[Glader]

System specs:

Windows 7 Pro 64bit provided by MSDNAA, most recent reinstall was exactly 2 months ago after I got my new graphics card (see below).

Intel i5 760 CPU running at stock 2.8Ghz
Gigabyte ATI Radeon 6870
Asus P7P55D LE motherboard
OCZ 700W PSU

PSU is just a week old, Graphics card 2 months, the rest is 6 months old.
RAM sticks have just been checked OK by memtest86

Ever since I got my new graphics card I've been experiencing a whole lot of system instability.
I've been getting BSOD's with increasing frequency (just today I think I'm up at 4, giving me an average of one every hour).

According to the bluescreen messages it's been for a number of reasons, in the beginning it was just atikmpag.sys so I suspected
the graphics drivers. I tried a whole bunch of different ones without success.
Then I started getting them because of dxgmms1.sys, tcpip.sys, wdf01000.sys and today for the first time I got it because of
MEMORY_MANAGEMENT.

I'm running driver verifier as we speak.

 

[VirGnarus]

I noticed there is a gap in your crashdumps concerning their dates, about a month in size. Plus that there's only 6 of em, 3 from january (19th, 21,26) and 3 from march (2, 2, 5). Doesn't really coincide with your statement that you get one around every hour, especially when none of them have today's date. If you are using any software that does any sort of 'cleaning' of any means (aside from antivirus/spyware removal), you may want to prevent using them (matter of fact, you should prolly uninstall them for various reasons).

What was your testing with the RAM? How many passes? Were you doing extended runs? Did you check slot positions as well?

What settings do you have active on DV? Also what drivers did you select? In order for DV to work most efficiently it should be a small amount (3-5) of drivers selected, and none of them should be MS/Windows drivers. Wait for it to crash, and if the bsod error code shows anything regarding any variation of Verifier, or if it mentions Special Pool, then you can be certain DV nabbed something. If not, select another small group of drivers and let it crash from those. Send all crashes to us, regardless of their error code.

If you say this started occurring shortly after your new video card, it's probably not coincidental. You may want to end up RMAing the sucker and getting a replacement. If you got a spare one, or you have onboard video as well, try using one of those temporarily (make sure to uninstall drivers for the graphics card you recently purchased). If your BSODs start clearing up after making the switch, you found your problem.

Does safe mode also experience these frequent crashes? If not, it most likely is a driver prob. If it does, it's more likely to be hardware.

Here's some more hardware tests:

CPU: Prime95
GPU: FurMark
Disk: Seatools

 

[Glader]

I just checked the system failure settings and for some reason the logging was turned off (I forgot that I had turned that off some time ago to turn off the swap file). I guess I'm gonna have to wait for a while before getting new BSOD's to report. I let memtest run overnight for 10 hours, during which it had done 7 passes. I did the testing with default settings and all RAM-sticks in (occupying all slots).

I followed the driver verifier instructions here on techsupport and used those settings (what to test, all non-win drivers etc). Since I started doing that I've had one BSOD (apparently not caught by DV since I did not see anything regarding verifier or special spool).

I currently have an RMA issue open with ATI/AMD and I've given them some dumps they asked for to check out any source but since the BSOD's have gotten weirder and weirder I've now started looking for alternative sources, and that's why I'm here asking for help. =)

I have not tried running in safe mode since I'm using the computer during the days, and I'm pretty sure that my CPU is not the issue since I have been able to run it under 100% load 24/7 on a 40% overclock for a couple of months without issues. Don't worry, I used aftermarket cooling and the little thing was at a comfy 60C temp.
And the first thing I did after getting the new GPU and fresh windows install in was to run Futuremark so temps are not an issue there either. I will however run it again for ****s and giggles (hey, here's hoping).

Also: No spare/on-board graphics card.

I've also run chkdsk and it did not give me any lip, as well as my system-disk's S.M.A.R.T. info not changing so I'm guessing that the disk is not an issue.


Right now I've noticed a curious thing though: My sound's acting up. I keep getting little freezes/buffer underruns resulting in sound looping for half a second-ish a couple of times per minute. I'm now switching from my PCI soundcard to the on-board and deactivating the PCI one. Don't know how/if this is related but I'm really hopeful that it's the sound card and not the graphics card that's broken - if any.

Anyway, I'm gonna do some additional testing as per your recommendations and upload any new crashdumps.

 

[dai]

what model is the 700w

 

[Glader]

The PSU is an OCZ ModXStream Pro 700w

The GPU has now passed a run at 3DMark Vantage,
the system drive has passed SeaTools's Long Generic and Long Drive self tests, and my CPU has been chewing [email protected] at 100% load on all 4 cores for 2 hours now and is enjoying itself at a nice 45C.

I have also limited DV to only testing 5 drivers at the same time, hoping it'll result in catching something.

Running Furmark for 20 minutes now. Temperature holding at 70C and I'm not noticing any problems yet.

Right. I let Furmark and [email protected] stay on for the night and nothing's blown up. For the time being I'm gonna call the PCI sound card the most likely culprit. Updates might follow.

Aaaand BAM! BSOD :upset:

At least I'm getting the minidumps now again, see attached.

