frame(struct min_context *self, uint8_t payload_len) {
return self->transport_fifo.n_frames < TRANSPORT_FIFO_MAX_FRAMES &&
self->transport_fifo.n_ring_buffer_bytes <= TRANSPORT_FIFO_MAX_FRAME_DATA - payload_len;
}
// Finds the frame in the window that was sent least recently
static struct transport_frame *find_retransmit_frame(struct min_context *self)
{
uint8_t window_size = self->transport_fifo.sn_max - self->transport_fifo.sn_min;
#ifdef ASSERTION_CHECKS
assert(window_size > 0);
assert(window_size <= self->transport_fifo.nframes);
#endif
// Start with the head of the queue and call this the oldest
struct transport_frame *oldest_frame = &self->transport_fifo.frames[self->transport_fifo.head_idx];
uint32_t oldest_elapsed_time = now - oldest_frame->last_sent_time_ms;
uint8_t idx = self->transport_fifo.head_idx;
for(uint8_t i = 0; i < window_size; i++) {
uint32_t elapsed = now - self->transport_fifo.frames[idx].last_sent_time_ms;
if(elapsed > oldest_elapsed_time) { // Strictly older only; otherwise the earlier frame is deemed the older
oldest_elapsed_time = elapsed;
oldest_frame = &self->transport_fifo.frames[idx];
}
idx++;
idx &= TRANSPORT_FIFO_SIZE_FRAMES_MASK;
}
return oldest_frame;
}
#endif // TRANSPORT_PROTOCOL
// This runs the receiving half of the transport protocol, acknowledging frames received, discarding
// duplicates received, and handling RESET requests.
static void valid_frame_received(struct min_context *self)
{
uint8_t id_control = self->rx_frame_id_control;
uint8_t *payload = self->rx_frame_payload_buf;
uint8_t payload_len = self->rx_control;
#ifdef TRANSPORT_PROTOCOL
uint8_t seq = self->rx_frame_seq;
uint8_t num_acked;
uint8_t num_nacked;
uint8_t num_in_window;
// When we receive anything we know the other end is still active and won't shut down
self->transport_fifo.last_received_anything_ms = now;
switch(id_control) {
case ACK:
// If we get an ACK then we remove all the acknowledged frames with seq < rn
// The payload byte specifies the number of NACKed frames: how many we want retransmitted because
// they have gone missing.
// But we need to make sure we don't accidentally ACK too many because of a stale ACK from an old session
num_acked = seq - self->transport_fifo.sn_min;
num_nacked = payload[0] - seq;
num_in_window = self->transport_fifo.sn_max - self->transport_fifo.sn_min;
if(num_acked <= num_in_window) {
self->transport_fifo.sn_min = seq;
#ifdef ASSERTION_CHECKING
assert(self->transport_fifo.n_frames >= num_in_window);
assert(num_in_window <= TRANSPORT_MAX_WINDOW_SIZE);
assert(num_nacked <= TRANSPORT_MAX_WINDOW_SIZE);
#endif
// Now pop off all the frames up to (but not including) rn
// The ACK contains Rn; all frames before Rn are ACKed and can be removed from the window
min_debug_print("Received ACK seq=%d, num_acked=%d, num_nacked=%d\n", seq, num_acked, num_nacked);
for(uint8_t i = 0; i < num_acked; i++) {
transport_fifo_pop(self);
}
uint8_t idx = self->transport_fifo.head_idx;
// Now retransmit the number of frames that were requested
for(uint8_t i = 0; i < num_nacked; i++) {
struct transport_frame *retransmit_frame = &self->transport_fifo.frames[idx];
transport_fifo_send(self, retransmit_frame);
idx++;
idx &= TRANSPORT_FIFO_SIZE_FRAMES_MASK;
}
}
else {
min_debug_print("Received spurious ACK seq=%d\n", seq);
self->transport_fifo.spurious_acks++;
}
break;
case RESET:
// If we get a RESET demand then we reset the transport protocol (empty the FIFO, reset the
// sequence numbers, etc.)
// We don't send anything, we just do it. The other end can send frames to see if this end is
// alive (pings, etc.) or just wait to get application frames.
self->transport_fifo.resets_received++;
transport_fifo_reset(self);
break;
default:
if (id_control & 0x80U) {
// Incoming application frames
// Reset the activity time (an idle connection will be stalled)
self->transport_fifo.last_received_frame_ms = now;
if (seq == self->transport_fifo.rn) {
// Accept this frame as matching the sequence number we were looking for
// Now looking for the next one in the sequence
self->transport_fifo.rn++;
// Always send an ACK back for the frame we received
// ACKs are short (should be about 9 microseconds to send on the wire) and
// this will cut the latency down.
