Solved Random different BSOD's

[rogerw99]

Very long story made short. This is a replacement laptop for the one that had many issues(Alienware 17R3 Late 2016) from Dell(also repaired & recieved back on 3/7/16).
BSOD's started 3/8/16 - Critical_Structure_Corruption BSOD, Reinstalled Windows from hidden partition(No dump file available).
3/12/16 - Memory_Management BSOD, dump file shows a date of 3/9/16, wierd or I wrote it down wrong.
Uninstalled all Dell specific software, Dell Support assistant, Dell Support Agent, Dell Foundation Services, Dell Digital Delivery, & backdated the Intel Rapid Storage Technology Driver to Version 14.8.0.1042
3/15/16 - Kmode_Exception_Not_Handled.
3/19/16 - System_Thread_Exception_Not_Handled.
3/20/16 - DPC_Watchdog_Violation, loading Skyrim at the time.
I do not overclock or modify the system in any way other than adding a powered USB hub & the two portable USB HDDs that caused no problems with my older Windows 7 Alienware M17R2 2010 model. My current system specs are in my profile if needed.

I think I included everything you need. Any help is very much appreciated.

 

[axe0]

Please update Kionix KXCNL Freefall Sensor Driver, Intel Rapid Storage driver & Nvidia display drivers.
Please make sure that switching between GPU's does NOT happen
Product Support | Dell US

3: kd> lmvm nvlddmkm
Browse full module list
start             end                 module name
fffff801`06c8a000 fffff801`07769000   nvlddmkm T (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: \SystemRoot\system32\DRIVERS\nvlddmkm.sys
    Image name: nvlddmkm.sys
    Browse all global symbols  functions  data
    Timestamp:        [COLOR="#FF0000"]Tue Oct 27 21:51:13 2015[/COLOR] (562FE3C1)
    CheckSum:         00AA09ED
    ImageSize:        00ADF000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

7: kd> lmvm iastora
Browse full module list
start             end                 module name
fffff800`75a9a000 fffff800`7600e000   iaStorA  T (no symbols)           
    Loaded symbol image file: iaStorA.sys
    Image path: iaStorA.sys
    Image name: iaStorA.sys
    Browse all global symbols  functions  data
    Timestamp:        [COLOR="#FF0000"]Wed Nov 04 11:27:49 2015[/COLOR] (5639DDA5)
    CheckSum:         00168F4C
    ImageSize:        00574000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

7: kd> lmvm kxdiskprot
Browse full module list
start             end                 module name
fffff800`757d0000 fffff800`757de000   kxdiskprot T (no symbols)           
    Loaded symbol image file: kxdiskprot.sys
    Image path: kxdiskprot.sys
    Image name: kxdiskprot.sys
    Browse all global symbols  functions  data
    Timestamp:        [COLOR="#FF0000"]Sun Sep 27 20:50:01 2015[/COLOR] (56083A59)
    CheckSum:         0000D16E
    ImageSize:        0000E000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

 

[rogerw99]

Updates.

Kionix Driver - Forgot to mention I had it disabled in Device Manager because an older driver caused Windows to take approx 3-4 minutes to boot to desktop. Now updated with ver 1.1.4.0 ,A01, 21 Oct 2015.
Intel Rapid Storage Driver - Also forgot to mention that I already had the very latest driver installed from Intel directly because the current latest from Dell causes massive iastora.sys BSOD's. ver 14.8.0.1042, 12/14/2015.
nVidea Driver - Downloaded and installed the latest from Dell, 10.18.13.5940 ,A03, 9 Mar 2016.

Switching between video card usage - There is no way that I know of to prevent Windows from booting up using the onboard Intel 530 Video built into the CPU itself. I cannot disable it in BIOS/UEFI. I have little to no control over most BIOS settings. I have not tried to disable it in Device Manager. I was afraid of having no video at all in Windows. Most of the time the switch to the 980M is automatic & if it isn't, I set the nVidia Manage 3D Settings to use of the 980M for that game. Global Settings are also set for the 980M. But Windows continues to use the Intel HD Graphics 530.

Waiting to see if there are any changes to BSOD's.

 

[rogerw99]

Another new different BSOD.

This time it said IRQL_NOT_LESS_OR_EQUAL caused by ntoskrnl.exe+14e3a0.

 

[axe0]

caused by ntoskrnl.exe+14e3a0.

This file is the last one to blame when nothing else is available to blame.

