BSOD

[Dutchmv]

Hey there,

About a month ago I build a new PC and since that time I have a lot of BSOD problems.
My OS is Win 7 pro x64. I got it from MSDN.

At first I thought it was an issue with conflicting programs so I threw away avg and bittorrent. This helped a lot but the bsod is still there, although not as frequent.

As far as I know all the drivers are up to date.
motherboard drivers:
GA-770TA-UD3 (rev. 1.0) - GIGABYTE
and graphics drivers:
ATI Radeon
I got the latest beta bios installed in the hope that would solve the problem, but no such luck.

I reïnstalled about 8 times now, but that doesn't solve anything anymore.

I ran memtest86+ for about a day. there were no errors.
I run speedfan from tome to time, it shows good temps. When the CPU runs at 100% the highest temperature that I've seen was about 42-44 degrees

Now most bsod come when I play starcraft 2 online, but it sometimes happens as well when I watch a movie on youtube.

And I don't know if it is related but a few windows installements back I had a lot of problems with corrupting files. I checked them on another pc and they weren't corrupt at all.

I really don't enough about pc's to solve this problem, so any help will be greatly appriciated.

 

[Capt.Jack Sparrow]

Hello and Welcome !

Seems like all the crashes are pointing to Direct X Driver which could be caused by Display card. I would recommend to run certain test http://www.sevenforums.com/748030-post2.html make sure you run all the tests.

Run a Hardware Diagnostic (Memory and Hard Drive) follow this link for instructions Hardware Diagnostic !! « Captain Debugger

Bugcheck:

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

Use !analyze -v to get detailed debugging information.

BugCheck 3B, {c0000005, fffff880042fcf0f, fffff88007c6ae70, 0}

Probably caused by : dxgmms1.sys ( dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7 )

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

1: kd> !analyze -v
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

SYSTEM_SERVICE_EXCEPTION (3b)
An exception happened while executing a system service routine.
Arguments:
Arg1: 00000000c0000005, Exception code that caused the bugcheck
Arg2: fffff880042fcf0f, Address of the exception record for the exception that caused the bugcheck
Arg3: fffff88007c6ae70, Address of the context record for the exception that caused the bugcheck
Arg4: 0000000000000000, zero.

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


EXCEPTION_CODE: (NTSTATUS) 0xc0000005 - The instruction at "0x%08lx" referenced memory at "0x%08lx". The memory could not be "%s".

FAULTING_IP: 
dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7
fffff880`042fcf0f 488b01          mov     rax,qword ptr [rcx]

CONTEXT:  fffff88007c6ae70 -- (.cxr 0xfffff88007c6ae70)
rax=fffffa800443a960 rbx=fffff8a008fc4000 rcx=dffff8a003a814b0
rdx=fffff8a003b422e0 rsi=0000000000000000 rdi=fffff8a003a81540
rip=fffff880042fcf0f rsp=fffff88007c6b848 rbp=fffff8a003a81540
 r8=fffffa80048ea820  r9=fffff8a003a81540 r10=fffff8a00151d701
r11=fffff8a0039a1000 r12=0000000000000000 r13=0000000000000001
r14=0000000000000000 r15=ffffffffc000000d
iopl=0         nv up ei pl nz na pe nc
cs=0010  ss=0018  ds=002b  es=002b  fs=0053  gs=002b             efl=00010202
dxgmms1!VidMmQueryAllocationSizeInSystemMemory+0x7:
fffff880`042fcf0f 488b01          mov     rax,qword ptr [rcx] ds:002b:dffff8a0`03a814b0=????????????????
Resetting default scope

CUSTOMER_CRASH_COUNT:  1

DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT

BUGCHECK_STR:  0x3B

PROCESS_NAME:  SC2.exe

CURRENT_IRQL:  0

LAST_CONTROL_TRANSFER:  from 0000000000000000 to fffff880042fcf0f

STACK_TEXT:  
fffff880`07c6b848 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : dxgmms1!VidMmQueryAllocationSizeInSystemMemory+0x7


FOLLOWUP_IP: 
dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7
fffff880`042fcf0f 488b01          mov     rax,qword ptr [rcx]

SYMBOL_STACK_INDEX:  0

SYMBOL_NAME:  dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7

FOLLOWUP_NAME:  MachineOwner

MODULE_NAME: dxgmms1

IMAGE_NAME:  dxgmms1.sys

DEBUG_FLR_IMAGE_TIMESTAMP:  4a5bc578

STACK_COMMAND:  .cxr 0xfffff88007c6ae70 ; kb

FAILURE_BUCKET_ID:  X64_0x3B_dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7

BUCKET_ID:  X64_0x3B_dxgmms1!VidMmQueryAllocationSizeInSystemMemory+7

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

Hope this helps,
Captain

 

[Dutchmv]

Thank you for replying so fast, I really appreciate it.

There was a new update for the graphics card available today, I don't know If this makes any difference, but anyway. Since these tests all have to run for a very long time, I'll update this post when I get some results, but for now I ran Prime.

Prime results:
First the blend test, after ten minutes core five had an error. temp of the cpu was 49 degrees max.
After that I ran the small ffts test. I ran it for 23 test and had no errors before I stopped it myself. Max cpu temp was 51 degrees.

