This reverts commit 01fe9c99.

EVGA/magewell devices seem to use the default system drivers rather than
custom drivers, which causes their audio to become decoupled and treated
as completely separate devices rather than as an audio pin on the video
device. Basically, this would cause those devices to not have audio by
default, and force the user to have to manually select the audio, which
is bad user experience.

We already had a workaround for this with elgato devices, so expand that
code to become a whitelist of devices, and include EVGA/magewell
devices.

We were performing a custom cursor draw for WGC, which we attempted
because we thought we could bring back hardware cursor responsiveness,
but MS says that it won't. Since the custom draw was bugged to cause
black screens anyway, just use the native cursor draw for now.

Rivatuner Statistics Server calls vkGetInstanceProcAddr for
vkCreateDevice with a NULL instance. According to spec, this should
return NULL, but RTSS will dereference this and crash.

This change returns function pointers for every intercepted function
regardless of instance, which is enough to keep RTSS alive.

ffmpeg-mux does not notice if ffmpeg returns an error from
av_interleaved_write_frame() which means that OBS never knows if there is a
problem in ffmpeg.

This is the biggest issue for cases like srt:// or tcp:// streams that can
regularly fail. Without this change OBS never knows that something went wrong.

If using ffmpeg-mux to send a network stream, signal a disconnect error code
on error to allow reconnections to take place.

The VolumeMeter widgets were apparently being drawn as transparent
widgets, which meant that it was necessary to redraw everything under
the widgets in order to properly draw the widgets, so the entire mixer
section of the window was being redrawn every time the meters updated.
This caused a significant perf hit when the only thing wanted/desired
was just to update the meters. This was more noticeable after changing
the audio meter update rate to 60hz from 30hz.

The fix is to simply mark them as opaque widgets, and paint the
background ourselves rather than rely on what's under the window. CPU
perf for painting the main window has been vastly improved because of
this; CPU usage of Qt in the main window is now reduced by at least half
of what it was, if not more.

FreeBSD removed the swig30 package and replaced it with just swig,
currently version 4.0:
https://bugs.freebsd.org/246613

A generally accepted path for DepsPath or LibfdkPath is the parent
directory that includes the 'bin' and 'include' directories. This fix
ensures FindLibfdk.cmake acts like other cmake helper scripts.

Closes obsproject/obs-studio#3474

Remove weabook.live from services.json

It seems like weabook.live is no longer a live streaming provider.

As with Windows and macOS there's no need to check for ChromeOS on
FreeBSD.

Fixes #3106

Fixes issues as described in https://bugreports.qt.io/browse/QTBUG-47660
and halves CPU usage of OBS' UI by about 40-50%.

Instead of having "Add [transition]" at the top of the combo box
blocking out the values, move the add values to the bottom of the box
instead for slightly better UX.

Fixes an issue where command prompts would pop up when registering the
virtual camera in the auto-updater

Fixes a race condition where a hash value could be overwritten from
different threads, causing corruption

If the tray is disabled in settings, we no longer allow the command line
parameter to override it and create a tray icon regardless. This matches
the behavior in the UI, where "minimize to tray on startup" becomes
unavailable if the "tray enabled" setting is unselected.

Remove a few unused variables and other minor changes. No functional
differences. Issues detected by PVS Studio.

This is a temporary fix as to not to delay the release, and will need to
be replaced with AddValue in future versions.

v4l2 pixfmt fourcc's are always stored little-endian.
ref: https://www.kernel.org/doc/html/latest/userspace-api/media/v4l/vidioc-enum-fmt.html?highlight=vidioc_enum_fmt

unlike other YUV based formats nv12 chroma plane has the same padding as
the luma plane.
ref: https://01.org/linuxgraphics/gfx-docs/drm/media/uapi/v4l/pixfmt-nv12.html

Due to the fact that a global was used on GenerateSpecifiedFilename to
save the remux file name, when a screenshot was made, it would overwrite
the filename being remuxed, because screenshots use the same function to
generate filenames as well.

This solves that problem by removing the global and the changes to
GeneratedSpecifiedFilename, and isolating that to the output handler.

Coincidentally, this bug probably also happened with replay buffers
under certain circumstances.

Fixes obsproject/obs-studio#3497
Closes obsproject/obs-studio#3498

Remove support for Vista as the platform update is no longer available
from Microsoft. Allows retrying certain error conditions such as when
OBS is already running. Moved in-use file checks to after the install
directory is selected. Added some comments.

deps/media-playback: Fix current time at not standard speed