Question Windows 11 freezes then BSOD - Hypervisor Error

[holoninja]

My laptop Asus ROG Strix G513QM keeps freezing 30-40 minutes after booting into windows and will eventually BSOD into hypervisor error. I have tried to factory reset, update drivers, run driver verifier but nothing worked. Running driver verifier still gave Hypervisor error. Before I factory reset, I did not get BSOD but it just froze forever until hard reset through power button. I have attached the dmp file from C:\Windows. Please help.

crash dump

 

[Colif]

Report - click read as fiddle to read (mostly for me)


File: zl9s95.dmp (Sep 24 2023 - 03:29:06)
BugCheck: [HYPERVISOR_ERROR (20001)]
Probably caused by: memory_corruption (Process: dwm.exe)
Uptime: 0 Day(s), 0 Hour(s), 05 Min(s), and 39 Sec(s)

Comment: 2 or more types of RAM are installed.
did it come with both sticks?

8192MB	3200MHz	Samsung	M471A1G44AB0-CWE
8192MB	3200MHz	Micron Technology	4ATF1G64HZ-3G2E2

Process named is victim.
DWM.exe = WIndows Desktop Manager. It sits between Applications and the GPU drivers... so... lets start there.

try updating gpu drivers - https://rog.asus.com/laptops/rog-strix/2021-rog-strix-g15-series/helpdesk_download/ (Or run MyAsus and use it to update)
notices you have AMD & Nvidia drivers... fun. AMD were the ones I was after... looks elsewhere - try these since Asus stopped a year ago - https://www.amd.com/en/support/apu/...processors-radeon-graphics/amd-ryzen-9-5900hx

 

[ubuysa]

I would strongly suggest you remove the RAM stick you have added and test it on just one stick (and test it one each stick). It may well be fine on each stick on their own, but not with both together. Mismatched RAM causes all sorts of problems, you want all of your RAM sticks to be exactly the same part number.

 

[Colif]

I asked if they added as its fairly common for laptops to have 2 different makes in them from factory.

 

[ubuysa]

I asked if they added as its fairly common for laptops to have 2 different makes in them from factory.

Oh, OK. I wasn't aware of that. It sounds dumb to me.

 

[holoninja]

I updated both NVidia and AMD drivers and reverted to windows 10 to check if it freezes. It does but no longer blue screen, just freezes until hard restart. Still happens every 10-20 mins. If I leave display on without working on laptop it usually doesn't freeze. Usually happens only when I am using the touchpad and using the laptop. I haven't touched RAM so it should be all factory. I only added a WD SSD but I doubt that would cause any issues since windows is installed on the factory drive.

 

[holoninja]

Update. So I reinstalled windows 11. I updated AMD drivers once again but did not fix the freeze. I also ran a memtest86 which showed no errors. Windows update for some reason overwrites radeon drivers with an older version so I disabled windows updates for now and have latest drivers. It still randomly freezes now however without any warning until I hard restart. Weird thing is laptop lighting works and if I close lid the screen does turn off. Reopening screen turns on but laptop is still frozen. I can't seem to find anyway to unfreeze except for hard restart through power button. I didn't get BSOD, just freeze. Someone please help this has been troubling me for a while.

 

[Colif]

curious, try running in safe mode and see if it still freezes... drivers won't work but if it only freezes in some actions, it could be those drivers.

go to settings/system/recovery
next to advanced startup, click restart now

1. this restarts PC in a blue menu
   
2. choose troubleshoot
   
3. choose advanced
   
4. choose startup settings
   
5. click the restart button
   
6. choose a safe mode (it doesn't matter which) by using number associated with it.
   
7. Pc will restart and load safe mode
   

maybe its touchpad drivers?

 

[holoninja]

curious, try running in safe mode and see if it still freezes... drivers won't work but if it only freezes in some actions, it could be those drivers.

go to settings/system/recovery
next to advanced startup, click restart now

1. this restarts PC in a blue menu
2. choose troubleshoot
3. choose advanced
4. choose startup settings
5. click the restart button
6. choose a safe mode (it doesn't matter which) by using number associated with it.
7. Pc will restart and load safe mode

maybe its touchpad drivers?

I tried in safe mode but it froze in 10 mins. The entire screen freezes, even the video I was playing on youtube so its not just touchpad i think.

 

[Colif]

freezing in 2 different OS (win 10 & 11) + in safe mode... sure looks like a hardware problem

How long has it been doing this for?

