[1/2] fs: fsnotify: account fsnotify metadata to kmemcg
diff mbox

Message ID 20180627191250.209150-2-shakeelb@google.com
State New
Headers show

Commit Message

Shakeel Butt June 27, 2018, 7:12 p.m. UTC
A lot of memory can be consumed by the events generated for the huge or
unlimited queues if there is either no or slow listener.  This can cause
system level memory pressure or OOMs.  So, it's better to account the
fsnotify kmem caches to the memcg of the listener.

However the listener can be in a different memcg than the memcg of the
producer and these allocations happen in the context of the event
producer. This patch introduces remote memcg charging API which the
producer can use to charge the allocations to the memcg of the listener.

There are seven fsnotify kmem caches and among them allocations from
dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
inotify_inode_mark_cachep happens in the context of syscall from the
listener.  So, SLAB_ACCOUNT is enough for these caches.

The objects from fsnotify_mark_connector_cachep are not accounted as they
are small compared to the notification mark or events and it is unclear
whom to account connector to since it is shared by all events attached to
the inode.

The allocations from the event caches happen in the context of the event
producer.  For such caches we will need to remote charge the allocations
to the listener's memcg.  Thus we save the memcg reference in the
fsnotify_group structure of the listener.

This patch has also moved the members of fsnotify_group to keep the size
same, at least for 64 bit build, even with additional member by filling
the holes.

Signed-off-by: Shakeel Butt <shakeelb@google.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Amir Goldstein <amir73il@gmail.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
---
Changelog since v7:
- Simplify the remote memcg charging API
- get_mem_cgroup_from_current() returns root_mem_cgroup if
  current->active_memcg is not NULL but css_tryget_online() fails.

Changelog since v6:
- Removed Jan's ACK as the code has changed a lot
- Squashed the separate remote charging API path into this one
- Removed kmalloc* & kmem_cache_alloc* APIs and only kept the scope API
- Changed fsnotify remote charging code to use scope API

Changelog since v5:
- None

Changelog since v4:
- Fixed the build for CONFIG_MEMCG=n

Changelog since v3:
- Rebased over Jan's patches.
- Some cleanup based on Amir's comments.

Changelog since v2:
- None

Changelog since v1:
- no more charging fsnotify_mark_connector objects
- Fixed the build for SLOB

 fs/notify/dnotify/dnotify.c          |  5 ++--
 fs/notify/fanotify/fanotify.c        | 14 ++++++++---
 fs/notify/fanotify/fanotify_user.c   |  5 +++-
 fs/notify/group.c                    |  3 +++
 fs/notify/inotify/inotify_fsnotify.c |  7 +++++-
 fs/notify/inotify/inotify_user.c     |  5 +++-
 include/linux/fsnotify_backend.h     | 12 ++++++---
 include/linux/memcontrol.h           | 10 +++++++-
 include/linux/sched.h                |  3 +++
 include/linux/sched/mm.h             | 37 ++++++++++++++++++++++++++++
 kernel/fork.c                        |  3 +++
 mm/memcontrol.c                      | 37 +++++++++++++++++++++++++---
 12 files changed, 123 insertions(+), 18 deletions(-)

Comments

Jan Kara June 28, 2018, 10:02 a.m. UTC | #1
On Wed 27-06-18 12:12:49, Shakeel Butt wrote:
> A lot of memory can be consumed by the events generated for the huge or
> unlimited queues if there is either no or slow listener.  This can cause
> system level memory pressure or OOMs.  So, it's better to account the
> fsnotify kmem caches to the memcg of the listener.
> 
> However the listener can be in a different memcg than the memcg of the
> producer and these allocations happen in the context of the event
> producer. This patch introduces remote memcg charging API which the
> producer can use to charge the allocations to the memcg of the listener.
> 
> There are seven fsnotify kmem caches and among them allocations from
> dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
> inotify_inode_mark_cachep happens in the context of syscall from the
> listener.  So, SLAB_ACCOUNT is enough for these caches.
> 
> The objects from fsnotify_mark_connector_cachep are not accounted as they
> are small compared to the notification mark or events and it is unclear
> whom to account connector to since it is shared by all events attached to
> the inode.
> 
> The allocations from the event caches happen in the context of the event
> producer.  For such caches we will need to remote charge the allocations
> to the listener's memcg.  Thus we save the memcg reference in the
> fsnotify_group structure of the listener.
> 
> This patch has also moved the members of fsnotify_group to keep the size
> same, at least for 64 bit build, even with additional member by filling
> the holes.

