Random BSODs and applications crashes on a fresh Win7 install

[R4ndom]

These are the results from last night, before changing the 'VDDNB Over Voltage' setting, as I ran the test before you replied. Will repeat it with new voltage settings as well.

View attachment 196848

There's a strange thing however, unless I understand something totally wrong: on the screenshot above, the CAS parameters shouldn't be 9-9-9-20 or am I mixing two different things? CPU-Z reports 9-9-9-20 - this is what I set in BIOS.

Also, when I try to get info on SPD during memtest, it tells me "SMbus controller unknown".

Is there any point in changing "chipset voltage"?

Thanks for help!

 

[Dave76]

Have you had any crashes in the last week?

If you haven't had any more crashes, then leave it as it is now.
If you are still having crashes then bump the 'VDDNB Over Voltage' to 1.250v.

Always post any new crash minidumps when you get them, they may have additional information that will help us.

Let us know your status.

 

[R4ndom]

Have you had any crashes in the last week?

Hi, I didn't have any crashes since I set the mems first to 8-8-8-24 @ 1.65v (which was roughly a week ago) and then after your suggestions to 9-9-9-20 @ 1.60v. No blue screens since then. However, 2 or 3 Firefox crashes (v.10 just released, maybe it was actually Mozilla's programme issue). [/quote]

If you haven't had any more crashes, then leave it as it is now.
If you are still having crashes then bump the 'VDDNB Over Voltage' to 1.250v.

I already bumped it to this value after your previous post. I ran memtest again and did 7 passes without any errors. Funny thing though, memtest would pass every time even when I was getting blue screens - i.e. sometime 2-3 weeks ago.

Another thing: several people report that memtest is showing incorrect timings values in their case - can anything be done about that? Isn't it that it's actually testing RAMs with wrong settings? CPU-Z reports correct values as I set them in BIOS and says also that I'm running dual-channel while memtest says 64bit.

I think, but I'm not 100% sure, that before timings in memtest were what was actually set in BIOS, then recently it would report wrong ones... Does it matter anyway?

Always post any new crash minidumps when you get them, they may have additional information that will help us.

Let us know your status.

Will do if I get BSODs again, hopefully not.

You mentioned that it's possible to get mems running @ 1600 with AMD - could you shed some more light on it? If everything stays fine as it is now, I may try some time to make them run as OCZ claims they should.

Thanks again for help!

 

[Dave76]

Good that your system is stable now.
Firefox has been known to have crashes with their latest updates.

Ram can still cause crashes even if passing memtest86+ (best RAM testing program available) and the RAM can fail a test after passing several times.
Just the way electronics are.

If your RAM timings are on 'Auto' in BIOS they can show different numbers in memtest86+.
You can manually set the first four RAM timings in BIOS, leave the rest in 'Auto', this should let memtest86+ run the test at the right timings.

I'll look for the AMD 1600MHz info and post back.

Since your system is running stable for the last week, you can mark this thread as 'Solved'.
Let us know if you have any questions.

 

[R4ndom]

Good that your system is stable now.
Firefox has been known to have crashes with their latest updates.

Ram can still cause crashes even if passing memtest86+ (best RAM testing program available) and the RAM can fail a test after passing several times.
Just the way electronics are.

Yes, it's been stable with both 8-8-8-24 @ 1.65 and 9-9-9-20 @ 1.60. Firefox seems to have become more stable with the 10. update.

If your RAM timings are on 'Auto' in BIOS they can show different numbers in memtest86+. You can manually set the first four RAM timings in BIOS, leave the rest in 'Auto', this should let memtest86+ run the test at the right timings.

That's what I did, as to get 8-8-8-24 I need to enter manual settings, so no auto settings for me - timings and voltage set manually, still memtest shows wrong timings info. When set to 9-9-9-20 (still manually, as on auto I get 9-9-9-24) also shows as 4-5-5-14. Strange.

The official memtest page states:

Memtest86 supports all types of memory. If fact the test has no knowledge of the memory or CPU type nor does it need to. Memtest86 is able to test all types and brands of memory. If the information reported regarding chipsets or CPU's is incorrect it will not affect testing in any way. This not a problem or bug but is listed here due to the many questions about this issue.

So maybe it's no issue, but still it's annoying and make me wonder if the test is reliable. Which memtest do you use, as I see there different versions around? The one I have now is this one

I'll look for the AMD 1600MHz info and post back.

