692685 - Constant crashes with npmozax.dll, blocklist it

Status: RESOLVED DUPLICATE of bug 692149

(Toolkit :: Blocklist Policy Requests, defect)

Description

[Rob]

User Agent: Mozilla/5.0 (Windows NT 5.1; rv:9.0a2) Gecko/20111006 Firefox/9.0a2
Build ID: 20111006042015

Steps to reproduce:

I had a crash MANY hours ago and wanted to see if a COUPLE (few?) additional Reports had appeared (after having percolated through the System).

I have not had a crash since approximately 10/6/2011 10:45 .

I typed about:crashes and TOO MANY Reports showed up, most of them not even bp-'ed (did not have a "bp-" prefix).

I clicked on the top of the List (was not bp-ed) and a "strober" appeared, then the Report was bp-ed. 

I clicked on the top of the List again and got this message:

"
Crash Not Found

We couldn't find the OOID you're after. If you recently submitted this crash, it may still be in the queue.

If you belive this message is an error, please submit a _Bugzilla_ _ticket_ describing what happened, and please include the URL for this page.
"


There is an URL to click in that message if there is an error so I clicked it and here I am. 

The URL (presumably) is for reporting that you think that the Crash Info ought to be available. I clicked it because I think the Entry ought NOT to have appeared in the List. I have not had a crash for hours. 


All these extra false Reports are not helpful if we are to include our "about:crashes" (when requested) in a Report. 

I do understand that the way it works is that WHEN a crash occurs that it MIGHT get split into a few Reports (that is not helpful (to me) either) BUT my complaint is about the TONS of (presumably) false Crash Report URLs.



Here is my latest about:crashes - REMEMBER that I have not had a crash since approximately 10/6/2011 10:45 .


