media/libaaudio/src/legacy/AudioStreamTrack.cpp - LeafOS-Project/android_frameworks_av - Gitiles

 /*
  * Copyright 2016 The Android Open Source Project
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
  * You may obtain a copy of the License at
  *
  *      http://www.apache.org/licenses/LICENSE-2.0
  *
  * Unless required by applicable law or agreed to in writing, software
  * distributed under the License is distributed on an "AS IS" BASIS,
  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */

 #define LOG_TAG "AudioStreamTrack"
 //#define LOG_NDEBUG 0
 #include <utils/Log.h>

 #include <stdint.h>
 #include <media/AudioTrack.h>

 #include <aaudio/AAudio.h>
 #include <system/audio.h>
 #include "utility/AudioClock.h"
 #include "legacy/AudioStreamLegacy.h"
 #include "legacy/AudioStreamTrack.h"
 #include "utility/FixedBlockReader.h"

 using namespace android;
 using namespace aaudio;

 // Arbitrary and somewhat generous number of bursts.
 #define DEFAULT_BURSTS_PER_BUFFER_CAPACITY     8

 /*
  * Create a stream that uses the AudioTrack.
  */
 AudioStreamTrack::AudioStreamTrack()
     : AudioStreamLegacy()
     , mFixedBlockReader(*this)
 {
 }

 AudioStreamTrack::~AudioStreamTrack()
 {
     const aaudio_stream_state_t state = getState();
     bool bad = !(state == AAUDIO_STREAM_STATE_UNINITIALIZED || state == AAUDIO_STREAM_STATE_CLOSED);
     ALOGE_IF(bad, "stream not closed, in state %d", state);
 }

 aaudio_result_t AudioStreamTrack::open(const AudioStreamBuilder& builder)
 {
     aaudio_result_t result = AAUDIO_OK;

     result = AudioStream::open(builder);
     if (result != OK) {
         return result;
     }

     const aaudio_session_id_t requestedSessionId = builder.getSessionId();
     const audio_session_t sessionId = AAudioConvert_aaudioToAndroidSessionId(requestedSessionId);

     // Try to create an AudioTrack
     // Use stereo if unspecified.
     int32_t samplesPerFrame = (getSamplesPerFrame() == AAUDIO_UNSPECIFIED)
                               ? 2 : getSamplesPerFrame();
     audio_channel_mask_t channelMask = samplesPerFrame <= 2 ?
                             audio_channel_out_mask_from_count(samplesPerFrame) :
                             audio_channel_mask_for_index_assignment_from_count(samplesPerFrame);

     audio_output_flags_t flags;
     aaudio_performance_mode_t perfMode = getPerformanceMode();
     switch(perfMode) {
         case AAUDIO_PERFORMANCE_MODE_LOW_LATENCY:
             // Bypass the normal mixer and go straight to the FAST mixer.
             // If the app asks for a sessionId then it means they want to use effects.
             // So don't use RAW flag.
             flags = (audio_output_flags_t) ((requestedSessionId == AAUDIO_SESSION_ID_NONE)
                     ? (AUDIO_OUTPUT_FLAG_FAST | AUDIO_OUTPUT_FLAG_RAW)
                     : (AUDIO_OUTPUT_FLAG_FAST));
             break;

         case AAUDIO_PERFORMANCE_MODE_POWER_SAVING:
             // This uses a mixer that wakes up less often than the FAST mixer.
             flags = AUDIO_OUTPUT_FLAG_DEEP_BUFFER;
             break;

         case AAUDIO_PERFORMANCE_MODE_NONE:
         default:
             // No flags. Use a normal mixer in front of the FAST mixer.
             flags = AUDIO_OUTPUT_FLAG_NONE;
             break;
     }

     size_t frameCount = (size_t)builder.getBufferCapacity();

     int32_t notificationFrames = 0;

     const audio_format_t format = (getFormat() == AUDIO_FORMAT_DEFAULT)
             ? AUDIO_FORMAT_PCM_FLOAT
             : getFormat();

     // Setup the callback if there is one.
     AudioTrack::callback_t callback = nullptr;
     void *callbackData = nullptr;
     // Note that TRANSFER_SYNC does not allow FAST track
     AudioTrack::transfer_type streamTransferType = AudioTrack::transfer_type::TRANSFER_SYNC;
     if (builder.getDataCallbackProc() != nullptr) {
         streamTransferType = AudioTrack::transfer_type::TRANSFER_CALLBACK;
         callback = getLegacyCallback();
         callbackData = this;

         // If the total buffer size is unspecified then base the size on the burst size.
         if (frameCount == 0
                 && ((flags & AUDIO_OUTPUT_FLAG_FAST) != 0)) {
             // Take advantage of a special trick that allows us to create a buffer
             // that is some multiple of the burst size.
             notificationFrames = 0 - DEFAULT_BURSTS_PER_BUFFER_CAPACITY;
         } else {
             notificationFrames = builder.getFramesPerDataCallback();
         }
     }
     mCallbackBufferSize = builder.getFramesPerDataCallback();

     ALOGD("open(), request notificationFrames = %d, frameCount = %u",
           notificationFrames, (uint)frameCount);

