Moving iOS Audio Device to sdk.

This change forks the existing iOS audio device module and audio device
from modules/audio_device/ into sdk/objc/Framework. It also updates
RTCPeerConnectionFactory to use the forked implementation.

The unit tests are re-implemented as XCTests.

(was: https://webrtc-review.googlesource.com/c/src/+/67300)

Bug: webrtc:9120
Change-Id: I07340505137b16c2dd487569ad0112f984557bba
Reviewed-on: https://webrtc-review.googlesource.com/75125
Reviewed-by: Kári Helgason <kthelgason@webrtc.org>
Commit-Queue: Peter Hanspers <peterhanspers@webrtc.org>
Cr-Commit-Position: refs/heads/master@{#23208}

sdk/objc/Framework/Native/src/audio/audio_device_ios.h

@@ -0,0 +1,293 @@
+/*
+ *  Copyright (c) 2012 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#ifndef MODULES_AUDIO_DEVICE_IOS_AUDIO_DEVICE_IOS_H_
+#define MODULES_AUDIO_DEVICE_IOS_AUDIO_DEVICE_IOS_H_
+
+#include <memory>
+
+#include "sdk/objc/Framework/Headers/WebRTC/RTCMacros.h"
+#include "modules/audio_device/audio_device_generic.h"
+#include "audio_session_observer.h"
+#include "voice_processing_audio_unit.h"
+#include "rtc_base/buffer.h"
+#include "rtc_base/thread.h"
+#include "rtc_base/thread_annotations.h"
+#include "rtc_base/thread_checker.h"
+
+RTC_FWD_DECL_OBJC_CLASS(RTCNativeAudioSessionDelegateAdapter);
+
+namespace webrtc {
+
+class FineAudioBuffer;
+
+namespace ios_adm {
+
+// Implements full duplex 16-bit mono PCM audio support for iOS using a
+// Voice-Processing (VP) I/O audio unit in Core Audio. The VP I/O audio unit
+// supports audio echo cancellation. It also adds automatic gain control,
+// adjustment of voice-processing quality and muting.
+//
+// An instance must be created and destroyed on one and the same thread.
+// All supported public methods must also be called on the same thread.
+// A thread checker will RTC_DCHECK if any supported method is called on an
+// invalid thread.
+//
+// Recorded audio will be delivered on a real-time internal I/O thread in the
+// audio unit. The audio unit will also ask for audio data to play out on this
+// same thread.
+class AudioDeviceIOS : public AudioDeviceGeneric,
+                       public AudioSessionObserver,
+                       public VoiceProcessingAudioUnitObserver,
+                       public rtc::MessageHandler {
+ public:
+  AudioDeviceIOS();
+  ~AudioDeviceIOS();
+
+  void AttachAudioBuffer(AudioDeviceBuffer* audioBuffer) override;
+
+  InitStatus Init() override;
+  int32_t Terminate() override;
+  bool Initialized() const override;
+
+  int32_t InitPlayout() override;
+  bool PlayoutIsInitialized() const override;
+
+  int32_t InitRecording() override;
+  bool RecordingIsInitialized() const override;
+
+  int32_t StartPlayout() override;
+  int32_t StopPlayout() override;
+  bool Playing() const override { return playing_; }
+
+  int32_t StartRecording() override;
+  int32_t StopRecording() override;
+  bool Recording() const override { return recording_; }
+
+  // These methods returns hard-coded delay values and not dynamic delay
+  // estimates. The reason is that iOS supports a built-in AEC and the WebRTC
+  // AEC will always be disabled in the Libjingle layer to avoid running two
+  // AEC implementations at the same time. And, it saves resources to avoid
+  // updating these delay values continuously.
+  // TODO(henrika): it would be possible to mark these two methods as not
+  // implemented since they are only called for A/V-sync purposes today and
+  // A/V-sync is not supported on iOS. However, we avoid adding error messages
+  // the log by using these dummy implementations instead.
+  int32_t PlayoutDelay(uint16_t& delayMS) const override;
+
+  // Native audio parameters stored during construction.
+  // These methods are unique for the iOS implementation.
+  int GetPlayoutAudioParameters(AudioParameters* params) const override;
+  int GetRecordAudioParameters(AudioParameters* params) const override;
+
+  // These methods are currently not fully implemented on iOS:
+
+  // See audio_device_not_implemented.cc for trivial implementations.
+  int32_t ActiveAudioLayer(
+      AudioDeviceModule::AudioLayer& audioLayer) const override;
+  int32_t PlayoutIsAvailable(bool& available) override;
+  int32_t RecordingIsAvailable(bool& available) override;
+  int16_t PlayoutDevices() override;
+  int16_t RecordingDevices() override;
+  int32_t PlayoutDeviceName(uint16_t index,
+                            char name[kAdmMaxDeviceNameSize],
+                            char guid[kAdmMaxGuidSize]) override;
+  int32_t RecordingDeviceName(uint16_t index,
+                              char name[kAdmMaxDeviceNameSize],
+                              char guid[kAdmMaxGuidSize]) override;
+  int32_t SetPlayoutDevice(uint16_t index) override;
+  int32_t SetPlayoutDevice(
+      AudioDeviceModule::WindowsDeviceType device) override;
+  int32_t SetRecordingDevice(uint16_t index) override;
+  int32_t SetRecordingDevice(
+      AudioDeviceModule::WindowsDeviceType device) override;
+  int32_t InitSpeaker() override;
+  bool SpeakerIsInitialized() const override;
+  int32_t InitMicrophone() override;
+  bool MicrophoneIsInitialized() const override;
+  int32_t SpeakerVolumeIsAvailable(bool& available) override;
+  int32_t SetSpeakerVolume(uint32_t volume) override;
+  int32_t SpeakerVolume(uint32_t& volume) const override;
+  int32_t MaxSpeakerVolume(uint32_t& maxVolume) const override;
+  int32_t MinSpeakerVolume(uint32_t& minVolume) const override;
+  int32_t MicrophoneVolumeIsAvailable(bool& available) override;
+  int32_t SetMicrophoneVolume(uint32_t volume) override;
+  int32_t MicrophoneVolume(uint32_t& volume) const override;
+  int32_t MaxMicrophoneVolume(uint32_t& maxVolume) const override;
+  int32_t MinMicrophoneVolume(uint32_t& minVolume) const override;
+  int32_t MicrophoneMuteIsAvailable(bool& available) override;
+  int32_t SetMicrophoneMute(bool enable) override;
+  int32_t MicrophoneMute(bool& enabled) const override;
+  int32_t SpeakerMuteIsAvailable(bool& available) override;
+  int32_t SetSpeakerMute(bool enable) override;
+  int32_t SpeakerMute(bool& enabled) const override;
+  int32_t StereoPlayoutIsAvailable(bool& available) override;
+  int32_t SetStereoPlayout(bool enable) override;
+  int32_t StereoPlayout(bool& enabled) const override;
+  int32_t StereoRecordingIsAvailable(bool& available) override;
+  int32_t SetStereoRecording(bool enable) override;
+  int32_t StereoRecording(bool& enabled) const override;
+
+  // AudioSessionObserver methods. May be called from any thread.
+  void OnInterruptionBegin() override;
+  void OnInterruptionEnd() override;
+  void OnValidRouteChange() override;
+  void OnCanPlayOrRecordChange(bool can_play_or_record) override;
+  void OnChangedOutputVolume() override;
+
+  // VoiceProcessingAudioUnitObserver methods.
+  OSStatus OnDeliverRecordedData(AudioUnitRenderActionFlags* flags,
+                                 const AudioTimeStamp* time_stamp,
+                                 UInt32 bus_number,
+                                 UInt32 num_frames,
+                                 AudioBufferList* io_data) override;
+  OSStatus OnGetPlayoutData(AudioUnitRenderActionFlags* flags,
+                            const AudioTimeStamp* time_stamp,
+                            UInt32 bus_number,
+                            UInt32 num_frames,
+                            AudioBufferList* io_data) override;
+
+  // Handles messages from posts.
+  void OnMessage(rtc::Message *msg) override;
+
+  bool IsInterrupted();
+
+ private:
+  // Called by the relevant AudioSessionObserver methods on |thread_|.
+  void HandleInterruptionBegin();
+  void HandleInterruptionEnd();
+  void HandleValidRouteChange();
+  void HandleCanPlayOrRecordChange(bool can_play_or_record);
+  void HandleSampleRateChange(float sample_rate);
+  void HandlePlayoutGlitchDetected();
+  void HandleOutputVolumeChange();
+
+  // Uses current |playout_parameters_| and |record_parameters_| to inform the
+  // audio device buffer (ADB) about our internal audio parameters.
+  void UpdateAudioDeviceBuffer();
+
+  // Since the preferred audio parameters are only hints to the OS, the actual
+  // values may be different once the AVAudioSession has been activated.
+  // This method asks for the current hardware parameters and takes actions
+  // if they should differ from what we have asked for initially. It also
+  // defines |playout_parameters_| and |record_parameters_|.
+  void SetupAudioBuffersForActiveAudioSession();
+
+  // Creates the audio unit.
+  bool CreateAudioUnit();
+
+  // Updates the audio unit state based on current state.
+  void UpdateAudioUnit(bool can_play_or_record);
+
+  // Configures the audio session for WebRTC.
+  bool ConfigureAudioSession();
+  // Unconfigures the audio session.
+  void UnconfigureAudioSession();
+
+  // Activates our audio session, creates and initializes the voice-processing
+  // audio unit and verifies that we got the preferred native audio parameters.
+  bool InitPlayOrRecord();
+
+  // Closes and deletes the voice-processing I/O unit.
+  void ShutdownPlayOrRecord();
+
+  // Ensures that methods are called from the same thread as this object is
+  // created on.
+  rtc::ThreadChecker thread_checker_;
+
+  // Native I/O audio thread checker.
+  rtc::ThreadChecker io_thread_checker_;
+
+  // Thread that this object is created on.
+  rtc::Thread* thread_;
+
+  // Raw pointer handle provided to us in AttachAudioBuffer(). Owned by the
+  // AudioDeviceModuleImpl class and called by AudioDeviceModule::Create().
+  // The AudioDeviceBuffer is a member of the AudioDeviceModuleImpl instance
+  // and therefore outlives this object.
+  AudioDeviceBuffer* audio_device_buffer_;
+
+  // Contains audio parameters (sample rate, #channels, buffer size etc.) for
+  // the playout and recording sides. These structure is set in two steps:
+  // first, native sample rate and #channels are defined in Init(). Next, the
+  // audio session is activated and we verify that the preferred parameters
+  // were granted by the OS. At this stage it is also possible to add a third
+  // component to the parameters; the native I/O buffer duration.
+  // A RTC_CHECK will be hit if we for some reason fail to open an audio session
+  // using the specified parameters.
+  AudioParameters playout_parameters_;
+  AudioParameters record_parameters_;
+
+  // The AudioUnit used to play and record audio.
+  std::unique_ptr<VoiceProcessingAudioUnit> audio_unit_;
+
+  // FineAudioBuffer takes an AudioDeviceBuffer which delivers audio data
+  // in chunks of 10ms. It then allows for this data to be pulled in
+  // a finer or coarser granularity. I.e. interacting with this class instead
+  // of directly with the AudioDeviceBuffer one can ask for any number of
+  // audio data samples. Is also supports a similar scheme for the recording
+  // side.
+  // Example: native buffer size can be 128 audio frames at 16kHz sample rate.
+  // WebRTC will provide 480 audio frames per 10ms but iOS asks for 128
+  // in each callback (one every 8ms). This class can then ask for 128 and the
+  // FineAudioBuffer will ask WebRTC for new data only when needed and also
+  // cache non-utilized audio between callbacks. On the recording side, iOS
+  // can provide audio data frames of size 128 and these are accumulated until
+  // enough data to supply one 10ms call exists. This 10ms chunk is then sent
+  // to WebRTC and the remaining part is stored.
+  std::unique_ptr<FineAudioBuffer> fine_audio_buffer_;
+
+  // Temporary storage for recorded data. AudioUnitRender() renders into this
+  // array as soon as a frame of the desired buffer size has been recorded.
+  // On real iOS devices, the size will be fixed and set once. For iOS
+  // simulators, the size can vary from callback to callback and the size
+  // will be changed dynamically to account for this behavior.
+  rtc::BufferT<int16_t> record_audio_buffer_;
+
+  // Set to 1 when recording is active and 0 otherwise.
+  volatile int recording_;
+
+  // Set to 1 when playout is active and 0 otherwise.
+  volatile int playing_;
+
+  // Set to true after successful call to Init(), false otherwise.
+  bool initialized_ RTC_GUARDED_BY(thread_checker_);
+
+  // Set to true after successful call to InitRecording() or InitPlayout(),
+  // false otherwise.
+  bool audio_is_initialized_;
+
+  // Set to true if audio session is interrupted, false otherwise.
+  bool is_interrupted_;
+
+  // Audio interruption observer instance.
+  RTCNativeAudioSessionDelegateAdapter* audio_session_observer_
+      RTC_GUARDED_BY(thread_checker_);
+
+  // Set to true if we've activated the audio session.
+  bool has_configured_session_ RTC_GUARDED_BY(thread_checker_);
+
+  // Counts number of detected audio glitches on the playout side.
+  int64_t num_detected_playout_glitches_ RTC_GUARDED_BY(thread_checker_);
+  int64_t last_playout_time_ RTC_GUARDED_BY(io_thread_checker_);
+
+  // Counts number of playout callbacks per call.
+  // The value isupdated on the native I/O thread and later read on the
+  // creating thread (see thread_checker_) but at this stage no audio is
+  // active. Hence, it is a "thread safe" design and no lock is needed.
+  int64_t num_playout_callbacks_;
+
+  // Contains the time for when the last output volume change was detected.
+  int64_t last_output_volume_change_time_ RTC_GUARDED_BY(thread_checker_);
+};
+}  // namespace ios_adm
+}  // namespace webrtc
+
+#endif  // MODULES_AUDIO_DEVICE_IOS_AUDIO_DEVICE_IOS_H_

