Android5.0Camera系统源码分析(1)：CameraService启动流程
1.前言
本文将分析Android系统源码，从frameworks层到hal层，暂不涉及app层和kernel层。由于某些函数比较复杂，在贴出代码时会适当对其进行简化。本文属于自己对源码的总结，仅仅是贯穿代码流程，不会深入分析各个细节。
分析android系统源码，需要对android系统的某些知识点有所了解
2.frameworks层
Android的各个子模块的启动都是从它们的Service的启动开始的，所以我们将从CameraService的启动开始分析。CameraService的启动就在MediaServer的main函数中，代码路径在：frameworks/av/media/mediaserver/main_mediaserver.cpp
[cpp]viewplaincopy
intmain(intargc__unused,charargv)
{
......
CameraService::instantiate();
......
}
CameraService类定义如下：
[cpp]viewplaincopy
classCameraService:
publicBinderService,
publicBnCameraService,
publicIBinder::DeathRecipient,
publiccamera_module_callbacks_t
{
staticcharconstgetServiceName(){return"media.camera";}
......
}
mediaserver的main函数中调用了CameraService的instantiate函数来创建实例，该函数的实现在其父类BinderService中实现
[cpp]viewplaincopy
template
classBinderService
{
staticstatus_tpublish(boolallowIsolated=false){
spsm(defaultServiceManager());
returnsm->addService(
String16(SERVICE::getServiceName()),
newSERVICE(),allowIsolated);
}
staticvoidinstantiate(){publish();}
}
1.instantiate函数只是简单的调用了publish函数
2.publish函数先构造CameraService，再通过addService函数将它注册到ServiceManager当中，而getServiceName函数获取到的值为“mediacamera”。这一切都是为了binder通信做准备
3.这里使用了c++模版，从上面的CameraService类定义中可以看出，这里的SERVICE等于CameraService，也就是说publish函数中的newSERVICE等于newCameraService
4.同时还使用了智能指针，也就是说除了调用CameraService的构造函数外，还会调用onFirstRef函数
[cpp]viewplaincopy
CameraService::CameraService()
:mSoundRef(0),mModule(0)
{
ALOGI("CameraServicestarted(pid=%d)",getpid());
gCameraService=this;
for(size_ti=0;i
mStatusList[i]=ICameraServiceListener::STATUS_PRESENT;
}
this->camera_device_status_change=android::camera_device_status_change;
}
voidCameraService::onFirstRef()
{
LOG1("CameraService::onFirstRef");
BnCameraService::onFirstRef();
if(hw_get_module(CAMERA_HARDWARE_MODULE_ID,
(consthw_module_t)&mModule)<0){
ALOGE("CouldnotloadcameraHALmodule");
mNumberOfCameras=0;
}
else{
ALOGI("Loaded\"%s\"cameramodule",mModule->common.name);
mNumberOfCameras=mModule->get_number_of_cameras();
if(mNumberOfCameras>MAX_CAMERAS){
ALOGE("Numberofcameras(%d)>MAX_CAMERAS(%d).",
mNumberOfCameras,MAX_CAMERAS);
mNumberOfCameras=MAX_CAMERAS;
}
for(inti=0;i
LOG1("setCameraFree(%d)",i);
setCameraFree(i);
}
if(mModule->common.module_api_version>=
CAMERA_MODULE_API_VERSION_2_1){
mModule->set_callbacks(this);
}
VendorTagDescriptor::clearGlobalVendorTagDescriptor();
if(mModule->common.module_api_version>=CAMERA_MODULE_API_VERSION_2_2){
setUpVendorTags();
}
CameraDeviceFactory::registerService(this);
}
}
第20行.通过hw_get_module函数加载了一个hw_module_t模块，这个模块是与hal层对接的接口，ID为CAMERA_HARDWARE_MODULE_ID，并将它保存在mModule成员变量中。
第27行.通过mModule->get_number_of_cameras函数进入到hal层，获取到了camera的个数。这个函数很重要，对于frameworks层来说只是拿到了camera的个数，但对于hal层和drivers层来说Camera的上电和初始化流程都是从这里开始的
3.hal层-基于MTK平台
先来看看mtkcameramodule的定义，代码路径在：vendor/mediatek/proprietary/hardware/mtkcam/module_hal/module/module.h
[cpp]viewplaincopy
static
camera_module
get_camera_module()
{
camera_modulemodule={
common:{
tag:HARDWARE_MODULE_TAG,
#if(PLATFORM_SDK_VERSION>=21)
module_api_version:CAMERA_MODULE_API_VERSION_2_3,
#else
module_api_version:CAMERA_DEVICE_API_VERSION_1_0,
