aboutsummaryrefslogtreecommitdiff
path: root/Marlin/src/HAL/DUE/HAL_SPI.cpp
diff options
context:
space:
mode:
authorGeorgiy Bondarenko <69736697+nehilo@users.noreply.github.com>2021-03-04 20:54:23 +0300
committerGeorgiy Bondarenko <69736697+nehilo@users.noreply.github.com>2021-03-04 20:54:23 +0300
commite8701195e66f2d27ffe17fb514eae8173795aaf7 (patch)
tree9f519c4abf6556b9ae7190a6210d87ead1dfadde /Marlin/src/HAL/DUE/HAL_SPI.cpp
downloadkp3s-lgvl-e8701195e66f2d27ffe17fb514eae8173795aaf7.tar.xz
kp3s-lgvl-e8701195e66f2d27ffe17fb514eae8173795aaf7.zip
Initial commit
Diffstat (limited to 'Marlin/src/HAL/DUE/HAL_SPI.cpp')
-rw-r--r--Marlin/src/HAL/DUE/HAL_SPI.cpp825
1 files changed, 825 insertions, 0 deletions
diff --git a/Marlin/src/HAL/DUE/HAL_SPI.cpp b/Marlin/src/HAL/DUE/HAL_SPI.cpp
new file mode 100644
index 0000000..342c373
--- /dev/null
+++ b/Marlin/src/HAL/DUE/HAL_SPI.cpp
@@ -0,0 +1,825 @@
+/**
+ * Marlin 3D Printer Firmware
+ * Copyright (c) 2020 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 <https://www.gnu.org/licenses/>.
+ *
+ */
+
+/**
+ * Software SPI functions originally from Arduino Sd2Card Library
+ * Copyright (c) 2009 by William Greiman
+ *
+ * Completely rewritten and tuned by Eduardo José Tagle in 2017/2018
+ * in ARM thumb2 inline assembler and tuned for maximum speed and performance
+ * allowing SPI clocks of up to 12 Mhz to increase SD card read/write performance
+ */
+
+/**
+ * HAL for Arduino Due and compatible (SAM3X8E)
+ *
+ * For ARDUINO_ARCH_SAM
+ */
+
+#ifdef ARDUINO_ARCH_SAM
+
+#include "../../inc/MarlinConfig.h"
+#include "../shared/Delay.h"
+
+// ------------------------
+// Public functions
+// ------------------------
+
+#if EITHER(DUE_SOFTWARE_SPI, FORCE_SOFT_SPI)
+
+ // ------------------------
+ // Software SPI
+ // ------------------------
+
+ // Make sure GCC optimizes this file.
+ // Note that this line triggers a bug in GCC which is fixed by casting.
+ // See the note below.
+ #pragma GCC optimize (3)
+
+ typedef uint8_t (*pfnSpiTransfer)(uint8_t b);
+ typedef void (*pfnSpiRxBlock)(uint8_t* buf, uint32_t nbyte);
+ typedef void (*pfnSpiTxBlock)(const uint8_t* buf, uint32_t nbyte);
+
+ /* ---------------- Macros to be able to access definitions from asm */
+ #define _PORT(IO) DIO ## IO ## _WPORT
+ #define _PIN_MASK(IO) MASK(DIO ## IO ## _PIN)
+ #define _PIN_SHIFT(IO) DIO ## IO ## _PIN
+ #define PORT(IO) _PORT(IO)
+ #define PIN_MASK(IO) _PIN_MASK(IO)
+ #define PIN_SHIFT(IO) _PIN_SHIFT(IO)
+
+ // run at ~8 .. ~10Mhz - Tx version (Rx data discarded)
+ static uint8_t spiTransferTx0(uint8_t bout) { // using Mode 0
+ uint32_t MOSI_PORT_PLUS30 = ((uint32_t) PORT(SD_MOSI_PIN)) + 0x30; /* SODR of port */
+ uint32_t MOSI_MASK = PIN_MASK(SD_MOSI_PIN);
+ uint32_t SCK_PORT_PLUS30 = ((uint32_t) PORT(SD_SCK_PIN)) + 0x30; /* SODR of port */
+ uint32_t SCK_MASK = PIN_MASK(SD_SCK_PIN);
+ uint32_t idx = 0;
+
+ /* Negate bout, as the assembler requires a negated value */
