Windows 10: BSOD Windows 10 build 1703 and 1709

Hi there.. please forgive me if i mess up any posting instructions.. i'm new here, and i've made an account to get some help with some BSOD issues on my system.. i've created the files as instructed in the BSOD posting instructions and i have attached the zip for you.

I'll give a little background, this is a pretty newly built system, and up til around a week ago was operating perfectly with zero crashes and not a single BSOD.

The first issue occured when the whole system locked up and gave me a corrupted bios and a system that wouldn't POST(happy to have a proper dual bios on this motherboard), i fixed the corrupted bios with a reflash from within the backup bios, and then the system worked normally, then the next day, for no reason at all, the system decided to POST with the backup bios, so i powered down and did a full bios reset(battery pulled for 30 minutes and cmos clear pins shorted.) and the system then proceeded to POST on the main bios, and was fine for a few days, til i got another hard system lockup, i restarted and the system worked normally afterwards.

After this, the system then started to give me BSODs. the first one was when i was still running the fall creators update, and sadly i lost the dump file when i wiped the OS so i could reinstall from build 1703 to rule out the fall creators update being a part of the problem. the reason that windows gave for that BSOD was memory_management. my ram has been tested and seems to be fine.

This brings me to the BSOD that is a part of the files in the zip i attached, this gave me a different error in the BSOD to the last one, this time it is WHEA_UNCORRECTABLE_ERROR.

I'm an experienced system builder, i've never encountered issues like this in the past that i couldn't pin point myself, but this time? i need some help figuring out the cause of the BSODs, so hopefully someone here can help, before i pull my hair out lol.

Also, here are my full specs.

Phanteks Enthoo Luxe Tempered Glass Edition Full Tower Chassis
Windows 10 x64 Pro Edition
AMD Ryzen 7 1700x
Gigabyte Aorus AX370 Gaming K7
16GB Corsair Vengeance LPX DDR4 3000mhz C15
Corsair H100i V2 with two Corsair SP120L Fans
Gigabyte Geforce GTX 1070 Xtreme Gaming 8GB
Samsung 850 EVO 500GB SSD
3TB Seagate Barracuda
Corsair 3rd Gen RM750 Modular PSU
Razer Deathadder Chroma
Coolermaster Masterkeys Pro L Cherry MX Brown RGB Keyboard

:)

 

Windows 10 build

Simple question. Let's assume i install Windows 10 enterprise build 1703. Does automatic updates will update it automaticly to build 1709? Asking because currently have problems (shutdowns..) with build 1709 (most probably) and i am gonna to install from
scratch build 1703 and don't want windows to update it automaticly to 1709 again. Does build is lets say completly diffrent of other build? How it works.

 

how correctly install the latest windows 10 build?

I do need to have the 1703 build installed to correctly upgrade to 1709 build

or I can just fresh install windows 10 directly with the latest build such the 1709??

 

Hi AliasOfMyself,

Welcome to 10forums *Smile

The dump is partly corrupted, the best I could get from it is that the 0x124 occured due to a TLB data error with the L0 cache.
As you've had some BIOS issues I would say that your motherboard is the issue.

 

Hi there axe0, thanks for the super fast response!

any ideas why the dump was corrupted? the install of windows is new, i forgot to get around to disabling auto restart on system failure so maybe it restarted before it finished creating it? i can read the dump fine on my end(i think) using windbg.

I have a friend with the same processor and motherboard as me, he has none of these issues. there is a newer bios for my motherboard that states it improves ram stability, do you think that updating it might help? in all honesty i'm already considering sending the board and ram back to the retailer and have them test it, but i need to be sure before i tear the entire system down.

also worth mentioning, the BSODs and crashes are not a daily occurance, it's once every few days since the bios corrupted, for example, i spent around 6 hours on the division yesterday and i didn't even get a game crash, let alone an OS bsod or system lockup... don't you just love issues like this.. lol

 

The dump is corrupted, as in the raw stack cannot be dumped.
Code: 8: kd> !rawstack <-- command from extern DLL FAILURE: Dump file is corrupt. nt!_KTHREAD: 0xffffa006`1a04c7c0 Stack Base: 0x00000000`00000000 Stack Limit: 0x00000000`00000000[/quote] A BIOS update usually improves things, definitely worth it *Smile

You would be surprised how many simlar situations there are that turn out to be a motherboard, CPU or PSU issue.

