[mirror_ubuntu-zesty-kernel.git] / kernel / taskstats.c

/*
 * taskstats.c - Export per-task statistics to userland
 *
 * Copyright (C) Shailabh Nagar, IBM Corp. 2006
 *           (C) Balbir Singh,   IBM Corp. 2006
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 */

#include <linux/kernel.h>
#include <linux/taskstats_kern.h>
#include <linux/tsacct_kern.h>
#include <linux/delayacct.h>
#include <linux/cpumask.h>
#include <linux/percpu.h>
#include <linux/slab.h>
#include <linux/cgroupstats.h>
#include <linux/cgroup.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/pid_namespace.h>
#include <net/genetlink.h>
#include <linux/atomic.h>

/*
 * Maximum length of a cpumask that can be specified in
 * the TASKSTATS_CMD_ATTR_REGISTER/DEREGISTER_CPUMASK attribute
 */
#define TASKSTATS_CPUMASK_MAXLEN	(100+6*NR_CPUS)

static DEFINE_PER_CPU(__u32, taskstats_seqnum);
static int family_registered;
struct kmem_cache *taskstats_cache;

static struct genl_family family;

static const struct nla_policy taskstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1] = {
	[TASKSTATS_CMD_ATTR_PID]  = { .type = NLA_U32 },
	[TASKSTATS_CMD_ATTR_TGID] = { .type = NLA_U32 },
	[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK] = { .type = NLA_STRING },
	[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },};

/*
 * We have to use TASKSTATS_CMD_ATTR_MAX here, it is the maxattr in the family.
 * Make sure they are always aligned.
 */
static const struct nla_policy cgroupstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1] = {
	[CGROUPSTATS_CMD_ATTR_FD] = { .type = NLA_U32 },
};

struct listener {
	struct list_head list;
	pid_t pid;
	char valid;
};

struct listener_list {
	struct rw_semaphore sem;
	struct list_head list;
};
static DEFINE_PER_CPU(struct listener_list, listener_array);

enum actions {
	REGISTER,
	DEREGISTER,
	CPU_DONT_CARE
};

static int prepare_reply(struct genl_info *info, u8 cmd, struct sk_buff **skbp,
				size_t size)
{
	struct sk_buff *skb;
	void *reply;

	/*
	 * If new attributes are added, please revisit this allocation
	 */
	skb = genlmsg_new(size, GFP_KERNEL);
	if (!skb)
		return -ENOMEM;

	if (!info) {
		int seq = this_cpu_inc_return(taskstats_seqnum) - 1;

		reply = genlmsg_put(skb, 0, seq, &family, 0, cmd);
	} else
		reply = genlmsg_put_reply(skb, info, &family, 0, cmd);
	if (reply == NULL) {
		nlmsg_free(skb);
		return -EINVAL;
	}

	*skbp = skb;
	return 0;
}

/*
 * Send taskstats data in @skb to listener with nl_pid @pid
 */
static int send_reply(struct sk_buff *skb, struct genl_info *info)
{
	struct genlmsghdr *genlhdr = nlmsg_data(nlmsg_hdr(skb));
	void *reply = genlmsg_data(genlhdr);

	genlmsg_end(skb, reply);

	return genlmsg_reply(skb, info);
}

/*
 * Send taskstats data in @skb to listeners registered for @cpu's exit data
 */
static void send_cpu_listeners(struct sk_buff *skb,
					struct listener_list *listeners)
{
	struct genlmsghdr *genlhdr = nlmsg_data(nlmsg_hdr(skb));
	struct listener *s, *tmp;
	struct sk_buff *skb_next, *skb_cur = skb;
	void *reply = genlmsg_data(genlhdr);
	int rc, delcount = 0;

	genlmsg_end(skb, reply);

	rc = 0;
	down_read(&listeners->sem);
	list_for_each_entry(s, &listeners->list, list) {
		skb_next = NULL;
		if (!list_is_last(&s->list, &listeners->list)) {
			skb_next = skb_clone(skb_cur, GFP_KERNEL);
			if (!skb_next)
				break;
		}
		rc = genlmsg_unicast(&init_net, skb_cur, s->pid);
		if (rc == -ECONNREFUSED) {
			s->valid = 0;
			delcount++;
		}
		skb_cur = skb_next;
	}
	up_read(&listeners->sem);

	if (skb_cur)
		nlmsg_free(skb_cur);

	if (!delcount)
		return;

	/* Delete invalidated entries */
	down_write(&listeners->sem);
	list_for_each_entry_safe(s, tmp, &listeners->list, list) {
		if (!s->valid) {
			list_del(&s->list);
			kfree(s);
		}
	}
	up_write(&listeners->sem);
}

static void fill_stats(struct user_namespace *user_ns,
		       struct pid_namespace *pid_ns,
		       struct task_struct *tsk, struct taskstats *stats)
{
	memset(stats, 0, sizeof(*stats));
	/*
	 * Each accounting subsystem adds calls to its functions to
	 * fill in relevant parts of struct taskstsats as follows
	 *
	 *	per-task-foo(stats, tsk);
	 */

	delayacct_add_tsk(stats, tsk);

	/* fill in basic acct fields */
	stats->version = TASKSTATS_VERSION;
	stats->nvcsw = tsk->nvcsw;
	stats->nivcsw = tsk->nivcsw;
	bacct_add_tsk(user_ns, pid_ns, stats, tsk);

	/* fill in extended acct fields */
	xacct_add_tsk(stats, tsk);
}

static int fill_stats_for_pid(pid_t pid, struct taskstats *stats)
{
	struct task_struct *tsk;

	rcu_read_lock();
	tsk = find_task_by_vpid(pid);
	if (tsk)
		get_task_struct(tsk);
	rcu_read_unlock();
	if (!tsk)
		return -ESRCH;
	fill_stats(current_user_ns(), task_active_pid_ns(current), tsk, stats);
	put_task_struct(tsk);
	return 0;
}

static int fill_stats_for_tgid(pid_t tgid, struct taskstats *stats)
{
	struct task_struct *tsk, *first;
	unsigned long flags;
	int rc = -ESRCH;