Report ID 						Date Submitted
bp-2589dd98-4d90-4dec-a59f-9600f2111006 	10/6/2011 17:57
bp-935ef166-6364-4b8d-b2f5-41e742111006 	10/6/2011 17:51
938ecd42-f7cb-48f0-a894-0e4af78d20b3 		10/6/2011 17:51
60b86cd0-a0c3-4d11-8a6e-79a32f31407a 		10/6/2011 17:51
e412e6d7-66a6-4c32-a517-3a640bb91e15 		10/6/2011 17:31
7159c7ee-7ec9-44ff-92a9-b4acdf6d1a86 		10/6/2011 17:31
34d0fe3e-bbb0-450d-9b3a-60870fb11272 		10/6/2011 17:31
71af6235-bf91-445f-9985-66522611592b 		10/6/2011 17:31
08dfff05-376e-4daf-a97c-03a6c0885420 		10/6/2011 17:11
77f5cc19-09fb-421c-ac1f-3d0e9e6e029b 		10/6/2011 17:10
95776f8f-c824-4490-a7d6-df01d5dc06c2 		10/6/2011 17:10
fd5d2731-040f-4f6a-8fde-7bf89b4fdfe9 		10/6/2011 17:10
dfecd32c-56c4-4873-a831-6b74c96437fd 		10/6/2011 16:50
7e882692-9f6c-4fa7-b314-87ab9b147dfe 		10/6/2011 16:50
0d3d5fd5-6efd-4ce1-9f5b-bccd31a82ebf 		10/6/2011 16:50
380b743f-9b55-450d-b053-5c69eb20f86a 		10/6/2011 16:50
9617045e-92d7-488d-ae52-e8d80d841b2d 		10/6/2011 16:30
928f632e-95f3-46e4-95b5-fdcf95b28ced 		10/6/2011 16:30
d0b583ce-757c-4453-962d-c3c3c54d26c4 		10/6/2011 16:30
ba8ca584-6074-48cd-80f2-58ce0fdcd59e 		10/6/2011 16:30
24f0e1dc-91a3-42b4-a468-76229ad1ded4 		10/6/2011 16:10
6755aaaf-907e-41b2-b10b-8ccf9976812a 		10/6/2011 16:09
d284a783-6365-467b-b92d-4504c7587b67 		10/6/2011 16:09
7fe1e13f-67d9-4a94-8e71-d62b5ca056e4 		10/6/2011 16:09
85ee10bf-9b8e-4f1f-9ddb-ff6599ded918 		10/6/2011 15:49
b62445c9-0dd7-4bbe-8988-fdc807b31257 		10/6/2011 15:49
551bfd02-f06c-4413-9fc4-763404a2aede 		10/6/2011 15:49
7af92aa8-0559-480d-83f5-79a9e72e90c9 		10/6/2011 15:49
92214c2d-1be6-4de3-9311-3b8843e64a4f 		10/6/2011 15:29
e1fe14ce-be41-4dd2-a482-4b9e8f1e6e5f 		10/6/2011 15:29
f15bc6d0-bd26-497c-8da9-13fcc8fb3db5 		10/6/2011 15:29
f2b8ff3a-591e-4b01-beec-48c32bc216a2 		10/6/2011 15:29
d7f98344-6272-4a83-bf2d-62027cb0a44e 		10/6/2011 15:09
ae8bfb98-9025-4833-b23e-3231ee815307 		10/6/2011 15:09
924b76ed-4610-4786-8dba-898d7fa3211c 		10/6/2011 15:09
34e74b53-9a17-4e8e-bb27-bc2fd671f19f 		10/6/2011 15:09
af4382e5-15bb-4b91-b7d4-c4fc67300b95 		10/6/2011 14:48
836fa4ad-f89e-44d2-8266-415172bfdec2 		10/6/2011 14:48
726ad1d9-89ed-47e5-bc40-27df06b9caae 		10/6/2011 14:48
b131797f-a11c-4d16-b077-0a10463200e2 		10/6/2011 14:48
e375667f-f8c9-4292-8b26-943bda01b1f7 		10/6/2011 14:28
c8d74ad7-b7a9-46dc-9a90-20506bdcebfb 		10/6/2011 14:28
28647a4f-4560-42b6-afcf-eabcb537bd2c 		10/6/2011 14:28
bab2c38a-4a04-433d-a0ff-d4a9f56f013f 		10/6/2011 14:28
138140b1-0ff8-406a-9cb0-831c0ac238fc 		10/6/2011 14:08
de3ac634-44cd-4cf1-ab4d-9c3b30c05e9f 		10/6/2011 14:08
feb91dab-d3ab-4856-a4cd-d03bdde797e7 		10/6/2011 14:08
b2079d87-3624-4d62-954c-4f1faa154c3f 		10/6/2011 14:08
28942bd2-62c8-4f36-9400-81a17cdbe25d 		10/6/2011 13:47
71f785e0-8985-4c22-9238-685e00caa68f 		10/6/2011 13:47
4fbccf8d-f56d-4bf1-b738-986ef621e70e 		10/6/2011 13:47
b630dcfb-aeb4-4b34-889f-9e1dcd8bdf13 		10/6/2011 13:47
00a448dc-49f3-4490-b867-07a38571b73c 		10/6/2011 13:27
cfc1c23b-3a9f-4e81-a32e-0cd63838a929 		10/6/2011 13:27
9e9fbeef-5702-4b6d-80c4-1e004a5ac38a 		10/6/2011 13:27
e2060f38-0ea4-4b7f-92ce-cbb5ecb8208d 		10/6/2011 13:27
193d9d24-efc6-4648-835a-79bbb42dc9c8 		10/6/2011 13:07
e5f89fc3-e52f-4bac-b7a1-38b3dffc5b03 		10/6/2011 13:07
65ca2930-b7f0-476a-bbbc-26eb943a4e27 		10/6/2011 13:07
fa2cf598-3bd8-4688-98d1-0933077a26bf 		10/6/2011 13:07
fd1bc7db-01f7-44d7-8c3d-b519b176ff4c 		10/6/2011 12:47
440fd397-3bc5-49e7-9139-8e0d23238687 		10/6/2011 12:47
ba8a237e-afff-406f-8493-6e0e408897d7 		10/6/2011 12:47
d139c48f-9f2b-4f28-a106-a3856cdf33a9 		10/6/2011 12:46
cecf8444-4570-466e-81e3-1be3fe499ac0 		10/6/2011 12:26
99c4d393-cad4-4b82-8c07-e47e63101d24 		10/6/2011 12:26
9d4de534-425c-4288-9879-9ae0637b73fb 		10/6/2011 12:26
864d5a62-5ecd-44a7-9eb9-52ad6a133df1 		10/6/2011 12:26
dc30f991-3ced-46ab-93c3-42b2e399c765 		10/6/2011 12:06
6005b4b6-a5ea-4fc7-863e-d01677d8a261 		10/6/2011 12:06
d9127c71-9461-45ac-a2a9-cc6cd6175e17 		10/6/2011 12:06
095940db-8da6-4e98-b649-f9e3dda60458 		10/6/2011 12:06
60a48abd-b9e3-42e3-9727-a9fe3daa1356 		10/6/2011 11:46
8faee278-ebcd-452e-977b-ec5596c9f5eb 		10/6/2011 11:46
85216a3b-129a-4e34-8715-5400a3386386 		10/6/2011 11:46
9a5723b0-cbcc-4bd0-a92f-de129a48b4ae 		10/6/2011 11:45
0ac933a3-f294-4626-9120-608bb5ed7ce4 		10/6/2011 11:28
c3ee9b98-51bd-4849-986c-0cdfeeac0b1d 		10/6/2011 11:28
92f82006-10bb-4898-8e34-e89802446ab8 		10/6/2011 11:25
0a443542-36cd-476e-b5ef-25f91a9df6eb 		10/6/2011 11:25
02a6b2c5-d04d-46f8-ae6d-9c330a634bbe 		10/6/2011 11:25
c910376f-4468-47fc-bfb3-7e5a17af8dff 		10/6/2011 11:24
0d75513a-0c30-41e1-9533-48b0bd66283d 		10/6/2011 11:05
da6e9b33-5451-4bdb-953f-91f6dd40e00a 		10/6/2011 11:05
f691ffbc-f8bb-4bfc-bb70-931bc57ae0af 		10/6/2011 11:05
feb51b99-4149-4764-9cbf-980d9380c49f 		10/6/2011 11:05
dead870e-6e79-4c2b-8ba8-0fa1920cf318 		10/6/2011 10:53
f1f4114f-4b52-453e-93ac-620c74d31570 		10/6/2011 10:53
0396ae6a-a529-4a2d-a07b-20b21d40e896 		10/6/2011 10:45
...


