New Hardware, Clean install

[CarlTR6]

Let's assume the driver is OK for the moment. It could be that the ESET driver was causing the interference. Third part antivirus programs and firewalls can cause the types of crashes by interfering with the graphics. Rename the ESET driver and reboot. Let's see what happens.

 

[Trinitrotoluene]

I have renamed eamonm.sys to .bak, screenshot. Ran furmark untill the temperature was leveled out, screenshot.

Then I turned on driver verifier and got 3 BSOD's. They didn't get dumped to the harddrive, so I took pictures of them (below). Is that normal? I hope the error codes are useful. Nevertheless I will still attach the normal .zip file here.

Also, I can't seem to acces this site, and some others, while driver verifier is active (I am in safemode right now just for this post). So I will turn it off now to see if that is really the case. EDIT: driver verifier off, and I can acces this site again. Perhaps that tells you something as well.

http://img818.imageshack.us/img818/1843/20101027101406.jpg
http://img716.imageshack.us/img716/9059/20101026235909.jpg
http://img100.imageshack.us/img100/4984/20101027103513.jpg


Also, do you require a perfmon /report everytime with the .zip file? Or is the Windows_NT6_BSOD_v3.03_jcgriff2_ app from now on enough?

 

[CarlTR6]

Good job on following through. Unfortunately, we need the Driver Verifier dumps. Hopefully the they point to the offending driver(s). Make sure your system is still set up for mini dumps. Check the last part of the first post in this tutorial: http://www.sevenforums.com/tutorials/83172-dump-files-upload-seven-forums.html. The error codes are right for for a Driver Verifier enabled dump, C4.

It is not necessary to run Perfmon every time unless you make a hardware change or a software change.

 

[Trinitrotoluene]

Unfortunely it doesn't continue to write the dump file to the hard drive upon crash with driver verifier enabled. I changed the write debugging information to "small memory dump (256kb)". Now I've also turned on automatically restart, but I doubt that would make a difference. Any suggestions? Or has this gone from bad to worse now :-(

 

[CarlTR6]

I do not know why it is not writing the dumps. Turn Driver Verifier off, run for a while and then turn it back on again.

 

[Trinitrotoluene]

Alright, will do and see where that brings me.

If necessary though; feeling as if I'm running out of options here, I decided to run memtest again a few moments ago. It turned up with some errors. So I rebooted the system, ran it again. And again errors, so I'm assuming they're hard errors.

The modules are new, so is the motherboard of course - but it's the second set of modules. Chances of them being broken AGAIN? I don't know. Remember I did a 16 hour run before, no errors. And now there are errors, so perhaps they turned faulty in the meanwhile. Though somehow, in this case, I find it more plausible that the motherboard is messed up in a way. But all this is still new for me.

Will do some testing in re-seating the modules, swapping them around. Perhaps another in-dept analysis of testing single modules, like before. I'll also go over to the OCZ forum again. On which I'll show this situation with bios screenshots of the memory timings and voltages and see if there's still a wrong setting. Perhaps it's even bend CPU pins on the motherboard - I don't know. I do know that I took GREAT care in installing everything.

I guess this is both good and bad news, as we're still getting closer to a solution. I ran out of luck a long long time ago, or so it seems...

Faulty motherboard memory slots, faulty modules again - it all IS in line with the BSOD's I'm getting. Especially considering I have already removed all the possible software that can mess up and have updated drivers.


Will keep you posted, please don't hold back any thoughts you still have on this issue. Thanks again!

 

[CarlTR6]

If Memetest found errors, you obviously have a RAM problem. Follow through with that. There is no need to work with Driver Verifier unless you have further BSOD's after you get your RAM problem ironed out. Do keep us updated, please. And good luck!

 

[Trinitrotoluene]

Hi again Carl,

I have been running prime95 blend the past 2 nights, without crashes. Also normal usage in windows and games without crashes. So memory modules, software side of things and CPU are OK.

Long story short, I checked all the memory slots on the motherboards for any faults. All connectors looked perfect to me, even used a magnifying glass. I did find a little piece of material in slot #1, which I removed. Not sure if that may have caused these problems. But since this type of hardware is quite delicate, it may just have messed things up - I sure hope so!

