Move AudioDecoder initialization into AudioRenderer.

Move AudioDecoder initialization into AudioRenderer.

More or less the audio equivalent of r151132. Even though we don't use multiple audio decoders today I added it anyway since it's not a lot of code and keeps things consistent.

BUG=145635

[scherkus (not reviewing), 2012-09-07 13:59:54]

You reviewed the VideoRenderer CL so here's the AudioRenderer variant!

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
File media/filters/audio_renderer_impl.cc (left):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:168: bool config_ok =
algorithm_->ValidateConfig(channels, sample_rate,
this was some miscellaneous cleanup

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
File media/filters/audio_renderer_impl.cc (right):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:518: CHECK(bytes_per_second);
calling algorithm_->ValidateConfig() should prevent this from becoming true

[Ami GONE FROM CHROMIUM, 2012-09-07 14:40:02]

I disagree with your "it's not a lot of code" assessment, and think simplicity
in the audio-renderer path is more useful than similarity w/ the video-renderer
path.

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/filter_c...
File media/base/filter_collection.h (right):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/filter_c...
media/base/filter_collection.h:49: void
SelectAudioDecoder(scoped_refptr<AudioDecoder>* out);
delete

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/filter_c...
File media/base/filter_collection_unittest.cc (right):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/filter_c...
media/base/filter_collection_unittest.cc:1: // Copyright (c) 2011 The Chromium
Authors. All rights reserved.
Delete this file?

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/pipeline...
File media/base/pipeline_unittest.cc (right):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/base/pipeline...
media/base/pipeline_unittest.cc:180: bool disable_after_init_cb = false) {
wat default value!

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
File media/filters/audio_renderer_impl.cc (left):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:52: DCHECK(sink_.get());
this is just random cleanup?

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
File media/filters/audio_renderer_impl.cc (right):

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:135: base::AutoLock l(lock_);
"l" is not stylish

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:148: statistics_cb_ = statistics_cb;
This is only used during the init dance, right?
Why not Bind it into the init calls, instead?

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:162: lock_.AssertAcquired();
this locking is making me very sad.
If you dropped the support for multiple decoders could you drop the locking
de-sad me?
It's kind of crazy to have all this code here considering it'll never other than
exactly one decoder to chew on.

https://chromiumcodereview.appspot.com/10918022/diff/5001/media/filters/audio...
media/filters/audio_renderer_impl.cc:165: scoped_refptr<AudioDecoder> decoder =
decoders->front();
put it straight into decoder_ instead?

[scherkus (not reviewing), 2012-10-18 00:31:01]

Sadly closing this as xhwang re-coded this as
http://codereview.chromium.org/11148011/ because it turns out we do need
multiple audio decoders for handling decryption.

[scherkus (not reviewing), 2012-10-18 00:49:34]

I salvaged my ARI cleanup + rebased on xhwang's:
https://codereview.chromium.org/11183048/