Windows 10: Random BSODs and Freezes, new Skylake build; 0xa, 0xc1, 0x3b, 0x1a,...

hey guys,


i built a pc around 2 months ago:


i5 6600k
Asrock Z170 Extreme4
G.Skill Ripjaws V 2x4GB (3000MHz)
XFX Radeon RX480 8GB
OCZ Trident 240GB SSD
Seasonic S12II-620W


i bought these components for overclocking, so i did that very early on. the first 2 weeks the system ran just fine, then it started to get BSODs and freezes. many times the mainboard diagnose leds would show after a crash the code '55' which is for memory not detected. I decided to step down from the overclock reducing the ram clockrate with no positive effect. i ended up getting crashes at stock settings aswell at that point. i got all the crashes at different moments, appearing very randomly. so this is the first part of the story.


i thought i had faulty ram so i decided to rma the g.skill and i got some Corsair Vengeance LPX 2x8GB (3000MHz). the system crashed less often but it still did. so i concluded it weren't the ram. i could leave the ram at stock settings and run a mild OC of the cpu at that time (still getting crashes occasionally). usually when i got one crash i would get another one very shortly after, like in the next 30mins. so i let it like that for some weeks dealing with the crashes and reducing also the cpu to stock settings... until it crashed again and didnt post anymore even after switching the dimm slots. i got first the code '62' and later the '79' on the mainboard led (Code 61-91 - Chipset initialization error. Please reset or clear CMOS.). i could make the system boot by trying it with a single ram stick. i tried it for all the slots with each of the 2 sticks and it worked. after already spending a lot of time and money into the system i thought if its not the ram then it has to be the motherboard....


so i rma'd the mobo and got a Gigabyte Z170X-Gaming 7 with the big hope that my issue is gonna be resolved. well sadly it wasn't. i got the gigabyte board this week,assembled my rig back again and installed windows 10. when that finished i installed a couple of drivers for the mobo aswell as the graphics card driver (i downloaded all from the manufacturers sites to be up to date). then i got a BSOD. i have to say that here i had XMP activated as i thought that the issue was on the old motherboard. okay so i disabled xmp again and decided to install some games and to see if i run in any more crashes. i did.


at that point i wanted to look deeper so i checked the dump files finding out that 'ntoskrnl.exe' is always involved. i did some driver verifier tests with different settings. i got MEMORY_CORRUPTION when i stressed only 'ntoskrnl.exe' through driver verifier. when i selected everything in the developer settings except low resources simulation and force pending i/o requests aswell as all non-windows drivers the system would not boot. the dump blames cthda.sys all the time. this is the soundcard driver from creative which is integrated on the gigabyte board. i can't update this one as i have the latest one, at least according to the gigabyte website and device manager.


okay so im pretty sure this has to be a hardware related issue and there aren't that many components left that i didnt change yet. the only check for the ssd i did was installing the manufacturers software which is monitoring it. it didnt find any issues on that but still i wanted to give a try to another drive. i installed a new downloaded iso of windows 10 on a hdd today. this time for some reason windows installed most of the drivers itself which kinda surprised me. i ran driver verifier again to see if the system boots this time and if the cthda.sys will still be an issue. this time the system booted and i was really happy and hoped that i finally solved it. just to kill my happiness it got a BSOD. it seems it was not triggered by the driver verifier as i got IRQL_NOT_LESS_OR_EQUAL. i started it again, saved the minidump and memory.dmp on an external drive and booted now the ssd again.


i ran memtest86+ last night and got 8 passes with 0 errors, i also ran prime95 on blend for a whole night without any issues. i didnt ran any special hard drive diagnose tool as i installed windows on 2 different drives. i didnt overclock any of the components for a while now.


im running out of ideas what it could be. in my opinion the most likely thing that is at fault might be the cpu but at this point i actually have no idea at all. its pretty much 3 new installs of win10, all new components and still i get random BSODs and freezes. the problem is i cant really reproduce the freezes and BSODs so troubleshooting is really tricky. i'd prefer to rma my stuff and end up with a stable system than fighting in software against a hardware problem (please correct me if i'm wrong on that), specially when the hardware is so new. moreover i spent a lot of money so i can overclock and now i'm having problems keeping a not overclocked system stable. i'm also still waiting for a new nvme drive so i definitely will be installing win10 again on a third drive.


i have to note that all the dumps i have are from the gigabyte board, i sadly dont have any from the time i was using the asrock board but i can tell there were a lot of different error messages (i do have some photos of BSODs at that time). i also saved a couple of memory.dmp files just in case they could be necessary (also all from the gigabyte board). please, someone help me out of this mess!




greetings,





Edit: Formatting fix

:)

 

Windows freeze and no reponse

Hi BBCar, NickCR1

I'm certain that this problem is caused by using Kaby Lake CPUs and BIOS 7.0.

My Deskmini with a Skylake CPU had been working fine for 8 months. After I upgraded CPU to a Kaby Lake CPU, it started having the same problem you mentioned and both of you use Kaby Lake CPUs

Once I swapped the Kaby lake for the old Skylake, the random freezes were gone.

Too bad no new BIOS since 7.0

 

Multiple BSOD Problems (maybe caused by Hardware, Driver)

I am reading this thread with interest as my issues highlighted in another thread:

Winsdows 10 - BSOD

were very alike.

I have not got to the bottom of it all myself and have handed my PC back to HP (under warranty) who after some investigations awaits spare parts apparently.

I have enquired what exactly is faulty and still await HP response, but seeing this and other threads with unexplained BSODs on brand new PCs with Windows 10 built on Skylake platform I started wondering whether this has not something to do with this?

https://www.theinquirer.net/inquire...ke-bug-that-freezes-windows-and-linux-systems

Has anybody tried testing against this?