     // Don't call mAudioTrack->setDeviceId() because it will be overwritten by set()!
     audio_port_handle_t selectedDeviceId = (getDeviceId() == AAUDIO_UNSPECIFIED)
                                            ? AUDIO_PORT_HANDLE_NONE
                                            : getDeviceId();

     const audio_content_type_t contentType =
             AAudioConvert_contentTypeToInternal(builder.getContentType());
     const audio_usage_t usage =
             AAudioConvert_usageToInternal(builder.getUsage());
     const audio_flags_mask_t attributesFlags =
         AAudioConvert_allowCapturePolicyToAudioFlagsMask(builder.getAllowedCapturePolicy());

     const audio_attributes_t attributes = {
             .content_type = contentType,
             .usage = usage,
             .source = AUDIO_SOURCE_DEFAULT, // only used for recording
             .flags = attributesFlags,
             .tags = ""
     };

     mAudioTrack = new AudioTrack();
     mAudioTrack->set(
             AUDIO_STREAM_DEFAULT,  // ignored because we pass attributes below
             getSampleRate(),
             format,
             channelMask,
             frameCount,
             flags,
             callback,
             callbackData,
             notificationFrames,
             0,       // DEFAULT sharedBuffer*/,
             false,   // DEFAULT threadCanCallJava
             sessionId,
             streamTransferType,
             NULL,    // DEFAULT audio_offload_info_t
             AUDIO_UID_INVALID, // DEFAULT uid
             -1,      // DEFAULT pid
             &attributes,
             // WARNING - If doNotReconnect set true then audio stops after plugging and unplugging
             // headphones a few times.
             false,   // DEFAULT doNotReconnect,
             1.0f,    // DEFAULT maxRequiredSpeed
             selectedDeviceId
     );

     // Did we get a valid track?
     status_t status = mAudioTrack->initCheck();
     if (status != NO_ERROR) {
         releaseCloseFinal();
         ALOGE("open(), initCheck() returned %d", status);
         return AAudioConvert_androidToAAudioResult(status);
     }

     doSetVolume();

     // Get the actual values from the AudioTrack.
     setSamplesPerFrame(mAudioTrack->channelCount());
     setFormat(mAudioTrack->format());
     setDeviceFormat(mAudioTrack->format());

     int32_t actualSampleRate = mAudioTrack->getSampleRate();
     ALOGW_IF(actualSampleRate != getSampleRate(),
              "open() sampleRate changed from %d to %d",
              getSampleRate(), actualSampleRate);
     setSampleRate(actualSampleRate);

     // We may need to pass the data through a block size adapter to guarantee constant size.
     if (mCallbackBufferSize != AAUDIO_UNSPECIFIED) {
         int callbackSizeBytes = getBytesPerFrame() * mCallbackBufferSize;
         mFixedBlockReader.open(callbackSizeBytes);
         mBlockAdapter = &mFixedBlockReader;
     } else {
         mBlockAdapter = nullptr;
     }

     setState(AAUDIO_STREAM_STATE_OPEN);
     setDeviceId(mAudioTrack->getRoutedDeviceId());

     aaudio_session_id_t actualSessionId =
             (requestedSessionId == AAUDIO_SESSION_ID_NONE)
             ? AAUDIO_SESSION_ID_NONE
             : (aaudio_session_id_t) mAudioTrack->getSessionId();
     setSessionId(actualSessionId);

     mAudioTrack->addAudioDeviceCallback(mDeviceCallback);

     // Update performance mode based on the actual stream flags.
     // For example, if the sample rate is not allowed then you won't get a FAST track.
     audio_output_flags_t actualFlags = mAudioTrack->getFlags();
     aaudio_performance_mode_t actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_NONE;
     // We may not get the RAW flag. But as long as we get the FAST flag we can call it LOW_LATENCY.
     if ((actualFlags & AUDIO_OUTPUT_FLAG_FAST) != 0) {
         actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_LOW_LATENCY;
     } else if ((actualFlags & AUDIO_OUTPUT_FLAG_DEEP_BUFFER) != 0) {
         actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_POWER_SAVING;
     }
     setPerformanceMode(actualPerformanceMode);

     setSharingMode(AAUDIO_SHARING_MODE_SHARED); // EXCLUSIVE mode not supported in legacy

     // Log warning if we did not get what we asked for.
     ALOGW_IF(actualFlags != flags,
              "open() flags changed from 0x%08X to 0x%08X",
              flags, actualFlags);
     ALOGW_IF(actualPerformanceMode != perfMode,
              "open() perfMode changed from %d to %d",
              perfMode, actualPerformanceMode);

     return AAUDIO_OK;
 }

 aaudio_result_t AudioStreamTrack::release_l() {
     if (getState() != AAUDIO_STREAM_STATE_CLOSING) {
         mAudioTrack->removeAudioDeviceCallback(mDeviceCallback);
         // TODO Investigate why clear() causes a hang in test_various.cpp
         // if I call close() from a data callback.
         // But the same thing in AudioRecord is OK!
         // mAudioTrack.clear();
         mFixedBlockReader.close();
         return AudioStream::release_l();
     } else {
         return AAUDIO_OK; // already released
     }
 }

 void AudioStreamTrack::processCallback(int event, void *info) {

     switch (event) {
         case AudioTrack::EVENT_MORE_DATA:
             processCallbackCommon(AAUDIO_CALLBACK_OPERATION_PROCESS_DATA, info);
             break;

