android+WM9714(AC97)調試

s3c6410平台，codec為WM9714(驅動可使用WM9713的),使用ALSA，android版本1.5. MID上移植android以來一直沒有聲音。

首先确認AC97的硬體連接配接，AC-LINK通訊是否正常，這裡使用到wince中的init寄存器組，在

sound/soc/codecs/wm9713.c中替換static const u16 wm9713_reg[]相應的寄存器，以适配硬體

配置。

做完這部分工作後，應該啟動時就會有噗噗聲，如果能聽到，說明硬體連接配接正常，軟體讀寫CODEC正常.

在網上查了很多資料，很多說要配置/system/etc/asound.conf，可我也不知道怎麼去配，找了一堆的資料學習，然後對照wm9713.c中

的參數和别人的asound.conf，終于弄了一份上去，不過似乎不起說明作用，用alsa_aplay z.wav還是提示錯誤"Unable to install

hw params:"，以下函數出錯：

err = snd_pcm_hw_params(handle, params);

也就是說真正要和硬體互動的時候錯了。。終端裡用logcat指令檢視log資訊，發現還是一直列印消息：

I/ServiceManager( 1185): Waiting for sevice media.audio_flinger...

W/AudioSystem( 1185): AudioFlinger not published, waiting...

I/ServiceManager( 1185): Waiting for sevice media.audio_flinger...

W/AudioSystem( 1185): AudioFlinger not published, waiting...

I/ServiceManager( 1185): Waiting for sevice media.audio_flinger...

這說明audio服務還是沒有正常啟動起來！于是查android源代碼，在frameworks/base/libs/audioflinger/AudioFlinger.cpp中，

void AudioFlinger::instantiate() {

LOGD("*******++AudioFlinger::instantiate()/n");

defaultServiceManager()->addService(

String16("media.audio_flinger"), new AudioFlinger());

LOGD("*******--AudioFlinger::instantiate()/n");

}

有調用此函數将自己添加到服務管理器，但是不知道什麼原因，系統的服務管理器并沒有

得到media.audio_flinger的管理權，是以就會出現上面的一直等待的情況。

跟進new AudioFlinger()

先建立一個audio硬體對象 mAudioHardware = AudioHardwareInterface::create();

AudioHardwareInterface* AudioHardwareInterface::create()

{

AudioHardwareInterface* hw = 0;

char value[PROPERTY_VALUE_MAX];

#ifdef GENERIC_AUDIO

hw = new AudioHardwareGeneric();

#else

// if running in emulation - use the emulator driver

if (property_get("ro.kernel.qemu", value, 0)) {

LOGD("Running in emulation - using generic audio driver");

hw = new AudioHardwareGeneric();

}

else {

LOGV("Creating Vendor Specific AudioHardware");

hw = createAudioHardware();

}

#endif

if (hw->initCheck() != NO_ERROR) {

LOGW("Using stubbed audio hardware. No sound will be produced.");

delete hw;

hw = new AudioHardwareStub();

}

#ifdef DUMP_FLINGER_OUT

// This code adds a record of buffers in a file to write calls made by AudioFlinger.

// It replaces the current AudioHardwareInterface object by an intermediate one which

// will record buffers in a file (after sending them to hardware) for testing purpose.

// This feature is enabled by defining symbol DUMP_FLINGER_OUT.

// The output file is FLINGER_DUMP_NAME. Pause are not recorded in the file.

hw = new AudioDumpInterface(hw); // replace interface

#endif

return hw;

}

這裡hw = createAudioHardware();才是真正的初始化主體：

android::AudioHardwareInterface *createAudioHardware(void) {

return new android::AudioHardwareALSA();

}

也就是AudioHardwareInterface::create()傳回的是實際的硬體初始化對象AudioHardwareALSA的指針。

接着跟進：

AudioHardwareALSA::AudioHardwareALSA() :

mOutput(0),

mInput(0),

mIPC(0) // sangsu fix : for IPC

{

snd_lib_error_set_handler(&ALSAErrorHandler);

mMixer = new ALSAMixer; //建立混合器對象

mIPC = new AudioHardwareIPC; // sangsu fix : IPC init

}

再看mIPC = new AudioHardwareIPC;

AudioHardwareIPC::AudioHardwareIPC() :

mClient(NULL)

{

LOGD("### %s", __func__);

int err = 0;

mClient = OpenClient_RILD();

if (mClient == NULL){

LOGE("[*] OpenClient_RILD() error/n");

err = 1;

}

if (RegisterRequestCompleteHandler(mClient, RIL_REQUEST_OEM_HOOK_RAW,

onRawReqComplete) != RIL_CLIENT_ERR_SUCCESS){

LOGE("[*] RegisterRequestCompleteHandler() error/n");

err = 1;

}

if (RegisterUnsolicitedHandler(mClient, 11004, onUnsol) !=

RIL_CLIENT_ERR_SUCCESS){

LOGE("[*] RegisterUnsolicitedHandler() error/n");

err = 1;

