The BSOD struggle continues.

[Aran]

This computer BSODs 3 or 4 times a day. I reinstalled windows last night, thinking that would probably clear up any software issues that would cause driver conflicts that may perhaps have led to one of my BSODs. No dice, though the install went perfectly smoothly.

Here are the BSODs i'm getting now. After I reinstalled, there were no issues while I installed programs and updated windows. Everything went off without a hitch. I got the first BSOD (PAGE_FAULT...) while downloading Gal Civ 2 and ripping an ISO. The others have followed within 1-2 hours of each other.

PAGE_FAULT_IN_NON_PAGED_AREA 0x50
This happened once. I did some research and I think I found teh culprit: old sound card drivers that I installed off of my motherboard disc. They've been upgraded. This is the first BSOD I've gotten after 10 hours of error-free computer operation.
PFN_LIST_CORRUPT 0x4E
This is the same as the one i got before i reinstalled windows, just different parameters. Windows says it's my RAM, or something is corrupting my IO drivers. What the hell? The only IO drivers i've installed are Microsoft's own Intellipoint drivers for my Microsoft Comfort Optical Mouse 3000.
IRQL_NOT_LESS_OR_EQUAL 0xA
This is another driver issue according to MS. I'm at a loss for what it could be. I'm going to try updating my chipset drivers first before rummaging through all of my other hardware drivers.

I spent 4 hours running memtest86. It ran though 6 complete batteries of tests and turned up 0 errors. Chances are, it's not my RAM that's the problem.

Any suggestions? I'm really at a loss here.

[Salem]

> I got the first BSOD (PAGE_FAULT...) while downloading Gal Civ 2 and ripping an ISO
Maybe a power supply problem - driving HD and CD at the same time?
Your memtest wouldn't be driving peripherals at the same time.

Have you overclocked your machine - if so, step back a bit and try again.

Also, do you have any temperature monitors running, CPU, case etc.
My guess, it just gets hotter when it's busy, and that just makes something trip and cause a fault.

If you leave the machine on overnight, not doing anything, does it BSOD then?

[computation]

Maybe a power supply problem - driving HD and CD at the same time?

that was my first thought. my computer did that all the time. when i changed my power supply it stopped...then again, my situation was slightly different. not only was i getting the BSOD constantly. i heard a buzzing sound coming from my computer and it would just shut off arbitrarily (the bad way). the buzzing sound turned out to be my p[ower supply fan. i'm guessing it was overheating.

my extremely unqualified suggestion: open your box and make sure the power supply fan is spinning. take note of the amount of time your computer has been turned off. then, take note of how long it stays on without crashing. my theory is that, in the case it is overheating, the longer the down time, the longer it runs without crashing. this is because it has time to cool off. however, sometimes it could be off for a week, and still crash immediatly. that's why it will take several times to see the pattern.

..by the way, i'm drunk, sorry if this doesn't make sense

[Aran]

I don't think the overheating holds. I've played Oblivion for 3 hour stretches without crashes and then exited the game without anything going wrong. I've played Civ 4 for several hours without incident. That probably pushes my comp as hard as it's going to be pushed.

My power supply is a 500 Watt Rosewill... I'll link you to its page on newegg:

http://www.newegg.com/product/produc...82E16817182027

I have not overclocked this machine at all. In fact, I underclocked my RAM thinking that may help the BSODs. It did allay them a little bit.

[VirtualAce]

Looks like mobo issues.

PAGE_FAULT_IN_NON_PAGED_AREA 0x50


This should NEVER happen. A page fault occurs when the page of memory an application asks for has been paged to disk. The page, although existing on disk, is not available in RAM and therefore the CPU pukes out a page fault.

The problem with this error message is that is says it occurred in a non-paged area? How can you have a page fault in an area of memory that has not been paged to disk? Very odd.

I'd be extremely interested to see the register dump, memory dump, process dump, and anything else you can get your hands on that might find out what the BSOD's are being caused by.

