[SOLVED] WHEA_UNCONTROLLABLE_ERROR on Win 10

[NUWANDA01]

Hi,
I've been having BSOD with the following error WHEA_UNCORRECTABLE_ERROR. I have had this BSOD with the exact same error multiple times since yesterday. It only seems to happen when the system is under stress during gaming or otherwise.
You can look at this for reference. This is how the screen looks during the crash.

Unfortunately, windows can't make a dump file for some reason. Searching through the event viewer I always find this error .
Windows not creating a dump is only an issue for this specific error (Make sure to read the updates below)

My laptop is a MSI GL63 9SEK gaming laptop with the following specifications.

Core i7 9750h (Undervolted -110, stable and never had any crashes)
RTX 2060
16gb DDR4 2666Mhz
Toshiba 256gb SSD NVMe + 1TB (7200RPM)

Alternately, you can find all the system specifications (to the full extent, I believe) here

Things I have tried so far :-

1. Reinstall windows.
2. Perform system repair through previous images and sfc/ scannow.
3. Update to the Latest drivers and also, moving back to the old ones.
4. Stress test on both storage devices (Had my HDD replaced last year). However, I find it odd that crystaldiskinfo shows SSD COMPOSITE TEMPERATURE. I'm not sure what exactly this means but I have never had this show up before. My ssd always runs hot, averaging 60 °c to 75 °c.
5. Removing undervolt on CPU (ThrottleStop)
6. Running MemTest64 to find bad memory.

Thanks

Update #1 I had BSOD with the same error as soon as my pc restarted from the previous BSOD. This changes the occurrence of the error from under any kind of load to just sitting idle. Also, I forgot to mention it originally but while restarting from BSOD my PC boots up in bios all the time. During this, I cannot exit the bios. It usually takes me holding down the power button to force shutdown (atleast thrice!!) to get out of this infinite loop of booting into bios.

Update #2 had another BSOD this time with the following error:

UNEXPECTED KERNEL MODE TRAP you can find a minidump for this specific crash here

Frequency of the BSOD has increased a lot.

 

[uWebb429]

Removing undervolt on CPU

Make sure the FIVR monitoring table shows +0.0000 in the Offset column from top to bottom. If you were undervolting before, you have to specifically tell ThrottleStop to reset your voltages. The FIVR monitoring table will confirm this.

 

[NUWANDA01]

Make sure the FIVR monitoring table shows +0.0000 in the Offset column from top to bottom. If you were undervolting before, you have to specifically tell ThrottleStop to reset your voltages. The FIVR monitoring table will confirm this.

Hi,
Thanks for the reply.
I can confirm that FIVR shows +0.0000 values

 

[Karadjgne]

Core i7 9750h (Undervolted -110

Says everything. Whea Errors are almost always due to lack of voltage at the cpu. Microsoft has a bad habit of changing things slightly, especially when partnered with Intel, who can and does slap in microcode inside Windows updates to 'fix' performance issues or bugs with the way a cpu works.

Those tweeks to performance will be based on a stock cpu with stock voltages and stock LLC etc, so your undervolt to bring the wattage down can very likely be no longer stable with certain instruction sets, but perfectly stable with others.

 

[NUWANDA01]

Says everything. Whea Errors are almost always due to lack of voltage at the cpu. Microsoft has a bad habit of changing things slightly, especially when partnered with Intel, who can and does slap in microcode inside Windows updates to 'fix' performance issues or bugs with the way a cpu works.

Those tweeks to performance will be based on a stock cpu with stock voltages and stock LLC etc, so your undervolt to bring the wattage down can very likely be no longer stable with certain instruction sets, but perfectly stable with others.

Hi,
I have already tried removing undervolt from my CPU. Unfortunately, that did not fix the BSOD.

 

[Karadjgne]

Did you remove the entire undervolt? There's multiple hidden settings in bios that get changed when you make a user accessible change. So when you applied the undervolt, some of those hidden settings had to also makes changes, happened as soon as you rebooted for the first time. That keeps all the voltages in correct proportions. But, by adding voltage when removing the undervolt, may not necessarily change those settings back up, if the cpu decides they are within tolerances.

For instance, if Intel says vccin needs to be lower than vcore, and vccin is 1.2v and vcore is 1.3v and you apply an undervolt dropping vcore to 1.1v, bios will reset vccin to 1.0v instead. Adding voltage back up to stock vcore of 1.3v still means vccin is within tolerances, it's still lower than vcore, so doesn't change.

I'd suggest you eliminate any software settings that change voltage / clock / bios overrides, and pull the battery/cmos battery and wait out a bios reset. It's the only way to reset the non-accessible bios settings back to factory stock.