Please limit the about:crashes Page to ACTUAL crashes. IF you must then we could have a FEW Reports surrounding the time in question but how useful is it to have TONS of Reports spewed to that Page. 

This MIGHT be the cause of Bug 568849, or Bug  579136 Comment 5. Whether it is a headache for the Developers or not it is not helpful for the end-user who is wanting to help and is asked for a 'bp' from their about:crashes.

Thanks.



PS: When submitting this Report (and "version" is by default blank) you get a message to specify the version or to use "unspecified", but "unspecified" is not a choice. After fixing that and clicking [Submit] you get a redirect that says to click [Back] and specify a "component", but "Bonsai" is already chosen (filed in) for us. Thus I had to choose "other" for "version" and "Socorro" for the "component" - I hope that is correct (please feel free to fix that problem too, and update my Report as needed). 

Now the Browser Window is returned to this Bug Report but there is no [Submit] Button (only a [Submitting Bug...] Button that is greyed out). Re-submitting this Bug Report. It was a problem with the [Back] Button ("<-") not going back in History but if you right-click on the Button and choose from the List if works and you don't have to re-enter your Report.

Comment 1

[(not currently active) Ted Mielczarek]

Hi Rob,

Everything in that list, believe it or not, is an actual crash report. There are a few things to note here:
1) Entries without a "bp-" prefix are simply crash reports that have not been submitted to the server. Clicking on one in about:crashes will attempt to submit it (hence the little spinner) and redirect you to the page.

2) Since we run plugins in a separate process now, if a plugin crashes it will generate a crash report but not automatically submit it. (The user interface for crashed plugins is being improved in bug 665196). I would guess that that long list of crash reports you have is all plugin crashes, although I can't prove that because I haven't been able to load any of them.

3) The fact that you can't load those two reports that you just submitted is an actual bug, and we'll figure out what's going on there.

Comment 2

[Laura Thomson :laura]

Thanks Ted.  Following up on those two reports:

Metrics: are the first two in HBase?  That's
bp-2589dd98-4d90-4dec-a59f-9600f2111006 	10/6/2011 17:57
bp-935ef166-6364-4b8d-b2f5-41e742111006 	10/6/2011 17:51

We experienced an outage around that time due to a load balancer issue and these two crashes may have gotten caught in that.

