H3C 4900G3 prompts error when configured in memory mode

[3257085208]

My server is H3C 4900G3
The CPU is 8259CL
When 4 DCPMMs are set to Memory Mode 100% in BIOS, a prompt appears MemoryMode capacity is unusable with the current ly selected platform BIOS volatile mode.
But I searched through the BIOS and couldn't find the volatile mode
I directly boot it up and enter it into the system
[ root@localhost ~]# sudo ipmctl show -dimm
DimmID | Capacity | LockState | HealthState | FWVersion

0x0101 | 126.422 GiB | Disabled | Healthy | 01.02.00.5444
0x0111 | 126.422 GiB | Disabled | Healthy | 01.02.00.5444
0x1001 | 126.422 GiB | Disabled | Healthy | 01.02.00.5444
0x1011 | 126.422 GiB | Disabled | Healthy | 01.02.00.5444
input
[ root@localhost ~]# ipmctl create -dimm 0x0101,0x0111,0x1001,0x1011 -goal MemoryMode=100
Create region configuration goal failed: Error 113 - Platform does not support Memory Mode
[ root@localhost ~]# ipmctl create -goal MemoryMode=100
Create region configuration goal failed: Error 113 - Platform does not support Memory Mode
Do you know what the problem is?

[3257085208]

https://images.080112.xyz/file/1753465725864_262d7813-4a05-4475-b096-dae4e369459a.png

[sscargal]

@3257085208 Do you have any DDR modules in the server? It's not clear from ipmctl show -dimm whether that's the complete output or you only showed the PMem modules.

Can you git clone https://github.com/sscargal/pmemchk and run the script, then post the terminal STDOUT output. Thanks.

[3257085208]

@sscargal
My configuration is 4 16G DDR4 2666+4 128G DCPMM
The following is the output of ./pmemchk
[root@localhost pmemchk]# ./pmemchk

Starting PMem Checker
pmemchk Version 0.1.0
Started: 2025年 07月 26日 星期六 05:30:36 CST

Operating System: TencentOS Server 3.1 (Final)
Kernel Version : 5.4.119-19.0009.44
Using NDCTL command: /usr/bin/ndctl
NDCTL version: 71.1
Using IPMCTL command: /usr/bin/ipmctl
IPMCTL version: 03.00.00.499
Using DAXCTL command: /usr/bin/daxctl
DAXCTL version: 71.1
INFO: cxl command not found! Use -c to specify the location.

Starting data collector

Collecting command outputs
100% (10 of 10) [====================]
Collecting files
100% (5 of 5) [====================]
Skipping CXL collector. 'cxl' command not found
Running DAXCTL Collector
100% (5 of 5) [====================]
Running IPMCTL Collector
100% (32 of 32) [====================]
Post-Processing Collected Data
100% (5 of 5) [====================]
Running NDCTL Collector
100% (8 of 8) [====================]

Data collector completed

=======================================================================
Starting analysis of the data

