[SOLVED] Clock watchdog timeout

[Tallion]

hello I would like to ask what could be caused by a PC with a new new motherboard with a new processor, RAM and GPU ran all well till yesterday and today when I played it fell twice and I threw it ==> Clock watchdog timeout
New OS: since September 13, 2019 the problem occurred only today around 12 th

MB : MSI b450 gaming plus
CPU : amd ryzen 5 3600
RAM: HyperX Fury Black 8GB DDR4 2400 ( 2x ) [16GB]
GPU : Zotac GeForce GTX 1660 GAMING, 6GB GDDR5
HDD 1 : HITACHI HUA7210SASUN 1 TB ( OS WIN 10)
HDD 2 : HITACHI HUA7210SASUN 1
PSU : Corsair tx850W ( 850W )

 

[PC Tailor]

I have run the dump file(s) and you can see the full report(s) here:
https://pste.eu/p/AXgA.html CLOCK_WATCHDOG_TIMEOUT
https://pste.eu/p/PCnH.html CLOCK_WATCHDOG_TIMEOUT
https://pste.eu/p/RuEW.html CLOCK_WATCHDOG_TIMEOUT
https://pste.eu/p/dHpv.html CLOCK_WATCHDOG_TIMEOUT
https://pste.eu/p/RZwK.html CLOCK_WATCHDOG_TIMEOUT

Summary of findings:

BugCheck 101, {10, 0, ffffd6812d2d1180, 9}
CLOCK_WATCHDOG_TIMEOUT
"This can come from misbehaving drivers, but are usually hardware. It is especially common in Virtual Machine environments if you are running them, it can also be hardware conflicts such as incompatibility, mixed RAM, or even firmware conflicts with the hardware.

• Confirm that hardware is compatible with your system.
• View the third party modules of the report to view which drivers may be misbehaving.
• Hardware testing may be necessary.

Some things to consider:
- Clock Watchdog timeouts are CPU hangs, and the majority of the time are hardware based errors. So I think we're on hardware testing due to the previous steps we've taken to update all the previous drivers.

I think @Colif is correct to suggest Prime95 as when all 5 dumps are not showing any clear third party drivers, I'd be going back to it being a CPU hang hardware problem.
I'd also potentially check for bent CPU pins for example.

 

[Colif]

I suspect they all caused by Nvidia drivers.
Try this and instead of getting drivers from Nvidia, run windows update after DDU - https://forums.tomshardware.com/faq...n-install-of-your-video-card-drivers.2402269/

all of the bsod mention nt!KeAccumulateTicks+1cbd90
looking that up on google leads to Nvidia. Well, eventually.

 

[PC Tailor]

I suspect they all caused by Nvidia drivers.
Try this and instead of getting drivers from Nvidia, run windows update after DDU - https://forums.tomshardware.com/faq...n-install-of-your-video-card-drivers.2402269/

all of the bsod mention nt!KeAccumulateTicks+1cbd90
looking that up on google leads to Nvidia. Well, eventually.

Well if it still occurred in safe mode however, full NVIDIA drivers shouldn't really be loaded.
In addition to that, it's much less common for clock timeouts to be driver based - happens, but certainly less likely

 

[Colif]

I think we got mixed up about the safe mode thing, I looked through thread earlier and all I saw was him saying it ran fine in safe mode and asking how long he should run in that mode. I think it only crashes in normal mode.

no reason to not run Prime95, I was one who suggested it. I just thought, its blaming a WIn 8 driver which we know includes win 10, so could still be either.

 

[PC Tailor]

Anyway, I have run some extra info on the clock watchdog.

So the Arg3 identifies the the PCRB address of the hung processor, when looking into it, it's referencing processor 8 (which can also be seen in P4). So processor 8 was the one with the hang.

0: kd> !prcb 8
[b]PRCB for Processor 8 at ffffc980afe79180:[/b]
Current IRQL -- 0
Threads--  Current ffff968b4c283080 Next ffff968b4c825080 Idle ffffc980afe8a340
Processor Index 8 Number (0, 8) GroupSetMember 100
Interrupt Count -- 00d97fff
Times -- Dpc    00000019 Interrupt 00000017
         Kernel 000b430a User      00000d8d

So I reconfigured the dump file to look at processor 8 (as it defaults to processor 0):

8: kd> kv
# Child-SP          RetAddr           : Args to Child                                                           : Call Site
00 000000f1`6d6fe668 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x00007ff9`a3d2c5a2

So can't see any third party modules there - and then processor 0 called the bugcheck. Stack of processor 0

