audio: fix deadlock on draining

The playback thread may obviously still fill the AO'S entire audio
buffer, which means it unset p->draining, which makes no sense and broke
ao_drain(). So just don't unset it here.

Not sure if this really fixes this, it was hard to reproduce. Regression
due to the recent changes. There are probably many more bugs like this.
Stupid asynchronous nightmare state machine. Give me a language that
supports formal verification (in presence of concurrency) or something.

audio/out/buffer.c

@@ -632,7 +632,7 @@ static void ao_play_data(struct ao *ao)
     bool ok = true;
     int written = 0;
     if (samples) {
-        p->draining = is_eof;
+        p->draining |= is_eof;
         MP_STATS(ao, "start ao fill");
         ok = ao->driver->write(ao, planes, samples);
         MP_STATS(ao, "end ao fill");