             // Stream got rerouted so we disconnect.
         case AudioTrack::EVENT_NEW_IAUDIOTRACK:
             processCallbackCommon(AAUDIO_CALLBACK_OPERATION_DISCONNECTED, info);
             break;

         default:
             break;
     }
     return;
 }

 aaudio_result_t AudioStreamTrack::requestStart() {
     if (mAudioTrack.get() == nullptr) {
         ALOGE("requestStart() no AudioTrack");
         return AAUDIO_ERROR_INVALID_STATE;
     }
     // Get current position so we can detect when the track is playing.
     status_t err = mAudioTrack->getPosition(&mPositionWhenStarting);
     if (err != OK) {
         return AAudioConvert_androidToAAudioResult(err);
     }

     // Enable callback before starting AudioTrack to avoid shutting
     // down because of a race condition.
     mCallbackEnabled.store(true);
     err = mAudioTrack->start();
     if (err != OK) {
         return AAudioConvert_androidToAAudioResult(err);
     } else {
         setState(AAUDIO_STREAM_STATE_STARTING);
     }
     return AAUDIO_OK;
 }

 aaudio_result_t AudioStreamTrack::requestPause() {
     if (mAudioTrack.get() == nullptr) {
         ALOGE("%s() no AudioTrack", __func__);
         return AAUDIO_ERROR_INVALID_STATE;
     }

     setState(AAUDIO_STREAM_STATE_PAUSING);
     mAudioTrack->pause();
     mCallbackEnabled.store(false);
     status_t err = mAudioTrack->getPosition(&mPositionWhenPausing);
     if (err != OK) {
         return AAudioConvert_androidToAAudioResult(err);
     }
     return checkForDisconnectRequest(false);
 }

 aaudio_result_t AudioStreamTrack::requestFlush() {
     if (mAudioTrack.get() == nullptr) {
         ALOGE("%s() no AudioTrack", __func__);
         return AAUDIO_ERROR_INVALID_STATE;
     }

     setState(AAUDIO_STREAM_STATE_FLUSHING);
     incrementFramesRead(getFramesWritten() - getFramesRead());
     mAudioTrack->flush();
     mFramesRead.reset32(); // service reads frames, service position reset on flush
     mTimestampPosition.reset32();
     return AAUDIO_OK;
 }

 aaudio_result_t AudioStreamTrack::requestStop() {
     if (mAudioTrack.get() == nullptr) {
         ALOGE("%s() no AudioTrack", __func__);
         return AAUDIO_ERROR_INVALID_STATE;
     }

     setState(AAUDIO_STREAM_STATE_STOPPING);
     mFramesRead.catchUpTo(getFramesWritten());
     mTimestampPosition.catchUpTo(getFramesWritten());
     mFramesRead.reset32(); // service reads frames, service position reset on stop
     mTimestampPosition.reset32();
     mAudioTrack->stop();
     mCallbackEnabled.store(false);
     return checkForDisconnectRequest(false);;
 }

 aaudio_result_t AudioStreamTrack::updateStateMachine()
 {
     status_t err;
     aaudio_wrapping_frames_t position;
     switch (getState()) {
     // TODO add better state visibility to AudioTrack
     case AAUDIO_STREAM_STATE_STARTING:
         if (mAudioTrack->hasStarted()) {
             setState(AAUDIO_STREAM_STATE_STARTED);
         }
         break;
     case AAUDIO_STREAM_STATE_PAUSING:
         if (mAudioTrack->stopped()) {
             err = mAudioTrack->getPosition(&position);
             if (err != OK) {
                 return AAudioConvert_androidToAAudioResult(err);
             } else if (position == mPositionWhenPausing) {
                 // Has stream really stopped advancing?
                 setState(AAUDIO_STREAM_STATE_PAUSED);
             }
             mPositionWhenPausing = position;
         }
         break;
     case AAUDIO_STREAM_STATE_FLUSHING:
         {
             err = mAudioTrack->getPosition(&position);
             if (err != OK) {
                 return AAudioConvert_androidToAAudioResult(err);
             } else if (position == 0) {
                 // TODO Advance frames read to match written.
                 setState(AAUDIO_STREAM_STATE_FLUSHED);
             }
         }
         break;
     case AAUDIO_STREAM_STATE_STOPPING:
         if (mAudioTrack->stopped()) {
             setState(AAUDIO_STREAM_STATE_STOPPED);
         }
         break;
     default:
         break;
     }
     return AAUDIO_OK;
 }

 aaudio_result_t AudioStreamTrack::write(const void *buffer,
                                       int32_t numFrames,
                                       int64_t timeoutNanoseconds)
 {
     int32_t bytesPerFrame = getBytesPerFrame();
     int32_t numBytes;
     aaudio_result_t result = AAudioConvert_framesToBytes(numFrames, bytesPerFrame, &numBytes);
     if (result != AAUDIO_OK) {
         return result;
     }

     if (getState() == AAUDIO_STREAM_STATE_DISCONNECTED) {
         return AAUDIO_ERROR_DISCONNECTED;
     }

