frequent random BSODs

[cluberti]

This is another set of dumps pointing at the CPU - one is showing the CPU going into a halt state and causing a bugcheck, and the second is the CPU trying to come back out of a halt state and causing a bugcheck (both fail in kernel because an interrupt failed to find the thread on the CPU that should have been available, but the problem isn't the kernel).

Let's have you try disabling the C3 power state via the registry, and see if that keeps the CPU from going so far to sleep that it won't come back:
reg add HKLM\System\CurrentControlSet\Control\Processor /v Capabilities /t REG_DWORD /d 0x0007c044
Reboot for the change to take effect, of course.

 

[thargas]

sleeping?

Hmm. I'd say that this shouldn't happen at all as I disabled the machine from power-saving by creating a profile which doesn't do it.

Anyway, I'll try this.

 

[thargas]

not sleeping

Here's a minidump from dvd creator

 

[Tews]

The dmp(s) point to ntkrpamp.exe or ntoskrnl.exe as the probable cause. We know this is not the case, as these are core OS files, and will definitely not cause a BSOD. They are just the default files the debuggers blame when they can't see a better cause. Enable driver verifier so we can stress the drivers..

- Go to Start and type in "verifier" (without the quotes) and press Enter
- Select "Create custom settings (for code developers)" and click "Next"
- Select "Select individual settings from a full list" and click "Next"
- Select everything EXCEPT FOR "Low Resource Simulation" and click "Next"
- Select "Select driver names from a list" and click "Next"
Then select all drivers NOT provided by Microsoft and click "Next"
- Select "Finish" on the next page.

Reboot the system and wait for it to crash to the Blue Screen. Continue to use your system normally, and if you know what causes the crash, do that repeatedly. The objective here is to get the system to crash because Driver Verifier is stressing the drivers out. If it doesn't crash for you, then let it run for at least 36 hours of continuous operation (an estimate on my part).

 

[cluberti]

Hmm. I'd say that this shouldn't happen at all as I disabled the machine from power-saving by creating a profile which doesn't do it.

Anyway, I'll try this.

You have to remember that the CPU goes into sleep states regardless of what your power options are - you're changing the way the *OS* does it's power management in the power config, but the CPU can still go into sleep states (C-states) when idle if designed to do so.

 

[thargas]

still bsod after disabling c3 power state

and setting verifier as per tews

 

[ttran]

Both of these point to your AMD processor driver. What have u done to it?

Microsoft (R) Windows Debugger Version 6.12.0002.633 X86
Copyright (c) Microsoft Corporation. All rights reserved.


Loading Dump File [C:\Users\TUANTR~1\AppData\Local\Temp\Rar$DI00.333\081610-73710-01.dmp]
Mini Kernel Dump File: Only registers and stack trace are available

Symbol search path is: SRV*e:\symbols*http://msdl.microsoft.com/download/symbols
Executable search path is: 
Windows 7 Kernel Version 7600 MP (4 procs) Free x64
Product: WinNt, suite: TerminalServer SingleUserTS
Built by: 7600.16617.amd64fre.win7_gdr.100618-1621
Machine Name:
Kernel base = 0xfffff800`02a0e000 PsLoadedModuleList = 0xfffff800`02c4be50
Debug session time: Mon Aug 16 08:46:02.689 2010 (UTC + 7:00)
System Uptime: 0 days 1:57:07.624
Loading Kernel Symbols
...............................................................
................................................................
........................
Loading User Symbols
Loading unloaded module list
.....
*******************************************************************************
*                                                                             *
*                        Bugcheck Analysis                                    *
*                                                                             *
*******************************************************************************

Use !analyze -v to get detailed debugging information.

BugCheck A, {ffffffffccfe00b0, d, 1, fffff800030078f0}

Probably caused by : amdppm.sys ( amdppm!C1Halt+2 )

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

 

[cluberti]

Agreed, probably good for the OP to answer if he's modified any of these files in any way. However, if not, the dump is showing that the CPU was in a C1Halt state at the time of the failure. This still looks 100% like a hardware issue.

