All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dennis Zhou <dennis@kernel.org>
To: Jens Axboe <axboe@kernel.dk>, Tejun Heo <tj@kernel.org>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Josef Bacik <josef@toxicpanda.com>
Cc: kernel-team@fb.com, linux-block@vger.kernel.org,
	cgroups@vger.kernel.org, linux-kernel@vger.kernel.org,
	Dennis Zhou <dennis@kernel.org>
Subject: [PATCH 12/13] blkcg: change blkg reference counting to use percpu_ref
Date: Mon, 26 Nov 2018 16:19:45 -0500	[thread overview]
Message-ID: <20181126211946.77067-13-dennis@kernel.org> (raw)
In-Reply-To: <20181126211946.77067-1-dennis@kernel.org>

Every bio is now associated with a blkg putting blkg_get, blkg_try_get,
and blkg_put on the hot path. Switch over the refcnt in blkg to use
percpu_ref.

Signed-off-by: Dennis Zhou <dennis@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
---
 block/blk-cgroup.c         | 41 ++++++++++++++++++++++++++++++++++++--
 include/linux/blk-cgroup.h | 15 +++++---------
 2 files changed, 44 insertions(+), 12 deletions(-)

diff --git a/block/blk-cgroup.c b/block/blk-cgroup.c
index dfd984bbed27..64ce424a78fd 100644
--- a/block/blk-cgroup.c
+++ b/block/blk-cgroup.c
@@ -81,6 +81,37 @@ static void blkg_free(struct blkcg_gq *blkg)
 	kfree(blkg);
 }
 
+static void __blkg_release(struct rcu_head *rcu)
+{
+	struct blkcg_gq *blkg = container_of(rcu, struct blkcg_gq, rcu_head);
+
+	percpu_ref_exit(&blkg->refcnt);
+
+	/* release the blkcg and parent blkg refs this blkg has been holding */
+	css_put(&blkg->blkcg->css);
+	if (blkg->parent)
+		blkg_put(blkg->parent);
+
+	wb_congested_put(blkg->wb_congested);
+
+	blkg_free(blkg);
+}
+
+/*
+ * A group is RCU protected, but having an rcu lock does not mean that one
+ * can access all the fields of blkg and assume these are valid.  For
+ * example, don't try to follow throtl_data and request queue links.
+ *
+ * Having a reference to blkg under an rcu allows accesses to only values
+ * local to groups like group stats and group rate limits.
+ */
+static void blkg_release(struct percpu_ref *ref)
+{
+	struct blkcg_gq *blkg = container_of(ref, struct blkcg_gq, refcnt);
+
+	call_rcu(&blkg->rcu_head, __blkg_release);
+}
+
 /**
  * blkg_alloc - allocate a blkg
  * @blkcg: block cgroup the new blkg is associated with
@@ -107,7 +138,6 @@ static struct blkcg_gq *blkg_alloc(struct blkcg *blkcg, struct request_queue *q,
 	blkg->q = q;
 	INIT_LIST_HEAD(&blkg->q_node);
 	blkg->blkcg = blkcg;
-	atomic_set(&blkg->refcnt, 1);
 
 	for (i = 0; i < BLKCG_MAX_POLS; i++) {
 		struct blkcg_policy *pol = blkcg_policy[i];
@@ -207,6 +237,11 @@ static struct blkcg_gq *blkg_create(struct blkcg *blkcg,
 		blkg_get(blkg->parent);
 	}
 
+	ret = percpu_ref_init(&blkg->refcnt, blkg_release, 0,
+			      GFP_NOWAIT | __GFP_NOWARN);
+	if (ret)
+		goto err_cancel_ref;
+
 	/* invoke per-policy init */
 	for (i = 0; i < BLKCG_MAX_POLS; i++) {
 		struct blkcg_policy *pol = blkcg_policy[i];
@@ -239,6 +274,8 @@ static struct blkcg_gq *blkg_create(struct blkcg *blkcg,
 	blkg_put(blkg);
 	return ERR_PTR(ret);
 
+err_cancel_ref:
+	percpu_ref_exit(&blkg->refcnt);
 err_put_congested:
 	wb_congested_put(wb_congested);
 err_put_css:
@@ -369,7 +406,7 @@ static void blkg_destroy(struct blkcg_gq *blkg)
 	 * Put the reference taken at the time of creation so that when all
 	 * queues are gone, group can be destroyed.
 	 */
-	blkg_put(blkg);
+	percpu_ref_kill(&blkg->refcnt);
 }
 
 /**
diff --git a/include/linux/blk-cgroup.h b/include/linux/blk-cgroup.h
index 6f3bb5e82a57..fc23bb758566 100644
--- a/include/linux/blk-cgroup.h
+++ b/include/linux/blk-cgroup.h
@@ -123,7 +123,7 @@ struct blkcg_gq {
 	struct blkcg_gq			*parent;
 
 	/* reference count */
-	atomic_t			refcnt;
+	struct percpu_ref		refcnt;
 