	/*
	 * Add additional stats from live tasks except zombie thread group
	 * leaders who are already counted with the dead tasks
	 */
	rcu_read_lock();
	first = find_task_by_vpid(tgid);

	if (!first || !lock_task_sighand(first, &flags))
		goto out;

	if (first->signal->stats)
		memcpy(stats, first->signal->stats, sizeof(*stats));
	else
		memset(stats, 0, sizeof(*stats));

	tsk = first;
	do {
		if (tsk->exit_state)
			continue;
		/*
		 * Accounting subsystem can call its functions here to
		 * fill in relevant parts of struct taskstsats as follows
		 *
		 *	per-task-foo(stats, tsk);
		 */
		delayacct_add_tsk(stats, tsk);

		stats->nvcsw += tsk->nvcsw;
		stats->nivcsw += tsk->nivcsw;
	} while_each_thread(first, tsk);

	unlock_task_sighand(first, &flags);
	rc = 0;
out:
	rcu_read_unlock();

	stats->version = TASKSTATS_VERSION;
	/*
	 * Accounting subsystems can also add calls here to modify
	 * fields of taskstats.
	 */
	return rc;
}

static void fill_tgid_exit(struct task_struct *tsk)
{
	unsigned long flags;

	spin_lock_irqsave(&tsk->sighand->siglock, flags);
	if (!tsk->signal->stats)
		goto ret;

	/*
	 * Each accounting subsystem calls its functions here to
	 * accumalate its per-task stats for tsk, into the per-tgid structure
	 *
	 *	per-task-foo(tsk->signal->stats, tsk);
	 */
	delayacct_add_tsk(tsk->signal->stats, tsk);
ret:
	spin_unlock_irqrestore(&tsk->sighand->siglock, flags);
	return;
}

static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
{
	struct listener_list *listeners;
	struct listener *s, *tmp, *s2;
	unsigned int cpu;
	int ret = 0;

	if (!cpumask_subset(mask, cpu_possible_mask))
		return -EINVAL;

	if (current_user_ns() != &init_user_ns)
		return -EINVAL;

	if (task_active_pid_ns(current) != &init_pid_ns)
		return -EINVAL;

	if (isadd == REGISTER) {
		for_each_cpu(cpu, mask) {
			s = kmalloc_node(sizeof(struct listener),
					GFP_KERNEL, cpu_to_node(cpu));
			if (!s) {
				ret = -ENOMEM;
				goto cleanup;
			}
			s->pid = pid;
			s->valid = 1;

			listeners = &per_cpu(listener_array, cpu);
			down_write(&listeners->sem);
			list_for_each_entry(s2, &listeners->list, list) {
				if (s2->pid == pid && s2->valid)
					goto exists;
			}
			list_add(&s->list, &listeners->list);
			s = NULL;
exists:
			up_write(&listeners->sem);
			kfree(s); /* nop if NULL */
		}
		return 0;
	}

	/* Deregister or cleanup */
cleanup:
	for_each_cpu(cpu, mask) {
		listeners = &per_cpu(listener_array, cpu);
		down_write(&listeners->sem);
		list_for_each_entry_safe(s, tmp, &listeners->list, list) {
			if (s->pid == pid) {
				list_del(&s->list);
				kfree(s);
				break;
			}
		}
		up_write(&listeners->sem);
	}
	return ret;
}

static int parse(struct nlattr *na, struct cpumask *mask)
{
	char *data;
	int len;
	int ret;

	if (na == NULL)
		return 1;
	len = nla_len(na);
	if (len > TASKSTATS_CPUMASK_MAXLEN)
		return -E2BIG;
	if (len < 1)
		return -EINVAL;
	data = kmalloc(len, GFP_KERNEL);
	if (!data)
		return -ENOMEM;
	nla_strlcpy(data, na, len);
	ret = cpulist_parse(data, mask);
	kfree(data);
	return ret;
}

static struct taskstats *mk_reply(struct sk_buff *skb, int type, u32 pid)
{
	struct nlattr *na, *ret;
	int aggr;

	aggr = (type == TASKSTATS_TYPE_PID)
			? TASKSTATS_TYPE_AGGR_PID
			: TASKSTATS_TYPE_AGGR_TGID;

	na = nla_nest_start(skb, aggr);
	if (!na)
		goto err;

	if (nla_put(skb, type, sizeof(pid), &pid) < 0) {
		nla_nest_cancel(skb, na);
		goto err;
	}
	ret = nla_reserve_64bit(skb, TASKSTATS_TYPE_STATS,
				sizeof(struct taskstats), TASKSTATS_TYPE_NULL);
	if (!ret) {
		nla_nest_cancel(skb, na);
		goto err;
	}
	nla_nest_end(skb, na);

	return nla_data(ret);
err:
	return NULL;
}

static int cgroupstats_user_cmd(struct sk_buff *skb, struct genl_info *info)
{
	int rc = 0;
	struct sk_buff *rep_skb;
	struct cgroupstats *stats;
	struct nlattr *na;
	size_t size;
	u32 fd;
	struct fd f;

	na = info->attrs[CGROUPSTATS_CMD_ATTR_FD];
	if (!na)
		return -EINVAL;

	fd = nla_get_u32(info->attrs[CGROUPSTATS_CMD_ATTR_FD]);
	f = fdget(fd);
	if (!f.file)
		return 0;

	size = nla_total_size(sizeof(struct cgroupstats));

	rc = prepare_reply(info, CGROUPSTATS_CMD_NEW, &rep_skb,
				size);
	if (rc < 0)
		goto err;

	na = nla_reserve(rep_skb, CGROUPSTATS_TYPE_CGROUP_STATS,
				sizeof(struct cgroupstats));
	if (na == NULL) {
		nlmsg_free(rep_skb);
		rc = -EMSGSIZE;
		goto err;
	}

	stats = nla_data(na);
	memset(stats, 0, sizeof(*stats));

	rc = cgroupstats_build(stats, f.file->f_path.dentry);
	if (rc < 0) {
		nlmsg_free(rep_skb);
		goto err;
	}

	rc = send_reply(rep_skb, info);

err:
	fdput(f);
	return rc;
}

static int cmd_attr_register_cpumask(struct genl_info *info)
{
	cpumask_var_t mask;
	int rc;