...

>  static int __init fanotify_user_setup(void)
>  {
> -	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
> +	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
> +					 SLAB_PANIC|SLAB_ACCOUNT);
>  	fanotify_event_cachep = KMEM_CACHE(fanotify_event_info, SLAB_PANIC);
>  	if (IS_ENABLED(CONFIG_FANOTIFY_ACCESS_PERMISSIONS)) {
>  		fanotify_perm_event_cachep =

Why don't you setup also fanotify_event_cachep and
fanotify_perm_event_cachep caches with SLAB_ACCOUNT and instead specify
__GFP_ACCOUNT manually? Otherwise the patch looks good to me.

								Honza
Shakeel Butt June 28, 2018, 7:21 p.m. UTC | #2
On Thu, Jun 28, 2018 at 12:03 PM Jan Kara <jack@suse.cz> wrote:
>
> On Wed 27-06-18 12:12:49, Shakeel Butt wrote:
> > A lot of memory can be consumed by the events generated for the huge or
> > unlimited queues if there is either no or slow listener.  This can cause
> > system level memory pressure or OOMs.  So, it's better to account the
> > fsnotify kmem caches to the memcg of the listener.
> >
> > However the listener can be in a different memcg than the memcg of the
> > producer and these allocations happen in the context of the event
> > producer. This patch introduces remote memcg charging API which the
> > producer can use to charge the allocations to the memcg of the listener.
> >
> > There are seven fsnotify kmem caches and among them allocations from
> > dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
> > inotify_inode_mark_cachep happens in the context of syscall from the
> > listener.  So, SLAB_ACCOUNT is enough for these caches.
> >
> > The objects from fsnotify_mark_connector_cachep are not accounted as they
> > are small compared to the notification mark or events and it is unclear
> > whom to account connector to since it is shared by all events attached to
> > the inode.
> >
> > The allocations from the event caches happen in the context of the event
> > producer.  For such caches we will need to remote charge the allocations
> > to the listener's memcg.  Thus we save the memcg reference in the
> > fsnotify_group structure of the listener.
> >
> > This patch has also moved the members of fsnotify_group to keep the size
> > same, at least for 64 bit build, even with additional member by filling
> > the holes.
>
> ...
>
> >  static int __init fanotify_user_setup(void)
> >  {
> > -     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
> > +     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
> > +                                      SLAB_PANIC|SLAB_ACCOUNT);
> >       fanotify_event_cachep = KMEM_CACHE(fanotify_event_info, SLAB_PANIC);
> >       if (IS_ENABLED(CONFIG_FANOTIFY_ACCESS_PERMISSIONS)) {
> >               fanotify_perm_event_cachep =
>
> Why don't you setup also fanotify_event_cachep and
> fanotify_perm_event_cachep caches with SLAB_ACCOUNT and instead specify
> __GFP_ACCOUNT manually? Otherwise the patch looks good to me.
>

Hi Jan, IMHO having a visible __GFP_ACCOUNT along with
memalloc_use_memcg() makes the code more explicit and readable that we
want to targeted/remote memcg charging. However if you think
otherwise, I will replace __GFP_ACCOUNT with SLAB_ACCOUNT.

