Running a couple of supermicro H8SGL-F recently upgraded processors (to 6300s)…
The 6338p upgrade went flawless, but on the other with a 6380
System:
Host: cassini Kernel: 6.12.12-1-lts arch: x86_64 bits: 64
Console: pty pts/0 Distro: EndeavourOS
Machine:
Type: Server System: Supermicro product: H8SGL v: 1234567890 serial: <superuser required>
Mobo: Supermicro model: H8SGL v: 1234567890 serial: <superuser required>
BIOS: American Megatrends v: 3.5b date: 03/18/2016
CPU:
Info: 8-core AMD Opteron 6380 [MT MCP] speed (MHz): avg: 1400 min/max: 1400/2500
...
Network:
Device-1: Intel 82599ES 10-Gigabit SFI/SFP+ Network driver: ixgbe
Device-2: Intel 82574L Gigabit Network driver: e1000e
Device-3: Intel 82574L Gigabit Network driver: e1000e
Drives:
Local Storage: total: raw: 473.24 GiB usable: 457.12 GiB used: 50.62 GiB (11.1%)
Info:
Memory: total: 128 GiB note: est. available: 125.76 GiB used: 8.74 GiB (6.9%)
Processes: 397 Uptime: 23h 10m Init: systemd Shell: Bash inxi: 3.3.37
I’m seeing the following log entry:
2025-02-09T00:59:13+01:00 kernel: mce: [Hardware Error]: Machine check events logged
2025-02-09T00:59:13+01:00 kernel: [Hardware Error]: Corrected error, no action required.
2025-02-09T00:59:13+01:00 kernel: [Hardware Error]: CPU:0 (15:2:0) MC4_STATUS[-|CE|MiscV|AddrV|-|CECC|-]: 0x9c3c44e0011f010b
2025-02-09T00:59:13+01:00 kernel: [Hardware Error]: Error Addr: 0x000000000001f140
2025-02-09T00:59:13+01:00 kernel: [Hardware Error]: MC4 Error (node 0): ECC Error in the Probe Filter directory.
2025-02-09T00:59:13+01:00 kernel: [Hardware Error]: cache level: L3/GEN, tx: GEN, mem-tx: GEN
Even though there is ‘no action required’, there’s nothing I hate more is finding turds in the log files.
The 6380 is one from a pair I got on Ebay that were initially installed in a dual processor system.
The only perceivable issue encountered thus far is upon occasion when issuing a reboot, the system doesn’t come back around and I have to do a manual reset.
My question is, if anybody has happened to come across this or knows about any possible issue, is this potentially a BIOS issue concerning supermicro, or an AMD processor issue (or non-issue). Apparently it involves HT-Assist aka Probe Filter.
