Re: [PATCH v2 2/2] vmscan,cgroup: apply mems_effective to reclaim

[Waiman Long <llong@redhat.com>]

On 4/18/25 11:27 PM, Gregory Price wrote:
> On Fri, Apr 18, 2025 at 05:06:20PM -1000, Tejun Heo wrote:
>> Hello,
>>
>> On Thu, Apr 17, 2025 at 11:13:52PM -0400, Gregory Price wrote:
>> ...
>>> +static inline bool mem_cgroup_node_allowed(struct mem_cgroup *memcg, int nid)
>>> +{
>>> +	return memcg ? cgroup_node_allowed(memcg->css.cgroup, nid) : true;
>>> +}
>>> +
>> ...
>>> +bool cgroup_node_allowed(struct cgroup *cgroup, int nid)
>>> +{
>>> +	return cpuset_node_allowed(cgroup, nid);
>>> +}
>> ...
>>> +bool cpuset_node_allowed(struct cgroup *cgroup, int nid)
>>> +{
>> What does the indirection through cgroup_node_allowed() add? Why not just
>> call cpuset directly?
>>
> This is an artifact of me trying to figure out how to get this to build
> with allconfig (matrix of CPUSET and MEM_CGROUP).
>
> I think you're right, I can probably drop it.  I was trying to write :
>
> bool cpuset_node_allowed(struct cpuset *cs, int nid);

The cpuset structure isn't exposed externally. So you can't use cpuset 
from outside cpuset.c. Passing the cgroup structure is the right approach.

Cheers,
Longman

>
> and just couldn't do it, so eventually landed on passing the cgroup into
> the cpuset function, which means I think I can drop the indirection now.
>
> Will push it and see if allconfig builds.
>
> Thanks
>
> ~Gregory
>