LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Greg Thelen <gthelen@google.com>
To: guro@fb.com, Johannes Weiner <hannes@cmpxchg.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Michal Hocko <mhocko@kernel.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>,
	Tejun Heo <tj@kernel.org>, Cgroups <cgroups@vger.kernel.org>,
	kernel-team@fb.com, Linux MM <linux-mm@kvack.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Greg Thelen <gthelen@google.com>
Subject: [RFC PATCH 2/2] memcg: add memory.min
Date: Sun, 22 Apr 2018 13:26:12 -0700	[thread overview]
Message-ID: <20180422202612.127760-3-gthelen@google.com> (raw)
In-Reply-To: <20180422202612.127760-1-gthelen@google.com>

The new memory.min limit is similar to memory.low, just no bypassing it
when reclaim is desparate.  Prefer oom kills before reclaim memory below
memory.min.  Sharing more code with memory_cgroup_low() is possible, but
the idea is posted here for simplicity.

Signed-off-by: Greg Thelen <gthelen@google.com>
---
 include/linux/memcontrol.h |  8 ++++++
 mm/memcontrol.c            | 58 ++++++++++++++++++++++++++++++++++++++
 mm/vmscan.c                |  3 ++
 3 files changed, 69 insertions(+)

diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index c46016bb25eb..22bb4a88653a 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -178,6 +178,7 @@ struct mem_cgroup {
 	struct page_counter tcpmem;
 
 	/* Normal memory consumption range */
+	unsigned long min;
 	unsigned long low;
 	unsigned long high;
 
@@ -281,6 +282,7 @@ static inline bool mem_cgroup_disabled(void)
 	return !cgroup_subsys_enabled(memory_cgrp_subsys);
 }
 
+bool mem_cgroup_min(struct mem_cgroup *root, struct mem_cgroup *memcg);
 bool mem_cgroup_low(struct mem_cgroup *root, struct mem_cgroup *memcg);
 
 int mem_cgroup_try_charge(struct page *page, struct mm_struct *mm,
@@ -726,6 +728,12 @@ static inline void mem_cgroup_event(struct mem_cgroup *memcg,
 {
 }
 
+static inline bool mem_cgroup_min(struct mem_cgroup *root,
+				  struct mem_cgroup *memcg)
+{
+	return false;
+}
+
 static inline bool mem_cgroup_low(struct mem_cgroup *root,
 				  struct mem_cgroup *memcg)
 {
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 9668f620203a..b2aaed4003b4 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -5058,6 +5058,36 @@ static u64 memory_current_read(struct cgroup_subsys_state *css,
 	return (u64)page_counter_read(&memcg->memory) * PAGE_SIZE;
 }
 
+static int memory_min_show(struct seq_file *m, void *v)
+{
+	struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
+	unsigned long min = READ_ONCE(memcg->min);
+
+	if (min == PAGE_COUNTER_MAX)
+		seq_puts(m, "max\n");
+	else
+		seq_printf(m, "%llu\n", (u64)min * PAGE_SIZE);
+
+	return 0;
+}
+
+static ssize_t memory_min_write(struct kernfs_open_file *of,
+				char *buf, size_t nbytes, loff_t off)
+{
+	struct mem_cgroup *memcg = mem_cgroup_from_css(of_css(of));
+	unsigned long min;
+	int err;
+
+	buf = strstrip(buf);
+	err = page_counter_memparse(buf, "max", &min);
+	if (err)
+		return err;
+
+	memcg->min = min;
+
+	return nbytes;
+}
+
 static int memory_low_show(struct seq_file *m, void *v)
 {
 	struct mem_cgroup *memcg = mem_cgroup_from_css(seq_css(m));
@@ -5288,6 +5318,12 @@ static struct cftype memory_files[] = {
 		.flags = CFTYPE_NOT_ON_ROOT,
 		.read_u64 = memory_current_read,
 	},
+	{
+		.name = "min",
+		.flags = CFTYPE_NOT_ON_ROOT,
+		.seq_show = memory_min_show,
+		.write = memory_min_write,
+	},
 	{
 		.name = "low",
 		.flags = CFTYPE_NOT_ON_ROOT,
@@ -5336,6 +5372,28 @@ struct cgroup_subsys memory_cgrp_subsys = {
 	.early_init = 0,
 };
 
+/**
+ * mem_cgroup_min returns true for a memcg below its min limit.  Such memcg are
+ * excempt from reclaim.
+ */
+bool mem_cgroup_min(struct mem_cgroup *root, struct mem_cgroup *memcg)
+{
+	if (mem_cgroup_disabled())
+		return false;
+
+	if (!root)
+		root = root_mem_cgroup;
+
+	if (memcg == root)
+		return false;
+
+	for (; memcg != root; memcg = parent_mem_cgroup(memcg)) {
+		if (page_counter_read(&memcg->memory) <= memcg->min)
+			return true; /* protect */
+	}
+	return false; /* !protect */
+}
+
 /**
  * mem_cgroup_low - check if memory consumption is below the normal range
  * @root: the top ancestor of the sub-tree being checked
diff --git a/mm/vmscan.c b/mm/vmscan.c
index cd5dc3faaa57..15ae19a38ad5 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -2539,6 +2539,9 @@ static bool shrink_node(pg_data_t *pgdat, struct scan_control *sc)
 			unsigned long reclaimed;
 			unsigned long scanned;
 
+			if (mem_cgroup_min(root, memcg))
+				continue;
+
 			if (mem_cgroup_low(root, memcg)) {
 				if (!sc->memcg_low_reclaim) {
 					sc->memcg_low_skipped = 1;
-- 
2.17.0.484.g0c8726318c-goog

  parent reply	other threads:[~2018-04-22 20:27 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-20 22:33 [RFC] mm: memory.low heirarchical behavior Roman Gushchin
2018-03-21 18:23 ` Johannes Weiner
2018-03-21 19:08   ` Roman Gushchin
2018-04-04 17:07     ` Johannes Weiner
2018-04-05 13:54       ` Roman Gushchin
2018-04-05 15:00         ` Johannes Weiner
2018-03-23 16:37   ` [PATCH v2] mm: memory.low hierarchical behavior Roman Gushchin
2018-04-22 20:26 ` [RFC PATCH 0/2] memory.low,min reclaim Greg Thelen
2018-04-22 20:26   ` [RFC PATCH 1/2] memcg: fix memory.low Greg Thelen
2018-04-22 20:26   ` Greg Thelen [this message]
2018-04-23 10:38   ` [RFC PATCH 0/2] memory.low,min reclaim Roman Gushchin
2018-04-24  0:56     ` Greg Thelen
2018-04-24 10:09       ` Roman Gushchin

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=20180422202612.127760-3-gthelen@google.com \
    --to=gthelen@google.com \
    --cc=akpm@linux-foundation.org \
    --cc=cgroups@vger.kernel.org \
    --cc=guro@fb.com \
    --cc=hannes@cmpxchg.org \
    --cc=kernel-team@fb.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mhocko@kernel.org \
    --cc=tj@kernel.org \
    --cc=vdavydov.dev@gmail.com \
    --subject='Re: [RFC PATCH 2/2] memcg: add memory.min' \
    /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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).