All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sebastian Andrzej Siewior <bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
To: Johannes Weiner <hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
	linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org,
	"Andrew Morton"
	<akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>,
	"Michal Hocko" <mhocko-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
	"Michal Koutný" <mkoutny-IBi9RG/b67k@public.gmane.org>,
	"Peter Zijlstra" <peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>,
	"Thomas Gleixner" <tglx-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>,
	"Vladimir Davydov"
	<vdavydov.dev-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>,
	"Waiman Long" <longman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
	"kernel test robot"
	<oliver.sang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Subject: Re: [PATCH v2 4/4] mm/memcg: Protect memcg_stock with a local_lock_t
Date: Wed, 16 Feb 2022 16:51:14 +0100	[thread overview]
Message-ID: <Yg0dctKholvzADYP@linutronix.de> (raw)
In-Reply-To: <YgqB77SaViGRAtgt-druUgvl0LCNAfugRpC6u6w@public.gmane.org>

On 2022-02-14 11:23:11 [-0500], Johannes Weiner wrote:
> Hi Sebastian,
Hi Johannes,

> This is a bit dubious in terms of layering. It's an objcg operation,
> but what's "locked" isn't the objcg, it's the underlying stock. That
> function then looks it up again, even though we have it right there.
> 
> You can open-code it and factor out the stock operation instead, and
> it makes things much simpler and clearer.
> 
> I.e. something like this (untested!):

This then:

------>8------

From: Johannes Weiner <hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
Date: Wed, 16 Feb 2022 13:25:49 +0100
Subject: [PATCH] mm/memcg: Opencode the inner part of
 obj_cgroup_uncharge_pages() in drain_obj_stock()

Provide the inner part of refill_stock() as __refill_stock() without
disabling interrupts. This eases the integration of local_lock_t where
recursive locking must be avoided.
Open code obj_cgroup_uncharge_pages() in drain_obj_stock() and use
__refill_stock(). The caller of drain_obj_stock() already disables
interrupts.

[bigeasy: Patch body around Johannes' diff ]

Signed-off-by: Johannes Weiner <hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
---
 mm/memcontrol.c | 25 +++++++++++++++++++------
 1 file changed, 19 insertions(+), 6 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 69130a5fe3d51..f574f2e1cc399 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2227,12 +2227,9 @@ static void drain_local_stock(struct work_struct *dummy)
  * Cache charges(val) to local per_cpu area.
  * This will be consumed by consume_stock() function, later.
  */
-static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
+static void __refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
 {
 	struct memcg_stock_pcp *stock;
-	unsigned long flags;
-
-	local_irq_save(flags);
 
 	stock = this_cpu_ptr(&memcg_stock);
 	if (stock->cached != memcg) { /* reset if necessary */
@@ -2244,7 +2241,14 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
 
 	if (stock->nr_pages > MEMCG_CHARGE_BATCH)
 		drain_stock(stock);
+}
 
+static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+	__refill_stock(memcg, nr_pages);
 	local_irq_restore(flags);
 }
 
@@ -3151,8 +3155,17 @@ static void drain_obj_stock(struct memcg_stock_pcp *stock)
 		unsigned int nr_pages = stock->nr_bytes >> PAGE_SHIFT;
 		unsigned int nr_bytes = stock->nr_bytes & (PAGE_SIZE - 1);
 
-		if (nr_pages)
-			obj_cgroup_uncharge_pages(old, nr_pages);
+		if (nr_pages) {
+			struct mem_cgroup *memcg;
+
+			memcg = get_mem_cgroup_from_objcg(old);
+
+			if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
+				page_counter_uncharge(&memcg->kmem, nr_pages);
+			__refill_stock(memcg, nr_pages);
+
+			css_put(&memcg->css);
+		}
 
