XP DRIVER_IRQL_NOT_LESS_OR_EQUAL BSOD shortly after boot

Status
Not open for further replies.
Shorly (about two minutes) after booting, my Windows XP SP2 system blue screens. Usually it is a DRIVER_IRQL_NOT_LESS_OR_EQUAL (d1) error. Occasionally it does not blue screen. If it doesn't BSOD shortly after booting, it will not BSOD until the next boot. It doesn't matter whether I log in or not.

I don't believe it's a hardware problem (power, memory, overheating) because the BSOD always happens at the same time. In addition, I'm using ECC RAM, so I doubt it's a memory problem.

Adding and removing hardware seems to have little effect on the BSOD.

I have followed the instructions for removing the page files, defragmenting, and recreating the page files.

I could not attach the minidumps I wanted to (too large), so instead I have uploaded them here: http://scarey.fastmail.fm/unlisted/rehevkor5dumps.zip Most of the minidumps are attributed to ntoskrnl (KeContextToKframes), but I did not include all of them. Instead, I included a wide variety of minidumps which have all happened in the same way.

Any help is greatly appreciated! If anyone wants more info, I'd be happy to provide it!
 
I just had a similar problem, about which I posted here. I'm no guru (in training my profile says), but I thought you might find my experience helpful. I also did not believe it was a RAM problem; it had passed extensive runs of Memtest86 with no issues and the BSoD error messages (and minidump files) seemed to indicate a video card driver, or my AV software, or something else. It wasn't until I tested each module of RAM in isolation that the faulty RAM finally revealed itself.

If you have two sticks of RAM, try running with just one, then the other. If you have just the one, see if you can borrow another one. I did do a lot of research reading other threads on this forum and using Google, and the majority of info I read indicated faulty RAM. It's worth making completely sure.

It might also be worth checking your PSU. Speedfan will give you details of your voltage fluctuations, and there are calculators that you can use to figure out if your PSU is powerful enough for your system. Here is one calculator:

http://www.extreme.outervision.com/psucalculator.jsp

Hopefully someone more adept at reading minidump files will come along soon and help you out.

Good luck!
 
Thanks a bunch for your input! I guess reinstalling my floppy drive to run Memtest86 is worth it. I'll take a look at your thread too.
 
I had a look at some of your dumps, they show crashes at "ntkrnlmp.exe", a windows kernal file, and memory corruption. You should use Memtest to check your ram.

I don't believe it's a hardware problem (power, memory, overheating) because the BSOD always happens at the same time.
You should still check your system temps, and whether your PSU is supplying enough power.
 
Well, I ran memtest and memtest+

I let memtest run while I was at work, and it went through 48 successful passes with no errors. ECC was off.

When I got home, I ran memtest+ through one successful pass with ECC off, then turned ECC on and ran 16 successful passes with no errors.

This was with only two RAM sticks installed (instead of the usual four), but I continued to get consistent BSOD when I booted. I'm not surprised by these results because memory errors shouldn't predictably BSOD about two minutes after boot about 75% of the time.

However, after unsuccessfully trying to run a couple of apps included with UBCD, a BIOS error came up saying "Uncorrectable memory error previously occurred in 000EF000h" This happened a couple times after rebooting from frozen DOS apps, but it's never happened before. My guess is that it's something whacky that the DOS apps are doing, and not the fault of my RAM.

Right now I'm preparing to reformat my computer (not going to reboot for a while!), but I'd still like to determine whether this is a hardware problem. Any input (heat and power are both fine)?
 
Looks like filesystem corruption to me. run cmd, then chkdsk C: /x and see what errors come up. if it wont run with the /x switch just run it with C:
 
Status
Not open for further replies.
Back