JorgeGzm opened a new pull request, #19592:
URL: https://github.com/apache/nuttx/pull/19592
# boards: follow the unified keyboard input in the two lvglterm
configurations
## Summary
This is the last of three parts. It names an application option, so it goes
with apache/nuttx-apps#NNNN.
Those two cannot be green at the same time, and it is worth saying why
before anybody reads the CI. Each repository is built against the other's
master, and there is no way to pair two branches, so this one sees an apps
master where `EXAMPLES_LVGLTERM_INPUT_KBD` and `SYSTEM_KBD` do not exist
yet, and the apps side sees board configurations that still name the option
it renamed. The failures on both sides are confined to the normalisation
step of the same two configurations, and both go green once both are in.
The pair was verified together, see Testing.
The LVGL terminal used to have three input sources to choose from, one per
kind of keyboard, and now has two: touch, or any keyboard. The two in-tree
configurations that named the old option are updated here.
* **Linum STM32H753BI.** Beyond the rename, the configuration now enables
`UINPUT_KEYBOARD` and `SYSTEM_KBD` so that the terminal can be driven from
the serial console with no keyboard plugged in, which is documented in the
board page. `INPUT_KEYBOARD_BYTESTREAM` is no longer needed there, since
the terminal reads events and this configuration has no other keyboard
consumer.
* **M5Stack Cardputer.** The rename only. Its driver was converted in
apache/nuttx#19586, so its cursor keys already arrive as keycodes.
The application documentation follows the same change: the lvglterm page
described the three sources and told the reader to pick the one matching the
hardware, the hidkbd page pointed at `HIDKBD_ENCODED`, which no longer
exists, and `system/kbd` had no page at all.
## Impact
Left as it was, the unknown option would have been dropped when the
configuration was normalised and both boards would have fallen back to the
touch variant, which is the default. Neither has a touchscreen, so the
terminal would have had no input at all, and the build would have said
nothing about it.
`UINPUT_KEYBOARD_BUFNUMBER` is raised to 128 on the Linum. It counts events
rather than keys, so the default of eight holds four keystrokes, and a
console hands over a whole line at once. The keyboard upper half overwrites
the oldest event when the buffer is full, so a typed line arrived with its
beginning silently missing.
`HIDKBD_NOGETREPORT` is enabled on the Linum as well. Sampling the keyboard
over the control pipe every 40 ms loses any key pressed and released between
two samples, which on this board meant most of them.
## Testing
Verified inside the CI container with this branch paired against the apps
branch, which is the state after both merge:
```
### Normalisation
clean linum-stm32h753bi:lvglterm_kbda clean sama5d4-ek:ipv6
clean esp32s3-m5-cardputer:lvglterm clean olimex-lpc1766stk:hidkbd
clean sama5d4-ek:nsh clean nucleo-h753zi:jumbo
clean sama5d4-ek:nxwm clean qemu-intel64:jumbo
clean sim:nxdoom
### Build
PASS linum-stm32h753bi:lvglterm_kbda PASS olimex-lpc1766stk:hidkbd
PASS esp32s3-m5-cardputer:lvglterm PASS sim:nxdoom
```
The documentation was built the way CI builds it, with warnings treated as
errors:
```
$ python -m sphinx -b html -W --keep-going -q . /tmp/doc
$ echo $?
0
```
On the Linum, the terminal was driven on the board's LCD from a real USB
keyboard, from the serial console through the uinput bridge, and from both
at once with the USB keyboard forwarded into the same virtual keyboard.
The Cardputer was not available, so that configuration is checked by build
only.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