The memtest I ran for about a day with no errors. But I did this about two weeks ago.

I had no time to run the hdd test yet.
The furmark site had maintenance so I couldn't dowload the program at this time.

I will post the other results as soon as possible, and again thanks for the help.

 

[Capt.Jack Sparrow]

Your Most Welcome ! You said you got some error while running Prime95 do you remember what was it ?

 

[Dutchmv]

hey,

the prime error was: FATAL ERROR: Rounding was 0.4985351563, expected less than 0.4
Hardware failure detected, consult stress.txt file.

I ran memtest today again for 8 hours, there were no errors.

Update:
I ran furmark, everything went fine. Max temp was 62 degrees and there were no problems.
I ran chkdsk as well but there were no problems either.
I tried to run the hdd from samsung but the HUTIL tool doesn't work.

 

[Wishmaster]

Are you or were you overclocking?

The fatal error means there was a error in the calculations.
Or in other words, the system is not stable.

Unstable system can and will cause corruption, BSODS etc. Quite simply, the CPU is making mistakes.
This could be CPU or RAM related, but seeing as MemTest is passing, it may just be a CPU issue..


I would let Prime go for 10Hrs. as a Blend test. But as always, keep in mind that just because Prime passes 10hrs, only means the system seems stable.
And 99.9% of the time, should be.

However, on occasions, some minor issues may pop up.

If its failing within a few minutes, this isn't good.


If the Rounding error comes up again, it may likely be related to the CPU itself.



If you are Overclocking, you may want to start backing down a little at a time and find a stable point.
Once you find that stable point, to OC any higher will require additional Vcore.

Also keep in mind, if Memtest passes, it means the RAM itself is good.
You can still get memory errors under Prime, related to NB for example.
And although its very rare, there are cases when MEMtest passes, and Prime fails and the culprit is a bad RAM module.

If this is such a case for you, testing with 1 RAM module at a time will help isolate the issue.

 

[Dutchmv]

hey,

Thanks for the reply.

As to overclocking, I left the system in its optimized defaults and never clocked anything over. I will test the ram one by one now to see if anything comes up but I have a question.
When you say "If the Rounding error comes up again, it may likely be related to the CPU itself." do you mean that the cpu is just broken and I need a new one, or that it can be fixed with a software update or something?

update:
I just ran prime and within seconds five of the six cores crashed wit these errors.

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

 

[Wishmaster]

In all honesty, I was assuming you were overclocking.

The fact its crashing so quickly is a stability issue. it almost certainly points to a piece of Bad hardware.
Being as you are not Overclocking, it makes it a bit tougher to pinpoint.
The problem Could be the CPU itself, Motherboard, or a RAM Module.

But, before I get ahead of myself, lets try this.

---> Your RAM is DDR3 1333 (9-9-9-24) Correct?
If Im looking at the correct ones here, you need to be sure its set at 1.5V @ 9-9-9-24 and 1333 Speeds or less.

---> Can you find out what version of bios you have?
It appears the minimum you need for the 6core CPU are the F3 bios (5/13/2010)
If your bios are F1 or F2 we need to update them.

Double check your Memory settings.
Make sure that the Voltage, Timings, Speed are set the way they should be.
(on rare occasions it can be set inappropriately)

Also, while your in the bios, double check the rest of your settings. Voltage, and speeds of CPU etc.


With optimized default settings, it should be all OK, but its certainly worth double checking.

In fact, if you want to go ahead and reset the bios to 'Optimized Defaults" again, it certainly would not hurt anything.


If all of the above is OK:

1) Pull ALL of your RAM out of the system.

2) Insert just 1 stick, in the first MOBO slot.

3) Test.
If it fails or passes, remove it it and set it aside.
(Just so you know which stick fails and which pass)

4) Put in the next Module and Test again.

Do this for each RAM module and let us know the results please.


If by chance they both fail, try again ..
Only this time, move to the next RAm slot on the Motherboard.


The point here is to see if any of the RAM modules will test as good by themselves.

 

[Dave76]

Excellent suggestions Wishmaster.

After running the above tests, D/L CPUZ and post the CPU, Mainboard, Memory and SPD tabs.

In the start menu search box, type 'Snipping Tool' then enter, select screen area you want to post, save to convenient place, add to your next post using this method, How to Post a Screenshot in Seven Forums.

This will give Wishmaster and others information on the details of your system.

Check the CPUZ memory tab if you are running in unganged mode.

If after the above tests and suggestions, your still having crashes, you may need to set your RAM to 1066MHz, as some AMD AM3 CPUs have integrated memory controller issues at those speeds.

 

[chev65]

According to the newegg reviews on this board it is required that you must update the bios F3 in order to run a 6 core processor. There are several complaints to that effect. Good call on that Wishmaster.

The problem is that, according to the complaints at Newegg, you can't flash the bios while the 6 core processor is installed. So this may be a consideration before the bios flash is attempted. If that isn't the problem then others mentioned it was required to adjust some of the bios voltage settings to get the board stable.

