diff options
author | Liyang HU <github.com@liyang.hu> | 2021-08-07 08:12:57 +0100 |
---|---|---|
committer | GitHub <noreply@github.com> | 2021-08-07 17:12:57 +1000 |
commit | 383fae55c5d53626eda210298f15fa6810510d5e (patch) | |
tree | 001eabe3cab3ff3e45f06e64e260ec8c99f32312 /keyboards/al1 | |
parent | a03aa301def77c867ae6c6c840f7fc82b26d91d6 (diff) |
Refactor `quantum/command.{c,h}` for code size & {read,maintain}ability (#11842)
* quantum/command.c: coalesce `print()`s in `command_common_help()` & `print_version()`
Also undo some damage by clang-format in b624f32f94
* quantum/command.c: replace `print(…); print_{,val_}{dec,hex}*(…);` sequences with single `xprintf(…)`
`print_{dec,hex}*(…)` are just `#define`s for `xprintf(…)` anyway.
Each additional `xprintf(…)` costs ~8 bytes: the call instructions,
plus an additional NUL terminator.
This _really_ adds up: this commit saves 814 bytes on my ATmega32.
* quantum/command.c: optimise `mousekey_console()` for size & legibility
Made various tweaks to the interface, but still functionally identical.
Assume `KC_1`…`KC_0` to be contiguous, and removed `numkey2num(…)` entirely.
It was exported in `command.h` by 1a0bac8bcc for no obvious reason, before
which it was `static`. I doubt anyone uses it.
`mousekey_console()` is now enabled regardless of `MK_3_SPEED`.
Needs fleshing out for things other than the X11 variant though.
This commit saves 638 bytes on my ATmega32.
Diffstat (limited to 'keyboards/al1')
0 files changed, 0 insertions, 0 deletions