Persistent ntkrnlmp.exe BSOD Win 7

Status
Not open for further replies.

gsaines

Posts: 8   +0
So I have literally been battling blue screens for weeks now. I have rebuilt my entire computer, ditched that, bought an OEM machine, and the problem still persists. Really the only thing has remained the same through my now 3 Windows reinstalls is the software I'm using. As such, I'm pretty convinced this is a software problem, but because the minidumps aren't pointing to consistent and specific driver .exes, I'm not sure where I should start pruning my software. In the past week it's blue screen 7 times, I've analyzed all of the dumps with Windows Debugger, and they all essentially say the same thing. I can't upload the minidumps because they are each 268kb, and the upload limit is 200kb, so I stitched them all together in Notepadd++ and uploaded the stripped down file instead. Also contained in that file is the everst report and hijack this report.

As per the FAQ in this thread, I read up on the error code on aumha.org/a/stop.htm], but didn't learn much that would be helpful. I also checked out osronline.com/article.cfm?article=335 to learn a little more about the error code, but again, it was a dead end (rather, it became too technical, and I wasn't able to see how determining whether a buffer underrun or overrun could help me fix the problem). What I have noticed is that I have a lot of general system instability. Firefox, photoshop, thunderbird, and chrome crash several times a day seemingly unprovoked. The error messages I get are generally cryptic and don't point to any obvious problem (I may just not be good at troubleshooting specific program crashes though).

I ran Memtest86+ from the bootable iso and it says my memory is A-okay, after more than a hour of testing, I've checked the internal temperature of the system (very well within reasonable limits), it's new so there's no dust in there to speak of, and I've had all of the HDDs tested at a computer shop here in town.

I haven't found a repeatable way to provoke the bluescreens, it seems to be fairly random. I'm always able to reboot without incident.

Yesterday my programs were crashing on load, and some were simply saying that .dlls were missing or corrupt. I restarted and it seems to have (at least temporarily) fixed the problem.

If I can't find a solution, I'll probably just reformat and reinstall XP again because it appears to like my hardware/software setup much better. Just in case it's helpful, here are my system specs:


Intel Core i3 2.93Ghz

4 GB RAM DDR3 1333mhz

64 bit Windows 7 Home Premium

500 GB SataII WD HD

Intel H55 Mobo chipset

ATI Radeon HD4350

This is a clean (and legit) install of Windows 7 on a newly formatted partition, so there shouldn't be any dual booting weirdness.

Any help you could lend would be very much appreciated, as I said, I've been wrestling with this for ages, and it feels like I'm getting nowhere.

-George
 

Attachments

  • Minidumps.txt
    47.1 KB · Views: 3
Severall of your errors are 0x50: PAGE_FAULT_IN_NONPAGED_AREA
Requested data was not in memory. An invalid system memory address was referenced. Defective memory (including main memory, L2 RAM cache, video RAM) or incompatible software (including remote control and antivirus software) might cause this Stop message, as may other hardware problems (e.g., incorrect SCSI termination or a flawed PCI card).

And every one cites memory corruption as the problem. First, Memtest needs to run for more than an hour. The rule is a minimum of 7 Passes and the more the better. It takes a LONG time but worth it. People often mistake that the individual 8 tests are just repeating themselves; they're not. They are instead throwing at the memory a vast array of differing patterns.

You should also check your BIOS and see if the motherboard set your memory voltage to the manufacture's specs. Many times what appears to be corrupted memory isn't; the real issue is wrong voltage settings for the RAM.

You should also check to see if your exact make and model of RAM is compatible with your motherboard.

* By the way, you can attach those larger files up to five in one zip file. The zip will compress them.
 
Thanks for checking this out Route44, and sorry that I overlooked the 7zip route, I hadn't thought that over properly. I actually signed up for Dropbox and got everything linked up before I realized that I couldn't add links to my posts yet. :)

At your suggestion, I left Memtest86 running over night, it completed 8 passes in 8.5 hours, and everything still checked out. It's not terribly expensive RAM, but at least according to Memtest86, things are working fine on the memory side.

Given that it passed that test, should I still go through the process of checking the voltages and RAM/Mobo compatibility?
 
That can be a good place to look at; also what is the make/model etc. of your H55 motherboard.
 
Hey Archean, it's a Gigabyte H55M-UD2H. It's about 2.5 weeks old and (I made sure of this) it's explicitly Windows 7 compatible.
 
Given that it passed that test, should I still go through the process of checking the voltages and RAM/Mobo compatibility?

Yes. There have been a significant number of memory corruption reads over the past two years but came to find out it wasn't the memory itself but the motherboard setting the voltages wrong.

Can you link us to your memory?
 
Okay, I opened up the case again and the RAM consists of two chinsey-looking (read as: doesn't have heat spreaders) Adata Premier Series DDR3 1333 240pin Unbuffered-DIMM CAS 9 Non-ECC modules. I did a quick google inquiry into compatibility between Gigabyte GA-H55M-UD2H and Adata RAM, but I couldn't find any data about the kind of memory that the board plays nice with.

