darkkirb/qmk_firmware
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
qmk_firmware/keyboards/bpiphany/frosty_flake/matrix.c

86 lines
3.3 KiB
C
Raw Normal View History

sketch out frosty flake matrix using custom matrix.c based on kitten paw
2017-03-18 23:19:10 +00:00
/*
Swap rows and columns to match the schematic
2017-03-20 03:33:41 +00:00
Copyright 2017 Gabriel Young <gabeplaysdrums@live.com>
sketch out frosty flake matrix using custom matrix.c based on kitten paw
2017-03-18 23:19:10 +00:00
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 2 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/>.
*/
#include <util/delay.h>
#include "matrix.h"
fix whitespace
2017-03-20 03:46:56 +00:00
static matrix_row_t scan_col(void) {
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
// Each of the 8 columns is read off pins as below
// 7 6 5 4 3 2 1 0
// ,--,--,--,--,--,--,--,--,
// |B0|B3|B2|B1|B6|B4|B5|C7|
// `--`--`--`--`--`--`--`--`
fix whitespace
2017-03-20 03:46:56 +00:00
return (
(PINC&(1<<7) ? 0 : ((matrix_row_t)1<<0)) |
(PINB&(1<<5) ? 0 : ((matrix_row_t)1<<1)) |
(PINB&(1<<4) ? 0 : ((matrix_row_t)1<<2)) |
(PINB&(1<<6) ? 0 : ((matrix_row_t)1<<3)) |
(PINB&(1<<1) ? 0 : ((matrix_row_t)1<<4)) |
(PINB&(1<<2) ? 0 : ((matrix_row_t)1<<5)) |
(PINB&(1<<3) ? 0 : ((matrix_row_t)1<<6)) |
(PINB&(1<<0) ? 0 : ((matrix_row_t)1<<7))
);
}
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
static void select_row(uint8_t row) {
switch (row) {
fix whitespace
2017-03-20 03:46:56 +00:00
case 0: PORTD = (PORTD & ~0b01111011) | 0b00011011; break;
case 1: PORTD = (PORTD & ~0b01111011) | 0b01000011; break;
case 2: PORTD = (PORTD & ~0b01111011) | 0b01101010; break;
case 3: PORTD = (PORTD & ~0b01111011) | 0b01111001; break;
case 4: PORTD = (PORTD & ~0b01111011) | 0b01100010; break;
case 5: PORTD = (PORTD & ~0b01111011) | 0b01110001; break;
case 6: PORTD = (PORTD & ~0b01111011) | 0b01100001; break;
case 7: PORTD = (PORTD & ~0b01111011) | 0b01110000; break;
case 8: PORTD = (PORTD & ~0b01111011) | 0b01100000; break;
case 9: PORTD = (PORTD & ~0b01111011) | 0b01101000; break;
case 10: PORTD = (PORTD & ~0b01111011) | 0b00101011; break;
case 11: PORTD = (PORTD & ~0b01111011) | 0b00110011; break;
case 12: PORTD = (PORTD & ~0b01111011) | 0b00100011; break;
case 13: PORTD = (PORTD & ~0b01111011) | 0b01111000; break;
case 14: PORTD = (PORTD & ~0b01111011) | 0b00010011; break;
case 15: PORTD = (PORTD & ~0b01111011) | 0b01101001; break;
case 16: PORTD = (PORTD & ~0b01111011) | 0b00001011; break;
case 17: PORTD = (PORTD & ~0b01111011) | 0b00111011; break;
}
}
sketch out frosty flake matrix using custom matrix.c based on kitten paw
2017-03-18 23:19:10 +00:00
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
void matrix_init_custom(void) {
Swap rows and columns to match the schematic
2017-03-20 03:33:41 +00:00
/* Row output pins */
DDRD |= 0b01111011;
/* Column input pins */
DDRC &= ~0b10000000;
DDRB &= ~0b01111111;
PORTC |= 0b10000000;
PORTB |= 0b01111111;
sketch out frosty flake matrix using custom matrix.c based on kitten paw
2017-03-18 23:19:10 +00:00
}
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
// matrix is 18 uint8_t.
// we select the row (one of 18), then read the column
bool matrix_scan_custom(matrix_row_t current_matrix[]) {
bool has_changed = false;
fix whitespace
2017-03-20 03:46:56 +00:00
for (uint8_t row = 0; row < MATRIX_ROWS; row++) {
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
matrix_row_t orig = current_matrix[row];
select_row(row);
_delay_us(3);
current_matrix[row] = scan_col();
has_changed |= (orig != current_matrix[row]);
successfully read data from the matrix
2017-03-19 22:59:21 +00:00
}
sketch out frosty flake matrix using custom matrix.c based on kitten paw
2017-03-18 23:19:10 +00:00
Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (#8243) * Keyboard: revamp frosty-flake leds This commit transitions bpiphany/frosty_flake to led_update_{kb,user} and rewrites the AVR bit twiddling logic to use the standard QMK GPIO API. * Keyboard: rewrite frosty_flake's matrix reader to be a lite custom matrix This commit replaces frosty_flake's custom matrix and debounce logic with a "lite" custom matrix. In addition to being somewhat clearer, this allows a consumer of the flake board to choose their own debouncing algorithm. The one closest to the implementation originally in use is sym_g, but this opens us up to supporting eager_pk and eager_pr. The original matrix code was 18 columns for 8 rows, but using a single row read and unpacking the bits into individual columns. To simplify, I've changed the key layout to be 8C 18R instead of 18C 8R: this lets us use a single read directly into the matrix _and_ drop down to a uint8_t instead of a uint32_t for matrix_row_t. Since we're no longer implementing our own debouncing and row unpacking, we save ~400 bytes on the final firmware image. Fully tested against a CM Storm QFR hosting the flake -- this commit message was written using the new matrix code. Firmware Sizes (assuming stock configuration as of 42d6270f2) Matrix+Debounce Size (bytes) --------------- ------------ original 17740 new + sym_g 17284 new + eager_pr 18106 new + eager_pk 18204 I expect that there are some scanning speed benefits as well. * Keyboard: update frosty_flake's UNUSED_PINS * Keyboard: Remove meaningless weak redefinitions from frosty These are not necessary (and all of them already live somewhere in Quantum).
2020-03-02 05:17:09 +00:00
return has_changed;
Fix frosty_flake warning when print is disabled
2017-04-09 19:19:52 +00:00
}
Reference in a new issue Copy permalink