sdk/objc/Framework/Native/src/audio/audio_device_module_ios.h

@@ -0,0 +1,143 @@
+/*
+ *  Copyright (c) 2012 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#ifndef SDK_IOS_NATIVE_API_AUDIO_DEVICE_MODULE_AUDIO_DEVICE_IOS_H_
+#define SDK_IOS_NATIVE_API_AUDIO_DEVICE_MODULE_AUDIO_DEVICE_IOS_H_
+
+#include <memory>
+
+#include "audio_device_ios.h"
+
+#include "modules/audio_device/audio_device_buffer.h"
+#include "modules/audio_device/include/audio_device.h"
+#include "rtc_base/checks.h"
+#include "rtc_base/criticalsection.h"
+
+
+namespace webrtc {
+
+class AudioDeviceGeneric;
+
+namespace ios_adm {
+
+  class AudioDeviceModuleIOS : public AudioDeviceModule {
+
+  public:
+
+    int32_t AttachAudioBuffer();
+
+    AudioDeviceModuleIOS();
+    ~AudioDeviceModuleIOS() override;
+
+    // Retrieve the currently utilized audio layer
+    int32_t ActiveAudioLayer(AudioLayer* audioLayer) const override;
+
+    // Full-duplex transportation of PCM audio
+    int32_t RegisterAudioCallback(AudioTransport* audioCallback) override;
+
+    // Main initializaton and termination
+    int32_t Init() override;
+    int32_t Terminate() override;
+    bool Initialized() const override;
+
+    // Device enumeration
+    int16_t PlayoutDevices() override;
+    int16_t RecordingDevices() override;
+    int32_t PlayoutDeviceName(uint16_t index,
+                              char name[kAdmMaxDeviceNameSize],
+                              char guid[kAdmMaxGuidSize]) override;
+    int32_t RecordingDeviceName(uint16_t index,
+                                char name[kAdmMaxDeviceNameSize],
+                                char guid[kAdmMaxGuidSize]) override;
+
+    // Device selection
+    int32_t SetPlayoutDevice(uint16_t index) override;
+    int32_t SetPlayoutDevice(WindowsDeviceType device) override;
+    int32_t SetRecordingDevice(uint16_t index) override;
+    int32_t SetRecordingDevice(WindowsDeviceType device) override;
+
+    // Audio transport initialization
+    int32_t PlayoutIsAvailable(bool* available) override;
+    int32_t InitPlayout() override;
+    bool PlayoutIsInitialized() const override;
+    int32_t RecordingIsAvailable(bool* available) override;
+    int32_t InitRecording() override;
+    bool RecordingIsInitialized() const override;
+
+    // Audio transport control
+    int32_t StartPlayout() override;
+    int32_t StopPlayout() override;
+    bool Playing() const override;
+    int32_t StartRecording() override;
+    int32_t StopRecording() override;
+    bool Recording() const override;
+
+    // Audio mixer initialization
+    int32_t InitSpeaker() override;
+    bool SpeakerIsInitialized() const override;
+    int32_t InitMicrophone() override;
+    bool MicrophoneIsInitialized() const override;
+
+    // Speaker volume controls
+    int32_t SpeakerVolumeIsAvailable(bool* available) override;
+    int32_t SetSpeakerVolume(uint32_t volume) override;
+    int32_t SpeakerVolume(uint32_t* volume) const override;
+    int32_t MaxSpeakerVolume(uint32_t* maxVolume) const override;
+    int32_t MinSpeakerVolume(uint32_t* minVolume) const override;
+
+    // Microphone volume controls
+    int32_t MicrophoneVolumeIsAvailable(bool* available) override;
+    int32_t SetMicrophoneVolume(uint32_t volume) override;
+    int32_t MicrophoneVolume(uint32_t* volume) const override;
+    int32_t MaxMicrophoneVolume(uint32_t* maxVolume) const override;
+    int32_t MinMicrophoneVolume(uint32_t* minVolume) const override;
+
+    // Speaker mute control
+    int32_t SpeakerMuteIsAvailable(bool* available) override;
+    int32_t SetSpeakerMute(bool enable) override;
+    int32_t SpeakerMute(bool* enabled) const override;
+
+    // Microphone mute control
+    int32_t MicrophoneMuteIsAvailable(bool* available) override;
+    int32_t SetMicrophoneMute(bool enable) override;
+    int32_t MicrophoneMute(bool* enabled) const override;
+
+    // Stereo support
+    int32_t StereoPlayoutIsAvailable(bool* available) const override;
+    int32_t SetStereoPlayout(bool enable) override;
+    int32_t StereoPlayout(bool* enabled) const override;
+    int32_t StereoRecordingIsAvailable(bool* available) const override;
+    int32_t SetStereoRecording(bool enable) override;
+    int32_t StereoRecording(bool* enabled) const override;
+
+    // Delay information and control
+    int32_t PlayoutDelay(uint16_t* delayMS) const override;
+
+    bool BuiltInAECIsAvailable() const override;
+    int32_t EnableBuiltInAEC(bool enable) override;
+    bool BuiltInAGCIsAvailable() const override;
+    int32_t EnableBuiltInAGC(bool enable) override;
+    bool BuiltInNSIsAvailable() const override;
+    int32_t EnableBuiltInNS(bool enable) override;
+
+#if defined(WEBRTC_IOS)
+    int GetPlayoutAudioParameters(AudioParameters* params) const override;
+    int GetRecordAudioParameters(AudioParameters* params) const override;
+#endif  // WEBRTC_IOS
+    private:
+    bool initialized_ = false;
+    std::unique_ptr<AudioDeviceIOS> audio_device_;
+    std::unique_ptr<AudioDeviceBuffer> audio_device_buffer_;
+
+  };
+} // namespace ios_adm
+} // namespace webrtc
+
+#endif  // SDK_IOS_NATIVE_API_AUDIO_DEVICE_MODULE_AUDIO_DEVICE_IOS_H_

