* Add a dedicated settings reset on start init priority and default
it to lower priority (high number) than default FLASH_INIT_PRIORITY
to be sure flash is initialized before we open the area.
Added a new CONFIG_ZMK_SETTINGS_RESET_ON_START option which enables init
code to call zmk_settings_erase(), and changed the settings_reset shield
to use it instead of CONFIG_ZMK_BLE_CLEAR_BONDS_ON_START, so it now
resets all settings instead of just clearing BLE bonds.
CONFIG_ZMK_BLE_CLEAR_BONDS_ON_START is left in place for now in case
someone still needs it. It may be replaced in the future once we find a
better way to repair a broken split connection.
Added a zmk_settings_erase() function to clear all saved settings. This
does not go through Zephyr's settings subsystem, but instead directly
clears the data from the setting storage backend, so a reboot is needed
for it to take effect.
* Move to local/stack allocated event API that doesn't require
dynamic allocation/freeing.
* Disable heap, we no longer use alloc/free unless using LVGL.
* Tons of refactors all over to account for the new event approach.
zmk_battery_start_reporting() may be called from battery_event_listener(), which
will result in a bus fault when attempting to read a battery that does not exist
such as on a dongle.
* If attempting to notify and getting an EPERM return value, request
upgrading the security of the connection at that moment, since it
likely means we got a connection to a bonded host but the connection
hasn't been upgraded to encrypted yet.
* Add security related tests to verify behavior when trying to read
a GATT characteristic from our peripheral with and without client
auto security request/retry.
* Auto security request actually makes macOS worse, so disable it,
and remove our early request in favor of using GATT enforcement
to ensure connections are secured.
Note there was one place where a non-strict prototype was actually being used
with an argument, in `zmk_hog_init`. In this case, the actual argument type was
added instead.
* Add ability to fetch and report peripheral battery levels
on split centrals.
* Add additional support for adding a new Battery Level
service to split centrals that exposes fetched peripheral
battery levels to connected hosts.
Co-authored-by: Peter Johanson <peter@peterjohanson.com>
Removed the label property from built-in behaviors, custom behaviors
defined in a few keymaps, and macros generated with ZMK_MACRO().
Now that node names are used to identify behaviors, and names only need
to be unique within the set of behaviors, the names of all behaviors
have been shortened to be similar to their original labels.
This means that any keymaps which reference behavior nodes by name
instead of by label will need to be updated. Keymaps typically use the
labels though, so most keymaps should be unaffected by this change.
Added BEHAVIOR_DT_DEFINE() and BEHAVIOR_DT_INST_DEFINE(), which work
exactly like the DEVICE_*_DEFINE() macros, except they also register the
device as a behavior by adding a pointer to it to a memory section.
Added zmk_behavior_get_binding(), which works like device_get_binding()
except that it only searches the devices that have been registered as
behaviors. This ensures that behaviors cannot have name collisions with
other devices defined by the SoC, which will be important when we remove
the label property from behaviors so they are given their node names.
As an added benefit, this is faster since it searches a smaller list.
Some basic benchmark code I wrote indicates it takes 30-70% as long,
depending on where the behavior is in the list and whether the name
string is an exact pointer match.
From now on, behaviors should use BEHAVIOR_*_DEFINe() instead of
DEVICE_*_DEFINE(), and any code that looks up a behavior by name should
use zmk_behavior_get_binding() instead of device_get_binding().
Changed the property used to define a layer name for displays from
"label" (which affects other things in Zephyr and is deprecated) to
"display-name". (It cannot be named simply "name", because that has
special meaning in newer versions of the devicetree compiler.)
"label" is still supported as a fallback, so no changes need to be made
to existing keymaps.
Changed the label property on zmk,ext-power-generic to be optional and
removed it from existing uses. Renamed the nodes for all non-development
boards to "EXT_POWER" to preserve user settings.
rgb_underglow.c now finds the correct device by finding the first
instance of zmk,ext-power-generic instead of looking for a node named
"EXT_POWER".
Changed all code (except for layer names) which used the label property
to use DEVICE_DT_NAME() instead, which uses the label if set or falls
back to the full node name. This matches how Zephyr determines the node
names used with device_get_binding() and allows us to start removing the
deprecated label property from things.
There is already a function to see if the peripheral is connected, a matching one for if it's bonded is a useful addition for displays/indicators. `is_bonded` gets reset to false in the advertising function in preparation for runtime peripheral bond clearing
* Don't propogate any key press events while in the
middle of passkey entry, avoid funky state on hosts.
* Handle passkey on release, not press, to ensure key *releases*
are not accidentally sent, especially the Enter release
at the very end of passkey entry, which can trigger
cancel in the dialog if the keyboard is connected
via USB to the same host.
* Properly handle the user enabling the
`CONFIG_BT_SMP_ALLOW_UNAUTH_OVERWRITE`
Zephyr flag and handle re-pairing to an existing taken
profile from the same address.
Adds new functionality and a behavior to disconnect an active BLE connection.
The motivation for this is that for some devices like phones, the presence of an
active BLE connection results in the onscreen keyboard being selected.
