leoconforti

I've included what I think are relevant bios settings bellow: Under Platform Misc Configuration: "PCI Express Native Power Management" -> Enabled "Native ASPM" -> Disabled Under PCH - PCI Express: "DMI Link ASPM Control" -> Disabled "ASPM" -> Auto "L1 Substates" -> Disabled Under SA - PCI Express: "DMI ASPM" -> Disabled "DMI gen3 ASPM" -> Disabled "PEG - ASPM" -> Disabled "PCIE Express Clock Gating" -> Enabled

I even tried "vfio-pci.disable_idle_d3=1" in my global syslinux configuration but it made no difference

I have a windows 10 vm that is set to autostart. When unraid boots, the vm starts fine and I usually put it to sleep when I am done and wake it form sleep just fine. Sometimes it gets shutdown though, at which point I can not start the vm again and the vm logs say "Unable to power on device, stuck in D3." This happens regardless of wether the vm's gpu is bound to vfio or just passed through diagnostics-20240430-1449.zip

The fix common problems plugin tells me that "your server has detected hardware errors. You should install mcelog via the NerdPack plugin, post your diagnostics and ask for assistance on the Unraid forums" Here is the contents of /var/log/mcelog mcelog: failed to prefill DIMM database from DMI data Kernel does not support page offline interface Hardware event. This is not a software error. MCE 0 CPU 7 BANK 1 MISC 86 ADDR 14fb211c0 TIME 1664041479 Sat Sep 24 12:44:39 2022 MCG status: MCi status: Error overflow Uncorrected error Error enabled MCi_MISC register valid MCi_ADDR register valid Processor context corrupt SRAR MCA: Data CACHE Level-1 Write Error STATUS ff80000000000124 MCGSTATUS 0 MCGCAP c0e APICID e SOCKETID 0 MICROCODE f0 CPUID Vendor Intel Family 6 Model 158 Step 13 will run a memtest for 24hr next, but does anything here jump out?