798858 - android.view.ViewRootImpl$CalledFromWrongThreadException: Only the original thread that created a view hierarchy can touch its views. at android.view.ViewRootImpl.checkThread(ViewRootImpl.java) at org.mozilla.gecko.AboutHomeSection.show

Status: RESOLVED FIXED

(Firefox for Android Graveyard :: General, defect, P1)

Description

[Scoobidiver (away)]

This startup crash is #1 top crasher in today's build. The regression range is:
http://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=2da1f2bde40e&tochange=9f677c2bb33d
It might be a regression from bug 797075.

android.view.ViewRootImpl$CalledFromWrongThreadException: Only the original thread that created a view hierarchy can touch its views.
	at android.view.ViewRootImpl.checkThread(ViewRootImpl.java:4607)
	at android.view.ViewRootImpl.invalidateChildInParent(ViewRootImpl.java:867)
	at android.view.ViewGroup.invalidateChild(ViewGroup.java:4066)
	at android.view.View.invalidate(View.java:10250)
	at android.view.View.setFlags(View.java:8366)
	at android.view.View.setVisibility(View.java:5512)
	at org.mozilla.gecko.AboutHomeSection.show(AboutHomeSection.java:87)
	at org.mozilla.gecko.AboutHomeContent.onQueryTabsComplete(AboutHomeContent.java:625)
	at org.mozilla.gecko.TabsAccessor$2.onPostExecute(TabsAccessor.java:159)
	at org.mozilla.gecko.TabsAccessor$2.onPostExecute(TabsAccessor.java:118)
	at org.mozilla.gecko.util.GeckoAsyncTask$BackgroundTaskRunnable$1.run(GeckoAsyncTask.java:35)
	at android.os.Handler.handleCallback(Handler.java:615)
	at android.os.Handler.dispatchMessage(Handler.java:92)
	at android.os.Looper.loop(Looper.java:137)
	at org.mozilla.gecko.util.GeckoBackgroundThread.run(GeckoBackgroundThread.java:31)

More reports at:
https://crash-stats.mozilla.com/report/list?signature=android.view.ViewRoot%24CalledFromWrongThreadException%3A+Only+the+original+thread+that+created+a+view+hierarchy+can+touch+its+views.+at+android.view.ViewRoot.checkThread%28ViewRoot.java%29
https://crash-stats.mozilla.com/report/list?signature=android.view.ViewRootImpl%24CalledFromWrongThreadException%3A+Only+the+original+thread+that+created+a+view+hierarchy+can+touch+its+views.+at+android.view.ViewRootImpl.checkThread%28ViewRootImpl.java%29

Comment 1

[Brian Nicholson (:bnicholson)]

(In reply to Scoobidiver from comment #0)
> It might be a regression from bug 797075.
> 
Bug 797075 was backed out. This looks like a regression from bug 798135.

Comment 2

[Ed Morley [:emorley]]

I've just backed bug 798135 out, so hopefully the Nightly generated in a bit (presuming the m-c backout is green enough in time for it to be auto-picked) will show this as no longer occurring.

Comment 3

[Axel Hecht [:Pike]]

My nightly from https://hg.mozilla.org/mozilla-central/rev/9738e5a0190a still crashes, which, II read right, is two revisions younger than the backout. That's on an ASUS transformer.

Comment 4

[Scoobidiver (away)]

(In reply to Axel Hecht [:Pike] from comment #3)
> My nightly from https://hg.mozilla.org/mozilla-central/rev/9738e5a0190a
> still crashes, which, II read right, is two revisions younger than the
> backout.
Are you sure it's this crash signature? I don't see it in today's nightly: https://crash-stats.mozilla.com/query/query?product=FennecAndroid&version=FennecAndroid%3A18.0a1&build_id=20121008031745&do_query=1

Comment 5

[Paul [pwd]]

I too am still seeing crashes, though they're not being caught by about:crashes

Comment 6

[Naoki Hirata :nhirata (please use needinfo instead of cc)]

Axel or Paul, could we get a logcat of your crashes please?

Comment 7

[Paul [pwd]]

Apologies nhirata but I uninstalled/reinstalled and am now stuck seeing bug 799141 causing the crashes.

Comment 8

[Chris Peterson [:cpeterson]]

Sriram, these crashes look like a regression from bug 798135. The stack traces all point to TabsAccessor.onPostExecute() running on GeckoBackgroundThread instead of the UI thread:

 android.view.ViewRoot$CalledFromWrongThreadException: Only the original thread that 
 created a view hierarchy can touch its views.
 at android.view.ViewRoot.checkThread(ViewRoot.java:3429)
 ...
 at android.view.View.setVisibility(View.java:3802)
 at org.mozilla.gecko.AboutHomeSection.show(AboutHomeSection.java:87)
 at org.mozilla.gecko.AboutHomeContent.onQueryTabsComplete(AboutHomeContent.java:625)
 at org.mozilla.gecko.TabsAccessor$2.onPostExecute(TabsAccessor.java:159)
 at org.mozilla.gecko.TabsAccessor$2.onPostExecute(TabsAccessor.java:118)
 at org.mozilla.gecko.util.GeckoAsyncTask$BackgroundTaskRunnable$1.run(GeckoAsyncTask.java:35)
 at android.os.Handler.handleCallback(Handler.java:587)
 at android.os.Handler.dispatchMessage(Handler.java:92)
 at android.os.Looper.loop(Looper.java:132)
 at org.mozilla.gecko.util.GeckoBackgroundThread.run(GeckoBackgroundThread.java:31)

Since your GeckoAsyncTask change "captures" the current thread's Handler, I suspect some chain of calls where AboutHomeSection is invoked from TabsAccessor is invoked from BrowserApp on the GeckoBackgroundThread. You may want to add `GeckoApp.assertOnUiThread()` calls before some of the `new GeckoAsyncTask()` calls to ensure the GeckoAsyncTask constructors are being called on the UI thread (when you believe they are).

Comment 9

[Chris Peterson [:cpeterson]]

(In reply to Axel Hecht [:Pike] from comment #3)
> My nightly from https://hg.mozilla.org/mozilla-central/rev/9738e5a0190a
> still crashes, which, II read right, is two revisions younger than the
> backout. That's on an ASUS transformer.

Alex, do you have STR for your ASUS Transformer crash?

Comment 10

[Axel Hecht [:Pike]]

I'm just surfing a few sites, and then let it sit, it seems to crash on being idle for a few seconds.

Comment 11

[Chris Peterson [:cpeterson]]

sriram, did backing out bug 798135 fix this crash?