Closed Bug 828911 Opened 9 years ago Closed 6 years ago

startup crash in gfxWindowsPlatform::VerifyD2DDevice @ nvwgf2um.dll@0xf648a (NVIDIA driver 310.90) with DisplayLink

Categories

(Core :: Graphics, defect)

Product:
Core
Component:
Graphics
Version:
18 Branch
Platform:
x86
Windows 7
Type:
defect
Priority:
Not set
Severity:
critical

Tracking

()

Status:
RESOLVED INCOMPLETE

People

(Reporter: scoobidiver, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: crash, Whiteboard: [startupcrash])

Crash Data

It's #15 top browser crasher in 18.0 and happens on startup.

It's correlated to NVIDIA driver 310.90:
    100% (60/60) vs.   5% (2754/60471) nvwgf2um.dll
        100% (60/60) vs.   1% (416/60471) 9.18.13.1090

Signature 	nvwgf2um.dll@0xf648a More Reports Search
UUID	be639eb9-76c2-4ae4-ad67-bb65f2130110
Date Processed	2013-01-10 00:48:19
Uptime	2
Last Crash	21.1 hours before submission
Install Age	2 seconds since version was first installed.
Install Time	2013-01-10 00:47:58
Product	Firefox
Version	18.0
Build ID	20130104151925
Release Channel	release
OS	Windows NT
OS Version	6.1.7601 Service Pack 1
Build Architecture	x86
Build Architecture Info	AuthenticAMD family 16 model 4 stepping 2
Crash Reason	EXCEPTION_ACCESS_VIOLATION_WRITE
Crash Address	0x64
User Comments	Probably has to do with the new nVidia driver.
App Notes 	
AdapterVendorID: 0x10de, AdapterDeviceID: 0x0e22, AdapterSubsysID: 23811462, AdapterDriverVersion: 9.18.13.1090
Processor Notes 	WARNING: JSON file missing Add-ons
EMCheckCompatibility	False
Adapter Vendor ID	0x10de
Adapter Device ID	0x0e22
Total Virtual Memory	4294836224
Available Virtual Memory	4085063680
System Memory Use Percentage	37
Available Page File	6431498240
Available Physical Memory	2691620864