Here we go again. This time because of good ol' atikmpag.sys

And here's today's minidump.

Bonus BSOD courtesy of usbhub.sys
Sorry about the frequent replies but I can't edit my old posts.

 

[reventon]

Hi,

Several BSODs mentioning graphics drivers, including one TDR failure BSOD (where the recovery process for graphics timeouts failed).

Did you get one of the BSODs while running Furmark? If so, then I agree 100% with:

If you say this started occurring shortly after your new video card, it's probably not coincidental. You may want to end up RMAing the sucker and getting a replacement.

BUGCHECK SUMMARY

[font=lucida console]
Built by: 7600.20826.amd64fre.win7_ldr.101026-1502
Debug session time: Sat Mar 19 11:25:55.264 2011 (GMT+13)
System Uptime: 0 days 1:10:29.295
Probably caused by : usbhub.sys ( usbhub!UsbhGenDispatch+4f )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
PROCESS_NAME:  System
BUGCHECK_STR:  0x7E
FAILURE_BUCKET_ID:  X64_0x7E_usbhub!UsbhGenDispatch+4f
Bugcheck code 1000007E
Arguments ffffffff`c0000005 fffff880`042f2faf fffff880`03185828 fffff880`03185090
SystemProductName = System Product Name
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.20826.amd64fre.win7_ldr.101026-1502
Debug session time: Sat Mar 19 10:14:06.435 2011 (GMT+13)
System Uptime: 1 days 4:47:24.466
*** WARNING: Unable to verify timestamp for atikmpag.sys
*** ERROR: Module load completed but symbols could not be loaded for atikmpag.sys
Probably caused by : atikmpag.sys ( atikmpag+73bc )
DEFAULT_BUCKET_ID:  GRAPHICS_DRIVER_TDR_FAULT
BUGCHECK_STR:  0x116
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_0x116_IMAGE_atikmpag.sys
Bugcheck code 00000116
Arguments fffffa80`0788c4e0 fffff880`03efd3bc 00000000`00000000 00000000`00000002
SystemProductName = System Product Name
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.20826.amd64fre.win7_ldr.101026-1502
Debug session time: Fri Mar 18 05:25:43.747 2011 (GMT+13)
System Uptime: 0 days 19:57:50.778
*** WARNING: Unable to verify timestamp for atikmpag.sys
*** ERROR: Module load completed but symbols could not be loaded for atikmpag.sys
Probably caused by : atikmpag.sys ( atikmpag+73bc )
DEFAULT_BUCKET_ID:  GRAPHICS_DRIVER_TDR_FAULT
BUGCHECK_STR:  0x116
PROCESS_NAME:  System
FAILURE_BUCKET_ID:  X64_0x116_IMAGE_atikmpag.sys
Bugcheck code 00000116
Arguments fffffa80`0abcf4e0 fffff880`03e073bc 00000000`00000000 00000000`00000002
SystemProductName = System Product Name
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``
Built by: 7600.20826.amd64fre.win7_ldr.101026-1502
Debug session time: Thu Mar 17 09:26:26.558 2011 (GMT+13)
System Uptime: 1 days 4:59:43.978
Probably caused by : dxgkrnl.sys ( dxgkrnl!DoublyLinkedList<DMMVIDPNSOURCEMODE>::DoublyLinkedList<DMMVIDPNSOURCEMODE>+c0 )
DEFAULT_BUCKET_ID:  VISTA_DRIVER_FAULT
PROCESS_NAME:  csrss.exe
BUGCHECK_STR:  0x7E
FAILURE_BUCKET_ID:  X64_0x7E_dxgkrnl!DoublyLinkedList_DMMVIDPNSOURCEMODE_::DoublyLinkedList_DMMVIDPNSOURCEMODE_+c0
Bugcheck code 1000007E
Arguments ffffffff`c0000005 fffff800`02bfa90d fffff880`07a37f28 fffff880`07a37790
SystemProductName = System Product Name
¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨¨``[/font]

 

[VirGnarus]

I want to make note that while 3dmark Vantage is an exceptional benchmarking utility (and now it's free! Huzzah!), it just doesn't do the job when it comes to testing stability of a GPU. It's designed to ensure efficiency to determine the quality of one's system, not so much catered to help test hardware stability. I've seen plenty of people stating that while 3Dmark ran through no prob in their 'stability test', other programs designed more for stability (such as OCCT and FurMark) conked out, and those ended up being most accurate on what was actually taking place.

The same applies for [email protected] It can suck up a lot on the CPU when it's given the opportunity too, but it's designed to be gentle on the CPU. Whereas Prime95 is designed to where there is absolutely positively no room for failure. Its Torture Test is pretty true to its name, and with the Small FFT setting, it's designed to work solely on the CPU and not even touch RAM or disk during the actual processing, which makes it great to isolating a prob to CPU-related troubles.

As for the sound hardware, often times the sound hw tries to work in tandem with the graphics card, and if one bugs out, the other one ends up doing the same to try to keep itself synched with what the graphics card is doing. It may well be the sound hw, but be aware that because you hear sound problems, it isn't exactly the sound hw or drivers that are being problematic; much like video problems can sometimes end up being from memory or cpu issues.