00 fffff802`51a75b08 fffff802`4cfee7b0 : 00000000`00000101 00000000`00000010 00000000`00000000 ffffc980`afe79180 : nt!KeBugCheckEx
01 fffff802`51a75b10 fffff802`4ce1f5dc : fffff780`00000320 fffff802`4a005180 00000000`000b5234 00000000`000b5234 : nt!KeAccumulateTicks+0x1cbd90
02 fffff802`51a75b70 fffff802`4cd5e4b7 : 00000000`00000001 ffffb907`9f339fe0 ffffb907`9f33a060 00000000`0000000c : nt!KeClockInterruptNotify+0x98c
03 fffff802`51a75f30 fffff802`4ce029e5 : 0000001a`fe01aacb fffff802`4cdcc0e0 fffff802`4cdcc190 00000000`00000000 : hal!HalpTimerClockInterrupt+0xf7
04 fffff802`51a75f60 fffff802`4cfc2cba : ffffb907`9f33a060 fffff802`4cdcc0e0 00000000`00000000 fffff802`4cdcc0e0 : nt!KiCallInterruptServiceRoutine+0xa5
05 fffff802`51a75fb0 fffff802`4cfc3227 : 00000000`1602d310 ffffb907`9f33a060 fffff802`4cdcc0e0 00000000`00000000 : nt!KiInterruptSubDispatchNoLockNoEtw+0xfa (TrapFrame @ fffff802`51a75e70)
06 ffffb907`9f339fe0 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiInterruptDispatchNoLockNoEtw+0x37 (TrapFrame @ ffffb907`9f339fe0)

 

[[C]oopeZz]

Hello,

Yes, BSOD never appeared in safe mode. His computer can run hours and able to play games. However, BSOD can appear even when he didn't play games. While writing this message, he got BSOD while using some sort of "stress test" so we can expect new minidump. I'll let him know about NVIDIA drivers.

 

[Tallion]

https://ufile.io/uu0a9hed

 

[PC Tailor]

What stress test was it?

 

[[C]oopeZz]

Prime 95, suggested by Colif.

 

[PC Tailor]

Prime 95, suggested by Colif.

And it failed during Prime95 test?
If so, I would check for the failure in the logs and that would explain the CPU hang.

I.e. faulty CPU.

 

[PC Tailor]

I have run the dump file(s) and you can see the full report(s) here: https://pste.eu/p/cHIY.html

Summary of findings:

CLOCK_WATCHDOG_TIMEOUT
"This can come typically from malfunctioning hardware, but in rare occasions can also come from some software. It is especially common in Virtual Machine environments if you are running them, it can also be hardware conflicts such as incompatibility, mixed RAM, or even firmware conflicts with the hardware.

• Confirm that hardware is compatible with your system.
• View the third party modules of the report to view which drivers may be misbehaving.
• Hardware testing may be necessary."

Some things to consider:
I would re-run the Prime95 test to confirm, and check the fail logs, if it repeats failure on a second run of Prime95 - I'd say faulty CPU as this is usually the reason behind a 101 bugcheck once software is eliminated.

It keeps failing on processor 8 also, so there is a common trend.

 

[Tallion]

according to Mr. Clif, I uninstalled the gpu drivers in safe mode through the DDU (Display Driver Uninstaller) and when installing the drivers directly on my gpu downloaded from (https://www.nvidia.com/Download/index.aspx) it wanted a reboot in some part so I reboot and the PC threw again when starting (CLOCK_WATCHDOG_TIMEOUT) and over and over and over, then I ran it in safe mode and reboot my pc and it normally started up and installed the GPU drivers pc of course created a new mini dump and also in those crashes 3 of them damaged of course here are the 2 new ones: https://ufile.io/4xlndsu6

 

[Colif]

All the BSOD (up until the last 2) mention - CLOCK_WATCHDOG_TIMEOUT_INVALID_CONTEXT_nt!KeAccumulateTicks
searching by that I get another case where the ram speed is wrong. Ram speed hasn't shown on any of the dumps, which is not unusual for PC's dumps but I am still not sure what its speed is, He has new bios now so it should show.

Prime95 also tests ram, its possible some ram settings are wrong?

did we ever run this? Try running memtest86 on each of your ram sticks, one stick at a time, up to 4 passes. Only error count you want is 0, any higher could be cause of the BSOD. Remove/replace ram sticks with errors.

If op can upload copy of the prime95 logs (believe text file goes to install folder) we can see what they have to show us.

 

[PC Tailor]

Also have you monitored CPU temps during the Prime95 test? Perfectly possible for temps to cause CPU hang bugchecks depending on the situation.