     // TODO add timeout to AudioTrack
     bool blocking = timeoutNanoseconds > 0;
     ssize_t bytesWritten = mAudioTrack->write(buffer, numBytes, blocking);
     if (bytesWritten == WOULD_BLOCK) {
         return 0;
     } else if (bytesWritten < 0) {
         ALOGE("invalid write, returned %d", (int)bytesWritten);
         // in this context, a DEAD_OBJECT is more likely to be a disconnect notification due to
         // AudioTrack invalidation
         if (bytesWritten == DEAD_OBJECT) {
             setState(AAUDIO_STREAM_STATE_DISCONNECTED);
             return AAUDIO_ERROR_DISCONNECTED;
         }
         return AAudioConvert_androidToAAudioResult(bytesWritten);
     }
     int32_t framesWritten = (int32_t)(bytesWritten / bytesPerFrame);
     incrementFramesWritten(framesWritten);

     result = updateStateMachine();
     if (result != AAUDIO_OK) {
         return result;
     }

     return framesWritten;
 }

 aaudio_result_t AudioStreamTrack::setBufferSize(int32_t requestedFrames)
 {
     // Do not ask for less than one burst.
     if (requestedFrames < getFramesPerBurst()) {
         requestedFrames = getFramesPerBurst();
     }
     ssize_t result = mAudioTrack->setBufferSizeInFrames(requestedFrames);
     if (result < 0) {
         return AAudioConvert_androidToAAudioResult(result);
     } else {
         return result;
     }
 }

 int32_t AudioStreamTrack::getBufferSize() const
 {
     return static_cast<int32_t>(mAudioTrack->getBufferSizeInFrames());
 }

 int32_t AudioStreamTrack::getBufferCapacity() const
 {
     return static_cast<int32_t>(mAudioTrack->frameCount());
 }

 int32_t AudioStreamTrack::getXRunCount() const
 {
     return static_cast<int32_t>(mAudioTrack->getUnderrunCount());
 }

 int32_t AudioStreamTrack::getFramesPerBurst() const
 {
     return static_cast<int32_t>(mAudioTrack->getNotificationPeriodInFrames());
 }

 int64_t AudioStreamTrack::getFramesRead() {
     aaudio_wrapping_frames_t position;
     status_t result;
     switch (getState()) {
     case AAUDIO_STREAM_STATE_STARTING:
     case AAUDIO_STREAM_STATE_STARTED:
     case AAUDIO_STREAM_STATE_STOPPING:
     case AAUDIO_STREAM_STATE_PAUSING:
     case AAUDIO_STREAM_STATE_PAUSED:
         result = mAudioTrack->getPosition(&position);
         if (result == OK) {
             mFramesRead.update32(position);
         }
         break;
     default:
         break;
     }
     return AudioStreamLegacy::getFramesRead();
 }

 aaudio_result_t AudioStreamTrack::getTimestamp(clockid_t clockId,
                                      int64_t *framePosition,
                                      int64_t *timeNanoseconds) {
     ExtendedTimestamp extendedTimestamp;
     status_t status = mAudioTrack->getTimestamp(&extendedTimestamp);
     if (status == WOULD_BLOCK) {
         return AAUDIO_ERROR_INVALID_STATE;
     } if (status != NO_ERROR) {
         return AAudioConvert_androidToAAudioResult(status);
     }
     int64_t position = 0;
     int64_t nanoseconds = 0;
     aaudio_result_t result = getBestTimestamp(clockId, &position,
                                               &nanoseconds, &extendedTimestamp);
     if (result == AAUDIO_OK) {
         if (position < getFramesWritten()) {
             *framePosition = position;
             *timeNanoseconds = nanoseconds;
             return result;
         } else {
             return AAUDIO_ERROR_INVALID_STATE; // TODO review, documented but not consistent
         }
     }
     return result;
 }

 status_t AudioStreamTrack::doSetVolume() {
     status_t status = NO_INIT;
     if (mAudioTrack.get() != nullptr) {
         float volume = getDuckAndMuteVolume();
         mAudioTrack->setVolume(volume, volume);
         status = NO_ERROR;
     }
     return status;
 }

 #if AAUDIO_USE_VOLUME_SHAPER

 using namespace android::media::VolumeShaper;

 binder::Status AudioStreamTrack::applyVolumeShaper(
         const VolumeShaper::Configuration& configuration,
         const VolumeShaper::Operation& operation) {

     sp<VolumeShaper::Configuration> spConfiguration = new VolumeShaper::Configuration(configuration);
     sp<VolumeShaper::Operation> spOperation = new VolumeShaper::Operation(operation);

     if (mAudioTrack.get() != nullptr) {
         ALOGD("applyVolumeShaper() from IPlayer");
         binder::Status status = mAudioTrack->applyVolumeShaper(spConfiguration, spOperation);
         if (status < 0) { // a non-negative value is the volume shaper id.
             ALOGE("applyVolumeShaper() failed with status %d", status);
         }
         return binder::Status::fromStatusT(status);
     } else {
         ALOGD("applyVolumeShaper()"
                       " no AudioTrack for volume control from IPlayer");
         return binder::Status::ok();
     }
 }
 #endif
	/*
	* Copyright 2016 The Android Open Source Project
	*
	* Licensed under the Apache License, Version 2.0 (the "License");
	* you may not use this file except in compliance with the License.
	* You may obtain a copy of the License at
	*
	* http://www.apache.org/licenses/LICENSE-2.0
	*
	* Unless required by applicable law or agreed to in writing, software
	* distributed under the License is distributed on an "AS IS" BASIS,
	* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
	* See the License for the specific language governing permissions and
	* limitations under the License.
	*/

