linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
To: cgroups@vger.kernel.org, linux-mm@kvack.org
Cc: "Andrew Morton" <akpm@linux-foundation.org>,
	"Johannes Weiner" <hannes@cmpxchg.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>,
	"Sebastian Andrzej Siewior" <bigeasy@linutronix.de>
Subject: [PATCH 4/4] mm/memcg: Allow the task_obj optimization only on non-PREEMPTIBLE kernels.
Date: Tue, 25 Jan 2022 17:43:37 +0100	[thread overview]
Message-ID: <20220125164337.2071854-5-bigeasy@linutronix.de> (raw)
In-Reply-To: <20220125164337.2071854-1-bigeasy@linutronix.de>

Based on my understanding the optimisation with task_obj for in_task()
mask sense on non-PREEMPTIBLE kernels because preempt_disable()/enable()
is optimized away. This could be then restricted to !CONFIG_PREEMPTION kernel
instead to only PREEMPT_RT.
With CONFIG_PREEMPT_DYNAMIC a non-PREEMPTIBLE kernel can also be
configured but these kernels always have preempt_disable()/enable()
present so it probably makes no sense here for the optimisation.

I did a micro benchmark with disabled interrupts and a loop of
100.000.000 invokcations of kfree(kmalloc()). Based on the results it
makes no sense to add an exception based on dynamic preemption.

Restrict the optimisation to !CONFIG_PREEMPTION kernels.

Link: https://lore.kernel.org/all/YdX+INO9gQje6d0S@linutronix.de
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---
 mm/memcontrol.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 2d8be88c00888..20ea8f28ad99b 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -2030,7 +2030,7 @@ struct memcg_stock_pcp {
 	local_lock_t stock_lock;
 	struct mem_cgroup *cached; /* this never be root cgroup */
 	unsigned int nr_pages;
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	/* Protects only task_obj */
 	local_lock_t task_obj_lock;
 	struct obj_stock task_obj;
@@ -2043,7 +2043,7 @@ struct memcg_stock_pcp {
 };
 static DEFINE_PER_CPU(struct memcg_stock_pcp, memcg_stock) = {
 	.stock_lock = INIT_LOCAL_LOCK(stock_lock),
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	.task_obj_lock = INIT_LOCAL_LOCK(task_obj_lock),
 #endif
 };
@@ -2132,7 +2132,7 @@ static void drain_local_stock(struct work_struct *dummy)
 	 * drain_stock races is that we always operate on local CPU stock
 	 * here with IRQ disabled
 	 */
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	local_lock(&memcg_stock.task_obj_lock);
 	old = drain_obj_stock(&this_cpu_ptr(&memcg_stock)->task_obj, NULL);
 	local_unlock(&memcg_stock.task_obj_lock);
@@ -2741,7 +2741,7 @@ static inline struct obj_stock *get_obj_stock(unsigned long *pflags,
 {
 	struct memcg_stock_pcp *stock;
 
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	if (likely(in_task())) {
 		*pflags = 0UL;
 		*stock_lock_acquried = false;
@@ -2759,7 +2759,7 @@ static inline struct obj_stock *get_obj_stock(unsigned long *pflags,
 static inline void put_obj_stock(unsigned long flags,
 				 bool stock_lock_acquried)
 {
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	if (likely(!stock_lock_acquried)) {
 		local_unlock(&memcg_stock.task_obj_lock);
 		return;
@@ -3177,7 +3177,7 @@ static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
 {
 	struct mem_cgroup *memcg;
 
-#ifndef CONFIG_PREEMPT_RT
+#ifndef CONFIG_PREEMPTION
 	if (in_task() && stock->task_obj.cached_objcg) {
 		memcg = obj_cgroup_memcg(stock->task_obj.cached_objcg);
 		if (memcg && mem_cgroup_is_descendant(memcg, root_memcg))
-- 
2.34.1



  parent reply	other threads:[~2022-01-25 16:43 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-01-25 16:43 [PATCH 0/4] mm/memcg: Address PREEMPT_RT problems instead of disabling it Sebastian Andrzej Siewior
2022-01-25 16:43 ` [PATCH 1/4] mm/memcg: Disable threshold event handlers on PREEMPT_RT Sebastian Andrzej Siewior
2022-01-26 14:40   ` Michal Hocko
2022-01-26 14:45     ` Sebastian Andrzej Siewior
2022-01-26 15:04       ` Michal Koutný
2022-01-27 13:36         ` Sebastian Andrzej Siewior
2022-01-26 15:21       ` Michal Hocko
2022-01-25 16:43 ` [PATCH 2/4] mm/memcg: Protect per-CPU counter by disabling preemption on PREEMPT_RT where needed Sebastian Andrzej Siewior
2022-01-26 10:06   ` Vlastimil Babka
2022-01-26 11:24     ` Sebastian Andrzej Siewior
2022-01-26 14:56   ` Michal Hocko
2022-01-25 16:43 ` [PATCH 3/4] mm/memcg: Add a local_lock_t for IRQ and TASK object Sebastian Andrzej Siewior
2022-01-26 15:20   ` Michal Hocko
2022-01-27 11:53     ` Sebastian Andrzej Siewior
2022-02-01 12:04       ` Michal Hocko
2022-02-01 12:11         ` Sebastian Andrzej Siewior
2022-02-01 15:29           ` Michal Hocko
2022-02-03  9:54             ` Sebastian Andrzej Siewior
2022-02-03 10:09               ` Michal Hocko
2022-02-03 11:09                 ` Sebastian Andrzej Siewior
2022-02-08 17:58                 ` Shakeel Butt
2022-02-09  9:17                   ` Michal Hocko
2022-01-26 16:57   ` Vlastimil Babka
2022-01-31 15:06     ` Sebastian Andrzej Siewior
2022-02-03 16:01       ` Vlastimil Babka
2022-02-08 17:17         ` Sebastian Andrzej Siewior
2022-02-08 17:28           ` Michal Hocko
2022-02-09  1:48   ` [mm/memcg] 86895e1e85: WARNING:possible_circular_locking_dependency_detected kernel test robot
2022-01-25 16:43 ` Sebastian Andrzej Siewior [this message]
2022-01-25 23:21 ` [PATCH 0/4] mm/memcg: Address PREEMPT_RT problems instead of disabling it Andrew Morton
2022-01-26  7:30   ` 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=20220125164337.2071854-5-bigeasy@linutronix.de \
    --to=bigeasy@linutronix.de \
    --cc=akpm@linux-foundation.org \
    --cc=cgroups@vger.kernel.org \
    --cc=hannes@cmpxchg.org \
    --cc=linux-mm@kvack.org \
    --cc=longman@redhat.com \
    --cc=mhocko@kernel.org \
    --cc=mkoutny@suse.com \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=vdavydov.dev@gmail.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox