308402 - After running for some time, attempting to clear the cache, or just closing the programme, will not only crash the programme, but Windows too. [@ ati3duag+12109f]

Status: RESOLVED INCOMPLETE

(Core Graveyard :: GFX: Win32, defect)

Description

[David E. Hockin]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.8) Gecko/20050511
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.8) Gecko/20050511

Running a Pc with a 3meg Athlon and WnXPpro2 with 2Gb ram and 256Mb ATI Saphire
videocard. After running Mozilla for a time, it frequently crashes itself when I
try to clear the cache before closing, and often just when I try to close the
programme (without touching the cache). It not only crashes itself, but the
whole PC too - with the PC switching right off, and then rebooting itself, with
a blue screen of death announcing that the Prefs have been buggered, as has the
Bookmark file, and other things I cannot recognise. Luckily I save the BM file
regularly to another drive to circumvent this. 

This has become an absolute pain in the butt! I'm losing my temper now with the
PC - its only a matter of time before I stick my boot into it, I'm so frustrated
with it. I always thought Mozilla (and Netscape) was lightyears better than the
dreaded IE, butI'm beginning to doubt this. Its wearing nme down - and I'm now
71, so I don't have long left!

Reproducible: Didn't try

Steps to Reproduce:
1. I don't need to try and reproduce it - I'm trying my damndest to avoid it!
2.
3.




There is no module report! It crashes the whole PC instantly! Blank screen and
power off. Self reboot.

Comment 1

[timeless]

see bug 307577, follow the steps (be sure to configure your system to dump a
core: win-break>advanced>startup and recovery|settings>write debugging
information <- the bigger the better)

the problem can be anything, but i'm willing to debug it until we show it's
something that you can direct elsewhere.

Comment 2

[OstGote!]

(In reply to comment #1)
> see bug 307577, follow the steps (be sure to configure your system to dump a
> core: win-break>advanced>startup and recovery|settings>write debugging
> information <- the bigger the better)

or: Right click on my 'My Computer', select properties > advanced > startup and
recovery | settings > write debugging information <- the bigger the better

Comment 3

[David E. Hockin]

Sorry Ostgote - I don't know what OS you are runninhg, but my WinXPpro2 has
nothing like you describe!

 Rt click My Computer>Properties>Advanced only offers a choice of "Run with
different credentials" or "Run in separate memory space"! 

Is what you are talking about actually in :-

Control Panel>System Properties>Advanced>Startup and recovery"?

Comment 4

[timeless]

yes

Comment 5

[David E. Hockin]

Its just crashed itself and the PC but there is no file called
%SystemRoot%\MEMORY.DMP to be found! I think it crashes too fast to record
anything - it appears to be instantaneous. 
I had set it to do a small 64K dump, but nothing had happened. 
When I now try to set it to do a complete memory dump, I get a message that if
this is done, and "the pagefile on Vol.C: is less than 2047Mb, then the system
may not be able to create a debugging information file if a stop occurs.
Continue anyway? Yes/No?" 
So I told it "No" and its left set to "kernel dump"
Ram is 2 gigabytes, (and video ram is 256Mb) so I don't see how its possible to
record anthying that big in the instant of a complete crash..

Comment 6

[timeless]

a few things:
1. increase your swap for c: to be at least 2048, that'd let you do a full dump.
2. the reboot is caused by the os, generally, which means the os chooses when 
to reboot, it's not really instantaneous, as such it can choose to write memory 
to disk before rebooting.
3. start>settings>control panel>administrative tools>event viewer>system. you 
should be able to see the entries that talk about rebooting

one source is "Save Dump", which should say something like:
The computer has rebooted from a bugcheck.  The bugcheck was: 0x1000007f 
(0x00000008, 0x80042000, 0x00000000, 0x00000000). A dump was saved in: 
C:\WINDOWS\Minidump\Mini081605-01.dmp.

For more information, see Help and Support Center at 
http://go.microsoft.com/fwlink/events.asp.

Please see if there's a Save Dump source.

the other event entry you should see near the crash is:
Microsoft (R) Windows (R) 5.01. 2600 Service Pack 2 Uniprocessor Free.

For more information, see Help and Support Center at 
http://go.microsoft.com/fwlink/events.asp.

or something like that. in really amusing cases, my computers have reported the 
computer restarting unexpectedly, if you have an entry like that, please copy 
it here.

Comment 7

[David E. Hockin]

Attachment: latest minidump of crashing Mozilla

Comment 8

[David E. Hockin]

pagefile.sys on C: is 786,432 kb

and there are no others on any of the other drives (D: to J:)

There is no longer any mention anywhere using the old term swap file, and I
can't find any means of changing it! Come back Win4WG3.11 ! All is forgiven...

Comment 9

[David E. Hockin]

The text with the minidump was :-
The computer has rebooted from a bugcheck.  The bugcheck was: 0x1000008e
(0xc0000005, 0xbfb8a09f, 0xa1465978, 0x00000000). A dump was saved in:
C:\WINXP2\Minidump\Mini091605-01.dmp.

Comment 10

[timeless]

kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

KERNEL_MODE_EXCEPTION_NOT_HANDLED_M (1000008e)
This is a very common bugcheck.  Usually the exception address pinpoints
the driver/function that caused the problem.  Always note this address
as well as the link date of the driver/image that contains this address.
Some common problems are exception code 0x80000003.  This means a hard
coded breakpoint or assertion was hit, but this system was booted
/NODEBUG.  This is not supposed to happen as developers should never have
hardcoded breakpoints in retail code, but ...
If this happens, make sure a debugger gets connected, and the
system is booted /DEBUG.  This will let us see why this breakpoint is
happening.
Arguments:
Arg1: c0000005, The exception code that was not handled
Arg2: bfb8a09f, The address that the exception occurred at
Arg3: a1465978, Trap Frame
Arg4: 00000000

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


EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at "0x%08lx" referenced 
memory at "0x%08lx". The memory could not be "%s".