Thanks!

Since your system is running stable for the last week, you can mark this thread as 'Solved'.
Let us know if you have any questions.

Will surely do.

 

[Dave76]

I use the same one, memtest86+ v4.20.
Haven't had that issue or seen it reported here.

AMD IMC and RAM frequency at or above 1600MHz

You can try posting in the Seven Forums Overclocking forum, several of the members with AMD CPUs have been doing this successfully.

Good luck with your 1600MHz RAM settings.

 

[R4ndom]

Thanks for help and a very useful link Dave, it explains a lot, also that the performance boost won't be very noticeable, but still worth a try, be it for fun.

I'll continue anyway testing with Memtest+ and tweaking RAM settings for 1600MHz, hopefully will get the right timings back in memtest.

 

[Dave76]

You're welcome.

Enjoy your tweaking, keep a log of your settings, pass and fail, will be a good reference.

 

[R4ndom]

I haven't overclocked, yet another BSOD occured when PC was on idle. PC had been stable for more than a week with timings suggested by you i.e. 9-9-9-20 @ 1.60v.

Event viewer recorded a crash of the comctl32.dll module, but WhoCrashed reported a severe memory management problem and windows explorer crashed as well several times in a row (I got this info from $msinfo32.nfo included in the griff zip), I had to reboot to get it work. Also problems with firefox, skype, windows management console. Again, this happens, as before, when the PC is not under load, but performing lightweigh tasks...

Feels like RAMs need some further tweaking (or eventually buying new ones), but don't know where to start. Should I post in another sub-forum for that?

I cannot upload the zip with dumps since yesterday, I keep getting the message "Upload of file failed."

 

[Dave76]

We have been having server issues, should be fixed now.
Try to upload again.

Do you have any other RAM you could borrow to test with?
This would be the easy way to rule out the RAM.

 

[R4ndom]

We have been having server issues, should be fixed now.
Try to upload again.

Do you have any other RAM you could borrow to test with?
This would be the easy way to rule out the RAM.

I attach dump files with the 2 latest crashes (+ all the others) which occured while I was doing absolutely nothing with my comp.

Unfortunately I have no other RAMs to swap with these at the moment.

Every time it's different modules that are reported as causing BSODs. This time WhoCrashed reports shareduserdata.sys...

Thanks.

 

[Dave76]

Bump the 'VDDNB Over Voltage' to 1.350v.

D/L and run Seagate SeaTools – Short Drive Self Test and Long Drive Self Test. Should work on your Samsung HDD.

What is 'i view 32'?

Can't read any of your event logs, take a look at them and tell me what the 'Errors' and 'Warning' events are.

This crash is STOP 0x0000001A: MEMORY_MANAGEMENT
Usual causes: Device driver, memory, kernel

*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

MEMORY_MANAGEMENT (1a)
    # Any other values for parameter 1 must be individually examined.
Arguments:
Arg1: 0000000000003452, The subtype of the bugcheck.
Arg2: 000007fefdedf000
Arg3: fffff70001084f98
Arg4: 9620000097426404

Debugging Details:
------------------


BUGCHECK_STR:  0x1a_3452

CUSTOMER_CRASH_COUNT:  1

DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT

PROCESS_NAME:  taskhost.exe

CURRENT_IRQL:  0

LAST_CONTROL_TRANSFER:  from fffff80002ef25f3 to fffff80002e83c80

STACK_TEXT:  
fffff880`0a777058 fffff800`02ef25f3 : 00000000`0000001a 00000000`00003452 000007fe`fdedf000 fffff700`01084f98 : nt!KeBugCheckEx
fffff880`0a777060 fffff800`02e54942 : fffffa80`04261160 fffffa80`00000000 fffff8a0`00000156 00000000`00000000 : nt! ?? ::FNODOBFM::`string'+0x356b3
fffff880`0a777910 fffff800`0315881a : fffff8a0`0bab5060 00000000`00000001 00000000`00000000 fffffa80`040f7060 : nt!MmCleanProcessAddressSpace+0x96
fffff880`0a777960 fffff800`0313a0d8 : 00000000`00000000 00000000`00000001 000007ff`fffde000 00000000`00000000 : nt!PspExitThread+0x56a
fffff880`0a777a60 fffff800`02e82f13 : fffffa80`04261160 fffff880`00000000 fffffa80`040f7060 00000000`00000001 : nt!NtTerminateProcess+0x138
fffff880`0a777ae0 00000000`774415da : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiSystemServiceCopyEnd+0x13
00000000`0024fc48 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x774415da


