Solved Intermittent ntoskrnl.exe+7c229 BSOD

[xandermac05]

memtest results

So, presumably I DO have a RAM issue (unless it's relating to the mobo or something). I take it there's no way of knowing which stick/slot is causing the issue until I run every stick in every damn slot, though, right?

Although... it seems to be in the same range every time. So might I gather that it's the second slot along? (They're 2GB sticks, so the error range is just outwith the range of stick 0/1.)

 

[usasma]

Follow the steps in this link: Memory Diagnostics
Start at step 7 and follow it to determine what's bad.

 

[xandermac05]

All individual sticks seem to check out OK in each slot, but...

1 stick: 0 errors
2 sticks: 0 errors
4 sticks: errors on test 6

Would this suggest a PSU problem? Or perhaps mobo failing to address 4 sticks at once?

 

[xandermac05]

The trouble is, this motherboard is garbage. I can't try fiddling with voltages or anything like that. So if it can't handle 2GB in 4 slots, I'm buggered.

The compatibility list claims the mobo fully supports my RAM, unless I'm making an incorrect assumption. The listing includes a Corsair "TW3X4G1333C9 G" set in either 2 or 4 slots, and I have two sets of TW3X4G1333C9A. As far as I understand, the A version just has a better heat spreader - but it operates at the same voltage as the TW3X4G1333C9, so surely that shouldn't be an issue? :huh:

 

[usasma]

Does the board have any BIOS updates available? Looks to me like there's 3 here (1002, 1101, and 1302): http://www.asus.com/Motherboards/Intel_Socket_1156/P7H55M/#download

I've got some sort of connection issue with the forums, so I can't check and see if you've updated your chipset drivers to the latest version available from http://downloadcenter.intel.com

PSU is possible. To check, remove as much stuff from the system as you can (so less power is consumed). If that improves things, then it may be the PSU.

Also, try this free test - it stress' the PSU (but be very, very careful!!!): http://www.ocbase.com/
Run the Power Supply test.

Good luck!

 

[xandermac05]

There are BIOS updates available, but they just fix issues I haven't been having and don't have reason to care about. Also, it's not actually that board, it's the SI version with the locked voltages etc.

I just got back to the flat to find that what I'd hoped would be the final memtest (all 4 sticks in) failed at passes 9 and 10. I've never seen it get anywhere near that far without a hitch then suddenly spit out errors. If it's able to come up with errors that late in the run then it's highly possible I okayed individual sticks which were actually bad (I ran 8 passes on each)... so I'm now running it with individual sticks again for more passes.

I'll check out the PSU if this doesn't come up with any results, methinks. Thanks for the info again.

 

[usasma]

There are undocumented "fixes" in some BIOS updates.
If the BIOS flash doesn't say it's needed, then I leave it until last (but it's still a possibility).
I've seen the undocumented fixes fix a few problems over the last few years.

 

[xandermac05]

OK, I'll give that a go as well then. For the time being, I'm running 20 passes on each stick, starting with the new ones (since the problem only appeared after I installed them).

 

[xandermac05]

Right. BIOS updated to latest. PSU apparently fine. Memtest86+ stopped spitting out errors. I've had more BSODs nonetheless.

I actually thought I might have sorted it when I discovered some dodgy config in foobar2000 pertaining to nonexistent plugins and an incorrect sound device selection (and had no BSODs for a long time afterwards). The config was copied from my old installation of Windows (on the same machine). Having said that, the whole reason I reinstalled Windows was to see if it'd fix the BSODs. I have been playing Mass Effect 2 on this system fairly non-stop for the past couple of days and it's never once crashed in-game; only when I'm using foobar2000 and/or browsing the web, as far as I recall. Just now, I only had the browser open when it died.

Other than running memtest for an extremely long period (perhaps with pairs of sticks in - old and new - to speed the diagnosis up somewhat) can you think of anything else at all here? Perhaps you'll spot something in the dumps; I've attached the last 5.

 

[usasma]

I'm thinking that this is looking more and more like a hardware problem.
STOP 0x109 are usually either driver problems or hardware problems - and the hardware problems are usually related to memory.

