AMD EPYC 7351P not working

[cyboerg42]

System: HPE ProLiant DL325 Gen10 Kernel: 5.0.15-1-pve x86_64 Processor: AMD EPYC 7351P 16-Core Processor (23/1/2) Distro : Proxmox 6.0 (Debian Buster)

if you need more info, just toss a message into my direction.

dmesg error :

[341189.548252] ACPI Error: No handler for Region [POWR] (000000008241729d) [IPMI] (20181213/evregion-132)
[341189.549001] ACPI Error: Region IPMI (ID=7) has no handler (20181213/exfldio-265)
[341189.549680] No Local Variables are initialized for Method [_PMM]
[341189.549680] No Arguments are initialized for method [_PMM]
[341189.549682] ACPI Error: Method parse/execution failed \_SB.PMI0._PMM, AE_NOT_EXIST (20181213/psparse-531)
[341189.550404] ACPI Error: AE_NOT_EXIST, Evaluating _PMM (20181213/power_meter-338)

sensors with k10temp :

k10temp-pci-00d3
Adapter: PCI adapter
Tdie:         +39.8°C  (high = +70.0°C)
Tctl:         +39.8°C  

k10temp-pci-00c3
Adapter: PCI adapter
Tdie:         +44.2°C  (high = +70.0°C)
Tctl:         +44.2°C  

ixgbe-pci-c401
Adapter: PCI adapter
loc2:         +34.0°C  (high = +105.0°C, crit = +95.0°C)

k10temp-pci-00db
Adapter: PCI adapter
Tdie:         +44.2°C  (high = +70.0°C)
Tctl:         +44.2°C  

k10temp-pci-00cb
Adapter: PCI adapter
Tdie:         +40.4°C  (high = +70.0°C)
Tctl:         +40.4°C  

power_meter-acpi-0
Adapter: ACPI interface
power1:        0.00 W  (interval = 300.00 s)

sensors with zenpower :

ixgbe-pci-c401
Adapter: PCI adapter
loc2:         +32.0°C  (high = +105.0°C, crit = +95.0°C)

power_meter-acpi-0
Adapter: ACPI interface
power1:        0.00 W  (interval = 300.00 s)

[cyboerg42]

acpidump > acpi_tables_dl325_g10_7351P.txt

acpi_tables_dl325_g10_7351P.txt

maybe this helps a bit :)

[cyboerg42]

kernel_smn_support = 1
0005a008 = 00000002
0005a00c = 0161004e
0005a010 = 01f70000
000598bc = 0fff00ff
0005994c = 00000000
00059954 = 00000a88
00059958 = 00000a84
0005995c = 00000a78
kernel_smn_support = 1
0005a008 = 00000002
0005a00c = 01400008
0005a010 = 01f70000
000598bc = 0fff00ff
0005994c = 00000000
00059954 = 00000000
00059958 = 00000000
0005995c = 00000000
kernel_smn_support = 1
0005a008 = 00000002
0005a00c = 00000000
0005a010 = 00000000
000598bc = 0fff00ff
0005994c = 00000000
00059954 = 00000000
00059958 = 00000000
0005995c = 00000000
kernel_smn_support = 1
0005a008 = 00000002
0005a00c = 00000000
0005a010 = 00000000
000598bc = 0fff00ff
0005994c = 00000000
00059954 = 00000000
00059958 = 00000000
0005995c = 00000000

zenpower debug output

[cyboerg42]

oh wow, now it's working. sorry for bothering... :)

zenpower-pci-00d3
Adapter: PCI adapter
SVI2_Core:    +1.55 V  
SVI2_SoC:     +1.55 V  
Tdie:         +30.8°C  (high = +70.0°C)
Tctl:         +30.8°C  
SVI2_P_Core:   0.00 W  
SVI2_P_SoC:    0.00 W  
SVI2_C_Core:  +0.00 A  
SVI2_C_SoC:   +0.00 A  

zenpower-pci-00c3
Adapter: PCI adapter
SVI2_Core:    +0.94 V  
SVI2_SoC:     +0.01 V  
Tdie:         +31.8°C  (high = +70.0°C)
Tctl:         +31.8°C  
SVI2_P_Core:  78.48 W  
SVI2_P_SoC:    0.00 W  
SVI2_C_Core: +83.14 A  
SVI2_C_SoC:   +0.00 A  

ixgbe-pci-c401
Adapter: PCI adapter
loc2:         +32.0°C  (high = +105.0°C, crit = +95.0°C)

zenpower-pci-00db
Adapter: PCI adapter
SVI2_Core:    +1.55 V  
SVI2_SoC:     +1.55 V  
Tdie:         +29.8°C  (high = +70.0°C)
Tctl:         +29.8°C  
SVI2_P_Core:   0.00 W  
SVI2_P_SoC:    0.00 W  
SVI2_C_Core:  +0.00 A  
SVI2_C_SoC:   +0.00 A  

zenpower-pci-00cb
Adapter: PCI adapter
SVI2_Core:    +1.15 V  
SVI2_SoC:     +0.01 V  
Tdie:         +31.5°C  (high = +70.0°C)
Tctl:         +31.5°C  
SVI2_P_Core:  13.15 W  
SVI2_P_SoC:    0.00 W  
SVI2_C_Core: +12.47 A  
SVI2_C_SoC:   +0.00 A  

power_meter-acpi-0
Adapter: ACPI interface
power1:        0.00 W  (interval = 300.00 s)

best regards, and thanks for this awesome module!

[ocerman]

There has been some reports before, that a restart is needed for some people. This was probably that case.

I am happy that it works. Even though I can see that some 0 values are being reported, which is propably not correct.

[cyboerg42]

Hm, i just reloaded the kernel module - and that fixed it somehow. No restart needed.

I thought, that those are the disabled dies on the EPYC Chip - but by the amount of L3, I guess all dies are active. Weird.