	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
		return -ENOMEM;
	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK], mask);
	if (rc < 0)
		goto out;
	rc = add_del_listener(info->snd_portid, mask, REGISTER);
out:
	free_cpumask_var(mask);
	return rc;
}

static int cmd_attr_deregister_cpumask(struct genl_info *info)
{
	cpumask_var_t mask;
	int rc;

	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
		return -ENOMEM;
	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK], mask);
	if (rc < 0)
		goto out;
	rc = add_del_listener(info->snd_portid, mask, DEREGISTER);
out:
	free_cpumask_var(mask);
	return rc;
}

static size_t taskstats_packet_size(void)
{
	size_t size;

	size = nla_total_size(sizeof(u32)) +
		nla_total_size_64bit(sizeof(struct taskstats)) +
		nla_total_size(0);

	return size;
}

static int cmd_attr_pid(struct genl_info *info)
{
	struct taskstats *stats;
	struct sk_buff *rep_skb;
	size_t size;
	u32 pid;
	int rc;

	size = taskstats_packet_size();

	rc = prepare_reply(info, TASKSTATS_CMD_NEW, &rep_skb, size);
	if (rc < 0)
		return rc;

	rc = -EINVAL;
	pid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_PID]);
	stats = mk_reply(rep_skb, TASKSTATS_TYPE_PID, pid);
	if (!stats)
		goto err;

	rc = fill_stats_for_pid(pid, stats);
	if (rc < 0)
		goto err;
	return send_reply(rep_skb, info);
err:
	nlmsg_free(rep_skb);
	return rc;
}

static int cmd_attr_tgid(struct genl_info *info)
{
	struct taskstats *stats;
	struct sk_buff *rep_skb;
	size_t size;
	u32 tgid;
	int rc;

	size = taskstats_packet_size();

	rc = prepare_reply(info, TASKSTATS_CMD_NEW, &rep_skb, size);
	if (rc < 0)
		return rc;

	rc = -EINVAL;
	tgid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_TGID]);
	stats = mk_reply(rep_skb, TASKSTATS_TYPE_TGID, tgid);
	if (!stats)
		goto err;

	rc = fill_stats_for_tgid(tgid, stats);
	if (rc < 0)
		goto err;
	return send_reply(rep_skb, info);
err:
	nlmsg_free(rep_skb);
	return rc;
}

static int taskstats_user_cmd(struct sk_buff *skb, struct genl_info *info)
{
	if (info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK])
		return cmd_attr_register_cpumask(info);
	else if (info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK])
		return cmd_attr_deregister_cpumask(info);
	else if (info->attrs[TASKSTATS_CMD_ATTR_PID])
		return cmd_attr_pid(info);
	else if (info->attrs[TASKSTATS_CMD_ATTR_TGID])
		return cmd_attr_tgid(info);
	else
		return -EINVAL;
}

static struct taskstats *taskstats_tgid_alloc(struct task_struct *tsk)
{
	struct signal_struct *sig = tsk->signal;
	struct taskstats *stats;

	if (sig->stats || thread_group_empty(tsk))
		goto ret;

	/* No problem if kmem_cache_zalloc() fails */
	stats = kmem_cache_zalloc(taskstats_cache, GFP_KERNEL);

	spin_lock_irq(&tsk->sighand->siglock);
	if (!sig->stats) {
		sig->stats = stats;
		stats = NULL;
	}
	spin_unlock_irq(&tsk->sighand->siglock);

	if (stats)
		kmem_cache_free(taskstats_cache, stats);
ret:
	return sig->stats;
}

/* Send pid data out on exit */
void taskstats_exit(struct task_struct *tsk, int group_dead)
{
	int rc;
	struct listener_list *listeners;
	struct taskstats *stats;
	struct sk_buff *rep_skb;
	size_t size;
	int is_thread_group;

	if (!family_registered)
		return;

	/*
	 * Size includes space for nested attributes
	 */
	size = taskstats_packet_size();

	is_thread_group = !!taskstats_tgid_alloc(tsk);
	if (is_thread_group) {
		/* PID + STATS + TGID + STATS */
		size = 2 * size;
		/* fill the tsk->signal->stats structure */
		fill_tgid_exit(tsk);
	}

	listeners = raw_cpu_ptr(&listener_array);
	if (list_empty(&listeners->list))
		return;

	rc = prepare_reply(NULL, TASKSTATS_CMD_NEW, &rep_skb, size);
	if (rc < 0)
		return;

	stats = mk_reply(rep_skb, TASKSTATS_TYPE_PID,
			 task_pid_nr_ns(tsk, &init_pid_ns));
	if (!stats)
		goto err;

	fill_stats(&init_user_ns, &init_pid_ns, tsk, stats);

	/*
	 * Doesn't matter if tsk is the leader or the last group member leaving
	 */
	if (!is_thread_group || !group_dead)
		goto send;

	stats = mk_reply(rep_skb, TASKSTATS_TYPE_TGID,
			 task_tgid_nr_ns(tsk, &init_pid_ns));
	if (!stats)
		goto err;

	memcpy(stats, tsk->signal->stats, sizeof(*stats));

send:
	send_cpu_listeners(rep_skb, listeners);
	return;
err:
	nlmsg_free(rep_skb);
}

static const struct genl_ops taskstats_ops[] = {
	{
		.cmd		= TASKSTATS_CMD_GET,
		.doit		= taskstats_user_cmd,
		.policy		= taskstats_cmd_get_policy,
		.flags		= GENL_ADMIN_PERM,
	},
	{
		.cmd		= CGROUPSTATS_CMD_GET,
		.doit		= cgroupstats_user_cmd,
		.policy		= cgroupstats_cmd_get_policy,
	},
};

static struct genl_family family __ro_after_init = {
	.name		= TASKSTATS_GENL_NAME,
	.version	= TASKSTATS_GENL_VERSION,
	.maxattr	= TASKSTATS_CMD_ATTR_MAX,
	.module		= THIS_MODULE,
	.ops		= taskstats_ops,
	.n_ops		= ARRAY_SIZE(taskstats_ops),
};

