1076917 - crash in npTNT2.dll

Status: RESOLVED FIXED

(Toolkit :: Blocklist Policy Requests, defect)

Description

[Tracy Walker [:tracy]]

[Tracking Requested - why for this release]:

This is a startup crash on 33beta in high enough volume to be considered for tracking. It is also present on release and in much less volume on Nightly and Aurora.

https://crash-stats.mozilla.com/report/index/14647b78-aee0-46ab-b700-570ce2140929

Frame 	Module 	Signature 	Source
Ø 0 	npTNT2.dll 	npTNT2.dll@0x2099 	
1 	xul.dll 	CallNPMethodInternal 	dom/plugins/base/nsJSNPRuntime.cpp
2 	xul.dll 	CallNPMethod 	dom/plugins/base/nsJSNPRuntime.cpp
3 	mozjs.dll 	js::Invoke(JSContext*, JS::CallArgs, js::MaybeConstruct) 	js/src/vm/Interpreter.cpp
4 	mozjs.dll 	Interpret 	js/src/vm/Interpreter.cpp
5 	mozjs.dll 	js::RunScript(JSContext*, js::RunState&) 	js/src/vm/Interpreter.cpp
6 	mozjs.dll 	js::Invoke(JSContext*, JS::CallArgs, js::MaybeConstruct) 	js/src/vm/Interpreter.cpp
7 	mozjs.dll 	js::Invoke(JSContext*, JS::Value const&, JS::Value const&, unsigned int, JS::Value const*, JS::MutableHandle<JS::Value>) 	js/src/vm/Interpreter.cpp
8 	mozjs.dll 	JS_CallFunctionValue(JSContext*, JS::Handle<JSObject*>, JS::Handle<JS::Value>, JS::HandleValueArray const&, JS::MutableHandle<JS::Value>) 	js/src/jsapi.cpp
9 	xul.dll 	nsXPCWrappedJSClass::CallMethod(nsXPCWrappedJS*, unsigned short, XPTMethodDescriptor const*, nsXPTCMiniVariant*) 	js/xpconnect/src/XPCWrappedJSClass.cpp

Comment 1

[Sylvestre Ledru [:Sylvestre]]

David, could you help here? thanks

Comment 2

[(Away)]

Attachment: nptnt2

Please mark as checkin-needed if you need this while I'm away.

Comment 3

[Sylvestre Ledru [:Sylvestre]]

Thanks David, We will take in the RC release
Enjoy your time off!

Comment 4

[Benjamin Smedberg]

Comment on attachment 8499220 [details] [diff] [review]
nptnt2

This is an NPAPI plugin, so the DLL blocklist isn't the right option, and AFAIK this is a plugin that some users still use on the Turner Television family of websites.

Comment 5

[Benjamin Smedberg]

It's odd that this is an in-process plugin in all cases. Tracy, can you check to see whether there is an addon in common across all these crashes? Here is a supersearch link for all the nptnt2.dll crashes:

https://crash-stats.mozilla.com/search/?signature=~nptnt2.dll&_facets=signature&_facets=process_type&_columns=date&_columns=signature&_columns=product&_columns=version&_columns=build_id&_columns=platform

Comment 6

[Tracy Walker [:tracy]]

These are not present in all the signatures, but in many that I scanned through. Common enough to make it stand out.

[e-mail]  2.0.0.1868
or
[e-mail]  2.0.0.1855

Comment 7

[Sylvestre Ledru [:Sylvestre]]

Looks like we are going to ship 33 with this bug. Maybe we could take a patch in 33.1 if we have a low risk patch.

Comment 8

[rctgamer3]

This DLL is part of the FindWide toolbar and is silently installed. Already filed bug 1082758 on Tuesday for blocking both the add-on and the plugin.

Comment 9

[Sylvestre Ledru [:Sylvestre]]

No activity, looks like that 33.1 will ship with this issue.
Tracking in 34.

Comment 10

[Liz Henry (:lizzard) (relman/hg->git project)]

#4 topcrash for Firefox 34.0b7 with 437/28000 crashes.   This is still a startup crash. 

I also noticed the toolbar extensions that Tracy mentions, findwide and freshy.

Comment 11

[Liz Henry (:lizzard) (relman/hg->git project)]

Should we be contacting findwide and/or Freshy about the plugin, since we're not blocklisting it and it is still associated with a startup crash?

Comment 12

[Lukas Blakk [:lsblakk] use ?needinfo]

Moving the ni? to Lawrence who is on point for 34.  Findwide/Freshy searches only point to potential malware so I'm not sure we've got an option to contact people here.

Comment 13

[Liz Henry (:lizzard) (relman/hg->git project)]

So, maybe we should either block the plugin despite the issues in Comment 4, or contact someone from Turner to see if they can help.

Comment 14

[(Away)]

I found the affected version of the binary but I haven't had any crashes despite various attempts. They may be crashing under some corner case. This is supported by crash-stats in that we're generally getting one crash per installtime, so at least it's not a permanent startup-crash. And we're getting reports going back to Firefox 3.6.

Comment 15

[Lawrence Mandel [:lmandel] (use needinfo)]

I have e-mailed the listed support address for FindWide ([e-mail]) about this issue. I'll report back if I get a response.

Comment 16

[FindWide]

Hello. We are the developers of the FindWide toolbar. We will look into the problem promptly.

Comment 17

[FindWide]

Hello. We have added more checks in an attempt to avoid this problem. We will release an update for new installations next week if it passes quality assurance. We have a release freeze now due to the holidays. Happy holidays!

Comment 18

[FindWide]

Hello. We released a new version of the FindWide toolbar on 2015-01-05 that added more checks in an attempt to avoid this problem. We plan another release next week to fix a specific NPAPI problem that we found during our audit. How can we find out if there are less crashes? Thank you!

Comment 19

[Lawrence Mandel [:lmandel] (use needinfo)]

Kairo - Can you help determine whether the FindWide fix reduced the number of crashes with this signature?

Comment 20

[FindWide]

Hello. Here is a download link to a newer version of the FindWide toolbar as of 2015-01-14 that fixes the NPAPI problem in comment #18.

http://mirror.mirror-files.com/tnt2/11287/FindWide-TB11287-e-f-s-w-r-p.exe
(Do not change the filename.)

Please let us know if we can be of any additional assistance. Thank you!

Comment 21

[Robert Kaiser]

(In reply to Lawrence Mandel [:lmandel] (use needinfo) from comment #19)
> Kairo - Can you help determine whether the FindWide fix reduced the number
> of crashes with this signature?

The crashes have seriously decreased and there seem to be none of this signature with versions 2.0.0.1923 and 2.0.0.1928 of npTNT2.dll any more.

Comment 22

[Jorge Villalobos [:jorgev] (he/him)]

This appears to have been resolved.