IRQL_NOT_LESS_OR_EQUAL 0xA
This sounds like an error occurring while Windows attempts to invoke an interrupt. 0xA would be int 11h but it could represent an offset from a base value for the error message. This is certainly not being caused by software or applications. It could be caused by low-level drivers running at high privilege levels because AFAIK no-one can invoke interrupts in Windows. The interrupt vector table has been replaced by the IDT or the interrupt descriptor table. Interrupts should be mapped by the OS into the IDT. Part of the intial stages of a 32-bit OS boot sequence is to init and setup the IDT, GDT, and/or LDT if needed.

These errors are certiainly low level which leads me to believe you may have some defective circuits on the mobo leading to very strange errors. This is only a guess, but I've not seen these types of errors very often.

My guess is the circuits heat up and are stressed and thus fail causing the BSOD's. It only takes the PIC or some other circuit to return an incorrect value once to hose the system.

My guess is also that the CPU may be triple faulting and Windows XP is actually catching the triple fault and causing the BSOD. Windows 98SE would not catch a triple fault and would just reboot. A triple fault on the CPU is a completely unrecoverable error which means something very bad has happened either on the mobo, CPU, memory, or some application did a very very bad thing.

[Aran]

Well, the PFN_LIST_CORRUPT error just happened again. This time i had disconnected my DVD drive as per Salem's idea. It happened after Oblivion crashed (it's incredibly unstable...) after an hour or so of gameplay.

The PFN_LIST_CORRUPT and IRQL_NOT_LESS_OR_EQUAL errors are the most common. The PAGE_FAULT_IN_NON_PAGED_AREA error only happened once.

I've also recently (before unplugging my DVD drive) received 0x1A MEMORY_MANAGEMENT and 0x8E errors.

Bubba:
Sounds like you have an idea of what is going on. Thanks for your commentary. Can you give me any steps I can take to try to narrow this problem down to my mobo? I don't have hardware laying around that I can swap for everything...

I've uploaded all the "minidumps" windows has generated on this install. There are several of them. Knock yourself out, bubba.

http://members.entropysink.com/aran/debug

EDIT: my mobo is an ASUS A8N5X.

[JaWiB]

Are you certain it's not a problem with RAM? I've had two of those errors on my laptop (the page fault and IRQL one) and then at one point my laptop would not boot up and gave an error code for memory not detected. For me, these errors have only occurred a few times (its been a few days since the last problem, and more than a week since the BSODs before that)

Have you tried pulling out one stick of RAM (if you have two) or maybe just reseating it?

[Aran]

I ran memtest86 for like 4 hours, completing SIX batteries of tests without a single failure. I highly doubt the RAM is faulty.

[bithub]

It sounds like you have a memory issue. That means the problem is either the CPU, RAM, or hard drive. The first thing I would do is make sure your CPU isn't overheating. There are plenty of free temperature monitors out there, I suggest you install one.

[Aran]

I'm only running one 120mm fan currrently (exhaust), because the other one died a week after I got this computer. The exhaust fan pulls air across the processor and from above the proc fan. The RAM is situated BEHIND that (if you are looking at the CPU from where the exhaust fan is), meaning it's not getting anywhere near the cooling it should be getting. Also, the RAM is very close to the CPU. Could the RAM be overheating? Does RAM even overheat?

[Aran]

I downloaded AMD's "Cool'n'Quiet Dashboard" to monitor my CPU temperature. After running Oblivion for an hour and occasionally checking the dashboard, either the dashboard doesn't work, or I'm having no heat problems with the CPU at all. There's no more than 3 degrees of variance.

[VirtualAce]

Perhaps memory timings. ASUS is bad about detecting the correct memory timings but it normally takes the 'safe' slower route by increasing the timings.

If you have an ASUS with Corsair, the timings are supposed to be:

6
3
3
2.5T
2.7v DDRq

It detects them as

8
3
3
3T
2.6v DDRq

Those memory dumps wont open. MSVC 2005 keeps popping up but I cannot view the dump data.

BIOS settings
You also may want to set the BIOS to a safe setup like Auto memory timings, auto detect CPU speed, auto DDR or by SPD, 2.6 volts on DDR instead of 2.7.

Software then hardware - External first, then internal
First thing you do is start with software which it seems you have done. Next thing is to do hardware and use only necessary cards.

