~ruther/qmk_firmware

ef8267d6a75c281e6981ea07520d369e2ef092fc — Joel Challis 4 years ago 6cdba3a
Refactor Drop boards to custom matrix lite (#13468)

M keyboards/massdrop/alt/matrix.c => keyboards/massdrop/alt/matrix.c +36 -142
@@ 16,167 16,61 @@ along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#include "alt.h"

#include "d51_util.h"
#include "debug.h"
#include "clks.h"
#include <string.h>

#ifndef MATRIX_IO_DELAY
#    define MATRIX_IO_DELAY 1
#endif

matrix_row_t mlatest[MATRIX_ROWS];
matrix_row_t mlast[MATRIX_ROWS];
matrix_row_t mdebounced[MATRIX_ROWS];

uint8_t row_ports[] = { MATRIX_ROW_PORTS };
uint8_t row_pins[] = { MATRIX_ROW_PINS };
uint8_t col_ports[] = { MATRIX_COL_PORTS };
uint8_t col_pins[] = { MATRIX_COL_PINS };
uint32_t row_masks[2]; //NOTE: If more than PA PB used in the future, adjust code to accomodate

__attribute__((weak)) void matrix_io_delay(void) { wait_us(MATRIX_IO_DELAY); }

__attribute__ ((weak))
void matrix_init_kb(void) {
    matrix_init_user();
}

__attribute__ ((weak))
void matrix_scan_kb(void) {
    matrix_scan_user();
}

__attribute__ ((weak))
void matrix_init_user(void) {
}

__attribute__ ((weak))
void matrix_scan_user(void) {
}

void matrix_init(void)
{
    memset(mlatest, 0, MATRIX_ROWS * sizeof(matrix_row_t));
    memset(mlast, 0, MATRIX_ROWS * sizeof(matrix_row_t));
    memset(mdebounced, 0, MATRIX_ROWS * sizeof(matrix_row_t));
const uint8_t row_ports[] = {MATRIX_ROW_PORTS};
const uint8_t row_pins[]  = {MATRIX_ROW_PINS};
const uint8_t col_ports[] = {MATRIX_COL_PORTS};
const uint8_t col_pins[]  = {MATRIX_COL_PINS};
uint32_t      row_masks[2];  // NOTE: If more than PA PB used in the future, adjust code to accommodate

void matrix_init_custom(void) {
    row_masks[PA] = 0;
    row_masks[PB] = 0;

    uint8_t row;
    for (row = 0; row < MATRIX_ROWS; row++)
    {
        PORT->Group[row_ports[row]].DIRCLR.reg = 1 << row_pins[row]; //Input
        PORT->Group[row_ports[row]].OUTCLR.reg = 1 << row_pins[row]; //Low
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.INEN = 1; //Input Enable,
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.PULLEN = 1; //Pull Enable
        row_masks[row_ports[row]] |= 1 << row_pins[row]; //Add pin to proper row mask
    for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
        PORT->Group[row_ports[row]].DIRCLR.reg                       = 1 << row_pins[row];  // Input
        PORT->Group[row_ports[row]].OUTCLR.reg                       = 1 << row_pins[row];  // Low
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.INEN   = 1;                   // Input Enable,
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.PULLEN = 1;                   // Pull Enable
        row_masks[row_ports[row]] |= 1 << row_pins[row];                                    // Add pin to proper row mask
    }

    uint8_t col;
    for (col = 0; col < MATRIX_COLS; col++)
    {
        PORT->Group[col_ports[col]].DIRSET.reg = 1 << col_pins[col]; //Output
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col]; //Low
    for (uint8_t col = 0; col < MATRIX_COLS; col++) {
        PORT->Group[col_ports[col]].DIRSET.reg = 1 << col_pins[col];  // Output
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col];  // Low
    }

    matrix_init_quantum();
}

uint64_t mdebouncing = 0;
uint8_t matrix_scan(void)
{
    uint8_t mchanged;
    uint8_t row;
    uint8_t col;
    uint32_t scans[2]; //PA PB

    if (timer_read64() < mdebouncing) return 1; //mdebouncing == 0 when no debouncing active
bool matrix_scan_custom(matrix_row_t current_matrix[]) {
    matrix_row_t raw[MATRIX_ROWS] = {0};
    uint32_t     scans[2];  // PA PB

