Closed
Bug 489970
Opened 16 years ago
Closed 16 years ago
crash upon gmail logout [@ user32.dll@0x1898d ] because googletalk unloaded without unregistering
Categories
(Firefox :: General, defect)
Tracking
()
VERIFIED
INVALID
People
(Reporter: dc, Unassigned)
References
()
Details
(Keywords: common-issue+, crash)
Crash Data
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.1b5pre) Gecko/20090424 Shiretoko/3.5b5pre
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.1b5pre) Gecko/20090424 Shiretoko/3.5b5pre
Upon logging out from gmail, there is a short popup mentioned about '... even thought the page is encrypted, some contents are ...) -> I hit ok to close the dialog box.
Right after the dialog box closed, firefox crashed.
Reproducible: Always
Steps to Reproduce:
1. Open http://www.gmail.com
2. Login as usual
3.a. Wait until all elements loaded (i.e.: the status on the left is fully loaded, gmail theme loaded, etc)
or
3.b Don't wait and go to the next step
4. Sign out from gmail
5. A dialog box mentioning that although the website is encrypted, some components are sent without encryption (or something similar with that)
6. Click on [OK] to close the dialog box
Actual Results:
Browser crash
Expected Results:
No crash, supposed to be brought back to gmail login page
Note: I remember that in the previous days, logging out from gmail did not provide such dialog box about the unencrypted contents sent over encrypted page (the warning). But I could have been using a fresh profile and it showed up.
I have tried the following combinations with the steps above and resulted in crash:
- firefox 3.0.9
- firefox 3.0.9 safe-mode (all add-ons disabled)
- firefox 3.0.9 blank profile
- firefox 3.0.9 safe-mode (all add-ons disabled) blank profile
- shiretoko 3.5b4
- shiretoko 3.5b4 safe-mode (all add-ons disabled)
- shiretoko 3.5b4 blank profile
- shiretoko 3.5b4 safe-mode (all add-ons disabled) blank profile
- shiretoko 3.5b5 (1240558102)
- shiretoko 3.5b5 (1240558102) blank profile
- minefield 3.6a1
- minefield 3.6a1 blank profile
Below are the list of crash ids (mixed up of above releases)
(date and time shown in US PDT)
eaadacc5-e1b1-4027-a768-047792090424 4/24/2009 2:44 AM
f2017cce-ebc1-471d-9fae-0cd8b2090424 4/24/2009 2:37 AM
fd5c9dc6-980c-4c42-83aa-7d4fa2090424 4/24/2009 2:25 AM
b1e109f3-4d89-483d-9daa-3f1722090424 4/24/2009 2:23 AM
a68a9100-4b64-42fa-aa0d-53fa42090424 4/24/2009 2:21 AM
f0f78e9c-2e05-4c5b-87f8-e8a1f2090424 4/24/2009 2:21 AM
41819c62-581e-45b1-9d76-6b8232090424 4/24/2009 2:20 AM
b982ddb4-b858-4b9f-8cbe-d438a2090424 4/24/2009 2:14 AM
Updated•16 years ago
|
Severity: normal → critical
Keywords: crash
Summary: crash upon gmail logout → crash upon gmail logout [@ user32.dll@0x1898d ]
Version: unspecified → Trunk
![]() |
||
Comment 1•16 years ago
|
||
Linkified crash stacks:
bp-eaadacc5-e1b1-4027-a768-047792090424
bp-f2017cce-ebc1-471d-9fae-0cd8b2090424
bp-fd5c9dc6-980c-4c42-83aa-7d4fa2090424
bp-b1e109f3-4d89-483d-9daa-3f1722090424
bp-a68a9100-4b64-42fa-aa0d-53fa42090424
bp-f0f78e9c-2e05-4c5b-87f8-e8a1f2090424
bp-41819c62-581e-45b1-9d76-6b8232090424
bp-b982ddb4-b858-4b9f-8cbe-d438a2090424
Comment 2•16 years ago
|
||
Which plugins are listed in about:plugins ?
Are there any google plugins found and which flash plugin is used ?
so, you have a flash 10, acrobat 9.1, cooliris (try disabling it -- or better, try safe mode), googledesktop.
please try disabling at least those 4, and preferably whichever other extensions you have installed.
Reporter | ||
Comment 4•16 years ago
|
||
Hi all,
I got it nailed down! The problem happened with googletalk plugin enabled!
Below are the crash ids.
