Random BSOD after windows 10

[vish_93]

Hey Guys,
I am trying to troubleshoot random BSOD that has been plaguing my pc after upgrading to windows 10. I also upgraded the bios after that. Now it seem that the windows will randomly BSOD. I have completely removed any overclock on my cpu (except RAM XMP profile) and gpu but the problem remains.

My PC specs

Core i7 4770K @ stock
MSI Z87-GD65 Gaming with latest bios (i believe V1.12)
16 GB Crucial ballistic RAM @ 1866 MHz
Asus R9 280X Matrix Platinum @ stock
480 GB Intel enterprise SSD (dont know the model number - OS drive)
120 GB Mushkin Chronos
3 TB WD Green
Asus Xonar D2X with uni drivers (as no asus drivers for this card on windows 10)

Using BSod viewer it seem the problem is sometimes caused by either ntoskrnl.exe or ndis.sys or werkernel.sys or afd.sys. But ntoskrnl appears in all of them.

After some research it turns out that ndis and afd is related to networking i removed the drivers i got from msi and downloaded the latest one from killer themselves but to no avail. I have also stress tested the cpu and ram using occt and gpu using unigine valley no crashes.

Any suggestion what should I do next? thanks alot for helping me

 

[tsnor]

"...Any suggestion what should I do next?..." Here are suggestions, hope they work for you.

First see if MSI has driver updates for the MSI Z87-GD65 MB for win10. If not I'd open a ticket with the support team. MSI released win10 drivers for my MB right at general availability of Win10. The drivers windows update finds will be generic 'hope this works' drivers, not the ones MSI tested for your MB.

Next I would physically remove the Xonar D2X from the PC and see if it stabilizes. If so you might need to wait for win10 drivers to appear to use the D2X.

If above doesn't help, I'd leave the D2X out of the PC and next tell windows to disable the killer ethernet port in device manager.

 

[johnbl]

put your memory dumps on a server and post a link.
yes afd.sys and ndis.sys are networking components
they support wifi wireless ethernet, wired ethernet and bluetooth networking.
(among other networking stuff)

getting bugchecks in ntoskrnl is not too useful, it provides the interface for the drivers in the system. most drivers break in their code or the code they call in ntoskrnl.
for example, if you call the ntoskrnl with a bad memory address from your driver ntoskrnl detects the bad address and calls a bugcheck but does not name the bad driver, just the bad memory address.

 

[vish_93]

First see if MSI has driver updates for the MSI Z87-GD65 MB for win10. If not I'd open a ticket with the support team. MSI released win10 drivers for my MB right at general availability of Win10. The drivers windows update finds will be generic 'hope this works' drivers, not the ones MSI tested for your MB.

Next I would physically remove the Xonar D2X from the PC and see if it stabilizes. If so you might need to wait for win10 drivers to appear to use the D2X.

If above doesn't help, I'd leave the D2X out of the PC and next tell windows to disable the killer ethernet port in device manager.

Hi I downloaded their live update and I have the latest drivers and bios apart from Asmedia controller. Not sure if it is needed. I guess it is a good idea to open a ticket I will do that now.

Also should I just remove the card or should i remove the drivers and the card?

 

[vish_93]

put your memory dumps on a server and post a link.
yes afd.sys and ndis.sys are networking components
they support wifi wireless ethernet, wired ethernet and bluetooth networking.
(among other networking stuff)

getting bugchecks in ntoskrnl is not too useful, it provides the interface for the drivers in the system. most drivers break in their code or the code they call in ntoskrnl.
for example, if you call the ntoskrnl with a bad memory address from your driver ntoskrnl detects the bad address and calls a bugcheck but does not name the bad driver, just the bad memory address.

https://drive.google.com/open?id=0B5pWqBg8cEbfZVl1X1VXY1hrU2M

there you go the latest one is 5859. Thnaks alot for taking time to help me out.

Also when I turned my pc on i got a message saying windows didnot load correctly and then gave me 2 options restart or repair. I restarted and it worked. Have any of you seen it before?

