- This was added early on as a hack to protect against some concurrency issues.
- It's not clear that this serves any purpose anymore, and if it does, individual components should be fixed rather than using a global recursive mutex.
MSVC supplied with VS2022 generates "warning C4189: 'CALIBRATION_THRESHOLD':
local variable is initialized but not referenced" which is treated as an
error.
Circumvent it by moving constexpr variable directly into body of lambda function.
I have `134850146304` bytes of ram and Yuzu was saying that I had `125.59 GB`
of ram. But `125.59` is actually the amount of gi**bi**bytes I have. In
gi**ga**bytes I would have `134.9`.
Additionally, I changed the `1024 / 1024 / 1024` here into the `_GiB`
user-literals that I added a while ago(#6519).
https://www.wolframalpha.com/input/?i=134850146304+bytes
SDL_HINT_JOYSTICK_HIDAPI_SWITCH_HOME_LED is no longer needed thanks to new default in SDL 2.0.18.
SDL_HINT_JOYSTICK_HIDAPI_XBOX is reported to cause conflicts with native driver Xbox driver on Linux, and Xbox controllers don't benefit from hidapi anyways.
CallbackStatus instances aren't the cheapest things to copy around
(relative to everything else), given that they're currently 520 bytes in
size and are currently copied numerous times when callbacks are invoked.
Instead, we can pass the status by const reference to avoid all the
copying.
Silences quite a bit of -Wdocumentation warnings, given the @param tag
is only intended to be used to identify function parameters, not what it
contains.
* this resolves the todo items in the CMakeLists.txt
* a version requirement check for ffmpeg is added to catch issues early
* for future-proof reasons, nasm/yasm is now only required when build on
x86/AMD64 systems
We were unconditionally accessing the keyboard_buttons array, even if the bottom_osk_index was for the numberpad, leading to an out of bounds array access. Fix this by accessing the proper array for the current button when the index is for the numberpad.