1147992 - crash in android.os.BadParcelableException: ClassNotFoundException when unmarshalling: com.facebook.graphql.model.GraphQLPrefetchInfo at android.os.Parcel.readParcelable(Parcel.java)

Status: VERIFIED FIXED

(Firefox for Android Graveyard :: General, defect)

Description

[Richard Newman [:rnewman]]

This bug was filed from the Socorro interface and is 
report bp-3de6067a-dce3-4b7f-871f-072352150321.
=============================================================

        final String args;
        if (context instanceof Activity) {
            args = ((Activity) context).getIntent().getStringExtra("args");
        } else {
            args = null;
        }

android.os.BadParcelableException: ClassNotFoundException when unmarshalling: com.facebook.graphql.model.GraphQLPrefetchInfo
	at android.os.Parcel.readParcelable(Parcel.java:2080)
	at android.os.Parcel.readValue(Parcel.java:1968)
	at android.os.Parcel.readMapInternal(Parcel.java:2229)
	at android.os.Bundle.unparcel(Bundle.java:223)
	at android.os.Bundle.getString(Bundle.java:1063)
	at android.content.Intent.getStringExtra(Intent.java:4492)
	at org.mozilla.gecko.GeckoProfile.get(GeckoProfile.java:98)
	at org.mozilla.gecko.BrowserApp.onCreate(BrowserApp.java:496)


Goddamnit, Android.

Comment 1

[Richard Newman [:rnewman]]

Line last touched in Bug 1074340.

Comment 2

[Richard Newman [:rnewman]]

Attachment: Be paranoid when reading strings from intents, yet again. v1

Comment 3

[Kevin Brosnan [Ex-Mozilla]]

[Tracking Requested - why for this release]: We might want this fix on 37. Crash if Firefox is not running when opening a link from Facebook app to Firefox.

Comment 4

[Richard Newman [:rnewman]]

https://hg.mozilla.org/integration/fx-team/rev/ca43c5951976

Comment 5

[Richard Newman [:rnewman]]

Comment on attachment 8583935 [details] [diff] [review]
Be paranoid when reading strings from intents, yet again. v1

Approval Request Comment
[Feature/regressing bug #]:
  N/A
  This happens whenever Android sends us an intent including a non-standard instance (in this case, Facebook is adding something), and the sending process dies before we can process the intent. Not our bug; this is a workaround.

[User impact if declined]:
  Crashes opening links from Facebook.

[Describe test coverage new/current, TreeHerder]:
  None.

[Risks and why]: 
  Very low risk: established workaround applied to a new case.

[String/UUID change made/needed]:
  None.

Comment 6

[Lawrence Mandel [:lmandel] (use needinfo)]

Comment on attachment 8583935 [details] [diff] [review]
Be paranoid when reading strings from intents, yet again. v1

We're not going to ship any more fixes for 36 and we have already built and qualified the 37 RC. We can consider this for a ridealong if we need to ship a 37 based point release. We should take this change on Aurora after it has hit m-c. Release- and Beta-

Comment 7

[Lawrence Mandel [:lmandel] (use needinfo)]

Tracking for 37 in case we do a point release.

Comment 8

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/ca43c5951976

Comment 9

[Kevin Brosnan [Ex-Mozilla]]

This is now the number 2 top crash in 36.0.4.

Comment 10

[Lawrence Mandel [:lmandel] (use needinfo)]

Any idea why this has spiked?

Comment 11

[Kevin Brosnan [Ex-Mozilla]]

Will this fix bug 1106090 as well?

