Frequent BSODs multiple errors Dell Optiflex 745 Win 7 32b

[LoisF]

Starting around the beginning of this month, I've been having frequent BSODs and spontaneous reboots without BSODs. Files created today are attached. The event seem to occur at random, while the computer is idle or performing some background processes.

The computer is primarily used as a media server, running Sage and associated operations. Sometimes the crashes occur during media conversions, but sometimes when idle other than the usual background processes.

So far, I've replaced one of the media capture boards, updated the display driver, and checked for viruses. The problem continues. I also ran some memory, hard drive, and video diagnostics. Again, nothing helpful (i.e. the tests show no problems except that HDDScan shows my primary hard drive as having a few (but stable number over the past few days; the number is 10 hex) of UltraDMA CRC Errors.

I saw some suggestions online about disabling memory cache via the bios setup, but my bios setup does not have that as an option.

Computer info in profile; log files attached.

 

[wither 2]

Please run msinfo32. Use the File, Save As to save the file. Attach the file here.

Did you run chkdsk and sfc?

 

[LoisF]

chkdsk reported no errors.
I ran sfc (apparently I had run it before, with no errors reported - at least that's how I read the log), which has several entries like this for the run I made today:

11: Move File: Source = [l:240{120}]"\SystemRoot\WinSxS\Temp\PendingRenames\cd972241b3dcd6010154000014147415.$$_inf_wsearchidxpi_0000_2e6e3f1caf9fca20.cdf-ms", Destination = [l:152{76}]"\SystemRoot\WinSxS\FileMaps\$$_inf_wsearchidxpi_0000_2e6e3f1caf9fca20.cdf-ms"

After some Googling, I found a Microsoft article describing a procedure to direct relevant strings to a file, which is also attached. It looks like it is not reporting any errors found/fixed - so I guess all the "move file" entries don't indicate any actual repairs.

I wasn't familiar with the msinfo32 tool, so ran it with the results attached.

Thanks for any insights into the problem that these might suggest.

 

[wither 2]

I'm not ignoring you. I'm not an expert so I've had to research a lot of what all these logs and dump files are saying and it's still up in the air. It looks like you tried to run the intel processor diagnostic tool and it faulted. Did you ever get it to run? In regards to sfc /scannow, you might try rebooting and running it again. The problem with the unicode might be directly related to what the scan is saying. There's also a number of problems with time resets. I think that's related to the fact that you have a client/server connection.

 

[LoisF]

The intel tool initially stopped on a "brand string" error, which appears to be a red herring, simply indicating that my processor is on the old side (which I already knew, of course); I told it not to stop on error and reran; nothing else failed there.

I just reran sfc /scannow (as I observed I'd had yet another spontaneous reboot early this morning. SFC reported that it "did not find any integrity violations."

I posted an updated log file, since there have been more reboots and at least one more bsod since I posted my previous logs.

Where are you seeing a unicode issue?

As for time, I have my system set to (1) update time from the internet, and (2) sync system clock with the sagetv EPG server. Could these somehow be conflicting? The settings haven't changed in forever, but could one or the other time-service be failing (where it wasn't before) and perhaps causing a problem? I've just turned off the Sage timesync to see if that makes any difference.

I'm contemplating replacing the system, BUT I will need to re-use some of the cards and it would be really helpful to figure out the nature of the problem so that I don't just move it to a new system - which will be a real pain to set up because of the media server setup.

Thanks again!

 

[wither 2]

This time it's a driver issue but there's no indication in the logs of which driver it was. There was a potential issue with the graphics card in some of the logs but they never occurred when the BSOD occurred. When this type of BSOD occurs, it's best to look at the bottom of the BSOD screen to see which file/driver is causing it.

At the time of this BSOD, nothing is happening except for those time synchronizations which occur frequently.

If I had to take a guess at the driver, I would say it's the one for your graphics card. You have a legacy graphics card for which the last driver to be provided is the AMD Radeon Software Crimson Edition Beta Driver. You can get it here- Download AMD Radeon Crimson Graphics Driver 16.2.1 Beta Hotfix Non-GCN 64-bit for Windows 7 64 bit, Windows 8.1 64 bit, Windows 10 64 bit. I suggest following the instructions there for removing the old software and installing the new since Crimson replaces the old type of software (for the life of me, I can't remember what it's called.)

 

[torchwood]

Hi Lois,

been lookin at the MSInfo report.

