The recommended substitute for clearTask of popUpTo with popUpToInclusive only works from start destination actions [116831650]

Fixed

Bug

Status Update

No update yet.

Description

[Deleted User]

created issue #1

Sep 28, 2018 10:32AM

I was previously using clearTask in my navigation actions to make sure that the user could not navigate back to the previous destination (useful for splash screens and login screens). After it was deprecated I decided to try the recommended approach of instead setting popUpTo pointing to the root of the graph together with setting popUpToInclusive="true", but this didn't have the effect I wanted.

This approach only seem to work on actions originating from the graph's start destination, and fails on actions between any other destinations. I created a sample project which reproduces the issue:

https://github.com/hsson/android-navigation-component-alpha06-bug

Here is the project's README detailing the issue:

# Bug in Android Navigation Component (alpha06)
(Note: I have only verified this in alpha06, I don't know if it exists in
previous versions)

## The problem
Basically the problem is that the NavOption "clearTask" was removed in
alpha02, and the recommended substitute was to use app:popUpTo pointing to
the root of the graph/the graph ID, together with app:popUpToInclusive="true".
However, this method does not work. It only seems to work for actions
going from the graph's start destination. If we have the following graph:

x y
A ---> B ---> C

Where A, B, and C are fragments, A is the start destination of the grapg,
and x, y are actions going from A to B and B to C respectively. Let's say
the graph's ID is "@+id/graph"

Then, if both actions x and y have
app:popUpTo="@+id/graph"
app:popUpToInclusive="true"

Then when navigating from A to B through x, and then hitting the back button
will exit the app (as expected). But if navigating from A to B to C through
x and y, and then hitting the back button we will navigate back to B — this
is not what was expected. The expected behavior with this graph is to always
exit the app when clicking the back button

## Environment
This was verified using:
Android Studio 3.2 RC 3
Build #AI-181.5540.7.32.4987877, built on August 31, 2018
JRE: 1.8.0_152-release-1136-b06 x86_64
JVM: OpenJDK 64-Bit Server VM by JetBrains s.r.o
macOS 10.13.6

Comments

[Deleted User] <[Deleted User]> #2Oct 1, 2018 08:14AM

Bringing setMaxSchedulerLimit to say 49 doesn't help. Same error.

il...@google.com <il...@google.com> Oct 1, 2018 10:56PM

Assigned to il...@google.com.

[Deleted User] <[Deleted User]> #3Oct 10, 2018 09:20AM

If you are using work-firebase then you should not be using JobScheduler directly. Can you please send us the full bug report so we can investigate ?

il...@google.com <il...@google.com> #4Oct 18, 2018 06:05PM

java.lang.IllegalStateException: Apps may not schedule more than 100 distinct jobs
at android.os.Parcel.readException(Parcel.java:1692)
at android.os.Parcel.readException(Parcel.java:1637)
at android.app.job.IJobScheduler$Stub$Proxy.schedule(IJobScheduler.java:158)
at android.app.JobSchedulerImpl.schedule(JobSchedulerImpl.java:42)
at androidx.work.impl.background.systemjob.SystemJobScheduler.scheduleInternal(SystemJobScheduler.java:126)
at androidx.work.impl.background.systemjob.SystemJobScheduler.schedule(SystemJobScheduler.java:95)
at androidx.work.impl.Schedulers.scheduleInternal(Schedulers.java:104)
at androidx.work.impl.Schedulers.schedule(Schedulers.java:73)
at androidx.work.impl.WorkerWrapper.setSucceededAndNotify(WorkerWrapper.java:393)
at androidx.work.impl.WorkerWrapper.handleResult(WorkerWrapper.java:255)
at androidx.work.impl.WorkerWrapper.run(WorkerWrapper.java:181)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1133)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:607)
at java.lang.Thread.run(Thread.java:761)

[Deleted User] <[Deleted User]> #5Oct 19, 2018 11:44AM

What do you mean I shouldn't use JobScheduler directly? Isn't work-firebase is just for <21 devices, otherwise JobScheduler always used?

mi...@gmail.com <mi...@gmail.com> #6Oct 19, 2018 12:22PM

I think i was confused by your previous comment. You are right, we only use work-firebase for API < 23.
JobScheduler is used for API >= 23.