sdk/objc/Framework/Native/src/audio/audio_device_module_ios.mm

@@ -0,0 +1,673 @@
+/*
+ *  Copyright (c) 2012 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include "audio_device_module_ios.h"
+
+#include "modules/audio_device/audio_device_config.h"
+#include "modules/audio_device/audio_device_generic.h"
+#include "rtc_base/checks.h"
+#include "rtc_base/logging.h"
+#include "rtc_base/refcount.h"
+#include "rtc_base/refcountedobject.h"
+#include "system_wrappers/include/metrics.h"
+
+#if defined(WEBRTC_IOS)
+#include "audio_device_ios.h"
+#endif
+
+#define CHECKinitialized_() \
+  {                         \
+    if (!initialized_) {    \
+      return -1;            \
+    };                      \
+  }
+
+#define CHECKinitialized__BOOL() \
+  {                              \
+    if (!initialized_) {         \
+      return false;              \
+    };                           \
+  }
+
+
+namespace webrtc {
+namespace ios_adm {
+
+  AudioDeviceModuleIOS::AudioDeviceModuleIOS() {
+    RTC_LOG(INFO) << "current platform is IOS";
+    RTC_LOG(INFO) << "iPhone Audio APIs will be utilized.";
+  }
+
+  int32_t AudioDeviceModuleIOS::AttachAudioBuffer() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    audio_device_->AttachAudioBuffer(audio_device_buffer_.get());
+    return 0;
+  }
+
+  AudioDeviceModuleIOS::~AudioDeviceModuleIOS() {
+    RTC_LOG(INFO) << __FUNCTION__;
+  }
+
+  int32_t AudioDeviceModuleIOS::ActiveAudioLayer(AudioLayer* audioLayer) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    AudioLayer activeAudio;
+    if (audio_device_->ActiveAudioLayer(activeAudio) == -1) {
+      return -1;
+    }
+    *audioLayer = activeAudio;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::Init() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    if (initialized_)
+      return 0;
+
+    audio_device_buffer_.reset(new webrtc::AudioDeviceBuffer());
+    audio_device_.reset(new ios_adm::AudioDeviceIOS());
+    RTC_CHECK(audio_device_);
+
+    this->AttachAudioBuffer();
+
+    AudioDeviceGeneric::InitStatus status = audio_device_->Init();
+    RTC_HISTOGRAM_ENUMERATION(
+        "WebRTC.Audio.InitializationResult", static_cast<int>(status),
+        static_cast<int>(AudioDeviceGeneric::InitStatus::NUM_STATUSES));
+    if (status != AudioDeviceGeneric::InitStatus::OK) {
+      RTC_LOG(LS_ERROR) << "Audio device initialization failed.";
+      return -1;
+    }
+    initialized_ = true;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::Terminate() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    if (!initialized_)
+      return 0;
+    if (audio_device_->Terminate() == -1) {
+      return -1;
+    }
+    initialized_ = false;
+    return 0;
+  }
+
+  bool AudioDeviceModuleIOS::Initialized() const {
+    RTC_LOG(INFO) << __FUNCTION__ << ": " << initialized_;
+    return initialized_;
+  }
+
+  int32_t AudioDeviceModuleIOS::InitSpeaker() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    return audio_device_->InitSpeaker();
+  }
+
+  int32_t AudioDeviceModuleIOS::InitMicrophone() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    return audio_device_->InitMicrophone();
+  }
+
+  int32_t AudioDeviceModuleIOS::SpeakerVolumeIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->SpeakerVolumeIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetSpeakerVolume(uint32_t volume) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << volume << ")";
+    CHECKinitialized_();
+    return audio_device_->SetSpeakerVolume(volume);
+  }
+
+  int32_t AudioDeviceModuleIOS::SpeakerVolume(uint32_t* volume) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    uint32_t level = 0;
+    if (audio_device_->SpeakerVolume(level) == -1) {
+      return -1;
+    }
+    *volume = level;
+    RTC_LOG(INFO) << "output: " << *volume;
+    return 0;
+  }
+
+  bool AudioDeviceModuleIOS::SpeakerIsInitialized() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    bool isInitialized = audio_device_->SpeakerIsInitialized();
+    RTC_LOG(INFO) << "output: " << isInitialized;
+    return isInitialized;
+  }
+
+  bool AudioDeviceModuleIOS::MicrophoneIsInitialized() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    bool isInitialized = audio_device_->MicrophoneIsInitialized();
+    RTC_LOG(INFO) << "output: " << isInitialized;
+    return isInitialized;
+  }
+
+  int32_t AudioDeviceModuleIOS::MaxSpeakerVolume(uint32_t* maxVolume) const {
+    CHECKinitialized_();
+    uint32_t maxVol = 0;
+    if (audio_device_->MaxSpeakerVolume(maxVol) == -1) {
+      return -1;
+    }
+    *maxVolume = maxVol;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::MinSpeakerVolume(uint32_t* minVolume) const {
+    CHECKinitialized_();
+    uint32_t minVol = 0;
+    if (audio_device_->MinSpeakerVolume(minVol) == -1) {
+      return -1;
+    }
+    *minVolume = minVol;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SpeakerMuteIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->SpeakerMuteIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetSpeakerMute(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    return audio_device_->SetSpeakerMute(enable);
+  }
+
+  int32_t AudioDeviceModuleIOS::SpeakerMute(bool* enabled) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool muted = false;
+    if (audio_device_->SpeakerMute(muted) == -1) {
+      return -1;
+    }
+    *enabled = muted;
+    RTC_LOG(INFO) << "output: " << muted;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::MicrophoneMuteIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->MicrophoneMuteIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetMicrophoneMute(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    return (audio_device_->SetMicrophoneMute(enable));
+  }
+
+  int32_t AudioDeviceModuleIOS::MicrophoneMute(bool* enabled) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool muted = false;
+    if (audio_device_->MicrophoneMute(muted) == -1) {
+      return -1;
+    }
+    *enabled = muted;
+    RTC_LOG(INFO) << "output: " << muted;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::MicrophoneVolumeIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->MicrophoneVolumeIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetMicrophoneVolume(uint32_t volume) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << volume << ")";
+    CHECKinitialized_();
+    return (audio_device_->SetMicrophoneVolume(volume));
+  }
+
+  int32_t AudioDeviceModuleIOS::MicrophoneVolume(uint32_t* volume) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    uint32_t level = 0;
+    if (audio_device_->MicrophoneVolume(level) == -1) {
+      return -1;
+    }
+    *volume = level;
+    RTC_LOG(INFO) << "output: " << *volume;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::StereoRecordingIsAvailable(
+      bool* available) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->StereoRecordingIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetStereoRecording(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    if (audio_device_->RecordingIsInitialized()) {
+      RTC_LOG(WARNING) << "recording in stereo is not supported";
+      return -1;
+    }
+    if (audio_device_->SetStereoRecording(enable) == -1) {
+      RTC_LOG(WARNING) << "failed to change stereo recording";
+      return -1;
+    }
+    int8_t nChannels(1);
+    if (enable) {
+      nChannels = 2;
+    }
+    audio_device_buffer_.get()->SetRecordingChannels(nChannels);
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::StereoRecording(bool* enabled) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool stereo = false;
+    if (audio_device_->StereoRecording(stereo) == -1) {
+      return -1;
+    }
+    *enabled = stereo;
+    RTC_LOG(INFO) << "output: " << stereo;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::StereoPlayoutIsAvailable(bool* available) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->StereoPlayoutIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetStereoPlayout(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    if (audio_device_->PlayoutIsInitialized()) {
+      RTC_LOG(LERROR)
+      << "unable to set stereo mode while playing side is initialized";
+      return -1;
+    }
+    if (audio_device_->SetStereoPlayout(enable)) {
+      RTC_LOG(WARNING) << "stereo playout is not supported";
+      return -1;
+    }
+    int8_t nChannels(1);
+    if (enable) {
+      nChannels = 2;
+    }
+    audio_device_buffer_.get()->SetPlayoutChannels(nChannels);
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::StereoPlayout(bool* enabled) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool stereo = false;
+    if (audio_device_->StereoPlayout(stereo) == -1) {
+      return -1;
+    }
+    *enabled = stereo;
