burn in has stop working

[Jimmy]

Dear Sir,

There is an error message when I test burn in under Vista,OS pop up an error message which showing"burn in has stop working",SW hang up,but OS is active,intel G31 and AMD RS740 all have the issue,and it is random,please help the to check and give your comment,many thanks.

[Ian (PassMark)]

Can you email us the crash address from the system event log to help [at] passmark [dot] com


This is often caused by BurnInTest exercising a faulty device driver.

I would start by checking you have the latest graphics card device driver - in our experience this is the most likely candidate.

If this does not resolve the problem, then start reducing the test set until you have the smallest test set that causes the problem (e.g. just the Network test running). This should help direct you to the problem hardware or device driver.

Can you describe the problem in more detail, such as what tests are running, how long into the tests does the problem occur? Which version of BurnInTest are you using, eg. 5.3.1014

Regards,
Ian

[jimyin]

Dear Sir:
Same issue find too,
Some units appear this fail symptom only run burnin program, and fail appear between 2~8 hours test. and fail rate about 3%.

BurnIn V5.3 1012 & 1014

Both fail on Intel and AMD platform:
Intel Platform: CPU: Intel(R) Core(TM)2 Duo P8600 & P9300
Chipset: Intel Cantiga (GM & PM)/ICH9M
OS: Windows Vista Ultimate, Service Pack 1
AMD Platform: CPU: AMD Turon QL60, M82
Chipset: AMD RS780M/SB700
OS: Windows Vista Ultimate, Service Pack 1

Test Items:
CPU-Maths
CPU-SIMD
Memory(RAM)
2D Graphics
3D Graphics
Disk
Sound

[passmark]

Did you try Ian's suggestion to reduce the test set to insolate the problem.

How many machines did you test to calculate the 3% rate? If you are seeing a 3% failure after testing 100's of machines, and all the machine are in exactly the same configuration, then it might be a real hardware issue. 3% failure wouldn't be too surprising for new PC's.

More details of the actual error would also help. I note that you post says you added a picture. But I don't see any picture.

[Ian (PassMark)]

In addition, for a "BurnInTest has stopped working" error we also need the following information:
- The specific build number: eg. BurnInTest Professional v5.3.1014
- The address related to this problem, eg 0x00412345. If this information is not available from a link from the "BurnInTest has stopped working" error window then you can normally get this information from the Windows Application event log. In XP (other operating systems will be similar) the Application event log is here:

Start-> right click "My Computer" and select Manage, double click "Event Viewer" on the left hand side of the window, double click "Application" and find the event of type "Error" around the time the problem occurred. This will contain information like:
"Faulting application bit.exe, version 5.0.1003.0, faulting module bit.exe, version 5.0.1003.0, fault address 0x00412345."

So if you could also send this information.

Regards,
Ian

[jimyin]

Dear Sir:
"Did you try Ian's suggestion to reduce the test set to insolate the problem" -->Not have this test yet, for the units on my hand is not enought and the symptom is not easy to reappear.

We have more than 5 project all appear the symptom. And HW design have huge different of the five project, so we don't think it is HW issue.

There is one of failure unit error messae
"Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x27262535, process id 0xc18, application start time 0x01c8452a70f1139d."

[jimyin]

BTW, did you the test burnIn program on OS: Windows Vista Ultimate, Service Pack 1

[Ian (PassMark)]

This address is outside the BurnInTest software address space and the faulting module is Unknown (not bit.exe). The address would typically imply that there is a fault in a dll (dynamic link library) that BurnInTest has loaded and exercised, e.g. a DirectX library such as d3d9.dll. Please check some of your other systems and see if the fault information (as posted) is the same for other systems. Also, when you next find a system with this problem, please start reducing the tests to find the minimum test set.

I would make sure you have the latest DirectX 9.0c installation and device drivers.

