2022-10-13 13:25:27

by Michal Hocko

[permalink] [raw]
Subject: Re: [External] Re: [RFC] mm: add new syscall pidfd_set_mempolicy()

On Thu 13-10-22 20:50:48, Zhongkun He wrote:
> > > Hi Michal
> > >
> > > Could we try to change the MPOL_F_SHARED flag to MPOL_F_STATIC to
> > > mark static mempolicy which cannot be freed, and mpol_needs_cond_ref
> > > can use MPOL_F_STATIC to avoid freeing the static mempolicy.
> >
> > Wouldn't it make more sense to get rid of a different treatment and
> > treat all memory policies the same way?
>
> I found a case, not sure if it makes sense. If there is no policy
> in task->mempolicy, the use of atomic_{inc,dec} can be skiped
> according to MPOL_F_STATIC. Atomic_{inc,dec} in hot path may reduces
> performance.

I would start with a simple conversion and do any potential
optimizations on top of that based on actual numbers. Maybe we can
special case default_policy to avoid reference counting a default (no
policy case). A simple check for pol == &default_policy should be
negligible.
--
Michal Hocko
SUSE Labs


2022-10-13 15:00:20

by Zhongkun He

[permalink] [raw]
Subject: Re: [External] Re: [RFC] mm: add new syscall pidfd_set_mempolicy()

> On Thu 13-10-22 20:50:48, Zhongkun He wrote:
>>>> Hi Michal
>>>>
>>>> Could we try to change the MPOL_F_SHARED flag to MPOL_F_STATIC to
>>>> mark static mempolicy which cannot be freed, and mpol_needs_cond_ref
>>>> can use MPOL_F_STATIC to avoid freeing the static mempolicy.
>>>
>>> Wouldn't it make more sense to get rid of a different treatment and
>>> treat all memory policies the same way?
>>
>> I found a case, not sure if it makes sense. If there is no policy
>> in task->mempolicy, the use of atomic_{inc,dec} can be skiped
>> according to MPOL_F_STATIC. Atomic_{inc,dec} in hot path may reduces
>> performance.
>
> I would start with a simple conversion and do any potential
> optimizations on top of that based on actual numbers. Maybe we can
> special case default_policy to avoid reference counting a default (no
> policy case). A simple check for pol == &default_policy should be
> negligible.

Got it, thanks for your reply and suggestions.