Diagnostic Test

RAM TEST

Run MemTest86+ to analyse your RAM. MemTest86+ - Test RAM - Windows 10 Forums

   Note

MemTest86+ needs to be run for at least 8 complete passes for conclusive results. Set it running before you go to bed and leave it overnight. We're looking for zero errors here. Even a single error will indicate RAM failure.

Addition:
If errors show up you can stop the test, remove all sticks but 1 and test this single stick in each slot for 8 passes or until errors show, switch sticks and repeat.
If errors show up and you see them a lot later, no problem, the errors don't affect the test.

 

[rogerw99]

Another squared new different BSOD.

Ok, passed all 8 passes with memtest86+ using bootable USB. And here is another dump file. This one was a BAD_POOL_CALLER caused by ntoskrnl.exe+14e3a0. All the BSOD's except one were caused by ntoskrnl.exe+14e3a0. When you say that this file is to blame, what does that mean I have to do?

 

[axe0]

The ntoskrnl.exe file is the Windows kernel, it basically is never the cause but blamed because nothing else is there to be blamed.
If the ntoskrnl.exe file is really the cause, one would experience a lot of more problems than only BSOD's.

And here is another dump file

Try to upload it again, I don't see a new upload.

 

[rogerw99]

Sorry forgot to attach the file.

Sorry forgot to attach the file. I edited my last entry and added the file. I'll put it in this one as well.

 

[axe0]

Diagnostic Test

DRIVER VERIFIER

   Warning

Please make a backup of your important files and get your rescue media or create one.
Please create a restore point.

Please follow this tutorial to run driver verifier.


Driver verifier stresses your drivers and will crash your pc if any driver fails due to a violation.


Driver verifier should be performed for a max of 48 hours, or when you have a bluescreen, whatever comes first.


If driver verifier has found a violation and you can't get back into windows normally, try to boot into safe mode and reset in safe mode driver verifier, or in the troubleshooting options open command prompt and type verifier /reset.

   Note

Your system will act very sluggishly while driver verifier is enabled, this is normal as your drivers will be being subjected to heavy testing in order to make them crash.

 

[rogerw99]

Ran Verifier, No issues arose.

Verifier cause no issues for about 45 hours. However before doing that I had a BSOD caused by my last nVidia driver. I then updated to the latest driver direct from nVidia since Dell had nothing as new/newer. So far, no BSOD's since Saturday the 26th at noon. But here is the dump file just in case you wish to look it over.

Thanks

 

[axe0]

Alright, until you have new BSOD's I won't post suggestions since it's been a few days

Just to let you know, I haven't yet seen a cause but did find pool page corruption although I can't find much more due to some limitations
To explain what pool page means (though this might be a bit technical):

• The memory manager of Windows creates the following memory pools that the system uses to allocate memory: nonpaged pool and paged pool.
• The paged pool consists of virtual memory that can be paged in and out of the system.
• The nonpaged pool consists of virtual memory addresses that are guaranteed to reside in physical memory as long as the corresponding kernel objects are allocated.

1: kd> !poolval ffffc0011dd6f010
GetPointerFromAddress: unable to read from fffff8032a5e5138
GetUlongPtrFromAddress: unable to read from fffff8032a5e5298
GetUlongPtrFromAddress: unable to read from fffff8032a5e5520
Pool page ffffc0011dd6f010 region is GetPointerFromAddress: unable to read from fffff8032a5e5138
GetUlongPtrFromAddress: unable to read from fffff8032a5e5298
GetUlongPtrFromAddress: unable to read from fffff8032a5e5520
Paged pool

Validating Pool headers for pool page: ffffc0011dd6f010

Pool page [ ffffc0011dd6f000 ] is __inVALID.

Analyzing linked list...
[ ffffc0011dd6f000 ]: invalid block size [ 0x0 ] should be [ 0x8d ]


Scanning for single bit errors...

None found

 

[rogerw99]

Good news and Bad news.

Thanks for your help in all this Axe0,
The good news, sort of, understood on page pooling. I'm a fight simulator technician that does troubleshooting & repair down to component level on circuit boards on many systems as well as hydraulics, mechanical & analog systems. However, I am a long way from ever being a programmer. That takes an entirely different mindset than what I was trained to do. But I do understand a good portion of it & appreciate the help your giving me. Debugging? Forget it.
The bad news,
Had another BSOD. This time while was responding to an email response to Dell Tech Support(using outlook web mail) on which they are woefully far behind on figuring out that many of their drivers/software were the cause of my multiple daily BSOD's after I informed them of such. Any way, I had just hovered over the dell support drop down and had just clicked on support by product when it happened, System_Service_Exception caused by afd.sys+46a78, hal.dll+669a and ntoskrnl.exe+159ee9. I just don't get why I am having so many BSOD's. And all are non-repeating except for once. Could this be a case of very poorly designed hardware by Dell/Alienware with major incompatibility issues here? I would not find this difficult to believe. This is my third(second replacement) system. The first was a 17R2, then a 17R3 and this 17R3 with updated nVidia graphics memory. All have had exactly the same or similar issues.