 

[johnbl]

you might consider confirming that your BIOS version is good (has not been removed by the vendor) update the drivers, reboot into BIOS, reset the BIOS to defaults then boot and run memtest86 to confirm your memory works as expected.

there is a chance you have a bad ram stick, each time windows loads it will load the drivers in different orders. if you have a bad ram stick, you get different bughcecks on each boot. Each of your bugchecks indicated a bad memory address but with different drivers. Run memtest on its own bood image to confirm your memory is ok.
-----------------


4th bughceck was a bad stack pointer in your graphic driver
third bugcheck, looks like a bad memory address given to a power managment routine

second bugcheck was in a network driver:
\SystemRoot\system32\DRIVERS\bwcW10x64.sys Tue Jul 07 11:27:00 2015
A bad memory address was used and passed to the kernel which resulted in a bugcheck.

I would try the updated drivers for killer network on the msi websited (came out aug 3 2015)
http://us.msi.com/support/mb/Z87-GD65-GAMING.html#down-driver&Win10 64
listed as onboard lan drivers.


-----------
most current bugcheck was in USB extensions, this tends to indicate you need a custom driver to be used with your usb support for your motherboard.
look here: http://us.msi.com/support/mb/Z87-GD65-GAMING.html#down-driver&Win10 64
i would guess it is in the system and chipset driver package.

Note: I could not find your BIOS version So confirm I picked the correct motherboard or select the correct one. (also make sure your BIOS was not removed because of a bug, it was newer than the ones listed)
Note: you have to change the memory dump type to kernel memory dump in order to see what is wrong with the USB ports. how to change the memory dump type:
https://www.sophos.com/en-us/support/knowledgebase/111474.aspx
kernel or full memory dumps will have the USB info, minidumps strip out the debug info for USB and don't save the error logs.

Most likely you will be looking for a update to a USB 3.0 chipset driver. (lots of versions of these chipsets, each with their own bugs in the chips, or in the BIOS setup for the usb 3.0 chipsets
note 2: you might also work around the problem by removing all of your devices from your usb 3.0 ports and plugging them into usb 2.x ports. in the stack trace it looks like some older (legacy USB device) was plugged into a usb 3.0 port (generally color coded with blue plastic on the connector) the device was told change its power state (sleep/wakeup) and the microsoft driver got a bad memory address and called a bugcheck. (hidusb.sys)

kernel memory dump I can tell you what the device was that made the bad call.




remove Asus Smart Doctor/ GPU tweak driver
http://www.sysnative.com/drivers/driver.php?id=IOMap64.sys Wed Oct 22 17:52:12 2014

update: Intel Smart Connect Technology Driver
\SystemRoot\System32\drivers\ISCTD64.sys Tue Nov 27 11:52:34 2012
maybe look here: http://www.intel.com/p/en_US/support/highlights/sftwr-prod/smrt_cnct

machine info:
BIOS Version V1.12
BIOS Starting Address Segment f000
BIOS Release Date 05/19/2015
Manufacturer MSI
Product Name MS-7845
Version 1.0
Product Z87-GD65 GAMING (MS-7845)
Version 1.0
Processor Version Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz
Processor Voltage 8ch - 1.2V
External Clock 100MHz
Max Speed 3800MHz
Current Speed 3500MHz

 

[vish_93]

thanks john for an extensive answer. I checked and I do have the latest bios. I changed memory dump to full memory dump to get as much information as possible. Then weirdly my computer started acting strangely it didnot freeze or anything but clicking on stuff will not do anything. couldnt even restart the computer. So before I do any of your other suggestion i decided to make a usb memetest and put it on test.

unfortunately I didnot reset the bios to default and already started the memetest. Is resetting absolutely necessary? I am currently 20 mins in with no errors. will keep you posted

Edit: I checked back in an hour and a half and it was showing 700 errors. So ram has gone bad. Is there a faster way to check which stick has gone bad. with out testing each individually?

 

[johnbl]

it is always better to update or reset the BIOS to defaults before running memtest, if you find errors you never know if it was corrupted/modified BIOS settings and you have to reset to defaults and re run the tests. If it fails, then you start looking at the memory timings and confirm they are correct. Or run on one stick and see if you can isolate a bad stick with problems.

thanks john for an extensive answer. I checked and I do have the latest bios. I changed memory dump to full memory dump to get as much information as possible. Then weirdly my computer started acting strangely it didnot freeze or anything but clicking on stuff will not do anything. couldnt even restart the computer. So before I do any of your other suggestion i decided to make a usb memetest and put it on test.

unfortunately I didnot reset the bios to default and already started the memetest. Is resetting absolutely necessary? I am currently 20 mins in with no errors. will keep you posted

Edit: I checked back in an hour and a half and it was showing 700 errors. So ram has gone bad. Is there a faster way to check which stick has gone bad. with out testing each individually?