Despite still having a number of 2009 drivers in your memory dumps, please run Prime95 3 times:
- Blend test
- Small FFT's
- Large FFT's

Run until errors occur - or 24 hours, whichever comes first.
If you can figure out which one of the 3 tests errors out fastest (and slowest) that'll help out later on.

Also, run several of these CPU stress tests: Hardware Diagnostics

Further info on BSOD error messages at: http://www.carrona.org/bsodindx.html
Info on how to troubleshoot BSOD's (DRAFT): http://www.carrona.org/userbsod.html
How I do it: http://www.carrona.org/howidoit.html

The following info is just FYI, I've already addressed the issues that I saw in the above paragraphs
3RD PARTY DRIVERS PRESENT IN THE DUMP FILES

[font=lucida console]
MpFilter.sys Fri Mar 09 06:05:26 2012 (4F59E3F6)
amdxata.sys  Fri Mar 19 12:18:18 2010 (4BA3A3CA)
iaStor.sys   Mon Oct 17 18:54:08 2011 (4E9CB210)
HECIx64.sys  Thu Sep 17 15:54:16 2009 (4AB293E8)
ASACPI.sys   Wed Jul 15 23:31:29 2009 (4A5E9F11)
walvhid.sys  Wed Aug 26 01:15:08 2009 (4A94C4DC)
Rt64win7.sys Fri Jun 10 02:33:15 2011 (4DF1BAAB)
moufiltr.sys Sun Mar 08 23:16:13 2009 (49B489FD)
nx6000.sys   Wed Jan 27 18:36:09 2010 (4B60CDE9)
viahduaa.sys Wed Aug 04 03:51:53 2010 (4C591C19)
nvlddmkm.sys Wed Feb 29 13:04:52 2012 (4F4E68C4)
iaStorV.sys  Thu Jun 10 20:46:19 2010 (4C11875B)
HssDrv.sys   Thu Sep 02 13:46:48 2010 (4C7FE308)
taphss.sys   Tue Sep 15 15:58:23 2009 (4AAFF1DF)
iaStor.sys   Tue Jun 08 13:33:03 2010 (4C0E7ECF)
viahduaa.sys Tue Oct 20 23:30:30 2009 (4ADE8056)
[/font]

http://www.carrona.org/drivers/driver.php?id=MpFilter.sys
http://www.carrona.org/drivers/driver.php?id=amdxata.sys
http://www.carrona.org/drivers/driver.php?id=iaStor.sys
http://www.carrona.org/drivers/driver.php?id=HECIx64.sys
http://www.carrona.org/drivers/driver.php?id=ASACPI.sys
http://www.carrona.org/drivers/driver.php?id=walvhid.sys
http://www.carrona.org/drivers/driver.php?id=Rt64win7.sys
http://www.carrona.org/drivers/driver.php?id=moufiltr.sys
http://www.carrona.org/drivers/driver.php?id=nx6000.sys
http://www.carrona.org/drivers/driver.php?id=viahduaa.sys
http://www.carrona.org/drivers/driver.php?id=nvlddmkm.sys
http://www.carrona.org/drivers/driver.php?id=iaStorV.sys
http://www.carrona.org/drivers/driver.php?id=HssDrv.sys
http://www.carrona.org/drivers/driver.php?id=taphss.sys
http://www.carrona.org/drivers/driver.php?id=iaStor.sys
http://www.carrona.org/drivers/driver.php?id=viahduaa.sys

