527540 - nvLsp.dll (NVIDIA Application Filter) 2.2.0.7111/2.2.0.7305/2.2.0.7313/2.2.0.7316/2.2.0.7325 crashes firefox when opening new tabs and loading sites

Status: RESOLVED INCOMPLETE

(Plugins Graveyard :: Nvidia ActiveArmor, defect)

Description

[4572301]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.1.5) Gecko/20091102 Firefox/3.5.5 GTB5 (.NET CLR 3.5.30729)
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.1.5) Gecko/20091102 Firefox/3.5.5 GTB5 (.NET CLR 3.5.30729)

Hangs when browsing or putting ona new toolbar.
3.5 update didn't help.

Reproducible: Sometimes

Steps to Reproduce:
1.Put on bars using middle mouse button from bookmarks.
2. Hangs on second one, or 5th... or 10th

Actual Results:  
Also hangs randomly when browsing. Especially when putting on a new toolbar


using:
kaspersky internet security 2009
vista windows mail
original vista home premium all updated
nvidia 650i chipset all updated

Comment 1

[Matthias Versen [:Matti]]

Do you already tried :
http://support.mozilla.com/en-US/kb/Safe+Mode

Comment 2

[4572301]

I've tried reseting settings but it didn't help.

Here is a bug report.

Add-ons: {972ce4c6-7e08-4474-a285-3208198ce6fd}:3.5.5
BuildID: 20091102152451
CrashTime: 1257847647
Email: [e-mail]
InstallTime: 1257785289
ProductName: Firefox
SecondsSinceLastCrash: 54
StartupTime: 1257847641
Theme: classic/1.0
Throttleable: 1
URL: about:blank
Vendor: Mozilla
Version: 3.5.5

This report also contains technical information about the state of the application when it crashed.

Comment 3

[Matthias Versen [:Matti]]

You reported a hang but you have posted a crash id.
Is comment #0 wrong or is this an unrelated crash ?

and we need the crash ID from about:crashes :
https://developer.mozilla.org/en/How_to_get_a_stacktrace_for_a_bug_report

Comment 4

[4572301]

Yeah it's just firefox that crashes. 
Sorry my mistake.

Comment 5

[4572301]

bp-a227aafd-fa96-4ec8-8962-799732091110	10/11/2009	18:38
bp-9d825109-98ff-497b-a917-eb8102091110	10/11/2009	18:37
bp-3edbba6a-3a34-423e-ba31-57d1b2091110	10/11/2009	17:52
bp-d38553f1-c61c-42be-8bd1-4b0002091110	10/11/2009	17:51
bp-9a242c77-61ad-492e-a885-b6f9d2091110	10/11/2009	16:55

Comment 6

[timeless]

so, 4/6 crashes are in a third party LSP, you should either:
A. uninstall the LSP
B. yell at the LSP vendor

we intend to block out all LSPs in the near future to protect you from this junk

Signature	RtlpWaitOnCriticalSection
UUID	a227aafd-fa96-4ec8-8962-799732091110
Time 	2009-11-10 09:38:42.89768
Uptime	80
Last Crash	85 seconds before submission
Product	Firefox
Version	3.5.5
Build ID	20091102152451
Branch	1.9.1
OS	Windows NT
OS Version	6.0.6001 Service Pack 1
CPU	x86
CPU Info	GenuineIntel family 6 model 15 stepping 6
Crash Reason	EXCEPTION_ACCESS_VIOLATION
Crash Address	0x14
User Comments	
Processor Notes 	
Related Bugs

Crashing Thread
Frame 	Module 	Signature [Expand] 	Source
0 	ntdll.dll 	RtlpWaitOnCriticalSection 	
1 	ntdll.dll 	EtwEventEnabled 	
2 	nvLsp.dll 	nvLsp.dll@0x5d0d 	
3 	ws2_32.dll 	GetProtocolStateForFamily 	
4 	ws2_32.dll 	LookupAddressForName 	
5 		@0x625fc57 	
6 	ws2_32.dll 	GetIp4Address 	
7 	ws2_32.dll 	getaddrinfo 	
8 	nspr4.dll 	PR_GetAddrInfoByName 	nsprpub/pr/src/misc/prnetdb.c:2042
9 	xul.dll 	nsHostResolver::ThreadFunc 	netwerk/dns/src/nsHostResolver.cpp:884
10 	nspr4.dll 	_PR_NativeRunThread 	nsprpub/pr/src/threads/combined/pruthr.c:426
11 	nspr4.dll 	pr_root 	nsprpub/pr/src/md/windows/w95thred.c:122
12 	mozcrt19.dll 	_callthreadstartex 	obj-firefox/memory/jemalloc/src/threadex.c:348
13 	mozcrt19.dll 	_threadstartex 	obj-firefox/memory/jemalloc/src/threadex.c:326
14 	kernel32.dll 	BaseThreadInitThunk 	
15 	ntdll.dll 	__RtlUserThreadStart 	
16 	ntdll.dll 	_RtlUserThreadStart 	

nvLsp.dll 	2.2.0.7316 	E20FEBDCC19A472ABA531FB44DC2265D1 	nvLSP.pdb

someone will mark this bug invalid in a bit once you've accepted the explanation.