 		/*
 		 * The leftover is flushed to the centralized per-memcg value.
-- 
2.34.1


WARNING: multiple messages have this Message-ID (diff)
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
To: Johannes Weiner <hannes@cmpxchg.org>
Cc: cgroups@vger.kernel.org, linux-mm@kvack.org,
	"Andrew Morton" <akpm@linux-foundation.org>,
	"Michal Hocko" <mhocko@kernel.org>,
	"Michal Koutný" <mkoutny@suse.com>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Vladimir Davydov" <vdavydov.dev@gmail.com>,
	"Waiman Long" <longman@redhat.com>,
	"kernel test robot" <oliver.sang@intel.com>
Subject: Re: [PATCH v2 4/4] mm/memcg: Protect memcg_stock with a local_lock_t
Date: Wed, 16 Feb 2022 16:51:14 +0100	[thread overview]
Message-ID: <Yg0dctKholvzADYP@linutronix.de> (raw)
In-Reply-To: <YgqB77SaViGRAtgt@cmpxchg.org>

On 2022-02-14 11:23:11 [-0500], Johannes Weiner wrote:
> Hi Sebastian,
Hi Johannes,

> This is a bit dubious in terms of layering. It's an objcg operation,
> but what's "locked" isn't the objcg, it's the underlying stock. That
> function then looks it up again, even though we have it right there.
> 
> You can open-code it and factor out the stock operation instead, and
> it makes things much simpler and clearer.
> 
> I.e. something like this (untested!):

This then:

------>8------

From: Johannes Weiner <hannes@cmpxchg.org>
Date: Wed, 16 Feb 2022 13:25:49 +0100
Subject: [PATCH] mm/memcg: Opencode the inner part of
 obj_cgroup_uncharge_pages() in drain_obj_stock()

Provide the inner part of refill_stock() as __refill_stock() without
disabling interrupts. This eases the integration of local_lock_t where
recursive locking must be avoided.
Open code obj_cgroup_uncharge_pages() in drain_obj_stock() and use
__refill_stock(). The caller of drain_obj_stock() already disables
interrupts.

[bigeasy: Patch body around Johannes' diff ]

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
 mm/memcontrol.c | 25 +++++++++++++++++++------
 1 file changed, 19 insertions(+), 6 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 69130a5fe3d51..f574f2e1cc399 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2227,12 +2227,9 @@ static void drain_local_stock(struct work_struct *dummy)
  * Cache charges(val) to local per_cpu area.
  * This will be consumed by consume_stock() function, later.
  */
-static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
+static void __refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
 {
 	struct memcg_stock_pcp *stock;
-	unsigned long flags;
-
-	local_irq_save(flags);
 
 	stock = this_cpu_ptr(&memcg_stock);
 	if (stock->cached != memcg) { /* reset if necessary */
@@ -2244,7 +2241,14 @@ static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
 
 	if (stock->nr_pages > MEMCG_CHARGE_BATCH)
 		drain_stock(stock);
+}
 
+static void refill_stock(struct mem_cgroup *memcg, unsigned int nr_pages)
+{
+	unsigned long flags;
+
+	local_irq_save(flags);
+	__refill_stock(memcg, nr_pages);
 	local_irq_restore(flags);
 }
 
@@ -3151,8 +3155,17 @@ static void drain_obj_stock(struct memcg_stock_pcp *stock)
 		unsigned int nr_pages = stock->nr_bytes >> PAGE_SHIFT;
 		unsigned int nr_bytes = stock->nr_bytes & (PAGE_SIZE - 1);
 
-		if (nr_pages)
-			obj_cgroup_uncharge_pages(old, nr_pages);
+		if (nr_pages) {
+			struct mem_cgroup *memcg;
+
+			memcg = get_mem_cgroup_from_objcg(old);
+
+			if (!cgroup_subsys_on_dfl(memory_cgrp_subsys))
+				page_counter_uncharge(&memcg->kmem, nr_pages);
+			__refill_stock(memcg, nr_pages);
+
+			css_put(&memcg->css);
+		}
 