/* Needed early in initialization */
void __init taskstats_init_early(void)
{
	unsigned int i;

	taskstats_cache = KMEM_CACHE(taskstats, SLAB_PANIC);
	for_each_possible_cpu(i) {
		INIT_LIST_HEAD(&(per_cpu(listener_array, i).list));
		init_rwsem(&(per_cpu(listener_array, i).sem));
	}
}

static int __init taskstats_init(void)
{
	int rc;

	rc = genl_register_family(&family);
	if (rc)
		return rc;

	family_registered = 1;
	pr_info("registered taskstats version %d\n", TASKSTATS_GENL_VERSION);
	return 0;
}

/*
 * late initcall ensures initialization of statistics collection
 * mechanisms precedes initialization of the taskstats interface
 */
late_initcall(taskstats_init);
Commit	Line	Data
c757249a SN	1	/*
	2	* taskstats.c - Export per-task statistics to userland
	3	*
	4	* Copyright (C) Shailabh Nagar, IBM Corp. 2006
	5	* (C) Balbir Singh, IBM Corp. 2006
	6	*
	7	* This program is free software; you can redistribute it and/or modify
	8	* it under the terms of the GNU General Public License as published by
	9	* the Free Software Foundation; either version 2 of the License, or
	10	* (at your option) any later version.
	11	*
	12	* This program is distributed in the hope that it will be useful,
	13	* but WITHOUT ANY WARRANTY; without even the implied warranty of
	14	* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
	15	* GNU General Public License for more details.
	16	*
	17	*/
	18
	19	#include <linux/kernel.h>
	20	#include <linux/taskstats_kern.h>
f3cef7a9	21	#include <linux/tsacct_kern.h>
6f44993f	22	#include <linux/delayacct.h>
f9fd8914 SN	23	#include <linux/cpumask.h>
f9fd8914 SN	24	#include <linux/percpu.h>
5a0e3ad6	25	#include <linux/slab.h>
846c7bb0 BS	26	#include <linux/cgroupstats.h>
	27	#include <linux/cgroup.h>
	28	#include <linux/fs.h>
	29	#include <linux/file.h>
4bd6e32a	30	#include <linux/pid_namespace.h>
c757249a	31	#include <net/genetlink.h>
60063497	32	#include <linux/atomic.h>
c757249a	33
f9fd8914 SN	34	/*
	35	* Maximum length of a cpumask that can be specified in
	36	* the TASKSTATS_CMD_ATTR_REGISTER/DEREGISTER_CPUMASK attribute
	37	*/
	38	#define TASKSTATS_CPUMASK_MAXLEN (100+6*NR_CPUS)
	39
b81f3ea9	40	static DEFINE_PER_CPU(__u32, taskstats_seqnum);
c757249a	41	static int family_registered;
e18b890b	42	struct kmem_cache *taskstats_cache;
c757249a	43
489111e5	44	static struct genl_family family;
c757249a	45
b54452b0	46	static const struct nla_policy taskstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1] = {
c757249a SN	47	[TASKSTATS_CMD_ATTR_PID] = { .type = NLA_U32 },
c757249a SN	48	[TASKSTATS_CMD_ATTR_TGID] = { .type = NLA_U32 },
f9fd8914 SN	49	[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK] = { .type = NLA_STRING },
	50	[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },};
	51
243d5212 WC	52	/*
	53	* We have to use TASKSTATS_CMD_ATTR_MAX here, it is the maxattr in the family.
	54	* Make sure they are always aligned.
	55	*/
	56	static const struct nla_policy cgroupstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1] = {
846c7bb0 BS	57	[CGROUPSTATS_CMD_ATTR_FD] = { .type = NLA_U32 },
	58	};
	59
f9fd8914 SN	60	struct listener {
	61	struct list_head list;
	62	pid_t pid;
bb129994	63	char valid;
c757249a SN	64	};
c757249a SN	65
f9fd8914 SN	66	struct listener_list {
	67	struct rw_semaphore sem;
	68	struct list_head list;
	69	};
	70	static DEFINE_PER_CPU(struct listener_list, listener_array);
	71
	72	enum actions {
	73	REGISTER,
	74	DEREGISTER,
	75	CPU_DONT_CARE
	76	};
c757249a SN	77
c757249a SN	78	static int prepare_reply(struct genl_info info, u8 cmd, struct sk_buff *skbp,
37167485	79	size_t size)
c757249a SN	80	{
	81	struct sk_buff *skb;
	82	void *reply;
	83
	84	/*
	85	* If new attributes are added, please revisit this allocation
	86	*/
3dabc715	87	skb = genlmsg_new(size, GFP_KERNEL);
c757249a SN	88	if (!skb)
	89	return -ENOMEM;
	90
	91	if (!info) {
cd85fc58	92	int seq = this_cpu_inc_return(taskstats_seqnum) - 1;
c757249a	93
17c157c8	94	reply = genlmsg_put(skb, 0, seq, &family, 0, cmd);
c757249a	95	} else
17c157c8	96	reply = genlmsg_put_reply(skb, info, &family, 0, cmd);
c757249a SN	97	if (reply == NULL) {
	98	nlmsg_free(skb);
	99	return -EINVAL;
	100	}
	101
	102	*skbp = skb;
c757249a SN	103	return 0;
	104	}
	105
f9fd8914 SN	106	/*
	107	* Send taskstats data in @skb to listener with nl_pid @pid
	108	*/
134e6375	109	static int send_reply(struct sk_buff skb, struct genl_info info)