* Add HID report/descriptor for a new report with mouse buttons,
and x/y/wheel deltas.
* New mouse key press behavior for press/release of mouse keys.
* Add constants for HID main item values (e.g. data/array/absolute)
* Define and use constants for our HID report IDs.
The battery reporting has been known to cause macOS computers to wakeup repeatedly. In some cases (e.g. display or custom lighting implementation) one might want to collect battery SOC without broadcasting over BLE
* Update docs/docs/config/battery.md
Co-authored-by: Cem Aksoylar <caksoylar@users.noreply.github.com>
* Previous version of multiple overrides of the default value of ZMK_USB
were problematic. Move to using board _defconfig files for the
defaults for those, along with proper `depends on` for ZMK_USB that
accounts for split and split roles.
* USB boot protocol support
* Use a single definition of a boot report, used for regular reports in
non-6KRO, and for rollover in all branches.
* Handle gaps in the zmk report when producing a boot report in HKRO mode. For
.example, if it was 8KRO, it would be possible to have the state 0 0 0 0 0 0 0
17 (by pressing 8 keys, and letting go of the first 7). Copying the first 6
bytes would not show up the single pressed key.
* Disable usb status change and callback on SOF events:
SOF events were introduced by the boot protocol changes, and required internally
by Zephyr's idle support, but are unused within ZMK itself. Ignore them in the
usb status callback.
---------
Co-authored-by: Andrew Childs <lorne@cons.org.nz>
This brings the 'global-quick-tap' functionality to combos by filtering
out candidate combos that fell within their own quick tap term.
I also replaced `return 0` with `return ZMK_EV_EVENT_BUBBLE` where appropriate.
(I assume this was done in past as it is similar to errno returning, but
being that this is to signify an event type I find this more clear)
Detaching the global-quick-tap functionality from the quick-tap term.
This makes way for two improvements:
1. This functionality can be added to combos under a unified name
'global-quick-tap-ms'.
2. This allows users to set a lower term for the 'global-quick-tap'
(typically ~100ms), and a higher term for the regular
quick-tap (typically ~200ms)
This deprecates the global-quick-tap option, however if it is set, the
quick-tap-ms value will be copied to global-quick-tap-ms.
Changed the endpoints code to rename the existing endpoint types to
"transport" and add the concept of "endpoint instances". A transport is
the method by which data is sent, while instances allow describing
multiple endpoints that use the same transport (e.g. bluetooth profiles)
Also added new APIs to get the total number of possible endpoint
instances and assign each instance a unique index, which can be used
for tracking separate state for each endpoint in other code files.
* Split centrals to scan with their identity so they receive direct
advertising packets.
* Split peripherals to use direct advertising if they have an existing
bond to a split central.
This commit adds a new GATT characteristics on the peripheral side
and wires it up to read sensor values. The central side subscribes
to this new characteristics and replays sensor values on its side.
Co-authored-by: Peter Johanson <peter@peterjohanson.com>
* Proper usage of bt_uuid_cmp.
* Central's don't start scanning for peripherals if
`ZMK_BLE_CLEAR_BONDS_ON_START` is enabled.
* Split peripherals don't advertize if
`ZMK_BLE_CLEAR_BONDS_ON_START` is enabled.
* Add two new compatibles for macros that
take one or two parameters when bound in
a keymap.
* Use `¯o_param_1to1`, `¯o_param_1to2`, `¯o_param_2to1`,
and `¯o_param_2to2` control entries in the bindings for the macro
to have the next binding entry have it's values substituted.
Co-authored-by: Cem Aksoylar <caksoylar@users.noreply.github.com>
* Refactor sensor events to include channel data,
necessary for prop split encoders, and avoiding duplicate calls,
to fetch channel data twice, etc.
* More consistent behavior driver API.
* Allow setting triggers per resolution at the behavior level optionally.
* Capture the last 6 entered digits, and then require pressing
Enter/Return to submit the entered digits. This matches the
messaging shown on hosts regarding how to complete pairing.
* Fix the wording on the Kconfig menu item to accurately describe
the feature.
Add CONFIG_ZMK_DISPLAY_INVERT Kconfig to invert colors
(black-on-white to white-on-black) on monochrome screens.
Currently applies only if CONFIG_LV_USE_THEME_MONO is selected,
which is the default unless user overrides it.
All displays currently used with ZMK are monochrome so it makes sense
to enable the mono theme by default, which can be disabled by the user
since we use the "imply" statement.
Without this theme setting, the small font size selection for widgets
at the bottom of the stock status screen does not work.
Updated existing pre-commit hooks and added some new hooks:
- Remove trailing whitespace
- Ensure every non-empty file ends with a new line
- Check YAML file validity
- Prevent adding large files
- Ensure any scripts with shebangs are executable
Added a GitHub action to run pre-commit on every commit. Removed any
existing actions which duplicate pre-commit.
Ran pre-commit on the codebase.
* Add new sensor behaviors that either take full bindings
add definition, or accept parameters when bound in the
keymap.
* Remove existing hard-coded key press sensor behavior
and instead leverage new generic sensor behaviors to
achieve the same functionality.