+ bout = ~bout;
+
+ /* The software SPI routine */
+ __asm__ __volatile__(
+ A(".syntax unified") // is to prevent CM0,CM1 non-unified syntax
+
+ /* Bit 7 */
+ A("ubfx %[idx],%[txval],#7,#1") /* Place bit 7 in bit 0 of idx*/
+
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#6,#1") /* Place bit 6 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 6 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#5,#1") /* Place bit 5 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 5 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#4,#1") /* Place bit 4 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 4 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#3,#1") /* Place bit 3 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 3 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#2,#1") /* Place bit 2 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 2 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#1,#1") /* Place bit 1 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 1 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[idx],%[txval],#0,#1") /* Place bit 0 in bit 0 of idx*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 0 */
+ A("str %[mosi_mask],[%[mosi_port], %[idx],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("nop") /* Result will be 0 */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ : [idx]"+r"( idx )
+ : [txval]"r"( bout ) ,
+ [mosi_mask]"r"( MOSI_MASK ),
+ [mosi_port]"r"( MOSI_PORT_PLUS30 ),
+ [sck_mask]"r"( SCK_MASK ),
+ [sck_port]"r"( SCK_PORT_PLUS30 )
+ : "cc"
+ );
+
+ return 0;
+ }
+
+ // Calculates the bit band alias address and returns a pointer address to word.
+ // addr: The byte address of bitbanding bit.
+ // bit: The bit position of bitbanding bit.
+ #define BITBAND_ADDRESS(addr, bit) \
+ (((uint32_t)(addr) & 0xF0000000) + 0x02000000 + ((uint32_t)(addr)&0xFFFFF)*32 + (bit)*4)
+
+ // run at ~8 .. ~10Mhz - Rx version (Tx line not altered)
+ static uint8_t spiTransferRx0(uint8_t) { // using Mode 0
+ uint32_t bin = 0;
+ uint32_t work = 0;
+ uint32_t BITBAND_MISO_PORT = BITBAND_ADDRESS( ((uint32_t)PORT(SD_MISO_PIN))+0x3C, PIN_SHIFT(SD_MISO_PIN)); /* PDSR of port in bitband area */
+ uint32_t SCK_PORT_PLUS30 = ((uint32_t) PORT(SD_SCK_PIN)) + 0x30; /* SODR of port */
+ uint32_t SCK_MASK = PIN_MASK(SD_SCK_PIN);
+
+ /* The software SPI routine */
+ __asm__ __volatile__(
+ A(".syntax unified") // is to prevent CM0,CM1 non-unified syntax
+
+ /* bit 7 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#7,#1") /* Store read bit as the bit 7 */
+
+ /* bit 6 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#6,#1") /* Store read bit as the bit 6 */
+
+ /* bit 5 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#5,#1") /* Store read bit as the bit 5 */
+
+ /* bit 4 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#4,#1") /* Store read bit as the bit 4 */
+
