1109257 - crash in _PR_MD_SEND | SocketSend | SocketWrite | PR_SetPollableEvent | nsSocketTransportService::OnDispatchedEvent(nsIThreadInternal*)

Status: RESOLVED DUPLICATE of bug 662330

(Core :: Networking: HTTP, defect)

Description

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

[Tracking Requested - why for this release]:

[Tracking Requested - why for this release]:

This bug was filed from the Socorro interface and is 
report bp-06101d1f-4c53-4841-b67d-43e922141208.
=============================================================

This is the #9 topcrash for 34.0.5 with 233/11721 crashes in the last 7 days.
It is often a startup crash. It's showing up in significant volume on 34.0b1 as well.  Comments express frustration with repeated startup crashes. Top urls are about:home, about:blank, about:sessionrestore. 

More reports: https://crash-stats.mozilla.com/report/list?product=Firefox&signature=_PR_MD_SEND+|+SocketSend+|+SocketWrite+|+PR_SetPollableEvent+|+nsSocketTransportService%3A%3AOnDispatchedEvent%28nsIThreadInternal*%29 

Crashing thread:

0 		@0xf4e84320 	
1 	nss3.dll 	_PR_MD_SEND 	nsprpub/pr/src/md/windows/w95sock.c
2 	nss3.dll 	SocketSend 	nsprpub/pr/src/io/prsocket.c
3 	nss3.dll 	SocketWrite 	nsprpub/pr/src/io/prsocket.c
4 	nss3.dll 	PR_SetPollableEvent 	nsprpub/pr/src/io/prpolevt.c
5 	xul.dll 	nsSocketTransportService::OnDispatchedEvent(nsIThreadInternal*) 	netwerk/base/src/nsSocketTransportService2.cpp
6 	xul.dll 	nsThread::Dispatch(nsIRunnable*, unsigned int) 	xpcom/threads/nsThread.cpp
7 	xul.dll 	nsSocketTransportService::Dispatch(nsIRunnable*, unsigned int) 	netwerk/base/src/nsSocketTransportService2.cpp
8 	xul.dll 	mozilla::net::nsHttpConnectionMgr::PostEvent(void ( mozilla::net::nsHttpConnectionMgr::*)(int, void*), int, void*) 	netwerk/protocol/http/nsHttpConnectionMgr.cpp
9 	xul.dll 	mozilla::net::nsHttpConnectionMgr::UpdateRequestTokenBucket(mozilla::net::EventTokenBucket*) 	netwerk/protocol/http/nsHttpConnectionMgr.cpp
10 	xul.dll 	mozilla::net::nsHttpHandler::MakeNewRequestTokenBucket() 	netwerk/protocol/http/nsHttpHandler.cpp
11 	xul.dll 	mozilla::net::nsHttpHandler::Init() 	netwerk/protocol/http/nsHttpHandler.cpp
12 	xul.dll 	mozilla::net::nsHttpHandlerConstructor 	netwerk/build/nsNetModule.cpp
13 	xul.dll 	mozilla::GenericFactory::CreateInstance(nsISupports*, nsID const&, void**) 	xpcom/glue/GenericFactory.cpp
14 	xul.dll 	nsComponentManagerImpl::CreateInstance(nsID const&, nsISupports*, nsID const&, void**) 	xpcom/components/nsComponentManager.cpp
15 	xul.dll 	nsComponentManagerImpl::GetService(nsID const&, nsID const&, void**) 	xpcom/components/nsComponentManager.cpp

Comment 1

[Tracy Walker [:tracy]]

pretty sure lizzard meant 35.0beta1 where it's currently 31st on the volume list.  but being a startup crasher, it should be weighted more heavily. Particularly since we see it on 33.* and 34.* in our release population.

Comment 2

[(Away)]

_PR_MD_SEND usually indicates external interference since bad guys like to hook the system network APIs.

In this case, the reports all have an adware "Wajam Web Enhancer" with a random binary name:
c:\program files\wajam web enhancer\dlls\xqptcfhihqgf.dll
c:\program files\WWE\dlls\ecfgtyorobwr.dll
C:\Program Files\WWE\dlls\repsivndckdw.dll

Comment 3

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

Tracking for all releases. There is no current plan to ship a point release. We can discuss options once we better understand this crash and whether we can fix it on our end.

Comment 4

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

I have also e-mailed Wajam and directed them to this bug report. It is in both of our best interests that this issue be resolved.

Comment 5

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

Heard back from Wajam. Their engineering team is looking into this issue.

Comment 6

[(Away)]

I would be curious to hear Wajam's explanation for why they disguise their DLL name (though I think I know the real reason).

Comment 7

[Lukas Blakk [:lsblakk] use ?needinfo]

Any updates from Wajam?

Comment 8

[Lukas Blakk [:lsblakk] use ?needinfo]

Fwiw, this has fallen out of the top 20 and is at #23 topcrasher in 35.0b

Comment 9

[Lukas Blakk [:lsblakk] use ?needinfo]

(In reply to Lukas Blakk [:lsblakk] use ?needinfo from comment #8)
> Fwiw, this has fallen out of the top 20 and is at #23 topcrasher in 35.0b

However it is showing as being on the rise...so that's concerning.

Comment 10

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

I didn't hear back from Wajam again. I sent a follow-up to Wajam about this bug today.

Comment 11

[Lukas Blakk [:lsblakk] use ?needinfo]

Nothing actionable for 35 at this point, wontfixing.

Comment 12

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

Follow-up from Wajam, "The build that fixes this is in QA as of today. It should be live very soon."

Comment 13

[(Away)]

Is that only for new Wajam downloads or will existing installations get updated?

> Fwiw, this has fallen out of the top 20 and is at #23 topcrasher in 35.0b
It's possible that we lost the affected users, this being a persistent startup crash.

Comment 14

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

(In reply to David Major [:dmajor] (UTC+13) from comment #13)
> Is that only for new Wajam downloads or will existing installations get
> updated?

Word is that a fix will be pushed for new downloads early next week and existing installations will be updated in 2-3 weeks.

Comment 15

[Sylvestre Ledru [:Sylvestre]]

36 & 37 seems to be fixed now.
I guess this bug can be fixed too.