812307 - crash in DataMngrHlpFF17.dll @ nsObserverEnumerator::Release

Status: RESOLVED INVALID

(Firefox :: Extension Compatibility, defect)

Description

[Marcia Knous [:marcia]]

This bug was filed from the Socorro interface and is 
report bp-7a07a39e-9268-4f7b-ada4-64fd12121115 .
============================================================= 

Seen while looking at early FF 17 B6 crash stats. 

Links to the crashes: https://crash-stats.mozilla.com/report/list?signature=arena_dalloc_small%20|%20je_free%20|%20nsVoidArray::~nsVoidArray%28%29.

Most of the reports I looked at had DataMngrHlpFF17.dll in the top part of the stack. Manual correlations are not yet picking up this signature, so likely tomorrow there will be more data to look at.

Frame 	Module 	Signature 	Source
0 	mozglue.dll 	arena_dalloc_small 	memory/mozjemalloc/jemalloc.c:4508
1 	mozglue.dll 	je_free 	memory/mozjemalloc/jemalloc.c:6565
2 	xul.dll 	nsVoidArray::~nsVoidArray 	obj-firefox/xpcom/build/nsVoidArray.cpp:344
3 	xul.dll 	nsObserverEnumerator::Release 	xpcom/ds/nsObserverList.cpp:113
4 	DataMngrHlpFF17.dll 	DataMngrHlpFF17.dll@0x584e 	
5 	xul.dll 	nsHttpHandler::NotifyObservers 	netwerk/protocol/http/nsHttpHandler.cpp:492
6 	xul.dll 	mozilla::net::nsHttpChannel::ProcessResponse 	netwerk/protocol/http/nsHttpChannel.cpp:1187
7 	xul.dll 	mozilla::net::nsHttpChannel::OnStartRequest 	netwerk/protocol/http/nsHttpChannel.cpp:4776
8 	xul.dll 	nsInputStreamPump::OnStateStart 	netwerk/base/src/nsInputStreamPump.cpp:417
9 	xul.dll 	nsInputStreamReadyEvent::Run 	xpcom/io/nsStreamUtils.cpp:82
10 	xul.dll 	nsThread::ProcessNextEvent 	xpcom/threads/nsThread.cpp:624
11 	xul.dll 	mozilla::ipc::MessagePump::Run 	ipc/glue/MessagePump.cpp:117
12 	xul.dll 	MessageLoop::RunHandler 	ipc/chromium/src/base/message_loop.cc:201
13 	xul.dll 	mozilla::Selection::Clear 	layout/generic/nsSelection.cpp:3624
14 	xul.dll 	nsAppShell::Run 	widget/windows/nsAppShell.cpp:232

Comment 1

[Scoobidiver (away)]

It seems correlated to the {1FD91A9C-410C-4090-BBCC-55D3450EF433}/DataMngr extension, a known crapware.

Comment 2

[Alex Keybl [:akeybl]]

We were previously unable to block this crapware in bug 665775 :(

Comment 3

[Scoobidiver (away)]

It's #6 top browser crasher in 17.0b6.
Here are correlations:
    100% (43/43) vs.   1% (44/8136) DataMngrHlpFF17.dll (1.0.0.1)
     95% (41/43) vs.   4% (335/8136) {1FD91A9C-410C-4090-BBCC-55D3450EF433} (DataMngr 1.0)
     60% (26/43) vs.   1% (76/8136) {f34c9277-6577-4dff-b2d7-7d58092f272f} (Search-Results Toolbar 1.0.0.12)

Comment 4

[Alex Keybl [:akeybl]]

1FD91A9C-410C-4090-BBCC-55D3450EF433 is the jzip.toolbar according to bug 665775.

Anthony, Juan, or Marcia - can you please try to reproduce using http://www.jzip.com/? Also including bsmedberg to try to take a stab at DLL blocking again (or discuss alternatives since this has bitten us twice).

Comment 5

[Scoobidiver (away)]

(In reply to Alex Keybl [:akeybl] from comment #4)
> 1FD91A9C-410C-4090-BBCC-55D3450EF433 is the jzip.toolbar according to bug
> 665775.
It's DataMngr (click More system details in https://support.mozilla.org/questions/816181) and comes bundled with many pieces of Bandoo's software including jZip.Toolbar (see http://forums.spybot.info/showthread.php?t=62634).
I asked to block the extension in bug 812445 but I am not confident that it will fix the crash as bug 665775 suggested it for other related extensions (MediaBar and Searchqu Toolbar).

