Windows 10 Frequent, Semi-random BSODs (Thread stuck in device driver)

[SpaceKitkat]

As the title says, almost* every time I launch a video game (specifically, a minecraft mod), I get a BSOD with the error "THREAD_STUCK_IN_DEVICE_DRIVER".
I'm also getting them more and more frequently just using google chrome, even just now while trying to get minidump log.
Checking the log, there was something about atikmdag.sys, and my problem seems to be very similar to this thread:
However, that thread's main solution was when someone checked the logs, they listed drivers that needed to be removed, and I don't think I have any of the same drivers.

After a bit of googling, a couple "easy" solutions were to update my BIOS and reinstall GPU drivers.
So I uninstalled all of my display drivers, deleted everything AMD, used DDU (Display Driver Uninstaller), restarted into my BIOS, updated my BIOS successfully, installed the new GPU drivers (Radeon Crimson edition software, for an R9 290), and restart my computer.

I've also performed several Windows diagnostics, including a RAM check, disk checks on all of my hard drives, and a "sfc /scannow" via CMD. All have come up with no errors.
I tried to add the minidump W7F log, but it said the file size was too large.. lol (RIP).
I've uploaded it to
but if it needs to be uploaded elsewhere, just let me know.
If any other logs are needed, I'll supply them ASAP. Any help is very much appreciated, as I work from home and this is currently my only available computer!

 

[nmsuk]

This will explain driver verifier

 

[sven31415]

I guess if you say, I feel it is in the hardware means you think the bahaviour is unique to my special R9 290x board right ?

 

[nmsuk]

It's odd yes as I ran a r9 290 and a 390 for most of this year using all the drivers as they came out with no problems at all. If the problem was happening to many people and it was the drivers it'd be fixed by now.

 

[sven31415]

I have a new theory. I think they changed the power managment from 15.12 to 16.1.1 and higher to fix this:

• [81856] Marginally increased power consumption may be observed during video playback

This is a listed issue on the 15.12 driver. It does not have a resolved issue notice on 16.1.1 but it is not on the know issues list any more.
They also changed the format of the issues list as if they won't expose everything anymore.

I feel that the fan is going more for video playback than for the higher drivers.

 

[kemical]

What is the driver verifier ?

When a driver crashes, which causes a bsod, one usually updates the driver and all is well again. Sometimes however updating the driver doesn't seem to work. and it could be that it's another driver causing the gpu driver to crash. The Driver verifier will, on start up, stress each driver in turn to see if a bsod occurs or not. If the verifier does cause a bsod then the offending driver should be captured by the dump file.

However.

The verifier can sometimes create a bsod start up loop.

This can be stopped by turning off the verifier although you'll need to enter safe mode in order to do so. As your probably aware Windows 10 safe mode is harder to get to unless your already in the system.
To enter safe mode you'll need to boot from a Link Removed, enter the recovery console and then safe mode.

The following guide will show you how to set the verifier up:
Driver Verifier - Windows 10, 8.1, 8, 7 & Vista (BSOD-related) - Tech Support Forum

Run for 24hrs only.

Post any resulting dump files.

 

[SpaceKitkat]

I didn't have a BSOD for about 2 days after I tried the Radeonmod registry tweak, however, I've gotten two in the last day. I can confirm it does not fix it.
I'm probably going to roll back my drivers to 15.11, and just deal with the lower optimization/FPS of certain games. It's not worth an FPS boost for BSODs every other day.

@nmsuk I've gone almost a week at a time without having BSODs, but they do come back. If you're serious about testing it, try testing it for a minimum of a week. Two would be better, as none of us in this thread have had an entire week free of BSODs (that have R9 290/x's)

 

[nmsuk]

I ran my r9 cards up until a fortnight ago without problem. Which is why I'm thinking this is hardware and not the drivers. That or you have another driver or piece of hardware conflicting with amd's drivers.

 

[SpaceKitkat]

I ran my r9 cards up until a fortnight ago without problem. Which is why I'm thinking this is hardware and not the drivers. That or you have another driver or piece of hardware conflicting with amd's drivers.

It must be a very common piece of hardware that's conflicting, if it is a hardware conflict. There are many, many more threads from people with r9 290s that are getting this exact same problem, all over the internet.

@sven31415 @Circinae Do either of you use more than one monitor? I'm currently using two, with the extended display. I remember skimming someone who posted about the same issue and they mentioned something about the windows' extended display causing issues, though I don't remember if it's related to the BSODs or not.

Edit: Testing out Clockblocker now. Included my most recent minidump, just for record's sake.

 

[kemical]

Included my most recent minidump, just for record's sake

Another bugcheck EA I'm afraid.

 

[sven31415]

Hello Spacekitcat,

good to hear from you.
The crashes for me have been more than for you. I think once per day. They are mostly during light load like video playback or load changes (game exit.). I never had a crash during playing a game. Or they were on idle screen.
Sometimes I have the strange feeling the crashes are somehow related to user interaction as i let youtube run the whole night and nothing bad happened and then i sat down at the computer and boom BSOD.