thanks,
Shakeel
Michal Hocko June 29, 2018, 9:52 a.m. UTC | #3
On Thu 28-06-18 12:21:26, Shakeel Butt wrote:
> On Thu, Jun 28, 2018 at 12:03 PM Jan Kara <jack@suse.cz> wrote:
> >
> > On Wed 27-06-18 12:12:49, Shakeel Butt wrote:
> > > A lot of memory can be consumed by the events generated for the huge or
> > > unlimited queues if there is either no or slow listener.  This can cause
> > > system level memory pressure or OOMs.  So, it's better to account the
> > > fsnotify kmem caches to the memcg of the listener.
> > >
> > > However the listener can be in a different memcg than the memcg of the
> > > producer and these allocations happen in the context of the event
> > > producer. This patch introduces remote memcg charging API which the
> > > producer can use to charge the allocations to the memcg of the listener.
> > >
> > > There are seven fsnotify kmem caches and among them allocations from
> > > dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
> > > inotify_inode_mark_cachep happens in the context of syscall from the
> > > listener.  So, SLAB_ACCOUNT is enough for these caches.
> > >
> > > The objects from fsnotify_mark_connector_cachep are not accounted as they
> > > are small compared to the notification mark or events and it is unclear
> > > whom to account connector to since it is shared by all events attached to
> > > the inode.
> > >
> > > The allocations from the event caches happen in the context of the event
> > > producer.  For such caches we will need to remote charge the allocations
> > > to the listener's memcg.  Thus we save the memcg reference in the
> > > fsnotify_group structure of the listener.
> > >
> > > This patch has also moved the members of fsnotify_group to keep the size
> > > same, at least for 64 bit build, even with additional member by filling
> > > the holes.
> >
> > ...
> >
> > >  static int __init fanotify_user_setup(void)
> > >  {
> > > -     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
> > > +     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
> > > +                                      SLAB_PANIC|SLAB_ACCOUNT);
> > >       fanotify_event_cachep = KMEM_CACHE(fanotify_event_info, SLAB_PANIC);
> > >       if (IS_ENABLED(CONFIG_FANOTIFY_ACCESS_PERMISSIONS)) {
> > >               fanotify_perm_event_cachep =
> >
> > Why don't you setup also fanotify_event_cachep and
> > fanotify_perm_event_cachep caches with SLAB_ACCOUNT and instead specify
> > __GFP_ACCOUNT manually? Otherwise the patch looks good to me.
> >
> 
> Hi Jan, IMHO having a visible __GFP_ACCOUNT along with
> memalloc_use_memcg() makes the code more explicit and readable that we
> want to targeted/remote memcg charging.

Agreed. If you had an implicit SLAB_ACCOUNT then you could get
inconsistencies when some allocations would get charged to the current
task while others would not.
Jan Kara June 29, 2018, 5:44 p.m. UTC | #4
On Thu 28-06-18 12:21:26, Shakeel Butt wrote:
> On Thu, Jun 28, 2018 at 12:03 PM Jan Kara <jack@suse.cz> wrote:
> >
> > On Wed 27-06-18 12:12:49, Shakeel Butt wrote:
> > > A lot of memory can be consumed by the events generated for the huge or
> > > unlimited queues if there is either no or slow listener.  This can cause
> > > system level memory pressure or OOMs.  So, it's better to account the
> > > fsnotify kmem caches to the memcg of the listener.
> > >
> > > However the listener can be in a different memcg than the memcg of the
> > > producer and these allocations happen in the context of the event
> > > producer. This patch introduces remote memcg charging API which the
> > > producer can use to charge the allocations to the memcg of the listener.
> > >
> > > There are seven fsnotify kmem caches and among them allocations from
> > > dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
> > > inotify_inode_mark_cachep happens in the context of syscall from the
> > > listener.  So, SLAB_ACCOUNT is enough for these caches.
> > >
> > > The objects from fsnotify_mark_connector_cachep are not accounted as they
> > > are small compared to the notification mark or events and it is unclear
> > > whom to account connector to since it is shared by all events attached to
> > > the inode.
> > >
> > > The allocations from the event caches happen in the context of the event
> > > producer.  For such caches we will need to remote charge the allocations
> > > to the listener's memcg.  Thus we save the memcg reference in the
> > > fsnotify_group structure of the listener.
> > >
> > > This patch has also moved the members of fsnotify_group to keep the size
> > > same, at least for 64 bit build, even with additional member by filling
> > > the holes.
> >
> > ...
> >
> > >  static int __init fanotify_user_setup(void)
> > >  {
> > > -     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
> > > +     fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
> > > +                                      SLAB_PANIC|SLAB_ACCOUNT);
> > >       fanotify_event_cachep = KMEM_CACHE(fanotify_event_info, SLAB_PANIC);
> > >       if (IS_ENABLED(CONFIG_FANOTIFY_ACCESS_PERMISSIONS)) {
> > >               fanotify_perm_event_cachep =
> >
> > Why don't you setup also fanotify_event_cachep and
> > fanotify_perm_event_cachep caches with SLAB_ACCOUNT and instead specify
> > __GFP_ACCOUNT manually? Otherwise the patch looks good to me.
> >
> 
> Hi Jan, IMHO having a visible __GFP_ACCOUNT along with
> memalloc_use_memcg() makes the code more explicit and readable that we
> want to targeted/remote memcg charging. However if you think
> otherwise, I will replace __GFP_ACCOUNT with SLAB_ACCOUNT.