	#define LOG_TAG "AudioStreamTrack"
	//#define LOG_NDEBUG 0
	#include <utils/Log.h>

	#include <stdint.h>
	#include <media/AudioTrack.h>

	#include <aaudio/AAudio.h>
	#include <system/audio.h>
	#include "utility/AudioClock.h"
	#include "legacy/AudioStreamLegacy.h"
	#include "legacy/AudioStreamTrack.h"
	#include "utility/FixedBlockReader.h"

	using namespace android;
	using namespace aaudio;

	// Arbitrary and somewhat generous number of bursts.
	#define DEFAULT_BURSTS_PER_BUFFER_CAPACITY 8

	/*
	* Create a stream that uses the AudioTrack.
	*/
	AudioStreamTrack::AudioStreamTrack()
	: AudioStreamLegacy()
	, mFixedBlockReader(*this)
	{
	}

	AudioStreamTrack::~AudioStreamTrack()
	{
	const aaudio_stream_state_t state = getState();
	bool bad = !(state == AAUDIO_STREAM_STATE_UNINITIALIZED \|\| state == AAUDIO_STREAM_STATE_CLOSED);
	ALOGE_IF(bad, "stream not closed, in state %d", state);
	}

	aaudio_result_t AudioStreamTrack::open(const AudioStreamBuilder& builder)
	{
	aaudio_result_t result = AAUDIO_OK;

	result = AudioStream::open(builder);
	if (result != OK) {
	return result;
	}

	const aaudio_session_id_t requestedSessionId = builder.getSessionId();
	const audio_session_t sessionId = AAudioConvert_aaudioToAndroidSessionId(requestedSessionId);

	// Try to create an AudioTrack
	// Use stereo if unspecified.
	int32_t samplesPerFrame = (getSamplesPerFrame() == AAUDIO_UNSPECIFIED)
	? 2 : getSamplesPerFrame();
	audio_channel_mask_t channelMask = samplesPerFrame <= 2 ?
	audio_channel_out_mask_from_count(samplesPerFrame) :
	audio_channel_mask_for_index_assignment_from_count(samplesPerFrame);

	audio_output_flags_t flags;
	aaudio_performance_mode_t perfMode = getPerformanceMode();
	switch(perfMode) {
	case AAUDIO_PERFORMANCE_MODE_LOW_LATENCY:
	// Bypass the normal mixer and go straight to the FAST mixer.
	// If the app asks for a sessionId then it means they want to use effects.
	// So don't use RAW flag.
	flags = (audio_output_flags_t) ((requestedSessionId == AAUDIO_SESSION_ID_NONE)
	? (AUDIO_OUTPUT_FLAG_FAST \| AUDIO_OUTPUT_FLAG_RAW)
	: (AUDIO_OUTPUT_FLAG_FAST));
	break;

	case AAUDIO_PERFORMANCE_MODE_POWER_SAVING:
	// This uses a mixer that wakes up less often than the FAST mixer.
	flags = AUDIO_OUTPUT_FLAG_DEEP_BUFFER;
	break;

	case AAUDIO_PERFORMANCE_MODE_NONE:
	default:
	// No flags. Use a normal mixer in front of the FAST mixer.
	flags = AUDIO_OUTPUT_FLAG_NONE;
	break;
	}

	size_t frameCount = (size_t)builder.getBufferCapacity();

	int32_t notificationFrames = 0;

	const audio_format_t format = (getFormat() == AUDIO_FORMAT_DEFAULT)
	? AUDIO_FORMAT_PCM_FLOAT
	: getFormat();

	// Setup the callback if there is one.
	AudioTrack::callback_t callback = nullptr;
	void *callbackData = nullptr;
	// Note that TRANSFER_SYNC does not allow FAST track
	AudioTrack::transfer_type streamTransferType = AudioTrack::transfer_type::TRANSFER_SYNC;
	if (builder.getDataCallbackProc() != nullptr) {
	streamTransferType = AudioTrack::transfer_type::TRANSFER_CALLBACK;
	callback = getLegacyCallback();
	callbackData = this;

	// If the total buffer size is unspecified then base the size on the burst size.
	if (frameCount == 0
	&& ((flags & AUDIO_OUTPUT_FLAG_FAST) != 0)) {
	// Take advantage of a special trick that allows us to create a buffer
	// that is some multiple of the burst size.
	notificationFrames = 0 - DEFAULT_BURSTS_PER_BUFFER_CAPACITY;
	} else {
	notificationFrames = builder.getFramesPerDataCallback();
	}
	}
	mCallbackBufferSize = builder.getFramesPerDataCallback();

	ALOGD("open(), request notificationFrames = %d, frameCount = %u",
	notificationFrames, (uint)frameCount);