+ /* bit 3 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#3,#1") /* Store read bit as the bit 3 */
+
+ /* bit 2 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#2,#1") /* Store read bit as the bit 2 */
+
+ /* bit 1 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#1,#1") /* Store read bit as the bit 1 */
+
+ /* bit 0 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#0,#1") /* Store read bit as the bit 0 */
+
+ : [bin]"+r"(bin),
+ [work]"+r"(work)
+ : [bitband_miso_port]"r"( BITBAND_MISO_PORT ),
+ [sck_mask]"r"( SCK_MASK ),
+ [sck_port]"r"( SCK_PORT_PLUS30 )
+ : "cc"
+ );
+
+ return bin;
+ }
+
+ // run at ~4Mhz
+ static uint8_t spiTransfer1(uint8_t b) { // using Mode 0
+ int bits = 8;
+ do {
+ WRITE(SD_MOSI_PIN, b & 0x80);
+ b <<= 1; // little setup time
+
+ WRITE(SD_SCK_PIN, HIGH);
+ DELAY_NS(125); // 10 cycles @ 84mhz
+
+ b |= (READ(SD_MISO_PIN) != 0);
+
+ WRITE(SD_SCK_PIN, LOW);
+ DELAY_NS(125); // 10 cycles @ 84mhz
+ } while (--bits);
+ return b;
+ }
+
+ // all the others
+ static uint32_t spiDelayCyclesX4 = (F_CPU) / 1000000; // 4µs => 125khz
+
+ static uint8_t spiTransferX(uint8_t b) { // using Mode 0
+ int bits = 8;
+ do {
+ WRITE(SD_MOSI_PIN, b & 0x80);
+ b <<= 1; // little setup time
+
+ WRITE(SD_SCK_PIN, HIGH);
+ __delay_4cycles(spiDelayCyclesX4);
+
+ b |= (READ(SD_MISO_PIN) != 0);
+
+ WRITE(SD_SCK_PIN, LOW);
+ __delay_4cycles(spiDelayCyclesX4);
+ } while (--bits);
+ return b;
+ }
+
+ // Pointers to generic functions for byte transfers
+
+ /**
+ * Note: The cast is unnecessary, but without it, this file triggers a GCC 4.8.3-2014 bug.
+ * Later GCC versions do not have this problem, but at this time (May 2018) Arduino still
+ * uses that buggy and obsolete GCC version!!
+ */
+ static pfnSpiTransfer spiTransferRx = (pfnSpiTransfer)spiTransferX;
+ static pfnSpiTransfer spiTransferTx = (pfnSpiTransfer)spiTransferX;
+
+ // Block transfers run at ~8 .. ~10Mhz - Tx version (Rx data discarded)
+ static void spiTxBlock0(const uint8_t* ptr, uint32_t todo) {
+ uint32_t MOSI_PORT_PLUS30 = ((uint32_t) PORT(SD_MOSI_PIN)) + 0x30; /* SODR of port */
+ uint32_t MOSI_MASK = PIN_MASK(SD_MOSI_PIN);
+ uint32_t SCK_PORT_PLUS30 = ((uint32_t) PORT(SD_SCK_PIN)) + 0x30; /* SODR of port */
+ uint32_t SCK_MASK = PIN_MASK(SD_SCK_PIN);
+ uint32_t work = 0;
+ uint32_t txval = 0;
+
+ /* The software SPI routine */
+ __asm__ __volatile__(
+ A(".syntax unified") // is to prevent CM0,CM1 non-unified syntax
+
+ L("loop%=")
+ A("ldrb.w %[txval], [%[ptr]], #1") /* Load value to send, increment buffer */
+ A("mvn %[txval],%[txval]") /* Negate value */
+
+ /* Bit 7 */
+ A("ubfx %[work],%[txval],#7,#1") /* Place bit 7 in bit 0 of work*/
+