CPU - AMD Ryzen 9 5900HX with Radeon Graphics
Ram - 16gb mixed ram... standard
GPU: NVIDIA® GeForce RTX™ 3060 Laptop GPU
Storage: 512GB PCIe® 3.0 NVMe™ M.2 SSD

rest of it is built onto motherboard... gpu is as well..

Weird thing is laptop lighting works and if I close lid the screen does turn off. Reopening screen turns on but laptop is still frozen.

Not sure if it would do that if cause is nvme... it might... tossing ideas around...

try runing Crystaldiskinfo (Blue icons) - https://crystalmark.info/en/
it will auto scan drives. you don't have to do anything.

software tests not perfect, can't see physical problems such as if its the controller chip.

 

[holoninja]

freezing in 2 different OS (win 10 & 11) + in safe mode... sure looks like a hardware problem

How long has it been doing this for?

CPU - AMD Ryzen 9 5900HX with Radeon Graphics
Ram - 16gb mixed ram... standard
GPU: NVIDIA® GeForce RTX™ 3060 Laptop GPU
Storage: 512GB PCIe® 3.0 NVMe™ M.2 SSD

rest of it is built onto motherboard... gpu is as well..



Not sure if it would do that if cause is nvme... it might... tossing ideas around...

try runing Crystaldiskinfo (Blue icons) - https://crystalmark.info/en/
it will auto scan drives. you don't have to do anything.

software tests not perfect, can't see physical problems such as if its the controller chip.

Its been over a month now. I already ran crystaldiskinfo earlier yesterday and everything seemed fine. Today I installed windows on my other SSD to see if my primary factory SSD might be bad but it still freezes even on my secondary SSD.

I have also noticed it happens a lot less with power cable

 

[holoninja]

Here's the crystaldiskinfo I ran now

 

[Colif]

there is a history of freezing on the model

tried running without the battery?

 

[holoninja]

there is a history of freezing on the model

tried running without the battery?

Yes I saw this reddit thread before, did not fix my issue sadly. I haven't tried removing battery yet.

 

[ubuysa]

I also ran a memtest86 which showed no errors.

Memtest is good but no RAM testing software can find 100% of RAM issues. I am still very uncomfortable with two different RAM sticks being installed. It would be extremely wise to remove one and see whether the problem remains. Try running on each stick separately. It's vital to be completely confident that the RAM is good before looking elsewhere.

 

[holoninja]

Memtest is good but no RAM testing software can find 100% of RAM issues. I am still very uncomfortable with two different RAM sticks being installed. It would be extremely wise to remove one and see whether the problem remains. Try running on each stick separately. It's vital to be completely confident that the RAM is good before looking elsewhere.

I understand. I'll try to remove and try running on both individually and update here.

 

[gardenman]

I agree that the RAM still should be checked first.

I am still very uncomfortable with two different RAM sticks being installed.

I understand how you feel. I've looked at 100's (maybe 1000's) of dumps on here. I'm no expert, but I've seen many laptops have 2 different brands of RAM installed. I assume one is installed or soldered onto the motherboard, and a 2nd one of a different type is added for more upgraded versions of the laptop. This is how they come from the manufacturer, not just added RAM by the end user. The 2 types are almost always a different brand name.

 

[ubuysa]

I'm not 100% convinced that's how they come from the (chassis) manufacturer, but I accept that's how they often come from the system integrator who supplied the laptop. I've seen many myself, but I didn't realise they'd been sold that way. Even so, it will continue to raise a red flag for me.

 

[holoninja]

Hi everyone, sorry I did not reply last couple days, had some work. However I have been using it on power cable and for some reason it I have noticed it does not freeze on power, at least it didn't when I used it. Yesterday it froze within 5 mins of booting when on battery but today I used it on power cable for an hour straight. I will definitely remove ram and try on each stick individually tonight or tomorrow morning as well. For the removing battery thing, is it safe to do so and try to run without battery? If so I will also try to do that.
Edit: I was using it on power cable and it just randomly restarted just now. After restart it froze again now on power cable. It was running fine till I changed the power mode from turbo to silent in g-helper to match the mode on battery. I will try to run on turbo mode on battery, see how that goes.

Update: No, still froze after 10 mins on turbo mode. It did however give me a BSOD with hypervisor error this time.
BSOD

 

[ubuysa]

Can you please upload the file C:\Windows\Memory.dmp to the cloud with a link to it here? That's the full kernel dump for your most recent BSOD (which should be the 0x20001 HYPERVISOR_ERROR).