BSOD BUGCHECK SUMMARY

[font=lucida console]
Loading Dump File [C:\Users\FUBAR\_jcgriff2_\dbug\__Kernel__\050612-14882-01.dmp]
Built by: 7601.17790.amd64fre.win7sp1_gdr.120305-1505
Debug session time: Sun May  6 09:57:29.181 2012 (UTC - 4:00)
System Uptime: 0 days 0:25:55.306
*** WARNING: Unable to verify timestamp for win32k.sys
*** ERROR: Module load completed but symbols could not be loaded for win32k.sys
Probably caused by : memory_corruption
BUGCHECK_STR:  0x109
DEFAULT_BUCKET_ID:  CODE_CORRUPTION
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_MEMORY_CORRUPTION_ONE_BIT
Bugcheck code 00000109
Arguments a3a039d8`98be4afe b3b7465e`eb3b1a54 fffff800`03308790 00000000`00000001
BiosVersion = 1102   
BiosReleaseDate = 05/11/2011
CPUID:        "Intel(R) Core(TM) i3 CPU         540  @ 3.07GHz"
MaxSpeed:     3070
CurrentSpeed: 3066
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Loading Dump File [C:\Users\FUBAR\_jcgriff2_\dbug\__Kernel__\050412-13774-01.dmp]
Built by: 7601.17790.amd64fre.win7sp1_gdr.120305-1505
Debug session time: Fri May  4 10:31:21.672 2012 (UTC - 4:00)
System Uptime: 0 days 0:36:09.186
*** WARNING: Unable to verify timestamp for win32k.sys
*** ERROR: Module load completed but symbols could not be loaded for win32k.sys
Probably caused by : memory_corruption
BUGCHECK_STR:  0x109
DEFAULT_BUCKET_ID:  CODE_CORRUPTION
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_MEMORY_CORRUPTION_ONE_BIT
Bugcheck code 00000109
Arguments a3a039d8`99478687 b3b7465e`ebc455dd fffff800`02f08880 00000000`00000001
BiosVersion = 1102   
BiosReleaseDate = 05/11/2011
CPUID:        "Intel(R) Core(TM) i3 CPU         540  @ 3.07GHz"
MaxSpeed:     3070
CurrentSpeed: 3066
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Loading Dump File [C:\Users\FUBAR\_jcgriff2_\dbug\__Kernel__\050312-13634-01.dmp]
Built by: 7601.17790.amd64fre.win7sp1_gdr.120305-1505
Debug session time: Thu May  3 16:36:45.222 2012 (UTC - 4:00)
System Uptime: 0 days 1:36:08.346
Probably caused by : ntkrnlmp.exe ( nt! ?? ::FNODOBFM::`string'+45a35 )
BUGCHECK_STR:  0x1a_41287
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
PROCESS_NAME:  foobar2000.exe
FAILURE_BUCKET_ID:  X64_0x1a_41287_nt!_??_::FNODOBFM::_string_+45a35
Bugcheck code 0000001A
Arguments 00000000`00041287 00000000`00000000 00000000`00000000 00000000`00000000
BiosVersion = 1102   
BiosReleaseDate = 05/11/2011
CPUID:        "Intel(R) Core(TM) i3 CPU         540  @ 3.07GHz"
MaxSpeed:     3070
CurrentSpeed: 3066
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Loading Dump File [C:\Users\FUBAR\_jcgriff2_\dbug\__Kernel__\050112-15334-01.dmp]
Built by: 7601.17790.amd64fre.win7sp1_gdr.120305-1505
Debug session time: Tue May  1 07:53:26.375 2012 (UTC - 4:00)
System Uptime: 0 days 10:14:00.000
*** WARNING: Unable to verify timestamp for win32k.sys
*** ERROR: Module load completed but symbols could not be loaded for win32k.sys
Probably caused by : memory_corruption
BUGCHECK_STR:  0x109
DEFAULT_BUCKET_ID:  CODE_CORRUPTION
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_MEMORY_CORRUPTION_ONE_BIT
Bugcheck code 00000109
Arguments a3a039d8`9e2e67f2 b3b7465e`f0ac37f4 fffff800`0220099c 00000000`00000001
BiosVersion = 0905   
BiosReleaseDate = 07/15/2010
CPUID:        "Intel(R) Core(TM) i3 CPU         540  @ 3.07GHz"
MaxSpeed:     3070
CurrentSpeed: 3066
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Loading Dump File [C:\Users\FUBAR\_jcgriff2_\dbug\__Kernel__\042912-13384-01.dmp]
Built by: 7601.17790.amd64fre.win7sp1_gdr.120305-1505
Debug session time: Sun Apr 29 08:54:17.464 2012 (UTC - 4:00)
System Uptime: 0 days 0:17:25.339
Probably caused by : memory_corruption
DEFAULT_BUCKET_ID:  CODE_CORRUPTION
BUGCHECK_STR:  0x3B
PROCESS_NAME:  SearchProtocol
FAILURE_BUCKET_ID:  X64_MEMORY_CORRUPTION_ONE_BIT
Bugcheck code 0000003B
Arguments 00000000`c0000005 fffff880`015b2b9a fffff880`09c50760 00000000`00000000
BiosVersion = 0905   
BiosReleaseDate = 07/15/2010
CPUID:        "Intel(R) Core(TM) i3 CPU         540  @ 3.07GHz"
MaxSpeed:     3070
CurrentSpeed: 3066
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
  