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#6,#1") /* Place bit 6 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 6 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#5,#1") /* Place bit 5 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 5 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#4,#1") /* Place bit 4 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 4 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#3,#1") /* Place bit 3 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 3 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#2,#1") /* Place bit 2 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 2 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#1,#1") /* Place bit 1 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 1 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ubfx %[work],%[txval],#0,#1") /* Place bit 0 in bit 0 of work*/
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+
+ /* Bit 0 */
+ A("str %[mosi_mask],[%[mosi_port], %[work],LSL #2]") /* Access the proper SODR or CODR registers based on that bit */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("subs %[todo],#1") /* Decrement count of pending words to send, update status */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bne.n loop%=") /* Repeat until done */
+
+ : [ptr]"+r" ( ptr ) ,
+ [todo]"+r" ( todo ) ,
+ [work]"+r"( work ) ,
+ [txval]"+r"( txval )
+ : [mosi_mask]"r"( MOSI_MASK ),
+ [mosi_port]"r"( MOSI_PORT_PLUS30 ),
+ [sck_mask]"r"( SCK_MASK ),
+ [sck_port]"r"( SCK_PORT_PLUS30 )
+ : "cc"
+ );
+ }
+
+ static void spiRxBlock0(uint8_t* ptr, uint32_t todo) {
+ uint32_t bin = 0;
+ uint32_t work = 0;
+ uint32_t BITBAND_MISO_PORT = BITBAND_ADDRESS( ((uint32_t)PORT(SD_MISO_PIN))+0x3C, PIN_SHIFT(SD_MISO_PIN)); /* PDSR of port in bitband area */
+ uint32_t SCK_PORT_PLUS30 = ((uint32_t) PORT(SD_SCK_PIN)) + 0x30; /* SODR of port */
+ uint32_t SCK_MASK = PIN_MASK(SD_SCK_PIN);
+
+ /* The software SPI routine */
+ __asm__ __volatile__(
+ A(".syntax unified") // is to prevent CM0,CM1 non-unified syntax
+
+ L("loop%=")
+
+ /* bit 7 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#7,#1") /* Store read bit as the bit 7 */
+
+ /* bit 6 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#6,#1") /* Store read bit as the bit 6 */
+
+ /* bit 5 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#5,#1") /* Store read bit as the bit 5 */
+
+ /* bit 4 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#4,#1") /* Store read bit as the bit 4 */
+
+ /* bit 3 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#3,#1") /* Store read bit as the bit 3 */
+
+ /* bit 2 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#2,#1") /* Store read bit as the bit 2 */
+
+ /* bit 1 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#1,#1") /* Store read bit as the bit 1 */
+
+ /* bit 0 */
+ A("str %[sck_mask],[%[sck_port]]") /* SODR */
+ A("ldr %[work],[%[bitband_miso_port]]") /* PDSR on bitband area for required bit: work will be 1 or 0 based on port */
+ A("str %[sck_mask],[%[sck_port],#0x4]") /* CODR */
+ A("bfi %[bin],%[work],#0,#1") /* Store read bit as the bit 0 */
+
