804493 - Win8 crash in atidxx32.dll 8.17.10.451, 8.17.10.453 and 8.17.10.470 (ATI driver versions 8.982.6.0, 8.982.7.0, 8.982.9.0, 8.982.10.8000, 9.0.0.0, 9.1.0.0, 9.2.0.0 and 9.10.8.0)

Status: RESOLVED INCOMPLETE

(Core :: Graphics, defect)

Description

[Scoobidiver (away)]

It's the same as bug 783517.
It's #2 top crasher on Windows 8: https://crash-analysis.mozilla.com/rkaiser/2012-10-21/2012-10-21.firefox.16.0.1.win8.topcrash.html

The latest AMD driver version is 9.2 (Catalyst 12.10 - see http://support.amd.com/us/kbarticles/Pages/AMDCatalystSoftwareSuiteVersion1210ReleaseNotes.aspx) but Windows 8 is officially supported from 8.982 (Catalyst 12.8 - see http://support.amd.com/us/kbarticles/Pages/AMDCatalystSoftwareSuiteVersion128ReleaseNotes.aspx).

The underlying issue should be fixed instead of gfx-blocklisting each new released driver version.

Signature 	atidxx32.dll@0x3c230c More Reports Search
UUID	d08d7aa4-a566-4cc4-a039-da9562121023
Date Processed	2012-10-23 02:48:02
Uptime	252
Last Crash	4.2 minutes before submission
Install Age	1.6 weeks since version was first installed.
Install Time	2012-10-11 23:48:08
Product	Firefox
Version	17.0
Build ID	20121010150351
Release Channel	beta
OS	Windows NT
OS Version	6.2.9200
Build Architecture	x86
Build Architecture Info	GenuineIntel family 6 model 42 stepping 7
Crash Reason	EXCEPTION_ACCESS_VIOLATION_READ
Crash Address	0x18e71000
App Notes 	
AdapterVendorID: 0x1002, AdapterDeviceID: 0x6719, AdapterSubsysID: 21fd1458, AdapterDriverVersion: 9.1.0.0
D2D? D2D+ DWrite? DWrite+ D3D10 Layers? D3D10 Layers+ 
EMCheckCompatibility	True
Adapter Vendor ID	0x1002
Adapter Device ID	0x6719
Total Virtual Memory	4294836224
Available Virtual Memory	3742285824
System Memory Use Percentage	21
Available Page File	28239351808
Available Physical Memory	10147028992

Frame 	Module 	Signature 	Source
0 	atidxx32.dll 	atidxx32.dll@0x3c230c 	
1 	atidxx32.dll 	atidxx32.dll@0x5a2a3 	
2 	atidxx32.dll 	atidxx32.dll@0x5a06f 	
3 	atidxx32.dll 	atidxx32.dll@0x5a52d 	
4 	atidxx32.dll 	atidxx32.dll@0xe067 	
5 	atidxx32.dll 	atidxx32.dll@0xd949 	
6 	atidxx32.dll 	atidxx32.dll@0x4633 	
7 	atidxx32.dll 	atidxx32.dll@0x222a6 	
8 	atiuxpag.dll 	atiuxpag.dll@0x44e7 	
9 	d3d11.dll 	NDXGI::CDevice::SetSharedResourceCreationArgs 	
10 	d3d11.dll 	TCLSWrappers<CTexture2D>::CLSFinalConstructFn 	
11 	d3d11.dll 	CTexture2D::FinalConstruct 	
12 	d3d11.dll 	CBridgeImpl<ID3D11LayeredDevice,ID3D11LayeredDevice,CLayeredObject<CDevice> >::C 	
13 	d3d11.dll 	NDXGI::CDeviceChild<IDXGIResource1>::FinalConstruct 	
14 	d3d11.dll 	NDXGI::CResource::FinalConstruct 	
15 	d3d11.dll 	NDXGI::CResource::LUCCompleteLayerConstruction 	
16 	d3d11.dll 	NOutermost::CDevice::CreateLayeredChild 	
17 	d3d11.dll 	CDevice::CreateTexture2D_Worker 	
18 	d3d11.dll 	CDevice::CreateTexture2D 	
19 	d2d1.dll 	CD3DDeviceCommon::CreateTexture 	
20 	d2d1.dll 	CD3DSurface::Create 	
21 	d2d1.dll 	BitmapRealization::CreateSurface 	
22 	d2d1.dll 	BitmapRealization::Initialize 	
23 	d2d1.dll 	CreateRefObject<BitmapRealization,win_scope::no_addref_release<CD3DDeviceLevel1> 	
24 	d2d1.dll 	CHwSurfaceRenderTarget::CreateCopiedBitmapFromMemory 	
25 	d2d1.dll 	DrawingContext::CreateBitmap 	
26 	d2d1.dll 	D2DDeviceContextBase<ID2D1HwndRenderTarget,ID2D1HwndRenderTarget,ID2D1DeviceCont 	
27 	gkmedias.dll 	_cairo_d2d_create_brush_for_pattern 	gfx/cairo/cairo/src/cairo-d2d-surface.cpp:2025
28 	gkmedias.dll 	_cairo_d2d_mask 	gfx/cairo/cairo/src/cairo-d2d-surface.cpp:3413
29 	gkmedias.dll 	_cairo_pattern_is_clear 	gfx/cairo/cairo/src/cairo-pattern.c:2032