Comment 3

[K Lars Lohn [:lars] [:klohn]]

2589dd98-4d90-4dec-a59f-9600f2111006 is a crashmover orphan.  We can assume that 935ef166-6364-4b8d-b2f5-41e742111006  shares that fate.  They were submitted during the Zeus outage yesterday and suffered the crashmover orphan generating bug.  That bug has been fixed in the crashmover and awaits deployment.

Neither of these crashes will be present in HBase.  They're in limbo in the collector05 primaryStorage.

If this crash is critical, we can resubmit it by running the orphansubmitter on collector05. However, since the orphan submitter hasn't been used in more than a year, we ought to examine the code for bit rot.  I believe that it can be run on a live system and we wouldn't have to disable collector05 during the run.  However, it would be a good idea to audit the code to make sure.

Comment 4

[(not currently active) Ted Mielczarek]

These aren't critical, no, they'd just be useful to help diagnose what's going on on his system. :)

Comment 5

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #1)
> Hi Rob,
> Everything in that list, believe it or not, is an actual crash report. There 
> are a few things to note here:
> 1) Entries without a "bp-" prefix are simply crash reports that have not been > submitted to the server. Clicking on one in about:crashes will attempt to 
> submit it (hence the little spinner) and redirect you to the page.
> ...
> 3) The fact that you can't load those two reports that you just submitted is 
> an actual bug, and we'll figure out what's going on there.

I figure I can not load 99.7% of them because they are spuriously generated. IE: I am not crashing and don't have Pages containing Flash (or other) Plugin's that would account for them.

BUT, if you think that is not so and in any event that if I can not click them (and wait patiently) at some point in time get a Report, then it is broke; that is OK too.



(In reply to Ted Mielczarek [:ted, :luser] from comment #4)
> These aren't critical, no, they'd just be useful to help diagnose what's
> going on on his system. :)

Actually they would be more helpful if they were ABSENT, with the exception of the FEW that are valid.

Yes some can be plugins. I had (previously) two crashes with perhaps 4 Tabs that had flash running so that accounts for eight, if we got 4 (excess) Reports for each ONE Crash then there would be no more than 32 (worst case and still excessive, but I understand that there is some 'splitting' that occurs).


I have had another TWO Crashes since my last Post (again with approx 4 Tabs holding Flash) and in addition I have had 2 "Flash Crashes" (without the Browser Crashing) - so perhaps the worst case is 32+16 more Reports - YET there are another 210 additional Reports in about:crashes since 10/6/2011 17:57 (shown in Comment 0).


If everyone's Browser (or "Error Reporter") does this to your Server it will place an unnecessary load on the Server and create excess noise for the Developer (and end user attempting to find a valid "bp-" to give you, in response to your inquiry).


PS: I do not need to access any crashes from the last few days so don't worry about trying to get them for me. Worry about making my (and the other many millions of people's) about:crashes Page show: "The Crash", "Plugin Crashes", and 'a couple of extras, if you like', but not over 80-100 Reports per crash, please.

Comment 6

[(not currently active) Ted Mielczarek]

I'm telling you that I believe these are all legitimate crash reports, but you don't have to believe me if you don't want. :)

We saw something similar to this in bug 690841. I filed bug 691432 for a more reasonable way to handle this. We certainly shouldn't be generating this volume of crashes just because you have a faulty plugin!

Comment 7

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #6)
> I'm telling you that I believe these are all legitimate crash reports, but
> you don't have to believe me if you don't want. :)

I only see the "Error Reporter" pop up a couple of times so the Crash Reports
are being sent without me choosing to send them. 

They can be called "legitimate crash reports" but they were not "permitted to be sent by me" allowing the "Error Reporter" to send them (I do not mind the 'privacy issue', but I do not want to SPAM your Server).

Also:
1. I can not access them.
2. There are so many; if I am ever asked for the 'bp-' (associated with a Bug Report that I file) which one would I send ?


> We saw something similar to this in bug 690841. I filed bug 691432 for a
> more reasonable way to handle this. We certainly shouldn't be generating
> this volume of crashes just because you have a faulty plugin!