 	/* is this blkg online? protected by both blkcg and q locks */
 	bool				online;
@@ -486,8 +486,7 @@ static inline int blkg_path(struct blkcg_gq *blkg, char *buf, int buflen)
  */
 static inline void blkg_get(struct blkcg_gq *blkg)
 {
-	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
-	atomic_inc(&blkg->refcnt);
+	percpu_ref_get(&blkg->refcnt);
 }
 
 /**
@@ -499,7 +498,7 @@ static inline void blkg_get(struct blkcg_gq *blkg)
  */
 static inline struct blkcg_gq *blkg_try_get(struct blkcg_gq *blkg)
 {
-	if (atomic_inc_not_zero(&blkg->refcnt))
+	if (percpu_ref_tryget(&blkg->refcnt))
 		return blkg;
 	return NULL;
 }
@@ -513,23 +512,19 @@ static inline struct blkcg_gq *blkg_try_get(struct blkcg_gq *blkg)
  */
 static inline struct blkcg_gq *blkg_try_get_closest(struct blkcg_gq *blkg)
 {
-	while (!atomic_inc_not_zero(&blkg->refcnt))
+	while (!percpu_ref_tryget(&blkg->refcnt))
 		blkg = blkg->parent;
 
 	return blkg;
 }
 
-void __blkg_release_rcu(struct rcu_head *rcu);
-
 /**
  * blkg_put - put a blkg reference
  * @blkg: blkg to put
  */
 static inline void blkg_put(struct blkcg_gq *blkg)
 {
-	WARN_ON_ONCE(atomic_read(&blkg->refcnt) <= 0);
-	if (atomic_dec_and_test(&blkg->refcnt))
-		call_rcu(&blkg->rcu_head, __blkg_release_rcu);
+	percpu_ref_put(&blkg->refcnt);
 }
 
 /**
-- 
2.17.1


  parent reply	other threads:[~2018-11-26 21:19 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-26 21:19 [PATCH 00/13 v4] block: always associate blkg and refcount cleanup Dennis Zhou
2018-11-26 21:19 ` [PATCH 01/13] blkcg: fix ref count issue with bio_blkcg() using task_css Dennis Zhou
2018-11-27 20:54   ` Josef Bacik
2018-11-26 21:19 ` [PATCH 02/13] blkcg: update blkg_lookup_create() to do locking Dennis Zhou
2018-11-27 20:56   ` Josef Bacik
2018-11-26 21:19 ` [PATCH 03/13] blkcg: convert blkg_lookup_create() to find closest blkg Dennis Zhou
2018-11-27 21:01   ` Josef Bacik
2018-11-26 21:19 ` [PATCH 04/13] blkcg: introduce common blkg association logic Dennis Zhou
2018-11-27 21:04   ` Josef Bacik
2018-11-29 15:49   ` Tejun Heo
2018-11-29 19:48     ` Dennis Zhou
2018-11-30  9:52   ` Christoph Hellwig
2018-12-03 20:58     ` Dennis Zhou
2018-11-26 21:19 ` [PATCH 05/13] blkcg: associate blkg when associating a device Dennis Zhou
2018-11-29 15:53   ` Tejun Heo
2018-11-29 19:54     ` Dennis Zhou
2018-11-30  9:54   ` Christoph Hellwig
2018-12-03 22:52     ` Dennis Zhou
2018-11-26 21:19 ` [PATCH 06/13] blkcg: consolidate bio_issue_init() to be a part of core Dennis Zhou
2018-11-26 21:19 ` [PATCH 07/13] blkcg: associate a blkg for pages being evicted by swap Dennis Zhou
2018-11-26 21:19 ` [PATCH 08/13] blkcg: associate writeback bios with a blkg Dennis Zhou
2018-11-26 21:19 ` [PATCH 09/13] blkcg: remove bio->bi_css and instead use bio->bi_blkg Dennis Zhou
2018-11-26 21:19 ` [PATCH 10/13] blkcg: remove additional reference to the css Dennis Zhou
2018-11-26 21:19 ` [PATCH 11/13] blkcg: remove bio_disassociate_task() Dennis Zhou
2018-11-29 15:54   ` Tejun Heo
2018-11-26 21:19 ` Dennis Zhou [this message]
2018-11-26 21:19 ` [PATCH 13/13] blkcg: rename blkg_try_get() to blkg_tryget() Dennis Zhou
2018-11-27 21:10 ` [PATCH 00/13 v4] block: always associate blkg and refcount cleanup Josef Bacik
2018-11-27 22:15   ` Dennis Zhou

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=20181126211946.77067-13-dennis@kernel.org \
    --to=dennis@kernel.org \
    --cc=axboe@kernel.dk \
    --cc=cgroups@vger.kernel.org \
    --cc=hannes@cmpxchg.org \
    --cc=josef@toxicpanda.com \
    --cc=kernel-team@fb.com \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=tj@kernel.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.