    memset(mlatest, 0, MATRIX_ROWS * sizeof(matrix_row_t)); //Zero the result buffer
    for (uint8_t col = 0; col < MATRIX_COLS; col++) {
        PORT->Group[col_ports[col]].OUTSET.reg = 1 << col_pins[col];  // Set col output

    for (col = 0; col < MATRIX_COLS; col++)
    {
        PORT->Group[col_ports[col]].OUTSET.reg = 1 << col_pins[col]; //Set col output
        matrix_io_delay();  // Delay for output

        matrix_io_delay(); //Delay for output
        scans[PA] = PORT->Group[PA].IN.reg & row_masks[PA];  // Read PA row pins data
        scans[PB] = PORT->Group[PB].IN.reg & row_masks[PB];  // Read PB row pins data

        scans[PA] = PORT->Group[PA].IN.reg & row_masks[PA]; //Read PA row pins data
        scans[PB] = PORT->Group[PB].IN.reg & row_masks[PB]; //Read PB row pins data
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col];  // Clear col output

        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col]; //Clear col output

        for (row = 0; row < MATRIX_ROWS; row++)
        {
            //Move scan bits from scans array into proper row bit locations
            if (scans[row_ports[row]] & (1 << row_pins[row]))
                mlatest[row] |= 1 << col;
        for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
            // Move scan bits from scans array into proper row bit locations
            if (scans[row_ports[row]] & (1 << row_pins[row])) {
                raw[row] |= 1 << col;
            }
        }
    }

    mchanged = 0; //Default to no matrix change since last

    for (row = 0; row < MATRIX_ROWS; row++)
    {
        if (mlast[row] != mlatest[row])
            mchanged = 1;
        mlast[row] = mlatest[row];
    }

    if (!mchanged)
    {
        for (row = 0; row < MATRIX_ROWS; row++)
            mdebounced[row] = mlatest[row];
        mdebouncing = 0;
    }
    else
    {
        //Begin or extend debounce on change
        mdebouncing = timer_read64() + DEBOUNCE;
    }

    matrix_scan_quantum();

    return 1;
}

matrix_row_t matrix_get_row(uint8_t row)
{
    return mdebounced[row];
}

void matrix_print(void)
{
    char buf[(MATRIX_COLS+8)*(MATRIX_ROWS+1)] = "R C";
    char *pbuf = buf+3;
    uint32_t cols;
    uint32_t rows;
    matrix_row_t row;

    for (cols = 1; cols <= MATRIX_COLS; cols++)
    {
        *pbuf = (cols%10)+48;
        pbuf++;
    }
    *pbuf = '\r'; pbuf++;
    *pbuf = '\n'; pbuf++;

    for (rows = 1; rows <= MATRIX_ROWS; rows++)
    {
        row = matrix_get_row(rows-1);
        if (rows < 10) { *pbuf = rows+48; pbuf++; *pbuf = ' '; pbuf++; *pbuf = ' '; pbuf++; }
        else { *pbuf = (rows/10)+48; pbuf++; *pbuf = (rows%10)+48; pbuf++; *pbuf = ' '; pbuf++; }
        for (cols = 0; cols < MATRIX_COLS; cols++)
        {
            if (row & 1 << cols) *pbuf = 'X';
            else                 *pbuf = '.';
            pbuf++;
    bool changed = false;
    for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
        if (current_matrix[row] != raw[row]) {
            current_matrix[row] = raw[row];
            changed             = true;
        }
        *pbuf = '\r'; pbuf++;
        *pbuf = '\n'; pbuf++;
    }
    *pbuf = 0;
    dprint(buf);

    return changed;
}

M keyboards/massdrop/alt/rules.mk => keyboards/massdrop/alt/rules.mk +1 -1
@@ 6,7 6,7 @@ SRC += config_led.c
ARM_ATSAM = SAMD51J18A
MCU = cortex-m4