}

if (Connect_RILD(mClient) != RIL_CLIENT_ERR_SUCCESS){

LOGE("[*] Connect_RILD() error/n");

err = 1;

}

if (!err) LOGD("Success Initializing IPC");

else LOGE("Failed Initializing IPC");

}

在logcat裡也确實有列印出這些出錯資訊，說明RIL服務沒有運作。

再次回到AudioFlinger::AudioFlinger()，建立音頻接口對象後，接着打開OutputStream

AudioStreamOut *hwOutput = mAudioHardware->openOutputStream(AudioSystem::PCM_16_BIT, 0, 0, &status);

實際調用函數：

AudioStreamOut *

AudioHardwareALSA::openOutputStream(int format,

int channelCount,

uint32_t sampleRate,

status_t *status)

{

AutoMutex lock(mLock);

LOGD("++AudioHardwareALSA::openOutputStream()/n");

// only one output stream allowed

if (mOutput) {

*status = ALREADY_EXISTS;

return 0;

}

AudioStreamOutALSA *out = new AudioStreamOutALSA(this);

*status = out->set(format, channelCount, sampleRate);

if (*status == NO_ERROR) {

mOutput = out;

// Some information is expected to be available immediately after

// the device is open.

uint32_t routes = mRoutes[mMode];

mOutput->setDevice(mMode, routes);

}

else {

delete out;

}

LOGD("--AudioHardwareALSA::openOutputStream()/n");

return mOutput;

}

進入new AudioStreamOutALSA(this)

AudioStreamOutALSA::AudioStreamOutALSA(AudioHardwareALSA *parent) :

mParent(parent),

mPowerLock(false)

{

static StreamDefaults _defaults = {

devicePrefix : "AndroidPlayback",

direction : SND_PCM_STREAM_PLAYBACK,

format : SND_PCM_FORMAT_S16_LE, // AudioSystem::PCM_16_BIT

channels : 2,

sampleRate : DEFAULT_SAMPLE_RATE,

latency : 250000, // Desired Delay in usec

// sangsu fix : optimizing output buffer size

bufferSize : 2048, // Desired Number of samples

};

setStreamDefaults(&_defaults);

}

這裡就是設定的預設OutStream參數了。。再看out->set(format, channelCount, sampleRate);

status_t set(int format = 0,

int channelCount = 0,

uint32_t sampleRate = 0) {

return ALSAStreamOps::set(format, channelCount, sampleRate);

}

status_t ALSAStreamOps::set(int format,

int channels,

uint32_t rate)

{

if (channels != 0)

mDefaults->channels = channels;

if (rate != 0)

mDefaults->sampleRate = rate;

switch(format) {

// format == 0

case AudioSystem::DEFAULT:

break;

case AudioSystem::PCM_16_BIT:

mDefaults->format = SND_PCM_FORMAT_S16_LE;

break;

case AudioSystem::PCM_8_BIT:

mDefaults->format = SND_PCM_FORMAT_S8;

break;

default:

LOGE("Unknown PCM format %i. Forcing default", format);

break;

}

return NO_ERROR;

}

也就是繼續設定default參數。然後調用mOutput->setDevice(mMode, routes);

status_t AudioStreamOutALSA::setDevice(int mode, uint32_t newDevice)

{

AutoMutex lock(mLock);

return ALSAStreamOps::setDevice(mode, newDevice);

}

status_t ALSAStreamOps::setDevice(int mode, uint32_t device)

{

// Close off previously opened device.

// It would be nice to determine if the underlying device actually

// changes, but we might be manipulating mixer settings (see asound.conf).

close();

const char *stream = streamName();

status_t status = open (mode, device);

int err;

if (status != NO_ERROR)

return status;

err = snd_pcm_hw_params_any(mHandle, mHardwareParams);

if (err < 0) {

LOGE("Unable to configure hardware: %s", snd_strerror(err));

return NO_INIT;

}

status = setPCMFormat(mDefaults->format);

// Set the interleaved read and write format.

err = snd_pcm_hw_params_set_access(mHandle, mHardwareParams,

SND_PCM_ACCESS_RW_INTERLEAVED);

if (err < 0) {

LOGE("Unable to configure PCM read/write format: %s",

snd_strerror(err));

return NO_INIT;

}

// Some devices do not have the default two channels. Force an error to

// prevent AudioMixer from crashing and taking the whole system down.

