Re: [PATCH v2 0/2] platform/x86/amd/pmc: Add support for AMD 1Ah Family 80h SoC

[Shyam Sundar S K <Shyam-sundar.S-k@amd.com>]

On 6/2/2026 19:28, Mario Limonciello wrote:
> 
> 
> On 6/1/26 06:21, Shyam Sundar S K wrote:
>> The series introduces the necessary device identification, refactors
>> SMU
>> mailbox register handling for better maintainability, and configures
>> the
>> SoC-specific register offsets required for SMU communication.
>>
>> Patch 1 refactors the per-SoC configuration by introducing
>> amd_pmc_cpu_info
>> struct that consolidates SMU register offsets, IP block table, and
>> OS hint.
>> The individual smu_msg, num_ips, and ips_ptr fields in amd_pmc_dev are
>> replaced with a single cpu_info pointer. Per-SoC instances are embedded
>> directly as driver_data in the PCI ID table via PCI_DEVICE_DATA(),
>> eliminating the amd_pmc_get_ip_info() and amd_pmc_get_os_hint() switch
>> statements. The AMD_CPU_ID_* macros are renamed to
>> PCI_DEVICE_ID_AMD_CPU_ID_* with compatibility aliases retained.
>>
>> Patch 2 adds support for the AMD 1Ah Family 80h SoC by defining its
>> distinct SMU register offsets (msg: 0xA10, arg: 0xA18, rsp: 0xA14),
>> adding
>> the amd_1ah_m80_cpu_info instance, and wiring it into the PCI ID table,
>> amd_pmc_set_cpu_info(), and amd_pmc_idlemask_read(). The corresponding
>> ACPI ID AMDI000C is also added.
>>
>> v2:
>>   - Consolidate 5-patch series into 2 patches
>>   - Use PCI_DEVICE_DATA() and maintain per SoC data.
>>   - Address other remarks as per Ilpo comments in v1
>>
>> Shyam Sundar S K (2):
>>    platform/x86/amd/pmc: Use per-SoC cpu_info struct for SMU mailbox
>> and
>>      IP info
>>    platform/x86/amd/pmc: Add PMC driver support for AMD 1Ah M80H SoC
>>
>>   drivers/platform/x86/amd/pmc/pmc.c | 176 +++++++++++++++++++
>> +---------
>>   drivers/platform/x86/amd/pmc/pmc.h |  75 ++++++++----
>>   2 files changed, 177 insertions(+), 74 deletions(-)
>>
> 
> Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
> 
> As a general statement though about patch 1:
> 
> If there ends up being too many "Special case" blocks in the future it
> may be better to just switch to explicit CPU ID range matching later.
> There are some other drivers that do this like k10temp:
> 
> https://github.com/torvalds/linux/
> blob/6f3ed7fec72fc8979b2a8c7219c0a9fcfc8d07b5/drivers/hwmon/
> k10temp.c#L490

Sure. Apologies, I missed this. Can you please add your Reviewed-by
tag on v3 which I just submitted now?

Thanks,
Shyam