1/6 of your dumps wasn't processable by our report system and one wasn't processed (not sure why).

Comment 7

[4572301]

Right...
So it must be nVIDIA network access manager than.
That is what I get for updating drivers...
I'll try to work around it somehow 

Thanks for your help

Comment 8

[timeless]

glad to be of service, but please do complain to nVidia.

Comment 9

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Module correlation data for RtlpWaitOnCriticalSection (currently the #4 Firefox 3.5.5 topcrash) shows:

     12% (135/1082) vs.   1% (562/86962) nvLsp.dll

In other words, nvLsp.dll is responsible for a bunch of crashes at this signature, plus, in:

  ntdll.dll@0x38c39|EXCEPTION_ACCESS_VIOLATION (150 crashes)
     76% (114/150) vs.   1% (562/86962) nvLsp.dll

and also:
  @0x1061c25|EXCEPTION_ACCESS_VIOLATION (10 crashes)
    100% (10/10) vs.   1% (562/86962) nvLsp.dll
    100% (10/10) vs.   3% (3022/86962) lgscroll.dll

in other words, that's showing likely responsibility for 0.3% of the crashes in Firefox 3.5.5.  Maybe fodder for DLL blocklisting?

Comment 10

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

(In reply to comment #9)
>   ntdll.dll@0x38c39|EXCEPTION_ACCESS_VIOLATION (150 crashes)
>      76% (114/150) vs.   1% (562/86962) nvLsp.dll

And, like the RtlpWaitOnCriticalSection crashes, nvLsp.dll shows up in the stacks of crashes with that signature (ntdll.dll@0x38c39).  I just loaded the 10 most recent such crashes, and 7 of them had ntdll.dll in the stack (the other three had the stack going through other third-party dlls):

bp-6182dcbc-49b5-447d-bc54-6a6bc2091111
bp-03d8ddcc-ca4a-4a84-8dfc-568ba2091111
bp-fb6362a2-b5fb-4936-b3a1-6c53d2091111
bp-0cdebc40-4dbb-40f9-96c3-84ccd2091111
bp-9d8288d8-48c5-4b99-82f6-872d12091111
bp-d5945389-5281-4240-ac2f-8a5ac2091111
bp-ca6d2b25-8f1f-4ad7-80ec-5b93a2091111

So nvLsp.dll seems like the cause and not something correlated with the cause.

So I'm going to reopen this as a *candidate* for DLL blocklisting, although I don't know what nvLsp.dll does, and we should find out before blocklisting.


Also, adding additional versions to the bug summary based on the interesting-modules-with-versions report in http://people.mozilla.com/crash_analysis/ ; I'd note that the versions reports show that the older versions seem to be ok, but the newer versions seem to be causing crashes:

     12% (135/1082) vs.   1% (562/86962) nvLsp.dll
          0% (1/1082) vs.   0% (91/86962) 2.2.0.465
          0% (0/1082) vs.   0% (32/86962) 2.2.0.6781
          0% (3/1082) vs.   0% (85/86962) 2.2.0.6793
          0% (0/1082) vs.   0% (72/86962) 2.2.0.6796
          0% (0/1082) vs.   0% (4/86962) 2.2.0.7109
          3% (30/1082) vs.   0% (39/86962) 2.2.0.7111
          3% (30/1082) vs.   0% (77/86962) 2.2.0.7305
          0% (0/1082) vs.   0% (13/86962) 2.2.0.7313
          7% (71/1082) vs.   0% (149/86962) 2.2.0.7316

...

     76% (114/150) vs.   1% (562/86962) nvLsp.dll
          0% (0/150) vs.   0% (91/86962) 2.2.0.465
          0% (0/150) vs.   0% (32/86962) 2.2.0.6781
          0% (0/150) vs.   0% (85/86962) 2.2.0.6793
          0% (0/150) vs.   0% (72/86962) 2.2.0.6796
          0% (0/150) vs.   0% (4/86962) 2.2.0.7109
          2% (3/150) vs.   0% (39/86962) 2.2.0.7111
         24% (36/150) vs.   0% (77/86962) 2.2.0.7305
          3% (4/150) vs.   0% (13/86962) 2.2.0.7313
         47% (71/150) vs.   0% (149/86962) 2.2.0.7316

Comment 11

[timeless]

dbaron, the reporter was very helpful in comment 7 "nvidia network access manager" which leads to http://groups.google.com/group/microsoft.public.windowsxp.network_web/browse_thread/thread/47ccfdec52028db3?pli=1

http://www.google.com/search?hl=en&safe=active&q=site%3AnVIDIA.com+network+access+manager+&aq=f&oq=&aqi=

I can't find *anything* good about this software :o.

http://forums.nvidia.com/index.php?showtopic=100637&pid=947981&st=60&#entry947981

Comment 12

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

After uninstalling the nVidia Network Manager, I have had no crashes. 

