741 lines
24 KiB
C++
741 lines
24 KiB
C++
|
/**
|
||
|
* Marlin 3D Printer Firmware
|
||
|
* Copyright (C) 2016 MarlinFirmware [https://github.com/MarlinFirmware/Marlin]
|
||
|
*
|
||
|
* Based on Sprinter and grbl.
|
||
|
* Copyright (C) 2011 Camiel Gubbels / Erik van der Zalm
|
||
|
*
|
||
|
* This program is free software: you can redistribute it and/or modify
|
||
|
* it under the terms of the GNU General Public License as published by
|
||
|
* the Free Software Foundation, either version 3 of the License, or
|
||
|
* (at your option) any later version.
|
||
|
*
|
||
|
* This program is distributed in the hope that it will be useful,
|
||
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
||
|
* GNU General Public License for more details.
|
||
|
*
|
||
|
* You should have received a copy of the GNU General Public License
|
||
|
* along with this program. If not, see <http://www.gnu.org/licenses/>.
|
||
|
*
|
||
|
*/
|
||
|
|
||
|
/**
|
||
|
* MarlinSerial.cpp - Hardware serial library for Wiring
|
||
|
* Copyright (c) 2006 Nicholas Zambetti. All right reserved.
|
||
|
*
|
||
|
* Modified 23 November 2006 by David A. Mellis
|
||
|
* Modified 28 September 2010 by Mark Sproul
|
||
|
* Modified 14 February 2016 by Andreas Hardtung (added tx buffer)
|
||
|
* Modified 01 October 2017 by Eduardo José Tagle (added XON/XOFF)
|
||
|
* Modified 10 June 2018 by Eduardo José Tagle (See #10991)
|
||
|
*/
|
||
|
|
||
|
// Disable HardwareSerial.cpp to support chips without a UART (Attiny, etc.)
|
||
|
|
||
|
#include "MarlinConfig.h"
|
||
|
|
||
|
#if USE_MARLINSERIAL && (defined(UBRRH) || defined(UBRR0H) || defined(UBRR1H) || defined(UBRR2H) || defined(UBRR3H))
|
||
|
|
||
|
#include "MarlinSerial.h"
|
||
|
#include "Marlin.h"
|
||
|
|
||
|
struct ring_buffer_r {
|
||
|
unsigned char buffer[RX_BUFFER_SIZE];
|
||
|
volatile ring_buffer_pos_t head, tail;
|
||
|
};
|
||
|
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
struct ring_buffer_t {
|
||
|
unsigned char buffer[TX_BUFFER_SIZE];
|
||
|
volatile uint8_t head, tail;
|
||
|
};
|
||
|
#endif
|
||
|
|
||
|
#if UART_PRESENT(SERIAL_PORT)
|
||
|
ring_buffer_r rx_buffer = { { 0 }, 0, 0 };
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
ring_buffer_t tx_buffer = { { 0 }, 0, 0 };
|
||
|
#endif
|
||
|
static bool _written;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_XON_XOFF)
|
||
|
constexpr uint8_t XON_XOFF_CHAR_SENT = 0x80, // XON / XOFF Character was sent
|
||
|
XON_XOFF_CHAR_MASK = 0x1F; // XON / XOFF character to send
|
||
|
// XON / XOFF character definitions
|
||
|
constexpr uint8_t XON_CHAR = 17, XOFF_CHAR = 19;
|
||
|
uint8_t xon_xoff_state = XON_XOFF_CHAR_SENT | XON_CHAR;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_DROPPED_RX)
|
||
|
uint8_t rx_dropped_bytes = 0;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_RX_BUFFER_OVERRUNS)
|
||
|
uint8_t rx_buffer_overruns = 0;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_RX_FRAMING_ERRORS)
|
||
|
uint8_t rx_framing_errors = 0;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_MAX_RX_QUEUED)
|
||
|
ring_buffer_pos_t rx_max_enqueued = 0;
|
||
|
#endif
|
||
|
|
||
|
// A SW memory barrier, to ensure GCC does not overoptimize loops
|
||
|
#define sw_barrier() asm volatile("": : :"memory");
|
||
|
|
||
|
#if ENABLED(EMERGENCY_PARSER)
|
||
|
#include "emergency_parser.h"
|
||
|
#endif
|
||
|
|
||
|
// "Atomically" read the RX head index value without disabling interrupts:
|
||
|
// This MUST be called with RX interrupts enabled, and CAN'T be called
|
||
|
// from the RX ISR itself!
