硬件抽象層是介於android內核kernel和上層之間的抽象出來的一層結構。他是對linux驅動的一個封裝,對上層提供統一接口,上層應用不必知道下層硬件具體怎麼實現工作的,它屏蔽了底層的實現細節。
它在整個android架構中的位置如下圖所示:
傳統的linux對硬件的操作基本上在內核空間的linux驅動程序中實現了,那麼現在爲什麼要多此一舉把對硬件的操作分爲HAL和linux驅動兩部分呢?而且HAL是屬於用戶空間,linux驅動屬於內核空間。
其實並不多餘,理由是很多的:
1.谷歌搭好了HAL的框架,爲上層framework通過JNI調用HAL提供了統一的API,硬件開發商或者移植人員只需要按照框架開發即可,無需話費精力在與上層的交互上的實現上,將精力放在hal層本身的實現上即可。
2.從商業角度,許多硬件廠商不願意將自己硬件相關一些核心的東西開源出去,假如將對自己硬件的驅動程序全部放入內核空間驅動程序實現,那麼必須遵循GPL協議,是必需開源的。有了HAL層之後,他們可以把一些核心的算法之類的東西的實現放在HAL層,而HAL層位於用戶空間,不屬於linux內核,和android源碼一樣遵循的是Apache協議,這個是可以不用開源的。
以上就是HAL層存在的意義,下面來根據HAL層源碼分析一下HAL到底是怎麼樣個架構和實現原理,深入剖析一下。
android HAL層的代碼主要位於/hardware/libhardware下面我們從上往下走。
在HAL層中,各類硬件的都是以硬件模塊的形式描述的,是用hw_module_t結構體來描述的,而每一類硬件模塊中又有各個獨立的硬件,是用hw_device_t結構體來描述的。
上層app通過JNI調用硬件時,首先得獲取到hw_module_t結構體,也即是硬件模塊,有了這個才能再對硬件進行操作。那麼我們來看看看看這兩個結構體定義是什麼樣子的。
它們的定義在/hardware/libhardware/include/hardware/hardware.h裏面。
A. hw_module_t表示硬件模塊,它主要包含了一些硬件模塊的信息,結構體的定義:
/** * Every hardware module must have a data structure named HAL_MODULE_INFO_SYM * and the fields of this data structure must begin with hw_module_t * followed by module specific information. */ typedef struct hw_module_t { /** tag must be initialized to HARDWARE_MODULE_TAG */ uint32_t tag; /** * The API version of the implemented module. The module owner is * responsible for updating the version when a module interface has * changed. * * The derived modules such as gralloc and audio own and manage this field. * The module user must interpret the version field to decide whether or * not to inter-operate with the supplied module implementation. * For example, SurfaceFlinger is responsible for making sure that * it knows how to manage different versions of the gralloc-module API, * and AudioFlinger must know how to do the same for audio-module API. * * The module API version should include a major and a minor component. * For example, version 1.0 could be represented as 0x0100. This format * implies that versions 0x0100-0x01ff are all API-compatible. * * In the future, libhardware will expose a hw_get_module_version() * (or equivalent) function that will take minimum/maximum supported * versions as arguments and would be able to reject modules with * versions outside of the supplied range. */ uint16_t module_api_version; #define version_major module_api_version /** * version_major/version_minor defines are supplied here for temporary * source code compatibility. They will be removed in the next version. * ALL clients must convert to the new version format. */ /** * The API version of the HAL module interface. This is meant to * version the hw_module_t, hw_module_methods_t, and hw_device_t * structures and definitions. * * The HAL interface owns this field. Module users/implementations * must NOT rely on this value for version information. * * Presently, 0 is the only valid value. */ uint16_t hal_api_version; #define version_minor hal_api_version /** Identifier of module */ const char *id; /** Name of this module */ const char *name; /** Author/owner/implementor of the module */ const char *author; /** Modules methods */ struct hw_module_methods_t* methods;//硬件模塊方法結構體 /** module's dso */ void* dso;//打開硬件模塊的庫時得到的句柄 #ifdef __LP64__ uint64_t reserved[32-7]; #else /** padding to 128 bytes, reserved for future use */ uint32_t reserved[32-7]; #endif } hw_module_t;