STACK_COMMAND:  kb

FOLLOWUP_IP: 
nt! ?? ::FNODOBFM::`string'+356b3
fffff800`02ef25f3 cc              int     3

SYMBOL_STACK_INDEX:  1

SYMBOL_NAME:  nt! ?? ::FNODOBFM::`string'+356b3

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: nt

IMAGE_NAME:  ntkrnlmp.exe

DEBUG_FLR_IMAGE_TIMESTAMP:  4ecdc9f2

FAILURE_BUCKET_ID:  X64_0x1a_3452_nt!_??_::FNODOBFM::_string_+356b3

BUCKET_ID:  X64_0x1a_3452_nt!_??_::FNODOBFM::_string_+356b3

Followup: MachineOwner
---------

This driver is a bit concerning.

tmp9720.tmp  Sat Jul 26 21:29:37 2008

Can't find any reference to it.

 

[R4ndom]

i view 32 is irfan view picture viewer app

I will bump the voltage and run the seagate tools and report back

what do you mean by "I can't read any of your event logs" - the file is corrupted or the difficulty is in the language?

as for the driver, something similar (same date, but different file name) was found on page 3 of this thread here and my reply below that in fact i never had LoL installed on my pc and that the file itself was nowhere to be found.

one thing for sure, this is my last amd-based rig

 

[Dave76]

Uninstall irfan viewer for testing, it's been crashing.

Language issue.

That driver is showing as loaded in the debugger, so it is there.
One possibility is that it's from a virus.

What Anti-virus are you using now?
If still Avira, uninstall it and run the Avira RegistryCleaner.
Download Avira RegistryCleaner

Then install MSE
Microsoft Security Essentials*- Free Antivirus for Windows
Update it through Windows Update.
Run a full scan.

Download and run Malwarebytes,
Update the virus definitions before running a full scan.

 

[R4ndom]

Thanks for replying Dave.

D/L and run Seagate SeaTools – Short Drive Self Test and Long Drive Self Test. Should work on your Samsung HDD.

The drive passed all tests with this tool and reported no errors.

Uninstall irfan viewer for testing, it's been crashing.

Uninstalled it, but if I continue, I would need to uninstall virtually everything on my PC, including windows explorer since it's been crashing as well together with loads of other native windows processes.

Language issue.

Should I translate what is in the $msinfo32.nfo file under windows error reporting contained in the griff zip?

That driver is showing as loaded in the debugger, so it is there. One possibility is that it's from a virus.

I cannot find any reference to that either.

What Anti-virus are you using now?
If still Avira, uninstall it and run the Avira RegistryCleaner.
Download Avira RegistryCleaner

I stopped using Avira a few weeks ago after I was advised to do so

Then install MSE
Microsoft Security Essentials*- Free Antivirus for Windows
Update it through Windows Update.
Run a full scan.

Done, with latest updates, nothing was detected.

Download and run Malwarebytes,
Update the virus definitions before running a full scan.

Also done and also nothing caught.

I bumped the voltage as suggested. Seems like there's no pattern whatsoever for crashes. Which is true however, BSODs have become much more rare than before, it's mainly apps that crash randomly. Sometimes it's all good during 3-4 days, then all starts again without me making any changes.

Will need to get some other RAM to test, but spending 50-60EUR just to test remains quite expensive. I'd like some RAM please @ US and not EU prices

Another funny thing: OCZ states on their website that this RAM is dedicated for AMD 6.. and 7... chipsets...

 

[Dave76]

Your previous CPUZ Mainboard tab is showing BIOS version 0802 and the Asus mobo site's latest BIOS version is 0801.
Where did you get the BIOS 0802 version?

That is good the crashes are becoming less, hope we are removing the drivers and apps that are causing or contributing to the crashes.
You may have to uninstall more apps until we find the cause. After your system is stable you can get new versions of the apps and install them a couple at a time, check for stability. Add or remove as needed.

Debugging is not a fine art, sometimes you can find the cause quickly and sometimes, well, it takes a while.

This could be an intermittent hardware fault or slight compatibility issue.
Usually RAM is the offending culprit, further down the list are graphics cards (or drivers), PSU and motherboard.

