In a memcg with even just moderate cache pressure, success rates for
transparent huge page allocations drop to zero, wasting a lot of effort
that the allocator puts into assembling these pages.

The reason for this is that the memcg reclaim code was never designed for
higher-order charges.  It reclaims in small batches until there is room
for at least one page.  Huge page charges only succeed when these batches
add up over a series of huge faults, which is unlikely under any
significant load involving order-0 allocations in the group.

Remove that loop on the memcg side in favor of passing the actual reclaim
goal to direct reclaim, which is already set up and optimized to meet
higher-order goals efficiently.

This brings memcg's THP policy in line with the system policy: if the
allocator painstakingly assembles a hugepage, memcg will at least make an
honest effort to charge it.  As a result, transparent hugepage allocation
rates amid cache activity are drastically improved:

                                      vanilla                 patched
pgalloc                 4717530.80 (  +0.00%)   4451376.40 (  -5.64%)
pgfault                  491370.60 (  +0.00%)    225477.40 ( -54.11%)
pgmajfault                    2.00 (  +0.00%)         1.80 (  -6.67%)
thp_fault_alloc               0.00 (  +0.00%)       531.60 (+100.00%)
thp_fault_fallback          749.00 (  +0.00%)       217.40 ( -70.88%)

[ Note: this may in turn increase memory consumption from internal
  fragmentation, which is an inherent risk of transparent hugepages.
  Some setups may have to adjust the memcg limits accordingly to
  accomodate this - or, if the machine is already packed to capacity,
  disable the transparent huge page feature. ]
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Reviewed-by: NVladimir Davydov <vdavydov@parallels.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Dave Hansen <dave@sr71.net>
Cc: Greg Thelen <gthelen@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When attempting to charge pages, we first charge the memory counter and
then the memory+swap counter.  If one of the counters is at its limit, we
enter reclaim, but if it's the memory+swap counter, reclaim shouldn't swap
because that wouldn't change the situation.  However, if the counters have
the same limits, we never get to the memory+swap limit.  To know whether
reclaim should swap or not, there is a state flag that indicates whether
the limits are equal and whether hitting the memory limit implies hitting
the memory+swap limit.

Just try the memory+swap counter first.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Reviewed-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Dave Hansen <dave@sr71.net>
Cc: Greg Thelen <gthelen@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

`While growing per memcg caches arrays, we jump between memcontrol.c and
slab_common.c in a weird way:

  memcg_alloc_cache_id - memcontrol.c
    memcg_update_all_caches - slab_common.c
      memcg_update_cache_size - memcontrol.c

There's absolutely no reason why memcg_update_cache_size can't live on the
slab's side though.  So let's move it there and settle it comfortably amid
per-memcg cache allocation functions.

Besides, this patch cleans this function up a bit, removing all the
useless comments from it, and renames it to memcg_update_cache_params to
conform to memcg_alloc/free_cache_params, which we already have in
slab_common.c.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

memcg_update_all_caches grows arrays of per-memcg caches, so we only need
to call it when memcg_limited_groups_array_size is increased.  However,
currently we invoke it each time a new kmem-active memory cgroup is
created.  Then it just iterates over all slab_caches and does nothing
(memcg_update_cache_size returns immediately).

This patch fixes this insanity.  In the meantime it moves the code dealing
with id allocations to separate functions, memcg_alloc_cache_id and
memcg_free_cache_id.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The only reason why they live in memcontrol.c is that we get/put css
reference to the owner memory cgroup in them.  However, we can do that in
memcg_{un,}register_cache.  OTOH, there are several reasons to move them
to slab_common.c.

First, I think that the less public interface functions we have in
memcontrol.h the better.  Since the functions I move don't depend on
memcontrol, I think it's worth making them private to slab, especially
taking into account that the arrays are defined on the slab's side too.

Second, the way how per-memcg arrays are updated looks rather awkward: it
proceeds from memcontrol.c (__memcg_activate_kmem) to slab_common.c
(memcg_update_all_caches) and back to memcontrol.c again
(memcg_update_array_size).  In the following patches I move the function
relocating the arrays (memcg_update_array_size) to slab_common.c and
therefore get rid this circular call path.  I think we should have the
cache allocation stuff in the same place where we have relocation, because
it's easier to follow the code then.  So I move arrays alloc/free
functions to slab_common.c too.

The third point isn't obvious.  I'm going to make the list_lru structure
per-memcg to allow targeted kmem reclaim.  That means we will have
per-memcg arrays in list_lrus too.  It turns out that it's much easier to
update these arrays in list_lru.c rather than in memcontrol.c, because all
the stuff we need is defined there.  This patch makes memcg caches arrays
allocation path conform that of the upcoming list_lru.