More reports at:
https://crash-stats.mozilla.com/query/query?product=Firefox&query_search=signature&query_type=contains&query=atidxx32.dll&do_query=1

Comment 1

[Alex Keybl [:akeybl]]

Thanks Scoobidiver. Assigning to Joe, since he resolved bug bug 783517.

(In reply to Scoobidiver from comment #0)
> The underlying issue should be fixed instead of gfx-blocklisting each new
> released driver version.

I'll let Joe speak to our options at this point.

Comment 2

[Joe Drew (not getting mail)]

(In reply to Alex Keybl [:akeybl] from comment #1)
> Thanks Scoobidiver. Assigning to Joe, since he resolved bug bug 783517.
> 
> (In reply to Scoobidiver from comment #0)
> > The underlying issue should be fixed instead of gfx-blocklisting each new
> > released driver version.
> 
> I'll let Joe speak to our options at this point.

Unfortunately the underlying issue seems to be AMD's bug.

We can blocklist more AMD drivers, but first: We now have a dialogue open with AMD, so we should have them tell us what they're going to do about it.

Benoit, would you ping them?

Comment 3

[Scoobidiver (away)]

I tried to be exhaustive with the latest crash stats.

Comment 4

[Lukas Blakk [:lsblakk] use ?needinfo]

tracking in bug 806991

Comment 5

[Scoobidiver (away)]

(In reply to Lukas Blakk [:lsblakk] from comment #4)
> tracking in bug 806991
Bug 806991 doesn't block 8.982.6.0, 8.982.9.0, 9.0.0.0, 9.2.0.0 and 9.10.8.0.

Comment 6

[Alex Keybl [:akeybl]]

(In reply to Scoobidiver from comment #5)
> (In reply to Lukas Blakk [:lsblakk] from comment #4)
> > tracking in bug 806991
> Bug 806991 doesn't block 8.982.6.0, 8.982.9.0, 9.0.0.0, 9.2.0.0 and 9.10.8.0.

Thanks scoobidiver, it does now.

Comment 7

[Scoobidiver (away)]

As Direct2D is force-enabled on Metro canceling the effects of bug 806991, crashes occur in Metro Firefox.

Comment 8

[Scoobidiver (away)]

It's a low volume crash on MetroFirefox because Direct2D is no longer force-enabled.

Comment 9

[Milan Sreckovic [:milan] (needinfo for best results)]

koi applies to B2G 1.2

Comment 10

[u279076]

Closing this bug since we've had no reports after June 24, 2016. Please reopen this bug report if you can reproduce the crash in a current version.