How to test your PC for the Skylake bug

 

Early build
You are still on a early build of windows 10 Windows 10 Pro 64-bit (10.0, Build 10240) (10240.th1_st1.161118-1836)

Update your system to the latest build and install all windows updates.
This will improve the stability of your system and if your lucky may even stops the BSODs.

 

thank you for your reply!

i installed all the updates now.
when i installed windows on this ssd i knew it will be only temporary as the nvme drive will be my main drive as soon it arrives. i was too focused on this being a hardware issue that i didnt pay too much attention to updates.

update: just got a freeze again. had rocket league, a stream over vlc, teamspeak and chrome running. i was looking for a minidump but the folder is just gone? does this happen when i collect the dumps with the tool from the forum or did something else delete it? i know that you usually dont get dumps when u get a freeze. the screen turned blue like for a bsod but i didnt get any message. i could see and hear the freeze on my second screen as the stream was running there.

 

Could you run the tool again and upload it.
Maybe we can get any clues from the other reports, like eventviewer, dxdiag and msf032 etc., if there are no dumpfiles inside.

 

sure. i included a photo in the zip of the bios message i got after the freeze. i had it already on default settings though...

thanks for your support!

 

Without dumpfiles it is difficult to analyse.
But in the eventviewer i found this
Code: Event[423]: Log Name: System Source: Service Control Manager Date: 2017-01-05T15:18:40.918 Event ID: 7034 Task: N/A Level: Error Opcode: N/A Keyword: Classic User: N/A User Name: N/A Computer: DESKTOP-4VDHCQS Description: The SB Recon3D Service service terminated unexpectedly. It has done this 1 time(s).[/quote] In the msinfo32 i found no references to a bugcheck, so i suspect that your system did not BSOd but only freezed.

Also as you already mentioned the previous dumpfiles of 4 jan all are blaming cthda.sys. part of soundblaster Recon 3d.
When googling there are many post about your mobo and onboard SB Recon 3d drivers problems/issues
GA-Z170X-Gaming 7 Audio Windows 10 and GA-Z170X-Gaming 7 Audio Windows 10 and in German https://eu.battle.net/forums/de/wow/topic/16885180658

So it could be that your problem/issue is not hardware related but software/driver related.

As a test if you disable the onboard sound(sb recon) in your BIOS could you check if your system then also still freezes( or get a BSOD).

If you get a new BSOD could you upload a new zip with the dumpfiles.

 

yes i've read already about the issues with the onboard sound but the most common thing was that the audio started to glitch, not that the whole system would crash.

anyway, i disabled the audio controller in the bios and had a short game session again with a stream running and teamspeak. this time no crash but unfortunately that doesn't mean a thing in this case.

so now i enabled driver verifier again, i'm stressing everything from the advanced section except force pending i/o requests and low resources simulation aswell as all non-windows drivers. the system is booting this time as the cthda.sys is not among the drivers being stressed. i will leave it running for now and see what comes out of this.

should i do any other stress tests on top of that (like prime95) or is that too much?

 

ok so it ran like this for over 24 hours without a crash, even though i experienced something that felt like mini-freezes when i had a faster download running (>5mb/s).
i decided to test an overclock with putting a little bit higher clock on the cpu and enabling xmp. i left the driver verifier up and turned prime95 on for an hour. temperatures were good, i could run a stream besides that and all was looking fine. i wanted to test the OC a little bit more so i turned off driver verifier. i restarted, checked cpu-z and went in the bench tab on stress cpu. i was doing this already on the asrock board just to bump the boostclock of the cpu and to check voltages. back then i could do everything besides while running that. well this time i realized very quickly that the system started to get mini freezes until it had a bsod. so i turned off OC and tried the same thing on stock settings. when i realized i'm getting the mini freezes again i stopped it before it would crash. as far as i know this is a very weak stress on the cpu and i only do this to check on the basic values before i get into longer stresstesting.
well, i decided to apply driver verifier again with the same settings as before but now with ntoskrnl.exe included. this time i was getting bsods while booting.
i can't really figure out much of this dumps but it says:

Probably caused by : ntkrnlmp.exe ( nt!IopProcessWorkItem+f0 )

i attached the zip again (i removed a dump i got from driver verifier due to cthda.sys as my bios settings got reseted and the soundcard got active again)
the dump where cpu-z is mentioned was while being on OC settings.

 

First i am not really a hardware guy and certainly not a overclocker.

But your intel extreme tuning utility has drivers from 2015, update to the latest version Download Intel(R) Extreme Tuning Utility (Intel(R) XTU)

Your Intel(R) Ethernet Connection (2) I219-V, although you have disabled it, has drivers from march 2016, maybe there is a newer driver version.

The 2 dumpfiles you mentioned with Probably caused by : ntkrnlmp.exe ( nt!IopProcessWorkItem+f0 )
are:
STOP*0x000000C4: DRIVER_VERIFIER_DETECTED_VIOLATION*** (go to top of page)
Usual causes:* Device driver - see parameters for more information
NOTE:* May occur in Win7 and later w/o Driver Verifier being enabled

In the other dumpfile with BugCheck 3B in the unloaded modules i found
fffff80b`84350000 fffff80b`84363000 xinputhid.sys
This seems to be related to xbox one.
Do you have a xbox one device connected to your system?

 

the thing that worries me is that when i stress in cpu-z on stock settings i get freezes. i dont know exactly how cpu-z's stresstest operates, but its just another indicator that somethings not right.
i installed now a newer driver for the intel nic and also the xtu. i'll do some testing today and see what happens.
yes i have a xbox one controller connected to the pc, i use it most of the time for gaming. what does it mean if its an unloaded module? did it cause the bsod?