Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling #8243
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
DHowett
changed the title
|
CI build failure seems unrelated to this change |
It was unrelated. |
fauxpark
requested changes
Feb 25, 2020
|
|
||
| led_set_user(usb_led); | ||
| return true; | ||
| } | ||
|
|
||
| __attribute__ ((weak)) |
There was a problem hiding this comment.
Everything from this line onwards can be deleted.
There was a problem hiding this comment.
Happy to. I kept most of these since they were in the original implementation, but I wasn’t sure if they were only valuable as a “here’s how to write a keyboard” template or had practical use.
There was a problem hiding this comment.
Done- I also pulled the two above which were matrix_{init,scan}_kb and were also present in quantum/
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.
…trix 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 42d6270) 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.
DHowett
force-pushed
the
frosty_rewrite
branch
from
a41a55f to
cace9e7
Compare
These are not necessary (and all of them already live somewhere in Quantum).
DHowett
force-pushed
the
frosty_rewrite
branch
from
cace9e7 to
6d18380
Compare
drashna
approved these changes
Mar 1, 2020
fauxpark
approved these changes
Mar 1, 2020
|
FWIW: I maintained the four commits indicated without intermediate "address PR feedback" commits for if you have a rebasing pull request workflow. 😄 Thanks for the review! |
zvecr
approved these changes
Mar 2, 2020
|
Thanks! |
Shinichi-Ohki
added a commit
to Shinichi-Ohki/qmk_firmware
that referenced
this pull request
Mar 3, 2020
* 'master' of https://github.com/qmk/qmk_firmware: (37 commits) Update Hungarian keymap and add sendstring LUT (qmk#8220) Remove "ugly hack in usb_main.c" comments (qmk#8296) Update encoder functions for Iris VIA keymap (qmk#8295) Reduce PROGMEM usage for sendstring LUT (qmk#8109) [Docs] Update ISP Flashing guide (qmk#8149) Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (qmk#8243) Add onekey keymap for testing reset to bootloader. (qmk#8288) Get the direction right on the S75 encoder (qmk#8287) Prune out pure software pwm && custom driver && remove wrapping BACKLIGHT_PIN (qmk#8041) Make a fix to savage65 and tmov2 for via (qmk#8286) format code according to conventions [skip ci] Short term fix for conflicting types for 'tfp_printf' (qmk#8157) Fix recent clang-format breaking quantum.c (qmk#8282) format code according to conventions [skip ci] Remove duplicate BRTG case (qmk#8277) Clean up includes for glcdfont headers (qmk#7745) Fix the Breaking Changes doc again [Docs] translated 'feature_tap_dance.md' to japanese. (qmk#8137) PWM DMA based RGB Underglow for STM32 (qmk#7928) Add VIA support to Prime_M. Clean up all files (qmk#8247) ...
HokieGeek
pushed a commit
to HokieGeek/qmk_firmware
that referenced
this pull request
Mar 5, 2020
…mk#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 42d6270) 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).
nesth
pushed a commit
to nesth/qmk_firmware
that referenced
this pull request
Mar 5, 2020
* upstream/master: (37 commits) Add RGB lighting through one of the free pins (qmk#8294) [Keymap] Adding the 4sStylZ xd75 (qmk#8285) YD60MQ refactor and Configurator layout support (qmk#8313) [Keyboard] Forget to ifdef Super16 led config (qmk#8314) [Keyboard] Switch to RGB Matrix for Super16 (qmk#8305) [Keymap] Keymap Update (qmk#8309) New Keyboard: SiddersKB Majbritt (Pronounced My Brit) (qmk#8260) [Keyboard] VIA Support: Tada68 (qmk#8289) [Keyboard] LFK78 refactor (qmk#7835) [Keymap] new userspace for ibnuda (qmk#8221) [Keymap] Add crd's equinox keymap (qmk#8251) [Keymap] Feature/alfrdmalr/keymap update (qmk#8174) Fix bootloader definition for namecard2x4 (qmk#8301) Update Hungarian keymap and add sendstring LUT (qmk#8220) Remove "ugly hack in usb_main.c" comments (qmk#8296) Update encoder functions for Iris VIA keymap (qmk#8295) Reduce PROGMEM usage for sendstring LUT (qmk#8109) [Docs] Update ISP Flashing guide (qmk#8149) Rewrite the Bathroom Epiphanies Frosty Flake matrix and LED handling (qmk#8243) Add onekey keymap for testing reset to bootloader. (qmk#8288) ...
c0psrul3
pushed a commit
to c0psrul3/qmk_firmware
that referenced
this pull request
Mar 23, 2020
…mk#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 42d6270) 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).
sowbug
pushed a commit
to sowbug/qmk_firmware
that referenced
this pull request
Apr 2, 2020
…mk#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 42d6270) 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).
kylekuj
pushed a commit
to kylekuj/qmk_firmware
that referenced
this pull request
Apr 21, 2020
…mk#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 42d6270) 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).
jakeisnt
pushed a commit
to jakeisnt/qmk_firmware
that referenced
this pull request
Aug 20, 2020
…mk#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 42d6270) 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).
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This pull request comprises three individual commits with their own descriptions, copied below.
dc994e883 - 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 42d6270)
I expect that there are some scanning speed benefits as well.
df14844619b - Keyboard: update frosty_flake's UNUSED_PINS
66bb555d - 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.
f3428f9be09 - Keyboards: Remove meaningless weak redefinitions from frosty
These are not necessary (and all of them already live somewhere in
Quantum).
Types of Changes
Checklist