The latest one is re-assurance that googletalk plugin is really causing it.
f160e1ed-e17c-4609-af34-565502090426 4/26/2009 1:43 AM
e663a6fa-9118-42ae-966e-3e1fb2090426 4/26/2009 1:38 AM
c5c779ec-e826-456e-9db0-fc0922090426 4/26/2009 1:35 AM
On the side note: i only experience this with my notebook. I have similar OS running on my desktop computer (srv 2008 x64 standard -> aka vista x64-ish)
Should this now be changed from 'critical' to 'major' ?
no, but please file a bug against google and provide a url to your complaint in this bug. it's highly unlikely we can fix it.
it /looks/ like they unloaded their library w/o first removing a window hook, which is an api violation that will crash. we can't do anything about it.
Status: UNCONFIRMED → RESOLVED
Closed: 16 years ago
Resolution: --- → INVALID
Summary: crash upon gmail logout [@ user32.dll@0x1898d ] → crash upon gmail logout [@ user32.dll@0x1898d ] because googletalk unloaded without unregistering
Reporter | ||
Comment 6•16 years ago
|
||
done!
well, technically, I didn't file the bug because there is already one filed to google here http://www.google.com/support/forum/p/Talk/thread?tid=187a50cb38266f2b&hl=en&fid=187a50cb38266f2b00046871fa8f8970
that'll work too. good luck. definitely try installing the latest version of their plugin in case it fixes the problem.
Status: RESOLVED → VERIFIED
Reporter | ||
Comment 8•16 years ago
|
||
no worries mate, even though I have no clue whether someone at Google would have a look at it (or how soon), as this could deter firefox users away due to their inability to access gmail with googletalk plugin installed (and enabled).
Comment 9•16 years ago
|
||
kev, any contacts at google that would help move this along?
Comment 13•16 years ago
|
||
I think resolved bugs are less likely to get any attention, perhaps this should be reopened?
Keywords: common-issue?
Comment 14•16 years ago
|
||
I agree, I have recently encountered this (https://bugzilla.mozilla.org/show_bug.cgi?id=509549) and it is significantly hampering my browsing experience. I am having to live with the current workaround of having to disable the plugin.
(In reply to comment #13)
> I think resolved bugs are less likely to get any attention, perhaps this should
> be reopened?
Comment 15•16 years ago
|
||
what sort of attention would you like?
put differently. imagine i sell you a sandbox made of wood,
http://www.backyardcity.com/images/OPC/SandboxWithRoof.jpg
the next guy over sells you a "ground spinner",
http://content.artofmanliness.com/uploads/2008/07/groundbloom.jpg
you decide to stick the ground spinner into the wooden sandbox i sold you (it's not a big sandbox, and yes, it's made of wood) ... without any sand, and then you light it.
the spinner spins and ignites the sandbox, and the sandbox burns.
when you come to me complaining that the wooden sandbox i sold you burned, what exactly do you expect me to say?
I'm willing to explain the physics and chemistry behind why using fireworks in a sandless sandbox is a bad idea. but it's not the fault of the sandbox that something like fireworks were used inside it.
to make the example better, imagine that the guy selling the ground spinner says "works well in sandboxes".
Who should you blame when the sandbox burns?
Comment 16•16 years ago
|
||
The kind of attention I am looking for is to investigate why this crash starts happening AFTER upgrading to v3.5.2. The least I expect is to have this bug in a open (or similar status) and not being dispensed off as being invalid.
I am already pursuing a bug report with Gtalk plugin team.
Comment 17•16 years ago
|
||
So what is this Google Talk plugin? Are we talking about this extension? https://addons.mozilla.org/en-US/firefox/addon/8241
Comment 18•16 years ago
|
||
It is the "Google Talk NPAPI Plugin" that exists as the file "npgoogletalk.dll" within the FF plugin folder. It has also been discussed here:
http://support.mozilla.com/tiki-view_forum_thread.php?locale=da&comments_parentId=265176&forumId=1
Comment 19•16 years ago
|
||
>The kind of attention I am looking for is to investigate why this crash starts
>happening AFTER upgrading to v3.5.2.
Ask the Google people why the plugin unloads without unregistering ?
How should the Mozilla.org developers know why it's doing that ?
>The least I expect is to have this bug in
>a open (or similar status) and not being dispensed off as being invalid.
