Random BSOD with brand new build

[farmerdils]

Greetings.

I'm getting random BSODs with a brand new build. I'm kind of a computer noob so I'm not entirely sure what to do. Three BSOD messages have been appearing.

Memory_management
page_fault_in_nonpaged_area
system_exception

The weird thing is that the BSOD occurs once and then my computer runs fine. Also worth noting that I had my hard drive plugged in when I installed my Windows OS, so although my OS is on my SSD, there is a system reserved partition on my HDD (500mb). I'm not sure if it is relevant or not, but there you go.

Edit: I guess you're going to need my specs.
OS: Windows 10 Home
Intel i7 6700k
Corsair h75 AIO
Gigabyte GAZ170MX Gaming 5
GeIL Potenza 16GB DDR4-2133
MSI GTX Geforce 1070
EVGA NEX750W 80+ Gold

Any help would be appreciated, thank you!

 

[farmerdils]

Updated the firmware. If I get a crash tomorrow, I will try to switch around my USB ports. I googled around some and disabled fast boot for Windows 10, apparently that was causing problems as well for some users. Will report back with results.

 

[farmerdils]

Something new has happened! I have driver verifier running and was playing Battlerite, and got a BSOD that I have never gotten before. APC_INDEX_MISMATCH.

Minidump: https://1drv.ms/u/s!AhcS4hvw2pjjgiEwZJF9QrHtrckt

Will update with memory.dmp if needed.

 

[johnbl]

the working set list is also corrupted in the memory dump file.
this is part of the mapping table between your pagefile.sys and where data is paged into ram.

you might want to run the intel processor diagnostics tool
-----------
one of the internal error logs for the usb 3 returned a error STATUS_NOT_SUPPORTED
the last command was still pending it was:
31: RootHub_Pdo_EvtDeviceEnableWakeAtBus - Controller 0xFFFFC80C387A7950 RootHub_Pdo_EvtDeviceEnableWakeAtBus

it looks like you have microsoft generic drivers for the usb 3 installed, it could be hitting a bug with the custom firmware you have for usb3.1
(ie a bug in the firmware for usb 3.1 that is in the utility directories on the gigabyte website)

bugs at this level can screw up a lot of subsystems when they share a common interface like the pci/e intrface.

 

[farmerdils]

That was installed recently, as in today. That won't have anything to do with my other BSODs though, right?

And how would I fix that? Do I just uninstall the USB 3.1 firmware?

edit: memory dump link: https://1drv.ms/u/s!AhcS4hvw2pjjgiIpYpL-ea3QRn0t

Seems like I'm adding issues onto my already problematic build.

Intel diagnostics tool comes back as pass for everything.

 

[johnbl]

Go to control panel and set your system power management to high performance just to see if you can avoid the bug check.
I will try and look at the memory dump in the morning if you upload it.


b]

Something new has happened! I have driver verifier running and was playing Battlerite, and got a BSOD that I have never gotten before. APC_INDEX_MISMATCH.

Minidump: https://1drv.ms/u/s!AhcS4hvw2pjjgiEwZJF9QrHtrckt

Will update with memory.dmp if needed.

 

[farmerdils]

The memory dump is uploaded, I updated the old post. The PC is now in high performance. I'll keep investigating.

Can't thank you enough john.

 

[johnbl]

looking a the usb 3 error logs, you have a
ProductId: Gaming Mouse G502
DeviceDescriptor: VID 046d ( Logitech Inc.) PID c332 REV 0302, Class: (0)(0) BcdUsb: 0200
on your usb 3 port 4
the port was reset for some reason then the device(mouse) returned a error 0xc0000001(STATUS_UNSUCCESSFUL)
then the microsoft driver failed the call because the client did not support it.

looks like logitech updated the windows 10 64 bit drivers on jan 4 2017
http://support.logitech.com/en_us/product/g502-proteus-core-tunable-gaming-mouse
*you will want to also look for a firmware update for the mouse

i assume the MemCompression executable is part of the new windows 10 feature https://www.tenforums.com/windows-10-news/17993-windows-10-memory-compression.html
if it is hitting a bug then that would explain the corrupted core windows file and still allow you to not have malware.
it could also be a bug in the debugger when it tries to verify a file that the new memory compression routines have compressed.

you might google on how to disable the windows 10 memory compression and see if it has an effect

 

[farmerdils]

I updated and the driver showed that it had caused a crash in BlueScreenView, so I rolled back to a system restore point.

Before I try anything else out however, I'm going to try different RAM sticks. I've just contacted Newegg and asked for a refund on my RAM, and I feel like this would be the best way to rule it out before I get into the nitty gritty of trying to figure out any software related issues. I checked the QVL for my MoBo and my RAM doesn't seem to be on there. Many others seem to be having problems with this company as well.

I will update once new RAM is in (which should be tomorrow) and post back here. I really hope it fixes my problem.

Thanks for your help john.

 

[farmerdils]

UPDATE: I ran Windows Defender... it found a keygen on my system that was tagged as Malware. I've removed it and am scanning once more. Additionally I will run the Windows Defender offline. Fingers crossed this is it...

For reference this was my latest minidump (it crashed mid work session, which it has never done before): https://1drv.ms/u/s!AhcS4hvw2pjjgjftVNif_9WLMObR

 

[farmerdils]

Replaced the RAM. Ended up costing a bit more but now I have cool RAM with red LEDs on it, so...

Before I mark this solved I'll give it a few more days, but things are looking good so far. I'll slowly start adding programs back in and seeing how they behave in the build.

I replaced it with Corsair Vengeance 3000 DDR4 2x8GB (XMP 2.0 enabled). If anyone is reading this who has had this problem: memtest isn't always correct. The only way to know for sure is to test with different RAM that has been proven to work.

I'll update with any new details and if I experience any crashes. Thank you for all your help, will definitely be coming back here for any other problems I might have.

UPDATE: Not sure what to pick as the best answer, but the problem seems to have gone away.