#endif
hal_api_version:HARDWARE_HAL_API_VERSION,
id:CAMERA_HARDWARE_MODULE_ID,
name:"MediaTekCameraModule",
author:"MediaTek",
methods:get_module_methods(),
dso:NULL,
reserved:{0},
},
get_number_of_cameras:get_number_of_cameras,
get_camera_info:get_camera_info,
set_callbacks:set_callbacks,
get_vendor_tag_ops:get_vendor_tag_ops,
#if(PLATFORM_SDK_VERSION>=21)
open_legacy:open_legacy,
#endif
reserved:{0},
};
returnmodule;
};
1.保存在frameworks层CameraService的成员变量mModule里面的就是上面这个module结构体
2.当frameworks层调用mModule->get_number_of_cameras函数时，实际就是调用上面结构体的get_number_of_cameras函数
[cpp]viewplaincopy
CamDeviceManagerImpgCamDeviceManager;
ICamDeviceManager
getCamDeviceManager()
{
return&gCamDeviceManager;
}
static
int
get_number_of_cameras(void)
{
returnNSCam::getCamDeviceManager()->getNumberOfDevices();
}
1.这里先通过getCamDeviceManager函数获取了CamDeviceManagerImp对象
2.CamDeviceManagerImp继承了CamDeviceManagerBase，这里的getNumberOfDevices方法将由父类CamDeviceManagerBase实现
[cpp]viewplaincopy
int32_t
CamDeviceManagerBase::
getNumberOfDevices()
{
mi4DeviceNum=enumDeviceLocked();
returnmi4DeviceNum;
}
这里只是调用了enumDeviceLocked函数，并将它的返回值（代表了camera的个数）返回到frameworks层。接着看enumDeviceLocked的实现
[cpp]viewplaincopy
int32_t
CamDeviceManagerImp::
enumDeviceLocked()
{
IHalSensorListconstpHalSensorList=IHalSensorList::get();
size_tconstsensorNum=pHalSensorList->searchSensors();
for(size_ti=0;i
{
int32_tconstdeviceId=i;
sppInfo=newEnumInfo;
mEnumMap.add(deviceId,pInfo);
IMetadataProvider>pMetadataProvider=IMetadataProvider::create(deviceId);
pInfo->pMetadata=pMetadataProvider->getStaticCharacteristics();
pInfo->iFacing=(pMetadataProvider->getDeviceFacing()==MTK_LENS_FACING_FRONT)
?CAMERA_FACING_FRONT
:CAMERA_FACING_BACK
;
pInfo->iWantedOrientation=pMetadataProvider->getDeviceWantedOrientation();
pInfo->iSetupOrientation=pMetadataProvider->getDeviceSetupOrientation();
i4DeviceNum++;
}
returni4DeviceNum;
}
第5-6行.这里需要重点关注pHalSensorList->searchSensors函数，它的返回值就是camera的个数
第8-24行.循环构造并初始化一个EnumInfo对象，并把它保存在mEnumMap中
[cpp]viewplaincopy
MUINT
HalSensorList::
enumerateSensor_Locked()
{
intret_count=0;
SensorDrvconstpSensorDrv=SensorDrv::get();
intconstiSensorsList=pSensorDrv->impSearchSensor(NULL);
if((iSensorsList&SENSOR_DEV_MAIN)==SENSOR_DEV_MAIN)
{
halSensorDev=SENSOR_DEV_MAIN;
pSensorInfo=pSensorDrv->getMainSensorInfo();
addAndInitSensorEnumInfo_Locked(halSensorDev,ret_count,mapToSensorType(pSensorInfo->GetType()),pSensorInfo->getDrvMacroName());
ret_count++;
}
if((iSensorsList&SENSOR_DEV_SUB)==SENSOR_DEV_SUB)
{
halSensorDev=SENSOR_DEV_SUB;
pSensorInfo=pSensorDrv->getSubSensorInfo();
addAndInitSensorEnumInfo_Locked(halSensorDev,ret_count,mapToSensorType(pSensorInfo->GetType()),pSensorInfo->getDrvMacroName());
ret_count++;
}
mEnumSensorCount=ret_count;
returnret_count;
}
MUINT
HalSensorList::
searchSensors()
{
returnenumerateSensor_Locked();
}
第33行.searchSensors函数只是调用了enumerateSensor_Locked函数，这里并没有贴出enumerateSensor_Locked函数的所有代码，删减了一些我们暂时不关注的东西
第7行.重点函数pSensorDrv->impSearchSensor，它的返回值决定了enumerateSensor_Locked的返回值，也就是camera的个数
[cpp]viewplaincopy
MINT32
ImgSensorDrv::impSearchSensor(pfExIdChkpExIdChkCbf)
{