Comment 6

[juan becerra [:juanb]]

I've been testing this on Fx17b6 by installing the software from jzip.com, which installs datamngr 1.0 (disabled in Fx17) and searchqu toolbar 4.6.1.01. I also installed the incredibar games and toolware. Testing involved browsing Flash sites, going into and out of private browsing mode, playing a Jewels type of game, upgrading from Fx17b5.

So far no luck.

Comment 7

[Scoobidiver (away)]

Juan, do you have DataMngrHlpFF17.dll installed somewhere on your test computer?

Comment 8

[juan becerra [:juanb]]

(In reply to Scoobidiver from comment #7)
> Juan, do you have DataMngrHlpFF17.dll installed somewhere on your test
> computer?

There isn't one corresponding to ...FF17.dll and there are ones for ff3 - ff16 so there might be an updated version somewhere which would have the ...ff17.dll although I haven't been able to find one or trigger an update.

Comment 9

[Alex Keybl [:akeybl]]

We should continue all possible testing here, but discussion of how to blocklist is ongoing in Scoobidiver's bug 812456. I'm concerned about blindly blocklisting this Extension, when it's used by so many add-ons (and could cause worse instability).

Comment 10

[Marcia Knous [:marcia]]

My first attempt at reproducing it on Win XP didn't work either - as Juan notes in Comment 6, datamngr is disabled, and I don't have DataMngrHlpFF17.dll on my machine either. Will have to hunt around to see if I can find a version of the malware that corresponds to that dll.

Some of the URLs look as if people were in the process of downloading something when they crashed, so that might be another avenue to pursue if we are able to get the right version.

Comment 11

[Scoobidiver (away)]

It seems to happen with an old version of datamngr.dll that we can't blocklist:
    100% (1752/1752) vs.   5% (1761/36281) DataMngrHlpFF17.dll (1.0.0.1)
     80% (1410/1752) vs.   7% (2368/36281) datamngr.dll
         80% (1410/1752) vs.   6% (2277/36281) 1.0.0.1
          0% (0/1752) vs.   0% (1/36281) 3.0.0.2736
          0% (0/1752) vs.   0% (1/36281) 3.0.0.2859
          0% (0/1752) vs.   0% (2/36281) 3.0.0.2950
          0% (0/1752) vs.   0% (12/36281) 3.0.0.3037
          0% (0/1752) vs.   0% (2/36281) 3.0.0.3171
          0% (0/1752) vs.   0% (28/36281) 3.0.0.3206
          0% (0/1752) vs.   0% (5/36281) 3.0.0.3369
          0% (0/1752) vs.   0% (5/36281) 3.0.0.42011
          0% (0/1752) vs.   0% (1/36281) 3.0.0.46577
          0% (0/1752) vs.   0% (6/36281) 3.0.0.46593
          0% (0/1752) vs.   0% (1/36281) 3.0.0.50653
          0% (0/1752) vs.   0% (4/36281) 3.0.0.51994
          0% (0/1752) vs.   0% (10/36281) 3.0.0.53061
          0% (0/1752) vs.   0% (2/36281) 3.0.0.55914
          0% (0/1752) vs.   0% (11/36281) 3.0.0.56544

Comment 12

[Robert Kaiser]

This now is the #2 crash on 17.0b6, I fear it will be just as high and bad on release, if not even worse, as the release population is more likely to have such those toolbars installed.

