Kernel crash 5.4.53

Just checking if anyone else has had kernel crashes on the latest -current kernel. I was working on my machine a short while again when the screen just went dark and then the machine rebooted. Some details:

- Slackware-current updated to this afternoon
- nVidia binary 440.82
- Asus B450M Prime bios 1804 (I'll be updating to the latest tomorrow)
- Ryzen 5 1500x
- 24GB memory across 4 slots (all Kingston)
- WD Blue SSSD 1TB, WD Black 1TB, Sandisk Ultra II 240GB

The error was:

Quote:

Jul 24 20:18:12 googly kernel: [ 0.104038] mce: [Hardware Error]: CPU 1: Machine Check: 0 Bank 5: bea0000000000108 Jul 24 20:18:12 googly kernel: [ 0.104038] mce: [Hardware Error]: TSC 0 ADDR 1ffffa216d05e MISC d012000100000000 SYND 4d000000 IPID 500b000000000 Jul 24 20:18:12 googly kernel: [ 0.104041] mce: [Hardware Error]: PROCESSOR 2:800f11 TIME 1595621869 SOCKET 0 APIC 2 microcode 8001138 Jul 24 20:18:12 googly kernel: [ 0.104058] #2 #3 #4 #5 #6 #7

I had some of these a while ago but the machine has been stable recently. There are quite a few of these issues mentioned in google searches. One fix mentioned was:

Quote:

I was having the same problem, after adding the kernel parameter idle=nomwait the problem is gone (ryzen 1600x)

There has also been some talk of this being an amdgpu driver issue:

Quote:

Thanks for the link. I tried the kernel parameters from the thread (amdgpu.ppfeaturemask=0xffffbffb amdgpu.dpm=0). So far zero crashes in 3 days. It's a pretty big performance hit, so not really usable long term. I went from 100+ fps and crashing 2-3 times a night, to 20 fps with no crashes in TW:Warhammer2. But still useful for identifying the gpu as the probable culprit in my case also.

... but I'm running NVidia, not AMDGPU.

I'd be interested to hear of any similar experiences.

Robby