m50sv vista x32 serious bsod frequency

x3/1bsa specs so the only real difference is a t8100 core2duo instead of a t9x00 , and a 1280x800 res screen vs 1440 or 1900.

factory software only it pulled a bsod during the configuring phase of vista first boot setup(bugcode_usb_driver). restart had me in vista normally, but the system would fault at totally random times with irql_not_less_or_equal or, and any attempt to use the fingerprint reader or camera lead to another bugcode_usb_driver 0x000000FE stop.

several factory restorations and a multitude of driver installation, windows update, and vista sp1 installs with varying orders and i fixed the bugcode_usb_driver, and the camera and fingerprint reader are usable(at least so far.

however, i am now getting a seemingly random smattering of bsod stops now. opening tabs in either firefox or ie randomly result in driver_irgl_not_less_or_equal or plain irql_not_less_or_equal stops typically in 0x000000D1 with dxgkrnl.sys and 0x0000000A with no file referenced respectively. additionally i've seen attempt_to_write_to_read_only_memory bsod in win32k.sys ( 0x0000000A) just changing my ip settings in the network card(lan not wireless).

the temp monitoring in the asus nb package report ok, and a bit of time running notebook hardware control showed perfectly acceptable temperatures in confirmation.

i can rule out errors from the initial boot bsod thank to the factory resets, and the irql/driver_irql errors are respectively tied to poor drivers accessing the hdd/ram while one of the cores is in a delay state(ie generating an interrupt when it cannot be handled), or system ram going bad.

as i noted i am using the latest driver and packages from asus, while vista has sp1 and all updates except silverlight.

the only peripheral connected is a wired usb mouse from inland in the back right usb port, the lan cable, and the power cable.

uac is disabled presently, and wireless is off.
the vista memory test tool reported no errors, and a disk check found nothing either. i'm hoping to burn memtest86 shortly and try that but i am seriously short on belief its as simple as a system ram problem(maybe if it tested the dedicated vram as well, i suppose that could generate a problem like this, but i know of no memtest packages for dedicated video memory).

i have until monday for the exchange policy on this unit, but given i have found no reports of this level of problems googling, i have to assume its something specific to this unit itself.

if anyone has an ideas short of installing xp(i only have an xp home disc i think i recall needing pro for dual cores, 2gb+ram . this might have changed in sp2 or sp3.. shrugs), i would much appreciate some feedback. i haven't seen this many issues since my win95 a/b/c days.. heck i couldn't even run the vista performance index without a bsod.

help. please. with a cherry on top and all..

 

Could you upload the dumps so I can have a more detailed look?

 

First thing I´d try is a memtest.

 

@axelm:"the vista memory test tool reported no errors, and a disk check found nothing either."
memtest86 from my fedora core 2 dvd reported ok as well, but may be deprecated enough a version to make a difference(?).

@wishmaker which stop error wold you like me to catch a sample of for you first? and do you want that handled with dbgx86 and debugwiz first(assuming yes).

edit: @wishmaker -in looking at the datestamps the main dump and minidump are both from the 2nd of this month and relate to the usb error, so apparently every bluescreen since then has been refusing to dump to disk even though the setting indicates it should have overwritten the old dump. worse, i checked and none of them are showing up in problem reporting, and the only evidence in the event logs is the notice telling me that the last time windows was running before that boot point it terminated unexpectedly. sigh. i think this is the point where i say good grief. i've thrown the old dumps and their resultant outputs into a zip, and moved em- i will try to reproduce one of the errors again and hope it creates a dump this time. if it doesn't any suggestions?

@general population: i went through the memory tests, both vista and the version of memtestx86 i had a couple times, especially since i knew the ms kb indicates these could be from bad ram, and the third and second tries at the aero index refresh it simply restarted itself just after getting to memory performance, no stop or bsod, just black screen then bios is starting to post..

alas both test types came back clean over and over. so that just leaves driver hell, or something even more obscure.