There is no further discussion needed because it's clear that they violate the API and this can not be fixed on our side. invalid= not a bug on our side
and don't forget that Bugzilla is not there to solve user issues
Comment 20•16 years ago
|
||
Sighh, this is going around in circles. Pl. take a moment to read my previous comment about my previous bug report (https://bugzilla.mozilla.org/show_bug.cgi?id=509549).
My addition to this bug report is that the plugin continued to perform normally in v3.5.1. But it seems to crash FF v3.5.2.
As said before, I am already chasing a similar conversation with the plugin folks, my request here is to identify what in the FF upgrade changed it's behavior causing it to crash.
Comment 21•16 years ago
|
||
mskadu, as matti said, it is not our fault that bad code in the plugin broke during a Firefox upgrade. If we tried to not break every plugin and addon, then you would never get any work done.
Comment 22•16 years ago
|
||
we do have partner outreach to work with plugin providers to make sure they stay compatible. kev, do you have contacts with the the gtalk plugin group?
Comment 23•16 years ago
|
||
Seeing many instances of this stack on support.
Keywords: common-issue? → common-issue+
Comment 24•16 years ago
|
||
that's odd because I see very few incoming reports on this exact signature over the last few weeks
report for the stack signature user32.dll@0x1898d
7 total crashes for user32.dll@0x1898d on 20090805-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090806-crashdata.csv
3 total crashes for user32.dll@0x1898d on 20090807-crashdata.csv
3 total crashes for user32.dll@0x1898d on 20090808-crashdata.csv
4 total crashes for user32.dll@0x1898d on 20090809-crashdata.csv
9 total crashes for user32.dll@0x1898d on 20090810-crashdata.csv
17 total crashes for user32.dll@0x1898d on 20090811-crashdata.csv
9 total crashes for user32.dll@0x1898d on 20090812-crashdata.csv
12 total crashes for user32.dll@0x1898d on 20090813-crashdata.csv
3 total crashes for user32.dll@0x1898d on 20090814-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090815-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090816-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090817-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090818-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090819-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090820-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090821-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090822-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090823-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090824-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090825-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090826-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090827-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090828-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090829-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090830-crashdata.csv
0 total crashes for user32.dll@0x1898d on 20090831-crashdata.csv
If I back off to look at a more generalized search for crashes at other addresses in this system .dll I get more crashes, but also a trend downward.
These other crashes at other address may, or may not be related to the bad google talk plugin code.
report for the stack signature user32.dll@0x
692 total crashes for user32.dll@0x on 20090805-crashdata.csv
357 total crashes for user32.dll@0x on 20090806-crashdata.csv
358 total crashes for user32.dll@0x on 20090807-crashdata.csv
397 total crashes for user32.dll@0x on 20090808-crashdata.csv
349 total crashes for user32.dll@0x on 20090809-crashdata.csv
1217 total crashes for user32.dll@0x on 20090810-crashdata.csv
1615 total crashes for user32.dll@0x on 20090811-crashdata.csv
1549 total crashes for user32.dll@0x on 20090812-crashdata.csv
2461 total crashes for user32.dll@0x on 20090813-crashdata.csv
940 total crashes for user32.dll@0x on 20090814-crashdata.csv
41 total crashes for user32.dll@0x on 20090815-crashdata.csv
49 total crashes for user32.dll@0x on 20090816-crashdata.csv
47 total crashes for user32.dll@0x on 20090817-crashdata.csv
46 total crashes for user32.dll@0x on 20090818-crashdata.csv
58 total crashes for user32.dll@0x on 20090819-crashdata.csv
72 total crashes for user32.dll@0x on 20090820-crashdata.csv
65 total crashes for user32.dll@0x on 20090821-crashdata.csv
60 total crashes for user32.dll@0x on 20090822-crashdata.csv
52 total crashes for user32.dll@0x on 20090823-crashdata.csv
65 total crashes for user32.dll@0x on 20090824-crashdata.csv
71 total crashes for user32.dll@0x on 20090825-crashdata.csv
73 total crashes for user32.dll@0x on 20090826-crashdata.csv
74 total crashes for user32.dll@0x on 20090827-crashdata.csv
60 total crashes for user32.dll@0x on 20090828-crashdata.csv
60 total crashes for user32.dll@0x on 20090829-crashdata.csv
54 total crashes for user32.dll@0x on 20090830-crashdata.csv
50 total crashes for user32.dll@0x on 20090831-crashdata.csv
Assignee | ||
Updated•14 years ago
|
Crash Signature: [@ user32.dll@0x1898d ]
You need to log in
before you can comment on or make changes to this bug.
Description
•