AMD Driver Failure on RX 480 8GB (17.2.1)

[StayFrostyy]

For whatever reason, my graphics card drivers have been malfunctioning since yesterday evening. I'm unsure what may have caused this to begin with. Basic video playback works fine, but just a few minutes into playing Overwatch (or any game, for that matter) and I will get a blue screen and be forced to restart. The GPU shouldn't be overheating, as it hardly has time to get up to even 65 degrees before blue-screening. Overwatch hardly takes more than 1GB of VRAM at any given time, either. It should not be a hardware issue (e.g. overheating, lack of VRAM, etc.)

 

I have used WhoCrashed to give those willing to help me an idea of what's going on. Here are the memory dumps and an analysis of each.

Spoiler

On Mon 2/27/2017 10:11:49 PM your computer crashed
crash dump file: C:\Windows\Minidump\022717-11512-01.dmp
This was probably caused by the following module: ntoskrnl.exe (nt+0x70400)
Bugcheck code: 0xA (0x48, 0x2, 0x0, 0xFFFFF80002ECCC79)
Error: IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\system32\ntoskrnl.exe
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: NT Kernel & System
Bug check description: This indicates that Microsoft Windows or a kernel-mode driver accessed paged memory at DISPATCH_LEVEL or above.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in the Windows kernel. Possibly this problem is caused by another driver that cannot be identified at this time.

 

At first, I assumed it was a driver issue and simply reinstalled my drivers. However, only a few minutes after I reinstalled the drivers, I got another blue screen.

 

On Mon 2/27/2017 10:59:05 PM your computer crashed
crash dump file: C:\Windows\Minidump\022817-11029-01.dmp
This was probably caused by the following module: atikmdag.sys (atikmdag+0x10B42EE)
Bugcheck code: 0x3B (0xC000001D, 0xFFFFF8800FED22EE, 0xFFFFF88007F54830, 0x0)
Error: SYSTEM_SERVICE_EXCEPTION
file path: C:\Windows\system32\drivers\atikmdag.sys
product: ATI Radeon Family
company: Advanced Micro Devices, Inc.
description: ATI Radeon Kernel Mode Driver
Bug check description: This indicates that an exception happened while executing a routine that transitions from non-privileged code to privileged code.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: atikmdag.sys (ATI Radeon Kernel Mode Driver, Advanced Micro Devices, Inc.).
Google query: Advanced Micro Devices, Inc. SYSTEM_SERVICE_EXCEPTION
 

This afternoon I reinstalled the drivers again, but seemingly to no avail.

 

On Tue 2/28/2017 4:03:40 PM your computer crashed
crash dump file: C:\Windows\Minidump\022817-16910-01.dmp
This was probably caused by the following module: msahci.sys (0xFFFFF88000E34B54)
Bugcheck code: 0xD1 (0xFFFFFFFF8B481275, 0xB, 0x0, 0xFFFFF88000E34B54)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\system32\drivers\msahci.sys
product: Microsoft® Windows® Operating System
company: Microsoft Corporation
description: MS AHCI 1.0 Standard Driver
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
The crash took place in a standard Microsoft module. Your system configuration may be incorrect. Possibly this problem is caused by another driver on your system that cannot be identified at this time.
 

Note that every crash occurred while playing Overwatch, and that no two blue screens share the exact same problem.

 

Is this a common problem and does anyone have any idea how to fix it?

[FiveNine]

Does it happen in other games? And have you tried an older driver version?

[DragonTamer1]

I suggest rolling back your drivers to the last version. AMD has been doing well with their drivers but it is still possible that sometimes bugs get through. The only real reason to update is if you need the driver to play a new game or there was an optimization for one of your existing games or your system is generally running poorly.

[StayFrostyy]

@DragonTamer1 @FiveNine I went into Counter-Strike: Global Offensive a few minutes ago and had the same issue occur. I then rolled my drivers all the way back to 16.11.2 manually (there was no rollback option available) and I'm still having issues.

[ivan134]

24 minutes ago, StayFrostyy said:

@DragonTamer1 @FiveNine I went into Counter-Strike: Global Offensive a few minutes ago and had the same issue occur. I then rolled my drivers all the way back to 16.11.2 manually (there was no rollback option available) and I'm still having issues.

Take out your GPU and use your iGPU and see if you still have problems.

[stconquest]

Chances are that it is the RAM.  If you have two sticks, take one out and use the PC.  If it crashes right away, switch the RAM stick out and try again.

 

It might be a storage device, but I would bet on some bad RAM sectors.

 

I have had the exact same symptoms on two separate machines with completely different hardware.  In both instances it was bad RAM.  The older PC RAM is dead and sitting in a drawer, and the newer PC's kit was RMA'd for a working kit.

[StayFrostyy]

2 hours ago, stconquest said:

Chances are that it is the RAM.  If you have two sticks, take one out and use the PC.  If it crashes right away, switch the RAM stick out and try again.

 

It might be a storage device, but I would bet on some bad RAM sectors.

