| Index: third_party/WebKit/Source/modules/webaudio/AudioNodeOutput.cpp
|
| diff --git a/third_party/WebKit/Source/modules/webaudio/AudioNodeOutput.cpp b/third_party/WebKit/Source/modules/webaudio/AudioNodeOutput.cpp
|
| index 3fa33d607efbf0be43734f71480cc488ef9ebda1..790fd0bb0c8afaf97ce69db01a74ed09ae087210 100644
|
| --- a/third_party/WebKit/Source/modules/webaudio/AudioNodeOutput.cpp
|
| +++ b/third_party/WebKit/Source/modules/webaudio/AudioNodeOutput.cpp
|
| @@ -70,11 +70,13 @@ void AudioNodeOutput::setNumberOfChannels(unsigned numberOfChannels) {
|
| m_desiredNumberOfChannels = numberOfChannels;
|
|
|
| if (deferredTaskHandler().isAudioThread()) {
|
| - // If we're in the audio thread then we can take care of it right away (we should be at the very start or end of a rendering quantum).
|
| + // If we're in the audio thread then we can take care of it right away (we
|
| + // should be at the very start or end of a rendering quantum).
|
| updateNumberOfChannels();
|
| } else {
|
| DCHECK(!m_didCallDispose);
|
| - // Let the context take care of it in the audio thread in the pre and post render tasks.
|
| + // Let the context take care of it in the audio thread in the pre and post
|
| + // render tasks.
|
| deferredTaskHandler().markAudioNodeOutputDirty(this);
|
| }
|
| }
|
| @@ -109,7 +111,8 @@ void AudioNodeOutput::propagateChannelCount() {
|
| ASSERT(deferredTaskHandler().isGraphOwner());
|
|
|
| if (isChannelCountKnown()) {
|
| - // Announce to any nodes we're connected to that we changed our channel count for its input.
|
| + // Announce to any nodes we're connected to that we changed our channel
|
| + // count for its input.
|
| for (AudioNodeInput* i : m_inputs)
|
| i->handler().checkNumberOfChannelsForInput(i);
|
| }
|
| @@ -119,11 +122,13 @@ AudioBus* AudioNodeOutput::pull(AudioBus* inPlaceBus, size_t framesToProcess) {
|
| DCHECK(deferredTaskHandler().isAudioThread());
|
| DCHECK(m_renderingFanOutCount > 0 || m_renderingParamFanOutCount > 0);
|
|
|
| - // Causes our AudioNode to process if it hasn't already for this render quantum.
|
| - // We try to do in-place processing (using inPlaceBus) if at all possible,
|
| - // but we can't process in-place if we're connected to more than one input (fan-out > 1).
|
| - // In this case pull() is called multiple times per rendering quantum, and the processIfNecessary() call below will
|
| - // cause our node to process() only the first time, caching the output in m_internalOutputBus for subsequent calls.
|
| + // Causes our AudioNode to process if it hasn't already for this render
|
| + // quantum. We try to do in-place processing (using inPlaceBus) if at all
|
| + // possible, but we can't process in-place if we're connected to more than one
|
| + // input (fan-out > 1). In this case pull() is called multiple times per
|
| + // rendering quantum, and the processIfNecessary() call below will cause our
|
| + // node to process() only the first time, caching the output in
|
| + // m_internalOutputBus for subsequent calls.
|
|
|
| m_isInPlace = inPlaceBus &&
|
| inPlaceBus->numberOfChannels() == numberOfChannels() &&
|
|
|