When the Adobe Flash Plugin crashes I get:
1. A (for want of a better description) black and grey (crosshatched?) Screen in the 'Plugin area' saying that the Plugin crashed.
2. A message saying to 'Click here to send an Error Report' (so I assume that I send ONE Report for the ONE Tab that I click the message for).
3. A message that I can reload that ONE Tab (no others reload - so I assume that no others send Reports either (unless I click in them)).


> ... volume of crashes just because you have a faulty plugin!
I suspect that most people "just use the Browser" (for various reasons) and that some actually use the 'Happy/Sad Feedback' (rather tersely in some cases) but fewer still use "about:crashes" and submit Reports utilizing the Info; thus they may well be generating TONs of Reports also, but do not know it.


My concerns are:
1. SPAMming the Server (multiplied by the millions of Users affected).
2. Having too many Entries on my "about:crashes" to sift through.
3. OTHER People have privacy concerns and we (Mozilla) have a Policy to protect them (they only want to send the Report IF they choose to do so (using the "Error Reporter"), instead they may be sending hundreds/thousands of Reports per day without their knowledge).


Should this be increased from Major, what is the noise like on your end ?

Comment 8

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #6)
> ...
> We saw something similar to this in bug 690841. I filed bug 691432 for a
> more reasonable way to handle this. We certainly shouldn't be generating
> this volume of crashes just because you have a faulty plugin!
Something similar to this (excess, inaccessible Reports) was FIXED in Bug 559226 .

Comment 9

[(not currently active) Ted Mielczarek]

(In reply to Rob from comment #7)
> I only see the "Error Reporter" pop up a couple of times so the Crash Reports
> are being sent without me choosing to send them. 
> 
> They can be called "legitimate crash reports" but they were not "permitted
> to be sent by me" allowing the "Error Reporter" to send them (I do not mind
> the 'privacy issue', but I do not want to SPAM your Server).

As I explained in comment 1, these crash reports have not been submitted. (Only entries with the bp- prefix in about:crashes have been submitted.) Plugin crashes are not automatically submitted, so they can just pile up like this currently.

> Also:
> 1. I can not access them.

This is a separate issue, there was a server problem (discussed in comment 2 and comment 3) that's preventing them from being loaded.

> 2. There are so many; if I am ever asked for the 'bp-' (associated with a
> Bug Report that I file) which one would I send ?

This is unfortunate, certainly, we need to fix bug 690841 and bug 691432 to prevent this many reports from accumulating.

> Should this be increased from Major, what is the noise like on your end ?

As I stated before, these crashes are not actually being submitted to our server.

Comment 10

[Rob]

I tried running FF 7.01 overnight. There were no 'Crash Reporter events' but I have lots and lots of 'about:crashes Reports' (that when I click on they will not "bp-", nor are they accessible).

- I tried the [Remove All Reports] Button. The first time I did that there was a brief moment of Drive clicking (like they were being erased) and then they were gone. 

- I clicked on the URL Bar to re-access the Page and then they were back. 

- I tried the [Remove All Reports] Button again and they were all gone instantly, no Drive access necessary.

- I clicked on the URL Bar again to re-access the Page and now they are all back. 


(In reply to Ted Mielczarek [:ted, :luser] from comment #6)
> I'm telling you that I believe these are all legitimate crash reports, but
> you don't have to believe me if you don't want. :)

If I can not doubt their existence can I doubt their deletion ?  :)

Comment 11

[(not currently active) Ted Mielczarek]

That bug is already known, it's bug 583738.

Comment 12

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #11)
> That bug is already known, it's bug 583738.
(And over a year old :( ).

I deleted over 6100 PENDING Reports in "C:\Documents and Settings\HP_Administrator.HOME\Application Data\Mozilla\Firefox\Crash Reports\pending" (took half a minute).

I reloaded about:crashes and now there 24 "bp-"ed Reports showing.

While I type this Report 4 more 'about:crashes Reports' automagically appeared in that Directory. (Obviously this occurred without me realizing I am crashing (now I sound like my neighbor); since the ER did not popup nor are other Tabs (with Flash) crashed).

Comment 13

[Rob]

I reloaded that Directory, I now have 8 Reports there.

I checked about:crashes and I have 28 Reports (4 are not bp-ed), thus Aurora is polling another Directory (or making up Reports on it's own) to derive it's about:crashes List.