 

[holoninja]

Can you please upload the file C:\Windows\Memory.dmp to the cloud with a link to it here? That's the full kernel dump for your most recent BSOD (which should be the 0x20001 HYPERVISOR_ERROR).

Yes,
MEMORY.DMP

 

[ubuysa]

In the kernel dump, where we can see all 16 processors and not just the one that failed, as in the minidump, it's clear that only processor #8 (the one that failed) is processing any real work. The other 15 processors are all running the idle loop.

The only processor we have to look at then is processor #8, the one in the minidump, and this BSOD is not caused by a rogue driver running on another processor. I hate to keep banging on about your RAM, but the indications in the full call stack in the kernel dump show that the BSOD was caused by a memory access fault. You can clearly see the call to the Microsoft nt!MmAccessFault function...

ffff8508`3b1de158  fffff807`20a401c2 nt!MmAccessFault+0x282

The nt!MmAccessFault function call indicates that there was a problem accessing memory (RAM), and that's the root cause of this BSOD. There is also a stack pointer error in this dump (most likely related to the access fault above), during a hypervisor function, and that's what eventually caused the bugcheck...

FAILURE_BUCKET_ID:  0x20001_13_0_STACKPTR_ERROR_nt!HvlSkCrashdumpCallbackRoutine

A stack pointer error is where a memory pointer become corrupted and no longer points anywhere sensible. In this case, as we shall see shortly, it was the instruction pointer (which points to the next instruction to be executed) that was in error.

In the call stack for processor #8 there is a call to an address, not a symbol. This is because we don't have a debugger symbol table entry for this address, and that's very suspicious (and unusual) in a kernel-mode stack. It follows a call to a hypervisor function (nt!HvcallpExtendedFastHypercall) which is why this was a 0x20001 bugcheck...

8: kd> knL
 # Child-SP          RetAddr               Call Site
00 ffff8601`1025eca8 fffff807`20d64b21     nt!KeBugCheckEx
01 ffff8601`1025ecb0 fffff807`20cc58b4     nt!HvlSkCrashdumpCallbackRoutine+0x81
02 ffff8601`1025ecf0 fffff807`20c3fd42     nt!KiProcessNMI+0x1df744
03 ffff8601`1025ed30 fffff807`20c3faae     nt!KxNmiInterrupt+0x82
04 ffff8601`1025ee70 fffff807`1eb90000     nt!KiNmiInterrupt+0x26e
05 ffff8508`3b1de858 fffff807`20c3e364     0xfffff807`1eb90000
06 ffff8508`3b1de860 fffff807`20bb48af     nt!HvcallpExtendedFastHypercall+0x54
07 ffff8508`3b1de870 fffff807`20bb472b     nt!HvlpFastFlushListTb+0xdf
08 ffff8508`3b1de980 fffff807`20bb468b     nt!HvlpFlushRangeListTb+0x83
09 ffff8508`3b1de9e0 fffff807`20a26e7b     nt!HvlFlushRangeListTb+0x33
0a ffff8508`3b1dea30 fffff807`20a265f3     nt!MiFlushTbList+0x51b
0b ffff8508`3b1deb10 fffff807`20a22958     nt!MiTerminateWsleCluster+0x373
0c ffff8508`3b1decc0 fffff807`20a22275     nt!MiDeletePteWsleCluster+0xb8
0d ffff8508`3b1dedf0 fffff807`20ed44f0     nt!MiDecommitPages+0x8f5
0e ffff8508`3b1df850 fffff807`20eb7deb     nt!MiDecommitRegion+0x80
0f ffff8508`3b1df900 fffff807`20eb7ab5     nt!MmFreeVirtualMemory+0x2fb
10 ffff8508`3b1dfa40 fffff807`20c46ee8     nt!NtFreeVirtualMemory+0x95
11 ffff8508`3b1dfaa0 00007ffb`cc38f114     nt!KiSystemServiceCopyEnd+0x28
12 00000073`5adfdaf8 00000000`00000000     0x00007ffb`cc38f114

Immediately following this rogue address is a kernel non-maskable interrupt (nt!KiNmiInterrupt) which is the beginning of the BSOD process.

If we take a closer look at the details of frame 5 (the one with the rogue function call) we can see that the instruction being pointed to by the instruction pointer (in the RIP register) is invalid (notice the ??? characters)...