c757249a	110	{
b529ccf2	111	struct genlmsghdr *genlhdr = nlmsg_data(nlmsg_hdr(skb));
f9fd8914	112	void *reply = genlmsg_data(genlhdr);
c757249a	113
053c095a	114	genlmsg_end(skb, reply);
c757249a	115
134e6375	116	return genlmsg_reply(skb, info);
c757249a SN	117	}
c757249a SN	118
f9fd8914 SN	119	/*
	120	* Send taskstats data in @skb to listeners registered for @cpu's exit data
	121	*/
115085ea ON	122	static void send_cpu_listeners(struct sk_buff *skb,
115085ea ON	123	struct listener_list *listeners)
f9fd8914	124	{
b529ccf2	125	struct genlmsghdr *genlhdr = nlmsg_data(nlmsg_hdr(skb));
f9fd8914 SN	126	struct listener s, tmp;
	127	struct sk_buff skb_next, skb_cur = skb;
	128	void *reply = genlmsg_data(genlhdr);
d94a0415	129	int rc, delcount = 0;
f9fd8914	130
053c095a	131	genlmsg_end(skb, reply);
f9fd8914 SN	132
f9fd8914 SN	133	rc = 0;
bb129994	134	down_read(&listeners->sem);
d94a0415	135	list_for_each_entry(s, &listeners->list, list) {
f9fd8914 SN	136	skb_next = NULL;
	137	if (!list_is_last(&s->list, &listeners->list)) {
	138	skb_next = skb_clone(skb_cur, GFP_KERNEL);
d94a0415	139	if (!skb_next)
f9fd8914	140	break;
f9fd8914	141	}
134e6375	142	rc = genlmsg_unicast(&init_net, skb_cur, s->pid);
d94a0415	143	if (rc == -ECONNREFUSED) {
bb129994 SN	144	s->valid = 0;
bb129994 SN	145	delcount++;
f9fd8914 SN	146	}
	147	skb_cur = skb_next;
	148	}
bb129994	149	up_read(&listeners->sem);
f9fd8914	150
d94a0415 SN	151	if (skb_cur)
	152	nlmsg_free(skb_cur);
	153
bb129994	154	if (!delcount)
d94a0415	155	return;
bb129994 SN	156
	157	/* Delete invalidated entries */
	158	down_write(&listeners->sem);
	159	list_for_each_entry_safe(s, tmp, &listeners->list, list) {
	160	if (!s->valid) {
	161	list_del(&s->list);
	162	kfree(s);
	163	}
	164	}
	165	up_write(&listeners->sem);
f9fd8914 SN	166	}
f9fd8914 SN	167
4bd6e32a EB	168	static void fill_stats(struct user_namespace *user_ns,
	169	struct pid_namespace *pid_ns,
	170	struct task_struct tsk, struct taskstats stats)
c757249a	171	{
51de4d90	172	memset(stats, 0, sizeof(*stats));
c757249a SN	173	/*
	174	* Each accounting subsystem adds calls to its functions to
	175	* fill in relevant parts of struct taskstsats as follows
	176	*
7d94dddd	177	* per-task-foo(stats, tsk);
c757249a SN	178	*/
c757249a SN	179
7d94dddd	180	delayacct_add_tsk(stats, tsk);
f3cef7a9 JL	181
f3cef7a9 JL	182	/* fill in basic acct fields */
6f44993f	183	stats->version = TASKSTATS_VERSION;
b663a79c MU	184	stats->nvcsw = tsk->nvcsw;
b663a79c MU	185	stats->nivcsw = tsk->nivcsw;
4bd6e32a	186	bacct_add_tsk(user_ns, pid_ns, stats, tsk);
6f44993f	187
9acc1853 JL	188	/* fill in extended acct fields */
9acc1853 JL	189	xacct_add_tsk(stats, tsk);
3d9e0cf1	190	}
9acc1853	191
3d9e0cf1 MH	192	static int fill_stats_for_pid(pid_t pid, struct taskstats *stats)
	193	{
	194	struct task_struct *tsk;
c757249a	195
3d9e0cf1 MH	196	rcu_read_lock();
	197	tsk = find_task_by_vpid(pid);
	198	if (tsk)
	199	get_task_struct(tsk);
	200	rcu_read_unlock();
	201	if (!tsk)
	202	return -ESRCH;
4bd6e32a	203	fill_stats(current_user_ns(), task_active_pid_ns(current), tsk, stats);
3d9e0cf1 MH	204	put_task_struct(tsk);
3d9e0cf1 MH	205	return 0;
c757249a SN	206	}
c757249a SN	207
3d9e0cf1	208	static int fill_stats_for_tgid(pid_t tgid, struct taskstats *stats)
c757249a	209	{
3d9e0cf1	210	struct task_struct tsk, first;
ad4ecbcb	211	unsigned long flags;
a98b6094	212	int rc = -ESRCH;
c757249a	213
ad4ecbcb SN	214	/*
	215	* Add additional stats from live tasks except zombie thread group
	216	* leaders who are already counted with the dead tasks
	217	*/
a98b6094	218	rcu_read_lock();
3d9e0cf1	219	first = find_task_by_vpid(tgid);
ad4ecbcb	220
a98b6094 ON	221	if (!first \|\| !lock_task_sighand(first, &flags))
a98b6094 ON	222	goto out;
ad4ecbcb	223
a98b6094 ON	224	if (first->signal->stats)
a98b6094 ON	225	memcpy(stats, first->signal->stats, sizeof(*stats));
51de4d90 ON	226	else
51de4d90 ON	227	memset(stats, 0, sizeof(*stats));
fca178c0	228
a98b6094	229	tsk = first;
c757249a	230	do {
d7c3f5f2	231	if (tsk->exit_state)
ad4ecbcb	232	continue;
c757249a	233	/*
ad4ecbcb	234	* Accounting subsystem can call its functions here to
c757249a SN	235	* fill in relevant parts of struct taskstsats as follows
c757249a SN	236	*
ad4ecbcb	237	* per-task-foo(stats, tsk);
c757249a	238	*/
ad4ecbcb	239	delayacct_add_tsk(stats, tsk);
6f44993f	240
b663a79c MU	241	stats->nvcsw += tsk->nvcsw;
b663a79c MU	242	stats->nivcsw += tsk->nivcsw;