Comment 14

[(not currently active) Ted Mielczarek]

Submitted reports are in the "submitted" directory next to the "pending" directory.

Can you try resubmitting (by clicking on it) one of the new crashes? I'd really like to find out what is crashing you. Alternately, if that doesn't work for some reason, if you could email me one of the files I could take a look.

Comment 15

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #14)
> Submitted reports are in the "submitted" directory next to the "pending"
> directory.
> 
> Can you try resubmitting (by clicking on it) one of the new crashes? I'd
> really like to find out what is crashing you. Alternately, if that doesn't
> work for some reason, if you could email me one of the files I could take a
> look.

Those deletions did not last long and the Directory is filling up again (despite only having a couple of crashes (that popped up the "Mozilla Crash Reporter") in the last few days).

As usual most are not accessible, here is one that is from a minute ago (and no I did not see the "Mozilla Crash Reporter" at that time):

bp-bb915bd8-8d94-499f-919d-778c22111016

Comment 16

[Rob]

(In reply to Rob from comment #15)
> (In reply to Ted Mielczarek [:ted, :luser] from comment #14)
> > Submitted reports are in the "submitted" directory next to the "pending"
> > directory.
> > 
> > Can you try resubmitting (by clicking on it) one of the new crashes? I'd
> > really like to find out what is crashing you. Alternately, if that doesn't
> > work for some reason, if you could email me one of the files I could take a
> > look.
> 
> Those deletions did not last long and the Directory is filling up again
> (despite only having a couple of crashes (that popped up the "Mozilla Crash
> Reporter") in the last few days).
> 
> As usual most are not accessible, here is one that is from a minute ago (and
> no I did not see the "Mozilla Crash Reporter" at that time):
> 
PR_FindFunctionSymbol
> bp-bb915bd8-8d94-499f-919d-778c22111016

I found a few I could open, some 300+ clicks later  :(  +  :(

PR_FindFunctionSymbol 
bp-a67bf4d2-e53f-4d23-9a72-71e602111016

Null Signature (This one DID pop up the "Mozilla Crash Reporter")
bp-e36d073c-a91b-4a1a-8f20-1293f2111016

AppendUTF16toUTF8(nsAString_internal const&, nsACString_internal&) Bug 691462 (This one DID pop up the "Mozilla Crash Reporter")
bp-caa414e2-e3cd-4a93-9bcb-e61f92111016

Null Signature (This one DID pop up the "Mozilla Crash Reporter")
bp-cb8b5aeb-066a-4ba7-8c0c-8dcba2111016

AppendUTF16toUTF8(nsAString_internal const&, nsACString_internal&) Bug 691462 (This one DID pop up the "Mozilla Crash Reporter")
bp-477c332c-d3eb-4a32-b0bd-334d32111015

AppendUTF16toUTF8(nsAString_internal const&, nsACString_internal&) Bug 691462 (This one DID pop up the "Mozilla Crash Reporter")
bp-b07f0ef3-5605-43a1-951d-201ec2111015


Those are the few accessible Reports (out of hundreds of Reports in the last 20 hours), with only a few (expected) Crash Reporter pop ups.


It seems that if a 'Reportable Event' (listed in the about:crashes List) is 'pre-bp-ed' then I can click on it and get a Report. If I must click-to-submit and await the "Throbber" to 'bp' it, then when I click on that result, I get a "Crash Not Found" Page.

Comment 17

[:aceman]

The top 2 crash reports from comment 0 are accessible now. Does that change anything?

I suggest using this bug for the "[@ PR_FindFunctionSymbol ]" signature. "AppendUTF16toUTF8" is already discussed in bug 691462, where the original reporter Rob is already commenting.

Comment 18

[(not currently active) Ted Mielczarek]

Rob: your crashes are all from npmozax.dll, FWIW. This is some sort of ActiveX plugin wrapper. Turns out that that's the exact same plugin as bug 690841...

Comment 19

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #18)
> Rob: your crashes are all from npmozax.dll, FWIW. This is some sort of
> ActiveX plugin wrapper. Turns out that that's the exact same plugin as bug
> 690841...
You renamed this Report (692685) but Bug 690841 Comment 9 seems to says it all.
Is there a reason that THIS Report was not marked as a dupe of Bug 690841 ?