OK, fair enough.

								Honza

Patch
diff mbox

diff --git a/fs/notify/dnotify/dnotify.c b/fs/notify/dnotify/dnotify.c
index e2bea2ac5dfb..a6365e6bc047 100644
--- a/fs/notify/dnotify/dnotify.c
+++ b/fs/notify/dnotify/dnotify.c
@@ -384,8 +384,9 @@  int fcntl_dirnotify(int fd, struct file *filp, unsigned long arg)
 
 static int __init dnotify_init(void)
 {
-	dnotify_struct_cache = KMEM_CACHE(dnotify_struct, SLAB_PANIC);
-	dnotify_mark_cache = KMEM_CACHE(dnotify_mark, SLAB_PANIC);
+	dnotify_struct_cache = KMEM_CACHE(dnotify_struct,
+					  SLAB_PANIC|SLAB_ACCOUNT);
+	dnotify_mark_cache = KMEM_CACHE(dnotify_mark, SLAB_PANIC|SLAB_ACCOUNT);
 
 	dnotify_group = fsnotify_alloc_group(&dnotify_fsnotify_ops);
 	if (IS_ERR(dnotify_group))
diff --git a/fs/notify/fanotify/fanotify.c b/fs/notify/fanotify/fanotify.c
index f90842efea13..6ff1f75d156d 100644
--- a/fs/notify/fanotify/fanotify.c
+++ b/fs/notify/fanotify/fanotify.c
@@ -11,6 +11,7 @@ 
 #include <linux/types.h>
 #include <linux/wait.h>
 #include <linux/audit.h>
+#include <linux/sched/mm.h>
 
 #include "fanotify.h"
 
@@ -140,8 +141,8 @@  struct fanotify_event_info *fanotify_alloc_event(struct fsnotify_group *group,
 						 struct inode *inode, u32 mask,
 						 const struct path *path)
 {
-	struct fanotify_event_info *event;
-	gfp_t gfp = GFP_KERNEL;
+	struct fanotify_event_info *event = NULL;
+	gfp_t gfp = GFP_KERNEL | __GFP_ACCOUNT;
 
 	/*
 	 * For queues with unlimited length lost events are not expected and
@@ -151,19 +152,22 @@  struct fanotify_event_info *fanotify_alloc_event(struct fsnotify_group *group,
 	if (group->max_events == UINT_MAX)
 		gfp |= __GFP_NOFAIL;
 
+	/* Whoever is interested in the event, pays for the allocation. */
+	memalloc_use_memcg(group->memcg);
+
 	if (fanotify_is_perm_event(mask)) {
 		struct fanotify_perm_event_info *pevent;
 
 		pevent = kmem_cache_alloc(fanotify_perm_event_cachep, gfp);
 		if (!pevent)
-			return NULL;
+			goto out;
 		event = &pevent->fae;
 		pevent->response = 0;
 		goto init;
 	}
 	event = kmem_cache_alloc(fanotify_event_cachep, gfp);
 	if (!event)
-		return NULL;
+		goto out;
 init: __maybe_unused
 	fsnotify_init_event(&event->fse, inode, mask);
 	event->tgid = get_pid(task_tgid(current));
@@ -174,6 +178,8 @@  init: __maybe_unused
 		event->path.mnt = NULL;
 		event->path.dentry = NULL;
 	}
+out:
+	memalloc_unuse_memcg();
 	return event;
 }
 
diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c
index ec4d8c59d0e3..0cf45041dc32 100644
--- a/fs/notify/fanotify/fanotify_user.c
+++ b/fs/notify/fanotify/fanotify_user.c
@@ -16,6 +16,7 @@ 
 #include <linux/uaccess.h>
 #include <linux/compat.h>
 #include <linux/sched/signal.h>
+#include <linux/memcontrol.h>
 