So let's move these functions to slab_common.c and make them static.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Christoph Lameter <cl@linux.com>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The cgroup iterators yield css objects that have not yet gone through
css_online(), but they are not complete memcgs at this point and so the
memcg iterators should not return them.  Commit d8ad3055 ("mm/memcg:
iteration skip memcgs not yet fully initialized") set out to implement
exactly this, but it uses CSS_ONLINE, a cgroup-internal flag that does
not meet the ordering requirements for memcg, and so the iterator may
skip over initialized groups, or return partially initialized memcgs.

The cgroup core can not reasonably provide a clear answer on whether the
object around the css has been fully initialized, as that depends on
controller-specific locking and lifetime rules.  Thus, introduce a
memcg-specific flag that is set after the memcg has been initialized in
css_online(), and read before mem_cgroup_iter() callers access the memcg
members.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Tejun Heo <tj@kernel.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: <stable@vger.kernel.org>	[3.12+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Dave Hansen reports a massive scalability regression in an uncontained
page fault benchmark with more than 30 concurrent threads, which he
bisected down to 05b84301 ("mm: memcontrol: use root_mem_cgroup
res_counter") and pin-pointed on res_counter spinlock contention.

That change relied on the per-cpu charge caches to mostly swallow the
res_counter costs, but it's apparent that the caches don't scale yet.

Revert memcg back to bypassing res_counters on the root level in order
to restore performance for uncontained workloads.
Reported-by: NDave Hansen <dave@sr71.net>
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Tested-by: NDave Hansen <dave.hansen@intel.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Acked-by: NVladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Charge migration currently disables IRQs twice to update the charge
statistics for the old page and then again for the new page.

But migration is a seamless transition of a charge from one physical
page to another one of the same size, so this should be a non-event from
an accounting point of view.  Leave the statistics alone.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Pages are now uncharged at release time, and all sources of batched
uncharges operate on lists of pages.  Directly use those lists, and
get rid of the per-task batching state.

This also batches statistics accounting, in addition to the res
counter charges, to reduce IRQ-disabling and re-enabling.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The memcg uncharging code that is involved towards the end of a page's
lifetime - truncation, reclaim, swapout, migration - is impressively
complicated and fragile.

Because anonymous and file pages were always charged before they had their
page->mapping established, uncharges had to happen when the page type
could still be known from the context; as in unmap for anonymous, page
cache removal for file and shmem pages, and swap cache truncation for swap
pages.  However, these operations happen well before the page is actually
freed, and so a lot of synchronization is necessary:

- Charging, uncharging, page migration, and charge migration all need
  to take a per-page bit spinlock as they could race with uncharging.

- Swap cache truncation happens during both swap-in and swap-out, and
  possibly repeatedly before the page is actually freed.  This means
  that the memcg swapout code is called from many contexts that make
  no sense and it has to figure out the direction from page state to
  make sure memory and memory+swap are always correctly charged.

- On page migration, the old page might be unmapped but then reused,
  so memcg code has to prevent untimely uncharging in that case.
  Because this code - which should be a simple charge transfer - is so
  special-cased, it is not reusable for replace_page_cache().

But now that charged pages always have a page->mapping, introduce
mem_cgroup_uncharge(), which is called after the final put_page(), when we
know for sure that nobody is looking at the page anymore.

For page migration, introduce mem_cgroup_migrate(), which is called after
the migration is successful and the new page is fully rmapped.  Because
the old page is no longer uncharged after migration, prevent double
charges by decoupling the page's memcg association (PCG_USED and
pc->mem_cgroup) from the page holding an actual charge.  The new bits
PCG_MEM and PCG_MEMSW represent the respective charges and are transferred
to the new page during migration.

mem_cgroup_migrate() is suitable for replace_page_cache() as well,
which gets rid of mem_cgroup_replace_page_cache().  However, care
needs to be taken because both the source and the target page can
already be charged and on the LRU when fuse is splicing: grab the page
lock on the charge moving side to prevent changing pc->mem_cgroup of a
page under migration.  Also, the lruvecs of both pages change as we
uncharge the old and charge the new during migration, and putback may
race with us, so grab the lru lock and isolate the pages iff on LRU to
prevent races and ensure the pages are on the right lruvec afterward.

