前文提要
如果发现错误,请在本文指出:https://www.jianshu.com/p/ac7b6a525b96
上一篇文章,跟随着源码深入了剖析了ActivityStack,TaskRecord在Activity启动的过程,怎么选择TaskRecord以及ActivityStack,以及是如何创建TaskRecord以及ActivityStack。当我们,确定的ActivityRecord,确定了TaskRecord,ActivityStack以及ActivityDisplay,已经决定好了ActivityRecord应该放在哪里,接下来就要通过Binder跨进程通信创建Activity。
正文
在我们操作完ActivityStack,TaskRecord在什么时候移动的前端,并且添加完ActivityRecord到TaskRecord之后。我们的数据结构已经处理完毕,还差跨进程启动Activity。但是启动Activity当然不可能凭空诞生,要我们自己编写逻辑,很容易能想到如下规则:
- 1.先暂定当前正在交互的Activity
- 2.检测Activity对应的进程是否存在
- 3.上述两步完成之后,才开始正在跨进程启动Activity。
接下来也是按照如下几步来处理。
resumeTopActivityInnerLocked 启动Activity
文件:/frameworks/base/services/core/java/com/android/server/am/ActivityStack.java
这里我们只关注核心的内容。
private boolean resumeTopActivityInnerLocked(ActivityRecord prev, ActivityOptions options) {
...
final ActivityRecord next = topRunningActivityLocked(true /* focusableOnly */);
final boolean hasRunningActivity = next != null;
// TODO: Maybe this entire condition can get removed?
if (hasRunningActivity && !isAttached()) {
return false;
}
....
boolean lastResumedCanPip = false;
ActivityRecord lastResumed = null;
final ActivityStack lastFocusedStack = mStackSupervisor.getLastStack();
...
final boolean resumeWhilePausing = (next.info.flags & FLAG_RESUME_WHILE_PAUSING) != 0
&& !lastResumedCanPip;
//停止Task内栈内的Activity
boolean pausing = mStackSupervisor.pauseBackStacks(userLeaving, next, false);
if (mResumedActivity != null) {
pausing |= startPausingLocked(userLeaving, false, next, false);
}
//判断到停止成功,则更新进程的lru数据
if (pausing && !resumeWhilePausing) {
...
if (next.app != null && next.app.thread != null) {
mService.updateLruProcessLocked(next.app, true, null);
}
if (lastResumed != null) {
lastResumed.setWillCloseOrEnterPip(true);
}
return true;
} else if (mResumedActivity == next && next.isState(RESUMED)
&& mStackSupervisor.allResumedActivitiesComplete()) {
.....
return true;
}
....
ActivityStack lastStack = mStackSupervisor.getLastStack();
if (next.app != null && next.app.thread != null) {
final boolean lastActivityTranslucent = lastStack != null
&& (lastStack.inMultiWindowMode()
|| (lastStack.mLastPausedActivity != null
&& !lastStack.mLastPausedActivity.fullscreen));
//判断下一个即将即启动的ActivityRecord数据本省存在app等数据,则尝试resume
synchronized(mWindowManager.getWindowManagerLock()) {
if (!next.visible || next.stopped || lastActivityTranslucent) {
next.setVisibility(true);
}
next.startLaunchTickingLocked();
ActivityRecord lastResumedActivity =
lastStack == null ? null :lastStack.mResumedActivity;
final ActivityState lastState = next.getState();
mService.updateCpuStats();
next.setState(RESUMED, "resumeTopActivityInnerLocked");
mService.updateLruProcessLocked(next.app, true, null);
updateLRUListLocked(next);
mService.updateOomAdjLocked();
....
try {
final ClientTransaction transaction = ClientTransaction.obtain(next.app.thread,
next.appToken);
// Deliver all pending results.
ArrayList<ResultInfo> a = next.results;
if (a != null) {
final int N = a.size();
if (!next.finishing && N > 0) {
transaction.addCallback(ActivityResultItem.obtain(a));
}
}
if (next.newIntents != null) {
transaction.addCallback(NewIntentItem.obtain(next.newIntents,
false /* andPause */));
}
next.notifyAppResumed(next.stopped);
next.sleeping = false;
mService.getAppWarningsLocked().onResumeActivity(next);
mService.showAskCompatModeDialogLocked(next);
next.app.pendingUiClean = true;
next.app.forceProcessStateUpTo(mService.mTopProcessState);
next.clearOptionsLocked();
transaction.setLifecycleStateRequest(
ResumeActivityItem.obtain(next.app.repProcState,
mService.isNextTransitionForward()));
mService.getLifecycleManager().scheduleTransaction(transaction);
} catch (Exception e) {
//resume失败,则尝试restart
next.setState(lastState, "resumeTopActivityInnerLocked");
if (lastResumedActivity != null) {
lastResumedActivity.setState(RESUMED, "resumeTopActivityInnerLocked");
}
if (!next.hasBeenLaunched) {
next.hasBeenLaunched = true;
} else if (SHOW_APP_STARTING_PREVIEW && lastStack != null
&& lastStack.isTopStackOnDisplay()) {
next.showStartingWindow(null /* prev */, false /* newTask */,
false /* taskSwitch */);
}
mStackSupervisor.startSpecificActivityLocked(next, true, false);
return true;
}
}
try {
next.completeResumeLocked();
} catch (Exception e) {
requestFinishActivityLocked(next.appToken, Activity.RESULT_CANCELED, null,
"resume-exception", true);
if (DEBUG_STACK) mStackSupervisor.validateTopActivitiesLocked();
return true;
}
} else {
// Whoops, need to restart this activity!
if (!next.hasBeenLaunched) {
next.hasBeenLaunched = true;
} else {
if (SHOW_APP_STARTING_PREVIEW) {
next.showStartingWindow(null /* prev */, false /* newTask */,
false /* taskSwich */);
}
}
mStackSupervisor.startSpecificActivityLocked(next, true, true);
}
...