CUSTOM_MATRIX = yes
CUSTOM_MATRIX = lite

# Build Options
#   comment out to disable the options.

M keyboards/massdrop/ctrl/matrix.c => keyboards/massdrop/ctrl/matrix.c +36 -142
@@ 16,167 16,61 @@ along with this program.  If not, see <http://www.gnu.org/licenses/>.
*/

#include "ctrl.h"

#include "d51_util.h"
#include "debug.h"
#include "clks.h"
#include <string.h>

#ifndef MATRIX_IO_DELAY
#    define MATRIX_IO_DELAY 1
#endif

matrix_row_t mlatest[MATRIX_ROWS];
matrix_row_t mlast[MATRIX_ROWS];
matrix_row_t mdebounced[MATRIX_ROWS];

uint8_t row_ports[] = { MATRIX_ROW_PORTS };
uint8_t row_pins[] = { MATRIX_ROW_PINS };
uint8_t col_ports[] = { MATRIX_COL_PORTS };
uint8_t col_pins[] = { MATRIX_COL_PINS };
uint32_t row_masks[2]; //NOTE: If more than PA PB used in the future, adjust code to accomodate

__attribute__((weak)) void matrix_io_delay(void) { wait_us(MATRIX_IO_DELAY); }

__attribute__ ((weak))
void matrix_init_kb(void) {
    matrix_init_user();
}

__attribute__ ((weak))
void matrix_scan_kb(void) {
    matrix_scan_user();
}

__attribute__ ((weak))
void matrix_init_user(void) {
}

__attribute__ ((weak))
void matrix_scan_user(void) {
}

void matrix_init(void)
{
    memset(mlatest, 0, MATRIX_ROWS * sizeof(matrix_row_t));
    memset(mlast, 0, MATRIX_ROWS * sizeof(matrix_row_t));
    memset(mdebounced, 0, MATRIX_ROWS * sizeof(matrix_row_t));
const uint8_t row_ports[] = {MATRIX_ROW_PORTS};
const uint8_t row_pins[]  = {MATRIX_ROW_PINS};
const uint8_t col_ports[] = {MATRIX_COL_PORTS};
const uint8_t col_pins[]  = {MATRIX_COL_PINS};
uint32_t      row_masks[2];  // NOTE: If more than PA PB used in the future, adjust code to accommodate

void matrix_init_custom(void) {
    row_masks[PA] = 0;
    row_masks[PB] = 0;

    uint8_t row;
    for (row = 0; row < MATRIX_ROWS; row++)
    {
        PORT->Group[row_ports[row]].DIRCLR.reg = 1 << row_pins[row]; //Input
        PORT->Group[row_ports[row]].OUTCLR.reg = 1 << row_pins[row]; //Low
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.INEN = 1; //Input Enable,
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.PULLEN = 1; //Pull Enable
        row_masks[row_ports[row]] |= 1 << row_pins[row]; //Add pin to proper row mask
    for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
        PORT->Group[row_ports[row]].DIRCLR.reg                       = 1 << row_pins[row];  // Input
        PORT->Group[row_ports[row]].OUTCLR.reg                       = 1 << row_pins[row];  // Low
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.INEN   = 1;                   // Input Enable,
        PORT->Group[row_ports[row]].PINCFG[row_pins[row]].bit.PULLEN = 1;                   // Pull Enable
        row_masks[row_ports[row]] |= 1 << row_pins[row];                                    // Add pin to proper row mask
    }

    uint8_t col;
    for (col = 0; col < MATRIX_COLS; col++)
    {
        PORT->Group[col_ports[col]].DIRSET.reg = 1 << col_pins[col]; //Output
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col]; //Low
    for (uint8_t col = 0; col < MATRIX_COLS; col++) {
        PORT->Group[col_ports[col]].DIRSET.reg = 1 << col_pins[col];  // Output
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col];  // Low
    }

    matrix_init_quantum();
}

uint64_t mdebouncing = 0;
uint8_t matrix_scan(void)
{
    uint8_t mchanged;
    uint8_t row;
    uint8_t col;
    uint32_t scans[2]; //PA PB

    if (timer_read64() < mdebouncing) return 1; //mdebouncing == 0 when no debouncing active
bool matrix_scan_custom(matrix_row_t current_matrix[]) {
    matrix_row_t raw[MATRIX_ROWS] = {0};
    uint32_t     scans[2];  // PA PB