 

[Glader]

reventon and VirGnarus: I ran Furmark overnight (for about 10 hours) without any problems a couple of days ago.

VirGnarus: I guess I'll be running Prime95 overnight today then

 

[Glader]

9h of prime95 with small FFTs passed.

 

[VirGnarus]

I perused further and noticed paging file and hard drive controller errors being rather prevalent. You may wanna look into that. Try Prime95 with mixed tests this time as well.

There's one bsod in your latest batch that mentioned usbhub, but that's the only one, the rest are directx. Again, while furmark doesn't show up anything either, I still wouldn't up on it being the graphics card, especially seeing it happened only after installation of it.

I also note that DV (Driver Verifier) has not been enabled on your past batch of crashdumps.

 

[Glader]

I find it very curious that DV isn't on since it keeps saying "No settings were changed" when I start it with the recommended settings and 5 drivers selected.

Since the last BSOD I have also installed newer drivers for the mobo's chipset manually and it seems they have finally been installed. Since this I have not yet had a BSOD for almost 2 days. I am once again cautiously optimistic.

 

[Glader]

Got another BSOD today, don't know the reason (came home to black monitors). I'm also changing the settings of driver verifier now, hopefully that'll turn the goddamn thing ON.

 

[VirGnarus]

You'll have to save settings then restart the PC successfully to lock in DV. You can verify if DV is on by going back to it via 'verifier' and looking at "display current settings". If it shows up empty, then DV ain't active. If the settings are displayed, it's on.

 

[Glader]

Just had another BSOD, this time the driver verifier should be on (ran it and checked its current settings to make sure).
It seems like as soon as I get hopeful that the problem is solved, I get hammered with BSOD's again...

 

[Glader]

Here's another one, same reason: Invalid_Kernel_Handle

 

[Glader]

Probably the last one for today. This one comes courtesy of IRQL_NOT_LESS_OR_EQUAL

 

[VirGnarus]

The ones with Invalid Kernel Handle are eyeballing Ad-aware as the culprit (its driver KernExplorer64.sys specifically) as it was trying to eyeball something that another process probably already got rid of (as in it came in late). That's the initial assumption, but I'm not sure if the object it tried to touch even existed in the first place. Either way, doesn't look like I can get enough info from just the minidumps.

The IRQL one occurred when memory written on RAM was being copied over to your paging file. When it went to notify the relevant thread/process that it finished this task, it apparently went to the wrong address and knocked on the wrong door. Dunno if it was wrong because someone else moved in or the other guy simply moved out and the house was abandoned. Can't even figure out who was supposed to live there in the first place, cuz - again - minidumps don't display that data (go figure). The address it was trying to reach looked rather legit however, so it doesn't seem like an obvious case of getting an address to the middle of nowhere.

To start off I'd recommend you take care of Ad-Aware for the time being. If you already did some thorough virus/malware scans (and I hope you have recently, as they can contribute or be the reason for the crashes), you'll want to put the realtime scanning engines on temporary hold. They often show up in crashdumps as being the cause of the crash, and often times they really aren't and often times they really are.

Oh, and DV didn't catch any of them. Strangely it doesn't appear that it's on (?), or it may not just be on any of the drivers related to the crash. Probably want to try another group. Make sure they're not MS-related of course.

 

[Glader]

Got 2 more BSOD's since last posting. I have rotated the drivers to be verified but I don't think that DV has caught anything (didn't see anything on the BS).

*EDIT* Also going to uninstall Ad-aware now to get rid of any unwanted unknowns (computer seems to be clean as far as malicious content goes).

 

[VirGnarus]

Both show up ad-aware yet again doing the same thing.

 

[nathna]

Hi Vir
i have analyze ur dump files ,the process windows destop manager (wdm.exe is getting effected due to the ati graphics driver "
Troubleshoting steps :
What is the exact error message?
Have updated all the device drivers?
DWM.exe is Desktop Window Manager.
It is a windows process which needs to be started for the aero and other features to work on windows 7.
DWM is that gives you all those pretty effects in Windows Vista or windows 7 Transparent windows, live taskbar thumbnails , and even the Flip3D switcher that you can disable and replace with Switcher.

I suggest that you check whether Desktop Service Manager is running .
To do this:
1. Click Start
2. Type "services" in the start search box.
3. Now Check status of the "Desktop Service Manager".

You may also try updating the graphic drivers and see if it makes any difference. You must be logged on as an administrator to perform these steps.
1. Open Device Manager by clicking the Start button, clicking Control Panel, clicking System and Maintenance, and then clicking Device Manager. If you are prompted for an administrator password or confirmation, type the password or provide confirmation.
2. In Device Manager, locate the Display adapter, and then double-click the device name.
3. Click the Driver tab, and then click Update Driver and follow the instructions.
Now try to follow the below provided link to Update a driver for hardware that isn't working properly and check for the issue.
http://windows.microsoft.com/en-us/windows7/Update-a-driver-for-hardware-that-isnt-working-properly
I also suggest that you update the BIOS to the latest version and check