return true;
}
在这个代码片段中,实际上可以看成在做两个步骤。
- 1.第一个调用Task栈内的所有Activity的onPause方法。
- 2.检测ActivityRecord是否为复用对象,是则回调onResume,不是则准备开始跨进程
调用Task栈内的所有Activity的onPause
在此时这个场景中,ActivityRecord next以及ActivityRecord prev是指同一个对象。因为在上一篇文章就有交代,在addOrReparentStartingActivity中会把ActivityRecord添加到Task的mActivities顶部。
因此此时从方法上传过来的和topRunningActivityLocked获取到的对象实际上都是我们新增加的ActivityRecord。
我们关注这一段代码
boolean pausing = mStackSupervisor.pauseBackStacks(userLeaving, next, false);
if (mResumedActivity != null) {
pausing |= startPausingLocked(userLeaving, false, next, false);
}
让我们坠重一下对应的pauseBackStacks方法。
文件:/frameworks/base/services/core/java/com/android/server/am/ActivityStackSupervisor.java
boolean pauseBackStacks(boolean userLeaving, ActivityRecord resuming, boolean dontWait) {
boolean someActivityPaused = false;
for (int displayNdx = mActivityDisplays.size() - 1; displayNdx >= 0; --displayNdx) {
final ActivityDisplay display = mActivityDisplays.valueAt(displayNdx);
for (int stackNdx = display.getChildCount() - 1; stackNdx >= 0; --stackNdx) {
final ActivityStack stack = display.getChildAt(stackNdx);
if (!isFocusedStack(stack) && stack.getResumedActivity() != null) {
if (DEBUG_STATES) Slog.d(TAG_STATES, "pauseBackStacks: stack=" + stack +
" mResumedActivity=" + stack.getResumedActivity());
someActivityPaused |= stack.startPausingLocked(userLeaving, false, resuming,
dontWait);
}
}
}
return someActivityPaused;
}
能看到,这里面实际上会循环获取ActivityStackSupervisor中的ActivityDisplay中的ActivityStack。去调用每个stack的startPausingLocked方法。能发现,调用Activity的onPause方法,都是通过startPausingLocked这个核心方法。
final boolean startPausingLocked(boolean userLeaving, boolean uiSleeping,
ActivityRecord resuming, boolean pauseImmediately) {
if (mPausingActivity != null) {
if (!shouldSleepActivities()) {
completePauseLocked(false, resuming);
}
}
ActivityRecord prev = mResumedActivity;
if (prev == null) {
if (resuming == null) {
mStackSupervisor.resumeFocusedStackTopActivityLocked();
}
return false;
}
if (prev == resuming) {
return false;
}
mPausingActivity = prev;
mLastPausedActivity = prev;
...
if (prev.app != null && prev.app.thread != null) {
try {
mService.updateUsageStats(prev, false);
mService.getLifecycleManager().scheduleTransaction(prev.app.thread, prev.appToken,
PauseActivityItem.obtain(prev.finishing, userLeaving,
prev.configChangeFlags, pauseImmediately));
} catch (Exception e) {
mPausingActivity = null;
mLastPausedActivity = null;
mLastNoHistoryActivity = null;
}
} else {
mPausingActivity = null;
mLastPausedActivity = null;
mLastNoHistoryActivity = null;
}
if (!uiSleeping && !mService.isSleepingOrShuttingDownLocked()) {
mStackSupervisor.acquireLaunchWakelock();
}
if (mPausingActivity != null) {
if (!uiSleeping) {
prev.pauseKeyDispatchingLocked();
} else if (DEBUG_PAUSE) {
Slog.v(TAG_PAUSE, "Key dispatch not paused for screen off");
}
if (pauseImmediately) {
completePauseLocked(false, resuming);
return false;
} else {
schedulePauseTimeout(prev);
return true;
}
} else {
if (resuming == null) {
mStackSupervisor.resumeFocusedStackTopActivityLocked();
}
return false;
}
}
我们能看到实际上在这段代码十分简单:
1.首先判断当前需要停止的,和需要交互的ActivityRecord是不是同一个。是同一个则说明不需要停止ActivityRecord对应App端的Activity。
2.如果mResumedActivity的app以及app.thread都是不为空,说明该ActivityRecord已经启动过了。那么就能尝试着跨进程调用ActivityRecord对应的Activity的onPause方法。核心方法是下面:
mService.getLifecycleManager().scheduleTransaction(prev.app.thread, prev.appToken,
PauseActivityItem.obtain(prev.finishing, userLeaving,
prev.configChangeFlags, pauseImmediately));
能看到的是,里面有一个关键的对象PauseActivityItem。这个对象是用来跨进程通信,记住这个对象,十分重要,这个方法就是跨进程的核心方法,稍后会回来分析。
- 3.假如停止的mPausingActivity不为空,则按照标志位,来判断是否需要调用resume当前传下来的ActivityRecord。
这里值得注意一下的,此时有一个mResumedActivity对象。这个对象实际上在addOrReparentStartingActivity方法中会执行这么一个步骤addActivityAtIndex,设置完ActivityRecord设置完为止,就会调用setTask方法绑定当前的Task,其中就是这个方法:
文件:/frameworks/base/services/core/java/com/android/server/am/TaskRecord.java
void setTask(TaskRecord task, boolean reparenting) {
// Do nothing if the {@link TaskRecord} is the same as the current {@link getTask}.
if (task != null && task == getTask()) {
return;
}
final ActivityStack oldStack = getStack();//从TaskRecord找到ActivityStack
final ActivityStack newStack = task != null ? task.getStack() : null;
// Inform old stack (if present) of activity removal and new stack (if set) of activity
// addition.
if (oldStack != newStack) {
if (!reparenting && oldStack != null) {
oldStack.onActivityRemovedFromStack(this);
}
if (newStack != null) {
newStack.onActivityAddedToStack(this);
}
}
this.task = task;
if (!reparenting) {
onParentChanged();
}
}
在ActivityRecord绑定TaskRecord的时候,能够发现当ActivityRecord切换了绑定Stack或者当ActivityRecord还没有绑定TaskRecord的时候,都会调用onActivityAddedToStack,设置mResumedActivity。也是精力这个步骤,才能确定在startPausingLocked方法中确定当前正在resume的是哪个Activity,那么pause就只需要停止这个Activity。
检测ActivityRecord是否为复用对象,是则回调onResume,不是则准备开始跨进程
next.app != null && next.app.thread != null
在下一段代码片段中我们首先能看到上述代码段。此时分为两种情况:
- 1.当此时判断到下一个要启动的ActivityRecord,不存在app(ProcessRecord)对象以及app.thread(IApplicationThread象征着ActivityRecord)。而绑定这个对象是在后面的步骤,因此此时就能判断到是一个复用的Activity,直接尝试调用
transaction.setLifecycleStateRequest(
ResumeActivityItem.obtain(next.app.repProcState,
mService.isNextTransitionForward()));
跨进程调用ActivityRecord对应Activity的onResume回调。
- 2.当判断到这两个对象为空,说明是新建的ActivityRecord,还没有和App端进行绑定。因此会调用
mStackSupervisor.startSpecificActivityLocked(next, true, true);
startSpecificActivityLocked检测ProcessRecord
文件:/frameworks/base/services/core/java/com/android/server/am/ActivityStackSupervisor.java
void startSpecificActivityLocked(ActivityRecord r,
boolean andResume, boolean checkConfig) {
ProcessRecord app = mService.getProcessRecordLocked(r.processName,
r.info.applicationInfo.uid, true);
getLaunchTimeTracker().setLaunchTime(r);
if (app != null && app.thread != null) {
try {
if ((r.info.flags&ActivityInfo.FLAG_MULTIPROCESS) == 0
|| !"android".equals(r.info.packageName)) {
app.addPackage(r.info.packageName, r.info.applicationInfo.longVersionCode,
mService.mProcessStats);
}
realStartActivityLocked(r, app, andResume, checkConfig);
return;
} catch (RemoteException e) {
}
}
mService.startProcessLocked(r.processName, r.info.applicationInfo, true, 0,
"activity", r.intent.getComponent(), false, false, true);
}
这一段的代码主要是检测Activity是否存在期待的进程。如果存在就调用realStartActivityLocked,开始跨进程。不存在则调用startProcessLocked,先通过socket联通Zygote,让Zygote孵化一个需要的进程,在重新进行startActivity。这个步骤我在很早就聊过了,这里不再赘述,我们把目光放到realStartActivityLocked跨进程启动Activity。
realStartActivityLocked跨进程启动Activity
在这个方法中,我们只想关注,我们需要关注的。
final boolean realStartActivityLocked(ActivityRecord r, ProcessRecord app,
boolean andResume, boolean checkConfig) throws RemoteException {
if (!allPausedActivitiesComplete()) {
return false;
}
final TaskRecord task = r.getTask();
final ActivityStack stack = task.getStack();
beginDeferResume();
try {
r.startFreezingScreenLocked(app, 0);
r.startLaunchTickingLocked();
r.setProcess(app);
if (getKeyguardController().isKeyguardLocked()) {
r.notifyUnknownVisibilityLaunched();
}
....