Quote below:
Cons: Board has one of the ugliest/oddest layouts. It is more of a molded mATX/ATX board and it is a rather pain to fit into some cases. It may state that it has 140W CPU support, but not after it has been flashed with a revised BIOS. It won't flash regardless with a X6, so if you plan on running an X6 with this board, (as I did on a budget), you need an X2 or X4 to flash it before inserting an X6 onto the board. The dual-bios isn't needed and on-board video wouldn't hurt.

http://www.newegg.com/Product/Produ...k=False&VendorMark=&IsFeedbackTab=true&Page=4

 

[Dutchmv]

Thanks for all the replies.

My bios version is F4B. In the beginning it was F1, I changed it with the help of another cpu to F3 a while back so it would recognize the AMD phenom. In the hope it would solve the bsod issue I changed it to F4B. But it didn't help.
update: I updated the bios with an AMD sempron the first time.

I just ran the first memtest with one stick in slot 1. It ran for 5:10 with 11 passes. There were three errors in pass 0,3,8. All with test 7.

I'll run the other memtests as soon as I can.

 

[Dave76]

Can you post the CPUZ tabs?

This information will help with further trouble shooting.

Can you also check your BIOS for NB or memory controller voltage, and let us know?

In your BIOS MB Intelligent Tweaker(M.I.T.) > NB Voltage Control

 

[Dutchmv]

hey,

I went into the BIOS and loaded optimized defaults again. After that I checked the memory settings before I started memtesting. They are the way Wishmaster said the should be.
I have know idea how to check if all the other voltages and speeds are good the way they are set now. I don't know enough about PC's for that.

memtests:
I just ran the first memtest with one stick in slot 1. It ran for 5:10 with 11 passes. There were three errors in pass 0,3,8. All with test 7.
In the second test I put the other stick in the same slot 1. It ran for about 4 hours with 9 passes. There were no errors.
Update: In the third test I put the memory stick with the errors in the second slot. It ran for 2.24 hours and had five passes. There was 1 error in pass 3 test 8.

CPU-Z:
Note that these pictures are taken with one stick in the memory, the stick that passed memtest with no errors. The dual memory picture was taken after a restart.
CPU:
View attachment 98622
Caches:
View attachment 98623
Mainboard:
View attachment 98624
Graphics:
View attachment 98625
Memory Single:
View attachment 98626
SPD:
View attachment 98627
Memory dual: (picture taken seperatly)
View attachment 98628

Other questions:
The NB voltage control is on auto, and the auto value is 1.100V.
The BIOS version is F4B.
As you can see in the dual memory picture. DC mode is unganged.

Should I do the memtests with the memory in the second slot, or is all the informetion here to pinpoint the problem?

 

[Dave76]

The information from CPUZ will help us know the hardware and settings.
SPD tab is the manufacturer recommended settings.
Memory tab shows the first four RAM timings, tRC and CR.

But not RAM voltage, Can you check the BIOS for the RAM voltage?
According to your SPD tab it should be 1.5v. Please confirm.

You have 2X2GB RAM cards, correct?

If the above is all correct, bump your NB to 1.2v and run memtest86+ with both RAM cards installed.

If you have an error, run the test with one card at a time, let us know the results.

If no errors, run Prime95 blend test for one hour or stop if you get any errors.

Let us know the results.

 

[Dutchmv]

The voltage of the memory is 1.5V
I do have 2x2gb RAM

I changed the nb to 1.2V and tested the memory with memtest. I ran it for 6.5 hours with 8 passes. there were no errors. Just as there were no erros when the nb was 1.1V

I ran a prime blend test two cores crashed with a rouding error within a minute.

I haven't got around testing the ram modules seperatly yet.

 

[Wishmaster]

1 stick passes, the other fails?
I would think that RAM module may be defective.


As previously suggested, at this point you may try lowering the RAM speed to 1066, see if that stabilizes the system.

 

[Dave76]

Agree with Wishmaster,

For now leave the NB at 1.2v.

Run Prime95 with one RAM stick at a time see if either fails.
I would RMA if one fails.

As another test, set to SPD settings JEDEC #3.
518 MHz
7-7-7-19 and 1.5v.

See if that will pass memtest86+ and Prime95.

Let us know the results.

 

[Dutchmv]

hey there,

I just applied for an RMA at the internet shop where I bought the stuff since there was no way to get it running without errors.

Thanks for all the help you guys gave me, no way I could have found it out by myself.
One last question thought. When I load optimized defaults in the bios, the nb stays at 1.2V is this normal and do I have to manualy put it at the right voltage from now on?

 

[Dave76]

Your welcome,

That's going to be the best approach, Hope the new RAM runs for you.

The NB should be 1.1v at default, optimized defults might stay at 1.2v depending on your BIOS.
I would put it back to 1.1v when you get the new RAM.

Let us know how your system is running with the new RAM.

 

[Dutchmv]

Everything is fixed

Hey there,

I send the memory ram away and the company tested it. It was a hardware defect and they send me new modules. Now, everything works fine. The prime blend test ran for two hour and there were no errors.

thansk again for all the help, and I did the rep thing.

hopefully, I won't be back soon.

bye