|
||
|
FORCE_INLINE ring_buffer_pos_t atomic_read_rx_head() {
|
||
|
#if RX_BUFFER_SIZE > 256
|
||
|
// Keep reading until 2 consecutive reads return the same value,
|
||
|
// meaning there was no update in-between caused by an interrupt.
|
||
|
// This works because serial RX interrupts happen at a slower rate
|
||
|
// than successive reads of a variable, so 2 consecutive reads with
|
||
|
// the same value means no interrupt updated it.
|
||
|
ring_buffer_pos_t vold, vnew = rx_buffer.head;
|
||
|
sw_barrier();
|
||
|
do {
|
||
|
vold = vnew;
|
||
|
vnew = rx_buffer.head;
|
||
|
sw_barrier();
|
||
|
} while (vold != vnew);
|
||
|
return vnew;
|
||
|
#else
|
||
|
// With an 8bit index, reads are always atomic. No need for special handling
|
||
|
return rx_buffer.head;
|
||
|
#endif
|
||
|
}
|
||
|
|
||
|
#if RX_BUFFER_SIZE > 256
|
||
|
static volatile bool rx_tail_value_not_stable = false;
|
||
|
static volatile uint16_t rx_tail_value_backup = 0;
|
||
|
#endif
|
||
|
|
||
|
// Set RX tail index, taking into account the RX ISR could interrupt
|
||
|
// the write to this variable in the middle - So a backup strategy
|
||
|
// is used to ensure reads of the correct values.
|
||
|
// -Must NOT be called from the RX ISR -
|
||
|
FORCE_INLINE void atomic_set_rx_tail(ring_buffer_pos_t value) {
|
||
|
#if RX_BUFFER_SIZE > 256
|
||
|
// Store the new value in the backup
|
||
|
rx_tail_value_backup = value;
|
||
|
sw_barrier();
|
||
|
// Flag we are about to change the true value
|
||
|
rx_tail_value_not_stable = true;
|
||
|
sw_barrier();
|
||
|
// Store the new value
|
||
|
rx_buffer.tail = value;
|
||
|
sw_barrier();
|
||
|
// Signal the new value is completely stored into the value
|
||
|
rx_tail_value_not_stable = false;
|
||
|
sw_barrier();
|
||
|
#else
|
||
|
rx_buffer.tail = value;
|
||
|
#endif
|
||
|
}
|
||
|
|
||
|
// Get the RX tail index, taking into account the read could be
|
||
|
// interrupting in the middle of the update of that index value
|
||
|
// -Called from the RX ISR -
|
||
|
FORCE_INLINE ring_buffer_pos_t atomic_read_rx_tail() {
|
||
|
#if RX_BUFFER_SIZE > 256
|
||
|
// If the true index is being modified, return the backup value
|
||
|
if (rx_tail_value_not_stable) return rx_tail_value_backup;
|
||
|
#endif
|
||
|
// The true index is stable, return it
|
||
|
return rx_buffer.tail;
|
||
|
}
|
||
|
|
||
|
// (called with RX interrupts disabled)
|
||
|
FORCE_INLINE void store_rxd_char() {
|
||
|
// Get the tail - Nothing can alter its value while this ISR is executing, but there's
|
||
|
// a chance that this ISR interrupted the main process while it was updating the index.
|
||
|
// The backup mechanism ensures the correct value is always returned.