+ A("subs %[todo],#1") /* Decrement count of pending words to send, update status */
+ A("strb.w %[bin], [%[ptr]], #1") /* Store read value into buffer, increment buffer pointer */
+ A("bne.n loop%=") /* Repeat until done */
+
+ : [ptr]"+r"(ptr),
+ [todo]"+r"(todo),
+ [bin]"+r"(bin),
+ [work]"+r"(work)
+ : [bitband_miso_port]"r"( BITBAND_MISO_PORT ),
+ [sck_mask]"r"( SCK_MASK ),
+ [sck_port]"r"( SCK_PORT_PLUS30 )
+ : "cc"
+ );
+ }
+
+ static void spiTxBlockX(const uint8_t* buf, uint32_t todo) {
+ do {
+ (void)spiTransferTx(*buf++);
+ } while (--todo);
+ }
+
+ static void spiRxBlockX(uint8_t* buf, uint32_t todo) {
+ do {
+ *buf++ = spiTransferRx(0xFF);
+ } while (--todo);
+ }
+
+ // Pointers to generic functions for block tranfers
+ static pfnSpiTxBlock spiTxBlock = (pfnSpiTxBlock)spiTxBlockX;
+ static pfnSpiRxBlock spiRxBlock = (pfnSpiRxBlock)spiRxBlockX;
+
+ #if MB(ALLIGATOR)
+ #define _SS_WRITE(S) WRITE(SD_SS_PIN, S)
+ #else
+ #define _SS_WRITE(S) NOOP
+ #endif
+
+ void spiBegin() {
+ SET_OUTPUT(SD_SS_PIN);
+ _SS_WRITE(HIGH);
+ SET_OUTPUT(SD_SCK_PIN);
+ SET_INPUT(SD_MISO_PIN);
+ SET_OUTPUT(SD_MOSI_PIN);
+ }
+
+ uint8_t spiRec() {
+ _SS_WRITE(LOW);
+ WRITE(SD_MOSI_PIN, HIGH); // Output 1s 1
+ uint8_t b = spiTransferRx(0xFF);
+ _SS_WRITE(HIGH);
+ return b;
+ }
+
+ void spiRead(uint8_t* buf, uint16_t nbyte) {
+ if (nbyte) {
+ _SS_WRITE(LOW);
+ WRITE(SD_MOSI_PIN, HIGH); // Output 1s 1
+ spiRxBlock(buf, nbyte);
+ _SS_WRITE(HIGH);
+ }
+ }
+
+ void spiSend(uint8_t b) {
+ _SS_WRITE(LOW);
+ (void)spiTransferTx(b);
+ _SS_WRITE(HIGH);
+ }
+
+ void spiSendBlock(uint8_t token, const uint8_t* buf) {
+ _SS_WRITE(LOW);
+ (void)spiTransferTx(token);
+ spiTxBlock(buf, 512);
+ _SS_WRITE(HIGH);
+ }
+
+ /**
+ * spiRate should be
+ * 0 : 8 - 10 MHz
+ * 1 : 4 - 5 MHz
+ * 2 : 2 - 2.5 MHz
+ * 3 : 1 - 1.25 MHz
+ * 4 : 500 - 625 kHz
+ * 5 : 250 - 312 kHz
+ * 6 : 125 - 156 kHz
+ */
+ void spiInit(uint8_t spiRate) {
+ switch (spiRate) {
+ case 0:
+ spiTransferTx = (pfnSpiTransfer)spiTransferTx0;
+ spiTransferRx = (pfnSpiTransfer)spiTransferRx0;
+ spiTxBlock = (pfnSpiTxBlock)spiTxBlock0;
+ spiRxBlock = (pfnSpiRxBlock)spiRxBlock0;
+ break;
+ case 1:
+ spiTransferTx = (pfnSpiTransfer)spiTransfer1;
+ spiTransferRx = (pfnSpiTransfer)spiTransfer1;
+ spiTxBlock = (pfnSpiTxBlock)spiTxBlockX;
+ spiRxBlock = (pfnSpiRxBlock)spiRxBlockX;
+ break;
+ default:
+ spiDelayCyclesX4 = ((F_CPU) / 1000000) >> (6 - spiRate);
+ spiTransferTx = (pfnSpiTransfer)spiTransferX;
+ spiTransferRx = (pfnSpiTransfer)spiTransferX;
+ spiTxBlock = (pfnSpiTxBlock)spiTxBlockX;
+ spiRxBlock = (pfnSpiRxBlock)spiRxBlockX;
+ break;
+ }
+
+ _SS_WRITE(HIGH);
+ WRITE(SD_MOSI_PIN, HIGH);
+ WRITE(SD_SCK_PIN, LOW);
+ }
+
+ /** Begin SPI transaction, set clock, bit order, data mode */
+ void spiBeginTransaction(uint32_t spiClock, uint8_t bitOrder, uint8_t dataMode) {
+ // TODO: to be implemented
+ }
+
+ #pragma GCC reset_options
+
+#else // !SOFTWARE_SPI
+