According to the Adata product page (adata-group.com/EN/product_show.php?ProductNo=AD31333U), the correct voltage should be 1.5v. I checked the mobo bios and the voltage has been set to 1.5v, which appears to be correct. As a precaution, I noticed on the bio memory power settings page that the memory was running in what the bios deamed "turbo" mode, which apparently is intended for overclocking. I've done my share of overclocking and what that said to me was: the memory is set to "unstable." So, I switched the setting to "Standard." However, due to the long cycles between blue screens, I don't know if that fixed anything.

I do have 2 sets of older DDR2 800 Corsair XMS modules lying around that I could try swapping in. They run at a different voltage than the current Adata set, and the two pairs aren't exactly the same (the CAS values on the two sets of memory are different, which I've heard isn't good for stability), but I could try just putting in the 2x 2GB sticks and running that for a while. I'm not sure if that would help or not.

As always, thanks a lot for your help guys, really appreciate it.
 
Good work by you. :grinthumb I would say let's see if stability continues since you set the memory at standard. If you get a Blue Screen then in your next post attach the minidump file.

A-Data certainly has stepped up their quality with their DDR3 offerings.

Keep us up-to-date; definitely will be interested in seeing if stability has been attained.
 
Well, that spree of optimism was short-lived. It bluescreened at 4:30 AM this morning, I'm attaching this most recent minidump. At least the failure types are remaining the same, this is still a PAGE_FAULT_IN_NONPAGED_AREA (50) error coming from VISTA_DRIVER_DEFAULT. A few minor details are different, but nothing very substantive. Do you think it's fair to conclude at this point that it's not a memory problem?
 

Attachments

  • Most Recent Minidump.7z
    17.7 KB · Views: 18
0x50 error and it specifically cited memory corruption as your issue. I went back a reread your previous psot and for whatever reason I missed this tibbit of very important infomation: the two pairs aren't exactly the same (the CAS values on the two sets of memory are different, which I've heard isn't good for stability)

Mismatched memory will cause instability.

Did this latest error come from the Adata or Corsair?
 
No, the current memory is matched, it's the old Corsair XMS modules that aren't matched. The current problem is getting thrown by the Adata modules, the Corsair memory I mentioned was just a backup in case I wanted to try swapping out the Adata sticks.
 
I looked your motherboard's specs up and considering it is designed for DDR 3 memory thus your Corsair DDR2 wouldn't work in any case.

The ADATA certainly is correct as 1333 but is it compatible with this Gigabyte board? That is the question. Gigabyte provides a qualified Vendors list for their boards. Go to their site, find your board and their recommended list of memory for said board.
 
Okay, I called Cyberpower (the folks who put my machine together) to get the P/N for the RAM modules, since I didn't get the proper documentation in the package that arrived, and they said that the modules had been chosen based on motherboard specs and match the list of qualified vendors. I don't have any other way to get the P/N numbers, so I'm sort of forced to accept their conclusion. I did a quick run through the Adata P/N chart for the modules that seemed to match mine, and some of them were listed on the Gigabyte qualified vendors list, some weren't.

Cyberpower recommended I run a check disk, which I had somehow overlooked doing, it only took 30 minutes, so I ran that, and it corrected some system files, but no bad sectors were reported. To make sure, I then got the diagnostic software made for this specific drive and did a quick test, which also completed successfully.

I've started to notice a trend with the blue screens, which is sort of encouraging in a dismal sort of way. They either happen when I am shutting down my computer (I hit "shutdown," the PC starts turning off, blue screens, and restarts, when windows reloads it tells me it has recovered from an unexpected shutdown) or it happens in the wee hours of the morning (typically around 4-5AM). What would happen around 4-5 AM consistently is beyond me, similarly, why it's pulling a BSOD on shut down is also beyond me.

When I was researching this issue prior to posting, I noticed that there were several forum threads about BSODs referencing ntkrnlmp.exe on shutdowns, but most of those seemed to have pretty easy resolutions (Macafee, a faulty logitech peripheral driver, etc).
 
Alright, I've run out of time on this computer. My refund window closes in 2 days, and I still haven't made any headway on this one, so I'm going to return this computer and recommission my old computer and it's trusty XP installation.

Before deciding to return it, I ran the driver verifier and managed to catch a BSOD, but the minidump was no more specific than the dumps without the driver verifier running. Whatever is plaguing this machine is something unholy difficult to figure out. If I was independently wealthy and had a lot of free time I'd try to fix this, but I just need the machine to work, so I'm abandoning ship on this one. Thanks a lot for your help. I really appreciate it, I'm only sorry there wasn't a happy end to this debugging tale!
 
I'm sorry too. :( Wish we could have found out and I think you are wise in shipping this back. If you find out the what the issue is please let us know. Thanks.
 
Status
Not open for further replies.
Back