|
||
|
const ring_buffer_pos_t t = atomic_read_rx_tail();
|
||
|
|
||
|
// Get the head pointer - This ISR is the only one that modifies its value, so it's safe to read here
|
||
|
ring_buffer_pos_t h = rx_buffer.head;
|
||
|
|
||
|
// Get the next element
|
||
|
ring_buffer_pos_t i = (ring_buffer_pos_t)(h + 1) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// This must read the M_UCSRxA register before reading the received byte to detect error causes
|
||
|
#if ENABLED(SERIAL_STATS_DROPPED_RX)
|
||
|
if (TEST(M_UCSRxA, M_DORx) && !++rx_dropped_bytes) --rx_dropped_bytes;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_RX_BUFFER_OVERRUNS)
|
||
|
if (TEST(M_UCSRxA, M_DORx) && !++rx_buffer_overruns) --rx_buffer_overruns;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_RX_FRAMING_ERRORS)
|
||
|
if (TEST(M_UCSRxA, M_FEx) && !++rx_framing_errors) --rx_framing_errors;
|
||
|
#endif
|
||
|
|
||
|
// Read the character from the USART
|
||
|
uint8_t c = M_UDRx;
|
||
|
|
||
|
#if ENABLED(EMERGENCY_PARSER)
|
||
|
emergency_parser.update(c);
|
||
|
#endif
|
||
|
|
||
|
// If the character is to be stored at the index just before the tail
|
||
|
// (such that the head would advance to the current tail), the RX FIFO is
|
||
|
// full, so don't write the character or advance the head.
|
||
|
if (i != t) {
|
||
|
rx_buffer.buffer[h] = c;
|
||
|
h = i;
|
||
|
}
|
||
|
#if ENABLED(SERIAL_STATS_DROPPED_RX)
|
||
|
else if (!++rx_dropped_bytes) --rx_dropped_bytes;
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_STATS_MAX_RX_QUEUED)
|
||
|
// Calculate count of bytes stored into the RX buffer
|
||
|
const ring_buffer_pos_t rx_count = (ring_buffer_pos_t)(h - t) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// Keep track of the maximum count of enqueued bytes
|
||
|
NOLESS(rx_max_enqueued, rx_count);
|
||
|
#endif
|
||
|
|
||
|
#if ENABLED(SERIAL_XON_XOFF)
|
||
|
// If the last char that was sent was an XON
|
||
|
if ((xon_xoff_state & XON_XOFF_CHAR_MASK) == XON_CHAR) {
|
||
|
|
||
|
// Bytes stored into the RX buffer
|
||
|
const ring_buffer_pos_t rx_count = (ring_buffer_pos_t)(h - t) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// If over 12.5% of RX buffer capacity, send XOFF before running out of
|
||
|
// RX buffer space .. 325 bytes @ 250kbits/s needed to let the host react
|
||
|
// and stop sending bytes. This translates to 13mS propagation time.
|
||
|
if (rx_count >= (RX_BUFFER_SIZE) / 8) {
|
||
|
|
||
|
// At this point, definitely no TX interrupt was executing, since the TX ISR can't be preempted.
|
||
|
// Don't enable the TX interrupt here as a means to trigger the XOFF char, because if it happens
|
||
|
// to be in the middle of trying to disable the RX interrupt in the main program, eventually the
|
||
|
// enabling of the TX interrupt could be undone. The ONLY reliable thing this can do to ensure
|
||
|
// the sending of the XOFF char is to send it HERE AND NOW.
|
||
|
|
||
|
// About to send the XOFF char
|
||
|
xon_xoff_state = XOFF_CHAR | XON_XOFF_CHAR_SENT;
|
||
|
|
||
|
// Wait until the TX register becomes empty and send it - Here there could be a problem
|
||
|
// - While waiting for the TX register to empty, the RX register could receive a new
|
||
|
// character. This must also handle that situation!
|
||
|
while (!TEST(M_UCSRxA, M_UDREx)) {
|
||
|
|
||
|
if (TEST(M_UCSRxA,M_RXCx)) {
|
||
|
// A char arrived while waiting for the TX buffer to be empty - Receive and process it!
|
||
|
|
||
|
i = (ring_buffer_pos_t)(h + 1) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// Read the character from the USART
|
||
|
c = M_UDRx;
|
||
|
|
||
|
#if ENABLED(EMERGENCY_PARSER)
|
||
|
emergency_parser.update(c);
|
||
|
#endif
|
||
|
|
||
|
// If the character is to be stored at the index just before the tail
|
||
|
// (such that the head would advance to the current tail), the FIFO is
|
||
|
// full, so don't write the character or advance the head.