FAULTING_IP: 
ati3duag+12109f
bfb8a09f 895024           mov     [eax+0x24],edx

TRAP_FRAME:  a1465978 -- (.trap ffffffffa1465978)
ErrCode = 00000002
eax=00000000 ebx=00000000 ecx=e1656010 edx=e4f34010 esi=e48f5010 edi=00000000
eip=bfb8a09f esp=a14659ec ebp=e48f5010 iopl=0         nv up ei pl nz na pe nc
cs=0008  ss=0010  ds=0023  es=0023  fs=0030  gs=0000             efl=00010202
ati3duag+0x12109f:
bfb8a09f 895024           mov     [eax+0x24],edx    ds:0023:00000024=????????
Resetting default scope

CUSTOMER_CRASH_COUNT:  1

DEFAULT_BUCKET_ID:  DRIVER_FAULT

BUGCHECK_STR:  0x8E

LAST_CONTROL_TRANSFER:  from 00000000 to bfb8a09f

STACK_TEXT:  
e48f5010 00000000 e48f5010 e1656010 e1657530 ati3duag+0x12109f


FOLLOWUP_IP: 
ati3duag+12109f
bfb8a09f 895024           mov     [eax+0x24],edx

SYMBOL_STACK_INDEX:  0

FOLLOWUP_NAME:  MachineOwner

SYMBOL_NAME:  ati3duag+12109f

MODULE_NAME:  ati3duag

IMAGE_NAME:  ati3duag.dll

DEBUG_FLR_IMAGE_TIMESTAMP:  40c91980

STACK_COMMAND:  .trap ffffffffa1465978 ; kb

FAILURE_BUCKET_ID:  0x8E_ati3duag+12109f

BUCKET_ID:  0x8E_ati3duag+12109f

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

this is the best i can do w/o the drivers, and even then i'd just get a 
slightly longer stack, you need to report the crash to ati or try to upgrade 
your driver. I don't believe we have contacts w/ ati.

please contact ati and offer them the dump. if you do get this resolved w/ ati, 
please follow up here (heck, if you get a ticket id, and are willing to post it 
here, that'd be good too).

Comment 11

[timeless]

fwiw, you can get to the swap file settings via: 
start>settings>control panel>system>advanced>performance>advanced>change

Comment 12

[David E. Hockin]

After a really good Mozilla crash, the system DID a full dump - 124Mb, which is
far bigger than ATI's file reporting limit of 6Mb. I've reported the crash to
them, and their website list of graphic cards doesn't even include mine, but
stupidly, their ticketing system demands an answer, even if its not correct! I
await their reply.....I've also asked ASUS - the makers of the card, for help. I
don't suppose anyone here really wants a 124Mb dump file...........Bed-time
reading?!

Comment 13

[timeless]

i read such things. hopefully you have a sever from which i can get it.
otherwise i'd have to arrange a drop box, which would be fairly hard atm.

anyway, i think from windbg (see other referenced bugs for steps to get it and
set it up), if you could do:
lm

that should get a list of the loaded modules. which would probably be helpful
for me to have when you make the dmp file available :). note that i won't
possibly get to it this weekend, and holidays are coming which disrupts my
normal work week and means i'll have to work more for normal work, so i probably
won't be able to look for a month. i'm slowly growing other people who will be
able to read these dumps, so maybe if you can arrange to have the files
available, one of them will be able to pick up your pieces and figure something out.

you should probably at least send the earlier !analyze -v output from comment 10
and the url to this bug report. possibly also the minidump since it's <6mb :)

Comment 14

[David E. Hockin]

Yesterday, after installing a pair of SATA mirror drives in my dual drive/dual
boot PC - one for Win and one for Suse9.2,  Mozilla tried a new one - it froze
the PC solid!. This resulted in the SATA mirrors being buggered, the MBR on
drive C: being ditto, and Grub also being very ditto.

Result - one unworkable PC! It only could get as far as "Grub.........Error 
Ceom" (or Geom - couldn't tell which letter). No matter, it didn't make any 
sense. Tried repairing via the LXF Suse CD. "Repaired", so it said. "No it 
wasn't" said the PC! Tried reinstalling Suse - OK till it came to reboot, 
and back to the error message. Tried it again but with the Sata drive chips 
disabled in Bios. No different. Tried reinstalling XP, equally - ditto. Both 
XP and Suse had done the majority of their install, but could not continue past 
the reboot stage.

Eventually a help list of Win commands was discovered on the Win 
setup/repair, and a couple gave "Fixboot" and "Fix MBR" - the latter with 
dire warnings of possible loss of info of Win partitions. Took deep breath 
and crossed all available fingers and toes, and hit FixMBR. "Done" it said. 
Suddenly XP setup returned to life, and after losing mouse, keyboard and 
other (non?)essentials, WinXP returned to life. This email is proof!

Now, I need some dutch courage before trying to:
1: resuscitate the SATA drives, and
2. re-reinstall Grub (plus water and relief!) and any other damage done to 
Suse.....

Its taken two days of cursing, hair pulling, blood sweat and tears, not to 
mention sheer fright.........I really DO hate computers, and have been 
hating them now for over 40 years.

I am giving up Mozilla and returning to Netscape 8.x - it hasn't had this
problem, and at 71, I'm too old for this hassle. I don't need troubles like this
with my data. Thanks guys for trying, but I'm giving up on this. Cheers to all.