 

I've no idea why its corrupted like that.. Windbg didn't give me any errors when i ran analyze -v after loading in the dump, but you obviously know a significant amount more about dump files than i do, i can usually find the culprit if it is a driver, but this isn't the case here, this much i can tell.

Oh i've seen a lot of different causes for a blue screen of death over the years, and when it's impossible to repeat the crash it's seriously hard to diagnose lol. I'll update the bios shortly, making sure i do a full cmos clear before and after the update so i can clear any memory training etc.

fyi, i don't do overclocks, i've not even attempted to overclock this system as i'm more than happy with the performance i'm getting, the system performs so well until it crashes once, then it's back to being perfect again, i hate it *Biggrin

 

Analyze -v shows a verbose analysis of the crash with a call stack.
When I run !rawstack, !rawstack runs !thread and dps {stack limit} {stack base} commands to get the raw stack. Since the thread data structure is missing it can't be used. This is unusual, because the thread where the crash occured should be available in minidumps.

 

Well i'm finally back from a bios update, to the latest and greatest from the gigabyte downloads page for my board lol. For some reason my pump wasn't being told to spin by its software afterwards and i finally fixed it by just re-applying a profile to jump start it, odd things are going on with this machine! *Biggrin

I'll give this bios update a few days, and if i get a bsod again, i'll upload it to this thread and update the thread accordingly *Smile should i just upload new dump files on their own now since you have all the info you need in the first zip? hopefully the next one is more useful!

Thanks for helping me out with this, it's very much so appreciated! *Smile

 

The eventlog, error reports and a few other logs are the main reasons why complete new logs are requested. Those logs may contain critical new information.

However, if the motherboard turns out to be bad, those logs will be full of errors and problems depending on where the motherboard is faulty. That will take a long time to go through and connect with each other, a few dumps only will be enough *Smile

 

So if the next few dumps are inconclusive then i'll upload the full log set from the crash after the next few dumps?

 

If it looks like more information is required it will be asked for *Smile

 

no worries, i'll reply here either way in the next few days or so *Smile

 

Right so i just got another BSOD.. a different one this time, i'm unsure if the dump is corrupt or not, but i've uploaded it in a zip file, i may have to send this board back!

 

Could you upload the kernel dump, in C:\Windows\MEMORY.dmp, to onedrive, google drive or dropbox and post a share link.
This one's a bit of an exceptional BSOD, although I've seen it quite often lately, there's not much of documentation online for proper troubleshooting this one so I'm wanting to make one.

Source: Determining the source of Bug Check 0x133 (DPC_WATCHDOG_VIOLATION) errors on Windows Server 2012 Ntdebugging Blog
Code: ******************************************************************************* * * * Bugcheck Analysis * * * ******************************************************************************* Use !analyze -v to get detailed debugging information. BugCheck 133, {1, 1e00, fffff8004445f348, 0} *** WARNING: Unable to verify timestamp for nvlddmkm.sys *** ERROR: Module load completed but symbols could not be loaded for nvlddmkm.sys Probably caused by : nvlddmkm.sys ( nvlddmkm+28df18 )[/quote]
It requires a kernel dump to properly troubleshoot it, mainly due to lack of information that is saved in the minidump, not because of corruption.
We need to know what DPC's / ISR's were running and which of them offended the DPC duration limit.
Code: 0: kd> !dpcs Cannot find the DPC Queue Information. This extension may not work on early builds. 0: kd> !irql nt!_KPRCB.DebuggerSavedIRQL not found, error : 0x3. 0: kd> !pcr Unable to read the PCR at fffff780ffff0000[/quote]