+    RTC_LOG(INFO) << "output: " << stereo;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::PlayoutIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->PlayoutIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::RecordingIsAvailable(bool* available) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    bool isAvailable = false;
+    if (audio_device_->RecordingIsAvailable(isAvailable) == -1) {
+      return -1;
+    }
+    *available = isAvailable;
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::MaxMicrophoneVolume(uint32_t* maxVolume) const {
+    CHECKinitialized_();
+    uint32_t maxVol(0);
+    if (audio_device_->MaxMicrophoneVolume(maxVol) == -1) {
+      return -1;
+    }
+    *maxVolume = maxVol;
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::MinMicrophoneVolume(uint32_t* minVolume) const {
+    CHECKinitialized_();
+    uint32_t minVol(0);
+    if (audio_device_->MinMicrophoneVolume(minVol) == -1) {
+      return -1;
+    }
+    *minVolume = minVol;
+    return 0;
+  }
+
+  int16_t AudioDeviceModuleIOS::PlayoutDevices() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    uint16_t nPlayoutDevices = audio_device_->PlayoutDevices();
+    RTC_LOG(INFO) << "output: " << nPlayoutDevices;
+    return (int16_t)(nPlayoutDevices);
+  }
+
+  int32_t AudioDeviceModuleIOS::SetPlayoutDevice(uint16_t index) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << index << ")";
+    CHECKinitialized_();
+    return audio_device_->SetPlayoutDevice(index);
+  }
+
+  int32_t AudioDeviceModuleIOS::SetPlayoutDevice(WindowsDeviceType device) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    return audio_device_->SetPlayoutDevice(device);
+  }
+
+  int32_t AudioDeviceModuleIOS::PlayoutDeviceName(
+      uint16_t index,
+      char name[kAdmMaxDeviceNameSize],
+      char guid[kAdmMaxGuidSize]) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << index << ", ...)";
+    CHECKinitialized_();
+    if (name == NULL) {
+      return -1;
+    }
+    if (audio_device_->PlayoutDeviceName(index, name, guid) == -1) {
+      return -1;
+    }
+    if (name != NULL) {
+      RTC_LOG(INFO) << "output: name = " << name;
+    }
+    if (guid != NULL) {
+      RTC_LOG(INFO) << "output: guid = " << guid;
+    }
+    return 0;
+  }
+
+  int32_t AudioDeviceModuleIOS::RecordingDeviceName(
+      uint16_t index,
+      char name[kAdmMaxDeviceNameSize],
+      char guid[kAdmMaxGuidSize]) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << index << ", ...)";
+    CHECKinitialized_();
+    if (name == NULL) {
+      return -1;
+    }
+    if (audio_device_->RecordingDeviceName(index, name, guid) == -1) {
+      return -1;
+    }
+    if (name != NULL) {
+      RTC_LOG(INFO) << "output: name = " << name;
+    }
+    if (guid != NULL) {
+      RTC_LOG(INFO) << "output: guid = " << guid;
+    }
+    return 0;
+  }
+
+  int16_t AudioDeviceModuleIOS::RecordingDevices() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    uint16_t nRecordingDevices = audio_device_->RecordingDevices();
+    RTC_LOG(INFO) << "output: " << nRecordingDevices;
+    return (int16_t)nRecordingDevices;
+  }
+
+  int32_t AudioDeviceModuleIOS::SetRecordingDevice(uint16_t index) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << index << ")";
+    CHECKinitialized_();
+    return audio_device_->SetRecordingDevice(index);
+  }
+
+  int32_t AudioDeviceModuleIOS::SetRecordingDevice(WindowsDeviceType device) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    return audio_device_->SetRecordingDevice(device);
+  }
+
+  int32_t AudioDeviceModuleIOS::InitPlayout() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    if (PlayoutIsInitialized()) {
+      return 0;
+    }
+    int32_t result = audio_device_->InitPlayout();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.InitPlayoutSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  int32_t AudioDeviceModuleIOS::InitRecording() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    if (RecordingIsInitialized()) {
+      return 0;
+    }
+    int32_t result = audio_device_->InitRecording();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.InitRecordingSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  bool AudioDeviceModuleIOS::PlayoutIsInitialized() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    return audio_device_->PlayoutIsInitialized();
+  }
+
+  bool AudioDeviceModuleIOS::RecordingIsInitialized() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    return audio_device_->RecordingIsInitialized();
+  }
+
+  int32_t AudioDeviceModuleIOS::StartPlayout() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    if (Playing()) {
+      return 0;
+    }
+    audio_device_buffer_.get()->StartPlayout();
+    int32_t result = audio_device_->StartPlayout();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.StartPlayoutSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  int32_t AudioDeviceModuleIOS::StopPlayout() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    int32_t result = audio_device_->StopPlayout();
+    audio_device_buffer_.get()->StopPlayout();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.StopPlayoutSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  bool AudioDeviceModuleIOS::Playing() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    return audio_device_->Playing();
+  }
+
+  int32_t AudioDeviceModuleIOS::StartRecording() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    if (Recording()) {
+      return 0;
+    }
+    audio_device_buffer_.get()->StartRecording();
+    int32_t result = audio_device_->StartRecording();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.StartRecordingSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  int32_t AudioDeviceModuleIOS::StopRecording() {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized_();
+    int32_t result = audio_device_->StopRecording();
+    audio_device_buffer_.get()->StopRecording();
+    RTC_LOG(INFO) << "output: " << result;
+    RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.StopRecordingSuccess",
+                          static_cast<int>(result == 0));
+    return result;
+  }
+
+  bool AudioDeviceModuleIOS::Recording() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    return audio_device_->Recording();
+  }
+
+  int32_t AudioDeviceModuleIOS::RegisterAudioCallback(
+      AudioTransport* audioCallback) {
+    RTC_LOG(INFO) << __FUNCTION__;
+    return audio_device_buffer_.get()->RegisterAudioCallback(audioCallback);
+  }
+
+  int32_t AudioDeviceModuleIOS::PlayoutDelay(uint16_t* delayMS) const {
+    CHECKinitialized_();
+    uint16_t delay = 0;
+    if (audio_device_->PlayoutDelay(delay) == -1) {
+      RTC_LOG(LERROR) << "failed to retrieve the playout delay";
+      return -1;
+    }
+    *delayMS = delay;
+    return 0;
+  }
+
+  bool AudioDeviceModuleIOS::BuiltInAECIsAvailable() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    bool isAvailable = audio_device_->BuiltInAECIsAvailable();
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return isAvailable;
+  }
+
+  int32_t AudioDeviceModuleIOS::EnableBuiltInAEC(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    int32_t ok = audio_device_->EnableBuiltInAEC(enable);
+    RTC_LOG(INFO) << "output: " << ok;
+    return ok;
+  }
+
+  bool AudioDeviceModuleIOS::BuiltInAGCIsAvailable() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    bool isAvailable = audio_device_->BuiltInAGCIsAvailable();
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return isAvailable;
+  }
+
+  int32_t AudioDeviceModuleIOS::EnableBuiltInAGC(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    int32_t ok = audio_device_->EnableBuiltInAGC(enable);
+    RTC_LOG(INFO) << "output: " << ok;
+    return ok;
+  }
+
+  bool AudioDeviceModuleIOS::BuiltInNSIsAvailable() const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    CHECKinitialized__BOOL();
+    bool isAvailable = audio_device_->BuiltInNSIsAvailable();
+    RTC_LOG(INFO) << "output: " << isAvailable;
+    return isAvailable;
+  }
+
+  int32_t AudioDeviceModuleIOS::EnableBuiltInNS(bool enable) {
+    RTC_LOG(INFO) << __FUNCTION__ << "(" << enable << ")";
+    CHECKinitialized_();
+    int32_t ok = audio_device_->EnableBuiltInNS(enable);
+    RTC_LOG(INFO) << "output: " << ok;
+    return ok;
+  }
+
+#if defined(WEBRTC_IOS)
+  int AudioDeviceModuleIOS::GetPlayoutAudioParameters(
+      AudioParameters* params) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    int r = audio_device_->GetPlayoutAudioParameters(params);
+    RTC_LOG(INFO) << "output: " << r;
+    return r;
+  }
+
+  int AudioDeviceModuleIOS::GetRecordAudioParameters(
+      AudioParameters* params) const {
+    RTC_LOG(INFO) << __FUNCTION__;
+    int r = audio_device_->GetRecordAudioParameters(params);
+    RTC_LOG(INFO) << "output: " << r;
+    return r;
+  }
+#endif  // WEBRTC_IOS
+
+}
+}