 #include <asm/ioctls.h>
 
@@ -756,6 +757,7 @@  SYSCALL_DEFINE2(fanotify_init, unsigned int, flags, unsigned int, event_f_flags)
 
 	group->fanotify_data.user = user;
 	atomic_inc(&user->fanotify_listeners);
+	group->memcg = get_mem_cgroup_from_mm(current->mm);
 
 	oevent = fanotify_alloc_event(group, NULL, FS_Q_OVERFLOW, NULL);
 	if (unlikely(!oevent)) {
@@ -957,7 +959,8 @@  COMPAT_SYSCALL_DEFINE6(fanotify_mark,
  */
 static int __init fanotify_user_setup(void)
 {
-	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark, SLAB_PANIC);
+	fanotify_mark_cache = KMEM_CACHE(fsnotify_mark,
+					 SLAB_PANIC|SLAB_ACCOUNT);
 	fanotify_event_cachep = KMEM_CACHE(fanotify_event_info, SLAB_PANIC);
 	if (IS_ENABLED(CONFIG_FANOTIFY_ACCESS_PERMISSIONS)) {
 		fanotify_perm_event_cachep =
diff --git a/fs/notify/group.c b/fs/notify/group.c
index aa5468f23e45..c03b83662876 100644
--- a/fs/notify/group.c
+++ b/fs/notify/group.c
@@ -22,6 +22,7 @@ 
 #include <linux/srcu.h>
 #include <linux/rculist.h>
 #include <linux/wait.h>
+#include <linux/memcontrol.h>
 
 #include <linux/fsnotify_backend.h>
 #include "fsnotify.h"
@@ -36,6 +37,8 @@  static void fsnotify_final_destroy_group(struct fsnotify_group *group)
 	if (group->ops->free_group_priv)
 		group->ops->free_group_priv(group);
 
+	mem_cgroup_put(group->memcg);
+
 	kfree(group);
 }
 
diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c
index 9ab6dde38a14..52e167d04b11 100644
--- a/fs/notify/inotify/inotify_fsnotify.c
+++ b/fs/notify/inotify/inotify_fsnotify.c
@@ -31,6 +31,7 @@ 
 #include <linux/types.h>
 #include <linux/sched.h>
 #include <linux/sched/user.h>
+#include <linux/sched/mm.h>
 
 #include "inotify.h"
 
@@ -98,7 +99,11 @@  int inotify_handle_event(struct fsnotify_group *group,
 	i_mark = container_of(inode_mark, struct inotify_inode_mark,
 			      fsn_mark);
 
-	event = kmalloc(alloc_len, GFP_KERNEL);
+	/* Whoever is interested in the event, pays for the allocation. */
+	memalloc_use_memcg(group->memcg);
+	event = kmalloc(alloc_len, GFP_KERNEL | __GFP_ACCOUNT);
+	memalloc_unuse_memcg();
+
 	if (unlikely(!event)) {
 		/*
 		 * Treat lost event due to ENOMEM the same way as queue
diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c
index 1cf5b779d862..749c46ababa0 100644
--- a/fs/notify/inotify/inotify_user.c
+++ b/fs/notify/inotify/inotify_user.c
@@ -38,6 +38,7 @@ 
 #include <linux/uaccess.h>
 #include <linux/poll.h>
 #include <linux/wait.h>
+#include <linux/memcontrol.h>
 
 #include "inotify.h"
 #include "../fdinfo.h"
@@ -636,6 +637,7 @@  static struct fsnotify_group *inotify_new_group(unsigned int max_events)
 	oevent->name_len = 0;
 
 	group->max_events = max_events;
+	group->memcg = get_mem_cgroup_from_mm(current->mm);
 
 	spin_lock_init(&group->inotify_data.idr_lock);
 	idr_init(&group->inotify_data.idr);
@@ -808,7 +810,8 @@  static int __init inotify_user_setup(void)
 
 	BUG_ON(hweight32(ALL_INOTIFY_BITS) != 21);
 