./analyzer/linux/full_file_system:行33: [[: 已用: 语法错误: 需要操作数 (错误符号是 "已用")
./analyzer/linux/full_file_system:行36: [[: 已用: 语法错误: 需要操作数 (错误符号是 "已用")
./analyzer/linux/full_file_system:行41: [[: 已用: 语法错误: 需要操作数 (错误符号是 "已用")
[ PASSED ] linux_check_full_file_systems : All file systems are <75% full.
[ PASSED ] linux_check_memmap : No 'memmap' entries found in the Linux boot command line (/proc/cmdline)
[ PASSED ] optane_check_dimm_fwupdatestatus : All PMem modules have no staged Firmware Update
[ PASSED ] optane_check_cpu_is_supported : CPU Model '' supports Intel Optane Persistent Memory Modules.
[ PASSED ] optane_check_dimm_extended_adr_enabled : All PMem modules have ExtendedAdrEnabled switch disabled
[ PASSED ] optane_check_media_temperature_injection_enabled : All PMem modules have thier MediaTemperature Injection Disabled
[ PASSED ] optane_check_dimm_media_temperature_injections_count : All PMem modules have 0 Media Temperature Injection Counter
[ INFO ] optane_check_dimm_ppc_extended_adr_enabled : One or more PMem modules have PpcExtendedAdrEnabled=False (Disabled)
[ PASSED ] optane_check_dimm_software_triggers_enabled_details : All PMem modules have software trigger disabled
[ PASSED ] optane_check_dimm_ait_dram_enabled : One or more PMem modules have AitDramEnabled=False (Disabled)
[ PASSED ] optane_check_dimm_arsstatus : All PMem module ARSStatus is Completed
[ PASSED ] optane_check_dimm_boot_status : All PMem modules have successful BootStatus
[ PASSED ] optane_check_dimm_capacity : All PMem Modules have the same capacity.
[ PASSED ] optane_check_dimm_configurationstatus : All PMem module ConfigurationStatus is 'Valid'
[ PASSED ] optane_check_dimm_error_injection_enabled : All PMem modules have error injection disabled
[ PASSED ] optane_check_dimm_fw_error_count : All PMem Modules have less than 10 FWError Counts
[ PASSED ] optane_check_dimm_firmware_version : All PMem Modules have the same firmware.
[ PASSED ] optane_check_dimm_health_status : All PMem Modules are Healthy
[ PASSED ] optane_check_dimm_lockstate : All PMem Module Security LockState is good
[ PASSED ] optane_check_dimm_manageabilitystate : One or more PMem modules reported ManageabilityState as Unmanageable
[ INFO ] optane_check_dimm_memoryboostfeature : Memory Bandwidth Boot Feature is Disabled
[ INFO ] optane_check_dimm_overwrite_status : 1 Overwrite Status found:
[ INFO ] optane_check_dimm_overwrite_status : Unknown = 0x0101,0x0111,0x1001,0x1011
[ PASSED ] optane_check_dimm_overwrite_status : All PMem modules have Overwrite Status == '0
[ PASSED ] optane_check_dimm_packagesparesavailable : All PMem modules have 1 PackageSparesAvailable
[ PASSED ] optane_check_dimm_packagesparingcapable : All PMem modules have PackageSparingCapable = True
[ PASSED ] optane_check_dimm_packagesparingenabled : All PMem modules have PackageSparingEnabled = True
[ PASSED ] optane_check_dimm_partnumber : All PMem modules have the same Part Number
[ INFO ] optane_check_dimm_percentage_remaining : DimmID 0x0101 has 97% life remaining
[ INFO ] optane_check_dimm_percentage_remaining : DimmID 0x0111 has 97% life remaining
[ INFO ] optane_check_dimm_percentage_remaining : DimmID 0x1011 has 97% life remaining
[ PASSED ] optane_check_dimm_percentage_remaining : All PMem Modules have 100% life remaining
[ PASSED ] optane_check_dimm_poision_error_injection_and_clear_counter : All PMem modules have equal PosionErrorInjectionCounter and PoisionErrorCounter
[ PASSED ] optane_check_dimm_population : Detected 4 PMem modules. Population looks good.
[ PASSED ] optane_check_dimm_reservedcapacity : No PMem modules have ReservedCapacity configured
[ PASSED ] optane_check_dimm_show_goal : There are no pending goals
[ PASSED ] optane_check_dimm_skuviolation : No PMem modules reported a SKU Violation
[ PASSED ] optane_check_dimm_software_trigger_counter : All PMem modules have Software Trigger Counter == 0
[ PASSED ] optane_check_dimm_thermalthrottlelosspercent : All PMem modules have zero ThermalThrottleLossPercent events
[ PASSED ] optane_check_dimm_viral_policy : All PMem modules have Viral Policy disabled
[ PASSED ] optane_check_dimm_viral_state : All PMem modules are not ViralState == 0
[ INFO ] optane_check_masterpassphraseenabled : Master Passphrase is Disabled
./analyzer/optane/check_region_capacity:行33: REGION_CAPACITY[${Capacity}]: 数组下标不正确
./analyzer/optane/check_region_freecapacity:行32: REGION_FREECAPACITY[${FreeCapacity}]: 数组下标不正确
./analyzer/optane/check_region_health:行37: REGION_HEALTH[${HealthState}]: 数组下标不正确
./analyzer/optane/check_region_persistentmemorytype:行33: REGION_TYPE[${PersistentMemoryType}]: 数组下标不正确

Data analysis completed

=======================================================================
Analysis Report Summary

[ PASSED ] = 34
[ FAILED ] = 0
[ SKIPPED ] = 1
[ INFO ] = 6
[ WARNING ] = 0
[ CRITICAL ] = 0
[ UNKNOWN ] = 0

PMem Checker Complete
Ended: 2025年 07月 26日 星期六 05:30:43 CST
Duration: 7 seconds
Results: ./pmemchk.localhost.localdomain.0726-0530
Logfile: ./pmemchk.localhost.localdomain.0726-0530/pmemchk.log

My operating system is TencentOS everything 3.1 he and Centos Stream 8 are very similar. Both belong to the RedHat series. I have successfully used the same system and memory combination in inspur SA 5112M5.