sdk/objc/Framework/Native/src/audio/audio_session_observer.h

@@ -0,0 +1,42 @@
+/*
+ *  Copyright 2016 The WebRTC Project Authors. All rights reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#ifndef MODULES_AUDIO_DEVICE_IOS_AUDIO_SESSION_OBSERVER_H_
+#define MODULES_AUDIO_DEVICE_IOS_AUDIO_SESSION_OBSERVER_H_
+
+#include "rtc_base/asyncinvoker.h"
+#include "rtc_base/thread.h"
+
+namespace webrtc {
+
+// Observer interface for listening to AVAudioSession events.
+class AudioSessionObserver {
+ public:
+  // Called when audio session interruption begins.
+  virtual void OnInterruptionBegin() = 0;
+
+  // Called when audio session interruption ends.
+  virtual void OnInterruptionEnd() = 0;
+
+  // Called when audio route changes.
+  virtual void OnValidRouteChange() = 0;
+
+  // Called when the ability to play or record changes.
+  virtual void OnCanPlayOrRecordChange(bool can_play_or_record) = 0;
+
+  virtual void OnChangedOutputVolume() = 0;
+
+ protected:
+  virtual ~AudioSessionObserver() {}
+};
+
+}  // namespace webrtc
+
+#endif  //  MODULES_AUDIO_DEVICE_IOS_AUDIO_SESSION_OBSERVER_H_