Prior to uninstalling, Firefox would crash every 10 minutes or so. (Many times it would crash within a minute of restarting from a previous crash)

Comment 13

[chris hofmann]

in early 3.6b5 data this has now ranked #2 crash overall and it appears concentrated on vista build 7600.  this could be mostly reflective of very early beta testers running mostly latest versions of nearly everything.

checking --- 20091217-crashdata.csv ntdll.dll@0x38c39
release total-crashes
              ntdll.dll@0x38c39 crashes

all     227066  371     0.00163389
3.0.15  17135   3       0.00017508
3.0.16  22875           0
3.5.5   47876   54      0.00112791
3.5.6   80444   150     0.00186465
3.6b5   990     24      0.0242424
3.6b4   21208   132     0.00622407
3.6b3   670             0
3.6b2   739     1       0.00135318
3.6b1   2144            0


os breakdown
370     0.997305        Windows NT6.1.7600
1       0.00269542      Windows NT6.1.7201

Comment 14

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

Pulled more recent data (from Jan 5th 2010):

3.5.6:

ntdll.dll@0x38c39 (topcrash #45)
     66% (163/248) vs.   1% (678/103611) nvLsp.dll
         14% (35/248) vs.   0% (77/103611) 2.2.0.7305
          1% (2/248) vs.   0% (5/103611) 2.2.0.7308
          4% (11/248) vs.   0% (27/103611) 2.2.0.7313
         46% (115/248) vs.   0% (235/103611) 2.2.0.7316

RtlpWaitOnCriticalSection | RtlEqualString (topcrash #140)
     48% (53/110) vs.   1% (678/103611) nvLsp.dll
         26% (29/110) vs.   0% (66/103611) 2.2.0.7111
          0% (0/110) vs.   0% (77/103611) 2.2.0.7305
          0% (0/110) vs.   0% (5/103611) 2.2.0.7308
          0% (0/110) vs.   0% (27/103611) 2.2.0.7313
         21% (23/110) vs.   0% (235/103611) 2.2.0.7316

RtlpWaitOnCriticalSection | RtlAddAccessAllowedAce | nvLsp.dll@0x5d0d (#300+)
    100% (50/50) vs.   1% (678/103611) nvLsp.dll
         48% (24/50) vs.   0% (66/103611) 2.2.0.7111
          0% (0/50) vs.   0% (77/103611) 2.2.0.7305
          0% (0/50) vs.   0% (5/103611) 2.2.0.7308
          0% (0/50) vs.   0% (27/103611) 2.2.0.7313
         52% (26/50) vs.   0% (235/103611) 2.2.0.7316


3.6b5:

ntdll.dll@0x38c39 (topcrash #6)
     77% (127/165) vs.   1% (317/21747) nvLsp.dll
          7% (11/165) vs.   0% (26/21747) 2.2.0.7305
          2% (4/165) vs.   0% (4/21747) 2.2.0.7308
          7% (12/165) vs.   0% (17/21747) 2.2.0.7313
         61% (100/165) vs.   1% (218/21747) 2.2.0.7316

RtlpWaitOnCriticalSection | RtlEqualString (topcrash #33)
     94% (75/80) vs.   1% (317/21747) nvLsp.dll
         14% (11/80) vs.   0% (18/21747) 2.2.0.7111
          0% (0/80) vs.   0% (26/21747) 2.2.0.7305
          0% (0/80) vs.   0% (4/21747) 2.2.0.7308
          0% (0/80) vs.   0% (17/21747) 2.2.0.7313
         80% (64/80) vs.   1% (218/21747) 2.2.0.7316

RtlpWaitOnCriticalSection | RtlpDeCommitFreeBlock | nvLsp.dll@0x5d0d (#268)
    100% (20/20) vs.   1% (317/21747) nvLsp.dll
        100% (20/20) vs.   1% (218/21747) 2.2.0.7316


There's a few others on the 3.6b5 branch that look like they might be correlated with nvLsp.dll, but there's to few crashes to know for sure.

Analysis to follow

Comment 15

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

I should also add that for the crashes listed in the previous comment, nvLsp.dll was the one with the heaviest correlation (apart from system libraries).

Seems like the same set of versions as already listed in the summary are still the ones causing crashes:

2.2.0.7111
2.2.0.7305
2.2.0.7313
2.2.0.7316


I'll see what we can do about contacting nVidia about this.