|
||
|
if (i != t) {
|
||
|
rx_buffer.buffer[h] = c;
|
||
|
h = i;
|
||
|
}
|
||
|
#if ENABLED(SERIAL_STATS_DROPPED_RX)
|
||
|
else if (!++rx_dropped_bytes) --rx_dropped_bytes;
|
||
|
#endif
|
||
|
}
|
||
|
sw_barrier();
|
||
|
}
|
||
|
|
||
|
M_UDRx = XOFF_CHAR;
|
||
|
|
||
|
// Clear the TXC bit -- "can be cleared by writing a one to its bit
|
||
|
// location". This makes sure flush() won't return until the bytes
|
||
|
// actually got written
|
||
|
SBI(M_UCSRxA, M_TXCx);
|
||
|
|
||
|
// At this point there could be a race condition between the write() function
|
||
|
// and this sending of the XOFF char. This interrupt could happen between the
|
||
|
// wait to be empty TX buffer loop and the actual write of the character. Since
|
||
|
// the TX buffer is full because it's sending the XOFF char, the only way to be
|
||
|
// sure the write() function will succeed is to wait for the XOFF char to be
|
||
|
// completely sent. Since an extra character could be received during the wait
|
||
|
// it must also be handled!
|
||
|
while (!TEST(M_UCSRxA, M_UDREx)) {
|
||
|
|
||
|
if (TEST(M_UCSRxA,M_RXCx)) {
|
||
|
// A char arrived while waiting for the TX buffer to be empty - Receive and process it!
|
||
|
|
||
|
i = (ring_buffer_pos_t)(h + 1) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// Read the character from the USART
|
||
|
c = M_UDRx;
|
||
|
|
||
|
#if ENABLED(EMERGENCY_PARSER)
|
||
|
emergency_parser.update(c);
|
||
|
#endif
|
||
|
|
||
|
// If the character is to be stored at the index just before the tail
|
||
|
// (such that the head would advance to the current tail), the FIFO is
|
||
|
// full, so don't write the character or advance the head.
|
||
|
if (i != t) {
|
||
|
rx_buffer.buffer[h] = c;
|
||
|
h = i;
|
||
|
}
|
||
|
#if ENABLED(SERIAL_STATS_DROPPED_RX)
|
||
|
else if (!++rx_dropped_bytes) --rx_dropped_bytes;
|
||
|
#endif
|
||
|
}
|
||
|
sw_barrier();
|
||
|
}
|
||
|
|
||
|
// At this point everything is ready. The write() function won't
|
||
|
// have any issues writing to the UART TX register if it needs to!
|
||
|
}
|
||
|
}
|
||
|
#endif // SERIAL_XON_XOFF
|
||
|
|
||
|
// Store the new head value - The main loop will retry until the value is stable
|
||
|
rx_buffer.head = h;
|
||
|
}
|
||
|
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
|
||
|
// (called with TX irqs disabled)
|
||
|
FORCE_INLINE void _tx_udr_empty_irq(void) {
|
||
|
|
||
|
// Read positions
|
||
|
uint8_t t = tx_buffer.tail;
|
||
|
const uint8_t h = tx_buffer.head;
|
||
|
|
||
|
#if ENABLED(SERIAL_XON_XOFF)
|
||
|
// If an XON char is pending to be sent, do it now
|
||
|
if (xon_xoff_state == XON_CHAR) {
|
||
|
|
||
|
// Send the character
|
||
|
M_UDRx = XON_CHAR;
|
||
|
|
||
|
// clear the TXC bit -- "can be cleared by writing a one to its bit
|
||
|
// location". This makes sure flush() won't return until the bytes
|
||
|
// actually got written
|
||
|
SBI(M_UCSRxA, M_TXCx);
|
||
|
|
||
|
// Remember we sent it.
|
||
|
xon_xoff_state = XON_CHAR | XON_XOFF_CHAR_SENT;
|
||
|
|
||
|
// If nothing else to transmit, just disable TX interrupts.