I'll keep the system running normally, probably turn it off tonight and then stress it again the next day from cold boot. It's looking better then before so lets hope it stays this way. Pretty weird problem though! The way it crashed and showed errors doesn't seem to be in line with the very small piece of material I found in the slot. But then again, you never know, it's still a computer. However, it might explain why reseating the modules did seem to help resolve errors in memtest. But if you add the first set of actually broken modules to this whole scenario, it makes you second guess everything.

It's probably safe to assume the software side of things is OK. Though I will not reinstall daemon tools after what I heard from you. And I might just keep avast as well - if it ain't broke, don't fix it. So that's where I'm at right now. I feel that it's just a bit too soon to officially call it, so lets see what happens for a bit longer.


May I ask about your experience with Hamachi on windows 7 (probably version 2.0?). Is it known to cause trouble like daemon tools and nod32 might do?

 

[CarlTR6]

I am sure glad to hear you have had a couple of days and nights without crashes. That is good news! A little piece of dust or debris in a RAM slot can most certainly cause a problem. It has happened before - many times. I, too, hope removing that debris is the solution. Good find and good work. It is a good idea to run Memtest from a cold boot. Let us know the results.

Regarding Logmein, I recall one thread where we suggested the OP uninstall it. What I do not remember is why - were the drivers obsolete or did a dump blame Logmein? If your system is stable, you can try it. If you start getting BSOD's, you can uninstall it. I have not had much experience with it other than that one BSOD thread I mentioned. You can see if it is on the Win 7 compatibility list:

Windows 7 Compatibility: Software Programs & Hardware Devices: Find Updates, Drivers, & Downloads

Wise call not to reinstall Daemon or Alcohol. You can use MagicDisc instead. It does not use the sptd.sys driver.

Freeware MagicISO Virtual CD/DVD-ROM(MagicDisc) Overview

We hope to hear a good report in a couple of days that you have no more crashes!

 

[Trinitrotoluene]

Everything has been fine since I removed that piece of debris from the slot. Though, this morning upon cold boot I did get errors in memtest. But NO problems whatsoever during normal usage, and that included once again blend prime95.

Will do a prime95 blend run tomorrow from cold boot and see what that brings. This cold boot thing seems to be a known phenomenon, but not necessarily a problem. If normal usage is all fine, I don't care about what memtest says during cold boot. So, looking good once again so far.

And it seems hamachi 2 for W7 should be just fine!

 

[CarlTR6]

Actually, the cold boot run of Memtest is significant. Sometimes that is when the RAM errors show up. I would venture to say that you do have a RAM problem. It may not be one that shows up unless the RAM is really stressed.

 

[Trinitrotoluene]

Yesterday when I shut down the PC I got a BSOD again, NTFS_FILE_SYSTEM. It might be related to what you said.
I haven't had any other problems yet, not even when running prime blend from cold boot, so far. How much more can one stress the RAM besides prime95 and games and applications. Memtest is great, but if nothing happens during real-world usage, I'm just reluctant to once again send the modules RMA or even take out the whole motherboard for RMA

edit: the BSOD might also be related to defraggler 2.00.221 beta. If I remember correctly, it was the first defrag I did after recent format and it was with that piece of software. So I'll keep my eyes on that.

edit2: just had a memory_management BSOD again. So there's still something wrong with the memory, like you said. Also firefox keeps crashing so often and W7 aero sometimes gets turned off and then on again. Which I think is worth mentioning, sounds familiar? Don't really know where to go from here, so I've also asked OCZ.
I'll try all available QPI/vtt voltages in between 1.200v and 1.360v and see what that does in memtest from coldboot. It's pretty much the last memory setting I can try fiddling around with. Besides trying 1.64 RAM voltage instead of 1.66v (1.65v unselectable). After that I guess it's either memory modules RMA (again) or motherboard RMA.

 

[Trinitrotoluene]

Hello again Carl. Still in the process of sorting this one out, though it's still looking like I have faulty memory. Still getting errors in memtest upon coldboot, today when I got into windows after memtesting and getting coldboot errors, my firefox would crash and later on windows with memory_management. So that one is pretty obvious.