 

I have had the exact same symptoms on two separate machines with completely different hardware.  In both instances it was bad RAM.  The older PC RAM is dead and sitting in a drawer, and the newer PC's kit was RMA'd for a working kit.

I ran both sticks by themselves and the computer booted just fine either way. I also ran chkdsk on my HDD to ensure that storage wasn't the problem. It wasn't.

Fortunately (and unfortunately), the computer simply reboots during my game, rather than throwing up an error from the game or driver or spitting out a BSoD.

[stconquest]

2 minutes ago, StayFrostyy said:

I ran both sticks by themselves and the computer booted just fine either way. I also ran chkdsk on my HDD to ensure that storage wasn't the problem. It wasn't.

Fortunately (and unfortunately), the computer simply reboots during my game, rather than throwing up an error from the game or driver or spitting out a BSoD.

Did you run a single stick for a good while?  Long enough for a crash to be likely to occur?

 

Bad RAM will boot just fine.  It will even function normally for a while... until it tries to use a bad sector.

[StayFrostyy]

4 minutes ago, stconquest said:

Did you run a single stick for a good while?  Long enough for a crash to be likely to occur?

 

Bad RAM will boot just fine.  It will even function normally for a while... until it tries to use a bad sector.

So I should attempt to play Overwatch or CS:GO with a single 4GB stick?

[StayFrostyy]

@stconquest I should also note that this set of RAM refused to run in dual-channel mode on my motherboard. The system refused to start until I put them in single-channel. I suspected that the motherboard was at fault but with this information it could very well be the RAM.

[stconquest]

2 minutes ago, StayFrostyy said:

So I should attempt to play Overwatch or CS:GO with a single 4GB stick?

Yes.

1 minute ago, StayFrostyy said:

@stconquest I should also note that this set of RAM refused to run in dual-channel mode on my motherboard. The system refused to start until I put them in single-channel. I suspected that the motherboard was at fault but with this information it could very well be the RAM.

hmmm

[StayFrostyy]

@stconquest I'll go ahead and do that now. I'll report back when I'm done with info.

[stconquest]

12 minutes ago, StayFrostyy said:

@stconquest I'll go ahead and do that now. I'll report back when I'm done with info.

You will also want to test each DIMM slot on the mobo in the same manner.  I know it may sound tedious, but will help detect any issue with the motherboard.

[StayFrostyy]

8 minutes ago, stconquest said:

You will also want to test each DIMM slot on the mobo in the same manner.  I know it may sound tedious, but will help detect any issue with the motherboard.

I will definitely do that, but that is a task for another day.

 

I tested the RAM that was originally in Slot 1 by itself and I played a full round of Casual CS:GO. The length of the game was long enough to let me know that the system wouldn't fail, as the game would ordinarily shut down within 10 minutes (usually closer to 2-5 minutes.

 

Testing the other stick now.

[stconquest]

1 minute ago, StayFrostyy said:

I will definitely do that, but that is a task for another day.

 

I tested the RAM that was originally in Slot 1 by itself and I played a full round of Casual CS:GO. The length of the game was long enough to let me know that the system wouldn't fail, as the game would ordinarily shut down within 10 minutes (usually closer to 2-5 minutes.

 

Testing the other stick now.

Hopefully this one fails quickly.  I can't believe I said that. 

[StayFrostyy]

15 minutes ago, stconquest said:

Hopefully this one fails quickly.  I can't believe I said that. 

Unfortunately, it didn't. It played through a full round. However, framerates were consistently lower across the board and I was getting stuttering left and right.

[stconquest]

23 minutes ago, StayFrostyy said:

Unfortunately, it didn't. It played through a full round. However, framerates were consistently lower across the board and I was getting stuttering left and right.

It should be a straight crash, BSOD.  I would try each one longer, until a crash occurs even if it takes a day.  If no crashes occur after many hours of utilizing the RAM, then switch again to see if the other stick fails.

 

One indicator for me was constant browser crashes (Firefox).  I noticed it on my old system, and when my newer system had one of the sticks (2x8GB) start failing... I noticed frequent Firefox crashes with BSODs mixed in.

 

The BSOD codes were the same as yours, both the IRQ one and the SSE one.  Same exact files:  ntoskrnl.exe and atikmdag.sys.

 

But not this one:  msahci.sys.  I don't think I had that one... maybe a Win10 thing, IDK.  You know what?  It might be your SSD/HDD (AHCI):  https://en.wikipedia.org/wiki/Advanced_Host_Controller_Interface

 

You can try updating the firmware for your SSD/HDD maybe?

 

@StayFrostyy  I added the part about the SSD/HDD/RAM host controller... worth looking into.  I still think the RAM is likely the problem, but that is because I have had it be the RAM on two separate PCs.  Never had an major issue with the host controller (mobo) or storage.

[FiveNine]

5 hours ago, StayFrostyy said:

Unfortunately, it didn't. It played through a full round. However, framerates were consistently lower across the board and I was getting stuttering left and right.