Would it have been best to dupe this (692685 -> 690841) and rename Bug 690841 

from:
"Pending crash reports from plugin crashes don't get cleaned up if the browser never crashes"

to:
"[@ PR_FindFunctionSymbol ] - Constant crashes with npmozax.dll, most are not submitted nor cleaned up."

?

Comment 20

[Rob]

In [Tools][Add-ons][Plugins] I have this Entry:

"
/!\ MetaStream 3 Plugin has been disabled due to security or stability issues [More Information] https://addons.mozilla.org/en-US/firefox/blocked/p32

MetaStream 3 Plugin 3.2.2.26 (Disabled)
MetaStream 3 Plugin r4
Last Updated			Friday, February 20, 2004
"


Why don't we do the same thing for:

"
Mozilla ActiveX control and plugin support 1.0.0.4 (Disabled)
Mozilla ActiveX control and plugin module
Last Updated			Monday, December 05, 2005
"

?

Comment 21

[(not currently active) Ted Mielczarek]

It's sort of arbitrary. :) We resummarized bug 690841 to be specifically about fixing the problem where a plugin that constantly crashes causes your pending crash reports directory to fill up. I've resummarized this bug to be about this particular bad plugin crashing. We might fix something in Gecko, or we might simply blacklist this plugin.

Comment 22

[Rob]

(In reply to Ted Mielczarek [:ted, :luser] from comment #21)
> It's sort of arbitrary. :) We resummarized bug 690841 to be specifically
> about fixing the problem where a plugin that constantly crashes causes your
> pending crash reports directory to fill up. I've resummarized this bug to be
> about this particular bad plugin crashing. We might fix something in Gecko,
> or we might simply blacklist this plugin.

Did you miss the prior comment ?, Comment 20 would be a fix ? Do we STILL need this 6 year old Plugin ?

I disabled the Plugin and was able to run WMP on the Page at: http://nunzioweb.com/streaming_audio-example.htm as well as the Video Examples.


There is a lengthy Thread about npmozax.dll here: http://forums.mozillazine.org/viewtopic.php?f=38&t=95187 which is where I got that test Link from.

Comment 23

[Benjamin Smedberg]

Not sure where plugin blocklist bugs go, but I'm taking a guess and cc'ing Kev/Johnath... we should just blocklist all versions of npmozax.dll since I'm almost positive that it relies on XPCOM features that mean it can't possibly work.

Comment 24

[Matthias Versen [:Matti]]

Blocklist requests seems to go into /dev/nul :-(

Comment 25

[Justin Scott [:fligtar]]

(In reply to Matthias Versen (Matti) from comment #24)
> Blocklist requests seems to go into /dev/nul :-(
> 
> *** This bug has been marked as a duplicate of bug 692149 ***

I manage the blocklist as one of about 1000 other responsibilities and don't have time to read tens of comments and decipher the problem. The best way to get something blocked is to file a new block request from the template at https://wiki.mozilla.org/Blocklisting that summarizes the exact problem, the severity of impact to users, and all of the needed metadata to block. Thanks!

Comment 26

[Matthias Versen [:Matti]]

Sorry Justin, i was just a little bit frustrated by the slow response for such an urgent problem but it should not be offending to you or any other responsible person.

I know that communicating with the various triagers is difficult but I'm sure that nobody from us knew that we should use the templates. I have created bug 695927 with the template for this issue.

Thanks in advance
 Matthias

Comment 27

[Rob]

Reporter: I am happy that this Bug Report has been resolved (in a manner likely to eventually be successful) by making a 'proper' "Blocklist Request" in Bug 695927.

Thank you.

Comment 28

[Rob]

Reporter: I am happy that this Bug Report has been resolved (in a manner likely to eventually be successful) by making a 'proper' "Blocklist Request" in Bug 695927.

Thank you.


PS: A quick fix for the end-user who may read this. 

This fixes everything (from the User's point of view) and prevents the Hard Drive from filling with (usually) non-submittable Bug Reports.


Simply go into [Tools][Add-ons][Plugins] and disable:

"
Mozilla ActiveX control and plugin support 1.0.0.4 (Disabled)
Mozilla ActiveX control and plugin module
Last Updated			Monday, December 05, 2005
"