sdk/objc/Framework/Native/src/audio/voice_processing_audio_unit.h

@@ -0,0 +1,139 @@
+/*
+ *  Copyright 2016 The WebRTC Project Authors. All rights reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#ifndef MODULES_AUDIO_DEVICE_IOS_VOICE_PROCESSING_AUDIO_UNIT_H_
+#define MODULES_AUDIO_DEVICE_IOS_VOICE_PROCESSING_AUDIO_UNIT_H_
+
+#include <AudioUnit/AudioUnit.h>
+
+namespace webrtc {
+namespace ios_adm {
+
+class VoiceProcessingAudioUnitObserver {
+ public:
+  // Callback function called on a real-time priority I/O thread from the audio
+  // unit. This method is used to signal that recorded audio is available.
+  virtual OSStatus OnDeliverRecordedData(AudioUnitRenderActionFlags* flags,
+                                         const AudioTimeStamp* time_stamp,
+                                         UInt32 bus_number,
+                                         UInt32 num_frames,
+                                         AudioBufferList* io_data) = 0;
+
+  // Callback function called on a real-time priority I/O thread from the audio
+  // unit. This method is used to provide audio samples to the audio unit.
+  virtual OSStatus OnGetPlayoutData(AudioUnitRenderActionFlags* io_action_flags,
+                                    const AudioTimeStamp* time_stamp,
+                                    UInt32 bus_number,
+                                    UInt32 num_frames,
+                                    AudioBufferList* io_data) = 0;
+
+ protected:
+  ~VoiceProcessingAudioUnitObserver() {}
+};
+
+// Convenience class to abstract away the management of a Voice Processing
+// I/O Audio Unit. The Voice Processing I/O unit has the same characteristics
+// as the Remote I/O unit (supports full duplex low-latency audio input and
+// output) and adds AEC for for two-way duplex communication. It also adds AGC,
+// adjustment of voice-processing quality, and muting. Hence, ideal for
+// VoIP applications.
+class VoiceProcessingAudioUnit {
+ public:
+  explicit VoiceProcessingAudioUnit(VoiceProcessingAudioUnitObserver* observer);
+  ~VoiceProcessingAudioUnit();
+
+  // TODO(tkchin): enum for state and state checking.
+  enum State : int32_t {
+    // Init() should be called.
+    kInitRequired,
+    // Audio unit created but not initialized.
+    kUninitialized,
+    // Initialized but not started. Equivalent to stopped.
+    kInitialized,
+    // Initialized and started.
+    kStarted,
+  };
+
+  // Number of bytes per audio sample for 16-bit signed integer representation.
+  static const UInt32 kBytesPerSample;
+
+  // Initializes this class by creating the underlying audio unit instance.
+  // Creates a Voice-Processing I/O unit and configures it for full-duplex
+  // audio. The selected stream format is selected to avoid internal resampling
+  // and to match the 10ms callback rate for WebRTC as well as possible.
+  // Does not intialize the audio unit.
+  bool Init();
+
+  VoiceProcessingAudioUnit::State GetState() const;
+
+  // Initializes the underlying audio unit with the given sample rate.
+  bool Initialize(Float64 sample_rate);
+
+  // Starts the underlying audio unit.
+  bool Start();
+
+  // Stops the underlying audio unit.
+  bool Stop();
+
+  // Uninitializes the underlying audio unit.
+  bool Uninitialize();
+
+  // Calls render on the underlying audio unit.
+  OSStatus Render(AudioUnitRenderActionFlags* flags,
+                  const AudioTimeStamp* time_stamp,
+                  UInt32 output_bus_number,
+                  UInt32 num_frames,
+                  AudioBufferList* io_data);
+
+ private:
+  // The C API used to set callbacks requires static functions. When these are
+  // called, they will invoke the relevant instance method by casting
+  // in_ref_con to VoiceProcessingAudioUnit*.
+  static OSStatus OnGetPlayoutData(void* in_ref_con,
+                                   AudioUnitRenderActionFlags* flags,
+                                   const AudioTimeStamp* time_stamp,
+                                   UInt32 bus_number,
+                                   UInt32 num_frames,
+                                   AudioBufferList* io_data);
+  static OSStatus OnDeliverRecordedData(void* in_ref_con,
+                                        AudioUnitRenderActionFlags* flags,
+                                        const AudioTimeStamp* time_stamp,
+                                        UInt32 bus_number,
+                                        UInt32 num_frames,
+                                        AudioBufferList* io_data);
+
+  // Notifies observer that samples are needed for playback.
+  OSStatus NotifyGetPlayoutData(AudioUnitRenderActionFlags* flags,
+                                const AudioTimeStamp* time_stamp,
+                                UInt32 bus_number,
+                                UInt32 num_frames,
+                                AudioBufferList* io_data);
+  // Notifies observer that recorded samples are available for render.
+  OSStatus NotifyDeliverRecordedData(AudioUnitRenderActionFlags* flags,
+                                     const AudioTimeStamp* time_stamp,
+                                     UInt32 bus_number,
+                                     UInt32 num_frames,
+                                     AudioBufferList* io_data);
+
+  // Returns the predetermined format with a specific sample rate. See
+  // implementation file for details on format.
+  AudioStreamBasicDescription GetFormat(Float64 sample_rate) const;
+
+  // Deletes the underlying audio unit.
+  void DisposeAudioUnit();
+
+  VoiceProcessingAudioUnitObserver* observer_;
+  AudioUnit vpio_unit_;
+  VoiceProcessingAudioUnit::State state_;
+};
+}  // namespace ios_adm
+}  // namespace webrtc
+
+#endif  // MODULES_AUDIO_DEVICE_IOS_VOICE_PROCESSING_AUDIO_UNIT_H_