+ #define WHILE_TX(N) while ((SPI0->SPI_SR & SPI_SR_TDRE) == (N))
+ #define WHILE_RX(N) while ((SPI0->SPI_SR & SPI_SR_RDRF) == (N))
+ #define FLUSH_TX() do{ WHILE_RX(1) SPI0->SPI_RDR; }while(0)
+
+ #if MB(ALLIGATOR)
+
+ // slave selects controlled by SPI controller
+ // doesn't support changing SPI speeds for SD card
+
+ // ------------------------
+ // hardware SPI
+ // ------------------------
+ static bool spiInitialized = false;
+
+ void spiInit(uint8_t spiRate) {
+ if (spiInitialized) return;
+
+ // 8.4 MHz, 4 MHz, 2 MHz, 1 MHz, 0.5 MHz, 0.329 MHz, 0.329 MHz
+ constexpr int spiDivider[] = { 10, 21, 42, 84, 168, 255, 255 };
+ if (spiRate > 6) spiRate = 1;
+
+ // Set SPI mode 1, clock, select not active after transfer, with delay between transfers
+ SPI_ConfigureNPCS(SPI0, SPI_CHAN_DAC,
+ SPI_CSR_CSAAT | SPI_CSR_SCBR(spiDivider[spiRate]) |
+ SPI_CSR_DLYBCT(1));
+ // Set SPI mode 0, clock, select not active after transfer, with delay between transfers
+ SPI_ConfigureNPCS(SPI0, SPI_CHAN_EEPROM1, SPI_CSR_NCPHA |
+ SPI_CSR_CSAAT | SPI_CSR_SCBR(spiDivider[spiRate]) |
+ SPI_CSR_DLYBCT(1));
+
+ // Set SPI mode 0, clock, select not active after transfer, with delay between transfers
+ SPI_ConfigureNPCS(SPI0, SPI_CHAN, SPI_CSR_NCPHA |
+ SPI_CSR_CSAAT | SPI_CSR_SCBR(spiDivider[spiRate]) |
+ SPI_CSR_DLYBCT(1));
+ SPI_Enable(SPI0);
+ spiInitialized = true;
+ }
+
+ void spiBegin() {
+ if (spiInitialized) return;
+
+ // Configure SPI pins
+ PIO_Configure(
+ g_APinDescription[SD_SCK_PIN].pPort,
+ g_APinDescription[SD_SCK_PIN].ulPinType,
+ g_APinDescription[SD_SCK_PIN].ulPin,
+ g_APinDescription[SD_SCK_PIN].ulPinConfiguration);
+ PIO_Configure(
+ g_APinDescription[SD_MOSI_PIN].pPort,
+ g_APinDescription[SD_MOSI_PIN].ulPinType,
+ g_APinDescription[SD_MOSI_PIN].ulPin,
+ g_APinDescription[SD_MOSI_PIN].ulPinConfiguration);
+ PIO_Configure(
+ g_APinDescription[SD_MISO_PIN].pPort,
+ g_APinDescription[SD_MISO_PIN].ulPinType,
+ g_APinDescription[SD_MISO_PIN].ulPin,
+ g_APinDescription[SD_MISO_PIN].ulPinConfiguration);
+
+ // set master mode, peripheral select, fault detection
+ SPI_Configure(SPI0, ID_SPI0, SPI_MR_MSTR | SPI_MR_MODFDIS | SPI_MR_PS);
+ SPI_Enable(SPI0);
+
+ SET_OUTPUT(DAC0_SYNC);
+ #if HAS_MULTI_EXTRUDER
+ SET_OUTPUT(DAC1_SYNC);
+ WRITE(DAC1_SYNC, HIGH);
+ #endif
+ SET_OUTPUT(SPI_EEPROM1_CS);
+ SET_OUTPUT(SPI_EEPROM2_CS);
+ SET_OUTPUT(SPI_FLASH_CS);
+ WRITE(DAC0_SYNC, HIGH);
+ WRITE(SPI_EEPROM1_CS, HIGH);
+ WRITE(SPI_EEPROM2_CS, HIGH);
+ WRITE(SPI_FLASH_CS, HIGH);
+ WRITE(SD_SS_PIN, HIGH);
+
+ OUT_WRITE(SDSS, LOW);
+
+ PIO_Configure(
+ g_APinDescription[SPI_PIN].pPort,
+ g_APinDescription[SPI_PIN].ulPinType,
+ g_APinDescription[SPI_PIN].ulPin,
+ g_APinDescription[SPI_PIN].ulPinConfiguration
+ );
+
+ spiInit(1);
+ }
+
+ // Read single byte from SPI
+ uint8_t spiRec() {
+ // write dummy byte with address and end transmission flag
+ SPI0->SPI_TDR = 0x000000FF | SPI_PCS(SPI_CHAN) | SPI_TDR_LASTXFER;