Frame 	Module 	Signature 	Source
0 	nvwgf2um.dll 	nvwgf2um.dll@0xf648a 	
1 	nvwgf2um.dll 	nvwgf2um.dll@0xf47f2 	
2 	nvwgf2um.dll 	nvwgf2um.dll@0x14435a 	
3 	nvwgf2um.dll 	nvwgf2um.dll@0x14382c 	
4 	nvwgf2um.dll 	nvwgf2um.dll@0x1700af 	
5 	nvwgf2um.dll 	nvwgf2um.dll@0x170582 	
6 	nvwgf2um.dll 	nvwgf2um.dll@0x1439b4 	
7 	nvwgf2um.dll 	nvwgf2um.dll@0x16fb4a 	
8 	nvwgf2um.dll 	nvwgf2um.dll@0x143add 	
9 	nvwgf2um.dll 	nvwgf2um.dll@0x16fb4a 	
10 	nvwgf2um.dll 	nvwgf2um.dll@0xf4550 	
11 	nvwgf2um.dll 	nvwgf2um.dll@0x1420b1 	
12 	nvwgf2um.dll 	nvwgf2um.dll@0x142334 	
13 	nvwgf2um.dll 	nvwgf2um.dll@0x1424dd 	
14 	nvwgf2um.dll 	nvwgf2um.dll@0x1406e6 	
15 	nvwgf2um.dll 	nvwgf2um.dll@0x6832 	
16 	nvwgf2um.dll 	nvwgf2um.dll@0xc6ab0 	
17 	dlumd32.dll 	dlumd32.dll@0x26f00 	
18 	dlumd32.dll 	dlumd32.dll@0x41e83 	
19 	dxgi.dll 	CD3D10Device::DestroyDriverInstance 	
20 	d3d10_1core.dll 	CDevice::LLOBeginLayerDestruction 	
21 	d3d10_1core.dll 	CBridgeImpl<ILayeredLockOwner,IDXGILayeredDevice,CLayeredObject<CDevice> >::LLOB 	
22 	dxgi.dll 	CD3D10Device::LLOBeginLayerDestruction 	
23 	dxgi.dll 	CBridgeImpl<ILayeredLockOwner,IDXGILayeredDevice,CLayeredObject<CD3D10Device> >: 	
24 	d3d10_1core.dll 	NMultithread::CDevice::LLOBeginLayerDestruction 	
25 	d3d10_1core.dll 	CBridgeImpl<ILayeredLockOwner,IDXGILayeredDevice,CLayeredObject<NMultithread::CD 	
26 	dxgi.dll 	NOutermost::CDevice::LLOBeginLayerDestruction 	
27 	dxgi.dll 	NOutermost::CDevice::FinalRelease 	
28 	dxgi.dll 	TComObject<NOutermost::CDevice>::~TComObject<NOutermost::CDevice> 	
29 	dxgi.dll 	TComObject<NOutermost::CDevice>::`scalar deleting destructor' 	
30 	dxgi.dll 	NOutermost::CDevice::FinalRelease 	
31 	d3d10_1core.dll 	CLayeredObject<NMultithread::CDevice>::CContainedObject::Release 	
32 	xul.dll 	nsRefPtr<ID3D10Effect>::assign_assuming_AddRef 	obj-firefox/dist/include/mozilla/StaticPtr.h:148
33 	xul.dll 	nsRefPtr<nsPluginArray>::assign_with_AddRef 	obj-firefox/dist/include/nsAutoPtr.h:846
34 	xul.dll 	gfxWindowsPlatform::VerifyD2DDevice 	gfx/thebes/gfxWindowsPlatform.cpp:664
35 	xul.dll 	gfxWindowsPlatform::UpdateRenderMode 	gfx/thebes/gfxWindowsPlatform.cpp:474
36 	xul.dll 	gfxWindowsPlatform::gfxWindowsPlatform 	gfx/thebes/gfxWindowsPlatform.cpp:365
37 	xul.dll 	gfxPlatform::Init 	gfx/thebes/gfxPlatform.cpp:292
38 	xul.dll 	gfxPlatform::GetPlatform 	gfx/thebes/gfxPlatform.cpp:239
39 	xul.dll 	nsDOMClassInfo::Init 	dom/base/nsDOMClassInfo.cpp:4634
40 	xul.dll 	NS_GetDOMClassInfoInstance 	dom/base/nsDOMClassInfo.cpp:5215
41 	xul.dll 	nsGlobalChromeWindow::QueryInterface 	dom/base/nsGlobalWindow.cpp:10806
42 	xul.dll 	xpcObjectHelper::GetClassInfo 	js/xpconnect/src/xpcObjectHelper.h:71
43 	xul.dll 	XPCWrappedNative::WrapNewGlobal 	js/xpconnect/src/XPCWrappedNative.cpp:303
44 	xul.dll 	nsXPConnect::InitClassesWithNewWrappedGlobal 	js/xpconnect/src/nsXPConnect.cpp:1168
45 	xul.dll 	CreateNativeGlobalForInner 	dom/base/nsGlobalWindow.cpp:1747
46 	xul.dll 	nsGlobalWindow::SetNewDocument 	dom/base/nsGlobalWindow.cpp:1949
47 	xul.dll 	DocumentViewerImpl::InitInternal 	layout/base/nsDocumentViewer.cpp:927
48 	xul.dll 	DocumentViewerImpl::Init 	layout/base/nsDocumentViewer.cpp:677
...

More reports at:
https://crash-stats.mozilla.com/report/list?signature=nvwgf2um.dll%400xf648a
Can not reproduce.

The scenario listed @ KB2488113 likely applies, It would help to ascertain if these users have other applications that utilise the IE backend open, such applications include WLM 2011, WLMail 2011, Skype
There isn't any correlation between hardware, either.

one user is using a Geforce 210
another is using a GTX 480
another using a GTX 650.