Comment 16

[Michael]

I have the same problem on my pc
I installed W7 and Firefox 3.5.7 on both computers.
Both have the newest NForce drivers installed with the same chipset.
On her PC firefox never crashes and on mine it does all the time.
After looking in the Nvidia control panel I found out First packet service was on on my pc and on hers it was off.
So after turning First packet service off on my pc the crashes are gone.
So no need to deinstall the network manager.

I hope this helps for some people..

Comment 17

[chris hofmann]

vlad,  do you have any contacts at NVIDIA  that we could check with?

Comment 18

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

Vlad has contacted nVidia and tried to get pointers to the right people in the right groups. So things are progressing here.

Comment 19

[chris hofmann]

#5 3.6 Rc1 signature looks like [@ RtlpWaitOnCriticalSection | RtlEqualString ] 

http://crash-stats.mozilla.com/report/list?range_value=2&range_unit=weeks&signature=RtlpWaitOnCriticalSection%20|%20RtlEqualString&version=Firefox%3A3.6

Comment 20

[Eric Lackner]

Based on the crash reports which led me to this thread, I did some google searches and found this other thread (below) which looks to have helped me with the crashes myself and seemingly the other people in this thread have had with Nvidia and Windows7 w/FF:

http://www.utorrent.com/forum/viewtopic.php?id=60266

Looks like nvLsp.dll is part of Nvidia's "FirstPacket" optimizer which is in their Nvidia Control Panel.  I disabled this feature (w/o rebooting), and so far I haven't had another browser crash.  I started opening tabs and switching the sites with my bookmarks as fast as possible and so far no problems.  I also opened up Google Chrome to cross-test since it was even less stable than FF, and it too did not crash at all.

What I haven't tested is specifically adding FF, etc., to the FirstPacket optimized program list.  I'll probably do that soon, but I want to enjoy the moment w/o crashes for a bit.  It's been a rough couple of months on Win7.  I was fearing I'd have to start using IE regularly (./shudder).

Hope this helps.  I'll update again if things go downhill.

Comment 21

[Joshua Rohr]

Eric, 

Ive been running W7 64 bit with FF as my main browser, and this issue has been occurring to much. 

I also have the Nvidia control panel, and will try adding FF to the "FirstPacket" priority list. 

I will also update the post when testing with the "firstpacket" totally disabled is complete.

Comment 22

[Richard]

What Eric mentioned about first packet seemed to have fixed my problem.

When i enable nvidia first packet, my browser crashes. (including google chrome)

when i disable it, i don't have any crashes at all.

The problem seems to be nvidia related. not firefox.

thanks this really helped.

Comment 23

[Kevin]

I am using Win7 64bit, 8500 GT. It was crashing like crazy in FF but after disabling FirstPacket stabilized it. Thank you!!

Comment 24

[chris hofmann]

anyone heard from NVIDIA folks lately?

Comment 26

[Whitedog Warren]

Using Win7 64bits, geforce 8300 on MB chipset, geforce GTS 250 graphic card.

FF crashes all the time, on some web sites I have to relaunch FF about 10 times in a row.

I looked into nvidia control panel : FirstPacket was "on".


I'm hopefully trying with FirstPacket disabled.
Thanks for your help.

Comment 27

[Whitedog Warren]

Seems to have no more trouble with FirstPacket disabled.

Comment 28

[Eric Lackner]

Sorry to hear that.  I haven't had a crash since the day I disabled FirstPacket.  Looks like you have some other problem then, unfortunately.

Comment 29

[Richard]

Disabling First Packet worked for me. No crashes for months now. Try updating your drivers and then disabling First Packet.

Comment 30

[Whitedog Warren]

I'm sorry I didn't express myself clearly ( I'm french ).
I disabled FirstPacket and have no more problem.
I tried few pages, full of flash and java stuff, that previously crashed firefox almost every time.
For the moment it works, even with many tabs.

I enjoy it a few days ( after months of nightmare ) before trying to add FF to the
"FirstPacket" priority list.

Thanks for your help.

Comment 33

[timeless]

We're now tracking such bugs. This doesn't mean it's something we can fix, merely something we hope to be able to point vendors to so they can investigate. This is an automated message.

Comment 34

[Hacker10001]

[Tracking Requested - why for this release]:

Comment 35

[Benjamin Smedberg]

Closing old bugs in the Plugins component. We aren't going to track issues in 3rd-party plugins in the Mozilla bug tracker. In addition, support for NPAPI plugins will be removed at the end of this year; for more details see the post at https://blog.mozilla.org/futurereleases/2015/10/08/npapi-plugins-in-firefox/

If there is a serious bug in Firefox, it needs to be filed in the "Core" product, "Plug-Ins" component.