+
+ WHILE_TX(0);
+ WHILE_RX(0);
+
+ //DELAY_US(1U);
+ return SPI0->SPI_RDR;
+ }
+
+ uint8_t spiRec(uint32_t chan) {
+
+ WHILE_TX(0);
+ FLUSH_RX();
+
+ // write dummy byte with address and end transmission flag
+ SPI0->SPI_TDR = 0x000000FF | SPI_PCS(chan) | SPI_TDR_LASTXFER;
+ WHILE_RX(0);
+
+ return SPI0->SPI_RDR;
+ }
+
+ // Read from SPI into buffer
+ void spiRead(uint8_t* buf, uint16_t nbyte) {
+ if (!nbyte) return;
+ --nbyte;
+ for (int i = 0; i < nbyte; i++) {
+ //WHILE_TX(0);
+ SPI0->SPI_TDR = 0x000000FF | SPI_PCS(SPI_CHAN);
+ WHILE_RX(0);
+ buf[i] = SPI0->SPI_RDR;
+ //DELAY_US(1U);
+ }
+ buf[nbyte] = spiRec();
+ }
+
+ // Write single byte to SPI
+ void spiSend(const byte b) {
+ // write byte with address and end transmission flag
+ SPI0->SPI_TDR = (uint32_t)b | SPI_PCS(SPI_CHAN) | SPI_TDR_LASTXFER;
+ WHILE_TX(0);
+ WHILE_RX(0);
+ SPI0->SPI_RDR;
+ //DELAY_US(1U);
+ }
+
+ void spiSend(const uint8_t* buf, size_t nbyte) {
+ if (!nbyte) return;
+ --nbyte;
+ for (size_t i = 0; i < nbyte; i++) {
+ SPI0->SPI_TDR = (uint32_t)buf[i] | SPI_PCS(SPI_CHAN);
+ WHILE_TX(0);
+ WHILE_RX(0);
+ SPI0->SPI_RDR;
+ //DELAY_US(1U);
+ }
+ spiSend(buf[nbyte]);
+ }
+
+ void spiSend(uint32_t chan, byte b) {
+ WHILE_TX(0);
+ // write byte with address and end transmission flag
+ SPI0->SPI_TDR = (uint32_t)b | SPI_PCS(chan) | SPI_TDR_LASTXFER;
+ WHILE_RX(0);
+ FLUSH_RX();
+ }
+
+ void spiSend(uint32_t chan, const uint8_t* buf, size_t nbyte) {
+ if (!nbyte) return;
+ --nbyte;
+ for (size_t i = 0; i < nbyte; i++) {
+ WHILE_TX(0);
+ SPI0->SPI_TDR = (uint32_t)buf[i] | SPI_PCS(chan);
+ WHILE_RX(0);
+ FLUSH_RX();
+ }
+ spiSend(chan, buf[nbyte]);
+ }
+
+ // Write from buffer to SPI
+ void spiSendBlock(uint8_t token, const uint8_t* buf) {
+ SPI0->SPI_TDR = (uint32_t)token | SPI_PCS(SPI_CHAN);
+ WHILE_TX(0);
+ //WHILE_RX(0);
+ //SPI0->SPI_RDR;
+ for (int i = 0; i < 511; i++) {
+ SPI0->SPI_TDR = (uint32_t)buf[i] | SPI_PCS(SPI_CHAN);
+ WHILE_TX(0);
+ WHILE_RX(0);
+ SPI0->SPI_RDR;
+ //DELAY_US(1U);
+ }
+ spiSend(buf[511]);
+ }
+
+ /** Begin SPI transaction, set clock, bit order, data mode */
+ void spiBeginTransaction(uint32_t spiClock, uint8_t bitOrder, uint8_t dataMode) {
+ // TODO: to be implemented
+ }
+
+ #else // U8G compatible hardware SPI
+
+ #define SPI_MODE_0_DUE_HW 2 // DUE CPHA control bit is inverted
+ #define SPI_MODE_1_DUE_HW 3
+ #define SPI_MODE_2_DUE_HW 0
+ #define SPI_MODE_3_DUE_HW 1
+
+ /**
+ * The DUE SPI controller is set up so the upper word of the longword
+ * written to the transmit data register selects which SPI Chip Select
+ * Register is used. This allows different streams to have different SPI
+ * settings.
+ *
+ * In practice it's spooky. Some combinations hang the system, while others
+ * upset the peripheral device.
+ *
+ * SPI mode should be the same for all streams. The FYSETC_MINI_12864 gets
+ * upset if the clock phase changes after chip select goes active.