|
||
|
if (h == t) CBI(M_UCSRxB, M_UDRIEx); // (Non-atomic, could be reenabled by the main program, but eventually this will succeed)
|
||
|
|
||
|
return;
|
||
|
}
|
||
|
#endif
|
||
|
|
||
|
// If nothing to transmit, just disable TX interrupts. This could
|
||
|
// happen as the result of the non atomicity of the disabling of RX
|
||
|
// interrupts that could end reenabling TX interrupts as a side effect.
|
||
|
if (h == t) {
|
||
|
CBI(M_UCSRxB, M_UDRIEx); // (Non-atomic, could be reenabled by the main program, but eventually this will succeed)
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
// There is something to TX, Send the next byte
|
||
|
const uint8_t c = tx_buffer.buffer[t];
|
||
|
t = (t + 1) & (TX_BUFFER_SIZE - 1);
|
||
|
M_UDRx = c;
|
||
|
tx_buffer.tail = t;
|
||
|
|
||
|
// Clear the TXC bit (by writing a one to its bit location).
|
||
|
// Ensures flush() won't return until the bytes are actually written/
|
||
|
SBI(M_UCSRxA, M_TXCx);
|
||
|
|
||
|
// Disable interrupts if there is nothing to transmit following this byte
|
||
|
if (h == t) CBI(M_UCSRxB, M_UDRIEx); // (Non-atomic, could be reenabled by the main program, but eventually this will succeed)
|
||
|
}
|
||
|
|
||
|
#ifdef M_USARTx_UDRE_vect
|
||
|
ISR(M_USARTx_UDRE_vect) { _tx_udr_empty_irq(); }
|
||
|
#endif
|
||
|
|
||
|
#endif // TX_BUFFER_SIZE
|
||
|
|
||
|
#ifdef M_USARTx_RX_vect
|
||
|
ISR(M_USARTx_RX_vect) { store_rxd_char(); }
|
||
|
#endif
|
||
|
|
||
|
// Public Methods
|
||
|
|
||
|
void MarlinSerial::begin(const long baud) {
|
||
|
uint16_t baud_setting;
|
||
|
bool useU2X = true;
|
||
|
|
||
|
#if F_CPU == 16000000UL && SERIAL_PORT == 0
|
||
|
// Hard-coded exception for compatibility with the bootloader shipped
|
||
|
// with the Duemilanove and previous boards, and the firmware on the
|
||
|
// 8U2 on the Uno and Mega 2560.
|
||
|
if (baud == 57600) useU2X = false;
|
||
|
#endif
|
||
|
|
||
|
if (useU2X) {
|
||
|
M_UCSRxA = _BV(M_U2Xx);
|
||
|
baud_setting = (F_CPU / 4 / baud - 1) / 2;
|
||
|
}
|
||
|
else {
|
||
|
M_UCSRxA = 0;
|
||
|
baud_setting = (F_CPU / 8 / baud - 1) / 2;
|
||
|
}
|
||
|
|
||
|
// assign the baud_setting, a.k.a. ubbr (USART Baud Rate Register)
|
||
|
M_UBRRxH = baud_setting >> 8;
|
||
|
M_UBRRxL = baud_setting;
|
||
|
|
||
|
SBI(M_UCSRxB, M_RXENx);
|
||
|
SBI(M_UCSRxB, M_TXENx);
|
||
|
SBI(M_UCSRxB, M_RXCIEx);
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
CBI(M_UCSRxB, M_UDRIEx);
|
||
|
#endif
|
||
|
_written = false;
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::end() {
|
||
|
CBI(M_UCSRxB, M_RXENx);
|
||
|
CBI(M_UCSRxB, M_TXENx);
|
||
|
CBI(M_UCSRxB, M_RXCIEx);
|
||
|
CBI(M_UCSRxB, M_UDRIEx);
|
||
|
}
|
||
|
|
||
|
int MarlinSerial::peek(void) {
|
||
|
const ring_buffer_pos_t h = atomic_read_rx_head(), t = rx_buffer.tail;
|
||
|
return h == t ? -1 : rx_buffer.buffer[t];
|
||
|
}
|
||
|
|
||
|
int MarlinSerial::read(void) {
|
||
|
const ring_buffer_pos_t h = atomic_read_rx_head();
|
||
|
|
||
|
// Read the tail. Main thread owns it, so it is safe to directly read it
|
||
|
ring_buffer_pos_t t = rx_buffer.tail;
|
||
|
|
||
|
// If nothing to read, return now
|
||
|
if (h == t) return -1;
|
||
|
|
||
|
// Get the next char
|
||
|
const int v = rx_buffer.buffer[t];
|
||
|
t = (ring_buffer_pos_t)(t + 1) & (RX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// Advance tail - Making sure the RX ISR will always get an stable value, even
|
||
|
// if it interrupts the writing of the value of that variable in the middle.