I'm using a GTX 680 myself and cannot reproduce the problem, i have KB2488113 installed, and cannot reproduce the crash what so ever.
Checked the modules list on a few of them and some of them actually have got that hotfix installed... atleast one of them is without sp1 and is running an original version of the affected dll

d3d10_1.dll 	6.1.7601.17514 > d3d10_1.dll 	6.1.7601.17544

That doesn't remove from the fact that my GPU is a prime candidate for the crash, yet i cannot reproduce it.

3/4 of that list is duplicate reports from the few users, so it would be negligent to blocklist the driver at this point.  There must be more to it.
Is this the same as bug 823077?
comment 7 (yours infact) of bug 823077? suggests it is.
bjacob, does it make sense to update bug 823966 to include the impacted versions here? Or do you have further investigation that you'd like to do first?
Assignee: nobody → bjacob
Sure, that seems to make sense! And while you're at it, if that's convenient to do, why not block everything in between these two versions. (If you can use >= and <= operators there).

Unassigning myself because I can't really commit to work on anything non-B2G at the moment, but I'm happy to answer questions like that.
Assignee: bjacob → nobody
(In reply to Danial Horton from comment #1)
> Can not reproduce.

Doesn't matter when it comes to a top crasher correlated with a driver version.

(In reply to Danial Horton from comment #2)
> There isn't any correlation between hardware, either.

So?

We're talking about blacklisting a specific driver version, not a specific hardware.
It's now #39 browser crasher in 18.0 so no longer a top crasher.
Its ranking is not stable because it's a startup crash.

One comment says:
"Hi, For some reason I can't use my Firefox without safe mode anymore. I can only use it with tha safe mode. Everytime I try to start Firefox, the Mozilla's crash program pops up and Firefox doesn't start even if I press "Start Firefox again". This problem become when I installed nVidia's latest drivers for my Geforce 660 Ti. But when I removed the drives, it didn't help to start Firefox normally. I really appreciate your help with this issue. Best regards, Reko"
tracking-firefox18: ? → ---
Keywords: topcrash
(In reply to Scoobidiver from comment #9)
> It's now #39 browser crasher in 18.0 so no longer a top crasher.
> Its ranking is not stable because it's a startup crash.
> 
> One comment says:
> "Hi, For some reason I can't use my Firefox without safe mode anymore. I can
> only use it with tha safe mode. Everytime I try to start Firefox, the
> Mozilla's crash program pops up and Firefox doesn't start even if I press
> "Start Firefox again". This problem become when I installed nVidia's latest
> drivers for my Geforce 660 Ti. But when I removed the drives, it didn't help
> to start Firefox normally. I really appreciate your help with this issue.
> Best regards, Reko"

Doesn't this imply that the issue is related to third party software other than the driver? Can we pull new correlations?
exactly, but it seems certain people here don't have a clue about correlation or "DRIVER" bugs.

If it was a bug in the driver(s), EVERYONE would have been experiencing it on any system they attempted to use the drivers on.

This issue exploded after Tuesday the 8th, and i have seen first hand a number of systems experiencing

-crashing
-failing to boot
-corruption because one of the updates failed to install in a timeley fashion, so the user pulled the plug -_-

So it seems those with systems already on the edge of failing are being pushed over by this update

Since some users, running the latest patches were not getting the crashes, its not specifically (atleast, solely) a fault in the hotfixes, but possibly a combination of software on the system.
It's now #55 browser crasher in 18.0.