Now I've just had a bad_pool_header BSOD. To be precise, my windows aero effect disabled out of nothing, I decided to restart the system, upon logging off etc. it would BSOD.
Usual cause device driver, but can it also be caused by bad RAM? We've already looked into device drivers so much. In case you need it, I have attached the dump.
Should I turn on SMART in my bios for my RAID0 drives (I have been told it would have a negative effect in case of RAID(0)) and do a check up of some sort just to make sure? Though I don't think SMART will be able to read out data in RAID0.

Update: Current memory kit has been shipped RMA to OCZ, they're in the process of testing them I suppose. I still think the only problem here is/was the memory, but you never know. Will update this once RMA has been finished. Thanks again.

 

[CarlTR6]

I apologize for the delay in getting back with you. I overlooked this thread. Please let us know what they say about the RAM.

Error code 19, BAD_POOL_HEADER - Usual causes: Device driver.

Error code 1A, MEMORY_MANAGEMENT. Usual causes: Device driver, memory, kernel.

You can see from the above that device driver is common to both error codes. We need to take this right down the line and eliminate any possible BSOD cause.. First uninstall Avast. Avast generally runs well with Win 7; but is known to cause BSOD's and some Win 7 systems. You can try it again after you get your system stable. You have MSE installed; so you are protected.

Next you should update or eliminate these obsolete drivers.

hamachi.sys Thu Feb 19 05:36:41 2009 - LogMeIn Hamachi Virtual Network interface driverhttps://secure.logmein.com/US/support/hamachi2/. Update this driver. If there is not a later driver. uninstall LogMeIn.

SSPORT.sys Thu Aug 11 19:07:32 2005 - Samsung printer driver Samsung. This is an XP driver. I do not believe there is an update. If you cannot update it, navigate to C:\Windows\System32\Drivers, locate this file and rename it from .sys to .bak. Reboot and this driver will not load. This will break the Samsung utility program; but your printer should still work.

Unistall Avast and take care of the above drivers. Let's see how your system runs. Please update us and let us know. Feel free to PM me if I miss your update.

 

[Trinitrotoluene]

Got a new kit of OCZ memory today, the one I RMA'd was also broken. So that's kit #2 in the bin! Now on my third, hopefully it's OK now.

When I first ran memtest after inserting all three sticks and correcting bios settings, memtest would give me 1018 errors at pass 0 test 6. All at 5000.0MB, as far I could see, because I couldn't (didn't) scroll. But for now I'll assume that was the only type of error.
I reseated all modules, then ran one pass without errors. Hopefully it was down to a module not being seated right. Will run prime 95 blend for a solid amount of hours and some coldboot memtest in a couple of days when I have time. I might not be around for a few days, but I think I'll be able to keep control with remote desktop.
I had one strange reboot, not even a BSOD, when I started prime blend test. I already started a couple of short ones before. I did get a report when it got into windows again, about CCC MOM not responding. But I'm not sure how I should interpret that report. As being the cause or as a notice of the app not responding due to the crash that was caused by something else.

Problem signature:
Problem Event Name: APPCRASH
Application Name: MOM.exe
Application Version: 2.0.0.0
Application Timestamp: 49ef8e68
Fault Module Name: ntdll.dll
Fault Module Version: 6.1.7600.16385
Fault Module Timestamp: 4a5be02b
Exception Code: c0000005
Exception Offset: 00000000000159ad
OS Version: 6.1.7600.2.0.0.256.1
Locale ID: 1033


Furthermore I have renamed ssport.sys and for now I will leave the PC running with avast still installed. Hamachi was on Windows 7 supported software list so it should also be OK. I'm just interested in seeing what happens with it still being on the system, if you don't mind. I have this feeling that faulty RAM caused all this. But on the other hand, I don't know what to trust anymore.
The second I get another BSOD with usual cause being driver related or anything unexplained, it will be uninstalled.

 

[CarlTR6]

MOM.exe is -

Catalyst Control Center: Monitoring program belongs to Catalyst Control Centre or Catalyst Control Centerhttp://www.google.com/search?q="Catalyst Control Center -"- or http://www.google.com/search?q=" MOM.exe"MOM.exe or Sony Ericsson Device Data or TurboTax 2008 WinPerProgramHelp by ATI Technologies Inc (www.ati.com) or Advanced Micro Devices Inc.

