Input系統之InputReader概要性實例分析
InputReader 的創建
從 InputManagerService: 創建與啟動 可知,Input 系統的主要功能,主要集中在 native 層,並且Input 系統的 native 層又包含 InputReader, InputClassifer, InputDispatcher 三個子模塊。本文來分析 InputReader 從創建到啟動的基本流程,為後續分析 InputReader 的每一個功能打好基礎。
從 InputManagerService: 創建與啟動 可知, InputReader 的創建過程如下
// InputReaderFactory.cpp sp<InputReaderInterface> createInputReader(const sp<InputReaderPolicyInterface>& policy, const sp<InputListenerInterface>& listener) { return new InputReader(std::make_unique<EventHub>(), policy, listener); }
InputReader 依賴 EventHub,因此首先要看下 EventHub 的創建過程
EventHub::EventHub(void) : mBuiltInKeyboardId(NO_BUILT_IN_KEYBOARD), mNextDeviceId(1), mControllerNumbers(), mNeedToSendFinishedDeviceScan(false), mNeedToReopenDevices(false), mNeedToScanDevices(true), // mNeedToScanDevices 初始化為 true,表示需要掃描輸入設備 mPendingEventCount(0), mPendingEventIndex(0), mPendingINotify(false) { ensureProcessCanBlockSuspend(); // 1. 創建 epoll mEpollFd = epoll_create1(EPOLL_CLOEXEC); // 2. 初始化 inotify mINotifyFd = inotify_init(); // 監聽 /dev/input/ 目錄項的創建與刪除,其實就是監聽輸入設備的創建與刪除 mInputWd = inotify_add_watch(mINotifyFd, DEVICE_PATH, IN_DELETE | IN_CREATE); // ... // 3. epoll 監聽 inotify 事件 // 可讀事件,表明有輸入設備的創建與刪除 struct epoll_event eventItem = {}; eventItem.events = EPOLLIN | EPOLLWAKEUP; eventItem.data.fd = mINotifyFd; int result = epoll_ctl(mEpollFd, EPOLL_CTL_ADD, mINotifyFd, &eventItem); // 4. 創建管道 int wakeFds[2]; result = pipe(wakeFds); mWakeReadPipeFd = wakeFds[0]; mWakeWritePipeFd = wakeFds[1]; // 設置管道兩端為非阻塞 result = fcntl(mWakeReadPipeFd, F_SETFL, O_NONBLOCK); result = fcntl(mWakeWritePipeFd, F_SETFL, O_NONBLOCK); // 5. epoll 監聽管道讀端的事件 // 可讀事件,表明需要喚醒 InputReader 線程,觸發條件一般為配置更新 eventItem.data.fd = mWakeReadPipeFd; result = epoll_ctl(mEpollFd, EPOLL_CTL_ADD, mWakeReadPipeFd, &eventItem); }
EventHub 創建過程如下
- 創建 epoll 實例。
- 初始化 inotify 實例,並用 epoll 監聽它的事件。當輸入設備添加/刪除時,epoll 就會收到 inotify 的可讀事件,因此 EventHub 和 InputReader 就可以動態地處理輸入設備的添加/刪除。
- 創建管道。
- epoll 監聽管道的讀端的事件。當配置更新時,會向管道的寫端寫入數據,epoll 就會收到管道的可讀事件,如果此時 InputReader 線程處於休眠狀態,那麼 InputReader 將被喚醒來處於配置更新。
epoll, inotify, pipe,它們的作用和使用方式,請讀者自行查閱 Unix/Linux 資料。
現在讓我們繼續看下 InputReader 的創建過程
InputReader::InputReader(std::shared_ptr<EventHubInterface> eventHub, const sp<InputReaderPolicyInterface>& policy, const sp<InputListenerInterface>& listener) : mContext(this), // mContext 代表 InputReader 的環境 mEventHub(eventHub), mPolicy(policy), mGlobalMetaState(0), mLedMetaState(AMETA_NUM_LOCK_ON), mGeneration(1), mNextInputDeviceId(END_RESERVED_ID), mDisableVirtualKeysTimeout(LLONG_MIN), mNextTimeout(LLONG_MAX), mConfigurationChangesToRefresh(0) { // InputReader 會把加工後的事件添加到 QueuedInputListener 隊列中,之後一起分發給 InputClassifier mQueuedListener = new QueuedInputListener(listener); { // acquire lock std::scoped_lock _l(mLock); // 刷新配置 // 其實就是更新 InputReader::mConfig refreshConfigurationLocked(0); // 更新 InputReader::mGlobalMetaState // 與鍵盤輸入設備的meta按鍵相關 updateGlobalMetaStateLocked(); } // release lock }
InputReader 的構造函數很簡單,就是成員變量的初始化。其中需要重點看下 refreshConfigurationLocked(0) 是如何刷新 InputReader 配置
// 註意,此時參數 changes 為 0 void InputReader::refreshConfigurationLocked(uint32_t changes) { // 通過 InputReaderPolicyInterface 獲取配置,保存到 InputReader::mConfig 中 mPolicy->getReaderConfiguration(&mConfig); // EventHub 保存排除的設備 mEventHub->setExcludedDevices(mConfig.excludedDeviceNames); if (!changes) return; // ... }