Co-authored-by: nick@conway.dev
Extended the virtual key position system from combos so that each sensor
also gets a virtual key position. This allows sensor behaviors to use
the behavior queue API.
Before this change, if a matrix position was not present in the transform,
various incorrect behaviors would happen:
1) In some cases out-of-bounds accesses:
Note that the size of the`transform[]` array does not necessarily match
the size of the matrix. So for example if key position
(ZMK_MATRIX_COLS-1, ZMK_MATRIX_ROWS-1) is not present in the transform,
but ends up being pressed, then the array will be accessed beyond its
size, and any data could be returned.
2) In other cases the 0th position in the keymap will be used because
the `transform[]` array is initialized to all zeros.
Fixed an issue where the sticky key behavior would call
ZMK_EVENT_RAISE_AFTER(), which would free the provided event, but then
it would keep using that now-freed event data.
* Move away from DT_LABEL.
* Move to DEVICE_DT_GET for non-behavior device access.
* Move various drivers to `gpio_spec_dt` and `DT` related macros.
* Remove mcp23017 while at it, since better upstream driver is
available.
* LV_ prefix from new LVGL official Kconfig now used.
* API changes for themes, container object removal, etc.
* Add our own Kconfig and code for theme default small font.
* Remove some hardcoded sizes.
Added ZMK_MACRO_DEFAULT_WAIT_MS and ZMK_MACRO_DEFAULT_TAP_MS to set
global defaults for the wait-ms and tap-ms properties of macros.
Also reduced the default timings for macros, since it's been reported
many times that 100 ms is too slow.
* Update docs for mod-morph
* Add unit tests for mod-morph
* Add keep-mods to DT binding
Co-authored-by: Martin Aumüller <aumuell@reserv.at>
Co-authored-by: Cem Aksoylar <caksoylar@users.noreply.github.com>
All credit for this one goes to @xudongzheng (thanks for helping debug
this!). Should fix the issue where keyboards go unresponsive after their
host machine wakes from sleep due to the USB driver entering an error
state. I was able to both reliably reproduce the issue before the patch
goes in and confirmed it no longer occurs post patch.
The `USB_DC_RESUME` state indicates the host event has resumed the connection.
Adding it to the list of valid connection states to prevent the error
when waking from sleep.
Zephyr API Link:
https://docs.zephyrproject.org/apidoc/latest/group____usb__device__controller__api.html#gac09e3e0af1a2b41a5bfbad91f900baf7fixes#1372
* This is a very simple fix to a rather complicated issue. Essentially,
hold-taps will "release" (raise) their captured keys before actually
telling the event manager they have captured a key. This means the event
manager ends up assigning the `last_listener_index` to the hold-tap
subscription rather than the combo. So when the combo calls
`ZMK_EVENT_RELEASE` it raises after the hold-tap instead of after the
combo as the combo code expects.
* The corresponding test (which fails without this change) has also been added.
* An event can be captured and released in the same event handler, before
the last_listener_index would have been updated. This causes some handlers
to be triggered multiple times.
* The solution is to update the last_listener_index before calling the next
event handler, so capturing and releasing within an event handler is harmless.
* Also see discussion at https://github.com/zmkfirmware/zmk/pull/1401
* If our handler dedides our undedided hold-tap,
return early before continuing.
* Fix incorrect pointer logic, resulting in combo
candidate filtering leaving incorrect timeout details.
Co-authored-by: Andrew Rae <ajrae.nv@gmail.com>
Co-authored-by: okke <okke@formsma.nl>
Two new options for functionality to enable/disable RGB for
USB status or idle events.
Co-authored-by: Pete Johanson <peter@peterjohanson.com>
Co-authored-by: ReFil <harryherring@gmail.com>
* Proporly handle multiple conditonal layers w/ the same target
`then-layer` values.
* Move handling to work callback, to avoid re-entrance for cascading
layers enabling other layers.
* Add new API/status to track state of the
peripheral connection to the central.
* Add new peripheral status widget for displaying
the current status of the connection to
the central.
battery.c now uses the zmk,battery chosen node to select a battery
sensor. Using the node labeled "BATTERY" is maintained for backwards
compatibility but is now deprecated. Custom boards should switch to
using the chosen node.
# Conflicts:
# app/boards/arm/bluemicro840/bluemicro840_v1.dts
# app/boards/arm/nice60/nice60.dts
# app/boards/arm/nrfmicro/nrfmicro_13.dts
# Conflicts:
# app/boards/arm/bluemicro840/bluemicro840_v1.dts
Peter Johanson
Theo Lemay
ReFil
Joel Spadin
Xudong Zheng
honorless
zhiayang
ctranstrum
Cem Aksoylar
Chris Andreae
moergo-sc
Gabor Hornyak
Alessandro Bortolin
Alexander Krikun
Andrew Rae
Flo Kempenich
Stephen Wan
Nick Winans
Nick Conway
Purdea Andrei
Robert U
Cem Aksoylar
Darryldh
John Drogo
ReFil
WSTRN
Caleb Goates
Okke Formsma
Andrew Rae
DoctorNefario