Commit 971ff493 authored by Zefan Li's avatar Zefan Li Committed by Tejun Heo
Browse files

cgroup: use a per-cgroup work for release agent 



Instead of using a global work to schedule release agent on removable
cgroups, we change to use a per-cgroup work to do this, which makes
the code much simpler.

v2: use a dedicated work instead of reusing css->destroy_work. (Tejun)

Signed-off-by: default avatarZefan Li <lizefan@huawei.com>
Signed-off-by: default avatarTejun Heo <tj@kernel.org>
parent 0c8fc2c1

include/linux/cgroup.h

+3 −7
Original line number Diff line number Diff line
@@ -233,13 +233,6 @@ struct cgroup { 
	 */

	struct list_head e_csets[CGROUP_SUBSYS_COUNT];



	/*

	 * Linked list running through all cgroups that can

	 * potentially be reaped by the release agent. Protected by

	 * release_list_lock

	 */

	struct list_head release_list;



	/*

	 * list of pidlists, up to two for each namespace (one for procs, one

	 * for tasks); created on demand.
@@ -249,6 +242,9 @@ struct cgroup { 


	/* used to wait for offlining of csses */

	wait_queue_head_t offline_waitq;



	/* used to schedule release agent */

	struct work_struct release_agent_work;

};



#define MAX_CGROUP_ROOT_NAMELEN 64

kernel/cgroup.c

+33 −75
Original line number Diff line number Diff line
@@ -392,12 +392,7 @@ static int notify_on_release(const struct cgroup *cgrp) 
			;						\

		else



/* the list of cgroups eligible for automatic release. Protected by

 * release_list_lock */

static LIST_HEAD(release_list);

static DEFINE_RAW_SPINLOCK(release_list_lock);

static void cgroup_release_agent(struct work_struct *work);

static DECLARE_WORK(release_agent_work, cgroup_release_agent);

static void check_for_release(struct cgroup *cgrp);



/*
@@ -1577,7 +1572,6 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp) 
	INIT_LIST_HEAD(&cgrp->self.sibling);

	INIT_LIST_HEAD(&cgrp->self.children);

	INIT_LIST_HEAD(&cgrp->cset_links);

	INIT_LIST_HEAD(&cgrp->release_list);

	INIT_LIST_HEAD(&cgrp->pidlists);

	mutex_init(&cgrp->pidlist_mutex);

	cgrp->self.cgroup = cgrp;
@@ -1587,6 +1581,7 @@ static void init_cgroup_housekeeping(struct cgroup *cgrp) 
		INIT_LIST_HEAD(&cgrp->e_csets[ssid]);



	init_waitqueue_head(&cgrp->offline_waitq);

	INIT_WORK(&cgrp->release_agent_work, cgroup_release_agent);

}



static void init_cgroup_root(struct cgroup_root *root,
@@ -4342,6 +4337,7 @@ static void css_free_work_fn(struct work_struct *work) 
		/* cgroup free path */

		atomic_dec(&cgrp->root->nr_cgrps);

		cgroup_pidlist_destroy_all(cgrp);

		cancel_work_sync(&cgrp->release_agent_work);



		if (cgroup_parent(cgrp)) {

			/*
@@ -4804,12 +4800,6 @@ static int cgroup_destroy_locked(struct cgroup *cgrp) 
	for_each_css(css, ssid, cgrp)

		kill_css(css);



	/* CSS_ONLINE is clear, remove from ->release_list for the last time */

	raw_spin_lock(&release_list_lock);

	if (!list_empty(&cgrp->release_list))

		list_del_init(&cgrp->release_list);

	raw_spin_unlock(&release_list_lock);



	/*

	 * Remove @cgrp directory along with the base files.  @cgrp has an

	 * extra ref on its kn.
@@ -5271,25 +5261,9 @@ void cgroup_exit(struct task_struct *tsk) 


static void check_for_release(struct cgroup *cgrp)

{

	if (cgroup_is_releasable(cgrp) && list_empty(&cgrp->cset_links) &&

	    !css_has_online_children(&cgrp->self)) {

		/*

		 * Control Group is currently removeable. If it's not

		 * already queued for a userspace notification, queue

		 * it now

		 */

		int need_schedule_work = 0;



		raw_spin_lock(&release_list_lock);

		if (!cgroup_is_dead(cgrp) &&

		    list_empty(&cgrp->release_list)) {

			list_add(&cgrp->release_list, &release_list);

			need_schedule_work = 1;

		}

		raw_spin_unlock(&release_list_lock);

		if (need_schedule_work)

			schedule_work(&release_agent_work);

	}

	if (cgroup_is_releasable(cgrp) && !cgroup_has_tasks(cgrp) &&

	    !css_has_online_children(&cgrp->self) && !cgroup_is_dead(cgrp))

		schedule_work(&cgrp->release_agent_work);

}



/*
@@ -5317,52 +5291,36 @@ static void check_for_release(struct cgroup *cgrp) 
 */

static void cgroup_release_agent(struct work_struct *work)

{

	BUG_ON(work != &release_agent_work);

	mutex_lock(&cgroup_mutex);

	raw_spin_lock(&release_list_lock);

	while (!list_empty(&release_list)) {

		char *argv[3], *envp[3];

		int i;

	struct cgroup *cgrp =

		container_of(work, struct cgroup, release_agent_work);

	char *pathbuf = NULL, *agentbuf = NULL, *path;

		struct cgroup *cgrp = list_entry(release_list.next,

						    struct cgroup,

						    release_list);

		list_del_init(&cgrp->release_list);

		raw_spin_unlock(&release_list_lock);

	char *argv[3], *envp[3];



	mutex_lock(&cgroup_mutex);



	pathbuf = kmalloc(PATH_MAX, GFP_KERNEL);

		if (!pathbuf)

			goto continue_free;

	agentbuf = kstrdup(cgrp->root->release_agent_path, GFP_KERNEL);

	if (!pathbuf || !agentbuf)

		goto out;



	path = cgroup_path(cgrp, pathbuf, PATH_MAX);

	if (!path)

			goto continue_free;

		agentbuf = kstrdup(cgrp->root->release_agent_path, GFP_KERNEL);

		if (!agentbuf)

			goto continue_free;

		goto out;



		i = 0;

		argv[i++] = agentbuf;

		argv[i++] = path;

		argv[i] = NULL;

	argv[0] = agentbuf;

	argv[1] = path;

	argv[2] = NULL;



		i = 0;

	/* minimal command environment */

		envp[i++] = "HOME=/";

		envp[i++] = "PATH=/sbin:/bin:/usr/sbin:/usr/bin";

		envp[i] = NULL;

	envp[0] = "HOME=/";

	envp[1] = "PATH=/sbin:/bin:/usr/sbin:/usr/bin";

	envp[2] = NULL;



		/* Drop the lock while we invoke the usermode helper,

		 * since the exec could involve hitting disk and hence

		 * be a slow process */

	mutex_unlock(&cgroup_mutex);

	call_usermodehelper(argv[0], argv, envp, UMH_WAIT_EXEC);

		mutex_lock(&cgroup_mutex);

 continue_free:

		kfree(pathbuf);

out:

	kfree(agentbuf);

		raw_spin_lock(&release_list_lock);

	}

	raw_spin_unlock(&release_list_lock);

	mutex_unlock(&cgroup_mutex);

	kfree(pathbuf);

}



static int __init cgroup_disable(char *str)