 

To your initial question, every error should have its own dump but you make it clear that there has been a problem with this. I wanted to have a look at the dump, line by line, to see if there are any files behaving naughty and every error message recorded. Usually, when you debug a dump you have detailed info about what went wrong. Usually . My only suggestion for the moment is to identify why it does not save dumps (try to set up a full dump and not overwrite) and to recreate the conditions giving you a BSOD.


Try Running PRIME95 for minimum 3 Hours!!! 8 is recommended but I have a feeling you will not get that far and BSODS will appear. Make sure you write down exactly what happens if the dump is not logged.

I wonder if you will get this error : FATAL ERROR: Rounding was 0.5, expected less than 0.4. Not many people know but RAMS can lead to BSODS on any component.


P.S.: Attach a shot with the EVENTLOG!

 

it just locked up, no bsod, no reboot, no dump, and only posthumous events referring to it, test 7 for progress. only had one copy of prime95 running at the time and was setting up the shortcut to run a second alongside it as per the instructions(i had just found them). sensors as soon as restart was done report temps still good, prime95 didn't dump anything directly, but there is a app event that seems relevant :/ even if it is nowhere near what we need to diagnose a software issue. it does lean my opinion towards hardware or both however..

Log Name: Application
Source: Windows Error Reporting
Date: 6/4/2008 10:02:31 AM
Event ID: 1001
Task Category: None
Level: Information
Keywords: Classic
User: N/A
Computer: starbreaker
Description:
Fault bucket 46110463, type 5
Event Name: RADAR_PRE_LEAK_32
Response: None
Cab Id: 0

Problem signature:
P1: prime95.exe
P2: 24.14.1.0
P3: 6.0.6001.2.1.0
P4:
P5:
P6:
P7:
P8:
P9:
P10:

Attached files:
C:\Users\steelblueskies\AppData\Local\Temp\RDR9197.tmp\empty.txt

These files may be available here:

Event Xml:
<Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
<System>
<Provider Name="Windows Error Reporting" />
<EventID Qualifiers="0">1001</EventID>
<Level>4</Level>
<Task>0</Task>
<Keywords>0x80000000000000</Keywords>
<TimeCreated SystemTime="2008-06-04T14:02:31.000Z" />
<EventRecordID>1933</EventRecordID>
<Channel>Application</Channel>
<Computer>starbreaker</Computer>
<Security />
</System>
<EventData>
<Data>46110463</Data>
<Data>5</Data>
<Data>RADAR_PRE_LEAK_32</Data>
<Data>None</Data>
<Data>0</Data>
<Data>prime95.exe</Data>
<Data>24.14.1.0</Data>
<Data>6.0.6001.2.1.0</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
</Data>
<Data>
C:\Users\steelblueskies\AppData\Local\Temp\RDR9197.tmp\empty.txt</Data>
<Data>
</Data>
</EventData>
</Event>

unfortunately both a search and manual check turn up the indicated directory to be nonexistent. any idea what i can try next, or should i be dragging this back as defective, seems a 50% cpu load over a few minutes shouldn't be locking up a dual core setup, and certainly not via heat.

 

I am looking into this. Did Prime manage to write a log? No need to send it back just yet. We need to see if there is a problem with the notebook or with some driver Vista does not like. If worst comes to worst, I will slipstream my XP PRO SP2 for you and put it on RS. I will lend you my account so you can download it without restrictions.


This might be a long shot, did you do a BIOS update? If you did, could you force an older version via a dos prompt?

 

rofl sadly the m50sv series only has one bios version available still and thats the original. prime95 failed to output the textfile it indicated it would, but this may be because i was using the torture test mode, as opposed to benchmark or normal gimps workloads, and the length was so short that may be a factor as well. i'm throwing wprime through stability test mode now since it handles logical cpu counts in app instead of double launches, then perhaps try prime 95 again if it survives.

if there is any other place prime95 might have left a log(since there was no memorydump, minidump, or debug generated in the usual windows locations, and no new files in the prime95 or root folders) any idea where it might be? i'm assuming if none of those exist then the answer will be no, but i could be wrong.. shrugs. more information shortly i think, wprimes up to 40% since i started writing this..