-	inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark, SLAB_PANIC);
+	inotify_inode_mark_cachep = KMEM_CACHE(inotify_inode_mark,
+					       SLAB_PANIC|SLAB_ACCOUNT);
 
 	inotify_max_queued_events = 16384;
 	init_user_ns.ucount_max[UCOUNT_INOTIFY_INSTANCES] = 128;
diff --git a/include/linux/fsnotify_backend.h b/include/linux/fsnotify_backend.h
index b38964a7a521..a0c4790c5302 100644
--- a/include/linux/fsnotify_backend.h
+++ b/include/linux/fsnotify_backend.h
@@ -84,6 +84,8 @@  struct fsnotify_event_private_data;
 struct fsnotify_fname;
 struct fsnotify_iter_info;
 
+struct mem_cgroup;
+
 /*
  * Each group much define these ops.  The fsnotify infrastructure will call
  * these operations for each relevant group.
@@ -127,6 +129,8 @@  struct fsnotify_event {
  * everything will be cleaned up.
  */
 struct fsnotify_group {
+	const struct fsnotify_ops *ops;	/* how this group handles things */
+
 	/*
 	 * How the refcnt is used is up to each group.  When the refcnt hits 0
 	 * fsnotify will clean up all of the resources associated with this group.
@@ -137,8 +141,6 @@  struct fsnotify_group {
 	 */
 	refcount_t refcnt;		/* things with interest in this group */
 
-	const struct fsnotify_ops *ops;	/* how this group handles things */
-
 	/* needed to send notification to userspace */
 	spinlock_t notification_lock;		/* protect the notification_list */
 	struct list_head notification_list;	/* list of event_holder this group needs to send to userspace */
@@ -160,6 +162,8 @@  struct fsnotify_group {
 	atomic_t num_marks;		/* 1 for each mark and 1 for not being
 					 * past the point of no return when freeing
 					 * a group */
+	atomic_t user_waits;		/* Number of tasks waiting for user
+					 * response */
 	struct list_head marks_list;	/* all inode marks for this group */
 
 	struct fasync_struct *fsn_fa;    /* async notification */
@@ -167,8 +171,8 @@  struct fsnotify_group {
 	struct fsnotify_event *overflow_event;	/* Event we queue when the
 						 * notification list is too
 						 * full */
-	atomic_t user_waits;		/* Number of tasks waiting for user
-					 * response */
+
+	struct mem_cgroup *memcg;	/* memcg to charge allocations */
 
 	/* groups can define private fields here or use the void *private */
 	union {
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 3607913032be..cb04b382c8d2 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -378,6 +378,8 @@  struct lruvec *mem_cgroup_page_lruvec(struct page *, struct pglist_data *);
 bool task_in_mem_cgroup(struct task_struct *task, struct mem_cgroup *memcg);
 struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p);
 
+struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm);
+
 static inline
 struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
 	return css ? container_of(css, struct mem_cgroup, css) : NULL;
@@ -385,7 +387,8 @@  struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css){
 
 static inline void mem_cgroup_put(struct mem_cgroup *memcg)
 {
-	css_put(&memcg->css);
+	if (memcg)
+		css_put(&memcg->css);
 }
 
 #define mem_cgroup_from_counter(counter, member)	\
@@ -857,6 +860,11 @@  static inline bool task_in_mem_cgroup(struct task_struct *task,
 	return true;
 }
 
+static inline struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
+{
+	return NULL;
+}
+
 static inline void mem_cgroup_put(struct mem_cgroup *memcg)
 {
 }
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 87bf02d93a27..9cba7f874443 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1149,6 +1149,9 @@  struct task_struct {
 
 	/* Number of pages to reclaim on returning to userland: */
 	unsigned int			memcg_nr_pages_over_high;
+
+	/* Used by memcontrol for targeted memcg charge: */
+	struct mem_cgroup		*active_memcg;
 #endif
 
 #ifdef CONFIG_UPROBES
diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h
index 44d356f5e47c..2ac55340cb73 100644
--- a/include/linux/sched/mm.h
+++ b/include/linux/sched/mm.h
@@ -248,6 +248,43 @@  static inline void memalloc_noreclaim_restore(unsigned int flags)
 	current->flags = (current->flags & ~PF_MEMALLOC) | flags;
 }
 