 

[axe0]

Please disable either of the GPU's in device manager as a test.

The drivers of both your GPU's have been flagged and the 0x3B is mostly caused by graphics drivers.

ffffd000`29f876d8  fffff801`d162a469Unable to load image igdkmd64.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for igdkmd64.sys
*** ERROR: Module load completed but symbols could not be loaded for igdkmd64.sys
 igdkmd64+0x346469

ffffd000`29f883d0  fffff801`d038bd5fUnable to load image nvkflt.sys, Win32 error 0n2
*** WARNING: Unable to verify timestamp for nvkflt.sys
*** ERROR: Module load completed but symbols could not be loaded for nvkflt.sys
 nvkflt+0x9d5f

 

[rogerw99]

Tried it...

Ok, I disabled the cpu built-in Intel 530 in device manager and it would not completely disable. It was still running the Windows desktop and ran any game I loaded(& ran badly). I also could no longer access the nVidia control panel. Error said that the nVidia graphics card was not attached to a dispaly. The nVidia being disabled gave essentially the same results but was fully disabled. There does not seem to be any way to disable the Intel Graphics even in BIOS. Dell provided no means to do so. That would have been my preference from the beginning. So I have returned everything back to default. The only reason I bought this was for mobile gaming.

 

[axe0]

Dell laptops have extremely limited options in the BIOS yes.
Lets try a GPU stress test on both GPU's.

Diagnostic test

GPU TEST

Run Furmark to test your GPU. Furmark tutorial

   Note

Run Furmark for around 30 minutes

   Warning

Your GPU temperatures will rise quickly while Furmark is running. Keep a keen eye on them and abort the test if temperatures rise too high

 

[rogerw99]

Ok, ran it & two instances of GPUZ, one for each GPU, for 30min. It tested only the 980m(61c max), no data on the HD530 in-program other than it existed. GPUZ showed HD530 @ ~50c probably due to CPU usage. To test the HD530, I had to use the nVidia Control Panel/Manage 3D Settings/Program Settings tab & set the FurMark program to use "Integrated Graphics". GPUZ reported a max of 56c for the HD530. Neither test crashed the laptop or showed any anomalies in rendering, though the HD530 ran poorly as expected.

 

[axe0]

A bit unusual from what I normally suggest, but try to run Furmark on either GPU's for ~1-2 hours.

 

[rogerw99]

Tried it, no change.

Sorry it took so long to get back. I tried it for two hours each. The only difference was a couple of degrees hotter for each GPU. No new BSODs. What makes me wonder is the complete randomness & almost always a different apparent cause.

 

[axe0]

What I've currently seen mostly with your crashes is the graphics cards being a pain.

adding a powered USB hub & the two portable USB HDDs

Are these connected when BSODs occur?

 

[rogerw99]

Yes & no

They have been connected most of the time when the BSODs occur because I use them extensively. But I have disconnected them while troubleshooting this issue with you at times & have had a couple of BSODs even then. The USB HDDs were used on an older previous Alienware Windows 7 laptop with no issues. This is my first Windows 8.1 system. The powered USB Hub is new & was plugged into the USB port with Power Share(for charging other USB devices even when the laptop is off, which I don't do). I just now moved it to a standard USB port to see if there is any change. Note that both of those ports are from a daughter card connected to the motherboard with a ribbon cable. I have one other USB type A port(directly mounted on the motherboard) & a Thunderbolt 3 (USB type C, USB 3.1 Gen2) port. The portable HDDs are a 2TB My Passport 0748 & a 2TB Seagate BUP Slim SL. Almost forgot, the Laptop Cooler Stand is USB powered & plugged into the powered USB Hub. My wireless mouse (Logitech m325) is plugged into the MB mounted USB port because if I plug it in to any other available port there is a noticeable lag or even a momentary pause in movement. I have not tried plugging the powered USB Hub into the MB mounted USB port & moving the mouse to it. Inconvenient but doable.