c757249a	243	} while_each_thread(first, tsk);
6f44993f	244
a98b6094 ON	245	unlock_task_sighand(first, &flags);
	246	rc = 0;
	247	out:
	248	rcu_read_unlock();
	249
	250	stats->version = TASKSTATS_VERSION;
c757249a	251	/*
3a4fa0a2	252	* Accounting subsystems can also add calls here to modify
ad4ecbcb	253	* fields of taskstats.
c757249a	254	*/
a98b6094	255	return rc;
ad4ecbcb SN	256	}
ad4ecbcb SN	257
ad4ecbcb SN	258	static void fill_tgid_exit(struct task_struct *tsk)
	259	{
	260	unsigned long flags;
	261
b8534d7b	262	spin_lock_irqsave(&tsk->sighand->siglock, flags);
ad4ecbcb SN	263	if (!tsk->signal->stats)
	264	goto ret;
	265
	266	/*
	267	* Each accounting subsystem calls its functions here to
	268	* accumalate its per-task stats for tsk, into the per-tgid structure
	269	*
	270	* per-task-foo(tsk->signal->stats, tsk);
	271	*/
	272	delayacct_add_tsk(tsk->signal->stats, tsk);
	273	ret:
b8534d7b	274	spin_unlock_irqrestore(&tsk->sighand->siglock, flags);
ad4ecbcb	275	return;
c757249a SN	276	}
c757249a SN	277
41c7bb95	278	static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd)
f9fd8914 SN	279	{
f9fd8914 SN	280	struct listener_list *listeners;
26c4caea	281	struct listener s, tmp, *s2;
f9fd8914	282	unsigned int cpu;
0d20633b	283	int ret = 0;
ad4ecbcb	284
41c7bb95	285	if (!cpumask_subset(mask, cpu_possible_mask))
f9fd8914 SN	286	return -EINVAL;
f9fd8914 SN	287
4bd6e32a EB	288	if (current_user_ns() != &init_user_ns)
	289	return -EINVAL;
	290
	291	if (task_active_pid_ns(current) != &init_pid_ns)
	292	return -EINVAL;
	293
f9fd8914	294	if (isadd == REGISTER) {
41c7bb95	295	for_each_cpu(cpu, mask) {
dfc428b6 ON	296	s = kmalloc_node(sizeof(struct listener),
dfc428b6 ON	297	GFP_KERNEL, cpu_to_node(cpu));
0d20633b CG	298	if (!s) {
0d20633b CG	299	ret = -ENOMEM;
f9fd8914	300	goto cleanup;
0d20633b	301	}
f9fd8914	302	s->pid = pid;
bb129994	303	s->valid = 1;
f9fd8914 SN	304
	305	listeners = &per_cpu(listener_array, cpu);
	306	down_write(&listeners->sem);
dfc428b6	307	list_for_each_entry(s2, &listeners->list, list) {
a7295898	308	if (s2->pid == pid && s2->valid)
dfc428b6	309	goto exists;
26c4caea	310	}
f9fd8914	311	list_add(&s->list, &listeners->list);
26c4caea	312	s = NULL;
dfc428b6	313	exists:
f9fd8914	314	up_write(&listeners->sem);
dfc428b6	315	kfree(s); /* nop if NULL */
f9fd8914 SN	316	}
	317	return 0;
	318	}
	319
	320	/* Deregister or cleanup */
	321	cleanup:
41c7bb95	322	for_each_cpu(cpu, mask) {
f9fd8914 SN	323	listeners = &per_cpu(listener_array, cpu);
	324	down_write(&listeners->sem);
	325	list_for_each_entry_safe(s, tmp, &listeners->list, list) {
	326	if (s->pid == pid) {
	327	list_del(&s->list);
	328	kfree(s);
	329	break;
	330	}
	331	}
	332	up_write(&listeners->sem);
	333	}
0d20633b	334	return ret;
f9fd8914 SN	335	}
f9fd8914 SN	336
41c7bb95	337	static int parse(struct nlattr na, struct cpumask mask)
f9fd8914 SN	338	{
	339	char *data;
	340	int len;
	341	int ret;
	342
	343	if (na == NULL)
	344	return 1;
	345	len = nla_len(na);
	346	if (len > TASKSTATS_CPUMASK_MAXLEN)
	347	return -E2BIG;
	348	if (len < 1)
	349	return -EINVAL;
	350	data = kmalloc(len, GFP_KERNEL);
	351	if (!data)
	352	return -ENOMEM;
	353	nla_strlcpy(data, na, len);
29c0177e	354	ret = cpulist_parse(data, mask);
f9fd8914 SN	355	kfree(data);
	356	return ret;
	357	}
	358
51de4d90	359	static struct taskstats mk_reply(struct sk_buff skb, int type, u32 pid)
68062b86	360	{
51de4d90	361	struct nlattr na, ret;
68062b86 ON	362	int aggr;
68062b86 ON	363
37167485 ON	364	aggr = (type == TASKSTATS_TYPE_PID)
	365	? TASKSTATS_TYPE_AGGR_PID
	366	: TASKSTATS_TYPE_AGGR_TGID;
68062b86 ON	367
68062b86 ON	368	na = nla_nest_start(skb, aggr);
37167485 ON	369	if (!na)
37167485 ON	370	goto err;
4be2c95d	371
3fa58266 CG	372	if (nla_put(skb, type, sizeof(pid), &pid) < 0) {
3fa58266 CG	373	nla_nest_cancel(skb, na);
51de4d90	374	goto err;
3fa58266	375	}
80df5542 ND	376	ret = nla_reserve_64bit(skb, TASKSTATS_TYPE_STATS,
80df5542 ND	377	sizeof(struct taskstats), TASKSTATS_TYPE_NULL);
3fa58266 CG	378	if (!ret) {
3fa58266 CG	379	nla_nest_cancel(skb, na);
51de4d90	380	goto err;
3fa58266	381	}
68062b86 ON	382	nla_nest_end(skb, na);
68062b86 ON	383
51de4d90 ON	384	return nla_data(ret);
	385	err:
	386	return NULL;
68062b86 ON	387	}
68062b86 ON	388
846c7bb0 BS	389	static int cgroupstats_user_cmd(struct sk_buff skb, struct genl_info info)
	390	{
	391	int rc = 0;
	392	struct sk_buff *rep_skb;
