340626 - BAD_POOL_HEADER crash going to myspace.com

Status: RESOLVED INVALID

(SeaMonkey :: General, defect)

Description

[jessestrachman]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.0.2) Gecko/20060404 SeaMonkey/1.0.1
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.0.2) Gecko/20060404 SeaMonkey/1.0.1

I opened a new tab by middle-clicking on the myspace.com logo in a previously opened tab.

Reproducible: Didn't try

Steps to Reproduce:
1. Go to the myspace.com "You must log in to do that!" screen.
2. Middle-click the myspace.com logo to open a new tab.

That's it.

Actual Results:  
Got a blue screen of death with a BAD_POOL_HEADER message. When the machine rebooted, all my saved passwords were gone. Very strange.

Expected Results:  
Should have opened a new tab and loaded the page.

I have the mini-dump for the crash. Unfortunately, because it was a crash in Windows, it might not be useful as I believe it will only contain minidump info for Windows and not for SeaMonkey. I hesitate to attach it to this bug because it may contain personal information. Contact me if you want it.

Comment 1

[jessestrachman]

I did try to reproduce this crash and could not.

Comment 2

[timeless]

congrats, your video card driver almost certainly killed your machine.

there are some nice bugs in bugzilla where i explain what to do.

first right click my computer and select properties, then go to advanced. startup and recovery settings.

change write debugging information to Complete memory dump.

remember that you only get one and it costs as much as you have memory, so you'll want to rename it as soon as your system reboots (if you have space to store a couple). once you have it, you can follow most of the instructions on http://developer.mozilla.org/en/docs/How_to_get_a_stacktrace_with_WinDbg to get windbg and setup symbols.

lastly, file>open crash dump and !analyze -v -f.

you can paste the output from that here and someone can help you figure out which vendor you should yell at (hint: it isn't us, we don't write hardware drivers which are essentially the only things that are allowed to trigger bluescreens).

Comment 3

[jessestrachman]

Seems unlikely that I will be able to do this; this box has a gig of RAM and not a lot of free disk space. If by some miracle it does happen again, I'll notify. In the meantime, you'll probably want to stash this bug away somewhere appropriate. FWIW, I feel like this is likely to be related to bug 335697 (a.k.a. bug 303823) in Firefox, which I also reported. I'm also very curious to know why all of my stored passwords would have been wiped. That seems like it would not be video card related, but then, stranger things have happened.

Comment 4

[timeless]

you can select a minidump instead of a full dump. i prefer full dumps, but minidumps work well enough for basic driver blame, they're just not as good when you want to convince a vendor to "fix the problem" same overall step, just a different item in that listbox.

don't bother notifying us if you reproduce it without a .dmp. it's a waste of my time.

Comment 5

[jessestrachman]

Here's the output. Certainly not trying to waste your time. Correct me if I'm wrong, but it appears to have been a problem with ibmfilter.sys, which (according to Google) seems to have a host of related problems. If it's all right with you, I feel like we should mark this bug as INVALID and I'll pursue a fix with IBM.

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

BAD_POOL_HEADER (19)
The pool is already corrupt at the time of the current request.
This may or may not be due to the caller.
The internal pool links must be walked to figure out a possible cause of
the problem, and then special pool applied to the suspect tags or the driver
verifier to a suspect driver.
Arguments:
Arg1: 00000020, a pool block header size is corrupt.
Arg2: 817e7000, The pool entry we were looking for within the page.
Arg3: 817e7408, The next pool entry.
Arg4: 0a810000, (reserved)

Debugging Details:
------------------

***** Kernel symbols are WRONG. Please fix symbols to do analysis.

*************************************************************************
***                                                                   ***
***                                                                   ***
***    Your debugger is not using the correct symbols                 ***
***                                                                   ***
***    In order for this command to work properly, your symbol path   ***
***    must point to .pdb files that have full type information.      ***
***                                                                   ***
***    Certain .pdb files (such as the public OS symbols) do not      ***
***    contain the required information.  Contact the group that      ***
***    provided you with these symbols if you need this command to    ***
***    work.                                                          ***
***                                                                   ***
***    Type referenced: nt!_KPRCB                                     ***
***                                                                   ***
*************************************************************************

