Date:   Mon, 30 Nov 2020 14:33:47 -0800
From:   Roman Gushchin <guro@...com>
To:     Yang Shi <shy828301@...il.com>
CC:     Vladimir Davydov <vdavydov.dev@...il.com>,
        Kirill Tkhai <ktkhai@...tuozzo.com>,
        Shakeel Butt <shakeelb@...gle.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linux MM <linux-mm@...ck.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] mm: list_lru: hold nlru lock to avoid reading transient
 negative nr_items

On Mon, Nov 30, 2020 at 12:57:47PM -0800, Yang Shi wrote:
> On Mon, Nov 30, 2020 at 12:09 PM Roman Gushchin <guro@...com> wrote:
> >
> > On Mon, Nov 30, 2020 at 10:45:14AM -0800, Yang Shi wrote:
> > > When investigating a slab cache bloat problem, significant amount of
> > > negative dentry cache was seen, but confusingly they neither got shrunk
> > > by reclaimer (the host has very tight memory) nor be shrunk by dropping
> > > cache.  The vmcore shows there are over 14M negative dentry objects on lru,
> > > but tracing result shows they were even not scanned at all.  The further
> > > investigation shows the memcg's vfs shrinker_map bit is not set.  So the
> > > reclaimer or dropping cache just skip calling vfs shrinker.  So we have
> > > to reboot the hosts to get the memory back.
> > >
> > > I didn't manage to come up with a reproducer in test environment, and the
> > > problem can't be reproduced after rebooting.  But it seems there is race
> > > between shrinker map bit clear and reparenting by code inspection.  The
> > > hypothesis is elaborated as below.
> > >
> > > The memcg hierarchy on our production environment looks like:
> > >                 root
> > >                /    \
> > >           system   user
> > >
> > > The main workloads are running under user slice's children, and it creates
> > > and removes memcg frequently.  So reparenting happens very often under user
> > > slice, but no task is under user slice directly.
> > >
> > > So with the frequent reparenting and tight memory pressure, the below
> > > hypothetical race condition may happen:
> > >
> > >     CPU A                            CPU B                         CPU C
> > > reparent
> > >     dst->nr_items == 0
> > >                                  shrinker:
> > >                                      total_objects == 0
> > >     add src->nr_items to dst
> > >     set_bit
> > >                                      retrun SHRINK_EMPTY
> > >                                      clear_bit
> > >                                                                   list_lru_del()
> > > reparent again
> > >     dst->nr_items may go negative
> > >     due to current list_lru_del()
> > >     on CPU C
> > >                                  The second run of shrinker:
> > >                                      read nr_items without any
> > >                                      synchronization, so it may
> > >                                      see intermediate negative
> > >                                      nr_items then total_objects
> > >                                      may return 0 conincidently
> > >
> > >                                      keep the bit cleared
> > >     dst->nr_items != 0
> > >     skip set_bit
> > >     add scr->nr_item to dst
> > >
> > > After this point dst->nr_item may never go zero, so reparenting will not
> > > set shrinker_map bit anymore.  And since there is no task under user
> > > slice directly, so no new object will be added to its lru to set the
> > > shrinker map bit either.  That bit is kept cleared forever.
> > >
> > > How does list_lru_del() race with reparenting?  It is because
> > > reparenting replaces childen's kmemcg_id to parent's without protecting
> > > from nlru->lock, so list_lru_del() may see parent's kmemcg_id but
> > > actually deleting items from child's lru, but dec'ing parent's nr_items,
> > > so the parent's nr_items may go negative as commit
> > > 2788cf0c401c268b4819c5407493a8769b7007aa ("memcg: reparent list_lrus and
> > > free kmemcg_id on css offline") says.

Also note that since the introduction of the slab reparenting, list_lru_from_kmem()
can return the parent lru.

> > >
> > > Can we move kmemcg_id replacement after reparenting?  No, because the
> > > race with list_lru_del() may result in negative src->nr_items, but it
> > > will never be fixed.  So the shrinker may never return SHRINK_EMPTY then
> > > keep the shrinker map bit set always.  The shrinker will be always
> > > called for nonsense.
> > >
> > > Can we synchronize list_lru_del() and reparenting?  Yes, it could be
> > > done.  But it seems we need introduce a new lock or use nlru->lock.  But
> > > it sounds complicated to move kmemcg_id replacement code under nlru->lock.
> > > And list_lru_del() may be called quite often to exacerbate some hot
> > > path, i.e. dentry kill.
> > >
> > > So, it sounds acceptable to synchronize reading nr_items to avoid seeing
> > > intermediate negative nr_items given the simplicity and it is typically
> > > just called by shrinkers when counting the freeable objects.
> > >
> > > The patch is tested with some shrinker intensive workloads, no
> > > noticeable regression is soptted.
> >
> > Hi Yang!
> >
> > It's really tricky, thank you for digging in! It's a perfect analysis!
> >
> > I wonder though, if it's better to just always set the shrinker bit on reparenting
> > if we do reparent some items? Then we'll avoid adding new synchronization
> > to the hot path. What do you think?
> 
> Thanks a lot for the suggestion. I was thinking about the same
> approach too, but I thought src->nr_items may go zero due to
> concurrent list_lru_del() at the first place. But I just rethought the
> whole thing, it seems impossible that dst->nr_items goes negative and
> src->nr_items goes zero at the same time.

Even if it would be possible, it seems less scary: the next reparenting
will likely set the bit. So we'll not get into the permanently bad state.

> list_lru_del() should just
> see either dst or src, it can't manipulate both lists simultaneously.
> So I think your suggestion should work. I will incarnate your
> suggestion in v2.
> 
> >
> > --
> >
> > @@ -534,7 +534,6 @@ static void memcg_drain_list_lru_node(struct list_lru *lru, int nid,
> >         struct list_lru_node *nlru = &lru->node[nid];
> >         int dst_idx = dst_memcg->kmemcg_id;
> >         struct list_lru_one *src, *dst;
> > -       bool set;
> >
> >         /*
> >          * Since list_lru_{add,del} may be called under an IRQ-safe lock,
> > @@ -546,9 +545,8 @@ static void memcg_drain_list_lru_node(struct list_lru *lru, int nid,
> >         dst = list_lru_from_memcg_idx(nlru, dst_idx);
> >
> >         list_splice_init(&src->list, &dst->list);
> > -       set = (!dst->nr_items && src->nr_items);
> >         dst->nr_items += src->nr_items;
> > -       if (set)
> > +       if (src->nr_items)
> >                 memcg_set_shrinker_bit(dst_memcg, nid, lru_shrinker_id(lru));
> >         src->nr_items = 0;
> >
> >
> > --
> >
> > Btw, it seems that the bug is quite old. I wonder why we haven't seen it before?
> > Any ideas?
> 
> It is not new, but not that old from my point of view. The
> shrinker_map thing was introduced since v4.19, I bet pre-v4.19 kernel
> may still dominate in production environment. And, it needs some
> conditions (i.e. nr_inode + nr_dentry == 0 coincidently, and there is
> not task under dst memcg directly, etc) to trigger, so it seems
> unlikely to hit.
> 
> And the consequence may be not noticeable to the most people at all.
> We happened to see frequent OOMs on a couple of small machines (32G
> memory w/o swap, but most memory was consumed by anonymous pages)
> recently and they were already up for long time (almost 300 days),
> then the investigation leads to this race condition.

I agree that most users will unlikely notice it.

But https://www.spinics.net/lists/cgroups/msg27295.html looks very similar
and can be caused by the same problem. Once you'll have v2, let's ask
them to test it too.

Thanks!

Powered by blists - more mailing lists