Comment 13

[Scoobidiver (away)]

The crash signature is different in 17.0 but as high volume as in 17.0 Beta: #4 top browser crasher.
Here is a stack trace:
Frame 	Module 	Signature 	Source
0 	mozalloc.dll 	mozalloc_abort 	memory/mozalloc/mozalloc_abort.cpp:23
1 	mozalloc.dll 	mozalloc_handle_oom 	memory/mozalloc/mozalloc_oom.cpp:27
2 	mozalloc.dll 	moz_xmalloc 	memory/mozalloc/mozalloc.cpp:59
3 	xul.dll 	nsTArray_base<nsTArrayDefaultAllocator>::EnsureCapacity 	obj-firefox/dist/include/nsTArray-inl.h:119
4 	xul.dll 	nsObserverList::FillObserverArray 	xpcom/ds/nsObserverList.cpp:71
5 	xul.dll 	nsObserverService::NotifyObservers 	xpcom/ds/nsObserverService.cpp:149
6 	DataMngrHlpFF17.dll 	DataMngrHlpFF17.dll@0x9fa4

In 16.0.2, it happens with DataMngrHlpFF16.dll and in 15.0 with DataMngrHlpFF15.dll but it is a low volume crash.

It started spiking in 17.0 Beta on November 15th at 12:40 UTC probably with a new version of a Discordia/Imesh/Bandoo Media/MusicLab's application to take into account 17.0.

More reports also at:
https://crash-stats.mozilla.com/report/list?signature=mozalloc_abort%28char+const*+const%29+|+mozalloc_handle_oom%28unsigned+int%29+|+moz_xmalloc+|+nsTArray_base%3CnsTArrayDefaultAllocator%3E%3A%3AEnsureCapacity%28unsigned+int%2C+unsigned+int%29+|+nsObserverList%3A%3AFillObserverArray%28nsCOMArray%3CnsIObserver%3E%26%29

Comment 14

[Alex Keybl [:akeybl]]

(In reply to Marcia Knous [:marcia] from comment #10)
> My first attempt at reproducing it on Win XP didn't work either - as Juan
> notes in Comment 6, datamngr is disabled, and I don't have
> DataMngrHlpFF17.dll on my machine either. Will have to hunt around to see if
> I can find a version of the malware that corresponds to that dll.
> 
> Some of the URLs look as if people were in the process of downloading
> something when they crashed, so that might be another avenue to pursue if we
> are able to get the right version.

Would you mind taking another look around today to see if we can find a toolbar that does install this DLL? Let's start with Scoobidiver's suggestion of testing "Discordia/Imesh/Bandoo Media/MusicLab".

Comment 15

[Scoobidiver (away)]

The DataMngr extension usually comes bundled with other toolbars. Try some of them:
     92% (1234/1338) vs.   9% (1706/18601) {1FD91A9C-410C-4090-BBCC-55D3450EF433} (DataMngr)
     59% (785/1338) vs.   5% (848/18601) {f34c9277-6577-4dff-b2d7-7d58092f272f} (Search-Results Toolbar)
     20% (262/1338) vs.   5% (860/18601) [e-mail] (Incredibar Toolbar)
     18% (235/1338) vs.   4% (744/18601) {336D0C35-8A85-403a-B9D2-65C292C39087} (Web Assistant)
     19% (254/1338) vs.   5% (1023/18601) [e-mail] (Yontoo Layers)
     16% (220/1338) vs.   6% (1138/18601) {635abd67-4fe9-1b23-4f01-e679fa7484c1} (Yahoo! Toolbar, https://addons.mozilla.org/addon/2032)
     10% (139/1338) vs.   2% (405/18601) [e-mail] (TornTV Toolbar?)
     12% (167/1338) vs.   5% (861/18601) {EEE6C361-6118-11DC-9C72-001320C79847} (SweetIM Toolbar for Firefox)
     89% (1194/1338) vs.  81% (15156/18601) [e-mail] (Mozilla Labs - Test Pilot, https://addons.mozilla.org/addon/13661)
     12% (156/1338) vs.   4% (797/18601) [e-mail] (IDM CC, https://addons.mozilla.org/addon/6973)
     98% (1310/1338) vs.  92% (17107/18601) {972ce4c6-7e08-4474-a285-3208198ce6fd} (Default, https://addons.mozilla.org/addon/8150)
      9% (115/1338) vs.   3% (534/18601) [e-mail] (Default Tab)

Comment 16

[Marcia Knous [:marcia]]

I have spent the last hour installing numerous toolbars from all the various sites listed above, as well as an exhaustive search looking through individual reports to see what users had installed, with no luck reproducing. I was not successful in getting that particular dll version to install.

I wasn't successful installing Bandoo V7 or V8, which may or may not be the delivery mechanism for the crashing dll. Juan notes that it gets "removed" in addons and there seems to be no way to get it back.