FAULTING_MODULE: 804d7000 nt

DEBUG_FLR_IMAGE_TIMESTAMP:  40b23a2e

BUGCHECK_STR:  0x19_20

POOL_ADDRESS:  817e7000 

CUSTOMER_CRASH_COUNT:  1

DEFAULT_BUCKET_ID:  DRIVER_FAULT

LAST_CONTROL_TRANSFER:  from 8054b741 to 80533336

STACK_TEXT:  
WARNING: Stack unwind information not available. Following frames may be wrong.
a9dd17b8 8054b741 00000019 00000020 817e7000 nt+0x5c336
a9dd1808 aa5367e2 817e7008 00000000 81929558 nt+0x74741
a9dd1a4c 804e37f7 a9dd1df0 81929500 81929548 ibmfilter+0x57e2
a9dd1b3c 8056316c 867e07b8 00000000 83627758 nt+0xc7f7
a9dd1bc4 8056729a 00000000 a9dd1c04 00000040 nt+0x8c16c
a9dd1c18 80570b73 00000000 00000000 a47d9801 nt+0x9029a
a9dd1c94 80570c42 0440782c 00100080 044077cc nt+0x99b73
a9dd1cf0 80570d78 0440782c 00100080 044077cc nt+0x99c42
a9dd1d30 804de7ec 0440782c 00100080 044077cc nt+0x99d78
a9dd1d64 7c90eb94 badb0d00 04407794 00000000 nt+0x77ec
a9dd1d68 badb0d00 04407794 00000000 00000000 0x7c90eb94
a9dd1d6c 04407794 00000000 00000000 00000000 0xbadb0d00
a9dd1d70 00000000 00000000 00000000 00000000 0x4407794


FOLLOWUP_IP: 
ibmfilter+57e2
aa5367e2 ??               ???

SYMBOL_STACK_INDEX:  2

FOLLOWUP_NAME:  MachineOwner

SYMBOL_NAME:  ibmfilter+57e2

MODULE_NAME:  ibmfilter

IMAGE_NAME:  ibmfilter.sys

STACK_COMMAND:  kb

BUCKET_ID:  WRONG_SYMBOLS

Followup: MachineOwner
---------

Comment 6

[timeless]

you probably didn't run .symfix+ c:\symbols before running !analyze -v, but yeah, that's close enough (although i don't mind seeing a bit more output).

yes, you can invalidate it when you feel like it. i don't really care too much about bookkeeping, so I'm just as soon to leave a bug open especially when I'm still willing/interested in talking to the reporter. usually I'll mark something invalid when i really don't think the reporter can add anything that would help the issue and it's not something we can do anything about.

I appreciate the effort it takes to do this and am always glad to do hand holding, it's just a strange balance as I need to do other things but like helping. Good luck getting IBM to resolve this. The last blue screen I dealt with from IBM involved IBM's branded RapidRestore, neither the rapidrestore vendor nor IBM were really interested in dealing with it, but most importantly, the laptop owner had no use for the product, so we resolved the issue by uninstalling it. So definitely find out if the driver is optional :).

Comment 7

[timeless]

i suppose i should clarify, a quote from:

http://www.msfn.org/board/index.php?act=Print&client=printer&f=70&t=31886
• You are using certain filter drivers. For example, the problem is known to occur if you are using the Ibmfilter.sys driver. This driver is installed by the IBM Rescue and Recovery with Rapid Restore software. The problem may also occur with other filter drivers."

Indicates that the blue screen you got probably *is* a match for the rapid restore blue screens we were getting.

unless you need rapid restore, i'd suggest you just uninstall it :(. you're of course welcome to try to fight their collective tech supports, but i gave up. and i'm fairly tolerant of tech support, since i was just listening to elevator music while i was hacking gecko or chasing other crashes.

Comment 8

[timeless]

oh, and while i'm causing email. i guess almost certainly isn't perfect. i've had mozilla / windows die because of virus scanners, video drivers, and rapid restore. the odds of a crash being rapid restore seemed really low and given that i didn't know your computer was an ibm laptop, it wasn't a good guess. had i known it was an ibm laptop i might not have said almost certainly. but you get a free refund with my almost certainly statement :).

Comment 9

[jessestrachman]

Yes, this problem actually started occurring with other browsers as well, so this is most definitely not a bug in Firefox. I don't think I need rapid restore, so I'll look into ways of uninstalling it. Thanks again for the help.