[PATCH cgroup/for-4.17 1/2] cgroup: Explicitly remove core interface files

[PATCH cgroup/for-4.17 1/2] cgroup: Explicitly remove core interface files

[Tejun Heo]

The "cgroup." core interface files bypass the usual interface removal
path and get removed recursively along with the cgroup itself.  While
this works now, the subtle discrepancy gets in the way of implementing
common mechanisms.

This patch updates cgroup core interface file handling so that it's
consistent with controller interface files.  When added, the css is
marked CSS_VISIBLE and they're explicitly removed before the cgroup is
destroyed.

This doesn't cause user-visible behavior changes.

Signed-off-by: Tejun Heo <tj@kernel.org>
---
 kernel/cgroup/cgroup.c |   35 ++++++++++++++++++++++-------------
 1 file changed, 22 insertions(+), 13 deletions(-)

--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -1573,8 +1573,17 @@ static void css_clear_dir(struct cgroup_
 
 	css->flags &= ~CSS_VISIBLE;
 
-	list_for_each_entry(cfts, &css->ss->cfts, node)
+	if (!css->ss) {
+		if (cgroup_on_dfl(cgrp))
+			cfts = cgroup_base_files;
+		else
+			cfts = cgroup1_base_files;
+
 		cgroup_addrm_files(css, cgrp, cfts, false);
+	} else {
+		list_for_each_entry(cfts, &css->ss->cfts, node)
+			cgroup_addrm_files(css, cgrp, cfts, false);
+	}
 }
 
 /**
@@ -1598,14 +1607,16 @@ static int css_populate_dir(struct cgrou
 		else
 			cfts = cgroup1_base_files;
 
-		return cgroup_addrm_files(&cgrp->self, cgrp, cfts, true);
-	}
-
-	list_for_each_entry(cfts, &css->ss->cfts, node) {
-		ret = cgroup_addrm_files(css, cgrp, cfts, true);
-		if (ret < 0) {
-			failed_cfts = cfts;
-			goto err;
+		ret = cgroup_addrm_files(&cgrp->self, cgrp, cfts, true);
+		if (ret < 0)
+			return ret;
+	} else {
+		list_for_each_entry(cfts, &css->ss->cfts, node) {
+			ret = cgroup_addrm_files(css, cgrp, cfts, true);
+			if (ret < 0) {
+				failed_cfts = cfts;
+				goto err;
+			}
 		}
 	}
 
@@ -5097,10 +5108,8 @@ static int cgroup_destroy_locked(struct
 	for_each_css(css, ssid, cgrp)
 		kill_css(css);
 
-	/*
-	 * Remove @cgrp directory along with the base files.  @cgrp has an
-	 * extra ref on its kn.
-	 */
+	/* clear and remove @cgrp dir, @cgrp has an extra ref on its kn */
+	css_clear_dir(&cgrp->self);
 	kernfs_remove(cgrp->kn);
 
 	if (parent && cgroup_is_threaded(cgrp))

[PATCH cgroup/for-4.17 2/2] Subject: cgroup: Limit event generation frequency

[Tejun Heo]

".events" files generate file modified event to notify userland of
possible new events.  Some of the events can be quite bursty
(e.g. memory high event) and generating notification each time is
costly and pointless.

This patch implements a event rate limit mechanism.  If a new
notification is requested before 10ms has passed since the previous
notification, the new notification is delayed till then.

As this only delays from the second notification on in a given close
cluster of notifications, userland reactions to notifications
shouldn't be delayed at all in most cases while avoiding notification
storms.

Signed-off-by: Tejun Heo <tj@kernel.org>
---
 include/linux/cgroup-defs.h |    2 ++
 kernel/cgroup/cgroup.c      |   25 +++++++++++++++++++++++--
 2 files changed, 25 insertions(+), 2 deletions(-)

--- a/include/linux/cgroup-defs.h
+++ b/include/linux/cgroup-defs.h
@@ -105,6 +105,8 @@ enum {
 struct cgroup_file {
 	/* do not access any fields from outside cgroup core */
 	struct kernfs_node *kn;
+	unsigned long notified_at;
+	struct timer_list notify_timer;
 };
 
 /*
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -61,6 +61,8 @@
 
 #define CGROUP_FILE_NAME_MAX		(MAX_CGROUP_TYPE_NAMELEN +	\
 					 MAX_CFTYPE_NAME + 2)
+/* let's not notify more than 100 times per second */
+#define CGROUP_FILE_NOTIFY_MIN_INTV	DIV_ROUND_UP(HZ, 100)
 
 /*
  * cgroup_mutex is the master lock.  Any modification to cgroup or its
@@ -1554,6 +1556,8 @@ static void cgroup_rm_file(struct cgroup
 		spin_lock_irq(&cgroup_file_kn_lock);
 		cfile->kn = NULL;
 		spin_unlock_irq(&cgroup_file_kn_lock);
+
+		del_timer_sync(&cfile->notify_timer);
 	}
 
 	kernfs_remove_by_name(cgrp->kn, cgroup_file_name(cgrp, cft, name));
@@ -3532,6 +3536,12 @@ static int cgroup_kn_set_ugid(struct ker
 	return kernfs_setattr(kn, &iattr);
 }
 
+static void cgroup_file_notify_timer(struct timer_list *timer)
+{
+	cgroup_file_notify(container_of(timer, struct cgroup_file,
+					notify_timer));
+}
+
 static int cgroup_add_file(struct cgroup_subsys_state *css, struct cgroup *cgrp,
 			   struct cftype *cft)
 {
@@ -3558,6 +3568,8 @@ static int cgroup_add_file(struct cgroup
 	if (cft->file_offset) {
 		struct cgroup_file *cfile = (void *)css + cft->file_offset;
 
+		timer_setup(&cfile->notify_timer, cgroup_file_notify_timer, 0);
+
 		spin_lock_irq(&cgroup_file_kn_lock);
 		cfile->kn = kn;
 		spin_unlock_irq(&cgroup_file_kn_lock);
@@ -3807,8 +3819,17 @@ void cgroup_file_notify(struct cgroup_fi
 	unsigned long flags;
 
 	spin_lock_irqsave(&cgroup_file_kn_lock, flags);
-	if (cfile->kn)
-		kernfs_notify(cfile->kn);
+	if (cfile->kn) {
+		unsigned long last = cfile->notified_at;
+		unsigned long next = last + CGROUP_FILE_NOTIFY_MIN_INTV;
+
+		if (time_in_range(jiffies, last, next)) {
+			timer_reduce(&cfile->notify_timer, next);
+		} else {
+			kernfs_notify(cfile->kn);
+			cfile->notified_at = jiffies;
+		}
+	}
 	spin_unlock_irqrestore(&cgroup_file_kn_lock, flags);
 }
 

Re: [PATCH cgroup/for-4.17 2/2] Subject: cgroup: Limit event generation frequency

[Tejun Heo]

On Fri, Mar 23, 2018 at 12:47:53PM -0700, Tejun Heo wrote:
> ".events" files generate file modified event to notify userland of
> possible new events.  Some of the events can be quite bursty
> (e.g. memory high event) and generating notification each time is
> costly and pointless.
> 
> This patch implements a event rate limit mechanism.  If a new
> notification is requested before 10ms has passed since the previous
> notification, the new notification is delayed till then.
> 
> As this only delays from the second notification on in a given close
> cluster of notifications, userland reactions to notifications
> shouldn't be delayed at all in most cases while avoiding notification
> storms.
> 
> Signed-off-by: Tejun Heo <tj@kernel.org>

Applied 1-2 to cgroup/for-4.18.

Thanks.

-- 
tejun