sdk/objc/Framework/Native/src/audio/voice_processing_audio_unit.mm

@@ -0,0 +1,470 @@
+/*
+ *  Copyright 2016 The WebRTC Project Authors. All rights reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#import "voice_processing_audio_unit.h"
+
+#include "rtc_base/checks.h"
+#include "rtc_base/system/fallthrough.h"
+#include "system_wrappers/include/metrics.h"
+
+#import "WebRTC/RTCLogging.h"
+#import "sdk/objc/Framework/Headers/WebRTC/RTCAudioSessionConfiguration.h"
+
+#if !defined(NDEBUG)
+static void LogStreamDescription(AudioStreamBasicDescription description) {
+  char formatIdString[5];
+  UInt32 formatId = CFSwapInt32HostToBig(description.mFormatID);
+  bcopy(&formatId, formatIdString, 4);
+  formatIdString[4] = '\0';
+  RTCLog(@"AudioStreamBasicDescription: {\n"
+          "  mSampleRate: %.2f\n"
+          "  formatIDString: %s\n"
+          "  mFormatFlags: 0x%X\n"
+          "  mBytesPerPacket: %u\n"
+          "  mFramesPerPacket: %u\n"
+          "  mBytesPerFrame: %u\n"
+          "  mChannelsPerFrame: %u\n"
+          "  mBitsPerChannel: %u\n"
+          "  mReserved: %u\n}",
+         description.mSampleRate, formatIdString,
+         static_cast<unsigned int>(description.mFormatFlags),
+         static_cast<unsigned int>(description.mBytesPerPacket),
+         static_cast<unsigned int>(description.mFramesPerPacket),
+         static_cast<unsigned int>(description.mBytesPerFrame),
+         static_cast<unsigned int>(description.mChannelsPerFrame),
+         static_cast<unsigned int>(description.mBitsPerChannel),
+         static_cast<unsigned int>(description.mReserved));
+}
+#endif
+
+namespace webrtc {
+namespace ios_adm {
+
+// Calls to AudioUnitInitialize() can fail if called back-to-back on different
+// ADM instances. A fall-back solution is to allow multiple sequential calls
+// with as small delay between each. This factor sets the max number of allowed
+// initialization attempts.
+static const int kMaxNumberOfAudioUnitInitializeAttempts = 5;
+// A VP I/O unit's bus 1 connects to input hardware (microphone).
+static const AudioUnitElement kInputBus = 1;
+// A VP I/O unit's bus 0 connects to output hardware (speaker).
+static const AudioUnitElement kOutputBus = 0;
+
+// Returns the automatic gain control (AGC) state on the processed microphone
+// signal. Should be on by default for Voice Processing audio units.
+static OSStatus GetAGCState(AudioUnit audio_unit, UInt32* enabled) {
+  RTC_DCHECK(audio_unit);
+  UInt32 size = sizeof(*enabled);
+  OSStatus result = AudioUnitGetProperty(audio_unit,
+                                         kAUVoiceIOProperty_VoiceProcessingEnableAGC,
+                                         kAudioUnitScope_Global,
+                                         kInputBus,
+                                         enabled,
+                                         &size);
+  RTCLog(@"VPIO unit AGC: %u", static_cast<unsigned int>(*enabled));
+  return result;
+}
+
+VoiceProcessingAudioUnit::VoiceProcessingAudioUnit(
+    VoiceProcessingAudioUnitObserver* observer)
+    : observer_(observer), vpio_unit_(nullptr), state_(kInitRequired) {
+  RTC_DCHECK(observer);
+}
+
+VoiceProcessingAudioUnit::~VoiceProcessingAudioUnit() {
+  DisposeAudioUnit();
+}
+
+const UInt32 VoiceProcessingAudioUnit::kBytesPerSample = 2;
+
+bool VoiceProcessingAudioUnit::Init() {
+  RTC_DCHECK_EQ(state_, kInitRequired);
+
+  // Create an audio component description to identify the Voice Processing
+  // I/O audio unit.
+  AudioComponentDescription vpio_unit_description;
+  vpio_unit_description.componentType = kAudioUnitType_Output;
+  vpio_unit_description.componentSubType = kAudioUnitSubType_VoiceProcessingIO;
+  vpio_unit_description.componentManufacturer = kAudioUnitManufacturer_Apple;
+  vpio_unit_description.componentFlags = 0;
+  vpio_unit_description.componentFlagsMask = 0;
+
+  // Obtain an audio unit instance given the description.
+  AudioComponent found_vpio_unit_ref =
+      AudioComponentFindNext(nullptr, &vpio_unit_description);
+
+  // Create a Voice Processing IO audio unit.
+  OSStatus result = noErr;
+  result = AudioComponentInstanceNew(found_vpio_unit_ref, &vpio_unit_);
+  if (result != noErr) {
+    vpio_unit_ = nullptr;
+    RTCLogError(@"AudioComponentInstanceNew failed. Error=%ld.", (long)result);
+    return false;
+  }
+
+  // Enable input on the input scope of the input element.
+  UInt32 enable_input = 1;
+  result = AudioUnitSetProperty(vpio_unit_, kAudioOutputUnitProperty_EnableIO,
+                                kAudioUnitScope_Input, kInputBus, &enable_input,
+                                sizeof(enable_input));
+  if (result != noErr) {
+    DisposeAudioUnit();
+    RTCLogError(@"Failed to enable input on input scope of input element. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Enable output on the output scope of the output element.
+  UInt32 enable_output = 1;
+  result = AudioUnitSetProperty(vpio_unit_, kAudioOutputUnitProperty_EnableIO,
+                                kAudioUnitScope_Output, kOutputBus,
+                                &enable_output, sizeof(enable_output));
+  if (result != noErr) {
+    DisposeAudioUnit();
+    RTCLogError(@"Failed to enable output on output scope of output element. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Specify the callback function that provides audio samples to the audio
+  // unit.
+  AURenderCallbackStruct render_callback;
+  render_callback.inputProc = OnGetPlayoutData;
+  render_callback.inputProcRefCon = this;
+  result = AudioUnitSetProperty(
+      vpio_unit_, kAudioUnitProperty_SetRenderCallback, kAudioUnitScope_Input,
+      kOutputBus, &render_callback, sizeof(render_callback));
+  if (result != noErr) {
+    DisposeAudioUnit();
+    RTCLogError(@"Failed to specify the render callback on the output bus. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Disable AU buffer allocation for the recorder, we allocate our own.
+  // TODO(henrika): not sure that it actually saves resource to make this call.
+  UInt32 flag = 0;
+  result = AudioUnitSetProperty(
+      vpio_unit_, kAudioUnitProperty_ShouldAllocateBuffer,
+      kAudioUnitScope_Output, kInputBus, &flag, sizeof(flag));
+  if (result != noErr) {
+    DisposeAudioUnit();
+    RTCLogError(@"Failed to disable buffer allocation on the input bus. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Specify the callback to be called by the I/O thread to us when input audio
+  // is available. The recorded samples can then be obtained by calling the
+  // AudioUnitRender() method.
+  AURenderCallbackStruct input_callback;
+  input_callback.inputProc = OnDeliverRecordedData;
+  input_callback.inputProcRefCon = this;
+  result = AudioUnitSetProperty(vpio_unit_,
+                                kAudioOutputUnitProperty_SetInputCallback,
+                                kAudioUnitScope_Global, kInputBus,
+                                &input_callback, sizeof(input_callback));
+  if (result != noErr) {
+    DisposeAudioUnit();
+    RTCLogError(@"Failed to specify the input callback on the input bus. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  state_ = kUninitialized;
+  return true;
+}
+
+VoiceProcessingAudioUnit::State VoiceProcessingAudioUnit::GetState() const {
+  return state_;
+}
+
+bool VoiceProcessingAudioUnit::Initialize(Float64 sample_rate) {
+  RTC_DCHECK_GE(state_, kUninitialized);
+  RTCLog(@"Initializing audio unit with sample rate: %f", sample_rate);
+
+  OSStatus result = noErr;
+  AudioStreamBasicDescription format = GetFormat(sample_rate);
+  UInt32 size = sizeof(format);
+#if !defined(NDEBUG)
+  LogStreamDescription(format);
+#endif
+
+  // Set the format on the output scope of the input element/bus.
+  result =
+      AudioUnitSetProperty(vpio_unit_, kAudioUnitProperty_StreamFormat,
+                           kAudioUnitScope_Output, kInputBus, &format, size);
+  if (result != noErr) {
+    RTCLogError(@"Failed to set format on output scope of input bus. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Set the format on the input scope of the output element/bus.
+  result =
+      AudioUnitSetProperty(vpio_unit_, kAudioUnitProperty_StreamFormat,
+                           kAudioUnitScope_Input, kOutputBus, &format, size);
+  if (result != noErr) {
+    RTCLogError(@"Failed to set format on input scope of output bus. "
+                 "Error=%ld.",
+                (long)result);
+    return false;
+  }
+
+  // Initialize the Voice Processing I/O unit instance.
+  // Calls to AudioUnitInitialize() can fail if called back-to-back on
+  // different ADM instances. The error message in this case is -66635 which is
+  // undocumented. Tests have shown that calling AudioUnitInitialize a second
+  // time, after a short sleep, avoids this issue.
+  // See webrtc:5166 for details.