int idx = app.activities.indexOf(r);
if (idx < 0) {
app.activities.add(r);
}
mService.updateLruProcessLocked(app, true, null);
mService.updateOomAdjLocked();
....
try {
if (app.thread == null) {
throw new RemoteException();
}
List<ResultInfo> results = null;
List<ReferrerIntent> newIntents = null;
if (andResume) {
results = r.results;
newIntents = r.newIntents;
}
if (r.isActivityTypeHome()) {
// Home process is the root process of the task.
mService.mHomeProcess = task.mActivities.get(0).app;
}
....
// Create activity launch transaction.
final ClientTransaction clientTransaction = ClientTransaction.obtain(app.thread,
r.appToken);
clientTransaction.addCallback(LaunchActivityItem.obtain(new Intent(r.intent),
System.identityHashCode(r), r.info,
// TODO: Have this take the merged configuration instead of separate global
// and override configs.
mergedConfiguration.getGlobalConfiguration(),
mergedConfiguration.getOverrideConfiguration(), r.compat,
r.launchedFromPackage, task.voiceInteractor, app.repProcState, r.icicle,
r.persistentState, results, newIntents, mService.isNextTransitionForward(),
profilerInfo));
// Set desired final state.
final ActivityLifecycleItem lifecycleItem;
if (andResume) {
lifecycleItem = ResumeActivityItem.obtain(mService.isNextTransitionForward());
} else {
lifecycleItem = PauseActivityItem.obtain();
}
clientTransaction.setLifecycleStateRequest(lifecycleItem);
// Schedule transaction.
mService.getLifecycleManager().scheduleTransaction(clientTransaction);
....
} catch (RemoteException e) {
if (r.launchFailed) {
mService.appDiedLocked(app);
stack.requestFinishActivityLocked(r.appToken, Activity.RESULT_CANCELED, null,
"2nd-crash", false);
return false;
}
r.launchFailed = true;
app.activities.remove(r);
throw e;
}
} finally {
endDeferResume();
}
r.launchFailed = false;
if (stack.updateLRUListLocked(r)) {
}
....
return true;
}
在真正的启动之前,AMS会先更新进程的在LRU中的缓存位置,接着会更新应用adj值,更新应用的优先级。做完这些行为之后才开始跨进程启动Activity。我们在这里在一次看到这几个类:ClientTransaction,LaunchActivityItem,ResumeActivityItem等等。
ClientTransactionItem
这几个类实际上是,代表着AMS控制App远程端生命周期抽象成的状态机以及状态。在AMS跨进程控制Activity生命周期中涉及到了如下几个类:
- 1.ClientTransaction 客户端事务控制者
- 2.ClientLifecycleManager 客户端的生命周期事务控制者
- 3.TransactionExecutor 远程通信事务执行者
- 4.LaunchActivityItem 远程App端的onCreate生命周期事务
- 5.ResumeActivityItem 远程App端的onResume生命周期事务
- 6.PauseActivityItem 远程App端的onPause生命周期事务
- 7.StopActivityItem 远程App端的onStop生命周期事务
- 8.DestroyActivityItem 远程App端onDestroy生命周期事务。
- 9.ClientTransactionHandler App端对ClientTransaction的处理。
仅仅是这样列出就能很简单的看出了google工程师对Android生命周期设计上的优化。并且能看到的是,在这里面我们并不能看到七大生命周期中的onStart以及onRestart。实际上,onRestart复用了LaunchActivityItem重新启动Activity,而onStart只是在onCreate之后,App客户端本地调用。
但是这样还不足以弄清楚整个结构什么,我们看看对应的UML图。
从这里我们清楚看到所有生命周期都是继承抽象出来的基类ClientTransactionItem。每当我们尝试着做着跨进程的操作,都会使用ClientTransactionItem这个基类。因此实际上还有ActivityResultItem,NewIntentItem等进行跨进程操作。
稍微理解了其中的类的结构,我们尝试着看看里面源码。我们抽出Activity启动的源码:
final ClientTransaction clientTransaction = ClientTransaction.obtain(app.thread,
r.appToken);
clientTransaction.addCallback(LaunchActivityItem.obtain(new Intent(r.intent),
System.identityHashCode(r), r.info,
// TODO: Have this take the merged configuration instead of separate global
// and override configs.
mergedConfiguration.getGlobalConfiguration(),
mergedConfiguration.getOverrideConfiguration(), r.compat,
r.launchedFromPackage, task.voiceInteractor, app.repProcState, r.icicle,
r.persistentState, results, newIntents, mService.isNextTransitionForward(),
profilerInfo));
final ActivityLifecycleItem lifecycleItem;
if (andResume) {
lifecycleItem = ResumeActivityItem.obtain(mService.isNextTransitionForward());
} else {
...
}
clientTransaction.setLifecycleStateRequest(lifecycleItem);
// Schedule transaction.