Comment 17

[Marcia Knous [:marcia]]

Archiving this email comment for future reference:

Note to QA:

Regarding the versions - Best product to check is www.ilivid.com

Comment 18

[Scoobidiver (away)]

A Bandoo's fix was pushed on November 22 at 15:30 GMT. Here are crashes per day in 17.0:
23 Nov 15:30 GMT - 24 Nov 15:30 GMT: 167
22 Nov 15:30 GMT - 23 Nov 15:30 GMT: 218
21 Nov 15:30 GMT - 22 Nov 15:30 GMT: 263
20 Nov 15:30 GMT - 21 Nov 15:30 GMT: 112
They are starting to decrease.

Comment 19

[Scoobidiver (away)]

Follow-up of comment 18:
25 Nov 15:30 GMT - 26 Nov 15:30 GMT: 134
24 Nov 15:30 GMT - 25 Nov 15:30 GMT: 157

The spike was sharp, the decrease is slow. I am wondering if it could be caused by users who left.

Comment 20

[Robert Kaiser]

Just as a note, the crash numbers are definitely dropping. Even though the absolute number of reports we recorded in the last two days is pretty much the same (slightly over 1400 processed crashes for both, and we process only 10% of all release reports), we increased active installations by ~30% between those days. I hope the absolute numbers start decreasing as well, but we're on the right track. The trend is also in the right direction longer-term: From 17th to 22nd, we had over 1000 crashes per million active daily installations (high point was 1400 on the 20th), we decreased in that measure every day since then, with yesterday (26th) being down to 400.
This is still the #7 top crash across Firefox 17, but as presented above, it's heavily trending in the right direction. Thanks for the fix!

Comment 21

[Robert Kaiser]

Oh, and in the last comment I forgot that in release it's a different signature than on beta, as comment #13 points out, but even that different one, which is at pretty low volume to begin with, is also in decline.

Comment 22

[Mihaela Velimiroviciu (:mihaelav)]

Could not reproduce the crash on the latest Firefox 18.0 beta 2, with some of the suggested add-ons and forced compatibility for DataMngr 1.0 (but I didn't find anything that contained DataMngrHlpFF17.dll). I tried playing flash videos and different scenarios involving files download, but no success.

Comment 23

[Scoobidiver (away)]

(In reply to Mihaela Velimiroviciu [QA] (:mihaela) from comment #22)
> (but I didn't find anything that contained DataMngrHlpFF17.dll).
Did you try with iLivid Free Download Manager (see comment 17)?

Comment 24

[Mihaela Velimiroviciu (:mihaelav)]

(In reply to Scoobidiver from comment #23)
> (In reply to Mihaela Velimiroviciu [QA] (:mihaela) from comment #22)
> > (but I didn't find anything that contained DataMngrHlpFF17.dll).
> Did you try with iLivid Free Download Manager (see comment 17)?

I installed iLivid Download Manager from www.ilivid.com and downloaded files while they were playing in Firefox, but Firefox didn't crash. Also, I opened in Firefox media files which were downloaded with iLivid, but still no crash.

Comment 25

[Marcia Knous [:marcia]]

One thing I noticed when looking at the manual correlations today for FF 17 is that 98% of users had Price Gong Version 1.0 installed for one of the signatures:

arena_dalloc_small | je_free | nsVoidArray::~nsVoidArray()|EXCEPTION_BREAKPOINT (296 crashes)
     98% (290/296) vs.   4% (1081/26794) {1FD91A9C-410C-4090-BBCC-55D3450EF433}
     67% (198/296) vs.   2% (403/26794) {f34c9277-6577-4dff-b2d7-7d58092f272f}
     89% (262/296) vs.  47% (12685/26794) [e-mail] (Mozilla Labs - Test Pilot, https://addons.mozilla.org/addon/13661)
     32% (94/296) vs.   3% (925/26794) [e-mail]
     22% (64/296) vs.   1% (225/26794) {336D0C35-8A85-403a-B9D2-65C292C39087}
     25% (73/296) vs.   4% (1071/26794) [e-mail]
     14% (40/296) vs.   2% (429/26794) [e-mail]
     13% (37/296) vs.   3% (676/26794) [e-mail] (IDM CC, https://addons.mozilla.org/addon/6973)
     14% (41/296) vs.   5% (1459/26794) {635abd67-4fe9-1b23-4f01-e679fa7484c1} (Yahoo! Toolbar, https://addons.mozilla.org/addon/2032)
     10% (29/296) vs.   2% (530/26794) [e-mail]
     99% (293/296) vs.  91% (24516/26794) {972ce4c6-7e08-4474-a285-3208198ce6fd} (Default, https://addons.mozilla.org/addon/8150)
     10% (30/296) vs.   4% (1069/26794) [e-mail]
      6% (19/296) vs.   1% (226/26794) [e-mail] (Flash Video Downloader, https://addons.mozilla.org/addon/6584)