Have you tried using the iGpu like the guy upstairs suggested? I'm curious about the dual channel not working, were the ram sticks bought separately or just mobo doesn't support it? Another option could be if you have a friend with a pc, to just put the gpu in his system and test it. If it doesn't crash, you can rule it out

[StayFrostyy]

17 hours ago, stconquest said:

It should be a straight crash, BSOD.  I would try each one longer, until a crash occurs even if it takes a day.  If no crashes occur after many hours of utilizing the RAM, then switch again to see if the other stick fails.

 

One indicator for me was constant browser crashes (Firefox).  I noticed it on my old system, and when my newer system had one of the sticks (2x8GB) start failing... I noticed frequent Firefox crashes with BSODs mixed in.

 

The BSOD codes were the same as yours, both the IRQ one and the SSE one.  Same exact files:  ntoskrnl.exe and atikmdag.sys.

 

But not this one:  msahci.sys.  I don't think I had that one... maybe a Win10 thing, IDK.  You know what?  It might be your SSD/HDD (AHCI):  https://en.wikipedia.org/wiki/Advanced_Host_Controller_Interface

 

You can try updating the firmware for your SSD/HDD maybe?

 

@StayFrostyy  I added the part about the SSD/HDD/RAM host controller... worth looking into.  I still think the RAM is likely the problem, but that is because I have had it be the RAM on two separate PCs.  Never had an major issue with the host controller (mobo) or storage.

I ran my system through a chkdsk and it found absolutely nothing wrong. So, I doubt it is the HDD.

 

However, both sticks of RAM crashed the PC within a few minutes of a game of Overwatch.

[stconquest]

Just now, StayFrostyy said:

I ran my system through a chkdsk and it found absolutely nothing wrong. So, I doubt it is the HDD.

 

However, both sticks of RAM crashed the PC within a few minutes of a game of Overwatch.

Each one, separately?

 

Darn.  If it is the motherboard, that is a bigger problem.  AsRock is not known for their customer support.  I have nothing against them, I am using an AsRock board right now.

 

You can try what Ivan suggested by removing the GPU and testing the PC with only the iGPU.

[StayFrostyy]

5 minutes ago, stconquest said:

Each one, separately?

 

Darn.  If it is the motherboard, that is a bigger problem.  AsRock is not known for their customer support.  I have nothing against them, I am using an AsRock board right now.

 

You can try what Ivan suggested by removing the GPU and testing the PC with only the iGPU.

I can't even install the iGPU drivers, so it doesn't look like I can try that.

 

I'd like to note that the problem isn't the GPU anymore, seemingly. The drivers were definitely screwed with somehow, but when I reverted to 16.12.2, my texture load times halved and my FPS went up by a solid 30 frames. No stuttering whatsoever, either.

 

I have a 550W SeaSonic power supply, so everything should be okay there.

 

My guess is ram at this point.

[stconquest]

4 minutes ago, StayFrostyy said:

I can't even install the iGPU drivers, so it doesn't look like I can try the former.

 

I'd like to note that the problem isn't the GPU anymore, seemingly. The drivers were definitely screwed with somehow, but when I reverted to 16.12.2, my texture load times halved and my FPS went up by a solid 30 frames. No stuttering whatsoever, either.

 

I have a 550W SeaSonic power supply, so everything should be okay there.

 

My guess is ram at this point.

It could the the host controller (AHCI) on the motherboard.  It affects the RAM and storage and stuff.

 

You could buy a single 8GB stick to test.  If it is the 2x4GB stick that is the problem, then you can use all 16GB when you RMA the bad kit.

 

@StayFrostyy  Have you changed any BIOS settings?  If yes, you might want to clear CMOS to reset the bios.

[StayFrostyy]

Just now, stconquest said:

It could the the host controller (AHCI) on the motherboard.  It affects the RAM and storage and stuff.

 

You could buy a single 8GB stick to test.  If it is the 2x4GB stick that is the problem, then you can use all 16GB when you RMA the bad kit.

My dad just ordered me a new 16GB kit; that should fix any RAM issues.

 

As for the AHCI, how would I go about checking on that or fixing it?

[stconquest]

1 minute ago, StayFrostyy said:

My dad just ordered me a new 16GB kit; that should fix any RAM issues.

 

As for the AHCI, how would I go about checking on that or fixing it?

Trial and error/deduction.  If everything else is likely fine, the mobo is the likely problem.  You would have to contact AsRock and tell them your story in a polite manner.  End with asking if there is anything they can do for you... to help you out.

 

Do you have another PC to test the RAM out in?  Do you have another HDD or SSD to test out?

[StayFrostyy]

Just now, stconquest said:

Trial and error/deduction.  If everything else is likely fine, the mobo is the likely problem.  You would have to contact AsRock and tell them your story in a polite manner.  End with asking if there is anything they can do for you... to help you out.

 

Do you have another PC to test the RAM out in?  Do you have another HDD or SSD to test out?

I wouldn't doubt that the motherboard is the issue, as I haven't been able to run dual channel since I've gotten it. The RAM was pulled from another system that I had been using for a year and had absolutely zero errors or blue-screens about anything, let alone memory. It was also running in perfectly-functional dual channel in my last system.