	393	struct cgroupstats *stats;
	394	struct nlattr *na;
	395	size_t size;
	396	u32 fd;
2903ff01	397	struct fd f;
846c7bb0 BS	398
	399	na = info->attrs[CGROUPSTATS_CMD_ATTR_FD];
	400	if (!na)
	401	return -EINVAL;
	402
	403	fd = nla_get_u32(info->attrs[CGROUPSTATS_CMD_ATTR_FD]);
2903ff01 AV	404	f = fdget(fd);
2903ff01 AV	405	if (!f.file)
f9615984	406	return 0;
846c7bb0	407
f9615984	408	size = nla_total_size(sizeof(struct cgroupstats));
846c7bb0	409
f9615984 AB	410	rc = prepare_reply(info, CGROUPSTATS_CMD_NEW, &rep_skb,
	411	size);
	412	if (rc < 0)
	413	goto err;
846c7bb0	414
f9615984 AB	415	na = nla_reserve(rep_skb, CGROUPSTATS_TYPE_CGROUP_STATS,
f9615984 AB	416	sizeof(struct cgroupstats));
25353b33	417	if (na == NULL) {
0324b5a4	418	nlmsg_free(rep_skb);
25353b33 AC	419	rc = -EMSGSIZE;
	420	goto err;
	421	}
	422
f9615984 AB	423	stats = nla_data(na);
f9615984 AB	424	memset(stats, 0, sizeof(*stats));
846c7bb0	425
b583043e	426	rc = cgroupstats_build(stats, f.file->f_path.dentry);
f9615984 AB	427	if (rc < 0) {
	428	nlmsg_free(rep_skb);
	429	goto err;
846c7bb0 BS	430	}
846c7bb0 BS	431
134e6375	432	rc = send_reply(rep_skb, info);
f9615984	433
846c7bb0	434	err:
2903ff01	435	fdput(f);
846c7bb0 BS	436	return rc;
	437	}
	438
93233125	439	static int cmd_attr_register_cpumask(struct genl_info *info)
c757249a	440	{
41c7bb95	441	cpumask_var_t mask;
93233125	442	int rc;
41c7bb95 RR	443
	444	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
	445	return -ENOMEM;
41c7bb95	446	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK], mask);
f9fd8914	447	if (rc < 0)
93233125	448	goto out;
15e47304	449	rc = add_del_listener(info->snd_portid, mask, REGISTER);
93233125 MH	450	out:
	451	free_cpumask_var(mask);
	452	return rc;
	453	}
	454
	455	static int cmd_attr_deregister_cpumask(struct genl_info *info)
	456	{
	457	cpumask_var_t mask;
	458	int rc;
f9fd8914	459
93233125 MH	460	if (!alloc_cpumask_var(&mask, GFP_KERNEL))
93233125 MH	461	return -ENOMEM;
41c7bb95	462	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK], mask);
f9fd8914	463	if (rc < 0)
93233125	464	goto out;
15e47304	465	rc = add_del_listener(info->snd_portid, mask, DEREGISTER);
93233125	466	out:
41c7bb95	467	free_cpumask_var(mask);
93233125 MH	468	return rc;
	469	}
	470
4be2c95d JM	471	static size_t taskstats_packet_size(void)
	472	{
	473	size_t size;
	474
	475	size = nla_total_size(sizeof(u32)) +
80df5542 ND	476	nla_total_size_64bit(sizeof(struct taskstats)) +
	477	nla_total_size(0);
	478
4be2c95d JM	479	return size;
	480	}
	481
93233125 MH	482	static int cmd_attr_pid(struct genl_info *info)
	483	{
	484	struct taskstats *stats;
	485	struct sk_buff *rep_skb;
	486	size_t size;
	487	u32 pid;
	488	int rc;
c757249a	489
4be2c95d	490	size = taskstats_packet_size();
c757249a	491
37167485	492	rc = prepare_reply(info, TASKSTATS_CMD_NEW, &rep_skb, size);
c757249a SN	493	if (rc < 0)
	494	return rc;
	495
51de4d90	496	rc = -EINVAL;
93233125 MH	497	pid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_PID]);
	498	stats = mk_reply(rep_skb, TASKSTATS_TYPE_PID, pid);
	499	if (!stats)
c757249a	500	goto err;
c757249a	501
3d9e0cf1	502	rc = fill_stats_for_pid(pid, stats);
93233125 MH	503	if (rc < 0)
93233125 MH	504	goto err;
134e6375	505	return send_reply(rep_skb, info);
c757249a SN	506	err:
	507	nlmsg_free(rep_skb);
	508	return rc;
	509	}
	510
93233125 MH	511	static int cmd_attr_tgid(struct genl_info *info)
	512	{
	513	struct taskstats *stats;
	514	struct sk_buff *rep_skb;
	515	size_t size;
	516	u32 tgid;
	517	int rc;
	518
4be2c95d	519	size = taskstats_packet_size();
93233125 MH	520
	521	rc = prepare_reply(info, TASKSTATS_CMD_NEW, &rep_skb, size);
	522	if (rc < 0)
	523	return rc;
	524
	525	rc = -EINVAL;
	526	tgid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_TGID]);
	527	stats = mk_reply(rep_skb, TASKSTATS_TYPE_TGID, tgid);
	528	if (!stats)
	529	goto err;
	530
3d9e0cf1	531	rc = fill_stats_for_tgid(tgid, stats);
93233125 MH	532	if (rc < 0)
	533	goto err;
	534	return send_reply(rep_skb, info);
	535	err:
	536	nlmsg_free(rep_skb);
	537	return rc;
	538	}
	539
	540	static int taskstats_user_cmd(struct sk_buff skb, struct genl_info info)
	541	{
	542	if (info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK])
	543	return cmd_attr_register_cpumask(info);
	544	else if (info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK])
	545	return cmd_attr_deregister_cpumask(info);
