STICKY SystemForegroundService crashes when targeting API 34 and location permissions are removed [333957914]

In Progress

Bug

Status Update

No update yet.

Description

st...@deliveroo.co.uk

created issue #1

Apr 12, 2024 03:47PM

Component used: WorkManager, CoroutineWorker
Version used: 2.9.0
Devices/Android versions reproduced on: API 34

When we target API 34, WorkManager is crashing our application on startup if the "location" permissions are removed whilst the application is running.

Everything works fine whilst the application has the "location" permissions. When the "location" permissions are removed (whilst the application is running), the OS is killing the application (OS behaviour) and because SystemForegroundService is a STICKY service, the OS tries to restart the service which causes a security exception because the "location" permissions are no longer present.

This is a realistic user scenario which will happen many times in our production app. We can not prevent our application from crashing as this relates to the STICKY SystemForegroundService. Our application, can check the "location" permissions before scheduling the "worker", but we have no control over the user removing the "location" permissions after the "worker" has been scheduled by the WorkManager.

This behaviour only happens with WorkManager when targeting API 34 as per the restrictions documented here:

https://developer.android.com/about/versions/14/changes/fgs-types-required

Here is the crash we are seeing on application startup:

Java.lang.SecurityException: Starting FGS with type location callerApp=ProcessRecord{34bbf76 32120:com.deliveroo.driverapp.test/u0a190} targetSDK=34 requires permissions: all of the permissions allOf=true [android.permission.FOREGROUND_SERVICE_LOCATION] any of the permissions allOf=false [android.permission.ACCESS_COARSE_LOCATION, android.permission.ACCESS_FINE_LOCATION] and the app must be in the eligible state/exemptions to access the foreground only permission

We are using the WorkManager to run a CoroutineWorker which accesses the device location. We are scheduling the Worker as follows:

val request = OneTimeWorkRequest.Builder(worker.java)
.setExpedited(OutOfQuotaPolicy.RUN_AS_NON_EXPEDITED_WORK_REQUEST)
.build()
WorkManager.getInstance(context).enqueueUniqueWork(

worker.java.name,
ExistingWorkPolicy.KEEP,
request,
)

We are specifying FOREGROUND_SERVICE_TYPE_LOCATION in the ForegroundInfo:

ForegroundInfo(ONLINE_NOTIFICATION_ID, notificationsManager.onlineNotification, ServiceInfo.FOREGROUND_SERVICE_TYPE_LOCATION)

Our Android manifest specifies the "location" foregroundServiceType:

<service
android:name="androidx.work.impl.foreground.SystemForegroundService"
android:foregroundServiceType="location"/>

As stated above, everything works fine until the user removes the "location" permissions after the "worker" has started. Then the OS kills our application and the OS tries to recover the STICKY SystemForegroundService which crashes because the app no longer as "location" permissions.

Please can you advise how we can avoid WorkManager crashing our application in the above scenario.

Comments

ra...@google.com <ra...@google.com> #2Apr 15, 2024 05:19PM

In SystemJobService, it can expect null, and able to tolerance null from WorkManagerImpl.getInstance, due to auto-backup reason.

@Override
public void onCreate() {
super.onCreate();
mWorkManagerImpl = WorkManagerImpl.getInstance(getApplicationContext());
if (mWorkManagerImpl == null) {
// This can occur if...
// 1. The app is performing an auto-backup. Prior to O, JobScheduler could erroneously
// try to send commands to JobService in this state (

b/32180780

). Since neither
// Application#onCreate nor ContentProviders have run, WorkManager won't be
// initialized. In this case, we should ignore all JobScheduler commands and tell it
// to retry.
But, current WorkManagerImpl.getInstance(Context context) implementation is that, it will never return null but throw IllegalStateException.