B. 下面我們再來看看hw_device_t結構體,這個結構體主要是用來描述模塊中硬件設備的屬性信息什麼的。一個硬件模塊可能有多個硬件設備。比如說,傳感器模塊,sensor_module,是一個硬件模塊,但是手機中的傳感器就對應的有好多種,比如加速度acc_sensor,磁傳感器M_sensor等,那麼他們都屬於sensor_module,但是他們有都有自己的hw_device_t結構體來描述。hw_device_t定義如下:
/** * Every device data structure must begin with hw_device_t * followed by module specific public methods and attributes. */ typedef struct hw_device_t { /** tag must be initialized to HARDWARE_DEVICE_TAG */ uint32_t tag; /** * Version of the module-specific device API. This value is used by * the derived-module user to manage different device implementations. * * The module user is responsible for checking the module_api_version * and device version fields to ensure that the user is capable of * communicating with the specific module implementation. * * One module can support multiple devices with different versions. This * can be useful when a device interface changes in an incompatible way * but it is still necessary to support older implementations at the same * time. One such example is the Camera 2.0 API. * * This field is interpreted by the module user and is ignored by the * HAL interface itself. */ uint32_t version; /** reference to the module this device belongs to */ struct hw_module_t* module;//本設備歸屬的硬件模塊 /** padding reserved for future use */ #ifdef __LP64__ uint64_t reserved[12]; #else uint32_t reserved[12]; #endif /** Close this device */ int (*close)(struct hw_device_t* device);//關閉設備的函數指針 } hw_device_t;
其中,第三個成員module指向的是這個設備歸屬的硬件模塊結構體。
最後一個函數指針close指向的肯定是關閉設備的函數。
到此,HAL的主要的兩個結構體講完了,下面我們繼續,將結合源碼,看看hal層到底是怎麼工作的,看看上層怎麼獲取到硬件模塊,硬件設備的,到底是怎麼加載解析動態共享庫的。
我們知道,一些硬件廠商不願意將自己的一些核心代碼開放出去,所以將這些代碼放到HAL層,但是怎麼保證它不開放呢?HAL層代碼不是也讓大家知道下載嗎?其實硬件廠商的HAL核心代碼是以共享庫的形式出現的,每次在需要的時候,HAL會自動加載調用相關共享庫。那麼是怎麼加載找到某一硬件設備對應的共享庫的呢?這也是我們都要說的。
上層APP通過JNI調用HAL層的hw_get_module函數獲取硬件模塊,這個函數是上層與HAL打交道的入口。所以如果我們以程序調用執行的流程去看源碼的話,這個函數就是HAL層第一個被調用的函數,下面我們就
沿着程序執行的流程走下去。比如說在frameworks\native\services\sensorservice\SensorDevice.cpp中
SensorDevice::SensorDevice() : mSensorDevice(0), mSensorModule(0) { status_t err = hw_get_module(SENSORS_HARDWARE_MODULE_ID, (hw_module_t const**)&mSensorModule); ALOGE_IF(err, "couldn't load %s module (%s)", SENSORS_HARDWARE_MODULE_ID, strerror(-err)); if (mSensorModule) { err = sensors_open_1(&mSensorModule->common, &mSensorDevice); ALOGE_IF(err, "couldn't open device for module %s (%s)", SENSORS_HARDWARE_MODULE_ID, strerror(-err)); if (mSensorDevice) { if (mSensorDevice->common.version == SENSORS_DEVICE_API_VERSION_1_1 || mSensorDevice->common.version == SENSORS_DEVICE_API_VERSION_1_2) { ALOGE(">>>> WARNING <<< Upgrade sensor HAL to version 1_3"); } sensor_t const* list; ssize_t count = mSensorModule->get_sensors_list(mSensorModule, &list); mActivationCount.setCapacity(count); Info model; for (size_t i=0 ; i<size_t(count) ; i++) { mActivationCount.add(list[i].handle, model); mSensorDevice->activate( reinterpret_cast<struct sensors_poll_device_t *>(mSensorDevice), list[i].handle, 0); } } } }
