Re: [PATCH v1 03/12] mm: memcontrol: make lruvec lock safe when LRU pages are reparented

[Roman Gushchin]

On Wed, Aug 18, 2021 at 12:28:02PM +0800, Muchun Song wrote:
> On Wed, Aug 18, 2021 at 11:18 AM Roman Gushchin <guro@xxxxxx> wrote:
> >
> > On Sat, Aug 14, 2021 at 01:25:10PM +0800, Muchun Song wrote:
> > > The diagram below shows how to make the folio lruvec lock safe when LRU
> > > pages are reparented.
> > >
> > > folio_lruvec_lock(folio)
> > >     retry:
> > >       lruvec = folio_lruvec(folio);
> > >
> > >         // The folio is reparented at this time.
> > >         spin_lock(&lruvec->lru_lock);
> > >
> > >         if (unlikely(lruvec_memcg(lruvec) != folio_memcg(folio)))
> > >             // Acquired the wrong lruvec lock and need to retry.
> > >             // Because this folio is on the parent memcg lruvec list.
> > >             goto retry;
> > >
> > >         // If we reach here, it means that folio_memcg(folio) is stable.
> > >
> > > memcg_reparent_objcgs(memcg)
> > >     // lruvec belongs to memcg and lruvec_parent belongs to parent memcg.
> > >     spin_lock(&lruvec->lru_lock);
> > >     spin_lock(&lruvec_parent->lru_lock);
> > >
> > >     // Move all the pages from the lruvec list to the parent lruvec list.
> > >
> > >     spin_unlock(&lruvec_parent->lru_lock);
> > >     spin_unlock(&lruvec->lru_lock);
> > >
> > > After we acquire the lruvec lock, we need to check whether the folio is
> > > reparented. If so, we need to reacquire the new lruvec lock. On the
> > > routine of the LRU pages reparenting, we will also acquire the lruvec
> > > lock (will be implemented in the later patch). So folio_memcg() cannot
> > > be changed when we hold the lruvec lock.
> > >
> > > Since lruvec_memcg(lruvec) is always equal to folio_memcg(folio) after
> > > we hold the lruvec lock, lruvec_memcg_debug() check is pointless. So
> > > remove it.
> > >
> > > This is a preparation for reparenting the LRU pages.
> > >
> > > Signed-off-by: Muchun Song <songmuchun@xxxxxxxxxxxxx>
> > > Acked-by: Roman Gushchin <guro@xxxxxx>
> >
> > Maybe it's mostly s/page/folio, but the patch looks quite differently
> > in comparison to the version I did ack. In general, please, drop acks
> > when there are significant changes between versions.
> 
> Got it. I'll drop all acks in the next version. Thanks.

Thank you!

The code look correct though. But the locking scheme becomes very complicated
(it was already complex). I wonder if there are better ideas. No ideas out
of my head, need to think more. If not, your approach looks ok.