[Morientes]

Hi!

I was just playing Trackmania United and in the background Mirc was running. When Mirc suddenly said "You are disconnected" (it plays a sound file saying that), the computer crashed and I got a BSOD saying machine_check_exception.

Now I have had the computer for a few month and have not had this happen before, so my question is: can it be caused by software, a bug in trackmania or mirc or some collision of ressources when mirc played that sound, or do I have some kind of hardware problem?

I have tested my RAM for about 4 hours with no problems. I have 2gb and my processor is a core 2 duo 6600. The graphic card is Nvidia geforce 7900gto.

I hope it's just software related!

 

[OldGrayGary]

Hi Morientes2000

This is usually a hardware condition. Here's Microsoft's article for the error, and a troubleshooting checklist (at the top, notice that if you are overclocking - this can cause such an error - so, too, can incorrect Bios settings, insufficient cpu/system cooling, and inadequate/faulty power supply, or faulty memory . . . you've already checked for faulty memory, so you'll be looking at the others) --- http://support.microsoft.com/kb/329284

Make sure you have all the right chipset drivers for your motherboard installed, too - as having dicey drivers can upset timings, & cause errors that look hardware related - when they are actually caused by poor or incorrect drivers.

Also, look in on the logs in EventViewer & see if that yields clues:
To Explore EventViewer
1) Click on Start/Control Panel/Performance and Maintenance/Administrative Tools/EventViewer.
2) Several logs are listed: simply select a log, and events will appear in a list in the window-pane to the right.
3) To view the details of an event, right-click the event, and select “Properties”.

... and should a BSOD or two keep occurring, send along the full error message, and we'll take a look.

Best of luck
. . . Gary

 

[Morientes]

Thank you for the reply!

I will try to check this when/if a new error appears.

I have searched a little on the net and, it seems, if it is software caused it should be only drivers that can cause it and not software like games?

I am currently away from home for about a day and a half and I have left my computer running, testing the RAM for an even longer period.

But an error like that, shouldnt it come up quite often if it is as serious as an error in the hardware - I mean a direct error in the CPU or motherboard should cause an error at a lot more occasions?

Damn, I hope I don't have to change any hardware. It's a new machine and the first motherboard I got when ordering it, was faulty. I can't be so unlucky that another component is also faulty!!

 

[OldGrayGary]

Hi again

I'd actually be fairly surprised if it was a problem with the cpu, motherboard, video card, or hard drive. You've tested the memory already, so that's not likely it. I'd be interested to hear what make/model/wattage power supply you're using.

I'd also be surprised if a program managed to cause such an error. The error is a "low-level" error - the sort usually caused by either a hardware problem, or a driver problem [since drivers operate in a very priviledged, low-level part of the operating system - whereas regular programs run in a higher-level, interacting with the core of the operating system in a safer, more removed way].

Remember that the hardware itself might not be faulty, but if the settings aren't quite right, or the driver isn't quite right - this kind of error can happen. Definitely check through your Bios settings to make sure they are all OK (you'll need your motherboard manual handy for that, and for more information on the right Bios settings to use, see Adrian's famous Bios Optimization Guide --- http://www.techarp.com/freebog.aspx ). What motherboard do you have? Is it the same model as the faulty one that you had to return?

Tell you what, let's have you run a tool that checks in a bit on your drivers: it's called the "Driver Verification Tool". It looks complicated to run, but it doesn't have to be - you can run it on the default settings. http://support.microsoft.com/kb/244617 . . . Here's a note to read if you happen to run Norton antivirus on your computer "If Norton Antivirus is installed, do not enable Driver Verifier's Deadlock Detection because of the recommendations in Microsoft Knowledge Base article 325672." The options that Verifier runs are covered in the section of that first Verifier article link entitled "Driver Verifier Manager".

Best of luck
. . . Gary

 

[Morientes]

Thanks again for the exellent reply!

I am just about to go to bed but I will check the things you said as soon as I have got the time to do it! Then I will report back...

Just finished the long memory test that I mentioned eariler and it confirmed that there is no problem with the memory...

Just to provide some new information I can say that my PSU is a "NorthQ 4300GP-FL". It is a fanless model, so completely silent but of course generating more heat. I am pretty sure the system is cool enough though. Right now the CPU is 54 degrees after running for 1½ days. At full load on both cores it lies at about 60-63 or something. My motherboard sits at 38-40 degrees.

 

[OldGrayGary]

Hi again


Thanks for the info on the power supply. Can you give the make/model of the motherboard too? [& if a BSOD comes by, see if you can capture the details, or perhaps come across a detailed error message in EventViewer].