Swap accounting is massively simplified: because the page is no longer
uncharged as early as swap cache deletion, a new mem_cgroup_swapout() can
transfer the page's memory+swap charge (PCG_MEMSW) to the swap entry
before the final put_page() in page reclaim.

Finally, page_cgroup changes are now protected by whatever protection the
page itself offers: anonymous pages are charged under the page table lock,
whereas page cache insertions, swapin, and migration hold the page lock.
Uncharging happens under full exclusion with no outstanding references.
Charging and uncharging also ensure that the page is off-LRU, which
serializes against charge migration.  Remove the very costly page_cgroup
lock and set pc->flags non-atomically.

[mhocko@suse.cz: mem_cgroup_charge_statistics needs preempt_disable]
[vdavydov@parallels.com: fix flags definition]
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Tested-by: NJet Chen <jet.chen@intel.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Tested-by: NFelipe Balbi <balbi@ti.com>
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

These patches rework memcg charge lifetime to integrate more naturally
with the lifetime of user pages.  This drastically simplifies the code and
reduces charging and uncharging overhead.  The most expensive part of
charging and uncharging is the page_cgroup bit spinlock, which is removed
entirely after this series.

Here are the top-10 profile entries of a stress test that reads a 128G
sparse file on a freshly booted box, without even a dedicated cgroup (i.e.
 executing in the root memcg).  Before:

    15.36%              cat  [kernel.kallsyms]   [k] copy_user_generic_string
    13.31%              cat  [kernel.kallsyms]   [k] memset
    11.48%              cat  [kernel.kallsyms]   [k] do_mpage_readpage
     4.23%              cat  [kernel.kallsyms]   [k] get_page_from_freelist
     2.38%              cat  [kernel.kallsyms]   [k] put_page
     2.32%              cat  [kernel.kallsyms]   [k] __mem_cgroup_commit_charge
     2.18%          kswapd0  [kernel.kallsyms]   [k] __mem_cgroup_uncharge_common
     1.92%          kswapd0  [kernel.kallsyms]   [k] shrink_page_list
     1.86%              cat  [kernel.kallsyms]   [k] __radix_tree_lookup
     1.62%              cat  [kernel.kallsyms]   [k] __pagevec_lru_add_fn

After:

    15.67%           cat  [kernel.kallsyms]   [k] copy_user_generic_string
    13.48%           cat  [kernel.kallsyms]   [k] memset
    11.42%           cat  [kernel.kallsyms]   [k] do_mpage_readpage
     3.98%           cat  [kernel.kallsyms]   [k] get_page_from_freelist
     2.46%           cat  [kernel.kallsyms]   [k] put_page
     2.13%       kswapd0  [kernel.kallsyms]   [k] shrink_page_list
     1.88%           cat  [kernel.kallsyms]   [k] __radix_tree_lookup
     1.67%           cat  [kernel.kallsyms]   [k] __pagevec_lru_add_fn
     1.39%       kswapd0  [kernel.kallsyms]   [k] free_pcppages_bulk
     1.30%           cat  [kernel.kallsyms]   [k] kfree

As you can see, the memcg footprint has shrunk quite a bit.

   text    data     bss     dec     hex filename
  37970    9892     400   48262    bc86 mm/memcontrol.o.old
  35239    9892     400   45531    b1db mm/memcontrol.o

This patch (of 4):

The memcg charge API charges pages before they are rmapped - i.e.  have an
actual "type" - and so every callsite needs its own set of charge and
uncharge functions to know what type is being operated on.  Worse,
uncharge has to happen from a context that is still type-specific, rather
than at the end of the page's lifetime with exclusive access, and so
requires a lot of synchronization.

Rewrite the charge API to provide a generic set of try_charge(),
commit_charge() and cancel_charge() transaction operations, much like
what's currently done for swap-in:

  mem_cgroup_try_charge() attempts to reserve a charge, reclaiming
  pages from the memcg if necessary.

  mem_cgroup_commit_charge() commits the page to the charge once it
  has a valid page->mapping and PageAnon() reliably tells the type.

  mem_cgroup_cancel_charge() aborts the transaction.

This reduces the charge API and enables subsequent patches to
drastically simplify uncharging.

As pages need to be committed after rmap is established but before they
are added to the LRU, page_add_new_anon_rmap() must stop doing LRU
additions again.  Revive lru_cache_add_active_or_unevictable().

[hughd@google.com: fix shmem_unuse]
[hughd@google.com: Add comments on the private use of -EAGAIN]
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NHugh Dickins <hughd@google.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Charge reclaim and OOM currently use the charge batch variable, but
batching is already disabled at that point.  To simplify the charge
logic, the batch variable is reset to the original request size when
reclaim is entered, so it's functionally equal, but it's misleading.