Regards,
Ian

[jimyin]

Dear sir:
For burinin program stop working issue, there are 12 units fail again. All the fail units have the same error code 0X005. (test items:CPU-Maths,CPU-SIMD,Memory(RAM),2D Graphics,3D Graphics,Disk,Sound)
We take the 12 units to two group to do below reducing test(GroupA: uints 1~6, GroupB: units 7~12) base on below test result, would you give us some sugestion?
-----------------------------------------------------------------------
Step1: GroupA: 2D+3D, 6h test PASS
GroupB: CPU-Maths+CPU-SIMD+Memory(RAM)+Disk+Sound, 6h test fail 5 units of 6 units.
-----------------------------------------------------------------------
Step2: GroupA:CPU-Maths+CPU-SIMD+Memory(RAM)+Disk, 6h test fail 3 units
GroupB:CPU-Maths+CPU-SIMD+Memory(RAM)+Sound, 6h test PASS
-----------------------------------------------------------------------
Step3: GroupA: CPU-Maths+CPU-SIMD+Disk, 6h test PASS
GroupB: Memory(RAM)+Disk, 6h test one unit fail
-----------------------------------------------------------------------
Step4: GroupA: Memory(RAM) 6h test PASS
GroupB: Disk 6h test PASS
-----------------------------------------------------------------------

[passmark]

All the fail units have the same error code 0X005

In your previous quote you said it was 0xC0000005. It would help to be precise. Does the fault offset address 0x27262535 also always remain the same for each crash or is it somewhat random?

If it is somewhat random then it might be a temperature related issue or EMI. If the fault offset address is always the same then it is more likely a software issue.

[Ian (PassMark)]

When you get "... has stopped working" then the useful information is:
- the software address where the problem occurs. On 32-bit Windows this is always a 32-bit number, i.e 0x12345678 (8 hexadecimal digits). This address can be in the software, BurnInTest (.exe), a device driver (.sys) , a Library (such as a DirectX dll) or the Windows Operating System. The previous address you sent was not in the BurnInTest address space and was not in the Windows Operating System address space (but rather appeared to be a dll). The new address is not complete, and I guess is not actually the address.
- the type of error, again this is a 32-bit number , i.e 0x12345678 (8 hexadecimal digits) (e.g. 0xC0000005 is an Access Violation).

If you could go through the Windows application event log on ALL systems that failed and post this information it may help (as described in a previous post). This should help determine if it is the same problem on all systems. The Number you sent does not help as it is not complete.

Your testing would seem to suggest that there is a problem when the Memory and Disk tests are run together. This may imply a problem like your Disk drive controller device driver failing under a low memory resource scenario. I would suggest looking at updating your Disk drive controller device drivers and RAID BIOS (if relevant).

Which Memory test are you running?
What type of Disk arrangement do you have?

Regards,
Ian

[jimyin]

Add another 8 units fail information:
1. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x2221203a, process id 0x4e8, application start time 0x01c84758cf1a031e.
2. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x2221203a, process id 0x874, application start time 0x01c8473921ff6fa9.
3. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x3f800021, process id 0xe7c, application start time 0x01c84751a7037476.
4. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x55594f87, process id 0xb98, application start time 0x01c847992464f3e6.
5. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x2221203a, process id 0xfa8, application start time 0x01c8474c1b87a996.
6. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x27262535, process id 0x200, application start time 0x01c8479983355285.
7. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x27262535, process id 0xe84, application start time 0x01c8479b911fa94c.
8. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x55594f87, process id 0xd14, application start time 0x01c84785affb3c28.
9. Faulting application bit.exe, version 5.3.1014.0, time stamp 0x479fa8c3, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x2221203a, process id 0xdb4, application start time 0x01c8479c2d2efc82.

[jimyin]

Dear Ian sir:
We have 5 project have the same failing, Memory and Disk not with the same type.
Memory: DDR2 (667)1G, 2G, 4G, DDR3 2G
Disk: Sata 120G,160G, 250G

[jimyin]

Had done below test:
Step5: 12 uints only test Memory (100% loading), 6h test PASS
Step6: 12 uints only test Disk (100% loading), 6h test PASS
Step7: 12 units test 2D+3D+CPU-Maths+CPU-SIMD+sound+Memory(RAM), 6h test pass
Step8: 12 units test 2D+3D+CPU-Maths+CPU-SIMD+sound+Disk, 6h test one unit fail

Base on above test, it seems Disk + other test items lead BurnIn program stop working.
Would you have some opinion?

Best regards
Jim

[Ian (PassMark)]

It is not conclusive, but yes it would seem that it is the disk test with another test. Are you submitting the error to Microsoft when you get the error (from the error Window, select the "Send Error Report" button) as this will give us more information about the problem.