	546	else if (info->attrs[TASKSTATS_CMD_ATTR_PID])
	547	return cmd_attr_pid(info);
	548	else if (info->attrs[TASKSTATS_CMD_ATTR_TGID])
	549	return cmd_attr_tgid(info);
	550	else
	551	return -EINVAL;
	552	}
	553
34ec1234 ON	554	static struct taskstats taskstats_tgid_alloc(struct task_struct tsk)
	555	{
	556	struct signal_struct *sig = tsk->signal;
	557	struct taskstats *stats;
	558
	559	if (sig->stats \|\| thread_group_empty(tsk))
	560	goto ret;
	561
	562	/* No problem if kmem_cache_zalloc() fails */
	563	stats = kmem_cache_zalloc(taskstats_cache, GFP_KERNEL);
	564
	565	spin_lock_irq(&tsk->sighand->siglock);
	566	if (!sig->stats) {
	567	sig->stats = stats;
	568	stats = NULL;
	569	}
	570	spin_unlock_irq(&tsk->sighand->siglock);
	571
	572	if (stats)
	573	kmem_cache_free(taskstats_cache, stats);
	574	ret:
	575	return sig->stats;
	576	}
	577
c757249a	578	/* Send pid data out on exit */
115085ea	579	void taskstats_exit(struct task_struct *tsk, int group_dead)
c757249a SN	580	{
c757249a SN	581	int rc;
115085ea	582	struct listener_list *listeners;
51de4d90	583	struct taskstats *stats;
c757249a	584	struct sk_buff *rep_skb;
c757249a SN	585	size_t size;
c757249a SN	586	int is_thread_group;
c757249a	587
4a279ff1	588	if (!family_registered)
c757249a SN	589	return;
c757249a SN	590
c757249a SN	591	/*
	592	* Size includes space for nested attributes
	593	*/
4be2c95d	594	size = taskstats_packet_size();
c757249a	595
34ec1234	596	is_thread_group = !!taskstats_tgid_alloc(tsk);
4a279ff1 ON	597	if (is_thread_group) {
	598	/* PID + STATS + TGID + STATS */
	599	size = 2 * size;
	600	/* fill the tsk->signal->stats structure */
	601	fill_tgid_exit(tsk);
	602	}
	603
4a32fea9	604	listeners = raw_cpu_ptr(&listener_array);
115085ea ON	605	if (list_empty(&listeners->list))
	606	return;
	607
37167485	608	rc = prepare_reply(NULL, TASKSTATS_CMD_NEW, &rep_skb, size);
c757249a	609	if (rc < 0)
51de4d90	610	return;
c757249a	611
4bd6e32a EB	612	stats = mk_reply(rep_skb, TASKSTATS_TYPE_PID,
4bd6e32a EB	613	task_pid_nr_ns(tsk, &init_pid_ns));
51de4d90	614	if (!stats)
37167485	615	goto err;
c757249a	616
4bd6e32a	617	fill_stats(&init_user_ns, &init_pid_ns, tsk, stats);
c757249a	618
c757249a	619	/*
ad4ecbcb	620	* Doesn't matter if tsk is the leader or the last group member leaving
c757249a	621	*/
68062b86	622	if (!is_thread_group \|\| !group_dead)
ad4ecbcb	623	goto send;
c757249a	624
4bd6e32a EB	625	stats = mk_reply(rep_skb, TASKSTATS_TYPE_TGID,
4bd6e32a EB	626	task_tgid_nr_ns(tsk, &init_pid_ns));
51de4d90	627	if (!stats)
37167485	628	goto err;
51de4d90 ON	629
51de4d90 ON	630	memcpy(stats, tsk->signal->stats, sizeof(*stats));
c757249a	631
ad4ecbcb	632	send:
115085ea	633	send_cpu_listeners(rep_skb, listeners);
ad4ecbcb	634	return;
37167485	635	err:
c757249a	636	nlmsg_free(rep_skb);
c757249a SN	637	}
c757249a SN	638
4534de83	639	static const struct genl_ops taskstats_ops[] = {
88d36a99 JB	640	{
	641	.cmd = TASKSTATS_CMD_GET,
	642	.doit = taskstats_user_cmd,
	643	.policy = taskstats_cmd_get_policy,
	644	.flags = GENL_ADMIN_PERM,
	645	},
	646	{
	647	.cmd = CGROUPSTATS_CMD_GET,
	648	.doit = cgroupstats_user_cmd,
	649	.policy = cgroupstats_cmd_get_policy,
	650	},
846c7bb0 BS	651	};
846c7bb0 BS	652
56989f6d	653	static struct genl_family family __ro_after_init = {
489111e5 JB	654	.name = TASKSTATS_GENL_NAME,
	655	.version = TASKSTATS_GENL_VERSION,
	656	.maxattr = TASKSTATS_CMD_ATTR_MAX,
	657	.module = THIS_MODULE,
	658	.ops = taskstats_ops,
	659	.n_ops = ARRAY_SIZE(taskstats_ops),
	660	};
	661
c757249a SN	662	/* Needed early in initialization */
	663	void __init taskstats_init_early(void)
	664	{
f9fd8914 SN	665	unsigned int i;
f9fd8914 SN	666
0a31bd5f	667	taskstats_cache = KMEM_CACHE(taskstats, SLAB_PANIC);
f9fd8914 SN	668	for_each_possible_cpu(i) {
	669	INIT_LIST_HEAD(&(per_cpu(listener_array, i).list));
	670	init_rwsem(&(per_cpu(listener_array, i).sem));
	671	}
c757249a SN	672	}
	673
	674	static int __init taskstats_init(void)
	675	{
	676	int rc;
	677
489111e5	678	rc = genl_register_family(&family);
c757249a SN	679	if (rc)
	680	return rc;
	681
c757249a	682	family_registered = 1;
f9b182e2	683	pr_info("registered taskstats version %d\n", TASKSTATS_GENL_VERSION);
c757249a	684	return 0;
c757249a SN	685	}
	686
	687	/*
	688	* late initcall ensures initialization of statistics collection
	689	* mechanisms precedes initialization of the taskstats interface
	690	*/
	691	late_initcall(taskstats_init);