8: kd> .frame /r 5
05 ffff8508`3b1de858 fffff807`20c3e364     0xfffff807`1eb90000
rax=0000000000000001 rbx=0000000000000100 rcx=0000000000000012
rdx=0000000000000000 rsi=0000000000000000 rdi=0000000000000000
rip=fffff8071eb90000 rsp=ffff85083b1de858 rbp=0000000000000000
 r8=0000000000000000  r9=0000000000000000 r10=0000000000000000
r11=0000000000000000 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0         nv up di ng nz na po nc
cs=0010  ss=0018  ds=002b  es=002b  fs=0053  gs=002b             efl=00000086
fffff807`1eb90000 2f              ???

This address in the call stack is thus an error, and we can tell that by looking at the addresses of the other function calls in this call stack; they are all in the range fffff807`20c46ee8 to fffff807`20bb48af, so this apparent call to
fffff807`1eb90000 is clearly in error, because it's well outside the range of addresses that we're calling.

Since all other function calls that precede it are Microsoft functions, this cannot be caused by a rogue third-party driver and thus must be a hardware error, and a RAM error is by far the most likely.

I know I've gone on at length here, but I want to clearly illustrate why I think that flaky RAM is the most likely cause of this BSOD.

I still think your wisest next step is to remove the stick of RAM that is removable and see whether the BSODs cease. Even if they don't, it's possible that any memory failure is in the soldered-in stick, in which case you have a problem.

 

[WrongRookie]

I am probably not the one with the most tech smarts in this but seeing as the mixing of ram is most likely the culprit and that it's more of a hardware problem, couldn't you send the laptop to RMA and get it under warranty if it's still new?

Because at this rate whatever tests you do keeps giving you fatal errors. In which case it's really best to send it to RMA or get it fixed.

 

[ubuysa]

@WrongRookie has a good point, but your problem will be providing concrete evidence of a hardware fault. One dump probably won't be enough. One thing you can do however, is to start Windows in Safe Mode and see whether it BSODs in that mode. Use the start Windows in 'Safe Mode without networking' option at first, if it won't BSOD in that mode, then try again using the 'Safe Mode with networking' option - you'll be able to use more features in that mode.

Safe Mode is a stripped-down version of Windows where only critical Windows features and services are loaded. Typically, no third-party drivers are loaded, and only those Windows components required to boot and run Windows are loaded. Because of this many devices may not work properly (or at all). Your display will be very low resolution for example, because you'll be using only the basic Windows display driver. (In Safe Mode with networking your third-party network adapter drivers will be loaded).

You will NOT be able to do any useful work in Safe Mode, the purpose of running it is only to see whether it will BSOD, so please leave it running in Safe Mode for long enough to have usually had a BSOD. You said 30-40 minutes in your OP, so leave it running in Safe Mode for a couple of hours.

If t BSODs in safe Mode then you have excellent evidence that this is a hardware problem, along with the 0x20001 bugcheck dump from here of course. If you cannot make it BSOD at all in Safe Mode then the problem may not be hardware and we'll try a different tack to locate the problem.

 

[WrongRookie]

@WrongRookie has a good point, but your problem will be providing concrete evidence of a hardware fault. One dump probably won't be enough. One thing you can do however, is to start Windows in Safe Mode and see whether it BSODs in that mode. Use the start Windows in 'Safe Mode without networking' option at first, if it won't BSOD in that mode, then try again using the 'Safe Mode with networking' option - you'll be able to use more features in that mode.

Safe Mode is a stripped-down version of Windows where only critical Windows features and services are loaded. Typically, no third-party drivers are loaded, and only those Windows components required to boot and run Windows are loaded. Because of this many devices may not work properly (or at all). Your display will be very low resolution for example, because you'll be using only the basic Windows display driver. (In Safe Mode with networking your third-party network adapter drivers will be loaded).

You will NOT be able to do any useful work in Safe Mode, the purpose of running it is only to see whether it will BSOD, so please leave it running in Safe Mode for long enough to have usually had a BSOD. You said 30-40 minutes in your OP, so leave it running in Safe Mode for a couple of hours.

If t BSODs in safe Mode then you have excellent evidence that this is a hardware problem, along with the 0x20001 bugcheck dump from here of course. If you cannot make it BSOD at all in Safe Mode then the problem may not be hardware and we'll try a different tack to locate the problem.

Thing is.. they already confirmed that its BSODing in safe mode within 10 minutes...

I tried in safe mode but it froze in 10 mins. The entire screen freezes, even the video I was playing on youtube so its not just touchpad i think.

So yeah...its pretty much RMA.