Regards,
Ian

[jimyin]

Dear Ian:
Did you think it is Microsoft software issue? Would you explain the reason?

thanks

[Ian (PassMark)]

To clarify, I still think the most likely issue is a fault in a software library (.dll) or device driver. When Software stops working like this, the Microsoft operating System has a mechnism to collect quite detailed technical information about this error. This information is sent to Microsoft when you select the "Send Error Report" button from the "BurnInTest has stopped working" Window. This sends a "MiniDump" of technical information about the failure to Microsoft. We get this information from Microsoft and using some tools can determine much more detail about where the problem is. So I was not suggesting it is a Microsoft software issue, but rather that you should sendn this information to Microsoft so we can get the MiniDump files to investigate further.

I hope that makes sense.

Regards,
Ian

[Ian (PassMark)]

If you send us am email at, help [at] passmark [dot] com, then I will send you a debug version of BurnInTest to try.

Regards,
Ian

[jimyin]

Hi, Ian:
20 units run this version (1019 debug) over 48h, no find failing...
Test items (CPU-maths, CPU-SIMD, 2D, 3D, Memory, disk, audio)

[Ian (PassMark)]

Can you try the release version of BurnInTest 5.3.1019, available here:
http://www.passmark.com/ftp/bitpro.exe

Thanks.
Ian

[jimyin]

Hello, Ian:
OK, I will try the release version (1019) program later.
but should feedback another lot units test result with the debug version program. Total test 90 units, no find failing "Stop working", but there are 5 units disk test fail.

Test platform and configuration:
Operating system: Windows Vista Business Edition Service Pack 1 build 6001
Number of CPUs: 1 (2 Core(s)/CPU, 1 Logical(s)/Core)
CPU manufacturer: GenuineIntel
CPU type: Intel(R) Core(TM)2 Duo CPU T9600 @ 2.80GHz
CPU features: MMX SSE SSE2 SSE3 DEP PAE
CPU1 speed: 2793.0 MHz
CPU L2 Cache: 6 MB
RAM: 3000 MB
Video card: Mobile Intel(R) GMA X4500HD (Resolution: 1024x768x32)
Disk drive: Model Hitachi HTS543232L9A300 ATA Device (Size: 298.1GB)

Fail information:
SERIOUS : 2007-12-25 20:49:49, Disk, C: Error while reading from the disk
LOG NOTE: 2007-12-25 20:49:49, Disk, File: C:\BurnInTest test files\~bittestC00086, Block Size: 32768, Block Num: 7105, Ops: 58558873600, ErrorCode: 0x45d
LOG NOTE: 2007-12-25 20:49:49, Disk , Disk Error (C:): Error while reading from the disk in Cycle 85 with pattern High Low freq overwrite

[Ian (PassMark)]

You should not use the Debug version for commercial testing purposes, only for debugging specific problems.

The disk error just looks like a real disk problem.

Regards,
Ian

[jimyin]

Use the release verison (1019) program still appear 12 units fail. fail rate:12/300.
BurnInTest VersionV5.3 Pro 1019Computer NameFACT_PREMP-PCDate05/12/08 22:57:14Operating systemWindows Vista Ultimate Edition Service Pack 1 build 6001Number of CPUs1 (2 Core(s)/CPU, 1 Logical(s)/Core)CPU manufacturerAuthenticAMDCPU typeAMD Turion(tm) X2 Dual-Core Mobile RM-70CPU1 speed1999.6 MHzCPU L2 Cache:512 KBCPU featuresMMX 3DNow! SSE SSE2 SSE3 DEP PAE RAM1790 MBVideo cardATI Radeon HD 3200 Graphic (Resolution: 1024x768x32)Disk driveModel WDC WD25 00BEVS-22UST0 SCSI Disk Device (Size: 232.9GB)

Faulting application bit.exe, version 5.3.1019.0, time stamp 0x482d11f5, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x27262535, process id 0x7d0, application start time 0x01c8b4c2e6e536b1.
Faulting application bit.exe, version 5.3.1019.0, time stamp 0x482d11f5, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x2c2b2a30, process id 0x5ec, application start time 0x01c8b4be2fb3cdde.

[Ian (PassMark)]

I still think the most likely issue is a fault in a software library (.dll) or device driver, maybe a Vista related issue. To date I have been unable to get the debug reports from Microsoft for this specific error to see if that helps explain the cause. I can tell you that it broadly falls under a category of errors, StackHash_1703. If you Google "StackHash_1703" you will see that there are many reports of software with problems with this category (module unknown, hash of stack = 1703). This implies to me that there is a good chance that the problem is a Windows or common 3rd party software component.

Out of interest do you have Nero installed on these systems?

In the latest build of BurnInTest we have added our own generation of "minidump" files. Can you start using the new build, as it should produce debug files when a crash occurs and you can send these to us.

