In order to improve security around the command processing, I plan to
upgrade the communication protocol to use a fixed structure with an
optional payload. As a first pass on this, I added a script to generate
the IDs automatically from the configuration, as well as a script to map
the module names and log levels to integer IDs.
This also replaces the manual lmap_* and array_get_index lookup tables,
so I don't have to always deal with creating an entry in constants.h and
command.c and making sure they are in sync.
This change is an initial step to adding support for profiles in x52d.
This will allow the user to create a keyboard layout in an easy to
read/write text based format, and have it compiled into a flat layout
that's easy for the daemon to parse and load into memory. This layout
can then be used to map a user's action key to the actual input usage
needed. This is necessary, because keyboards don't actually send the
character that is typed, but just the position on the keyboard. For
example, on a French AZERTY keyboard, the A key would actually send the
usage for `Keyboard_q_and_Q`. The OS would translate that into the
letter 'a' (or 'A' if Shift key is held) and pass that to the active
window.
This commit adds the full logic necessary for the layout loading,
validation and compiling, as well as tests for the compiler and loader.
The autotools build infrastructure was deprecated back in 0.3.3, and any
bugs in the Meson build infrastructure have been fixed now. This
eliminates the legacy build scripts which were hard to maintain.
This change add APIs to convert the string forms of the axis and button
names back to their corresponding enum identifiers. This is effectively
built such that a roundtrip of _to_str and _from_str will return the
same input. The _nocase variants handle case insensitive matching of the
names by folding of the ASCII alphabets A-Z and a-z only, so it doesn't
depend on localization.
Prior to this change, the mouse speed was controlled by an opaque
numeric value, that controlled both the speed and the delay between
updates. This caused a lot of choppy behavior with lower speeds, and the
really low speeds had as little as 1% speed difference between them in
the practical pixels/second speed, while there was effectively a 50%
jump in the speed between speed settings 11 and 12, due to the
hyperbolic relationship between steps. Post that, it was an even 25%
increase in sensitivity for every step.
This change modifies it so that the old Speed option is deprecated, it
is now replaced by the Sensitivity option, which is a direct percentage
scale from 10% to 500%. In addition, there is a CurveFactor option to
let it have fine control when there is little deflection, and move
faster when further away from the center. This also adds an
IsometricMode option which computes the speed as a function of the
cartesian distance from the center (`sqrt(dx^2 + dy^2)`). The default
behavior uses the existing linear speed which controls the speed of the
X and Y axes independently, but now uses the sensitivity and curve
factors to get better behavior. Also, the mouse events are consistently
reported every 10ms. This should make it a lot smoother.
Finally, this change also adds a Deadzone factor, which allows the user
to ignore small changes near the center of the joystick that can cause
mouse drift. This deadzone uses the total distance, so if just the X or
Y axis has moved, it will still allow suppressing any play in the thumb
stick.
Issue: #44
Previously, x52d had conditional compilation flags linking to libevdev
on Linux. However, I wanted to change this so that we use an abstraction
layer that will link to the appropriate backend (evdev on Linux only for
now). The idea is that we get rid of all conditional compilation blocks
and deal with the backend through the vkm library.
This new library handles the mouse scrolling, clicking, as well as
adding the ability to include keyboard events so that we can support the
profiles feature in a future commit.
When enabling --warnlevel=3 during Meson setup, the build threw up
several warnings, mostly related to either unused parameters, or
sometimes an integer type mismatch. This commit addresses all of those
changes and ensures that the build does not contain any unnecessary
warnings.
Prior to this change, the check was for an explicit -Dnls=enabled,
however, if the option was never set, it defaulted to disabled. With
this change, unless explicitly disabled, the Meson build system will
automatically build with NLS support.
The previous version of the Meson build files did not handle the po
directory correctly, and lost a lot of information. As part of the
migration away from Autotools, this is one more item that needs to be
checked off.
CI: [ci skip] [doxy skip]
Meson is a far more robust build framework, compared to autotools. This
greatly simplifies adding new features, since it's far easier to
maintain a set of meson.build files vs the autotools mishmash.
DEPRECATION NOTICE: Autotools based build is deprecated and will be
removed in the future.
SonarQube cloud identified a maintainability issue based on MISRA C
guidelines that prohibit backward jumps. While not a mandatory fix, it
helps to clean up the codebase and improves readability.
Ref. MISRA C:2012, 15.2 - The goto statement shall jump to a label
declared later in the same function.
This change adds logic to block signals on child threads. This is done
so that we can migrate signal handling to the main thread, even if we
add additional threads in the future.
This change adds a ReverseScroll parameter to the configuration, which
if set, will change the direction of the scroll wheel of the virtual
mouse.
Github-Issue: #45
This change adds a test suite for validating mouse configuration. This
only tests the logic for mouse thread enable/disable and mouse speed
calculations.
The configuration processing works as follows - the configuration
parameters are set in memory, and an apply function applies all the
changes directly to the device. While this works during startup and
reload, it adds a needless overhead when processing config set commands
from the command socket.
This change makes it such that a `config set` command would update the
configuration for that particular key, and then immediately call the
callback function for that key. This has the effect that individual
configuration changes are visible immediately, without having to reapply
every other configuration that hasn't changed.
This commit also removes the `config apply` command, since it is no
longer needed. The `config load` and `config reload` already handle
applying the configuration after reading it.
Prior to this change, the IO thread would always report a mouse button
event, even if the mouse was disabled in configuration, or the uinput
device was not created. This results in unexpected button/wheel events.
This change checks that the uinput device has been created and the mouse
has been enabled in the configuration before reporting a button or wheel
event.
This change adds the `logging show` and `logging set` commands to the
command processor module. This allows setting the log levels of the
individual modules via the socket.
The command allows the client to retrieve individual parameters from the
configuration. This follows a similar syntax to the 'config set'
command, with the client supplying the section and key, and if there is
a matching entry in the configuration, it will return the corresponding
value.
This change allows setting the configuration from the socket. It behaves
similar to the override, and requires the client to send the section,
key and value, and responds with an OK or error value.
This change adds the generic command parsing logic, as well as the
response for the following commands:
* config load <file>
* config reload
* config save
* config dump <file>
This change makes the protocol more strict, in that the buffer sent to
the daemon must be a series of NUL separated arguments. This makes it a
little easier to handle the strings, especially those that may have
embedded whitespace.
This change adds the logic to read a packet from the socket, accept
connections from clients, and close connections from clients that have
hung up. This commit does not yet have support for parsing and handling
the commands, and simply echoes the request back to the client.
This change makes X52 daemon listen on a Unix socket. This is in
preparation for changes that will read from the socket and allow clients
to communicate with and control the daemon.
Prior to this change, all the generated HTML documentation and locale
files were using the `x52pro-linux` name. That name is no longer
reflective of the project, since it works on macOS and with the non-Pro
X52 as well.
This change is also reflective of the patch used in the released PPA,
which takes care of this anyway, but is being added to the sources to
better integrate with PKGBUILD/AUR/MPR.
nirenjan