// Note that some devices will return an -EINVAL if the channel count

// is queried before it has been set. i.e. calling channelCount()

// before channelCount(channels) may return -EINVAL.

status = channelCount(mDefaults->channels);

if (status != NO_ERROR)

return status;

// Don't check for failure; some devices do not support the default

// sample rate.

sampleRate(mDefaults->sampleRate);

// Disable hardware resampling.

status = setHardwareResample(false);

if (status != NO_ERROR)

return status;

snd_pcm_uframes_t bufferSize = mDefaults->bufferSize;

unsigned int latency = mDefaults->latency;

// Make sure we have at least the size we originally wanted

err = snd_pcm_hw_params_set_buffer_size(mHandle, mHardwareParams, bufferSize);

if (err < 0) {

LOGE("Unable to set buffer size to %d: %s",

(int)bufferSize, snd_strerror(err));

return NO_INIT;

}

// Setup buffers for latency

err = snd_pcm_hw_params_set_buffer_time_near (mHandle, mHardwareParams,

&latency, NULL);

if (err < 0) {

unsigned int periodTime = latency / 4;

err = snd_pcm_hw_params_set_period_time_near (mHandle, mHardwareParams,

&periodTime, NULL);

if (err < 0) {

LOGE("Unable to set the period time for latency: %s", snd_strerror(err));

return NO_INIT;

}

snd_pcm_uframes_t periodSize;

err = snd_pcm_hw_params_get_period_size (mHardwareParams, &periodSize, NULL);

if (err < 0) {

LOGE("Unable to get the period size for latency: %s", snd_strerror(err));

return NO_INIT;

}

bufferSize = periodSize * 4;

if (bufferSize < mDefaults->bufferSize)

bufferSize = mDefaults->bufferSize;

err = snd_pcm_hw_params_set_buffer_size_near (mHandle, mHardwareParams, &bufferSize);

if (err < 0) {

LOGE("Unable to set the buffer size for latency: %s", snd_strerror(err));

return NO_INIT;

}

} else {

// OK, we got buffer time near what we expect. See what that did for bufferSize.

err = snd_pcm_hw_params_get_buffer_size (mHardwareParams, &bufferSize);

if (err < 0) {

LOGE("Unable to get the buffer size for latency: %s", snd_strerror(err));

return NO_INIT;

}

// Does set_buffer_time_near change the passed value? It should.

err = snd_pcm_hw_params_get_buffer_time (mHardwareParams, &latency, NULL);

if (err < 0) {

LOGE("Unable to get the buffer time for latency: %s", snd_strerror(err));

return NO_INIT;

}

unsigned int periodTime = latency / 4;

err = snd_pcm_hw_params_set_period_time_near (mHandle, mHardwareParams,

&periodTime, NULL);

if (err < 0) {

LOGE("Unable to set the period time for latency: %s", snd_strerror(err));

return NO_INIT;

}

LOGD("Buffer size: %d", (int)bufferSize);

LOGD("Latency: %d", (int)latency);

mDefaults->bufferSize = bufferSize;

mDefaults->latency = latency;

// Commit the hardware parameters back to the device.

err = snd_pcm_hw_params(mHandle, mHardwareParams);

if (err < 0) {

LOGE("Unable to set hardware parameters: %s", snd_strerror(err));

return NO_INIT;

}

status = setSoftwareParams();

return status;

}

在調用了一系列設定參數的函數後，最後調用err = snd_pcm_hw_params(mHandle, mHardwareParams);

将參數寫入到硬體。

完成這個過程後，接着建立 MixerThread ，AudioRecordThread ，至此media.audio_flinger服務初始化完畢。

在我們的MID上，wince6.0下有調試WM9714通過，可正常發聲和錄音。可android下使用alsa_aplay測試總是出現問題。

一氣之下重新解壓android包，編譯，燒寫ramdisk-uboot.img,system.img,userdata.img,偶然在android界面點開music player，找

到SD卡上的音樂，竟然能發出不連續的聲音，欣喜若狂！趕緊在kernel中關掉列印資訊，再次進入播放，悠揚的聲音出來了！

暈~~~ 竟然不用動什麼東西！也不需要配asound.conf，所需要注意的是測試在android界面裡用music player播放就好，

alsa_aplay還是在這種場合不合适，因為android加載，占用了音頻裝置。每次燒寫也應該帶上ramdisk-uboot.img，這個很重要。

我的alsa_aplay版本：

# alsa_aplay --version

aplay: version 1.0.19 by Jaroslav Kysela <[email protected]>

android+WM9714(AC97)調試

繼續閱讀

android常用工具類轉載自：http://www.trinea.cn/android/android-common-utils/Android常用工具類

Android常用工具類ToastUtilsSizeUtilsLogUtilsRetrofitManagerAnimUtil

轉換字元串成ＩＮＴ６４

c寫檔案

SIP Presence (二)RFC 3265 - Session Initiation Protocol (SIP)-Specific Event Notification

坐标系統和投影變換在桌面産品中的應用

Android Studio 0.8.0解決 Failure [INSTALL_FAILED_OLDER_SDK] 問題

Android – ListView 中添加按鈕，動态删除添加ItemView的操作

Kotlin的ListView與RecyclerView的簡單應用

關于公司的那些坑：Android Studio篇

android SQLite事務和kotlin協程

無法正常編譯錯誤Compilation failed to complete

關于Gradle配置的小結

AOP程式設計_Android優雅權限架構(1)概念基礎，2021金三銀四前言正文大綱正文

Android APK檔案反編譯Android APK檔案反編譯

Linux裝置模型（中）之上層容器