	// Don't call mAudioTrack->setDeviceId() because it will be overwritten by set()!
	audio_port_handle_t selectedDeviceId = (getDeviceId() == AAUDIO_UNSPECIFIED)
	? AUDIO_PORT_HANDLE_NONE
	: getDeviceId();

	const audio_content_type_t contentType =
	AAudioConvert_contentTypeToInternal(builder.getContentType());
	const audio_usage_t usage =
	AAudioConvert_usageToInternal(builder.getUsage());
	const audio_flags_mask_t attributesFlags =
	AAudioConvert_allowCapturePolicyToAudioFlagsMask(builder.getAllowedCapturePolicy());

	const audio_attributes_t attributes = {
	.content_type = contentType,
	.usage = usage,
	.source = AUDIO_SOURCE_DEFAULT, // only used for recording
	.flags = attributesFlags,
	.tags = ""
	};

	mAudioTrack = new AudioTrack();
	mAudioTrack->set(
	AUDIO_STREAM_DEFAULT, // ignored because we pass attributes below
	getSampleRate(),
	format,
	channelMask,
	frameCount,
	flags,
	callback,
	callbackData,
	notificationFrames,
	0, // DEFAULT sharedBuffer*/,
	false, // DEFAULT threadCanCallJava
	sessionId,
	streamTransferType,
	NULL, // DEFAULT audio_offload_info_t
	AUDIO_UID_INVALID, // DEFAULT uid
	-1, // DEFAULT pid
	&attributes,
	// WARNING - If doNotReconnect set true then audio stops after plugging and unplugging
	// headphones a few times.
	false, // DEFAULT doNotReconnect,
	1.0f, // DEFAULT maxRequiredSpeed
	selectedDeviceId
	);

	// Did we get a valid track?
	status_t status = mAudioTrack->initCheck();
	if (status != NO_ERROR) {
	releaseCloseFinal();
	ALOGE("open(), initCheck() returned %d", status);
	return AAudioConvert_androidToAAudioResult(status);
	}

	doSetVolume();

	// Get the actual values from the AudioTrack.
	setSamplesPerFrame(mAudioTrack->channelCount());
	setFormat(mAudioTrack->format());
	setDeviceFormat(mAudioTrack->format());

	int32_t actualSampleRate = mAudioTrack->getSampleRate();
	ALOGW_IF(actualSampleRate != getSampleRate(),
	"open() sampleRate changed from %d to %d",
	getSampleRate(), actualSampleRate);
	setSampleRate(actualSampleRate);

	// We may need to pass the data through a block size adapter to guarantee constant size.
	if (mCallbackBufferSize != AAUDIO_UNSPECIFIED) {
	int callbackSizeBytes = getBytesPerFrame() * mCallbackBufferSize;
	mFixedBlockReader.open(callbackSizeBytes);
	mBlockAdapter = &mFixedBlockReader;
	} else {
	mBlockAdapter = nullptr;
	}

	setState(AAUDIO_STREAM_STATE_OPEN);
	setDeviceId(mAudioTrack->getRoutedDeviceId());

	aaudio_session_id_t actualSessionId =
	(requestedSessionId == AAUDIO_SESSION_ID_NONE)
	? AAUDIO_SESSION_ID_NONE
	: (aaudio_session_id_t) mAudioTrack->getSessionId();
	setSessionId(actualSessionId);

	mAudioTrack->addAudioDeviceCallback(mDeviceCallback);

	// Update performance mode based on the actual stream flags.
	// For example, if the sample rate is not allowed then you won't get a FAST track.
	audio_output_flags_t actualFlags = mAudioTrack->getFlags();
	aaudio_performance_mode_t actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_NONE;
	// We may not get the RAW flag. But as long as we get the FAST flag we can call it LOW_LATENCY.
	if ((actualFlags & AUDIO_OUTPUT_FLAG_FAST) != 0) {
	actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_LOW_LATENCY;
	} else if ((actualFlags & AUDIO_OUTPUT_FLAG_DEEP_BUFFER) != 0) {
	actualPerformanceMode = AAUDIO_PERFORMANCE_MODE_POWER_SAVING;
	}
	setPerformanceMode(actualPerformanceMode);

	setSharingMode(AAUDIO_SHARING_MODE_SHARED); // EXCLUSIVE mode not supported in legacy

	// Log warning if we did not get what we asked for.
	ALOGW_IF(actualFlags != flags,
	"open() flags changed from 0x%08X to 0x%08X",
	flags, actualFlags);
	ALOGW_IF(actualPerformanceMode != perfMode,
	"open() perfMode changed from %d to %d",
	perfMode, actualPerformanceMode);

	return AAUDIO_OK;
	}

	aaudio_result_t AudioStreamTrack::release_l() {
	if (getState() != AAUDIO_STREAM_STATE_CLOSING) {
	mAudioTrack->removeAudioDeviceCallback(mDeviceCallback);
	// TODO Investigate why clear() causes a hang in test_various.cpp
	// if I call close() from a data callback.
	// But the same thing in AudioRecord is OK!
	// mAudioTrack.clear();
	mFixedBlockReader.close();
	return AudioStream::release_l();
	} else {
	return AAUDIO_OK; // already released
	}
	}

	void AudioStreamTrack::processCallback(int event, void *info) {

	switch (event) {
	case AudioTrack::EVENT_MORE_DATA:
	processCallbackCommon(AAUDIO_CALLBACK_OPERATION_PROCESS_DATA, info);
	break;