mService.getLifecycleManager().scheduleTransaction(clientTransaction);
在这段代码段中,AMS先创建一个LaunchActivityItem作为callback,ResumeActivityItem设置进lifeCycle的请求中。让我们跟踪一下源码scheduleTransaction。
文件:/frameworks/base/services/core/java/com/android/server/am/ClientLifecycleManager.java
void scheduleTransaction(ClientTransaction transaction) throws RemoteException {
final IApplicationThread client = transaction.getClient();
transaction.schedule();
if (!(client instanceof Binder)) {
transaction.recycle();
}
}
文件:/frameworks/base/core/java/android/app/servertransaction/ClientTransaction.java
public void schedule() throws RemoteException {
mClient.scheduleTransaction(this);
}
我们看到此时将会获取IApplicationThread对象,而这个IApplicationThread是一个Binder的存根对象。实现它的ActivityThread的内部类,ApplicationThread。而从上面的UML图,能看到ClientTransaction实现了Parcel,因此能做到跨进程传送该对象。
private class ApplicationThread extends IApplicationThread.Stub
可以清楚的知道此时ApplicationThread是一个本地Binder对象。到了AMS中这个接口就象征着远程代理对象。换句话说,此时会调用App端的scheduleTransaction方法。
实际上在Android高版本中
public final class ActivityThread extends ClientTransactionHandler
ActivityThread将会继承ClientTransactionHandler。就会在基类中处理事件。此时首先会到ApplicationThread中的方法:
文件:/frameworks/base/core/java/android/app/ActivityThread.java
@Override
public void scheduleTransaction(ClientTransaction transaction) throws RemoteException {
ActivityThread.this.scheduleTransaction(transaction);
}
此时我们能看到父类还是通过handler mH处理该方法:
void scheduleTransaction(ClientTransaction transaction) {
transaction.preExecute(this);
sendMessage(ActivityThread.H.EXECUTE_TRANSACTION, transaction);
}
在mH的handlerMessage中专门处理这个片段
case EXECUTE_TRANSACTION:
final ClientTransaction transaction = (ClientTransaction) msg.obj;
mTransactionExecutor.execute(transaction);
if (isSystem()) {
transaction.recycle();
}
break;
TransactionExecutor
此时,App端将会把启动Activity的事务交给TransactionExecutor处理。
文件:/frameworks/base/core/java/android/app/servertransaction/TransactionExecutor.java
public void execute(ClientTransaction transaction) {
final IBinder token = transaction.getActivityToken();
executeCallbacks(transaction);
executeLifecycleState(transaction);
mPendingActions.clear();
}
能看到这里分为两步:
- 1.首先处理传递过来的ClientTransaction中的callback
- 2.接着处理ClientTransaction中的LifecycleStateRequest。
代入当前情景,就是先处理LaunchActivityItem接着处理ResumeActivityItem。也就对应着onCreate以及onResume。实际上这个过程中,google工程师把每一个状态又分为两个步骤去执行,一个是execute,之后会执行postExecute。
我们分别看看这两个步骤分别完成了什么:
executeCallbacks处理LaunchActivityItem
public void executeCallbacks(ClientTransaction transaction) {
final List<ClientTransactionItem> callbacks = transaction.getCallbacks();
if (callbacks == null) {
return;
}
final IBinder token = transaction.getActivityToken();
ActivityClientRecord r = mTransactionHandler.getActivityClient(token);
final ActivityLifecycleItem finalStateRequest = transaction.getLifecycleStateRequest();
final int finalState = finalStateRequest != null ? finalStateRequest.getTargetState()
: UNDEFINED;
// Index of the last callback that requests some post-execution state.
final int lastCallbackRequestingState = lastCallbackRequestingState(transaction);
final int size = callbacks.size();
for (int i = 0; i < size; ++i) {
final ClientTransactionItem item = callbacks.get(i);
....
item.execute(mTransactionHandler, token, mPendingActions);
item.postExecute(mTransactionHandler, token, mPendingActions);
...
}
}
LaunchActivityItem 跨进程通信到ActivityThread
可以看到是,循环处理添加进去的callback,由于此时只有一个对象LaunchActivityItem,会先后执行execute,以及postExecute。
文件:/frameworks/base/core/java/android/app/servertransaction/LaunchActivityItem.java
public void execute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
ActivityClientRecord r = new ActivityClientRecord(token, mIntent, mIdent, mInfo,
mOverrideConfig, mCompatInfo, mReferrer, mVoiceInteractor, mState, mPersistentState,
mPendingResults, mPendingNewIntents, mIsForward,
mProfilerInfo, client);
client.handleLaunchActivity(r, pendingActions, null /* customIntent */);
Trace.traceEnd(TRACE_TAG_ACTIVITY_MANAGER);
}
此时的流程就和Android7.0的源码流程十分相似,client是指ActivityThread,生成一个AMS的ActivityRecord对应的ActivityClientRecord,调用handleLaunchActivity去调用onCreate方法。
到这里,就是我们十分熟悉的handleLaunchActivity方法了,实在不想说了。
private Activity performLaunchActivity(ActivityClientRecord r, Intent customIntent) {
ActivityInfo aInfo = r.activityInfo;
....
ContextImpl appContext = createBaseContextForActivity(r);
Activity activity = null;
try {
java.lang.ClassLoader cl = appContext.getClassLoader();
activity = mInstrumentation.newActivity(
cl, component.getClassName(), r.intent);
...
} catch (Exception e) {
}
try {
Application app = r.packageInfo.makeApplication(false, mInstrumentation);
if (activity != null) {
...
appContext.setOuterContext(activity);
activity.attach(appContext, this, getInstrumentation(), r.token,
r.ident, app, r.intent, r.activityInfo, title, r.parent,
r.embeddedID, r.lastNonConfigurationInstances, config,
r.referrer, r.voiceInteractor, window, r.configCallback);
...
activity.mCalled = false;
if (r.isPersistable()) {
mInstrumentation.callActivityOnCreate(activity, r.state, r.persistentState);
} else {
mInstrumentation.callActivityOnCreate(activity, r.state);
}
if (!activity.mCalled) {
throw new SuperNotCalledException(
"Activity " + r.intent.getComponent().toShortString() +
" did not call through to super.onCreate()");
}
r.activity = activity;
}
r.setState(ON_CREATE);
mActivities.put(r.token, r);
} catch (SuperNotCalledException e) {
throw e;
} catch (Exception e) {
...
}
return activity;
}
能看到这里有三个步骤:
- 反射生成Activity实例
- 2.获取当前的应用的Application对象并且调用attach绑定
- 3.最后通过Instrument调用callActivityOnCreate调用到Activity实例中的onCreate方法。
回到LaunchActivityItem,LaunchActivityItem的postExecute没有实现任何的事件,让我们看看TransactionExecutor的executeLifecycleState。
TransactionExecutor executeLifecycleState控制onResume生命周期
private void executeLifecycleState(ClientTransaction transaction) {
final ActivityLifecycleItem lifecycleItem = transaction.getLifecycleStateRequest();
if (lifecycleItem == null) {
return;
}
final IBinder token = transaction.getActivityToken();
final ActivityClientRecord r = mTransactionHandler.getActivityClient(token);
if (r == null) {
return;
}
cycleToPath(r, lifecycleItem.getTargetState(), true /* excludeLastState */);
lifecycleItem.execute(mTransactionHandler, token, mPendingActions);
lifecycleItem.postExecute(mTransactionHandler, token, mPendingActions);
}
此时调用getLifecycleStateRequest获取的正是之前设置进去的ResumeActivityItem。其思路和LaunchActivityItem十分相似,也是分两步执行,先执行execute,再执行postExecute。
调用onStart
但是请注意这个cycleToPath.
private void cycleToPath(ActivityClientRecord r, int finish,
boolean excludeLastState) {
final int start = r.getLifecycleState();
final IntArray path = mHelper.getLifecyclePath(start, finish, excludeLastState);
performLifecycleSequence(r, path);
}
这个方法最后会调用下面这个方法:
文件:/frameworks/base/core/java/android/app/servertransaction/TransactionExecutorHelper.java
public IntArray getLifecyclePath(int start, int finish, boolean excludeLastState) {
if (start == UNDEFINED || finish == UNDEFINED) {
throw new IllegalArgumentException("Can't resolve lifecycle path for undefined state");
}
if (start == ON_RESTART || finish == ON_RESTART) {
throw new IllegalArgumentException(
"Can't start or finish in intermittent RESTART state");
}
if (finish == PRE_ON_CREATE && start != finish) {
throw new IllegalArgumentException("Can only start in pre-onCreate state");
}
mLifecycleSequence.clear();
if (finish >= start) {
// just go there
for (int i = start + 1; i <= finish; i++) {
mLifecycleSequence.add(i);
}
} else { // finish < start, can't just cycle down
if (start == ON_PAUSE && finish == ON_RESUME) {
....
