761399 - RefCountedJavaObject::mRefCnt is not initialized

Status: RESOLVED INVALID

(Core Graveyard :: Widget: Android, defect)

Comment 1

[Brad Lassey [:blassey] (use needinfo?)]

Attachment: patch

Comment 2

[Julian Seward [:jseward]]

Occurs on every startup now; wasn't present a week ago.

Conditional jump or move depends on uninitialised value(s)
   at 0x37089218: nsAppShell::ProcessNextNativeEvent(bool) (AndroidJavaWrappers.h:54)
   by 0x3709D461: nsBaseAppShell::DoProcessNextNativeEvent(bool, unsigned int) (nsBaseAppShell.cpp:139)
   by 0x3709D519: nsBaseAppShell::OnProcessNextEvent(nsIThreadInternal*, bool, unsigned int) (nsBaseAppShell.cpp:280)
   by 0x37216953: nsThread::ProcessNextEvent(bool, bool*) (nsThread.cpp:586)
   by 0x371EEA05: NS_ProcessNextEvent_P(nsIThread*, bool) (nsThreadUtils.cpp:213)
   by 0x3712FE95: mozilla::ipc::MessagePump::Run(base::MessagePump::Delegate*) (MessagePump.cpp:82)
   by 0x372427FB: MessageLoop::RunInternal() (message_loop.cc:208)
   by 0x37242805: MessageLoop::RunHandler() (message_loop.cc:201)
   by 0x3724290D: MessageLoop::Run() (message_loop.cc:175)
   by 0x3709D0B7: nsBaseAppShell::Run() (nsBaseAppShell.cpp:163)
   by 0x36FA1891: nsAppStartup::Run() (nsAppStartup.cpp:256)
   by 0x368609B9: XREMain::XRE_mainRun() (nsAppRunner.cpp:3786)
   by 0x36860BF5: XREMain::XRE_main(int, char**, nsXREAppData const*) (nsAppRunner.cpp:3863)
   by 0x36860D97: XRE_main (nsAppRunner.cpp:3939)
   by 0x36864BE7: GeckoStart (nsAndroidStartup.cpp:73)
   by 0x2238AEE5: Java_org_mozilla_gecko_GeckoAppShell_nativeRun (APKOpen.cpp:970)
 Uninitialised value was created by a heap allocation
   at 0x48071D0: __wrap_malloc (vg_replace_malloc.c:275)
   by 0x364E272B: moz_xmalloc (mozalloc.cpp:54)
   by 0x3708AF8F: mozilla::AndroidGeckoEvent::Init(_JNIEnv*, _jobject*) (mozalloc.h:200)
   by 0x37094C83: Java_org_mozilla_gecko_GeckoAppShell_notifyGeckoOfEvent (AndroidJavaWrappers.h:556)
   by 0x2238AB59: Java_org_mozilla_gecko_GeckoAppShell_notifyGeckoOfEvent (APKOpen.cpp:290)
   by 0x540FBF3: dvmPlatformInvoke (CallEABI.S:258)

Comment 3

[Julian Seward [:jseward]]

(In reply to Brad Lassey [:blassey] from comment #1)
> Created attachment 629964 [details] [diff] [review]

Yeah, that seems to stop V complaining; +1 from me.

Comment 4

[Brad Lassey [:blassey] (use needinfo?)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/6e991c0ba5bd

Comment 5

[Brad Lassey [:blassey] (use needinfo?)]

backed out for talos panning regression https://hg.mozilla.org/integration/mozilla-inbound/rev/1ab8c859b19e

Comment 6

[Julian Seward [:jseward]]

I can't repro this any more.  OK to close as invalid?

Comment 7

[Kartikaya Gupta (email:[e-mail])]

This was re-landed as part of bug 755070 during one of the many landing iterations. So the code is actually in m-c already and this bug is invalid, yes.