+#ifdef CONFIG_MEMCG
+/**
+ * memalloc_use_memcg - Starts the remote memcg charging scope.
+ * @memcg: memcg to charge.
+ *
+ * This function marks the beginning of the remote memcg charging scope. All the
+ * __GFP_ACCOUNT allocations till the end of the scope will be charged to the
+ * given memcg.
+ *
+ * NOTE: This function is not nesting safe.
+ */
+static inline void memalloc_use_memcg(struct mem_cgroup *memcg)
+{
+	WARN_ON_ONCE(current->active_memcg);
+	current->active_memcg = memcg;
+}
+
+/**
+ * memalloc_unuse_memcg - Ends the remote memcg charging scope.
+ *
+ * This function marks the end of the remote memcg charging scope started by
+ * memalloc_use_memcg().
+ */
+static inline void memalloc_unuse_memcg(void)
+{
+	current->active_memcg = NULL;
+}
+#else
+static inline void memalloc_use_memcg(struct mem_cgroup *memcg)
+{
+}
+
+static inline void memalloc_unuse_memcg(void)
+{
+}
+#endif
+
 #ifdef CONFIG_MEMBARRIER
 enum {
 	MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY		= (1U << 0),
diff --git a/kernel/fork.c b/kernel/fork.c
index a64d0a19f174..0b07db6fef06 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -843,6 +843,9 @@  static struct task_struct *dup_task_struct(struct task_struct *orig, int node)
 	tsk->fail_nth = 0;
 #endif
 
+#ifdef CONFIG_MEMCG
+	tsk->active_memcg = NULL;
+#endif
 	return tsk;
 
 free_stack:
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 02e77c88967a..b25ca5c13196 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -678,9 +678,20 @@  struct mem_cgroup *mem_cgroup_from_task(struct task_struct *p)
 }
 EXPORT_SYMBOL(mem_cgroup_from_task);
 
-static struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
+/**
+ * get_mem_cgroup_from_mm: Obtain a reference on given mm_struct's memcg.
+ * @mm: mm from which memcg should be extracted. It can be NULL.
+ *
+ * Obtain a reference on mm->memcg and returns it if successful. Otherwise
+ * root_mem_cgroup is returned. However if mem_cgroup is disabled, NULL is
+ * returned.
+ */
+struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
 {
-	struct mem_cgroup *memcg = NULL;
+	struct mem_cgroup *memcg;
+
+	if (mem_cgroup_disabled())
+		return NULL;
 
 	rcu_read_lock();
 	do {
@@ -700,6 +711,24 @@  static struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm)
 	rcu_read_unlock();
 	return memcg;
 }
+EXPORT_SYMBOL(get_mem_cgroup_from_mm);
+
+/**
+ * If current->active_memcg is non-NULL, do not fallback to current->mm->memcg.
+ */
+static __always_inline struct mem_cgroup *get_mem_cgroup_from_current(void)
+{
+	if (unlikely(current->active_memcg)) {
+		struct mem_cgroup *memcg = root_mem_cgroup;
+
+		rcu_read_lock();
+		if (css_tryget_online(&current->active_memcg->css))
+			memcg = current->active_memcg;
+		rcu_read_unlock();
+		return memcg;
+	}
+	return get_mem_cgroup_from_mm(current->mm);
+}
 
 /**
  * mem_cgroup_iter - iterate over memory cgroup hierarchy
@@ -2260,7 +2289,7 @@  struct kmem_cache *memcg_kmem_get_cache(struct kmem_cache *cachep)
 	if (current->memcg_kmem_skip_account)
 		return cachep;
 
-	memcg = get_mem_cgroup_from_mm(current->mm);
+	memcg = get_mem_cgroup_from_current();
 	kmemcg_id = READ_ONCE(memcg->kmemcg_id);
 	if (kmemcg_id < 0)
 		goto out;
@@ -2344,7 +2373,7 @@  int memcg_kmem_charge(struct page *page, gfp_t gfp, int order)
 	if (memcg_kmem_bypass())
 		return 0;
 
-	memcg = get_mem_cgroup_from_mm(current->mm);
+	memcg = get_mem_cgroup_from_current();
 	if (!mem_cgroup_is_root(memcg)) {
 		ret = memcg_kmem_charge_memcg(page, gfp, order, memcg);
 		if (!ret)