    memset(mlatest, 0, MATRIX_ROWS * sizeof(matrix_row_t)); //Zero the result buffer
    for (uint8_t col = 0; col < MATRIX_COLS; col++) {
        PORT->Group[col_ports[col]].OUTSET.reg = 1 << col_pins[col];  // Set col output

    for (col = 0; col < MATRIX_COLS; col++)
    {
        PORT->Group[col_ports[col]].OUTSET.reg = 1 << col_pins[col]; //Set col output
        matrix_io_delay();  // Delay for output

        matrix_io_delay(); //Delay for output
        scans[PA] = PORT->Group[PA].IN.reg & row_masks[PA];  // Read PA row pins data
        scans[PB] = PORT->Group[PB].IN.reg & row_masks[PB];  // Read PB row pins data

        scans[PA] = PORT->Group[PA].IN.reg & row_masks[PA]; //Read PA row pins data
        scans[PB] = PORT->Group[PB].IN.reg & row_masks[PB]; //Read PB row pins data
        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col];  // Clear col output

        PORT->Group[col_ports[col]].OUTCLR.reg = 1 << col_pins[col]; //Clear col output

        for (row = 0; row < MATRIX_ROWS; row++)
        {
            //Move scan bits from scans array into proper row bit locations
            if (scans[row_ports[row]] & (1 << row_pins[row]))
                mlatest[row] |= 1 << col;
        for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
            // Move scan bits from scans array into proper row bit locations
            if (scans[row_ports[row]] & (1 << row_pins[row])) {
                raw[row] |= 1 << col;
            }
        }
    }

    mchanged = 0; //Default to no matrix change since last

    for (row = 0; row < MATRIX_ROWS; row++)
    {
        if (mlast[row] != mlatest[row])
            mchanged = 1;
        mlast[row] = mlatest[row];
    }

    if (!mchanged)
    {
        for (row = 0; row < MATRIX_ROWS; row++)
            mdebounced[row] = mlatest[row];
        mdebouncing = 0;
    }
    else
    {
        //Begin or extend debounce on change
        mdebouncing = timer_read64() + DEBOUNCE;
    }

    matrix_scan_quantum();

    return 1;
}

matrix_row_t matrix_get_row(uint8_t row)
{
    return mdebounced[row];
}

void matrix_print(void)
{
    char buf[(MATRIX_COLS+8)*(MATRIX_ROWS+1)] = "R C";
    char *pbuf = buf+3;
    uint32_t cols;
    uint32_t rows;
    matrix_row_t row;

    for (cols = 1; cols <= MATRIX_COLS; cols++)
    {
        *pbuf = (cols%10)+48;
        pbuf++;
    }
    *pbuf = '\r'; pbuf++;
    *pbuf = '\n'; pbuf++;

    for (rows = 1; rows <= MATRIX_ROWS; rows++)
    {
        row = matrix_get_row(rows-1);
        if (rows < 10) { *pbuf = rows+48; pbuf++; *pbuf = ' '; pbuf++; *pbuf = ' '; pbuf++; }
        else { *pbuf = (rows/10)+48; pbuf++; *pbuf = (rows%10)+48; pbuf++; *pbuf = ' '; pbuf++; }
        for (cols = 0; cols < MATRIX_COLS; cols++)
        {
            if (row & 1 << cols) *pbuf = 'X';
            else                 *pbuf = '.';
            pbuf++;
    bool changed = false;
    for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
        if (current_matrix[row] != raw[row]) {
            current_matrix[row] = raw[row];
            changed             = true;
        }
        *pbuf = '\r'; pbuf++;
        *pbuf = '\n'; pbuf++;
    }
    *pbuf = 0;
    dprint(buf);

    return changed;
}

M keyboards/massdrop/ctrl/rules.mk => keyboards/massdrop/ctrl/rules.mk +1 -1
@@ 6,7 6,7 @@ SRC += config_led.c
ARM_ATSAM = SAMD51J18A
MCU = cortex-m4

CUSTOM_MATRIX = yes
CUSTOM_MATRIX = lite

# Build Options
#   comment out to disable the options.