If you have another power supply of sufficient quality & wattage, I'd be interested to see if the trouble suddenly vanishes. NorthQ is not a common brand here in the States, though it seems to be fairly popular in Europe. I'm a little troubled by the two or three hardware-testing-website reviews that I found for NorthQ products: and even with just those three = on each supply, the 12v rail had trouble keeping a consistent strength. [Here's an example page --- http://www.hexus.net/content/item.php?item=6276&page=8 --- you'll notice that of the four premium supplies tested, the NorthQ supply had the worst effeciency and couldn't supply enough power on the 12v rail. Just that sort of problem can lead to the symptoms you are experiencing on your new build. It may or may not be the culprit: but it's a prime suspect, worth investigating.

Whenever I want unquestioned clean power, I prefer to drop in an Antec. At least to test with. They always deliver according to spec.

Here's a link to another hardware site that I belong to (along with my home here at Tech Support Forums) --- the following page is a list a recommended (and also of not-recommended) power supplies [NorthQ isn't listed at all - since it didn't have a strong market presence at the time the list was put together] --- http://forum.pcmech.com/showthread.php?t=131195

Your 'fan-less' NorthQ does have the temperature-controlled fan that kicks on in case of overheating, yes? (the model I looked at online did)

I'll be getting some rest myself soon -- as the sun's been down for a fair number of hours already here in California.
. . . Gary

 

[Morientes]

I dont have direct access to another power supply right now unfortunetly. But I can try to monitor the 12v on it.

It does have a fan in case of overheating.

My motherboard is an Asus P5N32-premium.

I just ran the driver verifier tool. It cant quite figure out what it is doing though. I ran it by typing verifier exe and using the standard settings and I said it should test all drivers. Then I rebooted and during boot I got a BSOD saying one of my drivers attempted to access memory not allocated to it...the code was 0x000000CD.

I rebooted and now I am inside windows, but everything seems to be running very slow. Is it because the verifier is running? For how long will it do that?


EDIT: I figured it out...just had to write verifier /reset in the command prompt. Then I had to reboot. When I did that I got another BSOD saying: "A device driver attempting to corrupt the system has been caught. The code was: 0x000000C4 (0x00000060, 0x00000070, 0x00000000, 0x00000001).

I guess these error mean that I have a driver not working properly. But how do I isolate it?

 

[OldGrayGary]

Hi again


Verifier has found an error, interestingly an error that can only occur when the "Pool Tracking" option is in use. Here's the MSDN tech note for 0x0C4 errors with the 0x00000060 2nd parameter:
"The driver is unloading without first freeing its pool allocations. A bug check with this parameter occurs only when the Pool Tracking option of Driver Verifier is active"

Check in your \Windows folder to see if there is a subfolder named \minidump -- there may be several minidumps in there, see if the same modules are indicated. The only error I've seen so far that's similiar to yours happened to be related to the Intel Active Monitor, or Intel hardware monitor, that the system was using. But that's just coincidental - check your minidump logs, and see if a particular module is specified.

I should have pointed to an easier to use method for running Verifier - using the graphical interface (rather than command line only) -- makes it easier to track errors. You can specify which driver to test (instead of all of them at once) -- you pick them from a list. Here's a link to instructions for using the graphical mode for Verifier in XP --- http://msdn.microsoft.com/library/d..._b55fe0e6-73a8-4d20-985d-14555e0048c8.xml.asp

EventViewer should have some details for you, too - that might point out which module is causing the trouble.

The graphical mode is also easier to use to Disable the Verifier when you are done with your troubleshooting (it has to be told to stop) - you use the "Delete Existing Settings" option = this clears the list of drivers being tested (so that Verifier doesn't load at boot time).

Best of luck
. . . Gary

[P.S. there's a known bug for systems running Norton Antivirus and Verifier - which can be solved by changing an option in Verifier -- but from the looks of your error messages, I don't think this will apply to your system]

 

[Morientes]

Once again thank you for your help, it is greatly appreciated!

I just checked the minidump folder. There are two files in there. I opened them with notepad but I must admit I can't make much out from them. It seems like a lot of code and then some word here and there. I could make out "Genuine Intel" one place in the first file. Also this line appears. It seems to mention some modules:

m i x e r . s y s « °"«  ÔðƇk m i x e r . s y s Ø« °Ú«  üðƇD M u s i c . s y s 0ð¬ ñ¬  $ñƇk m i x e r . s y s àm® �p®  LñƇd r m k a u d . s y s `èº pèº  tñƇD M u s i c . s y s �¯ `¯  œñƇa e c . s y s �p® Àr®  ÄñƇs w m i d i . s y s �¯ p¯  ìñƇs p l i t t e r . s y s @ݺ `ݺ  òƇP a r p o r t . S Y S ÐM® O®  <òƇk b d h i d . s y s 0>º p>º  dòƇk b d h i d . s y s ð=º 0>º  ŒòƇC d a u d i o . S Y S ¿º P¿º  ´òƇS f l o p p y . S Y S 0>º `>º

It mentions some that might point to the audio driver? But then again it also mentions floppy.sys...