 

[Jonathan_King]

Try this CPU test: http://www.sevenforums.com/tutorials/100352-cpu-stress-test-prime95.html

 

[thargas]

amdppm.sys

Nope. I've not modified it in any way. People have suggested that it should be upgraded, but I've not found a way to do that.

However, I finally put together some other info, i.e. that this is a MS provided driver, and started looking for amdppm.sys on the microsoft site. I didn't find an updated driver, but I did find people saying that the Athlon 64 processors should be using amdk8.sys not amdppm.sys and that this was a problem with windows update choosing the wrong driver.

I've looked on disk and both amdppm.sys and amdk8.sys appear to be present in the same directory. So the question I've got is: how do I make it use amdk8.sys instead? Given that it's a processor driver, I'm somewhat nervous about just renaming amdppm.sys to get it out of the way.

BTW, I think that the power-state hack seems to have made it more stable. I.E. it only crashes when I try to make it crash by making a DVD with windows DVD maker.

 

[thargas]

prime95 test

Hi Johnathan.

I've run prime95 before. Not as long as people have suggested because I was concerned about the temperature of the cpu. fanspeed said that it was up to 71C which sounded excessive to me. However, it didn't crash.

If someone can tell me what a normal temperature range for Athlon II 64 is, I can look into prime95 again.

 

[Jonathan_King]

Max temp for it is likely either 63 or 70C. Time to investigate some more cooling for your machine. Open the side of the case, clean out the dust, and then blow a fan inside while you work.

 

[thargas]

it's not temperature

or at least not only temperature. I can "reliably" bsod the machine by running windows dvd creator with the temperature never getting beyond 60C.

 

[Jonathan_King]

Could it be the DVD drive causing the issue? Try removing it and see what happens,

Have you tried a clean install or repair install?

 

[thargas]

now we're getting somewhere

I opened the box up and got out the *big* fan and set it up blowing straight into it on high. Idle, the temperature dropped to 37C. Then comes the test: running windows dvd creator. The temperature peaked at about 52C, almost 10C less than without the fan, and it finished creating the dvd instead of bsoding, which it always did before.

Now I have to figure out:

1) is this a problem with the processor, or
2) I just need more cooling

If 2, then I need to figure out how to get the effect of a huge high-speed fan into my case.

 

[ickymay]

I opened the box up and got out the *big* fan and set it up blowing straight into it on high. Idle, the temperature dropped to 37C. Then comes the test: running windows dvd creator. The temperature peaked at about 52C, almost 10C less than without the fan, and it finished creating the dvd instead of bsoding, which it always did before.

Now I have to figure out:

1) is this a problem with the processor, or
2) I just need more cooling

If 2, then I need to figure out how to get the effect of a huge high-speed fan into my case.

controlling temperatures is rarely about big fans that sound like a jet engine and is actually about airflow

a fan at the front of your box is often useful to create pressure inside the box and a fan at the back of your box often gives good results, given that
you need to create clean air flowing in and out of the box across all the components and so that the pressure inside is just slightly higher than outside ......

a good quality fan on the CPU is essential and taking warm air away from the GPU is important ...........

If you need any help maybe take a pic of your box from the side so we can see your fans and let us know the make and model of your CPU cooler ?

N.B. on my current system I use an Arctic Cooling Freezer 64 Pro and this runs at a happy and quiet 40deg core temp rising only by 10deg when stressed , the noisiest fan on my system is on my power supply which is quiet

 

[thargas]

fixed

Just for completeness, I figured I'd make a final note.

The problem appears to be fixed. I took the machine in to a local store, and after a while, they agreed with me that it was over-heating. They replaced the goop between the cpu and its heat sink and that seems to have done it.

Thanks to all the helpful folks here who helped me get there.

 

[CarlTR6]

That is good news, thargas. Thanks for reporting back and sharing your solution.