There are multiple errors with high resource usage services, now you have VERY little spare capacity, 400mb, which isnt helping. 3 options -- get more RAM or remove non required programs, and run Diskcleanup.

I can also see you have changed multiple service startup settings, i would reset them using BlackVipers recommendations
Black Viper's Windows 7 Service Pack 1 Service Configurations | Black Viper | www.blackviper.com

Windows updates have also been failing, please run KB947821, post its report, SURT, can be found here
C >> Windows >> logs >> CBS --- SURT

https://www.microsoft.com/en-us/download/details.aspx?id=3132

it could take a while depending on what it finds


Roy

 

[LoisF]

[MENTION=77976]torchwood[/MENTION]: Thanks, I'll look into those asap. I had changed service startups in an attempt to fix the problem, but obviously that didn't work! Also, I've generally told Windows Update to download but not install updates until I say it's ok (given that I keep hearing about Windows Update tending to break more than it fixes) - but I'll take a look. I have 4gb RAM, which is all a 32-bit system can handle, afaik - and I've been pruning what's on my C-drive (except that all the updates & diagnostics I've added have been taking up space, so I'll see what I can dump).

[MENTION=467481]withe[/MENTION]r-2: I had already installed non-whql-win8.1-win7-radeon-software-crimson-relive-16.2.1-sep20 to update my video card driver; the one you pointed to appears (from its filename - haven't followed the link yet) to be for a 64-bit system (mine is 32), and also in beta - still worth looking into?

Anyway, I'll look into both of your suggestions & report back in the next day or two.
Lois

 

[wither 2]

If you installed a 32 bit version, I don't see the need for the 64 bit version. AMD is not clear if one needs to use a 32 bit or 64 bit depending on the OS and the AMD web page where I normally tell people to go to get the update no longer has the links.

 

[wither 2]

@torchwood: Thanks, I'll look into those asap. I had changed service startups in an attempt to fix the problem, but obviously that didn't work! Also, I've generally told Windows Update to download but not install updates until I say it's ok (given that I keep hearing about Windows Update tending to break more than it fixes) - but I'll take a look. I have 4gb RAM, which is all a 32-bit system can handle, afaik - and I've been pruning what's on my C-drive (except that all the updates & diagnostics I've added have been taking up space, so I'll see what I can dump).

@wither-2: I had already installed non-whql-win8.1-win7-radeon-software-crimson-relive-16.2.1-sep20 to update my video card driver; the one you pointed to appears (from its filename - haven't followed the link yet) to be for a 64-bit system (mine is 32), and also in beta - still worth looking into?

Anyway, I'll look into both of your suggestions & report back in the next day or two.
Lois

Just re-reading your post. You alluded to "in beta" above. The Beta is the only one applicable to your card. The later 16.2.1 is not correct, if that's what you installed.

 

[LoisF]

Just re-reading your post. You alluded to "in beta" above. The Beta is the only one applicable to your card. The later 16.2.1 is not correct, if that's what you installed.

The one I installed was found here:
https://www.amd.com/en/support/kb/faq/gpu-630

They also referenced a "AMD Radeon Software Crimson Edition 16.2.1 Beta Driver for non GCN products" here:

https://www.amd.com/en/support/kb/release-notes/rn-rad-win-non-gcn-16-2-1-beta

I didn't set it up because (1) it is in beta and (2) I didn't know what "non-gcn" meant... ok, so I looked it up, and apparently there are 2 kinds of card architecture - GCN and non-GCN - but no obvious way to tell which is which, since it appeared that the same card model might come in either variant (or at least that's the impression I got at AMD's site, but maybe they just didn't know how to make it clear: why would a link pertaining to my card model number offer drivers called out as non-GCN and others where GCN or non-GCN isn't mentioned? Maybe when it isn't mentioned the driver could work for either type? I did, however, just download the non-gcn 16.2.1 beta to have it handy in case it's a more appropriate driver than the one I installed a week or so ago.

One other thing I tried since my last crash was to move my swap file to a drive with more space. So far so good, but it's too soon to tell, since I just did that a day or two ago.

 

[wither 2]

Based on this- AMD ends driver support for non-GCN Radeon cards - The Tech Report and a review here- Graphics Core Next - Wikipedia , you have a non-GCN card.

I don't know how I overlooked your lack of disk space, as identified by torchwood, since I identified that to be the cause (so far) of BSODs for someone else last week but, it's a good chance that may be the cause of at least some of your problems. Obviously, if you don't make enough space, it's going to cause other problems. I remember fighting that battle years ago when disk space was tiny.