This issue should have been resolved in `alpha03` and `alpha04` (we addressed different parts of the problem). If you can reproduce this on a device, then can you do:

1. adb shell dumpsys jobscheduler for your app.
2. Reboot your device.
3. After the reboot, can you re-run adb shell dumpsys jobscheduler ?

I want to check if you see any differences in the jobs created before and after the reboot.

Thanks for your patience.

[Deleted User] <[Deleted User]> #7Oct 19, 2018 12:24PM

Actually for step 2, the easier thing to do is to force stop your app. You don't need to fo a full reboot.

il...@google.com <il...@google.com> #8Oct 19, 2018 03:36PM

I don't have to to rummage through traces. Here's code snippet that sends app into a crash loop.

https://pastebin.com/NcTAnwjZ

mi...@gmail.com <mi...@gmail.com> #9Oct 19, 2018 03:50PM

I tried to reproduce this on alpha04. However, I cannot reproduce your issue.

For reference, this is what my activity looks like:

class MainActivity : AppCompatActivity() {
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_main)
findViewById<Button>(R.id.scheduler_limits_testcase1).setOnClickListener {
for (i in 1..100) {
val a1 = newWorkRequest("--- A1 ---")
val b1 = newWorkRequest("--- B1 ---")
val b2 = newWorkRequest("--- B2 ---")
val c1 = newWorkRequest("--- C1 ---")
WorkManager.getInstance()!!
.beginWith(a1).then(b1).enqueue()
WorkManager.getInstance()!!
.beginWith(a1).then(b2).enqueue()
WorkManager.getInstance()!!
.beginWith(b1, b2).then(c1).enqueue()
}

}
}

companion object {
private fun newWorkRequest(message: String): OneTimeWorkRequest {
val data = mapOf(TestWorker.KEY_MESSAGE to message).toWorkData()
return OneTimeWorkRequestBuilder<TestWorker>()
.setInputData(data)
.build()
}
}
}

This is what my Worker looks like:

class TestWorker : Worker() {
companion object {
private const val TAG = "TestWorker"
private const val MESSAGE_DEFAULT = "Default"
const val KEY_MESSAGE = "KEY_MESSAGE"
}

override fun doWork(): Result {
val message = inputData.getString(KEY_MESSAGE, MESSAGE_DEFAULT)
val random = Random()
val time = 500L + random.nextInt(300)
Log.d(TAG, "Before. $message")
Thread.sleep(time)
Log.d(TAG, "After. $message")
return Result.SUCCESS
}
}

il...@google.com <il...@google.com> #10Oct 19, 2018 04:13PM

Do you have another part of your codebase that is scheduling jobs with JobScheduler?

mi...@gmail.com <mi...@gmail.com> #11Oct 19, 2018 04:16PM

We have 1 job that schedules once on app start, that's it.

dc...@gmail.com <dc...@gmail.com> #12Oct 23, 2018 05:19PM

I created a sample app.

For some reason it rarely crashes on AVD Emulator(Api 27), but 100% crashes on Genymotion Api 26 and my Xiaomi Redmi 7.1.2

You can try both Activities.

da...@gmail.com <da...@gmail.com> #13Oct 26, 2018 10:58PM

Hi, I've tried numerous times and have been unable to get your example to exhibit this problem. Could you please update to alpha05 and reproduce it? If you can reliably reproduce this issue, could you please do so with verbose logging enabled in your Configuration object (introduced in alpha05) and send us a bugreport?

il...@google.com <il...@google.com> #14Nov 30, 2018 11:21PM

Marked as fixed.

I will some time later when I have time