I isolated the problem to be like this:
Every driver after 16.1.1 seems to cause the trouble.
Driver 15.12 (and probably lower) works.
Internal Intel on board GFX works, even if the Radeon is still in the system but deactivated.
I did not test ClockBlocker for long timem just 20 minutes, but Circinae did and he said it worked.
The drawback of ClockBlocker for me was ~62°C GPU temperatur while the fan was running at 42% on
the Windows desktop.

My guess is that the BSOD problem is related somehow to power managment as there is an open issue in 15.12 that is:

• [81856] Marginally increased power consumption may be observed during video playback

This issue does not appear on the issue list on 15.12. I guess they made that dynamic clocking more aggressive and this does not seem to work on all board the whole time.
But this is just a guess.

I relayed all that information to my AMD ticket and asked if they have some registry switches to disable the new power management in newer driver to test if the feature is responsible for the problem.

I have single monitor setup

BR

Sven

 

[nmsuk]

Just a thought. Disable hardware acceleration in your browser and see how that goes with video playback

Sent from my Nexus 6 using Windows Forums mobile app

 

[sven31415]

And the crashes on idle PC and game exit ?

 

[kemical]

I relayed all that information to my AMD ticket and asked if they have some registry switches to disable the new power management in newer driver to test if the feature is responsible for the problem.

I'll be very interested on what they have to say (if they reply).

Personally i still feel it's either the card or something else your running causing the card to TDR and then eventually BSOD.

 

[sven31415]

Hello kemical: How would you explain ClockBlocker to prevent the crashes then ?

 

[Circinae]

Hello kemical: How would you explain ClockBlocker to prevent the crashes then ?

It is absolutely the drivers and the clock speed control. Older drivers stop the problem. Clock blocker stops the problem. It is consistent, easy to reproduce and occurs across systems for a specific GFX card edition. You can even monitor the clock speeds to watch when it triggers.

I very much doubt, however, that AMD will do anything about it. It's an older model and as I mentioned earlier in the thread, there have been lengthy threads on their sub-reddit/forums that have discussed the issue without any inclination shown from AMD to remedy it. I originally only popped in to lend some prior knowledge of the matter, but am personally glad that this thread helped me find a better solution than older drivers, so for that, thank you folks.

 

[SpaceKitkat]

No offense @nmsuk but if you had read our previous posts, you'd know I've already tried disabling hardware acceleration, as well as having extensively tested each web browser (Chrome, Firefox, and Edge). This morning's BSOD was just from loading up a game and alt tabbing so that I could launch chrome and look something up. The trigger for the BSODs is almost *always* alt-tabbing. I cannot recall ever getting a BSOD in the last few months that wasn't related to alt tabbing, so I'm inclined to agree with Sven, it's an issue with something relating to switching between two different settings in the drivers provided by AMD. It's possible it could be a power management thing, or a clock speed difference between the 2D and 3D settings, as said by the user in this thread. I have not had a BSOD since using Clockblocker, however, it is much too early to tell.

All that said, with Clockblocker the temperatures for my GPU are slightly higher than idle (55-60), but they definitely do not reach their maximum (without Furmark). Tested while gaming and while just browsing the internet/watching youtube (although, like I previously stated, it hasn't even been a day.)

 

[sven31415]

I guess you guys have cards with more potent coolers. Mine has a stock cooler and it already kind a roars with ClockBlocker in Desktop idle mode.

 

[SpaceKitkat]

Hmm. Maybe my clockblocker isn't even working, and I just haven't gotten a BSOD, coincidentally? lol. I don't know. Mine has stock cooling, AFAIK.

 

[sven31415]

Hmm. Maybe my clockblocker isn't even working, and I just haven't gotten a BSOD, coincidentally? lol. I don't know. Mine has stock cooling, AFAIK.

Hello Space,

which settings did you use ? Which version of ClockBlocker ?

 

[kemical]

It is absolutely the drivers and the clock speed control. Older drivers stop the problem. Clock blocker stops the problem. It is consistent, easy to reproduce and occurs across systems for a specific GFX card edition. You can even monitor the clock speeds to watch when it triggers.

I very much doubt, however, that AMD will do anything about it. It's an older model and as I mentioned earlier in the thread, there have been lengthy threads on their sub-reddit/forums that have discussed the issue without any inclination shown from AMD to remedy it. I originally only popped in to lend some prior knowledge of the matter, but am personally glad that this thread helped me find a better solution than older drivers, so for that, thank you folks.

I have been looking around the net and yes there is indeed a number of users with this issue. I've only seen later cards with this problem so perhaps it's something they introduced for the latest cards but conflicts badly with older cards? If you guy's know of any contemporary cards with this issue please let me know.

I've advised another user to try the drivers without the CCC as well as checking various parts of the pc if only to rule them out.