The latest version of 32-bit BurnInTest Pro, v5.3.1020, is available here:
http://www.passmark.com/download/bit_download.htm

Regards,
Ian

[Michael]

Build 1019/1020 will not found "stop working" issue

But there is a sound error at random rate on P45/G45/X38/P35/965 Platform

Vista Ultimate SP1 32bit + nVidia VGA Driver 175.16 or Ati VGA Driver 8-5

Is this two Build Version is a Debug Version ? And which Version Burn-in Test we should to use under Vista Ultimate SP1 Recently ?

[passmark]

Michael, a few points.

1) If you have a new issue, like a sound card error, please start a new thread in the forum. By posting unrelated issues in this existing thread you are only confusing everyone.

2) I know English probably isn't your first language, but nevertheless, you need to write something we can quickly understand. If we need to re-read it 4 times and guess at the problem, you can't expect an accurate reply from us.

3) Regardless of your language problems, saying there is a "sound error" isn't a adequate description of whatever your problem is.

You should use the latest version of the software. See the post by Ian above.

[tjd]

I still think the most likely issue is a fault in a software library (.dll) or device driver, maybe a Vista related issue. To date I have been unable to get the debug reports from Microsoft for this specific error to see if that helps explain the cause.


Has there been any progress identifying this? There are a few threads about this sort of failure but none of them say whether it's been understood/fixed/worked around yet.

We're seeing this (and similar) bugs in a virtual-machine environment; normally I'd assume the bug is the hypervisor's fault but since it's happening on real hardware too it'd be great if we could work around it.

Cheers,

Tim.

[passmark]

tjd,

Like pretty much this entire thread, your post is too vague to provide any assistance.

We're seeing this (and similar) bugs..

This is NOT an adequate fault report. Especially in the context of this thread.

Put yourselves in our position. How on earth could we fix anything if the entire description of the problem is, "there is a bug"?



Here is a minimal list of information that is require for a fault report.
The software version.
The operating system version.
The hardware in question
A detailed description of the problem and how to reproduce it.
A copy of the result files / log files (if available).
Screen shots (if available)
What you have already tried to correct the problem.
Why you think the problem is with our software and not your hardware

[tjd]

tjd,
This is NOT an adequate fault report.


It wasn't intended to be a fault report, it was a question about a fault someone else reported. Burnintest turns up a bunch of hypervisor bugs for us (that's why we use it, it's great!) but right now we're having trouble knowing which of them are our fault and which might be blamed on this issue.

I can give you details of errors we've seen if you want, but I think it would make things more confusing for you, not less.

Cheers,

Tim.

[passmark]

The fact remains, the quality of our responses depends on the quality of the information provided.
Vague request = Vague response.

[tjd]

We have seen, when running 72-hour burnintests in Xen virtual machines:

On Vista SP1 32bit, running burnintest 1012:
Faulting application bit.exe, version 5.3.1012.0, time stamp 0x4715ab72, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x01010101, process id 0xefc, application start time 0x01c8c3d48f8decae.

On Vista SP1 32bit, running burnintest 1022:
Faulting application bit.exe, version 5.3.1022.0, time stamp 0x4850d46a, faulting module unknown, version 0.0.0.0, time stamp 0x00000000, exception code 0xc0000005, fault offset 0x28272634, process id 0xc94, application start time 0x01c8de315c210b78.

I'm afraid I don't have access to the burnintest log files from those runs right now.

Earlier posts from Ian suggest that you think errors like these are caused by something in Vista, or a driver, interacting poorly with burnintest. Is there anything more known about that?

We have also seen other failures that look like test patterns have overwritten something but that was on 2k3 and that's not something that's been reported on real hardware.

Once again: I don't think this is a bug in your software. But if there is a bug (in Windows?) that makes burnintest do this (as reported above by jimyin) it would be great to know whether it's understood, fixed, or can be worked around so we can get back to our own bugs.

Cheers,

Tim.

[passmark]

You should upgrade to the latest patch release to start with. Which is V5.3 Build 1026 as of today.

The fault offsets 0x01010101 and 0x28272634 don't match the posts above. You are also not using the the same software release and are not running in the same environment. So there is no immediate reason to think this is the same as one of the issues above.

Crashes at high addresses like 0x28272634 are often device driver faults however. Nothing more can be added based on the information provided.

Otherwise initially debugging steps have already been covered in the many posts above.