  [/font]

 

[xandermac05]

OK, pretty certain it's an issue with the mobo right enough. BurnTest throws up an instability warning after a few runs on Maximum, advising I "lower my overclock". I'm taking that to mean that the voltage to the RAM needs boosting. They're running at stock settings, however, and are supposed to be able to do so with this mobo. 2 sticks? No problem.

This all looks correct to me, but maybe I'm missing something:

I assume, then, that I have no choice other than

1. Deal with being limited to 4GB RAM, or
2. Get a new mobo?

 

[usasma]

Time for a specialist. Post over at the mobo manufacturer's forums for any additional advice that they may have.
We'd appreciate it if you'd post back with any results from there so we can advise folks about the resolution.

 

[xandermac05]

One other suggestion: would running two 4GB sticks instead of four 2GB sticks be more likely to work? The 4GB sticks aren't listed on the compatibility list for my mobo, but would I be correct in assuming they'll work since the board allegedly supports up to 16GB and the 4GB sticks run at the same timings/voltage as the 2GBs? The product page insists they'll work in "all DDR3 motherboards with two memory channels", which covers mine...

 

[usasma]

I'd presume that they'd work OK - but I'm not a RAM specialist.
I strongly suggest that you get further advice from someone who knows the in's and out's of this motherboard.

 

[xandermac05]

I've tried contacting ASUS support, but they're unresponsive (I understand their customer support is notoriously awful, so I'm not hugely surprised by this). On the other hand, I could just order a 2x4GB set and return it if it doesn't work. It's Amazon, who are excellent on the returns/service front. I'll continue to try to get word from ASUS or a mobo-specific forum to find out if I can do anything to get 4 sticks running, but Googling about suggests that all the boards in the P7H55-M series have issues with 4 sticks (despite what the official specs might claim).

Thanks for all your help and advice; it's been a rather useful learning experience for me, above all else. I appreciate it.

 

[usasma]

If you can afford to purchase the RAM (and then return it, if necessary) - then I'd think that it would be a valid troubleshooting procedure (and worth a try).

I really have no idea of why this system isn't functioning for you - and don't want to lead you in the wrong direction.
FWIW - I'd suspect issues with the mobo and would replace it if money's not an issue (based on the burn test results).

FYI - I had problems with my Asus P6T Deluxe v1.0 board for a long time (mostly BSOD's when going into sleep).
I finally gave up and bought a new mobo (P6T Deluxe v2.0) and haven't had a problem since!

 

[xandermac05]

Yeah, it looks like the board to me. Money definitely is an issue, unfortunately; I do intend to replace the mobo, but I can't really afford to at the moment. I had actually bought one on eBay but the seller then informed me that he'd mucked up the listing and it was a cheap ASrock (not the Asus advertised), so I've cancelled that.

For the time being, I've returned the 2x2GB RAM and ordered a 2x4GB set to see if that works (with the refund from the 2x2GB I'm only spending another £13). If it doesn't, I'll return that as well and just wait 'til I can afford a mobo upgrade...

 

[xandermac05]

Just to let you know: the two 4GB sticks are stable even under max-stress BurnTest. Furthermore, because I was curious, I shoved my original two 2GB sticks in the other slots and tested again. Guess what? Stable again. Presumably v2.12 of the 2GB stick is dodgy in some configurations; the original set I bought over a year ago is v2.2 (no idea why the "new" set is actually outdated, but they're returned anyway).

I think I'm still going to see if I can grab a bargain on a decent board, 'cause it seems a waste to not overclock the i3-540 (seen evidence it can hit 4.5GHz even on the stock cooler - it's at 3.06 currently!). I've got my eye on a Gigabyte P55-UD5 at the moment...

Thanks once again for all your help!

 

[usasma]

Good luck!
Thanks for keeping us informed!