This fixes a bug where selecting "Always on top" in the plugin editor
window in Cubase 11 on Windows 11 on a display with > 100% scale would
cause the editor to display with the wrong scale factor.

It seems that Cubase calls removed() then attached() on the plugin view,
destroying and recreating the editor wrapper component. The editor
opened with 100% scale, but requests from the host to set the "real"
scale factor were ignored because the JUCE wrapper thought that the
requested scale had already been applied.

Applying the cached scale factor directly after constructing the editor
keeps everything in a consistent state, and seems to resolve the issue.

This change allows mouse events (including enter and exit events) to
reach unfocused views on macOS. This matches the behaviour of unfocused
windows on Linux and Windows, where components paint in their "hovered"
states even when the application window is in the background.

As a byproduct of using tracking areas on macOS, we can remove the fake
mouse move generator.

Hosts such as REAPER normalise the program parameter value by dividing
the program value by the step count, rather than going via the
parameter's toNormalized function. To be compatible, we should use the
same scaling technique. At time of writing, the coversion process is
detailed under the heading "Conversion of normalized values" on this
page:

https://developer.steinberg.help/display/VST/Parameters+and+Automation

Adds a mechanism to notify the host that the plugin state needs saving,
using updateHostDisplay.

Also allows JUCE hosts to detect when a plugin needs its state saving
via the AudioProcessorListener.

Previously, updating the program or bypass parameter could cause an
out-of-bounds access into the parameter value cache. This is because
AudioProcessorParameter::getParameterIndex() was used to index into the
cache, but the parameter index could be negative for parameters that
had not been added to the AudioProcessor.

We now use the appropriate index in the cache for parameters that
have not been added to the AudioProcessor.

To use this feature, derive your AudioProcessor from
VST3ClientExtensions and override getPluginHasMainInput() to return
false. The main input bus will then be designated as an Aux bus, rather
than a Main bus.

This is mainly useful for synth plugins like vocoders, which may need a
sidechain audio input, but which should replace all audio on the channel
with the output of the synth, rather than mixing with the audio input.

See commit message of a1337a38 for more details

This function is required by the Steinberg VST3 validator from VST 3.7
onwards.

This fixes warnings that are emitted when building with the `-analyze`
flag enabled.

In cases where a VST3 plugin set a non-zero latency in its
prepareToPlay, and then set its latency to 0 later on, the host was not
notified about the latency change.

In older versions of the VST3SDK, hostContext is a raw pointer,
in newer versions it's a smart pointer. If we do manual
refcounting with the smart pointer, we may cause leaks.

Previously, IEditController parameter indices were being used to index
into the AudioProcessor parameter array, but these parameter indices are
not guaranteed to point to the same parameter (parameter groups may
cause reordering on JUCE's side). Now, we use the IEditController
indices universally.

Audition incorrectly increments the ref-count of the IPlugView* returned from IEditController::createView() and never fully releases it. This commit reverts e0306d25 for Audition to ensure that an editor is always returned, otherwise Audition displays a blank GUI.

Previously, all VST3 parameter changes were signalled to the host via
performEdit, regardless of the originating thread. This breaks the
contract of the IEditController interface, which requires that all calls
are made from the UI thread.

This change modifies the VST3 wrapper so that it will only call
performEdit when a parameter is set on the message thread. If a
parameter is set on the audio thread instead, then the parameter change
will be signalled to the host using
`ProcessData::outputParameterChanges`.

If your VST3 plugin uses a background thread to update parameters for
some reason, you'll encounter data races. Please don't do that.

When loading a JUCE plugin in Live, Live was producing diagnostic
messages in its Log.txt warning that restartComponent was being called
from a non-UI thread. This change should silence this warning, and
improve compliance with the VST3 spec.