distributed.nets client would probably murder this system too, but i will try that next meantime as i have a bit of experience with it, assuming i don't hear any better ideas before then.
no rush, my brain melted already trying to figure out what errors don't permit crashdumps but do stick at the blue stop error in microsofts kb for hours.(using a p200 running win98se with 64mb ram to do the searching, fun fun fun )


edit2- it survived wprime in 2 core mode finishing the speed test in 37.971 sec, and the stability test in 1231.877 sec. given that i'm going to try the cpu heavy prime95 mode, instead of the mixed or light modes, if it beats the previous run length i'll reverse into memory stress mode and compare runtimes.

 

You covered the right areas for logs or anything related. If there are none in those folders, then it did not write one because as you said, period was very short. Did it even calculate something?

L.E: we have insufficient info . We need a memory dump a log, anything that can give us more than your first post .

 

ran prime95 again two copies at once each set to a separate cpu affinity using the second torture test option(in place large fft's) with 256 mb ram allocated to each. left to make coffee, came back when it was done to a blue screen. irql_not_less_or_equal

stop: 0x0000000A( 0xC0F47DC0, 0x0000000B, 0x00000000, 0x822F840C)

there was no file listed after that, and it just stopped at "initializing for crash dump..." which explains the lack of output. the numlock/etc keys were unresponsive so its definitely a hard lockup occuring still, just happening after windows detects a problem and throws an error.

unfortunately this means we get no logs, except the notes saying windows can tell it didn't exit normally. and yes even with two prime95 copies running, one to each core, there was no log output from the application either :/

would the dump from the bugcode_usb_driver stop errors i was getting before i factory restored, updated to sp1, and reinstalled all the drivers again potentially help? perhaps the same issue is still the root cause, and is now merely leading into other faults(ie if two stop error causing faults occur simultaneously windows can only report one in the bsod )?
that was also the last time i actually was getting crash/bsod/stop dumps.

 

Unfortunately, I have to go to a concert in 20 minutes . When I get back, I will have a look and see what we can do.

We could try and dissect that .


Can you RESET the BIOS on that thing? Sorry,PC habit . Speaking of RESET, let's do that to all your USBs

OPEN NOTEPAD...copy and paste the text between the lines ...

click on file...then...save as.. on the left side click on desktop...then in the file name box type in FIX.REG then in the box below that save as type..all files...then click save...you should see a new icon on your desktop ...double click it to run it..
should look like mine


-------------------------- Use text after this line --------------------------------
REGEDIT4

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Session Manager\Environment]

"DEVMGR_SHOW_NONPRESENT_DEVICES"="1"
-------------------------- Use text before this line --------------------------------

Double click on FIX.REG and say yes to the Merge Into Registry question

Unplug ALL USB devices.
Open Device Manager.
View, Show Hidden Devices.
Uninstall all devices under USB Controllers.
Uninstall any HP or unknown device under Printers.
If you have a yellow ? with unknown devices, uninstall all of the entries there as well.

When this is done, reboot TWICE....

NOTE: If you have a USB keyboard and/or mouse, you'll have to modify the instructions and leave enough parts for those to function.


I'm off , will be back .

 

bah it started getting weird after removing the usb controller and hubs etc, ie just locking up/freezing- broke down and dragged the old unit in for exchange early, the new unit of the same model has so far experienced a grand sum of zero lockups/bsod/etc. i'll be trying prime95 and a few other things shortly, but i'm set on it being either a bad factory copy of the install/recovery leading to persistent corruption, or hardware fault. granted the new unit exhibits the missing keystroke problem, but frankly i'll take that over what happened with the last unit, or even the risk of all that hell over again..

i thank and salute you wishmaker, and whereas it would have been good to work out the exact root cause for all that, i was rapidly running out of time to get the system set up for class and such.

 

Mate, I am sorry we could not identify the problem. You are most welcome and I hope the new machine does not yield the same result. Let me know how it performs in memtest and prime .