(In reply to Alex Keybl [:akeybl] from comment #10)
> Doesn't this imply that the issue is related to third party software other
> than the driver? Can we pull new correlations?
You're right. At first glance, I though it was a driver DLL, but it's not. It's DisplayLink Core Software (http://www.displaylink.com/support/index.php):
    100% (52/52) vs.   0% (103/75925) dlumd32.dll
          0% (0/52) vs.   0% (1/75925) 5.2.22271.0
          0% (0/52) vs.   0% (5/75925) 8.5.4.26772
          0% (0/52) vs.   0% (1/75925) 8.5.4.28579
          0% (0/52) vs.   0% (2/75925) 8.5.5.29055
          0% (0/52) vs.   0% (1/75925) 8.5.5.30661
          0% (0/52) vs.   0% (2/75925) 8.5.6.30904
          8% (4/52) vs.   0% (17/75925) 8.5.6.31854
          0% (0/52) vs.   0% (4/75925) 8.6.0.34621
          4% (2/52) vs.   0% (5/75925) 8.6.1.35667
          0% (0/52) vs.   0% (1/75925) 8.6.1.35912
          6% (3/52) vs.   0% (12/75925) 8.6.1.36484
          0% (0/52) vs.   0% (2/75925) 8.6.3.38355
         15% (8/52) vs.   0% (15/75925) 8.6.3.40660
         58% (30/52) vs.   0% (30/75925) 8.7.0.41409
          4% (2/52) vs.   0% (2/75925) 8.7.0.43526
          6% (3/52) vs.   0% (3/75925) 8.7.1.44152
Summary: startup crash in gfxWindowsPlatform::VerifyD2DDevice @ nvwgf2um.dll@0xf648a (NVIDIA driver 310.90) → startup crash in gfxWindowsPlatform::VerifyD2DDevice @ nvwgf2um.dll@0xf648a (NVIDIA driver 310.90) with DisplayLink
(In reply to Alex Keybl [:akeybl] from comment #10)
> Doesn't this imply that the issue is related to third party software other
> than the driver? Can we pull new correlations?

No, safe mode always deactivates hardware acceleration. If anything, this confirms that the blocklist should work to fix the crash (even though it will make Firefox slower on a number of things for all people with those drivers).
Blocking this driver version for all NVIDIA users seems overkilling as only DisplayLink users are affected.

In addition, it continues dropping off as it's now #72 browser crasher in 18.0.
A number of reports have beem nade on the forums that the Tuesday updates have indeed introduced crashing in the Display Link drivers.

An updated display link driver is reportedly on the way,  Mozilla should be able to block the affected display link driver right?
(In reply to Danial Horton from comment #15)
> An updated display link driver is reportedly on the way,  Mozilla should be
> able to block the affected display link driver right?
It no longer meets blocklisting criteria: https://wiki.mozilla.org/Blocklisting#Blocklisting_Policy
(In reply to Danial Horton from comment #11)
> exactly, but it seems certain people here don't have a clue about
> correlation or "DRIVER" bugs.

Alright, un-CC-ing myself here.
I can confirm that after a lot of investigating and crash, I can say that it depends from new Nvidia drivers 310.90 . I didn't try yet the new version 313.95-notebook-win8-win7-winvista-64bit-international-beta 

What I can say is that happened on Asus Notebook U36S series with Optimus Technology and dual vga:

Intel(R) HD Graphics 3000

Report Date:			1/28/2013
Report Time[hh:mm:ss]:		22:44:04
Driver Version:			9.17.10.2932
Operating System:		Windows 7  Service Pack 1(6.1.7601)
Default Language:		Italian (Italy)
DirectX* Version:		11.0
Physical Memory:		4001 MB
Processor:			Intel64 Family 6 Model 42 Stepping 7
Processor Speed:		2693 MHz
Vendor ID:			8086
Device ID:			0126
Device Revision:		09

*   Processor Graphics Information   *


Processor Graphics in Use:		Intel(R) HD Graphics 3000
Video BIOS:			2117.0
Current Graphics Mode:		1366 by 768

That doesn't give any problem.

The problems begin immediately when I install drivers for nVIDIA GeForce GT 520M (GF119M).
submit the crash if you haven't via about:crashes and post the link to the relevant info

I can't confirm your situation under either of the optimus laptops i have available to me here.
I am closing this as incomplete since I can't find recent reports to investigate. Please reopen this bug report and provide crash report IDs if you are able to reproduce it. Thanks.
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.