這裏調用hw_get_module()得到硬件模塊結構體hw_module_t.有了hw_module_t,通過其內部的method open就能打開硬件模塊對應的設備了,通過結構體中的一些方法就能操作硬件設備了。
接下來我們看hw_get_module函數具體是怎麼實現的,
hw_get_module函數定義在hardware\libhardware\hardware.c中,打開這個文件可以看到定義如下:
int hw_get_module(const char *id, const struct hw_module_t **module) { return hw_get_module_by_class(id, NULL, module); }
這是其函數原型,傳入的參數id對應硬件設備的id,這是一個字符串,比如我們比較熟悉的sensor的id是SENSORS_HARDWARE_MODULE_ID,這個跟HAL_MODULE_INFO_SYM結構體中id的定義一樣,這個將在下面講到。hw_get_module函數根據這個id去查找匹配和這個id對應的硬件模塊結構體的。如果找到了對應的hw_module_t結構體,會將其指針放入*module中。
然後我們進入hw_get_module_by_class()看看它具體實現:
int hw_get_module_by_class(const char *class_id, const char *inst, const struct hw_module_t **module) { int i = 0; char prop[PATH_MAX] = {0}; char path[PATH_MAX] = {0}; char name[PATH_MAX] = {0}; char prop_name[PATH_MAX] = {0}; if (inst) snprintf(name, PATH_MAX, "%s.%s", class_id, inst); else strlcpy(name, class_id, PATH_MAX);//拷貝hardware id,name爲sensor /* * Here we rely on the fact that calling dlopen multiple times on * the same .so will simply increment a refcount (and not load * a new copy of the library). * We also assume that dlopen() is thread-safe. */ /* First try a property specific to the class and possibly instance */ snprintf(prop_name, sizeof(prop_name), "ro.hardware.%s", name);//先嚐試獲取ro.hardware.sensor的屬性值 if (property_get(prop_name, prop, NULL) > 0) { if (hw_module_exists(path, sizeof(path), name, prop) == 0) {//查看是否存在對應的庫 goto found; } } /* Loop through the configuration variants looking for a module */ for (i=0 ; i<HAL_VARIANT_KEYS_COUNT; i++) { if (property_get(variant_keys[i], prop, NULL) == 0) {//獲取數組variant_keys裏面的屬性值 continue; } if (hw_module_exists(path, sizeof(path), name, prop) == 0) { goto found; } } /* Nothing found, try the default */ //當所有變種名稱形式的包都不存在時,就以"id.default.so"形式包名查找是否存在 if (hw_module_exists(path, sizeof(path), name, "default") == 0) { goto found; } return -ENOENT; found: /* load the module, if this fails, we're doomed, and we should not try * to load a different variant. */ return load(class_id, path, module);//裝載庫,得到module }
prop這個變量的值是前面property_get(variant_keys[i], prop, NULL)函數獲取到的,其實這個函數是通過ariant_keys數組的屬性查找到系統中對應的變種名稱。不同的平臺獲取到prop值是不一樣的。然後將參數傳入hw_module_exists()函數進行判斷該庫是否存在。
hw_module_exists如下:
/* * Check if a HAL with given name and subname exists, if so return 0, otherwise * otherwise return negative. On success path will contain the path to the HAL. */ static int hw_module_exists(char *path, size_t path_len, const char *name, const char *subname) { snprintf(path, path_len, "%s/%s.%s.so",//查找動態鏈接庫的兩個路徑 vendor/lib/hw/ system/lib/hw HAL_LIBRARY_PATH2, name, subname); if (access(path, R_OK) == 0) return 0; snprintf(path, path_len, "%s/%s.%s.so", HAL_LIBRARY_PATH1, name, subname); if (access(path, R_OK) == 0) return 0; return -ENOENT; }
該函數先拼湊出path,結合 HAL_LIBRARY_PATH2 爲"/vendor/lib/hw",假設獲取到的prop值是tout,需要獲取的硬件模塊的id是leds,這種情況下path拼出來的值是/vender/lib/hw/leds.tout.so,然後在判斷文件是否存在。存在返回0.