+ *
+ * SPI_CSR_CSAAT should be set for all streams. If not the WHILE_TX(0)
+ * macro returns immediately which can result in the SPI chip select going
+ * inactive before all the data has been sent.
+ *
+ * The TMC2130 library uses SPI0->SPI_CSR[3].
+ *
+ * The U8G hardware SPI uses SPI0->SPI_CSR[0]. The system hangs and/or the
+ * FYSETC_MINI_12864 gets upset if lower baud rates are used and the SD card
+ * is inserted or removed.
+ *
+ * The SD card uses SPI0->SPI_CSR[3]. Efforts were made to use [1] and [2]
+ * but they all resulted in hangs or garbage on the LCD.
+ *
+ * The SPI controlled chip selects are NOT enabled in the GPIO controller.
+ * The application must control the chip select.
+ *
+ * All of the above can be avoided by defining FORCE_SOFT_SPI to force the
+ * display to use software SPI.
+ */
+
+ void spiInit(uint8_t spiRate=6) { // Default to slowest rate if not specified)
+ // Also sets U8G SPI rate to 4MHz and the SPI mode to 3
+
+ // 8.4 MHz, 4 MHz, 2 MHz, 1 MHz, 0.5 MHz, 0.329 MHz, 0.329 MHz
+ constexpr int spiDivider[] = { 10, 21, 42, 84, 168, 255, 255 };
+ if (spiRate > 6) spiRate = 1;
+
+ // Enable PIOA and SPI0
+ REG_PMC_PCER0 = (1UL << ID_PIOA) | (1UL << ID_SPI0);
+
+ // Disable PIO on A26 and A27
+ REG_PIOA_PDR = 0x0C000000;
+ OUT_WRITE(SDSS, HIGH);
+
+ // Reset SPI0 (from sam lib)
+ SPI0->SPI_CR = SPI_CR_SPIDIS;
+ SPI0->SPI_CR = SPI_CR_SWRST;
+ SPI0->SPI_CR = SPI_CR_SWRST;
+ SPI0->SPI_CR = SPI_CR_SPIEN;
+
+ // TMC2103 compatible setup
+ // Master mode, no fault detection, PCS bits in data written to TDR select CSR register
+ SPI0->SPI_MR = SPI_MR_MSTR | SPI_MR_PS | SPI_MR_MODFDIS;
+ // SPI mode 3, 8 Bit data transfer, baud rate
+ SPI0->SPI_CSR[3] = SPI_CSR_SCBR(spiDivider[spiRate]) | SPI_CSR_CSAAT | SPI_MODE_3_DUE_HW; // use same CSR as TMC2130
+ SPI0->SPI_CSR[0] = SPI_CSR_SCBR(spiDivider[1]) | SPI_CSR_CSAAT | SPI_MODE_3_DUE_HW; // U8G default to 4MHz
+ }
+
+ void spiBegin() { spiInit(); }
+
+ static uint8_t spiTransfer(uint8_t data) {
+ WHILE_TX(0);
+ SPI0->SPI_TDR = (uint32_t)data | 0x00070000UL; // Add TMC2130 PCS bits to every byte (use SPI0->SPI_CSR[3])
+ WHILE_TX(0);
+ WHILE_RX(0);
+ return SPI0->SPI_RDR;
+ }
+
+ uint8_t spiRec() { return (uint8_t)spiTransfer(0xFF); }
+
+ void spiRead(uint8_t* buf, uint16_t nbyte) {
+ for (int i = 0; i < nbyte; i++)
+ buf[i] = spiTransfer(0xFF);
+ }
+
+ void spiSend(uint8_t data) { spiTransfer(data); }
+
+ void spiSend(const uint8_t* buf, size_t nbyte) {
+ for (uint16_t i = 0; i < nbyte; i++)
+ spiTransfer(buf[i]);
+ }
+
+ void spiSendBlock(uint8_t token, const uint8_t* buf) {
+ spiTransfer(token);
+ for (uint16_t i = 0; i < 512; i++)
+ spiTransfer(buf[i]);
+ }
+
+ #endif // !ALLIGATOR
+#endif // !SOFTWARE_SPI
+
+#endif // ARDUINO_ARCH_SAM