This fixes an issue where a decoding backend can report configurations
outside of miniaudio's limits.

Public issues:
  * https://github.com/dr-soft/miniaudio/issues/166
  * https://github.com/dr-soft/miniaudio/issues/167

This is uninitializing the ring buffer before the device which results
in a situation where a stray data callback may try touching the buffer
after it's be uninitialized.

This is an internal change and does not affect any public APIs. This
change is in preparation for future lock-free high level APIs.

This updates ma_linear_resampler_get_expected_output_frame_count() to
calculate the expected output frame count in constant time rather than
linear time. This should improve performances of anything using
ma_calculate_frame_count_after_resampling() which includes
ma_decoder_get_length_in_pcm_frames().

These are completely untested. Nothing here is wired up to the main
system - they're just for experimenting with some ideas.

This is in preparation for self-managed message handling.

When a decoding backend is not present, initialization should abort
immediately with MA_NO_BACKEND, however it is including code for no
reason. This commit excludes that useless code from the build.

The thread priority can be set via ma_thread_create() and can be set in
the context config along side the thread priority for configuring the
size of the stack for the audio thread.

This addresses some 64-bit issues with stdio with older versions of VC.

This needs to be #define-d before the header *and* implementation
because it affects the size of publicly visible structures.

A streaming data source keeps in memory only two pages of audio data
and dynamically loads data from a background thread. It is essentially
a double buffering system - as one page is playing, the other is being
loaded by the async thread.

The size of a single page is defined by the following macro:

    MA_RESOURCE_MANAGER_PAGE_SIZE_IN_MILLISECONDS

By default this is currently set to 1 second of audio data. This means
each page has 1 second to load which should be plenty of time. If you
need additional time, the only way to do it is increase the size of the
page by changing the value of the above macro.

  * ma_resource_manager_uninit() has been implemented.
  * Bug fixes and inserting and removing data buffers from the BST.
  * Some old experimental code has been removed.
  * Minor whitespace clean up.