|
||
|
atomic_set_rx_tail(t);
|
||
|
|
||
|
#if ENABLED(SERIAL_XON_XOFF)
|
||
|
// If the XOFF char was sent, or about to be sent...
|
||
|
if ((xon_xoff_state & XON_XOFF_CHAR_MASK) == XOFF_CHAR) {
|
||
|
// Get count of bytes in the RX buffer
|
||
|
const ring_buffer_pos_t rx_count = (ring_buffer_pos_t)(h - t) & (ring_buffer_pos_t)(RX_BUFFER_SIZE - 1);
|
||
|
if (rx_count < (RX_BUFFER_SIZE) / 10) {
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
// Signal we want an XON character to be sent.
|
||
|
xon_xoff_state = XON_CHAR;
|
||
|
// Enable TX ISR. Non atomic, but it will eventually enable them
|
||
|
SBI(M_UCSRxB, M_UDRIEx);
|
||
|
#else
|
||
|
// If not using TX interrupts, we must send the XON char now
|
||
|
xon_xoff_state = XON_CHAR | XON_XOFF_CHAR_SENT;
|
||
|
while (!TEST(M_UCSRxA, M_UDREx)) sw_barrier();
|
||
|
M_UDRx = XON_CHAR;
|
||
|
#endif
|
||
|
}
|
||
|
}
|
||
|
#endif
|
||
|
|
||
|
return v;
|
||
|
}
|
||
|
|
||
|
ring_buffer_pos_t MarlinSerial::available(void) {
|
||
|
const ring_buffer_pos_t h = atomic_read_rx_head(), t = rx_buffer.tail;
|
||
|
return (ring_buffer_pos_t)(RX_BUFFER_SIZE + h - t) & (RX_BUFFER_SIZE - 1);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::flush(void) {
|
||
|
|
||
|
// Set the tail to the head:
|
||
|
// - Read the RX head index in a safe way. (See atomic_read_rx_head.)
|
||
|
// - Set the tail, making sure the RX ISR will always get a stable value, even
|
||
|
// if it interrupts the writing of the value of that variable in the middle.
|
||
|
atomic_set_rx_tail(atomic_read_rx_head());
|
||
|
|
||
|
#if ENABLED(SERIAL_XON_XOFF)
|
||
|
// If the XOFF char was sent, or about to be sent...
|
||
|
if ((xon_xoff_state & XON_XOFF_CHAR_MASK) == XOFF_CHAR) {
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
// Signal we want an XON character to be sent.
|
||
|
xon_xoff_state = XON_CHAR;
|
||
|
// Enable TX ISR. Non atomic, but it will eventually enable it.
|
||
|
SBI(M_UCSRxB, M_UDRIEx);
|
||
|
#else
|
||
|
// If not using TX interrupts, we must send the XON char now
|
||
|
xon_xoff_state = XON_CHAR | XON_XOFF_CHAR_SENT;
|
||
|
while (!TEST(M_UCSRxA, M_UDREx)) sw_barrier();
|
||
|
M_UDRx = XON_CHAR;
|
||
|
#endif
|
||
|
}
|
||
|
#endif
|
||
|
}
|
||
|
|
||
|
#if TX_BUFFER_SIZE > 0
|
||
|
void MarlinSerial::write(const uint8_t c) {
|
||
|
_written = true;
|
||
|
|
||
|
// If the TX interrupts are disabled and the data register
|
||
|
// is empty, just write the byte to the data register and
|
||
|
// be done. This shortcut helps significantly improve the
|
||
|
// effective datarate at high (>500kbit/s) bitrates, where
|
||
|
// interrupt overhead becomes a slowdown.