Comment 26

[Scoobidiver (away)]

(In reply to Marcia Knous [:marcia] (Away 12/5-12/7) from comment #25)
> One thing I noticed when looking at the manual correlations today for FF 17
> is that 98% of users had Price Gong Version 1.0 installed for one of the
> signatures:
{1FD91A9C-410C-4090-BBCC-55D3450EF433} is the extension ID for DataMgr and has nothing to do with the Price Gong extension that has {8A9386B4-E958-4c4c-ADF4-8F26DB3E4829} as ID.

Comment 27

[Marcia Knous [:marcia]]

Sorry, I misread the report - Price Gong was listed in another crash.

(In reply to Scoobidiver from comment #26)
> (In reply to Marcia Knous [:marcia] (Away 12/5-12/7) from comment #25)
> > One thing I noticed when looking at the manual correlations today for FF 17
> > is that 98% of users had Price Gong Version 1.0 installed for one of the
> > signatures:
> {1FD91A9C-410C-4090-BBCC-55D3450EF433} is the extension ID for DataMgr and
> has nothing to do with the Price Gong extension that has
> {8A9386B4-E958-4c4c-ADF4-8F26DB3E4829} as ID.

Comment 28

[Robert Kaiser]

None of the two signatures are in the top 300 crashes of 18.0b2, and way out of topcrash range on 17.0.1 - I think we should untrack for 18, and we might even for 17, possibly actually marking this FIXED as Bandoo apparently deployed a fix successfully.

Comment 29

[Alex Keybl [:akeybl]]

Thanks KaiRo - agreed.

Comment 30

[Robert Kaiser]

Unfortunately, over the last two days, another signature with this rose sharply:

  mozalloc_abort(char const* const) | mozalloc_handle_oom(unsigned int) | moz_xmalloc | nsTArray_base<nsTArrayDefaultAllocator>::EnsureCapacity(unsigned int, unsigned int) | nsObserverList::FillObserverArray(nsCOMArray<nsIObserver>&)|EXCEPTION_BREAKPOINT (468 crashes)
    100% (468/468) vs.   4% (1624/45641) DataMngrHlpFF17.dll
        100% (468/468) vs.   4% (1614/45641) 1.0.0.1
          0% (0/468) vs.   0% (10/45641) 3.0.0.3703
     78% (363/468) vs.   5% (2392/45641) datamngr.dll
         78% (363/468) vs.   5% (2290/45641) 1.0.0.1
          0% (0/468) vs.   0% (5/45641) 3.0.0.2699
          0% (0/468) vs.   0% (1/45641) 3.0.0.2736
          0% (0/468) vs.   0% (3/45641) 3.0.0.2790
          0% (0/468) vs.   0% (8/45641) 3.0.0.2859
          0% (0/468) vs.   0% (1/45641) 3.0.0.2950
          0% (0/468) vs.   0% (4/45641) 3.0.0.3037
          0% (0/468) vs.   0% (10/45641) 3.0.0.3206
          0% (0/468) vs.   0% (2/45641) 3.0.0.3289
          0% (0/468) vs.   0% (14/45641) 3.0.0.3369
          0% (0/468) vs.   0% (14/45641) 3.0.0.3703
          0% (0/468) vs.   0% (5/45641) 3.0.0.42011
          0% (0/468) vs.   0% (5/45641) 3.0.0.46577
          0% (0/468) vs.   0% (6/45641) 3.0.0.46593
          0% (0/468) vs.   0% (1/45641) 3.0.0.49236
          0% (0/468) vs.   0% (1/45641) 3.0.0.50653
          0% (0/468) vs.   0% (3/45641) 3.0.0.51994
          0% (0/468) vs.   0% (6/45641) 3.0.0.53061
          0% (0/468) vs.   0% (4/45641) 3.0.0.55914
          0% (0/468) vs.   0% (2/45641) 3.0.0.56274
          0% (0/468) vs.   0% (7/45641) 3.0.0.56544


It's topcrash #9 in yesterday's data for 17.0.* release.