Isn't there's conflict among these 2 SystemJobService.onCreate logic, and WorkManagerImpl.getInstance logic?

st...@deliveroo.co.uk <st...@deliveroo.co.uk> #3Apr 16, 2024 09:53AM

Project: platform/frameworks/support
Branch: androidx-master-dev

commit e14104eb7008f82edd50b37e16bd119b307e0554
Author: Sumir Kataria <sumir@google.com>
Date: Mon Jun 24 10:50:57 2019

Fix for JobScheduler calls when backup is in progress.

The original fix (ag/Ia1604b7c4b98f6b31869e1161a0c1efce7fe46f2) relied
on a null-check, but the new version of WorkManager.getInstance(Context)
with the Context param doesn't return null; it throws an Exception. This
CL updates the three places where we explicitly check for the return value
internally:

1. SystemJobService (the main place where this would manifest)
2. RescheduleReceiver
3. WorkManagerTestInitHelper (for consistency)

Fixes: 135858602
Test: Ran existing tests
Change-Id: Ia232869712d0c78fd80e91d0a863b8c50a080400

M work/workmanager-testing/src/main/java/androidx/work/testing/WorkManagerTestInitHelper.java
M work/workmanager/src/main/java/androidx/work/impl/background/systemalarm/RescheduleReceiver.java
M work/workmanager/src/main/java/androidx/work/impl/background/systemjob/SystemJobService.java

https://android-review.googlesource.com/988026

https://goto.google.com/android-sha1/e14104eb7008f82edd50b37e16bd119b307e0554

da...@google.com <da...@google.com> #4Apr 16, 2024 01:59PM

Thanks you so much for the speedy fixed!

May I know when will it be ready for production? As, currently, this IllegalStateException is impacting our users.

ra...@google.com <ra...@google.com> #5Apr 16, 2024 05:56PM

Accepted by ra...@google.com.

Hello, please look for this later in the week. If you want to work around this bug specifically and don't need on-demand initialization, you can use the old (deprecated) WorkManager.getInstance() method for now.

Message last modified on Apr 16, 2024 06:22PM

st...@deliveroo.co.uk <st...@deliveroo.co.uk> #6Apr 17, 2024 03:25PM

Hello. Thanks for respond. I surely will keep a pair of eye on the next release :)

Yes. We don't need on-demand initialization. May I know how does using old (deprecated) WorkManager.getInstance() in our app code, able to bypass this problem? Isn't the crash is origin from library internal SystemJobService?

We once thinking to bypass this problem, by switching back to Version 2.0.1 (The version just before WorkManager.getInstance(Context) is introduced). However, we notice Room version is being changed in between. Downgrading WorkManager version, might cause additional SQLite problem.

st...@deliveroo.co.uk <st...@deliveroo.co.uk> #7Jul 15, 2024 09:42AM

Sorry, you are correct that moving to the deprecated method will not help you here. I would advise waiting for the release later this week. And thank you very much for finding this bug and doing the research to find the underlying issue.

I am curious: do you actually have reports of this bug on Android 9? It should've been fixed in the platform layer at that point. If possible, can you share any Android 9 specific logs/bugreports/stacktraces with device models?

va...@gmail.com <va...@gmail.com> #8Jul 19, 2024 12:50PM

OK. Finger crossed users will not leave our app. I will follow the Green Dinosaur in Twitter, to wait for his announcement.

Based on your code comment, I thought this problem shouldn't occur in new Android but it happens some how. This is the log

Asus ZenFone 5 (ZE620KL) (ASUS_X00QD), Android 9 (Happen 2 times)
Asus ZenFone 5Z (ZS620KL/ZS621KL) (ASUS_Z01R_1), Android 9 (Happen 1 time)