|
||
|
// Yes, there is a race condition between the sending of the
|
||
|
// XOFF char at the RX ISR, but it is properly handled there
|
||
|
if (!TEST(M_UCSRxB, M_UDRIEx) && TEST(M_UCSRxA, M_UDREx)) {
|
||
|
M_UDRx = c;
|
||
|
|
||
|
// clear the TXC bit -- "can be cleared by writing a one to its bit
|
||
|
// location". This makes sure flush() won't return until the bytes
|
||
|
// actually got written
|
||
|
SBI(M_UCSRxA, M_TXCx);
|
||
|
return;
|
||
|
}
|
||
|
|
||
|
const uint8_t i = (tx_buffer.head + 1) & (TX_BUFFER_SIZE - 1);
|
||
|
|
||
|
// If global interrupts are disabled (as the result of being called from an ISR)...
|
||
|
if (!ISRS_ENABLED()) {
|
||
|
|
||
|
// Make room by polling if it is possible to transmit, and do so!
|
||
|
while (i == tx_buffer.tail) {
|
||
|
|
||
|
// If we can transmit another byte, do it.
|
||
|
if (TEST(M_UCSRxA, M_UDREx)) _tx_udr_empty_irq();
|
||
|
|
||
|
// Make sure compiler rereads tx_buffer.tail
|
||
|
sw_barrier();
|
||
|
}
|
||
|
}
|
||
|
else {
|
||
|
// Interrupts are enabled, just wait until there is space
|
||
|
while (i == tx_buffer.tail) { sw_barrier(); }
|
||
|
}
|
||
|
|
||
|
// Store new char. head is always safe to move
|
||
|
tx_buffer.buffer[tx_buffer.head] = c;
|
||
|
tx_buffer.head = i;
|
||
|
|
||
|
// Enable TX ISR - Non atomic, but it will eventually enable TX ISR
|
||
|
SBI(M_UCSRxB, M_UDRIEx);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::flushTX(void) {
|
||
|
// No bytes written, no need to flush. This special case is needed since there's
|
||
|
// no way to force the TXC (transmit complete) bit to 1 during initialization.
|
||
|
if (!_written) return;
|
||
|
|
||
|
// If global interrupts are disabled (as the result of being called from an ISR)...
|
||
|
if (!ISRS_ENABLED()) {
|
||
|
|
||
|
// Wait until everything was transmitted - We must do polling, as interrupts are disabled
|
||
|
while (tx_buffer.head != tx_buffer.tail || !TEST(M_UCSRxA, M_TXCx)) {
|
||
|
|
||
|
// If there is more space, send an extra character
|
||
|
if (TEST(M_UCSRxA, M_UDREx))
|
||
|
_tx_udr_empty_irq();
|
||
|
|
||
|
sw_barrier();
|
||
|
}
|
||
|
|
||
|
}
|
||
|
else {
|
||
|
// Wait until everything was transmitted
|
||
|
while (tx_buffer.head != tx_buffer.tail || !TEST(M_UCSRxA, M_TXCx)) sw_barrier();
|
||
|
}
|
||
|
|
||
|
// At this point nothing is queued anymore (DRIE is disabled) and
|
||
|
// the hardware finished transmission (TXC is set).
|
||
|
}
|
||
|
|
||
|
#else // TX_BUFFER_SIZE == 0
|
||
|
|
||
|
void MarlinSerial::write(const uint8_t c) {
|
||
|
_written = true;
|
||
|
while (!TEST(M_UCSRxA, M_UDREx)) sw_barrier();
|
||
|
M_UDRx = c;
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::flushTX(void) {
|
||
|
// No bytes written, no need to flush. This special case is needed since there's
|
||
|
// no way to force the TXC (transmit complete) bit to 1 during initialization.
|
||
|
if (!_written) return;
|
||
|
|
||
|
// Wait until everything was transmitted
|
||
|
while (!TEST(M_UCSRxA, M_TXCx)) sw_barrier();
|
||
|
|
||
|
// At this point nothing is queued anymore (DRIE is disabled) and
|
||
|
// the hardware finished transmission (TXC is set).