	// Stream got rerouted so we disconnect.
	case AudioTrack::EVENT_NEW_IAUDIOTRACK:
	processCallbackCommon(AAUDIO_CALLBACK_OPERATION_DISCONNECTED, info);
	break;

	default:
	break;
	}
	return;
	}

	aaudio_result_t AudioStreamTrack::requestStart() {
	if (mAudioTrack.get() == nullptr) {
	ALOGE("requestStart() no AudioTrack");
	return AAUDIO_ERROR_INVALID_STATE;
	}
	// Get current position so we can detect when the track is playing.
	status_t err = mAudioTrack->getPosition(&mPositionWhenStarting);
	if (err != OK) {
	return AAudioConvert_androidToAAudioResult(err);
	}

	// Enable callback before starting AudioTrack to avoid shutting
	// down because of a race condition.
	mCallbackEnabled.store(true);
	err = mAudioTrack->start();
	if (err != OK) {
	return AAudioConvert_androidToAAudioResult(err);
	} else {
	setState(AAUDIO_STREAM_STATE_STARTING);
	}
	return AAUDIO_OK;
	}

	aaudio_result_t AudioStreamTrack::requestPause() {
	if (mAudioTrack.get() == nullptr) {
	ALOGE("%s() no AudioTrack", __func__);
	return AAUDIO_ERROR_INVALID_STATE;
	}

	setState(AAUDIO_STREAM_STATE_PAUSING);
	mAudioTrack->pause();
	mCallbackEnabled.store(false);
	status_t err = mAudioTrack->getPosition(&mPositionWhenPausing);
	if (err != OK) {
	return AAudioConvert_androidToAAudioResult(err);
	}
	return checkForDisconnectRequest(false);
	}

	aaudio_result_t AudioStreamTrack::requestFlush() {
	if (mAudioTrack.get() == nullptr) {
	ALOGE("%s() no AudioTrack", __func__);
	return AAUDIO_ERROR_INVALID_STATE;
	}

	setState(AAUDIO_STREAM_STATE_FLUSHING);
	incrementFramesRead(getFramesWritten() - getFramesRead());
	mAudioTrack->flush();
	mFramesRead.reset32(); // service reads frames, service position reset on flush
	mTimestampPosition.reset32();
	return AAUDIO_OK;
	}

	aaudio_result_t AudioStreamTrack::requestStop() {
	if (mAudioTrack.get() == nullptr) {
	ALOGE("%s() no AudioTrack", __func__);
	return AAUDIO_ERROR_INVALID_STATE;
	}

	setState(AAUDIO_STREAM_STATE_STOPPING);
	mFramesRead.catchUpTo(getFramesWritten());
	mTimestampPosition.catchUpTo(getFramesWritten());
	mFramesRead.reset32(); // service reads frames, service position reset on stop
	mTimestampPosition.reset32();
	mAudioTrack->stop();
	mCallbackEnabled.store(false);
	return checkForDisconnectRequest(false);;
	}

	aaudio_result_t AudioStreamTrack::updateStateMachine()
	{
	status_t err;
	aaudio_wrapping_frames_t position;
	switch (getState()) {
	// TODO add better state visibility to AudioTrack
	case AAUDIO_STREAM_STATE_STARTING:
	if (mAudioTrack->hasStarted()) {
	setState(AAUDIO_STREAM_STATE_STARTED);
	}
	break;
	case AAUDIO_STREAM_STATE_PAUSING:
	if (mAudioTrack->stopped()) {
	err = mAudioTrack->getPosition(&position);
	if (err != OK) {
	return AAudioConvert_androidToAAudioResult(err);
	} else if (position == mPositionWhenPausing) {
	// Has stream really stopped advancing?
	setState(AAUDIO_STREAM_STATE_PAUSED);
	}
	mPositionWhenPausing = position;
	}
	break;
	case AAUDIO_STREAM_STATE_FLUSHING:
	{
	err = mAudioTrack->getPosition(&position);
	if (err != OK) {
	return AAudioConvert_androidToAAudioResult(err);
	} else if (position == 0) {
	// TODO Advance frames read to match written.
	setState(AAUDIO_STREAM_STATE_FLUSHED);
	}
	}
	break;
	case AAUDIO_STREAM_STATE_STOPPING:
	if (mAudioTrack->stopped()) {
	setState(AAUDIO_STREAM_STATE_STOPPED);
	}
	break;
	default:
	break;
	}
	return AAUDIO_OK;
	}

	aaudio_result_t AudioStreamTrack::write(const void *buffer,
	int32_t numFrames,
	int64_t timeoutNanoseconds)
	{
	int32_t bytesPerFrame = getBytesPerFrame();
	int32_t numBytes;
	aaudio_result_t result = AAudioConvert_framesToBytes(numFrames, bytesPerFrame, &numBytes);
	if (result != AAUDIO_OK) {
	return result;
	}

	if (getState() == AAUDIO_STREAM_STATE_DISCONNECTED) {
	return AAUDIO_ERROR_DISCONNECTED;
	}