Comment 31

[u279076]

I don't know if this is helpful or not. I installed BandooV8 from http://download.bandooo.com/o/2/r/0/BandooV8.exe which installs...

Applictions:
 * Bandoo 8.0.0.130452
 * Searchqu 1.0.0.12

Firefox Add-ons:
 * Bandoo for Firefox 5.1
 * Searchqu 1.0.0.12
 * Datamngr 1.0 (hardblock disabled due to incompatibility)

DLLs:
 * C:\Program Files\SRToolbar\Datamngr\datamngr.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF3.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF4.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF5.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF6.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF7.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF8.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF9.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF10.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF11.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF12.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF13.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF14.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF15.dll
 * C:\Program Files\SRToolbar\Datamngr\FirefoxExtension\components\DataMngrHlpFF16.dll

All of the above DLLs have the same file description...
 * File Description: Data Manager
 * Type: Application extension
 * File Version: 1.0.0.1
 * Original Filename: DatamngrHelper.dll

Finally, after watching Firefox.exe in Process Monitor for 15 minutes I see the following DLLs appear:
 * C:\Program Files\Mozilla Firefox\xul.dll
 * C:\Program Files\Mozilla Firefox\mozjs.dll
 * C:\Program Files\Mozilla Firefox\gkmedias.dll
 * C:\Windows\System32\usp10.dll
 * C:\Users\mozilla\AppData\Roaming\Mozilla\Firefox\Profiles\qbpce7bz.default\extensions\[e-mail]\components\FFPluginV17.dll
 * C:\Program Files\Bandoo\BandooRes.dll
 * C:\Program Files\Mozilla Firefox\msvcr100.dll
 * C:\Program Files\Mozilla Firefox\ui\SwDRM.dll
 * C:\Program Files\Mozilla Firefox\breakpadinjector.dll
 * C:\Windows\System32\icm32.dllC:\Windows\System32\ole32.dll
 * C:\Program Files\Mozilla Firefox\mozsqlite3.dll

The Search Results toolbar (provided by Bandoo) just appears as an empty toolbar in Firefox 17.0.1 and there is no DataMngrHlpFF17.dll on my system. I'm at a loss as to where people are getting DataMngrHlpFF17.dll and how they are getting datamngr.dll to be enabled in Firefox. The only way I understand this crash could happen is if these DLLs are enabled and loaded, but I'm failing to find a way to do that.

Comment 32

[u279076]

I tried disabling add-on compatibility checking and installed http://download.ilivid.com/iLividSetup.exe which has given me a DataMngrHlpFF17.dll on my system. Now to see if I can trigger a crash somehow.

Comment 33

[u279076]

I installed all of the top-50 liked Conduit apps in the toolbar and tried a few of the things (some mentioned in recent comments):
 * playing games
 * using webmail
 * facebook integration
 * downloading flash videos from Youtube
 * restarting Firefox
 * using iLivid's video conversion tools

I've yet to experience a crash. Sorry but I don't have anymore time to spend on this today. Please advise if there's other testing I can try and I'll try to find time to do it early next week. At least I have to reported DLL now.

Comment 34

[Marcia Knous [:marcia]]

https://crash-stats.mozilla.com/report/list?signature=InitLog shows up in Firefox 17.0.1 correlated to the same, dll, so I am adding this signature so it shows up associated with a bug.

Comment 35

[Robert Kaiser]

Looks like this dropped out of topcrash range now, it's #38 over the last week of 17.0.1 data.

Comment 36

[u279076]

Also dropping qawanted, we've been unsuccessful in finding any concrete leads.

Comment 37

[Andy McKay]

With WebExtensions being the only valid way of doing extensions in Firefox 57, I don't think this bug is still relevant.