+  int failed_initalize_attempts = 0;
+  result = AudioUnitInitialize(vpio_unit_);
+  while (result != noErr) {
+    RTCLogError(@"Failed to initialize the Voice Processing I/O unit. "
+                 "Error=%ld.",
+                (long)result);
+    ++failed_initalize_attempts;
+    if (failed_initalize_attempts == kMaxNumberOfAudioUnitInitializeAttempts) {
+      // Max number of initialization attempts exceeded, hence abort.
+      RTCLogError(@"Too many initialization attempts.");
+      return false;
+    }
+    RTCLog(@"Pause 100ms and try audio unit initialization again...");
+    [NSThread sleepForTimeInterval:0.1f];
+    result = AudioUnitInitialize(vpio_unit_);
+  }
+  if (result == noErr) {
+    RTCLog(@"Voice Processing I/O unit is now initialized.");
+  }
+
+  // AGC should be enabled by default for Voice Processing I/O units but it is
+  // checked below and enabled explicitly if needed. This scheme is used
+  // to be absolutely sure that the AGC is enabled since we have seen cases
+  // where only zeros are recorded and a disabled AGC could be one of the
+  // reasons why it happens.
+  int agc_was_enabled_by_default = 0;
+  UInt32 agc_is_enabled = 0;
+  result = GetAGCState(vpio_unit_, &agc_is_enabled);
+  if (result != noErr) {
+    RTCLogError(@"Failed to get AGC state (1st attempt). "
+                 "Error=%ld.",
+                (long)result);
+    // Example of error code: kAudioUnitErr_NoConnection (-10876).
+    // All error codes related to audio units are negative and are therefore
+    // converted into a postive value to match the UMA APIs.
+    RTC_HISTOGRAM_COUNTS_SPARSE_100000(
+        "WebRTC.Audio.GetAGCStateErrorCode1", (-1) * result);
+  } else if (agc_is_enabled) {
+    // Remember that the AGC was enabled by default. Will be used in UMA.
+    agc_was_enabled_by_default = 1;
+  } else {
+    // AGC was initially disabled => try to enable it explicitly.
+    UInt32 enable_agc = 1;
+    result =
+        AudioUnitSetProperty(vpio_unit_,
+                             kAUVoiceIOProperty_VoiceProcessingEnableAGC,
+                             kAudioUnitScope_Global, kInputBus, &enable_agc,
+                             sizeof(enable_agc));
+    if (result != noErr) {
+      RTCLogError(@"Failed to enable the built-in AGC. "
+                   "Error=%ld.",
+                  (long)result);
+      RTC_HISTOGRAM_COUNTS_SPARSE_100000(
+          "WebRTC.Audio.SetAGCStateErrorCode", (-1) * result);
+    }
+    result = GetAGCState(vpio_unit_, &agc_is_enabled);
+    if (result != noErr) {
+      RTCLogError(@"Failed to get AGC state (2nd attempt). "
+                   "Error=%ld.",
+                  (long)result);
+      RTC_HISTOGRAM_COUNTS_SPARSE_100000(
+          "WebRTC.Audio.GetAGCStateErrorCode2", (-1) * result);
+    }
+  }
+
+  // Track if the built-in AGC was enabled by default (as it should) or not.
+  RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.BuiltInAGCWasEnabledByDefault",
+                        agc_was_enabled_by_default);
+  RTCLog(@"WebRTC.Audio.BuiltInAGCWasEnabledByDefault: %d",
+         agc_was_enabled_by_default);
+  // As a final step, add an UMA histogram for tracking the AGC state.
+  // At this stage, the AGC should be enabled, and if it is not, more work is
+  // needed to find out the root cause.
+  RTC_HISTOGRAM_BOOLEAN("WebRTC.Audio.BuiltInAGCIsEnabled", agc_is_enabled);
+  RTCLog(@"WebRTC.Audio.BuiltInAGCIsEnabled: %u",
+         static_cast<unsigned int>(agc_is_enabled));
+
+  state_ = kInitialized;
+  return true;
+}
+
+bool VoiceProcessingAudioUnit::Start() {
+  RTC_DCHECK_GE(state_, kUninitialized);
+  RTCLog(@"Starting audio unit.");
+
+  OSStatus result = AudioOutputUnitStart(vpio_unit_);
+  if (result != noErr) {
+    RTCLogError(@"Failed to start audio unit. Error=%ld", (long)result);
+    return false;
+  } else {
+    RTCLog(@"Started audio unit");
+  }
+  state_ = kStarted;
+  return true;
+}
+
+bool VoiceProcessingAudioUnit::Stop() {
+  RTC_DCHECK_GE(state_, kUninitialized);
+  RTCLog(@"Stopping audio unit.");
+
+  OSStatus result = AudioOutputUnitStop(vpio_unit_);
+  if (result != noErr) {
+    RTCLogError(@"Failed to stop audio unit. Error=%ld", (long)result);
+    return false;
+  } else {
+    RTCLog(@"Stopped audio unit");
+  }
+
+  state_ = kInitialized;
+  return true;
+}
+
+bool VoiceProcessingAudioUnit::Uninitialize() {
+  RTC_DCHECK_GE(state_, kUninitialized);
+  RTCLog(@"Unintializing audio unit.");
+
+  OSStatus result = AudioUnitUninitialize(vpio_unit_);
+  if (result != noErr) {
+    RTCLogError(@"Failed to uninitialize audio unit. Error=%ld", (long)result);
+    return false;
+  } else {
+    RTCLog(@"Uninitialized audio unit.");
+  }
+
+  state_ = kUninitialized;
+  return true;
+}
+
+OSStatus VoiceProcessingAudioUnit::Render(AudioUnitRenderActionFlags* flags,
+                                          const AudioTimeStamp* time_stamp,
+                                          UInt32 output_bus_number,
+                                          UInt32 num_frames,
+                                          AudioBufferList* io_data) {
+  RTC_DCHECK(vpio_unit_) << "Init() not called.";
+
+  OSStatus result = AudioUnitRender(vpio_unit_, flags, time_stamp,
+                                    output_bus_number, num_frames, io_data);
+  if (result != noErr) {
+    RTCLogError(@"Failed to render audio unit. Error=%ld", (long)result);
+  }
+  return result;
+}
+
+OSStatus VoiceProcessingAudioUnit::OnGetPlayoutData(
+    void* in_ref_con,
+    AudioUnitRenderActionFlags* flags,
+    const AudioTimeStamp* time_stamp,
+    UInt32 bus_number,
+    UInt32 num_frames,
+    AudioBufferList* io_data) {
+  VoiceProcessingAudioUnit* audio_unit =
+      static_cast<VoiceProcessingAudioUnit*>(in_ref_con);
+  return audio_unit->NotifyGetPlayoutData(flags, time_stamp, bus_number,
+                                          num_frames, io_data);
+}
+
+OSStatus VoiceProcessingAudioUnit::OnDeliverRecordedData(
+    void* in_ref_con,
+    AudioUnitRenderActionFlags* flags,
+    const AudioTimeStamp* time_stamp,
+    UInt32 bus_number,
+    UInt32 num_frames,
+    AudioBufferList* io_data) {
+  VoiceProcessingAudioUnit* audio_unit =
+      static_cast<VoiceProcessingAudioUnit*>(in_ref_con);
+  return audio_unit->NotifyDeliverRecordedData(flags, time_stamp, bus_number,
+                                               num_frames, io_data);
+}
+
+OSStatus VoiceProcessingAudioUnit::NotifyGetPlayoutData(
+    AudioUnitRenderActionFlags* flags,
+    const AudioTimeStamp* time_stamp,
+    UInt32 bus_number,
+    UInt32 num_frames,
+    AudioBufferList* io_data) {
+  return observer_->OnGetPlayoutData(flags, time_stamp, bus_number, num_frames,
+                                     io_data);
+}
+
+OSStatus VoiceProcessingAudioUnit::NotifyDeliverRecordedData(
+    AudioUnitRenderActionFlags* flags,
+    const AudioTimeStamp* time_stamp,
+    UInt32 bus_number,
+    UInt32 num_frames,
+    AudioBufferList* io_data) {
+  return observer_->OnDeliverRecordedData(flags, time_stamp, bus_number,
+                                          num_frames, io_data);
+}
+
+AudioStreamBasicDescription VoiceProcessingAudioUnit::GetFormat(
+    Float64 sample_rate) const {
+  // Set the application formats for input and output:
+  // - use same format in both directions
+  // - avoid resampling in the I/O unit by using the hardware sample rate
+  // - linear PCM => noncompressed audio data format with one frame per packet
+  // - no need to specify interleaving since only mono is supported
+  AudioStreamBasicDescription format;
+  RTC_DCHECK_EQ(1, kRTCAudioSessionPreferredNumberOfChannels);
+  format.mSampleRate = sample_rate;
+  format.mFormatID = kAudioFormatLinearPCM;
+  format.mFormatFlags =
+      kLinearPCMFormatFlagIsSignedInteger | kLinearPCMFormatFlagIsPacked;
+  format.mBytesPerPacket = kBytesPerSample;
+  format.mFramesPerPacket = 1;  // uncompressed.
+  format.mBytesPerFrame = kBytesPerSample;
+  format.mChannelsPerFrame = kRTCAudioSessionPreferredNumberOfChannels;
+  format.mBitsPerChannel = 8 * kBytesPerSample;
+  return format;
+}
+
+void VoiceProcessingAudioUnit::DisposeAudioUnit() {
+  if (vpio_unit_) {
+    switch (state_) {
+      case kStarted:
+        Stop();
+        // Fall through.
+        RTC_FALLTHROUGH();
+      case kInitialized:
+        Uninitialize();
+        break;
+      case kUninitialized:
+        RTC_FALLTHROUGH();
+      case kInitRequired:
+        break;
+    }
+
+    RTCLog(@"Disposing audio unit.");
+    OSStatus result = AudioComponentInstanceDispose(vpio_unit_);
+    if (result != noErr) {
+      RTCLogError(@"AudioComponentInstanceDispose failed. Error=%ld.",
+                  (long)result);
+    }
+    vpio_unit_ = nullptr;
+  }
+}
+
+}  // namespace ios_adm
+}  // namespace webrtc