	// TODO add timeout to AudioTrack
	bool blocking = timeoutNanoseconds > 0;
	ssize_t bytesWritten = mAudioTrack->write(buffer, numBytes, blocking);
	if (bytesWritten == WOULD_BLOCK) {
	return 0;
	} else if (bytesWritten < 0) {
	ALOGE("invalid write, returned %d", (int)bytesWritten);
	// in this context, a DEAD_OBJECT is more likely to be a disconnect notification due to
	// AudioTrack invalidation
	if (bytesWritten == DEAD_OBJECT) {
	setState(AAUDIO_STREAM_STATE_DISCONNECTED);
	return AAUDIO_ERROR_DISCONNECTED;
	}
	return AAudioConvert_androidToAAudioResult(bytesWritten);
	}
	int32_t framesWritten = (int32_t)(bytesWritten / bytesPerFrame);
	incrementFramesWritten(framesWritten);

	result = updateStateMachine();
	if (result != AAUDIO_OK) {
	return result;
	}

	return framesWritten;
	}

	aaudio_result_t AudioStreamTrack::setBufferSize(int32_t requestedFrames)
	{
	// Do not ask for less than one burst.
	if (requestedFrames < getFramesPerBurst()) {
	requestedFrames = getFramesPerBurst();
	}
	ssize_t result = mAudioTrack->setBufferSizeInFrames(requestedFrames);
	if (result < 0) {
	return AAudioConvert_androidToAAudioResult(result);
	} else {
	return result;
	}
	}

	int32_t AudioStreamTrack::getBufferSize() const
	{
	return static_cast<int32_t>(mAudioTrack->getBufferSizeInFrames());
	}

	int32_t AudioStreamTrack::getBufferCapacity() const
	{
	return static_cast<int32_t>(mAudioTrack->frameCount());
	}

	int32_t AudioStreamTrack::getXRunCount() const
	{
	return static_cast<int32_t>(mAudioTrack->getUnderrunCount());
	}

	int32_t AudioStreamTrack::getFramesPerBurst() const
	{
	return static_cast<int32_t>(mAudioTrack->getNotificationPeriodInFrames());
	}

	int64_t AudioStreamTrack::getFramesRead() {
	aaudio_wrapping_frames_t position;
	status_t result;
	switch (getState()) {
	case AAUDIO_STREAM_STATE_STARTING:
	case AAUDIO_STREAM_STATE_STARTED:
	case AAUDIO_STREAM_STATE_STOPPING:
	case AAUDIO_STREAM_STATE_PAUSING:
	case AAUDIO_STREAM_STATE_PAUSED:
	result = mAudioTrack->getPosition(&position);
	if (result == OK) {
	mFramesRead.update32(position);
	}
	break;
	default:
	break;
	}
	return AudioStreamLegacy::getFramesRead();
	}

	aaudio_result_t AudioStreamTrack::getTimestamp(clockid_t clockId,
	int64_t *framePosition,
	int64_t *timeNanoseconds) {
	ExtendedTimestamp extendedTimestamp;
	status_t status = mAudioTrack->getTimestamp(&extendedTimestamp);
	if (status == WOULD_BLOCK) {
	return AAUDIO_ERROR_INVALID_STATE;
	} if (status != NO_ERROR) {
	return AAudioConvert_androidToAAudioResult(status);
	}
	int64_t position = 0;
	int64_t nanoseconds = 0;
	aaudio_result_t result = getBestTimestamp(clockId, &position,
	&nanoseconds, &extendedTimestamp);
	if (result == AAUDIO_OK) {
	if (position < getFramesWritten()) {
	*framePosition = position;
	*timeNanoseconds = nanoseconds;
	return result;
	} else {
	return AAUDIO_ERROR_INVALID_STATE; // TODO review, documented but not consistent
	}
	}
	return result;
	}

	status_t AudioStreamTrack::doSetVolume() {
	status_t status = NO_INIT;
	if (mAudioTrack.get() != nullptr) {
	float volume = getDuckAndMuteVolume();
	mAudioTrack->setVolume(volume, volume);
	status = NO_ERROR;
	}
	return status;
	}

	#if AAUDIO_USE_VOLUME_SHAPER

	using namespace android::media::VolumeShaper;

	binder::Status AudioStreamTrack::applyVolumeShaper(
	const VolumeShaper::Configuration& configuration,
	const VolumeShaper::Operation& operation) {

	sp<VolumeShaper::Configuration> spConfiguration = new VolumeShaper::Configuration(configuration);
	sp<VolumeShaper::Operation> spOperation = new VolumeShaper::Operation(operation);

	if (mAudioTrack.get() != nullptr) {
	ALOGD("applyVolumeShaper() from IPlayer");
	binder::Status status = mAudioTrack->applyVolumeShaper(spConfiguration, spOperation);
	if (status < 0) { // a non-negative value is the volume shaper id.
	ALOGE("applyVolumeShaper() failed with status %d", status);
	}
	return binder::Status::fromStatusT(status);
	} else {
	ALOGD("applyVolumeShaper()"
	" no AudioTrack for volume control from IPlayer");
	return binder::Status::ok();
	}
	}
	#endif