Switch reclaim/OOM to nr_pages, which is the original request size.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Kmem page charging and uncharging is serialized by means of exclusive
access to the page.  Do not take the page_cgroup lock and don't set
pc->flags atomically.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Acked-by: NVladimir Davydov <vdavydov@parallels.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There is a write barrier between setting pc->mem_cgroup and
PageCgroupUsed, which was added to allow LRU operations to lookup the
memcg LRU list of a page without acquiring the page_cgroup lock.

But ever since commit 38c5d72f ("memcg: simplify LRU handling by new
rule"), pages are ensured to be off-LRU while charging, so nobody else
is changing LRU state while pc->mem_cgroup is being written, and there
are no read barriers anymore.

Remove the unnecessary write barrier.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Due to an old optimization to keep expensive res_counter changes at a
minimum, the root_mem_cgroup res_counter is never charged; there is no
limit at that level anyway, and any statistics can be generated on
demand by summing up the counters of all other cgroups.

However, with per-cpu charge caches, res_counter operations do not even
show up in profiles anymore, so this optimization is no longer
necessary.

Remove it to simplify the code.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When mem_cgroup_try_charge() returns -EINTR, it bypassed the charge to
the root memcg.  But move precharging does not catch this and treats
this case as if no charge had happened, thus leaking a charge against
root.  Because of an old optimization, the root memcg's res_counter is
not actually charged right now, but it's still an imbalance and
subsequent patches will charge the root memcg again.

Catch those bypasses to the root memcg and properly cancel them before
giving up the move.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The move precharge function does some baroque things: it tries raw
res_counter charging of the entire amount first, and then falls back to
a loop of one-by-one charges, with checks for pending signals and
cond_resched() batching.

Just use mem_cgroup_try_charge() without __GFP_WAIT for the first bulk
charge attempt.  In the one-by-one loop, remove the signal check (this
is already checked in try_charge), and simply call cond_resched() after
every charge - it's not that expensive.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

For the page allocator, __GFP_NORETRY implies that no OOM should be
triggered, whereas memcg has an explicit parameter to disable OOM.

The only callsites that want OOM disabled are THP charges and charge
moving.  THP already uses __GFP_NORETRY and charge moving can use it as
well - one full reclaim cycle should be plenty.  Switch it over, then
remove the OOM parameter.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There is no reason why oom-disabled and __GFP_NOFAIL charges should try
to reclaim only once when every other charge tries several times before
giving up.  Make them all retry the same number of times.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Currently, __GFP_NORETRY tries charging once and gives up before even
trying to reclaim.  Bring the behavior on par with the page allocator
and reclaim at least once before giving up.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The charging path currently starts out with OOM condition checks when
OOM is the rarest possible case.

Rearrange this code to run OOM/task dying checks only after trying the
percpu charge and the res_counter charge and bail out before entering
reclaim.  Attempting a charge does not hurt an (oom-)killed task as much
as every charge attempt having to check OOM conditions.  Also, only
check __GFP_NOFAIL when the charge would actually fail.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

These patches rework memcg charge lifetime to integrate more naturally
with the lifetime of user pages.  This drastically simplifies the code
and reduces charging and uncharging overhead.  The most expensive part
of charging and uncharging is the page_cgroup bit spinlock, which is
removed entirely after this series.

Here are the top-10 profile entries of a stress test that reads a 128G
sparse file on a freshly booted box, without even a dedicated cgroup
(i.e. executing in the root memcg).  Before:

    15.36%              cat  [kernel.kallsyms]   [k] copy_user_generic_string
    13.31%              cat  [kernel.kallsyms]   [k] memset
    11.48%              cat  [kernel.kallsyms]   [k] do_mpage_readpage
     4.23%              cat  [kernel.kallsyms]   [k] get_page_from_freelist
     2.38%              cat  [kernel.kallsyms]   [k] put_page
     2.32%              cat  [kernel.kallsyms]   [k] __mem_cgroup_commit_charge
     2.18%          kswapd0  [kernel.kallsyms]   [k] __mem_cgroup_uncharge_common
     1.92%          kswapd0  [kernel.kallsyms]   [k] shrink_page_list
     1.86%              cat  [kernel.kallsyms]   [k] __radix_tree_lookup
     1.62%              cat  [kernel.kallsyms]   [k] __pagevec_lru_add_fn

After:

    15.67%           cat  [kernel.kallsyms]   [k] copy_user_generic_string
    13.48%           cat  [kernel.kallsyms]   [k] memset
    11.42%           cat  [kernel.kallsyms]   [k] do_mpage_readpage
     3.98%           cat  [kernel.kallsyms]   [k] get_page_from_freelist
     2.46%           cat  [kernel.kallsyms]   [k] put_page
     2.13%       kswapd0  [kernel.kallsyms]   [k] shrink_page_list
     1.88%           cat  [kernel.kallsyms]   [k] __radix_tree_lookup
     1.67%           cat  [kernel.kallsyms]   [k] __pagevec_lru_add_fn
     1.39%       kswapd0  [kernel.kallsyms]   [k] free_pcppages_bulk
     1.30%           cat  [kernel.kallsyms]   [k] kfree

As you can see, the memcg footprint has shrunk quite a bit.

   text    data     bss     dec     hex filename
  37970    9892     400   48262    bc86 mm/memcontrol.o.old
  35239    9892     400   45531    b1db mm/memcontrol.o

This patch (of 13):

This function was split out because mem_cgroup_try_charge() got too big.
But having essentially one sequence of operations arbitrarily split in
half is not good for reworking the code.  Fold it back in.
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Paul Furtado has reported the following GPF:

  general protection fault: 0000 [#1] SMP
  Modules linked in: ipv6 dm_mod xen_netfront coretemp hwmon x86_pkg_temp_thermal crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel ablk_helper cryptd lrw gf128mul glue_helper aes_x86_64 microcode pcspkr ext4 jbd2 mbcache raid0 xen_blkfront
  CPU: 3 PID: 3062 Comm: java Not tainted 3.16.0-rc5 #1
  task: ffff8801cfe8f170 ti: ffff8801d2ec4000 task.ti: ffff8801d2ec4000
  RIP: e030:mem_cgroup_oom_synchronize+0x140/0x240
  RSP: e02b:ffff8801d2ec7d48  EFLAGS: 00010283
  RAX: 0000000000000001 RBX: ffff88009d633800 RCX: 000000000000000e
  RDX: fffffffffffffffe RSI: ffff88009d630200 RDI: ffff88009d630200
  RBP: ffff8801d2ec7da8 R08: 0000000000000012 R09: 00000000fffffffe
  R10: 0000000000000000 R11: 0000000000000000 R12: ffff88009d633800
  R13: ffff8801d2ec7d48 R14: dead000000100100 R15: ffff88009d633a30
  FS:  00007f1748bb4700(0000) GS:ffff8801def80000(0000) knlGS:0000000000000000
  CS:  e033 DS: 0000 ES: 0000 CR0: 000000008005003b
  CR2: 00007f4110300308 CR3: 00000000c05f7000 CR4: 0000000000002660
  Call Trace:
    pagefault_out_of_memory+0x18/0x90
    mm_fault_error+0xa9/0x1a0
    __do_page_fault+0x478/0x4c0
    do_page_fault+0x2c/0x40
    page_fault+0x28/0x30
  Code: 44 00 00 48 89 df e8 40 ca ff ff 48 85 c0 49 89 c4 74 35 4c 8b b0 30 02 00 00 4c 8d b8 30 02 00 00 4d 39 fe 74 1b 0f 1f 44 00 00 <49> 8b 7e 10 be 01 00 00 00 e8 42 d2 04 00 4d 8b 36 4d 39 fe 75
  RIP  mem_cgroup_oom_synchronize+0x140/0x240

Commit fb2a6fc5 ("mm: memcg: rework and document OOM waiting and
wakeup") has moved mem_cgroup_oom_notify outside of memcg_oom_lock
assuming it is protected by the hierarchical OOM-lock.

Although this is true for the notification part the protection doesn't
cover unregistration of event which can happen in parallel now so
mem_cgroup_oom_notify can see already unlinked and/or freed
mem_cgroup_eventfd_list.

Fix this by using memcg_oom_lock also in mem_cgroup_oom_notify.

Addresses https://bugzilla.kernel.org/show_bug.cgi?id=80881

Fixes: fb2a6fc5 (mm: memcg: rework and document OOM waiting and wakeup)
Signed-off-by: NMichal Hocko <mhocko@suse.cz>
Reported-by: NPaul Furtado <paulfurtado91@gmail.com>
Tested-by: NPaul Furtado <paulfurtado91@gmail.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: <stable@vger.kernel.org>	[3.12+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Until now, cftype arrays carried files for both the default and legacy
hierarchies and the files which needed to be used on only one of them
were flagged with either CFTYPE_ONLY_ON_DFL or CFTYPE_INSANE.  This
gets confusing very quickly and we may end up exposing interface files
to the default hierarchy without thinking it through.