The other minidump is quite long and mentions alot of things, but I cant really find a pattern in the things it mentions...

The event viewer only shows the error code of the system errors, it doesnt seem to mention what modules that are involved. I am looking under system and then doubble-clicking on the error, when clicking on the "For more information.." support center link it says that a matching error might appear with id 1002, but I can't really find it.

I noticed that in the event viewer there seems to be quite alot of errors concerning the disks. There is an error saying: "The device, \Device\Scsi\iteraid1, did not respond within the timeout period." and later a series of errors saying: "The device, \Device\Scsi\iteraid1, is not ready for access yet.", "The driver detected a controller error on \Device\Harddisk2\D". These messages repeat 3 times.
Finally there is also an error that says: "The device, \Device\Ide\IdePort0, did not respond within the timeout period".

I don't know if these errors are normal...

I have attached the two minidump files, maybe you can make something out of them that I can't.

 

[OldGrayGary]

Hi again

Look through the Verifier documentation on msdn in my last reply, and see if you can follow their tips on how to run Verifier to only test one driver at a time. That allows you to know ahead of time which driver isn't passing the strict tests that Verifier uses.

Is your system set up with a RAID array? If so, which kind? Are your drives parallel-ATA or serial ATA? RAID-0 or RAID-1? The timings on an advanced motherboard like yours can be tricky, and if either the timings themselves, the components involved, or the drivers running the show aren't just right = errors crop up. So - if you haven't done so already, go through the Bios settings with your Asus manual in hand (and perhaps Adrian's Bios Optimization Guide, too). Read over the RAID setup instructions & make sure everything checks out. The Setup in the Bios is just as important as having the right Windows XP drivers setup. If you aren't using a RAID setup, make sure that the settings are set for that, too.

Did you have any trouble while installing the motherboard drivers? (especially for the drive controllers)? Check on the ASUS website for any updates or patches for those, just in case.

Are you running a Hardware Monitoring program? (either from Asus or Intel)? If so, try disabling it, especially during the testiing. Also, if you are using the ASUS "AI" built into the motherboard, disable that as well (it's for over-clocking = you don't want any overclocking when troubleshooting - & truthfully, that system is plenty fast without any overclocking at all). I looked at a few reviews from overclockers - and the board seems very sensitive, a liable to BSODs, when overclocked even modestly.

I'll have to look at the minidumps a bit more - they aren't the easiest thing to have to make sense of. (And, luckily, I don't have to rely on them very often).

So, check on the settings, disable software hardware monitors, check on the drivers, and run the Veriifier so that it tests one driver at a time

Keep at it
. . . Gary

 

[Morientes]

Ok, thanks, I will try the things you suggested tomorrow...

On the motherboard there is only room for one ata plug and I use that for my two CD-ROM drives. My primary harddisk (from which I boot Windows and Linux) is a sata disk so it is plugged directly into one of the sata-plugs on the motherboard. Other that that I have three old ata disks that are plugged in via a RAID-card. They DO NOT run a RAID configuration though. They are just connected via the plugs on that card...

But as I said I will report back tommorow when I have tried your other tips!

 

[Morientes]

Ok, I have now isolated two faulty drivers by using verifier.

They are: aegisp.sys (causes a DRIVER_IRQ_NOT_LESS_THAN_OR_EQUAL stop error as soon as windows is loaded - at the login screen) and sjypkt.sys (causes a C4 stop-error when rebooting (or logging out of) windows).

I would, of course, like to remove these drivers, but how do I do that? Do I just delete the two .sys files?

Also do you think any of those could be the culprit in my original MACHINE_CHECK_EXCEPTION?

PS: I looked in my BIOS and I can see the "AI tuning" option is set to "Auto". Do you know if that means that it will actually overclock sometimes? Because in that case I should set it to "Standard" I think...

 

[OldGrayGary]

Hi again

The aegisp.sys file seems to be a part of an 802.1x wireless protocol, and the sjypkt.sys seems to also be a part of networking, related to NDIS 5.0 Protocol Driver (NDIS=Network Driver Interface Specification) --- also possibly part of a wireless driver. I'm not sure if your motherboard has built-in wireless (I didn't notice it in the specs). If you do have a wireless device installed in your system, you could try updating the drivers for that. -- If the wireless is built onto the motherboard, you might be able to grab a newer driver from Asus - or try the website of the wireless device that Asus placed on the board. If you installed an add-on wireless device (PCI, Firewire, or USB) - you'd visit the manufacturer's website for newer drivers.