3 incidents are having same log trace as below

java.lang.RuntimeException:
at android.app.ActivityThread.handleCreateService (ActivityThread.java:3570)
at android.app.ActivityThread.access$1300 (ActivityThread.java:200)
at android.app.ActivityThread$H.handleMessage (ActivityThread.java:1672)
at android.os.Handler.dispatchMessage (Handler.java:106)
at android.os.Looper.loop (Looper.java:198)
at android.app.ActivityThread.main (ActivityThread.java:6732)
at java.lang.reflect.Method.invoke (Method.java)
at com.android.internal.os.RuntimeInit$MethodAndArgsCaller.run (RuntimeInit.java:493)
at com.android.internal.os.ZygoteInit.main (ZygoteInit.java:858)
Caused by: java.lang.IllegalStateException:
at androidx.work.impl.WorkManagerImpl.getInstance (WorkManagerImpl.java:142)
at androidx.work.impl.background.systemjob.SystemJobService.onCreate (SystemJobService.java:53)
at android.app.ActivityThread.handleCreateService (ActivityThread.java:3558)
at android.app.ActivityThread.access$1300 (ActivityThread.java:200)
at android.app.ActivityThread$H.handleMessage (ActivityThread.java:1672)
at android.os.Handler.dispatchMessage (Handler.java:106)
at android.os.Looper.loop (Looper.java:198)
at android.app.ActivityThread.main (ActivityThread.java:6732)
at java.lang.reflect.Method.invoke (Method.java)
at com.android.internal.os.RuntimeInit$MethodAndArgsCaller.run (RuntimeInit.java:493)
at com.android.internal.os.ZygoteInit.main (ZygoteInit.java:858)

I attach the screenshot of Google Play Console as well. (I expect the stack trace should print programmer's message side-by-side with IllegalStateException. But it doesn't)

Feel free to let me know if you need any information from my side. Happy to help.

Message last modified on Jul 19, 2024 12:51PM

da...@google.com <da...@google.com> #9Jul 24, 2024 02:01PM

Thanks, we'll investigate this internally.

ap...@google.com <ap...@google.com> #10Jul 25, 2024 12:03AM

Hello, wondering whether the fix will be released in another one, or two days. As, users start to complain and leaving our app, due to increased crash rate :(

Thanks.

ap...@google.com <ap...@google.com> #11Jul 25, 2024 12:03AM

Hi, we're still aiming to get this out this week.

rw...@gmail.com <rw...@gmail.com> #12Sep 24, 2024 06:23AM

still have bug:

trace:
5 java.lang.IllegalStateException:WorkManager needs to be initialized via a ContentProvider#onCreate() or an Application#onCreate().
6 androidx.work.impl.background.systemjob.SystemJobService.onCreate(SystemJobService.java:68)
7 android.app.ActivityThread.handleCreateService(ActivityThread.java:3133)
8 android.app.ActivityThread.access$1900(ActivityThread.java:196)
9 android.app.ActivityThread$H.handleMessage(ActivityThread.java:1622)
10 com.android.tony.defenselib.hook.HookHandle$1.handleMessage(HookHandle.java:30)
11 android.os.Handler.dispatchMessage(Handler.java:107)
12 android.os.Looper.loop(Looper.java:210)
13 android.app.ActivityThread.main(ActivityThread.java:5982)
14 java.lang.reflect.Method.invoke(Native Method)
15 com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:852)
16 com.android.internal.os.ZygoteInit.main(ZygoteInit.java:742)

this code cause crash:

if (!Application.class.equals(getApplication().getClass())) {
// During auto-backup, we don't get a custom Application subclass. This code path
// indicates we are either performing auto-backup or the user never used a custom
// Application class (or both).
throw new IllegalStateException("WorkManager needs to be initialized via a "
+ "ContentProvider#onCreate() or an Application#onCreate().");
}

my analysis:

code in handleCreateService() in ActivityThread.java:
" Application app = packageInfo.makeApplication(false, mInstrumentation);
service.attach(context, this,

data.info.name, data.token, app, ActivityManager.getService())
"

handleCreateService--->. Application app = packageInfo.makeApplication(false, mInstrumentation); ---> forceDefaultAppClass is always false-->if have custom Application,will make custom Application ---> service.attach(context, this,

data.info.name, data.token, app, ActivityManager.getService()); ---> so getApplication() in service get custom application
--> if have customApplication, the code : "Application.class.equals(getApplication().getClass()" in Service will be false , even it's in backup mode