 		/*
 		 * The leftover is flushed to the centralized per-memcg value.
-- 
2.34.1



  parent reply	other threads:[~2022-02-16 15:51 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-11 22:35 [PATCH v2 0/4] mm/memcg: Address PREEMPT_RT problems instead of disabling it Sebastian Andrzej Siewior
2022-02-11 22:35 ` Sebastian Andrzej Siewior
     [not found] ` <20220211223537.2175879-1-bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-11 22:35   ` [PATCH v2 1/4] mm/memcg: Revert ("mm/memcg: optimize user context object stock access") Sebastian Andrzej Siewior
2022-02-11 22:35     ` Sebastian Andrzej Siewior
     [not found]     ` <20220211223537.2175879-2-bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-14 16:23       ` Johannes Weiner
2022-02-14 16:23         ` Johannes Weiner
2022-02-14 19:45       ` Roman Gushchin
2022-02-14 19:45         ` Roman Gushchin
2022-02-11 22:35   ` [PATCH v2 2/4] mm/memcg: Disable threshold event handlers on PREEMPT_RT Sebastian Andrzej Siewior
2022-02-11 22:35     ` Sebastian Andrzej Siewior
     [not found]     ` <20220211223537.2175879-3-bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-14 16:23       ` Johannes Weiner
2022-02-14 16:23         ` Johannes Weiner
2022-02-14 19:46       ` Roman Gushchin
2022-02-14 19:46         ` Roman Gushchin
2022-02-11 22:35   ` [PATCH v2 3/4] mm/memcg: Protect per-CPU counter by disabling preemption on PREEMPT_RT where needed Sebastian Andrzej Siewior
2022-02-11 22:35     ` Sebastian Andrzej Siewior
     [not found]     ` <20220211223537.2175879-4-bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-14 16:46       ` Johannes Weiner
2022-02-14 16:46         ` Johannes Weiner
     [not found]         ` <YgqHSIa/WvJSXERe-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
2022-02-14 19:53           ` Roman Gushchin
2022-02-14 19:53             ` Roman Gushchin
2022-02-15 18:01           ` Sebastian Andrzej Siewior
2022-02-15 18:01             ` Sebastian Andrzej Siewior
2022-02-11 22:35   ` [PATCH v2 4/4] mm/memcg: Protect memcg_stock with a local_lock_t Sebastian Andrzej Siewior
2022-02-11 22:35     ` Sebastian Andrzej Siewior
     [not found]     ` <20220211223537.2175879-5-bigeasy-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-14 16:23       ` Johannes Weiner
2022-02-14 16:23         ` Johannes Weiner
     [not found]         ` <YgqB77SaViGRAtgt-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
2022-02-16 15:51           ` Sebastian Andrzej Siewior [this message]
2022-02-16 15:51             ` Sebastian Andrzej Siewior
     [not found]             ` <Yg0dctKholvzADYP-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org>
2022-02-16 18:08               ` Johannes Weiner
2022-02-16 18:08                 ` Johannes Weiner
     [not found]                 ` <Yg09t/j5Z0X9L7aX-druUgvl0LCNAfugRpC6u6w@public.gmane.org>
2022-02-17  9:28                   ` Sebastian Andrzej Siewior
2022-02-17  9:28                     ` Sebastian Andrzej Siewior

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Yg0dctKholvzADYP@linutronix.de \
    --to=bigeasy-hfztesqfncyowbw4kg4ksq@public.gmane.org \
    --cc=akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org \
    --cc=cgroups-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
    --cc=hannes-druUgvl0LCNAfugRpC6u6w@public.gmane.org \
    --cc=linux-mm-Bw31MaZKKs3YtjvyW6yDsg@public.gmane.org \
    --cc=longman-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
    --cc=mhocko-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
    --cc=mkoutny-IBi9RG/b67k@public.gmane.org \
    --cc=oliver.sang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org \
    --cc=peterz-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org \
    --cc=tglx-hfZtesqFncYOwBW4kG4KsQ@public.gmane.org \
    --cc=vdavydov.dev-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.