You also might want to point your antimalware tools at those files, to make sure they scan clean. They should be in the \Windows\System32\drivers subfolder.

In fact, you should thoroughly scan for malware - and include a rootkit scan as well, as that first file has known to have been compromised by some attacks.

I believe "Auto" for the AI does indeed automatically overclock. There are so many overclocking features on that board, it can make you dizzy. (Just look at the brief summary of them in the opening pages of the manual! During troubleshooting, try to get back to stock timings and voltages). If you aren't sure what the "Standard" setting does, you can likely call or email Asus [it's been a few months since I've had an Asus board in for repair - so I can't remember all the overclocking options].

I still wonder if trying a different power supply might help, too.

Any more BSODs lately? Any modules listed other than the wireless related modules? . . . And - are the disk errors continuing to have the three-errors-in-a-row pattern in the EventViewer?

Best of luck
. . . Gary

 

[Morientes]

I have not had any BSODs since the MACHINE_CHECK_EXCEPTION.

As far as I can see there has been no disk error either, I don't know why they were there before...

Hm I will try to set the board to "Standard". Its a little difficult finding out what the "Auto" setting really does, I think...

My motherboard does have build in wireless so I have installed the forceware network drivers now. The two .sys files are still there though so I doubt anything is changed from that. Suppose I could run the verifier again and see if it does the same thing...

I have scanned the system using "Windows Defender" and it says that it is clean.

Dont really know what my next step might be. But as I said I guess I could try the verifier again! But I dont think the drivers are changed. They also still say "Unknown" under Manufacturer.

 

[Morientes]

I just set the AI tuning to Standard. To check if anyone knows for sure what the two options do, I just created this thread: http://www.techsupportforum.com/f15/asus-p5n32-ai-tuning-options-135578.html#post762477

 

[Morientes]

hehe, ups. It seems I have two logins...well, it was me who wrote that message just before!

 

[OldGrayGary]

Good Morning

Well - you might be OK after all. It's good to know you made sure that no malware was lurking about. Good to hear, too, that the disk errors are no longer showing up.

Might be time to try a round of your game again! -- In fact, try several games (not all at the same time, of course). And if you get no errors at all - Shout HooRay. And if you only get errors when playing one particular game -- go to that game's website, and check for either patches, or workarounds for errors such as you see while playing the game.

Regarding the AI options, I wish I could remember how I set things on that last ASUS board -- It might have been Standard, but whatever it was, I went through the settings one-by-one, setting the timings to what I knew was 'standard' for the device involved (bus speeds, cpu timings, memory timings -- I tend to know the stock speeds from spec sheets). It was a very fine board I was working on -- as is yours = those top ASUS boards ride first class.

You're just about ready for a technician's certificate
. . . Gary

 

[Morientes]

Good evening!

Well, I just checked with the verifier and it still crashes with these two drivers. I would like to try to just remove them. But how do I do that?
Isn't there some place where you could uninstall the .sys files? Or are they part of something?

Thanks again for the help! I think this migh be my last question - luckly it is not a serious error and it has actually only occured this one time.
I have learned a lot from this and I am pretty convinced it's not an error with my hardware which was what I was fearing initially.

Removing these drivers that I now KNOW might cause trouble will give me some peace of mind though!

 

[OldGrayGary]

Hi again


The good thing about those two files is that they so far have only crashed in Verifier, which can be pretty strict. I'm not certain of their origin, the first one seems to be a third-party file, while the second might be a Microsoft file. You could right-click on them (they should be in the \Windows\System32\drivers subfolder) and select "Properties" to see who created them, and if they are "signed" drivers. They likely are installed either as part of XP's wireless support, or are installed when drivers for a wireless device are installed.

Deleting them might cause trouble, replacing them with Verifier-friendly versions would be nicest = but it's also possible that they are OK, and that it's a little overkill on Verifier's part. Time will tell. There doesn't seem to be a lot of information on just those two files.

One thing you could do, in case they are files that might have recent new versions = is to visit Windows Update, and see if there are any patches available there. If you are behind on any Critical Updates - install those right away, as they often help close vulnerabilities in XP, and can make your system a lot more secure.

Before updating a driver, it's a good idea to create a "restore point" in System Restore, so that if things go badly, you can return without much fuss.
Start/All Programs/Accessories/System Tools/System Restore - "create a restore point".

If we're lucky, that will be your only crash of 2007
(I like to think positive!)
. . . Gary

 

[Morientes]

I am up-to-date with Windows updates...

The Aegisp.sys comes from something called "Meetinghouse Data Communications". The sjypkt.sys is from "Windows(R) 2000 DDK Provider".

Hm, when you say they only crash from verifier does that mean that you do not suspect any of them to be involved in the MACHINE_CHECK_EXCEPTION?

Would be nice if this is the only crash of 2007!