 

[LoisF]

Unfortunately, I had another crash, this time with a blue screen log. I guess I can try the beta 16-2-1 non-GCN driver to see if that makes any difference. I'm still working through the other leads from a couple days ago, but so far didn't see anything applicable on the Black Viper's site; given that I didn't mess that much with my services startups (just taking out some applications that I wasn't using anymore - after the crashes started). Any chance this is a hardware issue, and if so, which hardware?

 

[wither 2]

If you had a blue screen, you should be able to post the information. Please do.

I mentioned the graphics card because in one of your early logs, you were getting a DVD-OV error. A similar error occurred in this thread- amdkmdag DVD_OV in Event Viewer and a registry fix was identified. Elsewhere, there was a mention of capture cards such as a Hauppauge affecting this. Since you had other errors, I couldn't attribute the BSOD to this particular problem.

 

[LoisF]

I followed the thread, and found in my logs a number of "DVD-OV" events in my log, but they're all marked "information," not "error" or "critical." Does that matter? The thread referenced essentially says, to get rid of those event notifications, turn off event-logging - which clearly doesn't address whatever underlying problem might be present.

I also found in Event Viewer an ACEEvent log, with numerous events, typically occurring in batches of hundreds, occurring at one-minute intervals, like this:

0000003210: 2021-01-02 00:16:54:488 Exception <Killing CCC0 Process>: Process with an Id of 2988 is not running.
Exception Called by: ATI.ACE.MOM.Implementation.CCC0StateMachine::KillMe processID:02532 threadIDMOM_STM_DelegateThread10) domainNameMOM.exe ) assemblyNameMOM.Implementation, Version=4.5.0.0, Culture=neutral, PublicKeyToken=90ba9c70f846762e)
************************************************************************************************************************

Edit: These ran from 12:45pm 1/1 to just before the last crash around 1:43am on 1/2. They seem to be related to the ATI driver, but I'm not convinced they point to a cause of the crash, given how long they persisted beforehand. But they haven't resumed (so far) since the crash & reboot.

After the post-crash reboot, several additional events occur, which are not all the same. Here's one:

0000000042: 2021-01-02 01:47:33:432 Could not find Type [ATI.ACE.CLI.AIB.TutorialInfoCentre.Tutorial.Dashboard.DB_TutorialInfoCentre] from [CLI.AIB.TutorialInfoCentre.Tutorial.Dashboard] in assembly [CLI.AIB.TutorialInfoCentre.Tutorial.Dashboard]
Error Called by: ATI.ACE.LOG.Foundation.Services::LoadAssembly processID:03640 threadIDCCCThreadNewashboard25) domainNameCCC.exe ) assemblyNameLOG.Foundation, Version=4.5.5693.41895, Culture=neutral, PublicKeyToken=90ba9c70f846762e)
------------------------------------------------------------------------------------------------------------------------

These only started around Jan 1, and I changed display drivers over a week ago, so why are they just starting? (or perhaps the log only keeps a certain number of events before flushing itself?) Also, since I had crashes before & after updating the ATI driver, are these "information" events linked to the problem or are they just "noise"?

My latest log is attached. Thanks for your continuing help!

 

[wither 2]

The ATI messages appear to be related to Catalyst. According to the logs, you have the Catalyst driver installed. Go to Control Panel, Programs and Features, uninstall the the AMD Catalyst Install Manager. After that, I would also uninstall the driver for your graphics card and install the Beta driver.

There are no specific causes for the BSOD (other than the description of what happened) in the files you submitted. Drivers could certainly be a cause of your BSOD.

 

[LoisF]

Ok I did the uninstall/installs (it took a while and there were some crashes along the way. Now I have to wait to see what happens (longest time between crashes recently was 2-3 days). I also uninstalled a couple of other things I'm not using while I was at it.

 

[wither 2]

I'm glad you got it installed. This morning I was thinking I might have given you the wrong sequence because I didn't look it up. The alternate would have been to uninstall the driver and then, the Manager. I'll have to be more careful to check for the proper procedure in the future.

 

[wither 2]

Is your problem solved?

 

[LoisF]

Sadly, no. I thought for a few days that installing a couple of "important" Windows updates had fixed it, since it behaved for a few days, but then another crash. This is after swapping out RAM & video cards, neither of which had any effect. It had trouble installing some of the other updates that were listed as important & available.