原來 InputReader::mConfig 代表的就是 InputReader 的配置,並且是通過 InputReaderPolicyInterface mPolicy 獲取配置的。
從 InputManagerService: 創建與啟動 可知,InputReaderPolicyInterface 接口的實現者是 NativeInputManager ,而 NativeInputManager 是 Input 系統的上層與底層溝通的橋梁,因此 InputReader 必定是通過 NativeInputManager 向上層獲取配置
void NativeInputManager::getReaderConfiguration(InputReaderConfiguration* outConfig) { ATRACE_CALL(); JNIEnv* env = jniEnv(); // 1. 通過JNI,向上層 InputManagerService 獲取配置,並保存到 outConfig 中 jint virtualKeyQuietTime = env->CallIntMethod(mServiceObj, gServiceClassInfo.getVirtualKeyQuietTimeMillis); if (!checkAndClearExceptionFromCallback(env, "getVirtualKeyQuietTimeMillis")) { outConfig->virtualKeyQuietTime = milliseconds_to_nanoseconds(virtualKeyQuietTime); } outConfig->excludedDeviceNames.clear(); jobjectArray excludedDeviceNames = jobjectArray(env->CallStaticObjectMethod( gServiceClassInfo.clazz, gServiceClassInfo.getExcludedDeviceNames)); if (!checkAndClearExceptionFromCallback(env, "getExcludedDeviceNames") && excludedDeviceNames) { jsize length = env->GetArrayLength(excludedDeviceNames); for (jsize i = 0; i < length; i++) { std::string deviceName = getStringElementFromJavaArray(env, excludedDeviceNames, i); outConfig->excludedDeviceNames.push_back(deviceName); } env->DeleteLocalRef(excludedDeviceNames); } // Associations between input ports and display ports // The java method packs the information in the following manner: // Original data: [{'inputPort1': '1'}, {'inputPort2': '2'}] // Received data: ['inputPort1', '1', 'inputPort2', '2'] // So we unpack accordingly here. outConfig->portAssociations.clear(); jobjectArray portAssociations = jobjectArray(env->CallObjectMethod(mServiceObj, gServiceClassInfo.getInputPortAssociations)); if (!checkAndClearExceptionFromCallback(env, "getInputPortAssociations") && portAssociations) { jsize length = env->GetArrayLength(portAssociations); for (jsize i = 0; i < length / 2; i++) { std::string inputPort = getStringElementFromJavaArray(env, portAssociations, 2 * i); std::string displayPortStr = getStringElementFromJavaArray(env, portAssociations, 2 * i + 1); uint8_t displayPort; // Should already have been validated earlier, but do it here for safety. bool success = ParseUint(displayPortStr, &displayPort); if (!success) { ALOGE("Could not parse entry in port configuration file, received: %s", displayPortStr.c_str()); continue; } outConfig->portAssociations.insert({inputPort, displayPort}); } env->DeleteLocalRef(portAssociations); } outConfig->uniqueIdAssociations.clear(); jobjectArray uniqueIdAssociations = jobjectArray( env->CallObjectMethod(mServiceObj, gServiceClassInfo.getInputUniqueIdAssociations)); if (!checkAndClearExceptionFromCallback(env, "getInputUniqueIdAssociations") && uniqueIdAssociations) { jsize length = env->GetArrayLength(uniqueIdAssociations); for (jsize i = 0; i < length / 2; i++) { std::string inputDeviceUniqueId = getStringElementFromJavaArray(env, uniqueIdAssociations, 2 * i); std::string displayUniqueId = getStringElementFromJavaArray(env, uniqueIdAssociations, 2 * i + 1); outConfig->uniqueIdAssociations.insert({inputDeviceUniqueId, displayUniqueId}); } env->DeleteLocalRef(uniqueIdAssociations); } jint hoverTapTimeout = env->CallIntMethod(mServiceObj, gServiceClassInfo.getHoverTapTimeout); if (!checkAndClearExceptionFromCallback(env, "getHoverTapTimeout")) { jint doubleTapTimeout = env->CallIntMethod(mServiceObj, gServiceClassInfo.getDoubleTapTimeout); if (!checkAndClearExceptionFromCallback(env, "getDoubleTapTimeout")) { jint longPressTimeout = env->CallIntMethod(mServiceObj, gServiceClassInfo.getLongPressTimeout); if (!checkAndClearExceptionFromCallback(env, "getLongPressTimeout")) { outConfig->pointerGestureTapInterval = milliseconds_to_nanoseconds(hoverTapTimeout); // We must ensure that the tap-drag interval is significantly shorter than // the long-press timeout because the tap is held down for the entire duration // of the double-tap timeout. jint tapDragInterval = max(min(longPressTimeout - 100, doubleTapTimeout), hoverTapTimeout); outConfig->pointerGestureTapDragInterval = milliseconds_to_nanoseconds(tapDragInterval); } } } jint hoverTapSlop = env->CallIntMethod(mServiceObj, gServiceClassInfo.getHoverTapSlop); if (!checkAndClearExceptionFromCallback(env, "getHoverTapSlop")) { outConfig->pointerGestureTapSlop = hoverTapSlop; } // 2. 從 NativeInputManager::mLocked 更新配置,保存到 outConfig 中 // NativeInputManager::mLocked 的數據是上層經由 InputManagerService 傳入的 { // acquire lock AutoMutex _l(mLock); outConfig->pointerVelocityControlParameters.scale = exp2f(mLocked.pointerSpeed * POINTER_SPEED_EXPONENT); outConfig->pointerGesturesEnabled = mLocked.pointerGesturesEnabled; outConfig->showTouches = mLocked.showTouches; outConfig->pointerCapture = mLocked.pointerCapture; outConfig->setDisplayViewports(mLocked.viewports); outConfig->defaultPointerDisplayId = mLocked.pointerDisplayId; outConfig->disabledDevices = mLocked.disabledInputDevices; } // release lock }
從整體看,獲取 InputReader 配置的方式有兩種
- 通過 JNI 向上層的 InputManagerService 獲取配置。
- 從 NativeInputManager::mLocked 獲取配置。
從 InputManagerService: 創建與啟動 可知,NativeInputManager::mLocked 是在 NativeInputManager 的構造函數中進行初始化的,但是它並不是不變的,而是上層經由 InputManagerService 進行操控的。
例如,mLocked.showTouches 對應開發者模式下的 Show taps 功能,InputManagerService 會監聽這個開關的狀態,相應地改變 mLocked.showTouches,並且會通知 InputReader 配置改變瞭,InputReader 在處理配置改變的過程時,會重新獲取 mLocked.showTouches 這個配置。
有 一部分 的配置是可以通過 adb shell dumpsys input 命令進行查看的
Input Manager State: Interactive: true System UI Lights Out: false Pointer Speed: 0 Pointer Gestures Enabled: true Show Touches: false Pointer Capture: Disabled, seq=0
而另外一部分配置,由於會對輸入設備進行配置,因此可以從 dump 出的輸入設備的信息中查看。
InputReader 的運行
從 InputManagerService: 創建與啟動 可知,InputReader 通過線程,循環調用 InputReader::loopOnce() 執行任務
void InputReader::loopOnce() { int32_t oldGeneration; int32_t timeoutMillis; bool inputDevicesChanged = false; std::vector<InputDeviceInfo> inputDevices; { // acquire lock std::scoped_lock _l(mLock); oldGeneration = mGeneration; timeoutMillis = -1; // 1. 如果配置有改變,那麼就刷新配置 uint32_t changes = mConfigurationChangesToRefresh; if (changes) { mConfigurationChangesToRefresh = 0; timeoutMillis = 0; // 刷新配置 refreshConfigurationLocked(changes); } else if (mNextTimeout != LLONG_MAX) { nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC); timeoutMillis = toMillisecondTimeoutDelay(now, mNextTimeout); } } // release lock // 2. 從 EventHub 獲取事件 size_t count = mEventHub->getEvents(timeoutMillis, mEventBuffer, EVENT_BUFFER_SIZE); { // acquire lock std::scoped_lock _l(mLock); mReaderIsAliveCondition.notify_all(); // 3. 