RAM seated fully?
I'm still thinking mobo here but ASUS does not normally have any issues like this. Perhaps the memory is seated just enough to boot, but not enough to be stable. Check all internal components to make sure they are seated in their slots.

Non-essential card removal
My recommendations are to take out all non-essential cards like LAN, sound, etc, etc. All you want is the video card. Boot and ignore Windows yelling at you about this device or that. If it does not BSOD during games or after games, one of the cards you took out is the problem.

Check for BSOD's
If it still BSOD's then you know it's something else. My guess would be between memory, faulty chipset, faulty chipset drivers, faulty CPU, faulty hard drive causing erroneous reads and writes, fault data cables leading to devices (these short out a lot). A bad read or write to disk could be caused by a short in the power cable leading to your hard drive. If the drive loses power for an instant, it will auto-reset. If you have read-ahead cache enabled, you will lose and/or corrupt data.

BSOD with only video installed - RAM/CPU/Mobo issue
If you only have a video card installed and it BSOD's and your board has on-board video, attempt to use it and remove your video card. If it still BSOD's I'd say you have a hardware issue either in RAM, Mobo/chipset, and/or CPU. Note that all of these are highly unlikely. In 20 years I've only had 1 hardware issue when my Voodoo 3 went out on me in my P3 600/ASUS board. I even cracked the top portion of my AMD 1.333 GHz and she still runs today as if nothing ever happened. I cracked it by installing a Thermaltake fan on it - damn those things install tight.

ASUS overclocks well
I've got an ASUS A78NXE-Deluxe with an AMD 3200+ XP and 1GB Corsair DDR 400. I've got the multiplier at 12.5 and the core at 200 and I have no problems. ASUS boards can handle a lot of speed and overclock extremely well as well as AMD CPUs. I'm just not sure why you are having all these issues.

It doesn't make sense. With an ASUS board you should be having these types of issues.

[Aran]

Thanks for the great reply, bubba!

Regarding the dumps: I didn't touch them. I uploaded them right out of the minidump folder without even trying to view them. Should I try re-uploading them?

How could the memory be seated improperly and still work perfectly fine for 4+ hours while memtest is performing rigorous tests on it?

I don't have anything in this computer, really.

Here is all the hardware i currently have installed:
ASUS A8N5X Socket 939 NVIDIA nForce4 ATX AMD Motherboard - Retail
Rosewill RP550S-2MK ATX 2.01 550W Modular Power Supply /Black - Retail
AMD Athlon 64 X2 4400+ Toledo 2000MHz HT Socket 939 Dual Core Processor Model ADA4400CDBOX - Retail
2x G.SKILL 1GB 184-Pin DDR SDRAM DDR 400 (PC 3200) Unbuffered System Memory Model F1-3200PHU1-1GBNS - Retail
NEC 16X DVD±R DVD Burner Beige IDE Model ND-3550A - OEM
VGA XFX 7900GT 256MB PV-T71G-UDE7 - Retail
HD 74GB|WD 10,000RPM 8MB WD740GD % - OEM

And a 150GB Maxtor drive that i use for media.

I'll do what you suggested tomorrow. Thanks again.

[Aran]

Here's my testing procedure from here on out.

For each hardware change I make, I will document it here and use the computer until it BSODs. If it does not BSOD for 48 hours, I will consider that hardware change to be the solution.

My first change: I swapped the SATA and power cables for my harddrives, entirely removing the second, Maxtor drive.

I'll report back if there's another BSOD.

[Aran]

Another BSOD at 12:41AM while playing Oblivion with my exhaust fan on max speed.

PFN_LIST_CORRUPT 0x0000004e (0x0000008f, 0x0000e2cd, 0x0000e2cf, 0x00000000)

The last several BSODs i've recieved have been similar to this one: PFN_LIST_CORRUPT 0x4e. MS says it's a RAM issue. I tried to download and use their memory testing app. I installed it using their installer onto a blank floppy, their installer claims the disk creation was successful, then i get a disk read error when I try to boot to the disk.

Current attempt: I reseated my RAM and swapped the position of the sticks (both are still in my machine).

My next step if this fails: Removing one of the sticks.

Here's hoping i just had to reseat my RAM to make these BSODs disappear...