} else if (start <= ON_STOP && finish >= ON_START) {
....
} else {
...
}
...
return mLifecycleSequence;
}
此时的start是ON_CREATE,finish是ON_RESUME,因此此时会为mLifecycleSequence添加一个中间值ON_START方法。
此时performLifecycleSequence,将会执行ActivityThread中的handleStartActivity方法。
public void handleStartActivity(ActivityClientRecord r,
PendingTransactionActions pendingActions) {
final Activity activity = r.activity;
if (r.activity == null) {
// TODO(lifecycler): What do we do in this case?
return;
}
if (!r.stopped) {
throw new IllegalStateException("Can't start activity that is not stopped.");
}
if (r.activity.mFinished) {
// TODO(lifecycler): How can this happen?
return;
}
// Start
activity.performStart("handleStartActivity");
r.setState(ON_START);
if (pendingActions == null) {
// No more work to do.
return;
}
// Restore instance state
if (pendingActions.shouldRestoreInstanceState()) {
if (r.isPersistable()) {
if (r.state != null || r.persistentState != null) {
mInstrumentation.callActivityOnRestoreInstanceState(activity, r.state,
r.persistentState);
}
} else if (r.state != null) {
mInstrumentation.callActivityOnRestoreInstanceState(activity, r.state);
}
}
....
}
此时将调用Activity的performStart方法。该方法将会调用Fragment的onStart以及通过Instrument调用onStart方法。
ResumeActivityItem execute
@Override
public void execute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
client.handleResumeActivity(token, true /* finalStateRequest */, mIsForward,
"RESUME_ACTIVITY");
}
此时就转移到了ActivityThread的handleResumeActivity方法中,而这个方法又调用了performResumeActivity
public ActivityClientRecord performResumeActivity(IBinder token, boolean finalStateRequest,
String reason) {
final ActivityClientRecord r = mActivities.get(token);
....
try {
r.activity.onStateNotSaved();
r.activity.mFragments.noteStateNotSaved();
checkAndBlockForNetworkAccess();
if (r.pendingIntents != null) {
deliverNewIntents(r, r.pendingIntents);
r.pendingIntents = null;
}
if (r.pendingResults != null) {
deliverResults(r, r.pendingResults, reason);
r.pendingResults = null;
}
r.activity.performResume(r.startsNotResumed, reason);
r.state = null;
r.persistentState = null;
r.setState(ON_RESUME);
} catch (Exception e) {
...
}
return r;
}
我能看到如果pendingIntent的数据不为空则发送一个pendingIntent数据回到onActivityResult,接着调用Activity的onResume方法。
ResumeActivityItem的postExecute
@Override
public void postExecute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
try {
// TODO(lifecycler): Use interface callback instead of AMS.
ActivityManager.getService().activityResumed(token);
} catch (RemoteException ex) {
throw ex.rethrowFromSystemServer();
}
}
此时service是指ActivityManagerService。
public final void activityResumed(IBinder token) {
final long origId = Binder.clearCallingIdentity();
synchronized(this) {
ActivityRecord.activityResumedLocked(token);
mWindowManager.notifyAppResumedFinished(token);
}
Binder.restoreCallingIdentity(origId);
}
能看到的是,此时把ActivityRecord中状态设置为resume状态,把对应的windowManager设置成resume完成状态。
这样子就完成了Activity的onResume方法。还记得我上面写的PauseActivityItem吗?此时代表着pause的生命周期,接下来我们探索一下onPause的流程。
PauseActivityItem 处理execute
回顾一下上面的代码段:
mService.getLifecycleManager().scheduleTransaction(prev.app.thread, prev.appToken,
PauseActivityItem.obtain(prev.finishing, userLeaving,
prev.configChangeFlags, pauseImmediately));
当我们执行启动一个新的Activity时候,将会借助TransactionLifecycleManager其启动跨进程通信,那么原理和上面一致,我们可以直接到PauseActivityItem的execute方法中。
@Override
public void execute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
client.handlePauseActivity(token, mFinished, mUserLeaving, mConfigChanges, pendingActions,
"PAUSE_ACTIVITY_ITEM");
}
此时将会到达Activity的onPause方法。handlePauseActivity最终会达到performPauseActivity
private Bundle performPauseActivity(ActivityClientRecord r, boolean finished, String reason,
PendingTransactionActions pendingActions) {
if (r.paused) {
if (r.activity.mFinished) {
return null;
}
...
}
if (finished) {
r.activity.mFinished = true;
}
final boolean shouldSaveState = !r.activity.mFinished && r.isPreHoneycomb();
if (shouldSaveState) {
callActivityOnSaveInstanceState(r);
}
performPauseActivityIfNeeded(r, reason);
ArrayList<OnActivityPausedListener> listeners;
synchronized (mOnPauseListeners) {
listeners = mOnPauseListeners.remove(r.activity);
}
int size = (listeners != null ? listeners.size() : 0);
for (int i = 0; i < size; i++) {
listeners.get(i).onPaused(r.activity);
}
final Bundle oldState = pendingActions != null ? pendingActions.getOldState() : null;
if (oldState != null) {
if (r.isPreHoneycomb()) {
r.state = oldState;
}
}
return shouldSaveState ? r.state : null;
}
能看到的是,如果此时需要保存当前的状态,将会在onPause中save起来。并且唤醒那些在监听onPause事件的接口。通过performPauseActivityIfNeeded会调到Activity中。并且通知Fragment中所有状态设置为onPause状态。
PauseActivityItem 处理postExecute
@Override
public void postExecute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
if (mDontReport) {
return;
}
try {
ActivityManager.getService().activityPaused(token);
} catch (RemoteException ex) {
throw ex.rethrowFromSystemServer();
}
}
能看到的是此时会回调到AMS中的activityPaused,并且通知ActivityStack调用activityPausedLocked。提一句这个token是IApplicationToken.Stub用来唯一标示对应的ActivityRecord,并且通知对应的WindowManager做后续的处理。
调用完onPause,通知AMS调用onStop方法。
在这个activityPaused方法中,当执行完pause方法之后会执行completePauseLocked方法,执行onStop的一个步骤,之后将会调用Activity的onStop方法。
private void completePauseLocked(boolean resumeNext, ActivityRecord resuming) {
ActivityRecord prev = mPausingActivity;
if (prev != null) {
prev.setWillCloseOrEnterPip(false);
final boolean wasStopping = prev.isState(STOPPING);
prev.setState(PAUSED, "completePausedLocked");
if (prev.finishing) {
prev = finishCurrentActivityLocked(prev, FINISH_AFTER_VISIBLE, false,
"completedPausedLocked");
} else if (prev.app != null) {
if (mStackSupervisor.mActivitiesWaitingForVisibleActivity.remove(prev)) {
}
if (prev.deferRelaunchUntilPaused) {
prev.relaunchActivityLocked(false /* andResume */,
prev.preserveWindowOnDeferredRelaunch);
} else if (wasStopping) {
prev.setState(STOPPING, "completePausedLocked");
} else if (!prev.visible || shouldSleepOrShutDownActivities()) {
prev.setDeferHidingClient(false);
addToStopping(prev, true /* scheduleIdle */, false /* idleDelayed */);
}
} else {
prev = null;
}
...