variant_keys定義如下:
/** * There are a set of variant filename for modules. The form of the filename * is "<MODULE_ID>.variant.so" so for the led module the Dream variants * of base "ro.product.board", "ro.board.platform" and "ro.arch" would be: * * led.trout.so * led.msm7k.so * led.ARMV6.so * led.default.so */ static const char *variant_keys[] = { "ro.hardware", /* This goes first so that it can pick up a different file on the emulator. */ "ro.product.board", "ro.board.platform", "ro.arch" };
上述代碼主要是搜索庫並獲取動態鏈接庫的路徑,然後調用load函數去加載指定路徑下的庫文件,load函數是關鍵所在。下面我們看看load():
/** * Load the file defined by the variant and if successful * return the dlopen handle and the hmi. * @return 0 = success, !0 = failure. */ static int load(const char *id, const char *path, const struct hw_module_t **pHmi)//傳入硬件模塊id和庫所在路徑,獲取到硬件模塊結構體 { int status = -EINVAL; void *handle = NULL; struct hw_module_t *hmi = NULL; /* * load the symbols resolving undefined symbols before * dlopen returns. Since RTLD_GLOBAL is not or'd in with * RTLD_NOW the external symbols will not be global */ handle = dlopen(path, RTLD_NOW);//打開動態連接庫 if (handle == NULL) { char const *err_str = dlerror(); ALOGE("load: module=%s\n%s", path, err_str?err_str:"unknown"); status = -EINVAL; goto done; } /* Get the address of the struct hal_module_info. */ const char *sym = HAL_MODULE_INFO_SYM_AS_STR;//這個變量很重要,HMI就是模塊結構的符號標誌 hmi = (struct hw_module_t *)dlsym(handle, sym);//根據符號返回符號對應的地址 if (hmi == NULL) { ALOGE("load: couldn't find symbol %s", sym); status = -EINVAL; goto done; } /* Check that the id matches */ if (strcmp(id, hmi->id) != 0) {//匹配解析出硬件模塊的id和傳入我們實際想要得到的模塊id是否一致。 ALOGE("load: id=%s != hmi->id=%s", id, hmi->id); status = -EINVAL; goto done; } hmi->dso = handle;//將打開庫得到的句柄傳給硬件模塊的dso /* success */ status = 0; done: if (status != 0) { hmi = NULL; if (handle != NULL) { dlclose(handle); handle = NULL; } } else { ALOGV("loaded HAL id=%s path=%s hmi=%p handle=%p", id, path, *pHmi, handle); } *pHmi = hmi;//返回找到的模塊結構,賦值給module return status; }
可以看到load函數傳入的幾個參數,第一個參數就是需要加載的硬件模塊對應動態庫的硬件模塊的id,
第二個參數就是動態庫存放的路徑,就是在hw_get_module函數前部分搜索庫得到的path,
第三個參數就是我們需要得到的硬件模塊結構體,通過它傳給hw_get_module,hw_get_module函數在通過參數傳給jni。
首先調用dlopen打開共享庫,該函數通過傳入的庫的路徑找到庫,並且打開它,傳回一個操作句柄handle,然後再調用dlsym函數解析這個打開的庫,得到庫中包含的對應的硬件模塊結構體,並將它返回回來。所有硬件廠商或者硬件移植者都必須根據HAL的這個架構去實現填充這個和自己硬件相關的硬件模塊結構體hw_module_t,供使用。
通過dlsym解析之後就得到了hw_module_t,將從庫中解析得到的結構體中的id和傳入的id做比較,看是否一致。如果一致則證明就是得到正確的硬件模塊了。
最後將hw_module_t結構體指針傳給第三個參數,傳給hw_get_module函數。
提醒一下,HAL_MODULE_INFO_SYM_AS_STR這個宏很重要:
/** * Name of the hal_module_info as a string */ #define HAL_MODULE_INFO_SYM_AS_STR "HMI"
硬件抽象層中的每一個模塊都必須存在一個導出符號HAL_MODULE_IFNO_SYM,即「HMI」,它指向一個自定義的硬件抽象層模塊結構體。以sensor爲例,在\hardware\libhardware\modules\sensors\multihal.cpp下