This patch makes cgroup core provide separate sets of interfaces for
cftype handling so that the cftypes for the default and legacy
hierarchies are clearly distinguished.  The previous two patches
renamed the existing ones so that they clearly indicate that they're
for the legacy hierarchies.  This patch adds the interface for the
default hierarchy and apply them selectively depending on the
hierarchy type.

* cftypes added through cgroup_subsys->dfl_cftypes and
  cgroup_add_dfl_cftypes() only show up on the default hierarchy.

* cftypes added through cgroup_subsys->legacy_cftypes and
  cgroup_add_legacy_cftypes() only show up on the legacy hierarchies.

* cgroup_subsys->dfl_cftypes and ->legacy_cftypes can point to the
  same array for the cases where the interface files are identical on
  both types of hierarchies.

* This makes all the existing subsystem interface files legacy-only by
  default and all subsystems will have no interface file created when
  enabled on the default hierarchy.  Each subsystem should explicitly
  review and compose the interface for the default hierarchy.

* A boot param "cgroup__DEVEL__legacy_files_on_dfl" is added which
  makes subsystems which haven't decided the interface files for the
  default hierarchy to present the legacy files on the default
  hierarchy so that its behavior on the default hierarchy can be
  tested.  As the awkward name suggests, this is for development only.

* memcg's CFTYPE_INSANE on "use_hierarchy" is noop now as the whole
  array isn't used on the default hierarchy.  The flag is removed.

v2: Updated documentation for cgroup__DEVEL__legacy_files_on_dfl.

v3: Clear CFTYPE_ONLY_ON_DFL and CFTYPE_INSANE when cfts are removed
    as suggested by Li.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NNeil Horman <nhorman@tuxdriver.com>
Acked-by: NLi Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Aristeu Rozanski <aris@redhat.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>

Currently, cftypes added by cgroup_add_cftypes() are used for both the
unified default hierarchy and legacy ones and subsystems can mark each
file with either CFTYPE_ONLY_ON_DFL or CFTYPE_INSANE if it has to
appear only on one of them.  This is quite hairy and error-prone.
Also, we may end up exposing interface files to the default hierarchy
without thinking it through.

cgroup_subsys will grow two separate cftype addition functions and
apply each only on the hierarchies of the matching type.  This will
allow organizing cftypes in a lot clearer way and encourage subsystems
to scrutinize the interface which is being exposed in the new default
hierarchy.

In preparation, this patch adds cgroup_add_legacy_cftypes() which
currently is a simple wrapper around cgroup_add_cftypes() and replaces
all cgroup_add_cftypes() usages with it.

While at it, this patch drops a completely spurious return from
__hugetlb_cgroup_file_init().

This patch doesn't introduce any functional differences.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NNeil Horman <nhorman@tuxdriver.com>
Acked-by: NLi Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>

Currently, cgroup_subsys->base_cftypes is used for both the unified
default hierarchy and legacy ones and subsystems can mark each file
with either CFTYPE_ONLY_ON_DFL or CFTYPE_INSANE if it has to appear
only on one of them.  This is quite hairy and error-prone.  Also, we
may end up exposing interface files to the default hierarchy without
thinking it through.

cgroup_subsys will grow two separate cftype arrays and apply each only
on the hierarchies of the matching type.  This will allow organizing
cftypes in a lot clearer way and encourage subsystems to scrutinize
the interface which is being exposed in the new default hierarchy.

In preparation, this patch renames cgroup_subsys->base_cftypes to
cgroup_subsys->legacy_cftypes.  This patch is pure rename.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NNeil Horman <nhorman@tuxdriver.com>
Acked-by: NLi Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Aristeu Rozanski <aris@redhat.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>

sane_behavior has been used as a development vehicle for the default
unified hierarchy.  Now that the default hierarchy is in place, the
flag became redundant and confusing as its usage is allowed on all
hierarchies.  There are gonna be either the default hierarchy or
legacy ones.  Let's make that clear by removing sane_behavior support
on non-default hierarchies.

This patch replaces cgroup_sane_behavior() with cgroup_on_dfl().  The
comment on top of CGRP_ROOT_SANE_BEHAVIOR is moved to on top of
cgroup_on_dfl() with sane_behavior specific part dropped.