I have seen different RAM solve these kind of tough cases before.


Let us know if you get any more crashes.

 

[writhziden]

Another method for determining applications causing crashes is to http://www.sevenforums.com/tutorial...ation-conflicts-performing-clean-startup.html.

 

[R4ndom]

Your previous CPUZ Mainboard tab is showing BIOS version 0802 and the Asus mobo site's latest BIOS version is 0801.
Where did you get the BIOS 0802 version?

Hi, the version you're referring to is for another MB, the M5A78L-M-USB3 ASUSTeK Computer Inc. - Motherboards- ASUS M5A78L-M/USB3, while mine is M5A78L/USB3 ASUSTeK Computer Inc. - Motherboards- ASUS M5A78L/USB3

That is good the crashes are becoming less, hope we are removing the drivers and apps that are causing or contributing to the crashes.
You may have to uninstall more apps until we find the cause. After your system is stable you can get new versions of the apps and install them a couple at a time, check for stability. Add or remove as needed.

Debugging is not a fine art, sometimes you can find the cause quickly and sometimes, well, it takes a while.

This could be an intermittent hardware fault or slight compatibility issue.
Usually RAM is the offending culprit, further down the list are graphics cards (or drivers), PSU and motherboard.

I have seen different RAM solve these kind of tough cases before.

Let us know if you get any more crashes.

To cut the story shorter (at least that's what I hoped, but read on) I got new RAMs: Corsair XMS3 - 8GB 1600MHz PC3-12800 DDR3 - 240 pin - 1.65V - CL9 - Kit: CMX8GX3M2A1600C9 - Komplett.be

At first the computer would not boot up, then it would after I reset the BIOS. All good so far, but once I launched memtest, it froze after a couple of minutes, so I rebooted and re-run it, it spit out errors and after 1hour or so, froze again.

BIOS sets the RAMs to 9-9-9-24 right away, which are the specification's speeds.

Symptoms are the same though, application crashes and a couple of BSODs as well. But again, nothing crashes under load, but on default settings and in low-resources consuming situations.

Starting to think it might not be a RAM issue after all? Unless this CPU/MB combination is so extremely picky on RAMs. Corsair guarantees full compatibility with all aMD and Intel systems and the sticks are on ASUS' qualified vendors list - although they list 2x2GB and not 4x4, I couldn't get the 2x2 anymore.

I cannot upload any dumps now, will do later. Any ideas will be more that appreaciated.

 

[Dave76]

What is the RAM frequency?

If not 1333MHz then manually set it to that.
Manually set the RAM to manufacturers specs, 9-9-9-24 and Command Rate of 2T.
DRAM voltage to 1.65v manually.
Set the 'VDDNB Over Voltage' to 1.250v.

If that's not stable, we'll bump the CPU voltage a little.
Can you upload the CPUZ CPU tab under full load?
Go to 'CPU Over Voltage' and tell us the options there.
Start with a +0.10v.

 

[R4ndom]

Hi Dave,

Didn't have time to reply these days, and to test.

So, the current situation is as follows: with the Corsair RAMs the PC won't even boot up - sometimes I can't even get to BIOS! And once I succeed, after lowering the speed to 1333 and setting the rest as suggested, Windows won't start, it just stays blocked on the "windows loading" screen and nothing happens.
I should stress that I reset the BIOS before putting new mems.

As I had stated before as well, the Corsair mems show errors in Memtest after a couple of minutes, and after another quarter of an hour, memtest freezes.

So I stuck back the OCZ, the PC boots up, but I'm still getting constant app crashes, but only a couple of blue screens.

So in the end it seems I'll have to RMA the corsair mems, but what puzzles me is that the problem actually could be with the MB and not RAMs - what do you think?
Damaged DIMM slots, short circuit on the MB, what else?

Thanks.

PS Will upload recent dump files and once I have some time, I'll translate the errors in the event log as precisely as I can.

I also noticed that when launching call of duty, every time it says "some parameters of your PC have changed since last usage, would you like to auto-adjust the settings?" although I don't change anything at all!

Starcraft suggests from time to time file corruption due to graphic problems, but I tested the GPU for several hours with a stress test, which showed no errors.

Seems like RAM settings are inconsistent and change randomly without me doing anything whatsoever.

MB fault? BIOS fault? RAM fault, I ran out of ideas.