CCC can be a real PIA on some Win 7 systems. You can install just the video drivers and uninstall CCC.

I agree that Hamachi and Avast should be OK. When troubleshooting, we have point out any potential BSOD cause. I think your problem was faulty RAM. And then CCC caused and app crash. You have done very well in troubleshooting and following suggestions.

 

[Trinitrotoluene]

Thanks Carl, I think and hope you are right about it just being a RAM problem. Hate to get more trouble going on after returning two RAM kits already.

I have now also deleted CCC, ran driversweeper, installed the whole package again except for CCC and the WoW trial of course. I take it audio HDMI stuff and the other Windows 7 support features are OK.
Hamachi is now also permanently deleted, hamachi.sys is gone and so is ssport.sys (double checked). Avast is also deleted for the time being.

This morning I ran memtest from coldboot, the first 2 tries getting into memtest failed: the memtest screen would stay blank (blue), no info. Third time worked though, and I did a 2.5 hour 3 pass run without errors. Not sure if that was actually a cold boot anymore by that time though. And I had that blank screen before on the other kit as well... So I don't know... Is this something significant? We might happen to run into something motherboard related here, considering this is my 3rd memory kit. Now, I do assume the RMA department checked the previous 2 kits and also found them corrupt. But you never know...
Also, some days ago when I wasn't at home and wanted to check back on the system with teamviewer, my PC just froze a couple of times. No errors or BSOD, just a freeze. I think prime95 was on, but I don't remember seeing any errors in the workers (didn't check to be honest...). Will try to replicate this.


Will use the system normally and keep doing coldboot checks as well. I'm sorry for this wall of text. I'll keep you posted.


p.s. when logging in and out of windows, the screen fades, you know? When it does, the picture on my screen vibrates, it gets off-center to the left I suppose. Don't think I had that before. Having samsungs monitor identification software installed or not doesn't make a difference (currently uninstalled again, whats the use please?).

 

[CarlTR6]

Try running Prime95 to test both your CPU and your RAM. Follow the instructions in this tutorial: http://www.sevenforums.com/tutorials/100352-cpu-stress-test-prime95.html. Run 3 separate tests, one on each of the settings (Blend, Small FFTs, Large FFTs).

Try this free CPU stress test: |MG| CPU Stability Test 6.0 Download

 

[Trinitrotoluene]

Will do!

So far no BSOD, but I haven't tested for so long either (don't mean prime). I did get a couple of freezes, starting some days ago (that is on the new ram) Happened again just a minute ago. I noticed that during this freeze, my USB mouse and USB keyboard lights turned off as well. A clue perhaps? My PSU is brand new, Corsair HX850. But the previous two ram kits were also new, we all know how well those ran. The freezes I've had so far occurred during prime, gaming and idle windows.

I locked my CPU vCore at 1.23125v, if I remember correctly, the "normal" value. I did this because my motherboard suffers from coil whine when it varies voltages for power saving goals. The sound used to come from my old PSU. Now that I RMA'd that one, it comes from the CPU socket. Quite annoying, but locking the voltage fixed it. It is a known fault in the X58A UD3R motherboard and there's a fix for it. But that would mean another RMA and I'm getting sick and tired of that.
Maybe it (voltage locking in combination with other stuff) is the cause of the freezes, C1E and EIST are both on I suppose, didn't change those. Could try to replicate the freezes with vCore on auto of course. Or maybe it's still about trying to get the optimal QPI/VTT voltage (currently 1.34v). Pffff, back to the drawing board I guess...

Your thoughts please

Update
Blend test: Test completed, 22 tests in 54 minutes - 1 error, 0 warnings (worker #7) other workers were still running OK. I'm happy to see at least an error instead of a freeze. Though interesting that I didn't get errors before out of all the prime's I already ran.
And from the results log file:

FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.

Now running small FFT's, 8 threads...

 

[CarlTR6]

Since Prime has found an error and your motherboard has a known fault, I recommend RMAing it. I would do so if it were my computer.