struct sensors_module_t HAL_MODULE_INFO_SYM = { common :{ tag : HARDWARE_MODULE_TAG, version_major : 1, version_minor : 1, id : SENSORS_HARDWARE_MODULE_ID, name : "MultiHal Sensor Module", author : "Google, Inc", methods : &sensors_module_methods, dso : NULL, reserved : {0}, }, get_sensors_list : module__get_sensors_list };
這裏定義了一個名爲HAL_MODULE_INFO_SYM的sensors_module_t的結構體,事實上每個硬件模塊都必須定義一個變量名爲HAL_MODULE_INFO_SYM的結構體,而且common成員爲hw_module_t類型(這段話在定義hw_module_t結構體時有說明)。注意這裏的HAL_MODULE_INFO_SYM變量必須爲這個名字,這樣編譯器纔會將這個結構體的導出符號變爲「HMI」,dlsym函數才能找到這個結構體,才能根據符號返回符號對應的模塊地址。
另外在hw_module_t結構體中定義了struct hw_module_methods_t* methods; hw_module_methods_t
這個指針methods它指向的是與本硬件模塊相關的方法的結構體,裏面不用看可以猜出肯定有一些函數指針,但是它裏面只有一個函數指針。可以看看定義:
typedef struct hw_module_methods_t { /** Open a specific device */ int (*open)(const struct hw_module_t* module, const char* id, struct hw_device_t** device); } hw_module_methods_t;
我們可以看到確實只有一個函數指針,open它是打開硬件模塊中硬件設備的函數。
以sensors爲例,在hardware\libhardware\modules\sensors\multihal.cpp中有關於sensors打開設備的方法定義代碼如下
static struct hw_module_methods_t sensors_module_methods = { open : open_sensors };
我們進入open_sensors,看看其具體實現:
static int open_sensors(const struct hw_module_t* hw_module, const char* name, struct hw_device_t** hw_device_out) { ALOGV("open_sensors begin..."); lazy_init_modules(); // Create proxy device, to return later. sensors_poll_context_t *dev = new sensors_poll_context_t(); memset(dev, 0, sizeof(sensors_poll_device_1_t)); dev->proxy_device.common.tag = HARDWARE_DEVICE_TAG; dev->proxy_device.common.version = SENSORS_DEVICE_API_VERSION_1_3; dev->proxy_device.common.module = const_cast<hw_module_t*>(hw_module); dev->proxy_device.common.close = device__close; dev->proxy_device.activate = device__activate; dev->proxy_device.setDelay = device__setDelay; dev->proxy_device.poll = device__poll; dev->proxy_device.batch = device__batch; dev->proxy_device.flush = device__flush; dev->nextReadIndex = 0; // Open() the subhal modules. Remember their devices in a vector parallel to sub_hw_modules. for (std::vector<hw_module_t*>::iterator it = sub_hw_modules->begin(); it != sub_hw_modules->end(); it++) { sensors_module_t *sensors_module = (sensors_module_t*) *it; struct hw_device_t* sub_hw_device; int sub_open_result = sensors_module->common.methods->open(*it, name, &sub_hw_device); if (!sub_open_result) { if (!HAL_VERSION_IS_COMPLIANT(sub_hw_device->version)) { ALOGE("SENSORS_DEVICE_API_VERSION_1_3 is required for all sensor HALs"); ALOGE("This HAL reports non-compliant API level : %s", apiNumToStr(sub_hw_device->version)); ALOGE("Sensors belonging to this HAL will get ignored !"); } dev->addSubHwDevice(sub_hw_device); } } // Prepare the output param and return *hw_device_out = &dev->proxy_device.common; ALOGV("...open_sensors end"); return 0; }
可以看到裏面對設備的一些賦值
以上是我這兩天學習硬件抽象層知識的一些總結。