MUINT32SensorEnum=(MUINT32)DUAL_CAMERA_MAIN_SENSOR;
MUINT32i,id[KDIMGSENSOR_MAX_INVOKE_DRIVERS]={0,0};
MINT32sensorDevs=SENSOR_NONE;
GetSensorInitFuncList(&m_pstSensorInitFunc);
m_fdSensor=::open("/dev/kd_camera_hw",O_RDWR);
for(SensorEnum=DUAL_CAMERA_MAIN_SENSOR;SensorEnum<=DUAL_CAMERA_SUB_SENSOR;SensorEnum<<=1){
for(i=0;i
//endofdriverlist
if(m_pstSensorInitFunc[i].getCameraDefault==NULL){
LOG_MSG("m_pstSensorInitFunc[i].getCameraDefaultisNULL:%d\n",i);
break;
}
id[KDIMGSENSOR_INVOKE_DRIVER_0]=(SensorEnum<
err=ioctl(m_fdSensor,KDIMGSENSORIOC_X_SET_DRIVER,&id[KDIMGSENSOR_INVOKE_DRIVER_0]);
err=ioctl(m_fdSensor,KDIMGSENSORIOC_T_CHECK_IS_ALIVE);
if(err<0||err2<0){
LOG_MSG("sensorIDmismatch\n");
continue;
}
if(SensorEnum==DUAL_CAMERA_MAIN_SENSOR){
m_mainSensorDrv.index[m_mainSensorDrv.number]=i;
m_mainSensorDrv.type[m_mainSensorDrv.number]=sensorType;
m_mainSensorDrv.position=socketPos;
m_mainSensorDrv.sensorID=m_pstSensorInitFunc[m_mainSensorDrv.index[m_mainSensorDrv.number]].SensorId;
m_mainSensorDrv.nuwww.shanxiwang.netmber++;
}elseif(SensorEnum==DUAL_CAMERA_SUB_SENSOR){
m_subSensorDrv.index[m_subSensorDrv.number]=i;
m_subSensorDrv.type[m_subSensorDrv.number]=sensorType;
m_subSensorDrv.position=socketPos;
m_subSensorDrv.sensorID=m_pstSensorInitFunc[m_subSensorDrv.index[m_subSensorDrv.number]].SensorId;
m_subSensorDrv.number++;
}
}
}
if(BAD_SENSOR_INDEX!=m_mainSensorDrv.index[0]){
m_mainSensorId=m_mainSensorDrv.sensorID;
m_mainSensorIdx=m_mainSensorDrv.index[0];
sensorDevs|=SENSOR_MAIN;
}
if(BAD_SENSOR_INDEX!=m_subSensorDrv.index[0]){
m_subSensorId=m_subSensorDrv.sensorID;
m_subSensorIdx=m_subSensorDrv.index[0];
sensorDevs|=SENSOR_SUB;
}
returnsensorDevs;
}
这个函数比较长，所以只贴出关键代码
第8行,调用GetSensorInitFuncList函数来获取hal层的sersors列表，并把它保存在m_pstSensorInitFunc变量中
第9行,通过系统调用open函数打开camera的设备节点，后面会通过这个节点来进入到kernel层
第11-12行,通过两个for循环来遍历sensorlist中所有可能存在的camera
第20行,通过ioctl下达setDriver指令，并下传正在遍历的sensorlist中的ID。Driver层根据这个ID，挂载Driver层sensorlist中对应的操作接口
第21行,通过ioctl下达checkID指令，Driver层为对应sensor上电，通过I2C读取预存在寄存器中的sensorid。然后比较读取结果，如果不匹配returnerror后继续遍历
第29-41行,将sensor相关的信息保存在m_mainSensorDrv和m_subSensorDrv中
第45-56行,给sensroDevs变量赋值，并将它返回给上一级
这里暂不分析kernel层的代码，先来看看GetSensorInitFuncList函数，代码在sensorlist.cpp中
[cpp]viewplaincopy
MSDK_SENSOR_INIT_FUNCTION_STRUCTSensorList[]=
{
#ifdefined(IMX175_MIPI_RAW)
RAW_INFO(IMX175_SENSOR_ID,SENSOR_DRVNAME_IMX175_MIPI_RAW,NULL),
#endif
#ifdefined(IMX179_MIPI_RAW)
RAW_INFO(IMX179_SENSOR_ID,SENSOR_DRVNAME_IMX179_MIPI_RAW,NULL),
#endif
#ifdefined(IMX219_MIPI_RAW)
RAW_INFO(IMX219_SENSOR_ID,SENSOR_DRVNAME_IMX219_MIPI_RAW,NULL),
#endif
#ifdefined(IMX214_MIPI_RAW)
RAW_INFO(IMX214_SENSOR_ID,SENSOR_DRVNAME_IMX214_MIPI_RAW,NULL),
#endif
#ifdefined(GC2235_RAW)
RAW_INFO(GC2235_SENSOR_ID,SENSOR_DRVNAME_GC2235_RAW,NULL),
#endif
#ifdefined(GC2035_YUV)
YUV_INFO(GC2035_SENSOR_ID,SENSOR_DRVNAME_GC2035_YUV,NULL),
#endif
......
}
UINT32GetSensorInitFuncList(MSDK_SENSOR_INIT_FUNCTION_STRUCTppSensorList)
{
ppSensorList=&SensorList[0];
returnMHAL_NO_ERROR;
}
hal层的sensorList，再熟悉不过的代码，需要注意的是hal层sensorList和kernel层的sensorList顺序必须保持一致
4.总结
至此，除kernel层外，简述了CameraService的启动流程，大概过程如下图所示
|
|
