620924 - crash [@ atioglxx.dll@0xd772c ] with ATI driver version 6.14.10.6747

Status: RESOLVED WORKSFORME

(Core :: Graphics, defect)

Description

[Scoobidiver (away)]

It is #132 top crasher in 4.0b7 for the last week.
It happens only with ATI driver version 6.14.10.6747.

It is the OpenGL driver of ATI.

Comments say:
"i can't fucking go to vanswarpedtour.com/bands it crashes every time without fail"
"I'm being bombed with pop ups wherever i go. Before when i was just using the Beta, i never ever had this problem."

So it is not related to WebGL and probably related to D2D/D3D10.
I think this ATI driver version should be blocklist for D2D/D3D10 features.

Signature	atioglxx.dll@0xd772c
UUID	d3e3f6ed-bdda-4699-bf9a-95dcd2101221
Time 	2010-12-21 15:49:53.847916
Uptime	17229
Last Crash	1175506 seconds (1.9 weeks) before submission
Install Age	31267 seconds (8.7 hours) since version was first installed.
Product	Firefox
Version	4.0b9pre
Build ID	20101221030401
Branch	2.0
OS	Windows NT
OS Version	6.1.7600
CPU	x86
CPU Info	AuthenticAMD family 15 model 104 stepping 2
Crash Reason	EXCEPTION_ACCESS_VIOLATION_EXEC
Crash Address	0x2047aec0
User Comments	
App Notes 	AdapterVendorID: 1002, AdapterDeviceID: 791f
MSAFD Tcpip [TCP/IP] : 2 : 1 :
MSAFD Tcpip [UDP/IP] : 2 : 2 : %SystemRoot%\system32\mswsock.dll
MSAFD Tcpip [RAW/IP] : 2 : 3 :
MSAFD Tcpip [TCP/IPv6] : 2 : 1 : %SystemRoot%\system32\mswsock.dll
MSAFD Tcpip [UDP/IPv6] : 2 : 2 :
MSAFD Tcpip [RAW/IPv6] : 2 : 3 : %SystemRoot%\system32\mswsock.dll
RSVP TCPv6 Service Provider : 2 : 1 :
RSVP TCP Service Provider : 2 : 1 : %SystemRoot%\system32\mswsock.dll
RSVP UDPv6 Service Provider : 2 : 2 :
RSVP UDP Service Provider : 2 : 2 : %SystemRoot%\system32\mswsock.dll
MSAFD NetBIOS [\Device\NetBT_Tcpip_{2812637C-6AD0-4D12-8F7A-331F5DC9EC89}] SEQPACKET 0 : 2 : 5 :
MSAFD NetBIOS [\Device\NetBT_Tcpip_{2812637C-6AD0-4D12-8F7A-331F5DC9EC89}] DATAGRAM 0 : 2 : 2 : %SystemRoot%\system32\mswsock.dll
MSAFD NetBIOS [\Device\NetBT_Tcpip_{BD700D70-4407-43E8-AA8B-DDA1E4D7C854}] SEQPACKET 1 : 2 : 5 :
MSAFD NetBIOS [\Device\NetBT_Tcpip_{BD700D70-4407-43E8-AA8B-DDA1E4D7C854}] DATAGRAM 1 :

Frame 	Module 	Signature [Expand] 	Source
0 		@0x2047aec0 	
1 	atioglxx.dll 	atioglxx.dll@0xd772c 	
2 	atioglxx.dll 	atioglxx.dll@0xd7751 	
3 	atioglxx.dll 	atioglxx.dll@0xd7e74 	
4 	atioglxx.dll 	atioglxx.dll@0xd76ef 	
5 	atioglxx.dll 	atioglxx.dll@0x8be49 	
6 	atioglxx.dll 	atioglxx.dll@0x7d07d 	
7 	atioglxx.dll 	atioglxx.dll@0x7b9f6 	
8 	atioglxx.dll 	atioglxx.dll@0x78079 	
9 	atioglxx.dll 	atioglxx.dll@0x7a1e9 	
10 	atioglxx.dll 	atioglxx.dll@0x8227b 	
11 	atioglxx.dll 	atioglxx.dll@0x7b0fe 	
12 	atioglxx.dll 	atioglxx.dll@0x7b054 	
13 	atioglxx.dll 	atioglxx.dll@0x7b18e 	
14 	atioglxx.dll 	atioglxx.dll@0x3ff97b 

More reports at:
http://crash-stats.mozilla.com/report/list?range_value=4&range_unit=weeks&signature=atioglxx.dll%400xd772c

Comment 1

[Bas Schouten (:bas.schouten)]

Don't some websites that do fancy stuff for HTML5 create an OpenGL context for checking the presence of some features?

I really don't think D3D10/D2D load that driver.

Comment 2

[Benoit Jacob [:bjacob] (mostly away)]

Yes, for example modernizr does create a WebGL context just because it can.

Comment 3

[Benoit Jacob [:bjacob] (mostly away)]

However, the stack traces don't point to WebGL:

0 		@0x102eb540 	
1 	atioglxx.dll 	atioglxx.dll@0xd772c 	
2 	atioglxx.dll 	atioglxx.dll@0xd7751 	
3 	atioglxx.dll 	atioglxx.dll@0xd7e74 	
4 	atioglxx.dll 	atioglxx.dll@0xd76ef 	
5 	xul.dll 	nsThebesRegion::QueryInterface 	gfx/src/thebes/nsThebesRegion.cpp:42


The code at nsThebesRegion.cpp:42 is

   NS_IMPL_ISUPPORTS1(nsThebesRegion, nsIRegion)


I don't know how that can result in a call to atioglxx.dll. It would be very helpful if someone could reproduce this in a debug build.

Comment 4

[Bas Schouten (:bas.schouten)]

(In reply to comment #3)
> However, the stack traces don't point to WebGL:
> 
> 0         @0x102eb540     
> 1     atioglxx.dll     atioglxx.dll@0xd772c     
> 2     atioglxx.dll     atioglxx.dll@0xd7751     
> 3     atioglxx.dll     atioglxx.dll@0xd7e74     
> 4     atioglxx.dll     atioglxx.dll@0xd76ef     
> 5     xul.dll     nsThebesRegion::QueryInterface    
> gfx/src/thebes/nsThebesRegion.cpp:42
> 
> 
> The code at nsThebesRegion.cpp:42 is
> 
>    NS_IMPL_ISUPPORTS1(nsThebesRegion, nsIRegion)
> 
> 
> I don't know how that can result in a call to atioglxx.dll. It would be very
> helpful if someone could reproduce this in a debug build.

Not having symbols for atioglxx.dll might very well cause frames below it to be invalid if they compile with frame pointer omission. It's still very likely to be inside the ATI GL drivers.

Comment 5

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

I can't find the page that has what driver version 6.14.10.6747 corresponds to, but let's just blocklist this and move on.

Comment 6

[Benoit Jacob [:bjacob] (mostly away)]

Only one guy crashed with beta11, and he force-enabled (his driver version 8.401.x.x is blacklisted). So this crash is fixed by the blacklisting of old ATI drivers.

@ Bas ---> see, this is another good reason to blacklist old drivers.