Re: [PATCH for 6.19 0/4] Revise the EM YNL spec to be clearer

[Donald Hunter]

> On 30 Dec 2025, at 09:44, Lukasz Luba <lukasz.luba@xxxxxxx> wrote:
> 
> Hi Changwoo,
> 
>> On 12/25/25 04:01, Changwoo Min wrote:
>> This patch set addresses all the concerns raised at [1] to make the EM YNL spec
>> clearer. It includes the following changes:
>> - Fix the lint errors (1/4).
>> - Rename em.yaml to dev-energymodel.yaml (2/4).  “dev-energymodel” was used
>>   instead of “device-energy-model”, which was originally proposed [2], because
>>   the netlink protocol name cannot exceed GENL_NAMSIZ(16). In addition, docs
>>   strings and flags attributes were added.
>> - Change cpus' type from string to u64 array of CPU ids (3/4).
>> - Add dump to get-perf-domains in the EM YNL spec (4/4). A user can fetch
>>   either information about a specific performance domain with do or information
>>   about all performance domains with dump.
>> This can be tested using the tool, tools/net/ynl/pyynl/cli.py, for example,
>> with the following commands:
>>   $> tools/net/ynl/pyynl/cli.py \
>>      --spec Documentation/netlink/specs/dev-energymodel.yaml \
>>      --dump get-perf-domains
>>   $> tools/net/ynl/pyynl/cli.py \
>>      --spec Documentation/netlink/specs/dev-energymodel.yaml \
>>      --do get-perf-domains --json '{"perf-domain-id": 0}'
>>   $> tools/net/ynl/pyynl/cli.py \
>>      --spec Documentation/netlink/specs/dev-energymodel.yaml \
>>      --do get-perf-table --json '{"perf-domain-id": 0}'
>>   $> tools/net/ynl/pyynl/cli.py \
>>      --spec Documentation/netlink/specs/dev-energymodel.yaml \
>>      --subscribe event  --sleep 10
>> [1] https://lore.kernel.org/lkml/CAD4GDZy-aeWsiY=-ATr+Y4PzhMX71DFd_mmdMk4rxn3YG8U5GA@xxxxxxxxxxxxxx/
>> [2] https://lore.kernel.org/lkml/CAJZ5v0gpYQwC=1piaX-PNoyeoYJ7uw=DtAGdTVEXAsi4bnSdbA@xxxxxxxxxxxxxx/
> 
> My apologies, I've missed those conversations (not the best season).
> 
> So what would be the procedure here for the review?
> Could Folks from netlink help here?

I will review, hopefully later today. 

What hardware can it be tested on?

> I will do my bit for the EM related stuff (to double-check them).
> 
> Regards,
> Lukasz