[sscargal]

Thanks @3257085208. There's nothing obvious from this output to point us at the cause. Given that the issue is reproducible in the BIOS, it leads us to conclude that this is a BIOS/Hardware issue. You can try factory resetting the PMem modules in case there's any lingering configuration in the LSA (Label Storage Area) that's causing configuration issues. Instructions can be found in How to Securely Erase Data on Intel® Optane™ Persistent Memory

[3257085208]

@sscargal
Is it effective for me to simply remove the BIOS battery and reinstall it? That might be more convenient

[3257085208]

I tried to input ipmctl delete -dimm and ipmctl delete -f -dimm -pcd
but the result prompted

[root@localhost ~]# ipmctl delete -dimm
    Delete the region configuration goal from one or more PMem modules.
    delete  [OPTIONS][ -dimm[DimmIDs]]  -goal [ -socket[SocketIDs]] 

[OPTIONS]
   [-help|-h] : Display Help for the command
   [-verbose|-v] : Change the Debug Level Message Display
   [-ddrt] : Used to specify DDRT as the desired transport protocol
   [-smbus] : Used to specify SMBUS as the desired transport protocol
   [-lpmb] : Used to specify large transport payload size
   [-spmb] : Used to specify small transport payload size
   [-output|-o(text|nvmxml)] : Changes the output format.
                    

    Clear the namespace LSA partition on one or more PMem modules.
    delete  [OPTIONS][ -dimm[DimmIDs]]  -pcd[Config] 

[OPTIONS]
   [-help|-h] : Display Help for the command
   [-verbose|-v] : Change the Debug Level Message Display
   [-ddrt] : Used to specify DDRT as the desired transport protocol
   [-smbus] : Used to specify SMBUS as the desired transport protocol
   [-lpmb] : Used to specify large transport payload size
   [-spmb] : Used to specify small transport payload size
   [-force|-f] : Suppresses the confirmation from the User to use this operation
   [-output|-o(text|nvmxml)] : Changes the output format.
                    

[root@localhost ~]# ipmctl delete -f -dimm -pcd

Clear Config partition(s) on PMem module 0x0101: Success
Clear Config partition(s) on PMem module 0x0111: Success
Clear Config partition(s) on PMem module 0x1001: Success
Clear Config partition(s) on PMem module 0x1011: Success


Data dependencies may result in other commands being affected. A system reboot is required before all changes will take effect.

[sscargal]

@sscargal Is it effective for me to simply remove the BIOS battery and reinstall it? That might be more convenient

No. That will only reset the BIOS. Data and configuration will still persist on the Optane Persistent Memory (PMem) modules.

ipmctl delete -f -dimm -pcd

Clear Config partition(s) on PMem module 0x0101: Success
Clear Config partition(s) on PMem module 0x0111: Success
Clear Config partition(s) on PMem module 0x1001: Success
Clear Config partition(s) on PMem module 0x1011: Success

This was successful and removed the PCD, which is one of the areas we keep the config. This will have factory reset the PMem modules.

After a reboot, does this solve the issue? The system should default to Memory Mode when the BIOS fails to detect an AppDirect configuration on any of the modules. You can check the same BIOS screen as before to see if the message has gone away, or ipmctl show -memoryresources to see if the system is in Memory Mode or App Direct.

• If yes, great. No further action is required.
• I no, then I recommend contacting H3C for additional hardware/BIOS support.

[3257085208]

@sscargal After Reboot, the memory is still 64G (DDR capacity) Enter free -h prompt

[root@localhost ~]# free -h
              total        used        free      shared  buff/cache   available
Mem:           62Gi       1.8Gi        59Gi        21Mi       979Mi        59Gi
Swap:            0B          0B          0B

Enter ipmctl show -memoryresources prompt

MemoryType   | DDR        | PMemModule  | Total       
=======================================================
 Volatile     | 64.000 GiB | 0.000 GiB   | 64.000 GiB
 AppDirect    | -          | 0.000 GiB   | 0.000 GiB
 Cache        | 0.000 GiB  | -           | 0.000 GiB
 Inaccessible | 0.000 GiB  | 505.689 GiB | 505.689 GiB
 Physical     | 64.000 GiB | 505.689 GiB | 569.689 GiB

[sscargal]

Thanks @3257085208

This still appears to be a hardware/BIOS issue. Please ensure your system is running the latest available BIOS update.

Here are some additional tasks you can try:

Check for existing goals:

ipmctl show -goals

If you see an unapplied goal, clear any failed or pending goals.

ipmctl delete -goal

The previous ipmctl delete -dimm pcd should have removed any goal config from the PMem modules themselves, allowing the BIOS to manage it.

If that works:

Set a new allocation goal.

For example, to set all PMem as AppDirect (persistent memory):

ipmctl create -goal PersistentMemoryType=AppDirect

Or, for 100% Memory Mode:

ipmctl create -goal MemoryMode=100