android.os.BadParcelableException: ClassNotFoundException when unmarshalling: com.facebook.graphql.model.GraphQLPrefetchInfo
	at android.os.Parcel.readParcelableCreator(Parcel.java:2289)
	at android.os.Parcel.readParcelable(Parcel.java:2239)
	at android.os.Parcel.readValue(Parcel.java:2146)
	at android.os.Parcel.readArrayMapInternal(Parcel.java:2479)
	at android.os.BaseBundle.unparcel(BaseBundle.java:221)
	at android.os.BaseBundle.getString(BaseBundle.java:918)
	at android.content.Intent.getStringExtra(Intent.java:5378)
	at org.mozilla.gecko.GeckoProfile.get(GeckoProfile.java:98)
	at org.mozilla.gecko.BrowserApp.onCreate(BrowserApp.java:496)
	at android.app.Activity.performCreate(Activity.java:6221)
	at android.app.Instrumentation.callActivityOnCreate(Instrumentation.java:1119)
	at android.app.ActivityThread.performLaunchActivity(ActivityThread.java:2614)
	at android.app.ActivityThread.handleLaunchActivity(ActivityThread.java:2728)
	at android.app.ActivityThread.access$900(ActivityThread.java:172)
	at android.app.ActivityThread$H.handleMessage(ActivityThread.java:1421)
	at android.os.Handler.dispatchMessage(Handler.java:102)
	at android.os.Looper.loop(Looper.java:145)
	at android.app.ActivityThread.main(ActivityThread.java:5837)
	at java.lang.reflect.Method.invoke(Native Method)
	at java.lang.reflect.Method.invoke(Method.java:372)
	at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:1388)
	at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:1183)

Comment 12

[Richard Newman [:rnewman]]

(In reply to Lawrence Mandel [:lmandel] (use needinfo) from comment #10)
> Any idea why this has spiked?

New version of Facebook, perhaps?


(In reply to Kevin Brosnan [:kbrosnan] from comment #11)
> Will this fix bug 1106090 as well?

Yeah, straight dupe.

Comment 14

[Richard Newman [:rnewman]]

This is hitting Chrome, too:

https://code.google.com/p/chromium/issues/detail?id=469219

Comment 15

[Richard Newman [:rnewman]]

I have a question out to my Facebook friends.

Comment 16

[Richard Newman [:rnewman]]

Previous bug: Bug 1077645.
Previous Chromium issue: https://code.google.com/p/chromium/issues/detail?id=412527

which was for the Facebook "ActionSource" class.

Comment 17

[Richard Newman [:rnewman]]

Apparently FB has turned off the server-side feature that was driving this client behavior, so our crash rate should nosedive. Kevin and Aaron, would you keep an eye on this, please?

Comment 18

[Lawrence Mandel [:lmandel] (use needinfo)]

If that's the case, we won't need to respin for 37 but should ensure that this fix gets in 38 soon and should keep this tracked for 37 in case this crash spikes again before we release 38.

Comment 19

[Kevin Brosnan [Ex-Mozilla]]

Hard to tell volume in less than a day. Still seeing crashes for example one at 2015-03-28 21:49:49 where it is currently 22:26.

Comment 20

[Richard Newman [:rnewman]]

(In reply to Lawrence Mandel [:lmandel] (use needinfo) from comment #18)
> If that's the case, we won't need to respin for 37 but should ensure that
> this fix gets in 38 soon and should keep this tracked for 37 in case this
> crash spikes again before we release 38.

Still waiting for approval for uplift to 38.

Comment 21

[Sylvestre Ledru [:Sylvestre]]

Comment on attachment 8583935 [details] [diff] [review]
Be paranoid when reading strings from intents, yet again. v1

I was waiting for the patch to land in m-c.

Comment 22

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/36249d8296a4

Comment 23

[Kevin Brosnan [Ex-Mozilla]]

Looks like we are back to background noise levels in the 3 day view.

Comment 24

[Mark Finkle (:mfinkle) (use needinfo?)]

Comment on attachment 8583935 [details] [diff] [review]
Be paranoid when reading strings from intents, yet again. v1

Reopening the request for uplift to the point release. This is low risk and prevents a crash that is controlled by a third party. If we can close up the crash ourselves, I'd feel more confident.

Comment 25

[Lawrence Mandel [:lmandel] (use needinfo)]

Comment on attachment 8583935 [details] [diff] [review]
Be paranoid when reading strings from intents, yet again. v1

Agreed about low risk. We'll take this as a ride along in the 37.0.1 point release. Release+

Comment 26

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-release/rev/7abec92ce5f9

Comment 27

[Ioana Chiorean]

I have tried different workaround and different apps - I am not hitting it. I will keep an eye on it too for the crashstats.