}
...
}
我们能看到的是,此时一旦判断当前ActivityRecord已经绑定了App端的数据,说明已经启动了,并且当前的ActivityRecord的visible为false,或者点击了锁屏使其睡眠,都会调用addToStopping,把当前的ActivityRecord设置为onStop。具体表现上就是当Activity默认不是window透明的时候,这个标志就会为false。场景经常在Activity的Dialog中。
在addToStopping中,将会透过ActivityStackSupervisorHandler 这个Handler处理onStop方法。
final ActivityRecord activityIdleInternalLocked(final IBinder token, boolean fromTimeout,
boolean processPausingActivities, Configuration config) {
if (DEBUG_ALL) Slog.v(TAG, "Activity idle: " + token);
ArrayList<ActivityRecord> finishes = null;
ArrayList<UserState> startingUsers = null;
int NS = 0;
int NF = 0;
boolean booting = false;
boolean activityRemoved = false;
ActivityRecord r = ActivityRecord.forTokenLocked(token);
...
for (int i = 0; i < NS; i++) {
r = stops.get(i);
final ActivityStack stack = r.getStack();
if (stack != null) {
if (r.finishing) {
stack.finishCurrentActivityLocked(r, ActivityStack.FINISH_IMMEDIATELY, false,
"activityIdleInternalLocked");
} else {
stack.stopActivityLocked(r);
}
}
}
....
return r;
}
在ActivityStack中将会跨进程调用onStop
final void stopActivityLocked(ActivityRecord r) {
...
if (r.app != null && r.app.thread != null) {
adjustFocusedActivityStack(r, "stopActivity");
r.resumeKeyDispatchingLocked();
try {
r.stopped = false;
r.setState(STOPPING, "stopActivityLocked");
if (!r.visible) {
r.setVisible(false);
}
mService.getLifecycleManager().scheduleTransaction(r.app.thread, r.appToken,
StopActivityItem.obtain(r.visible, r.configChangeFlags));
if (shouldSleepOrShutDownActivities()) {
r.setSleeping(true);
}
Message msg = mHandler.obtainMessage(STOP_TIMEOUT_MSG, r);
mHandler.sendMessageDelayed(msg, STOP_TIMEOUT);
} catch (Exception e) {
r.stopped = true;
if (DEBUG_STATES) Slog.v(TAG_STATES, "Stop failed; moving to STOPPED: " + r);
r.setState(STOPPED, "stopActivityLocked");
if (r.deferRelaunchUntilPaused) {
destroyActivityLocked(r, true, "stop-except");
}
}
}
}
在这里面我再一次看到了核心的生命周期的类StopActivityItem。我们看看其execute以及postExecute方法。
StopActivityItem的execute
@Override
public void execute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
client.handleStopActivity(token, mShowWindow, mConfigChanges, pendingActions,
true /* finalStateRequest */, "STOP_ACTIVITY_ITEM");
}
可以看到的是,此时将会执行ActivityThread的handleStopActivity,而它会调用如下方法:
private void performStopActivityInner(ActivityClientRecord r, StopInfo info, boolean keepShown,
boolean saveState, boolean finalStateRequest, String reason) {
if (r != null) {
if (!keepShown && r.stopped) {
if (r.activity.mFinished) {
return;
}
if (!finalStateRequest) {
...
}
}
performPauseActivityIfNeeded(r, reason);
if (info != null) {
try {
info.setDescription(r.activity.onCreateDescription());
} catch (Exception e) {
if (!mInstrumentation.onException(r.activity, e)) {
throw new RuntimeException(
"Unable to save state of activity "
+ r.intent.getComponent().toShortString()
+ ": " + e.toString(), e);
}
}
}
if (!keepShown) {
callActivityOnStop(r, saveState, reason);
}
}
}
当没有进行onPause的时候,将会先惊醒onPause方法,接着再调用,callActivityOnStop调用performStop,调用所有Fragment的onStop以及设置WindowManagerGlobal的状态为暂停。最后通过Instrument调用Activity的onStop方法。
执行AMS的activityStopped,在activityStoppedLocked
里面判断ActivityRecord是否通过makeFinishingLocked设置了finishing为true,从而判断是否需要执行后续的周期。
当调用Activity的finish
很常见的一个情况,就是当Activity调用finish的时候,将会最终会调用onDestory的生命周期,让我们稍微探索源码。
当开发者调用finish的方法最终会调用到AMS中finishActivity方法
public final boolean finishActivity(IBinder token, int resultCode, Intent resultData,
int finishTask) {
....
synchronized(this) {
ActivityRecord r = ActivityRecord.isInStackLocked(token);
if (r == null) {
return true;
}
TaskRecord tr = r.getTask();
ActivityRecord rootR = tr.getRootActivity();
....
final long origId = Binder.clearCallingIdentity();
try {
boolean res;
final boolean finishWithRootActivity =
finishTask == Activity.FINISH_TASK_WITH_ROOT_ACTIVITY;
if (finishTask == Activity.FINISH_TASK_WITH_ACTIVITY
|| (finishWithRootActivity && r == rootR)) {
...
} else {
res = tr.getStack().requestFinishActivityLocked(token, resultCode,
resultData, "app-request", true);
}
return res;
} finally {
Binder.restoreCallingIdentity(origId);
}
}
}
此时我们的finishtask一般不是FINISH_TASK_WITH_ACTIVITY,因此一般会走到下面的分支。此时将会拿到当前taskRecord对应的ActivityStack,开始调用requestFinishActivityLocked。
在requestFinishActivityLocked会通过token(上文介绍的IApplicationToken)找到唯一标示的ActivityRecord。接着调用finishActivityLocked
final boolean finishActivityLocked(ActivityRecord r, int resultCode, Intent resultData,
String reason, boolean oomAdj, boolean pauseImmediately) {
if (r.finishing) {
return false;
}
mWindowManager.deferSurfaceLayout();
try {
r.makeFinishingLocked();
final TaskRecord task = r.getTask();
final ArrayList<ActivityRecord> activities = task.mActivities;
final int index = activities.indexOf(r);
if (index < (activities.size() - 1)) {
task.setFrontOfTask();
if ((r.intent.getFlags() & Intent.FLAG_ACTIVITY_CLEAR_WHEN_TASK_RESET) != 0) {
ActivityRecord next = activities.get(index+1);
next.intent.addFlags(Intent.FLAG_ACTIVITY_CLEAR_WHEN_TASK_RESET);
}
}
//停止事件的分发
r.pauseKeyDispatchingLocked();
//调整焦点ActivityStack
adjustFocusedActivityStack(r, "finishActivity");
finishActivityResultsLocked(r, resultCode, resultData);
final boolean endTask = index <= 0 && !task.isClearingToReuseTask();
final int transit = endTask ? TRANSIT_TASK_CLOSE : TRANSIT_ACTIVITY_CLOSE;
if (mResumedActivity == r) {
....