On the default and legacy hierarchies w/o sane_behavior, this
shouldn't cause any behavior differences.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NVivek Goyal <vgoyal@redhat.com>
Acked-by: NLi Zefan <lizefan@huawei.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>

Currently, the blkio subsystem attributes all of writeback IOs to the
root.  One of the issues is that there's no way to tell who originated
a writeback IO from block layer.  Those IOs are usually issued
asynchronously from a task which didn't have anything to do with
actually generating the dirty pages.  The memory subsystem, when
enabled, already keeps track of the ownership of each dirty page and
it's desirable for blkio to piggyback instead of adding its own
per-page tag.

cgroup now has a mechanism to express such dependency -
cgroup_subsys->depends_on.  This patch declares that blkcg depends on
memcg so that memcg is enabled automatically on the default hierarchy
when available.  Future changes will make blkcg map the memcg tag to
find out the cgroup to blame for writeback IOs.

As this means that a memcg may be made invisible, this patch also
implements css_reset() for memcg which resets its basic
configurations.  This implementation will probably need to be expanded
to cover other states which are used in the default hierarchy.

v2: blkcg's dependency on memcg is wrapped with CONFIG_MEMCG to avoid
    build failure.  Reported by kbuild test robot.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NLi Zefan <lizefan@huawei.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Jens Axboe <axboe@kernel.dk>

Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Jianyu Zhan <nasa4836@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Memcg zoneinfo lookup sites have either the page, the zone, or the node
id and zone index, but sites that only have the zone have to look up the
node id and zone index themselves, whereas sites that already have those
two integers use a function for a simple pointer chase.

Provide mem_cgroup_zone_zoneinfo() that takes a zone pointer and let
sites that already have node id and zone index - all for each node, for
each zone iterators - use &memcg->nodeinfo[nid]->zoneinfo[zid].

Rename page_cgroup_zoneinfo() to mem_cgroup_page_zoneinfo() to match.
Signed-off-by: NJianyu Zhan <nasa4836@gmail.com>
Signed-off-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Memory reclaim always uses swappiness of the reclaim target memcg
(origin of the memory pressure) or vm_swappiness for global memory
reclaim.  This behavior was consistent (except for difference between
global and hard limit reclaim) because swappiness was enforced to be
consistent within each memcg hierarchy.

After "mm: memcontrol: remove hierarchy restrictions for swappiness and
oom_control" each memcg can have its own swappiness independent of
hierarchical parents, though, so the consistency guarantee is gone.
This can lead to an unexpected behavior.  Say that a group is explicitly
configured to not swapout by memory.swappiness=0 but its memory gets
swapped out anyway when the memory pressure comes from its parent with a
It is also unexpected that the knob is meaningless without setting the
hard limit which would trigger the reclaim and enforce the swappiness.
There are setups where the hard limit is configured higher in the
hierarchy by an administrator and children groups are under control of
somebody else who is interested in the swapout behavior but not
necessarily about the memory limit.

From a semantic point of view swappiness is an attribute defining anon
vs.
 file proportional scanning of LRU which is memcg specific (unlike
charges which are propagated up the hierarchy) so it should be applied
to the particular memcg's LRU regardless where the memory pressure comes
from.

This patch removes vmscan_swappiness() and stores the swappiness into
the scan_control structure.  mem_cgroup_swappiness is then used to
provide the correct value before shrink_lruvec is called.  The global
vm_swappiness is used for the root memcg.

[hughd@google.com: oopses immediately when booted with cgroup_disable=memory]
Signed-off-by: NMichal Hocko <mhocko@suse.cz>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: NHugh Dickins <hughd@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

mem_cgroup_force_empty_list() can iterate a large number of pages on an
lru and mem_cgroup_move_parent() doesn't return an errno unless certain
criteria, none of which indicate that the iteration may be taking too
long, is met.

We have encountered the following stack trace many times indicating
"need_resched set for > 51000020 ns (51 ticks) without schedule", for
example:

	scheduler_tick()
	<timer irq>
	mem_cgroup_move_account+0x4d/0x1d5
	mem_cgroup_move_parent+0x8d/0x109
	mem_cgroup_reparent_charges+0x149/0x2ba
	mem_cgroup_css_offline+0xeb/0x11b
	cgroup_offline_fn+0x68/0x16b
	process_one_work+0x129/0x350

If this iteration is taking too long, we still need to do cond_resched()
even when an individual page is not busy.

[rientjes@google.com: changelog]
Signed-off-by: NHugh Dickins <hughd@google.com>
Signed-off-by: NDavid Rientjes <rientjes@google.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Current names are rather inconsistent. Let's try to improve them.