處理事件 if (count) { processEventsLocked(mEventBuffer, count); } if (mNextTimeout != LLONG_MAX) { nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC); if (now >= mNextTimeout) { mNextTimeout = LLONG_MAX; timeoutExpiredLocked(now); } } // 4. 處理輸入設備改變 // 4.1 輸入設備改變,重新獲取輸入設備信息 if (oldGeneration != mGeneration) { inputDevicesChanged = true; inputDevices = getInputDevicesLocked(); } } // release lock // 4.2 通知監聽者,輸入設備改變瞭 if (inputDevicesChanged) { mPolicy->notifyInputDevicesChanged(inputDevices); } // 5. 刷新隊列中緩存的事件 // 其實就是把事件分發給 InputClassifier mQueuedListener->flush(); }
InputReader 所做的事情如下
- 如果配置改變瞭,那麼就更新配置。
- 從 EventHub 獲取事件,並處理獲取到的事件。在處理事件的過程中,InputReader 會對事件進行加工,然後保存到 QueuedInputListener 緩存隊列中。
- 如果設備發生改變,那麼重新獲取新的設備信息,並通知監聽者。
- QueuedInputListener 刷新緩存的事件,其實就是把 InputReader 加工後的事件分發給 InputClassifer。
EventHub 提供事件
InputReader 的本質就是處理從 EventHub 獲取的事件,然後分發給下一環。因為我們必須瞭解 EventHub::getEvents() 是如何為 InputReader 提供事件的
// EventHub.cpp size_t EventHub::getEvents(int timeoutMillis, RawEvent* buffer, size_t bufferSize) { ALOG_ASSERT(bufferSize >= 1); std::scoped_lock _l(mLock); struct input_event readBuffer[bufferSize]; RawEvent* event = buffer; size_t capacity = bufferSize; bool awoken = false; for (;;) { nsecs_t now = systemTime(SYSTEM_TIME_MONOTONIC); // Reopen input devices if needed. if (mNeedToReopenDevices) { // ... } // Report any devices that had last been added/removed. for (auto it = mClosingDevices.begin(); it != mClosingDevices.end();) { // ... } // 掃描輸入設備 if (mNeedToScanDevices) { mNeedToScanDevices = false; scanDevicesLocked(); mNeedToSendFinishedDeviceScan = true; } // 為掃描後打開的每一個輸入設備,填充一個類型為 DEVICE_ADDED 的事件 while (!mOpeningDevices.empty()) { std::unique_ptr<Device> device = std::move(*mOpeningDevices.rbegin()); mOpeningDevices.pop_back(); event->when = now; event->deviceId = device->id == mBuiltInKeyboardId ? 0 : device->id; event->type = DEVICE_ADDED; event += 1; // Try to find a matching video device by comparing device names for (auto it = mUnattachedVideoDevices.begin(); it != mUnattachedVideoDevices.end(); it++) { // ... } // 每次填充完事件,就把設備 Device 保存到 mDevices 中 auto [dev_it, inserted] = mDevices.insert_or_assign(device->id, std::move(device)); if (!inserted) { ALOGW("Device id %d exists, replaced.", device->id); } // 表明你需要發送設備掃描完成事件 mNeedToSendFinishedDeviceScan = true; if (--capacity == 0) { break; } } // 填充設備掃描完成事件 if (mNeedToSendFinishedDeviceScan) { mNeedToSendFinishedDeviceScan = false; event->when = now; event->type = FINISHED_DEVICE_SCAN; event += 1; if (--capacity == 0) { break; } } // Grab the next input event. bool deviceChanged = false; // 處理 epoll 事件 while (mPendingEventIndex < mPendingEventCount) { // 處理 inotify 事件,表明輸入設備新增或者刪除 const struct epoll_event& eventItem = mPendingEventItems[mPendingEventIndex++]; if (eventItem.data.fd == mINotifyFd) { if (eventItem.events & EPOLLIN) { mPendingINotify = true; } else { ALOGW("Received unexpected epoll event 0x%08x for INotify.", eventItem.events); } continue; } // 處理管道事件,這是用來喚醒 InputReader 線程 if (eventItem.data.fd == mWakeReadPipeFd) { if (eventItem.events & EPOLLIN) { ALOGV("awoken after wake()"); awoken = true; char wakeReadBuffer[16]; ssize_t nRead; do { nRead = read(mWakeReadPipeFd, wakeReadBuffer, sizeof(wakeReadBuffer)); } while ((nRead == -1 && errno == EINTR) || nRead == sizeof(wakeReadBuffer)); } else { ALOGW("Received unexpected epoll event 0x%08x for wake read pipe.", eventItem.events); } continue; } // 接下來是處理設備的輸入事件 Device* device = getDeviceByFdLocked(eventItem.data.fd); if (device == nullptr) { continue; } if (device->videoDevice && eventItem.data.fd == device->videoDevice->getFd()) { // ... } if (eventItem.events & EPOLLIN) { // 讀取輸入事件以及數量 int32_t readSize = read(device->fd, readBuffer, sizeof(struct input_event) * capacity); if (readSize == 0 || (readSize < 0 && errno == ENODEV)) { // Device was removed before INotify noticed. ALOGW("could not get event, removed? (fd: %d size: %" PRId32 " bufferSize: %zu capacity: %zu errno: %d)\n", device->fd, readSize, bufferSize, capacity, errno); deviceChanged = true; closeDeviceLocked(*device); } else if (readSize < 0) { if (errno != EAGAIN && errno != EINTR) { ALOGW("could not get event (errno=%d)", errno); } } else if ((readSize % sizeof(struct input_event)) != 0) { ALOGE("could not get event (wrong size: %d)", readSize); } else { int32_t deviceId = device->id == mBuiltInKeyboardId ? 0 : device->id; // 為每一個輸入事件,填充一個事件 size_t count = size_t(readSize) / sizeof(struct input_event); for (size_t i = 0; i < count; i++) { struct input_event& iev = readBuffer[i]; event->when = processEventTimestamp(iev); event->readTime = systemTime(SYSTEM_TIME_MONOTONIC); event->deviceId = deviceId; event->type = iev.type; event->code = iev.code; event->value = iev.value; event += 1; capacity -= 1; } if (capacity == 0) { // The result buffer is full. Reset the pending event index // so we will try to read the device again on the next iteration. mPendingEventIndex -= 1; break; } } } else if (eventItem.events & EPOLLHUP) { ALOGI("Removing device %s due to epoll hang-up event.", device->identifier.name.c_str()); deviceChanged = true; closeDeviceLocked(*device); } else { ALOGW("Received unexpected epoll event 0x%08x for device %s.", eventItem.events, device->identifier.name.c_str()); } } // 處理設備改變 // mPendingEventIndex >= mPendingEventCount 表示處理完所有的輸入事件後,再處理設備的改變 if (mPendingINotify && mPendingEventIndex >= mPendingEventCount) { mPendingINotify = false; readNotifyLocked(); deviceChanged = true; } // 設備發生改變,那麼跳過當前循環,在下一個循環的開頭處理設備改變 if (deviceChanged) { continue; } // 如果有事件,或者被喚醒,那麼終止循環,接下來 InputReader 會處理事件或者更新配置 if (event != buffer || awoken) { break; } mPendingEventIndex = 0; mLock.unlock(); // release lock before poll // 此時沒有事件,並且也沒有被喚醒,那麼超時等待 epoll 事件 int pollResult = epoll_wait(mEpollFd, mPendingEventItems, EPOLL_MAX_EVENTS, timeoutMillis); mLock.lock(); // reacquire lock after poll if (pollResult == 0) { // 處理超時... } if (pollResult < 0) { // 處理錯誤... } else { // 保存待處理事件的數量 mPendingEventCount = size_t(pollResult); } } // 返回事件的數量 return event - buffer; }
EventHub::getEvent() 提供事件的過程很長,但是現在我們不必去瞭解所有的細節,我們要有從整體看局部的眼光。EventHub 其實隻生成瞭兩類事件
- 設備的添加/刪除事件。這種事件不是通過操作設備而產生的,系統稱之為合成事件。
- 輸入事件。這種事件是通過操作設備產生的,例如手指在觸摸屏上滑動,系統稱之為元輸入事件。
看來我們得分兩部分來分析這兩類事件的生成以及處理過程,因此下一篇文章,我們分析合成事件的生成以及處理過程。
以上就是Input系統之InputReader概要性實例分析的詳細內容,更多關於Input系統InputReader概要性的資料請關註WalkonNet其它相關文章!
推薦閱讀:
- 如何用C寫一個web服務器之I/O多路復用
- 通俗易通講解Android藍牙鍵值適配
- python實現電腦操控安卓手機
- pytorch中.to(device) 和.cuda()的區別說明
- 解析Linux高性能網絡IO和Reactor模型