r.setVisibility(false);
if (mPausingActivity == null) {
startPausingLocked(false, false, null, pauseImmediately);
}
if (endTask) {
mService.getLockTaskController().clearLockedTask(task);
}
} else if (!r.isState(PAUSING)) {
....
final int finishMode = (r.visible || r.nowVisible) ? FINISH_AFTER_VISIBLE
: FINISH_AFTER_PAUSE;
final boolean removedActivity = finishCurrentActivityLocked(r, finishMode, oomAdj,
"finishActivityLocked") == null;
....
return removedActivity;
} else {
}
return false;
} finally {
mWindowManager.continueSurfaceLayout();
}
}
从finishActivityLocked方法中,实际上在finish的时候,按照2个情况处理。当
- 1.当前要finish的Activity刚好就是当前的正在交互的Activity,则调用onPause,在activityIdleInternalLocked调用onStop过程中,实际上还有这么一段代码段,
for (int i = 0; i < NF; i++) {
r = finishes.get(i);
final ActivityStack stack = r.getStack();
if (stack != null) {
activityRemoved |= stack.destroyActivityLocked(r, true, "finish-idle");
}
}
从这里能看到会调用destroyActivityLocked,调用onDestroy方法。
- 2.当finish的Activity不是onPause,尝试调用finishCurrentActivityLocked,finish对应的Activity。
final ActivityRecord finishCurrentActivityLocked(ActivityRecord r, int mode, boolean oomAdj,
String reason) {
final ActivityRecord next = mStackSupervisor.topRunningActivityLocked(
true /* considerKeyguardState */);
if (mode == FINISH_AFTER_VISIBLE && (r.visible || r.nowVisible)
&& next != null && !next.nowVisible) {
if (!mStackSupervisor.mStoppingActivities.contains(r)) {
addToStopping(r, false /* scheduleIdle */, false /* idleDelayed */);
}
if (DEBUG_STATES) Slog.v(TAG_STATES,
"Moving to STOPPING: "+ r + " (finish requested)");
r.setState(STOPPING, "finishCurrentActivityLocked");
if (oomAdj) {
mService.updateOomAdjLocked();
}
return r;
}
// make sure the record is cleaned out of other places.
mStackSupervisor.mStoppingActivities.remove(r);
mStackSupervisor.mGoingToSleepActivities.remove(r);
mStackSupervisor.mActivitiesWaitingForVisibleActivity.remove(r);
final ActivityState prevState = r.getState();
if (DEBUG_STATES) Slog.v(TAG_STATES, "Moving to FINISHING: " + r);
r.setState(FINISHING, "finishCurrentActivityLocked");
final boolean finishingActivityInNonFocusedStack
= r.getStack() != mStackSupervisor.getFocusedStack()
&& prevState == PAUSED && mode == FINISH_AFTER_VISIBLE;
if (mode == FINISH_IMMEDIATELY
|| (prevState == PAUSED
&& (mode == FINISH_AFTER_PAUSE || inPinnedWindowingMode()))
|| finishingActivityInNonFocusedStack
|| prevState == STOPPING
|| prevState == STOPPED
|| prevState == ActivityState.INITIALIZING) {
r.makeFinishingLocked();
boolean activityRemoved = destroyActivityLocked(r, true, "finish-imm:" + reason);
....
return activityRemoved ? null : r;
}
....
return r;
}
能看到,此时将会尝试的调用addToStopping,会调到onStop方法,接着也会调用destroyActivityLocked,进行finish的核心操作。
destroyActivityLocked
final boolean destroyActivityLocked(ActivityRecord r, boolean removeFromApp, String reason) {
if (r.isState(DESTROYING, DESTROYED)) {
return false;
}
EventLog.writeEvent(EventLogTags.AM_DESTROY_ACTIVITY,
r.userId, System.identityHashCode(r),
r.getTask().taskId, r.shortComponentName, reason);
boolean removedFromHistory = false;
cleanUpActivityLocked(r, false, false);
final boolean hadApp = r.app != null;
if (hadApp) {
if (removeFromApp) {
r.app.activities.remove(r);
if (mService.mHeavyWeightProcess == r.app && r.app.activities.size() <= 0) {
mService.mHeavyWeightProcess = null;
mService.mHandler.sendEmptyMessage(
ActivityManagerService.CANCEL_HEAVY_NOTIFICATION_MSG);
}
if (r.app.activities.isEmpty()) {
...
mService.updateLruProcessLocked(r.app, false, null);
mService.updateOomAdjLocked();
}
}
boolean skipDestroy = false;
try {
mService.getLifecycleManager().scheduleTransaction(r.app.thread, r.appToken,
DestroyActivityItem.obtain(r.finishing, r.configChangeFlags));
} catch (Exception e) {
if (r.finishing) {
removeActivityFromHistoryLocked(r, reason + " exceptionInScheduleDestroy");
removedFromHistory = true;
skipDestroy = true;
}
}
r.nowVisible = false;
if (r.finishing && !skipDestroy) {
r.setState(DESTROYING,
"destroyActivityLocked. finishing and not skipping destroy");
Message msg = mHandler.obtainMessage(DESTROY_TIMEOUT_MSG, r);
mHandler.sendMessageDelayed(msg, DESTROY_TIMEOUT);
} else {
r.setState(DESTROYED,
"destroyActivityLocked. not finishing or skipping destroy");
r.app = null;
}
} else {
if (r.finishing) {
removeActivityFromHistoryLocked(r, reason + " hadNoApp");
removedFromHistory = true;
} else {
r.setState(DESTROYED, "destroyActivityLocked. not finishing and had no app");
r.app = null;
}
}
r.configChangeFlags = 0;
return removedFromHistory;
}
我能够看到熟悉DestroyActivityItem这个代表着调用Activity的onResume的类。通过这个类做完跨进程通信之后,调用removeActivityFromHistoryLocked,清除TaskRecord中ActivityRecord,清除ActivityRecord中的Window对象,ProcessRecord等对象。
这里一样看看DestroyActivityItem的execute方法。
DestroyActivityItem execute
public void execute(ClientTransactionHandler client, IBinder token,
PendingTransactionActions pendingActions) {
client.handleDestroyActivity(token, mFinished, mConfigChanges,
false /* getNonConfigInstance */, "DestroyActivityItem");
}
此时将调用ActivityThread的handleDestroyActivity。
public void handleDestroyActivity(IBinder token, boolean finishing, int configChanges,
boolean getNonConfigInstance, String reason) {
ActivityClientRecord r = performDestroyActivity(token, finishing,
configChanges, getNonConfigInstance, reason);
if (r != null) {
cleanUpPendingRemoveWindows(r, finishing);
WindowManager wm = r.activity.getWindowManager();
View v = r.activity.mDecor;
if (v != null) {
if (r.activity.mVisibleFromServer) {
mNumVisibleActivities--;
}
IBinder wtoken = v.getWindowToken();
if (r.activity.mWindowAdded) {
if (r.mPreserveWindow) {
