Re: global_reclaim() and code documentation (was: Re: [PATCH v3 3/3] mm: vmscan: ignore non-LRU-based reclaim in memcg reclaim

[Yu Zhao]

On Wed, Apr 5, 2023 at 2:01 PM Johannes Weiner <hannes@xxxxxxxxxxx> wrote:
>
> On Fri, Mar 31, 2023 at 12:30:11AM -0700, Yosry Ahmed wrote:
> > On Fri, Mar 31, 2023 at 12:25 AM Yu Zhao <yuzhao@xxxxxxxxxx> wrote:
> > > On Fri, Mar 31, 2023 at 1:08 AM Yosry Ahmed <yosryahmed@xxxxxxxxxx> wrote:
> > >
> > > ...
> > >
> > > > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > > > index a3e38851b34ac..bf9d8e175e92a 100644
> > > > --- a/mm/vmscan.c
> > > > +++ b/mm/vmscan.c
> > > > @@ -533,7 +533,35 @@ EXPORT_SYMBOL(mm_account_reclaimed_pages);
> > > >  static void flush_reclaim_state(struct scan_control *sc,
> > > >                                 struct reclaim_state *rs)
> > > >  {
> > > > -       if (rs) {
> > > > +       /*
> > > > +        * Currently, reclaim_state->reclaimed includes three types of pages
> > > > +        * freed outside of vmscan:
> > > > +        * (1) Slab pages.
> > > > +        * (2) Clean file pages from pruned inodes.
> > > > +        * (3) XFS freed buffer pages.
> > > > +        *
> > > > +        * For all of these cases, we have no way of finding out whether these
> > > > +        * pages were related to the memcg under reclaim. For example, a freed
> > > > +        * slab page could have had only a single object charged to the memcg
> > > > +        * under reclaim. Also, populated inodes are not on shrinker LRUs
> > > > +        * anymore except on highmem systems.
> > > > +        *
> > > > +        * Instead of over-reporting the reclaimed pages in a memcg reclaim,
> > > > +        * only count such pages in system-wide reclaim. This prevents
> > > > +        * unnecessary retries during memcg charging and false positive from
> > > > +        * proactive reclaim (memory.reclaim).
> > >
> > > What happens when writing to the root memory.reclaim?
> > >
> > > > +        *
> > > > +        * For uncommon cases were the freed pages were actually significantly
> > > > +        * charged to the memcg under reclaim, and we end up under-reporting, it
> > > > +        * should be fine. The freed pages will be uncharged anyway, even if
> > > > +        * they are not reported properly, and we will be able to make forward
> > > > +        * progress in charging (which is usually in a retry loop).
> > > > +        *
> > > > +        * We can go one step further, and report the uncharged objcg pages in
> > > > +        * memcg reclaim, to make reporting more accurate and reduce
> > > > +        * under-reporting, but it's probably not worth the complexity for now.
> > > > +        */
> > > > +       if (rs && !cgroup_reclaim(sc)) {
> > >
> > > To answer the question above, global_reclaim() would be preferred.
> >
> > Great point, global_reclaim() is fairly recent. I didn't see it
> > before. Thanks for pointing it out. I will change it for v4 -- will
> > wait for more feedback before respinning.
>
> I didn't realize it came back, I deleted it a while ago:
>
> commit b5ead35e7e1d3434ce436dfcb2af32820ce54589
> Author: Johannes Weiner <hannes@xxxxxxxxxxx>
> Date:   Sat Nov 30 17:55:40 2019 -0800
>
>     mm: vmscan: naming fixes: global_reclaim() and sane_reclaim()
>
>     Seven years after introducing the global_reclaim() function, I still have
>     to double take when reading a callsite.  I don't know how others do it,
>     this is a terrible name.
>
>     Invert the meaning and rename it to cgroup_reclaim().
>
> Could you shed some light on why it was brought back? It's not clear
> to me from the changelog in a579086c99ed70cc4bfc104348dbe3dd8f2787e6.

Sorry about the confusion. I was asking Yosry to post an RFC to clear that up.

> We also now have this:
>
> static bool cgroup_reclaim(struct scan_control *sc)
> {
>         return sc->target_mem_cgroup;
> }
>
> static bool global_reclaim(struct scan_control *sc)
> {
>         return !sc->target_mem_cgroup || mem_cgroup_is_root(sc->target_mem_cgroup);
> }
>
> The name suggests it's the same thing twice, with opposite
> polarity. But of course they're subtly different, and not documented.
>
> When do you use which?

The problem I saw is that target_mem_cgroup is set when writing to the
root memory.reclaim. And for this case, a few places might prefer
global_reclaim(), e.g., in shrink_lruvec(), in addition to where it's
being used.

If this makes sense, we could 1) document it (or rename it) and apply
it to those places, or 2) just unset target_mem_cgroup for root and
delete global_reclaim(). Option 2 might break ABI but still be
acceptable.

If we don't want to decide right now, I can rename global_reclaim() to
root_reclaim() and move it within #ifdef CONFIG_LRU_GEN and probably
come back and revisit later.