|
||
|
}
|
||
|
#endif // TX_BUFFER_SIZE == 0
|
||
|
|
||
|
/**
|
||
|
* Imports from print.h
|
||
|
*/
|
||
|
|
||
|
void MarlinSerial::print(char c, int base) {
|
||
|
print((long)c, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(unsigned char b, int base) {
|
||
|
print((unsigned long)b, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(int n, int base) {
|
||
|
print((long)n, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(unsigned int n, int base) {
|
||
|
print((unsigned long)n, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(long n, int base) {
|
||
|
if (base == 0) write(n);
|
||
|
else if (base == 10) {
|
||
|
if (n < 0) { print('-'); n = -n; }
|
||
|
printNumber(n, 10);
|
||
|
}
|
||
|
else
|
||
|
printNumber(n, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(unsigned long n, int base) {
|
||
|
if (base == 0) write(n);
|
||
|
else printNumber(n, base);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::print(double n, int digits) {
|
||
|
printFloat(n, digits);
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(void) {
|
||
|
print('\r');
|
||
|
print('\n');
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(const String& s) {
|
||
|
print(s);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(const char c[]) {
|
||
|
print(c);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(char c, int base) {
|
||
|
print(c, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(unsigned char b, int base) {
|
||
|
print(b, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(int n, int base) {
|
||
|
print(n, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(unsigned int n, int base) {
|
||
|
print(n, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(long n, int base) {
|
||
|
print(n, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(unsigned long n, int base) {
|
||
|
print(n, base);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::println(double n, int digits) {
|
||
|
print(n, digits);
|
||
|
println();
|
||
|
}
|
||
|
|
||
|
// Private Methods
|
||
|
|
||
|
void MarlinSerial::printNumber(unsigned long n, uint8_t base) {
|
||
|
if (n) {
|
||
|
unsigned char buf[8 * sizeof(long)]; // Enough space for base 2
|
||
|
int8_t i = 0;
|
||
|
while (n) {
|
||
|
buf[i++] = n % base;
|
||
|
n /= base;
|
||
|
}
|
||
|
while (i--)
|
||
|
print((char)(buf[i] + (buf[i] < 10 ? '0' : 'A' - 10)));
|
||
|
}
|
||
|
else
|
||
|
print('0');
|
||
|
}
|
||
|
|
||
|
void MarlinSerial::printFloat(double number, uint8_t digits) {
|
||
|
// Handle negative numbers
|
||
|
if (number < 0.0) {
|
||
|
print('-');
|
||
|
number = -number;
|
||
|
}
|
||
|
|
||
|
// Round correctly so that print(1.999, 2) prints as "2.00"
|
||
|
double rounding = 0.5;
|
||
|
for (uint8_t i = 0; i < digits; ++i)
|
||
|
rounding *= 0.1;
|
||
|
|
||
|
number += rounding;
|
||
|
|
||
|
// Extract the integer part of the number and print it
|
||
|
unsigned long int_part = (unsigned long)number;
|
||
|
double remainder = number - (double)int_part;
|
||
|
print(int_part);
|
||
|
|
||
|
// Print the decimal point, but only if there are digits beyond
|
||
|
if (digits) {
|
||
|
print('.');
|
||
|
// Extract digits from the remainder one at a time
|
||
|
while (digits--) {
|
||
|
remainder *= 10.0;
|
||
|
int toPrint = int(remainder);
|
||
|
print(toPrint);
|
||
|
remainder -= toPrint;
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
// Preinstantiate
|
||
|
MarlinSerial customizedSerial;
|
||
|
|
||
|
#endif // USE_MARLINSERIAL && (UBRRH || UBRR0H || UBRR1H || UBRR2H || UBRR3H)
|
||
|
|
||
|
// For AT90USB targets use the UART for BT interfacing
|
||
|
#if !USE_MARLINSERIAL && ENABLED(BLUETOOTH)
|
||
|
HardwareSerial bluetoothSerial;
|
||
|
#endif
|