r.mPendingRemoveWindow = r.window;
r.mPendingRemoveWindowManager = wm;
r.window.clearContentView();
} else {
wm.removeViewImmediate(v);
}
}
if (wtoken != null && r.mPendingRemoveWindow == null) {
WindowManagerGlobal.getInstance().closeAll(wtoken,
r.activity.getClass().getName(), "Activity");
} else if (r.mPendingRemoveWindow != null) {
WindowManagerGlobal.getInstance().closeAllExceptView(token, v,
r.activity.getClass().getName(), "Activity");
}
r.activity.mDecor = null;
}
if (r.mPendingRemoveWindow == null) {
WindowManagerGlobal.getInstance().closeAll(token,
r.activity.getClass().getName(), "Activity");
}
Context c = r.activity.getBaseContext();
if (c instanceof ContextImpl) {
((ContextImpl) c).scheduleFinalCleanup(
r.activity.getClass().getName(), "Activity");
}
}
if (finishing) {
try {
ActivityManager.getService().activityDestroyed(token);
} catch (RemoteException ex) {
throw ex.rethrowFromSystemServer();
}
}
mSomeActivitiesChanged = true;
}
能看到当我们通过performDestroyActivity调用Activity的OnDestroy之后,将会清空Activity中设置的window数据以及设置的ContentView,最后通过activityDestroyed通知AMS。
最后通信到AMS中,发送一个Handler消息,10秒后把ActivityRecord从TaskRecord 中移除,如果此时TaskRecord已经不存在ActivityRecord,则从ActivityStack移除。
performDestroyActivity
ActivityClientRecord performDestroyActivity(IBinder token, boolean finishing,
int configChanges, boolean getNonConfigInstance, String reason) {
ActivityClientRecord r = mActivities.get(token);
Class<? extends Activity> activityClass = null;
if (localLOGV) Slog.v(TAG, "Performing finish of " + r);
if (r != null) {
activityClass = r.activity.getClass();
r.activity.mConfigChangeFlags |= configChanges;
if (finishing) {
r.activity.mFinished = true;
}
performPauseActivityIfNeeded(r, "destroy");
if (!r.stopped) {
callActivityOnStop(r, false /* saveState */, "destroy");
}
if (getNonConfigInstance) {
try {
r.lastNonConfigurationInstances
= r.activity.retainNonConfigurationInstances();
} catch (Exception e) {
if (!mInstrumentation.onException(r.activity, e)) {
throw new RuntimeException(
"Unable to retain activity "
+ r.intent.getComponent().toShortString()
+ ": " + e.toString(), e);
}
}
}
try {
r.activity.mCalled = false;
mInstrumentation.callActivityOnDestroy(r.activity);
if (!r.activity.mCalled) {
throw new SuperNotCalledException(
"Activity " + safeToComponentShortString(r.intent) +
" did not call through to super.onDestroy()");
}
if (r.window != null) {
r.window.closeAllPanels();
}
} catch (SuperNotCalledException e) {
throw e;
} catch (Exception e) {
if (!mInstrumentation.onException(r.activity, e)) {
throw new RuntimeException(
"Unable to destroy activity " + safeToComponentShortString(r.intent)
+ ": " + e.toString(), e);
}
}
r.setState(ON_DESTROY);
}
mActivities.remove(token);
StrictMode.decrementExpectedActivityCount(activityClass);
return r;
}
能看到的是,此时当没有调用过onPause以及onStop将会调用一次,但是实际上这种情况不可能出现,都在上面的对应的PauseActivityItem以及StopActivityItem中调用过一次了。最后通Instrument调用callActivityOnDestroy,回调到Activity的performDestroy,最后关闭Activity的键盘,清空在ActivityThread的缓存。
在performDestroy调用Fragment的onDestroy,以及回调Activity的onDestroy。
这样就完成Activity的onDestroy流程。
onRestart
肯定有人觉得奇怪,七大声明周期之一的onRestart呢?他其实和onStart一样隐藏在TransactionExecutor 中。
来看看/frameworks/base/core/java/android/app/servertransaction/TransactionExecutorHelper.java
:
public IntArray getLifecyclePath(int start, int finish, boolean excludeLastState) {
if (start == UNDEFINED || finish == UNDEFINED) {
throw new IllegalArgumentException("Can't resolve lifecycle path for undefined state");
}
if (start == ON_RESTART || finish == ON_RESTART) {
throw new IllegalArgumentException(
"Can't start or finish in intermittent RESTART state");
}
if (finish == PRE_ON_CREATE && start != finish) {
throw new IllegalArgumentException("Can only start in pre-onCreate state");
}
mLifecycleSequence.clear();
if (finish >= start) {
// just go there
...
} else { // finish < start, can't just cycle down
if (start == ON_PAUSE && finish == ON_RESUME) {
// Special case when we can just directly go to resumed state.
mLifecycleSequence.add(ON_RESUME);
} else if (start <= ON_STOP && finish >= ON_START) {
// Restart and go to required state.
// Go to stopped state first.
for (int i = start + 1; i <= ON_STOP; i++) {
mLifecycleSequence.add(i);
}
// Restart
mLifecycleSequence.add(ON_RESTART);
// Go to required state
for (int i = ON_START; i <= finish; i++) {
mLifecycleSequence.add(i);
}
} else {
// Relaunch and go to required state
// Go to destroyed state first.
for (int i = start + 1; i <= ON_DESTROY; i++) {
mLifecycleSequence.add(i);
}
// Go to required state
for (int i = ON_CREATE; i <= finish; i++) {
mLifecycleSequence.add(i);
}
}
}
...
return mLifecycleSequence;
}
注意这里的参数start是指当前Activity的状态,finish是指经过TransactionExecutor执行后,每一个ActivityLifecycleItem
对应的目标Activity需要达到什么声明周期。
1.如果当前的Activity是
ON_PAUSE
状态,目标是ON_RESUME
,此时只需要执行一个ON_RESUME
2.如果此时的状态是
ON_STOP
之后的状态,且目标是ON_START
.一般来说此时都是执行的是ResumeActivityItem
需要从AMS让此时的Activity转化为可见。此时的Activity
已经执行了onStop,就会把小于ON_STOP
的状态添加进来(没有就跳过了),再把ON_RESTART
声明周期添加进来,最后把onStart
和onResume
(因为ResumeActivityItem
目标就是onResume
)添加进来3.最后到达了TransactionExecutor中执行每一个
ActivityLifecycleItem
的生命周期,从而执行了ActivityThread的onRestart
后执行,onStart
,onResume
总结
本文着重说了Activity的生命周期的流程,所有的生命周期流程,就不是直接通过Binder跨进程通信,而是每一个需要通行都Activity处理的事务都抽象成了一个ClientTransactionItem处理,并且交由ClientTransaction统一分发处理。每一个ClientTransactionItem,都会把执行分为两步,execute以及postExecute。
可以说,Android对Activity的生命周期的思考从来没有停过,通过抽象以及实现Parcel灵活的实现了生命周期状态机的管理。
阅读了本文之后,就能知道我之前写的插件化,没有办法直接hook Android O的系统,因为在插件话基础模型中,有一个关键的步骤,就是hook ActivityThread的mH中的handlermessage方法。然而此时mH将不处理生命周期,因此会出现没办法偷梁换柱的情况的。但是处理的方式十分简单,就是hook handlerMessage中的处理事务的信号即可。