// We also periodically send an ACK in case the ACK was lost, and in any case
// frames are re-sent.
send_ack(self);
// Now ready to pass this up to the application handlers
// Pass frame up to application handler to deal with
min_debug_print("Incoming app frame seq=%d, id=%d, payload len=%d\n", seq, id_control & (uint8_t)0x3fU, payload_len);
min_application_handler(id_control & (uint8_t)0x3fU, payload, payload_len, self->port);
} else {
// Discard this frame because we aren't looking for it: it's either a dupe because it was
// retransmitted when our ACK didn't get through in time, or else it's further on in the
// sequence and others got dropped.
self->transport_fifo.sequence_mismatch_drop++;
}
}
else {
// Not a transport frame
min_application_handler(id_control & (uint8_t)0x3fU, payload, payload_len, self->port);
}
break;
}
#else // TRANSPORT_PROTOCOL
min_application_handler(id_control & (uint8_t)0x3fU, payload, payload_len, self->port);
#endif // TRANSPORT_PROTOCOL
}
static void rx_byte(struct min_context *self, uint8_t byte)
{
// Regardless of state, three header bytes means "start of frame" and
// should reset the frame buffer and be ready to receive frame data
//
// Two in a row in over the frame means to expect a stuff byte.
uint32_t crc;
if(self->rx_header_bytes_seen == 2) {
self->rx_header_bytes_seen = 0;
if(byte == HEADER_BYTE) {
self->rx_frame_state = RECEIVING_ID_CONTROL;
return;
}
if(byte == STUFF_BYTE) {
/* Discard this byte; carry on receiving on the next character */
return;
}
else {
/* Something has gone wrong, give up on this frame and look for header again */
self->rx_frame_state = SEARCHING_FOR_SOF;
return;
}
}
if(byte == HEADER_BYTE) {
self->rx_header_bytes_seen++;
}
else {
self->rx_header_bytes_seen = 0;
}
switch(self->rx_frame_state) {
case SEARCHING_FOR_SOF:
break;
case RECEIVING_ID_CONTROL:
self->rx_frame_id_control = byte;
self->rx_frame_payload_bytes = 0;
crc32_init_context(&self->rx_checksum);
crc32_step(&self->rx_checksum, byte);
if(byte & 0x80U) {
#ifdef TRANSPORT_PROTOCOL
self->rx_frame_state = RECEIVING_SEQ;
#else
// If there is no transport support compiled in then all transport frames are ignored
self->rx_frame_state = SEARCHING_FOR_SOF;
#endif // TRANSPORT_PROTOCOL
}
else {
self->rx_frame_seq = 0;
self->rx_frame_state = RECEIVING_LENGTH;
}
break;
case RECEIVING_SEQ:
self->rx_frame_seq = byte;
crc32_step(&self->rx_checksum, byte);
self->rx_frame_state = RECEIVING_LENGTH;
break;
case RECEIVING_LENGTH:
self->rx_frame_length = byte;
self->rx_control = byte;
crc32_step(&self->rx_checksum, byte);
if(self->rx_frame_length > 0) {
// Can reduce the RAM size by compiling limits to frame sizes
if(self->rx_frame_length <= MAX_PAYLOAD) {
self->rx_frame_state = RECEIVING_PAYLOAD;
}
else {
// Frame dropped because it's longer than any frame we can buffer
self->rx_frame_state = SEARCHING_FOR_SOF;
}
}
else {
self->rx_frame_state = RECEIVING_CHECKSUM_3;
}
break;
case RECEIVING_PAYLOAD:
self->rx_frame_payload_buf[self->rx_frame_payload_bytes++] = byte;
crc32_step(&self->rx_checksum, byte);
if(--self->rx_frame_length == 0) {
self->rx_frame_state = RECEIVING_CHECKSUM_3;
}
break;
case RECEIVING_CHECKSUM_3:
self->rx_frame_checksum = ((uint32_t)byte) << 24;
self->rx_frame_state = RECEIVING_CHECKSUM_2;
break;
case RECEIVING_CHECKSUM_2:
self->rx_frame_checksum |= ((uint32_t)byte) << 16;
self->rx_frame_state = RECEIVING_CHECKSUM_1;
break;
case RECEIVING_CHECKSUM_1:
self->rx_frame_checksum |= ((uint32_t)byte) << 8;
self->rx_frame_state = RECEIVING_CHECKSUM_0;
break;
case RECEIVING_CHECKSUM_0:
self->rx_frame_checksum |= byte;
crc = crc32_finalize(&self->rx_checksum);
if(self->rx_frame_checksum != crc) {
// Frame fails the checksum and so is dropped
self->rx_frame_state = SEARCHING_FOR_SOF;
}
else {
// Checksum passes, go on to check for the end-of-frame marker
self->rx_frame_state = RECEIVING_EOF;
}
break;
case RECEIVING_EOF:
if(byte == 0x55u) {
// Frame received OK, pass up data to handler
valid_frame_received(self);
}
// else discard
// Look for next frame */
self->rx_frame_state = SEARCHING_FOR_SOF;
break;
default:
// Should never get here but in case we do then reset to a safe state
self->rx_frame_state = SEARCHING_FOR_SOF;
break;
}
}
// API call: sends received bytes into a MIN context and runs the transport timeouts
void min_poll(struct min_context *self, uint8_t *buf, uint32_t buf_len)
{
for(uint32_t i = 0; i < buf_len; i++) {
rx_byte(self, buf[i]);
}
#ifdef TRANSPORT_PROTOCOL
uint8_t window_size;
now = min_time_ms();
bool remote_connected = (now - self->transport_fifo.last_received_anything_ms < TRANSPORT_IDLE_TIMEOUT_MS);
bool remote_active = (now - self->transport_fifo.last_received_frame_ms < TRANSPORT_IDLE_TIMEOUT_MS);
// This sends one new frame or resends one old frame
window_size = self->transport_fifo.sn_max - self->transport_fifo.sn_min; // Window size
if((window_size < TRANSPORT_MAX_WINDOW_SIZE) && (self->transport_fifo.n_frames > window_size)) {
// There are new frames we can send; but don't even bother if there's no buffer space for them
struct transport_frame *frame = transport_fifo_get(self, window_size);
if(ON_WIRE_SIZE(frame->payload_len) <= min_tx_space(self->port)) {
frame->seq = self->transport_fifo.sn_max;
transport_fifo_send(self, frame);
// Move window on
self->transport_fifo.sn_max++;
}
}
else {
// Sender cannot send new frames so resend old ones (if there's anyone there)
if((window_size > 0) && remote_connected) {
// There are unacknowledged frames. Can re-send an old frame. Pick the least recently sent one.
struct transport_frame *oldest_frame = find_retransmit_frame(self);
if(now - oldest_frame->last_sent_time_ms >= TRANSPORT_FRAME_RETRANSMIT_TIMEOUT_MS) {
// Resending oldest frame if there's a chance there's enough space to send it
if(ON_WIRE_SIZE(oldest_frame->payload_len) <= min_tx_space(self->port)) {
transport_fifo_send(self, oldest_frame);
}
}
}
}
#ifndef DISABLE_TRANSPORT_ACK_RETRANSMIT
// Periodically transmit the ACK with the rn value, unless the line has gone idle
if(now - self->transport_fifo.last_sent_ack_time_ms > TRANSPORT_ACK_RETRANSMIT_TIMEOUT_MS) {
if(remote_active) {
send_ack(self);
}
}
#endif // DISABLE_TRANSPORT_ACK_RETRANSMIT
#endif // TRANSPORT_PROTOCOL
}
void min_init_context(struct min_context *self, uint8_t port)
{
// Initialize context
self->rx_header_bytes_seen = 0;
self->rx_frame_state = SEARCHING_FOR_SOF;
self->port = port;
#ifdef TRANSPORT_PROTOCOL
// Counters for diagnosis purposes
self->transport_fifo.spurious_acks = 0;
self->transport_fifo.sequence_mismatch_drop = 0;
self->transport_fifo.dropped_frames = 0;
self->transport_fifo.resets_received = 0;
self->transport_fifo.n_ring_buffer_bytes_max = 0;
self->transport_fifo.n_frames_max = 0;
transport_fifo_reset(self);
#endif // TRANSPORT_PROTOCOL
}
// Sends an application MIN frame on the wire (do not put into the transport queue)
void min_send_frame(struct min_context *self, uint8_t min_id, uint8_t *payload, uint8_t payload_len)
{
if((ON_WIRE_SIZE(payload_len) <= min_tx_space(self->port))) {
on_wire_bytes(self, min_id & (uint8_t) 0x3fU, 0, payload, 0, 0xffffU, payload_len);
}
}