Brief change log:

** old name **                          ** new name **

kmem_cache_create_memcg                 memcg_create_kmem_cache
memcg_kmem_create_cache                 memcg_regsiter_cache
memcg_kmem_destroy_cache                memcg_unregister_cache

kmem_cache_destroy_memcg_children       memcg_cleanup_cache_params
mem_cgroup_destroy_all_caches           memcg_unregister_all_caches

create_work                             memcg_register_cache_work
memcg_create_cache_work_func            memcg_register_cache_func
memcg_create_cache_enqueue              memcg_schedule_register_cache
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

It isn't worth complicating the code by allocating it on the first access,
because it only takes 256 bytes.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Instead of calling back to memcontrol.c from kmem_cache_create_memcg in
order to just create the name of a per memcg cache, let's allocate it in
place.  We only need to pass the memcg name to kmem_cache_create_memcg for
that - everything else can be done in slab_common.c.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: NQiang Huang <h.huangqiang@huawei.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

It is only used in __mem_cgroup_begin_update_page_stat(), the name is
confusing and 2 routines for one thing also confuse people, so fold this
function seems more clear.

[akpm@linux-foundation.org: fix typo, per Michal]
Signed-off-by: NQiang Huang <h.huangqiang@huawei.com>
Acked-by: NMichal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

static values are automatically initialized to NULL
Signed-off-by: NFabian Frederick <fabf@skynet.be>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Replace places where __get_cpu_var() is used for an address calculation
with this_cpu_ptr().
Signed-off-by: NChristoph Lameter <cl@linux.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

At present, we have the following mutexes protecting data related to per
memcg kmem caches:

 - slab_mutex.  This one is held during the whole kmem cache creation
   and destruction paths.  We also take it when updating per root cache
   memcg_caches arrays (see memcg_update_all_caches).  As a result, taking
   it guarantees there will be no changes to any kmem cache (including per
   memcg).  Why do we need something else then?  The point is it is
   private to slab implementation and has some internal dependencies with
   other mutexes (get_online_cpus).  So we just don't want to rely upon it
   and prefer to introduce additional mutexes instead.

 - activate_kmem_mutex.  Initially it was added to synchronize
   initializing kmem limit (memcg_activate_kmem).  However, since we can
   grow per root cache memcg_caches arrays only on kmem limit
   initialization (see memcg_update_all_caches), we also employ it to
   protect against memcg_caches arrays relocation (e.g.  see
   __kmem_cache_destroy_memcg_children).

 - We have a convention not to take slab_mutex in memcontrol.c, but we
   want to walk over per memcg memcg_slab_caches lists there (e.g.  for
   destroying all memcg caches on offline).  So we have per memcg
   slab_caches_mutex's protecting those lists.

The mutexes are taken in the following order:

   activate_kmem_mutex -> slab_mutex -> memcg::slab_caches_mutex

Such a syncrhonization scheme has a number of flaws, for instance:

 - We can't call kmem_cache_{destroy,shrink} while walking over a
   memcg::memcg_slab_caches list due to locking order.  As a result, in
   mem_cgroup_destroy_all_caches we schedule the
   memcg_cache_params::destroy work shrinking and destroying the cache.

 - We don't have a mutex to synchronize per memcg caches destruction
   between memcg offline (mem_cgroup_destroy_all_caches) and root cache
   destruction (__kmem_cache_destroy_memcg_children).  Currently we just
   don't bother about it.

This patch simplifies it by substituting per memcg slab_caches_mutex's
with the global memcg_slab_mutex.  It will be held whenever a new per
memcg cache is created or destroyed, so it protects per root cache
memcg_caches arrays and per memcg memcg_slab_caches lists.  The locking
order is following:

   activate_kmem_mutex -> memcg_slab_mutex -> slab_mutex

This allows us to call kmem_cache_{create,shrink,destroy} under the
memcg_slab_mutex.  As a result, we don't need memcg_cache_params::destroy
work any more - we can simply destroy caches while iterating over a per
memcg slab caches list.

Also using the global mutex simplifies synchronization between concurrent
per memcg caches creation/destruction, e.g.  mem_cgroup_destroy_all_caches
vs __kmem_cache_destroy_memcg_children.

The downside of this is that we substitute per-memcg slab_caches_mutex's
with a hummer-like global mutex, but since we already take either the
slab_mutex or the cgroup_mutex along with a memcg::slab_caches_mutex, it
shouldn't hurt concurrency a lot.
Signed-off-by: NVladimir Davydov <vdavydov@parallels.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Pekka Enberg <penberg@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>