Fix SEGV and deadlock when stopping async reads

[charlescochran]

As of mercuryapi-1.31.4.35, when tm_reader_async.c's TMR_stopReading()
is called by Reader_stop_reading(), it ultimately ends up waiting for
reader->readState to become TMR_READ_STATE_DONE while holding the GIL.
This happens in tm_reader_async.c's do_background_reads(), but not until
the tag queue is emptied by read callbacks. These are triggered by
invoke_read_callback(), but this function blocks until it can acquire
the GIL. This frequently results in deadlock upon attempting to stop an
asynchronous read.

Previously, this was addressed by temporarily setting self->readCallback
(and self->statsCallback) to NULL in Reader_stop_reading(), but this
causes a SEGV in invoke_read_callback() if it occurs after
self->readCallback is verified to not be NULL but before it is called.
Instead of temporarily setting these function pointers to NULL, release
the GIL in Reader_stop_reading(), allowing invoke_read_callback() to
empty the tag queue and thereby avoiding deadlock. Then, use a mutex to
prevent concurrent calls to TMR_stopReading(), since this was previously
prevented by the GIL.

Temporarily setting self->statsCallback to NULL in Reader_stop_reading()
also prevented stats callbacks from occurring in subsequent asynchronous
reads. This functionality is now fixed.

[charlescochran]

@gotthardp @natcl

Can we merge this? I think it is beneficial to the library, and it may fix the root cause of a few of the currently open issues. See this discussion for detailed explanations concerning a different but-related upstream bug which also exists.

[gotthardp]

@charlescochran Thank you! Looks reasonable, but I was not able to test it though.