Read the Fucking Source Code
上周在Fabric的Crashlytics榜首出现了一个关于Fragment状态引发的崩溃日志:
android.app.FragmentManagerImpl.checkStateLoss (FragmentManager.java:1858)
android.app.FragmentManagerImpl.enqueueAction (FragmentManager.java:1881)
android.app.BackStackRecord.commitInternal (BackStackRecord.java:688)
android.app.BackStackRecord.commit (BackStackRecord.java:646)
android.app.DialogFragment.show (DialogFragment.java:230)
刚开始以为是DialogFragment调用show的时候,没有判断Activity是否已经Finish了。结果发现在项目代码里是做了判断的,看崩溃的人数和次数还是一个非必现的bug,瞬间就感觉事情可能不是那么简单了,只能硬着头皮去RTFSC了....
这次我们反向追踪(就是这么皮)
FragmentManagerImpl.checkStateLoss
private void checkStateLoss() {
//异常抛出的地方
if (mStateSaved) {
throw new IllegalStateException(
"Can not perform this action after onSaveInstanceState");
}
if (mNoTransactionsBecause != null) {
throw new IllegalStateException(
"Can not perform this action inside of " + mNoTransactionsBecause);
}
}
FragmentManagerImpl. enqueueAction
/**
* Adds an action to the queue of pending actions.
*
* @param action the action to add
* @param allowStateLoss whether to allow loss of state information
* @throws IllegalStateException if the activity has been destroyed
*/
public void enqueueAction(OpGenerator action, boolean allowStateLoss) {
//注意这里的allowStateLoss 它决定了checkStateLoss方法是否进行
if (!allowStateLoss) {
checkStateLoss();
}
synchronized (this) {
if (mDestroyed || mHost == null) {
throw new IllegalStateException("Activity has been destroyed");
}
if (mPendingActions == null) {
mPendingActions = new ArrayList<>();
}
mPendingActions.add(action);
scheduleCommit();
}
}
BackStackRecord.commitInternal
int commitInternal(boolean allowStateLoss) {
if (mCommitted) throw new IllegalStateException("commit already called");
if (FragmentManagerImpl.DEBUG) {
Log.v(TAG, "Commit: " + this);
LogWriter logw = new LogWriter(TAG);
PrintWriter pw = new PrintWriter(logw);
dump(" ", null, pw, null);
pw.close();
}
mCommitted = true;
if (mAddToBackStack) {
mIndex = mManager.allocBackStackIndex(this);
} else {
mIndex = -1;
}
//重点
mManager.enqueueAction(this, allowStateLoss);
return mIndex;
}
BackStackRecord.commit
public int commit() {
//这里传入的就是allowStateLoss参数
return commitInternal(false);
}
DialogFragment.show
public void show(FragmentManager manager, String tag) {
mDismissed = false;
mShownByMe = true;
FragmentTransaction ft = manager.beginTransaction();
ft.add(this, tag);
//没想到的时候DialogFragment的show方法会调用commit() 导火索
ft.commit();
}
目前来看是因为DialogFragment的show方法导致强行检查Fragment的状态,而恰好在检查的时候状态已经被保存导致FragmentManager的标记位mStateSaved = true,那么去查mStateSaved = true的调用时机一切都会水落石出了..
---------------------------------------查找中-------------------------------------------
FragmentManager中mStateSaved = true总共出现了三次,最后确定了其中一条路线是本次崩溃走过的路...
---------------------------------------复现中-------------------------------------------
情景再现
当用户在进入首页模块时直接按Home键或者遇到内存不足等特殊情况下触发了onSaveInstanceState导致FragmentManager进行“智能”保存当前Activity中Fragment的状态,因为DialogFragment(广告弹窗)需要先从服务器拉取数据,所以之后在show的时候去checkStateLoss时触发了异常...
解决方案:
1.将广告弹窗实现改为纯Dialog,不再使用DialogFragment.
2.如果自己的项目里面不需要对Fragment的相关状态进行保存和维护,可以在相关Activity中复写onSaveInstance不进行super回调即可.
网友评论