taskstats.c 13.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
/*
 * 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>
21
#include <linux/tsacct_kern.h>
22
#include <linux/delayacct.h>
23
#include <linux/tsacct_kern.h>
24 25
#include <linux/cpumask.h>
#include <linux/percpu.h>
26 27 28
#include <net/genetlink.h>
#include <asm/atomic.h>

29 30 31 32 33 34
/*
 * 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)

35 36 37 38 39 40 41 42 43 44 45 46 47 48 49
static DEFINE_PER_CPU(__u32, taskstats_seqnum) = { 0 };
static int family_registered;
kmem_cache_t *taskstats_cache;

static struct genl_family family = {
	.id		= GENL_ID_GENERATE,
	.name		= TASKSTATS_GENL_NAME,
	.version	= TASKSTATS_GENL_VERSION,
	.maxattr	= TASKSTATS_CMD_ATTR_MAX,
};

static struct nla_policy taskstats_cmd_get_policy[TASKSTATS_CMD_ATTR_MAX+1]
__read_mostly = {
	[TASKSTATS_CMD_ATTR_PID]  = { .type = NLA_U32 },
	[TASKSTATS_CMD_ATTR_TGID] = { .type = NLA_U32 },
50 51 52 53 54 55
	[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK] = { .type = NLA_STRING },
	[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK] = { .type = NLA_STRING },};

struct listener {
	struct list_head list;
	pid_t pid;
56
	char valid;
57 58
};

59 60 61 62 63 64 65 66 67 68 69
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
};
70 71 72 73 74 75 76 77 78 79

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

	/*
	 * If new attributes are added, please revisit this allocation
	 */
80
	skb = nlmsg_new(genlmsg_total_size(size), GFP_KERNEL);
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
	if (!skb)
		return -ENOMEM;

	if (!info) {
		int seq = get_cpu_var(taskstats_seqnum)++;
		put_cpu_var(taskstats_seqnum);

		reply = genlmsg_put(skb, 0, seq,
				family.id, 0, 0,
				cmd, family.version);
	} else
		reply = genlmsg_put(skb, info->snd_pid, info->snd_seq,
				family.id, 0, 0,
				cmd, family.version);
	if (reply == NULL) {
		nlmsg_free(skb);
		return -EINVAL;
	}

	*skbp = skb;
	*replyp = reply;
	return 0;
}

105 106 107 108
/*
 * Send taskstats data in @skb to listener with nl_pid @pid
 */
static int send_reply(struct sk_buff *skb, pid_t pid)
109 110
{
	struct genlmsghdr *genlhdr = nlmsg_data((struct nlmsghdr *)skb->data);
111
	void *reply = genlmsg_data(genlhdr);
112 113 114 115 116 117 118 119 120 121 122
	int rc;

	rc = genlmsg_end(skb, reply);
	if (rc < 0) {
		nlmsg_free(skb);
		return rc;
	}

	return genlmsg_unicast(skb, pid);
}

123 124 125
/*
 * Send taskstats data in @skb to listeners registered for @cpu's exit data
 */
126
static void send_cpu_listeners(struct sk_buff *skb, unsigned int cpu)
127 128 129 130 131 132
{
	struct genlmsghdr *genlhdr = nlmsg_data((struct nlmsghdr *)skb->data);
	struct listener_list *listeners;
	struct listener *s, *tmp;
	struct sk_buff *skb_next, *skb_cur = skb;
	void *reply = genlmsg_data(genlhdr);
133
	int rc, delcount = 0;
134 135 136 137

	rc = genlmsg_end(skb, reply);
	if (rc < 0) {
		nlmsg_free(skb);
138
		return;
139 140 141 142
	}

	rc = 0;
	listeners = &per_cpu(listener_array, cpu);
143
	down_read(&listeners->sem);
144
	list_for_each_entry(s, &listeners->list, list) {
145 146 147
		skb_next = NULL;
		if (!list_is_last(&s->list, &listeners->list)) {
			skb_next = skb_clone(skb_cur, GFP_KERNEL);
148
			if (!skb_next)
149 150
				break;
		}
151 152
		rc = genlmsg_unicast(skb_cur, s->pid);
		if (rc == -ECONNREFUSED) {
153 154
			s->valid = 0;
			delcount++;
155 156 157
		}
		skb_cur = skb_next;
	}
158
	up_read(&listeners->sem);
159

160 161 162
	if (skb_cur)
		nlmsg_free(skb_cur);

163
	if (!delcount)
164
		return;
165 166 167 168 169 170 171 172 173 174

	/* 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);
175 176
}

177 178 179
static int fill_pid(pid_t pid, struct task_struct *pidtsk,
		struct taskstats *stats)
{
180
	int rc = 0;
181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198
	struct task_struct *tsk = pidtsk;

	if (!pidtsk) {
		read_lock(&tasklist_lock);
		tsk = find_task_by_pid(pid);
		if (!tsk) {
			read_unlock(&tasklist_lock);
			return -ESRCH;
		}
		get_task_struct(tsk);
		read_unlock(&tasklist_lock);
	} else
		get_task_struct(tsk);

	/*
	 * Each accounting subsystem adds calls to its functions to
	 * fill in relevant parts of struct taskstsats as follows
	 *
199
	 *	per-task-foo(stats, tsk);
200 201
	 */

202
	delayacct_add_tsk(stats, tsk);
203 204

	/* fill in basic acct fields */
205
	stats->version = TASKSTATS_VERSION;
206
	bacct_add_tsk(stats, tsk);
207

208 209 210
	/* fill in extended acct fields */
	xacct_add_tsk(stats, tsk);

211
	/* Define err: label here if needed */
212 213 214 215 216 217 218 219 220
	put_task_struct(tsk);
	return rc;

}

static int fill_tgid(pid_t tgid, struct task_struct *tgidtsk,
		struct taskstats *stats)
{
	struct task_struct *tsk, *first;
221
	unsigned long flags;
222

223 224 225 226
	/*
	 * Add additional stats from live tasks except zombie thread group
	 * leaders who are already counted with the dead tasks
	 */
227 228
	first = tgidtsk;
	if (!first) {
229
		read_lock(&tasklist_lock);
230 231 232 233 234
		first = find_task_by_pid(tgid);
		if (!first) {
			read_unlock(&tasklist_lock);
			return -ESRCH;
		}
235 236 237 238 239 240 241 242 243 244 245
		get_task_struct(first);
		read_unlock(&tasklist_lock);
	} else
		get_task_struct(first);

	/* Start with stats from dead tasks */
	spin_lock_irqsave(&first->signal->stats_lock, flags);
	if (first->signal->stats)
		memcpy(stats, first->signal->stats, sizeof(*stats));
	spin_unlock_irqrestore(&first->signal->stats_lock, flags);

246
	tsk = first;
247
	read_lock(&tasklist_lock);
248
	do {
249 250
		if (tsk->exit_state == EXIT_ZOMBIE && thread_group_leader(tsk))
			continue;
251
		/*
252
		 * Accounting subsystem can call its functions here to
253 254
		 * fill in relevant parts of struct taskstsats as follows
		 *
255
		 *	per-task-foo(stats, tsk);
256
		 */
257
		delayacct_add_tsk(stats, tsk);
258

259 260
	} while_each_thread(first, tsk);
	read_unlock(&tasklist_lock);
261 262
	stats->version = TASKSTATS_VERSION;

263
	/*
264 265
	 * Accounting subsytems can also add calls here to modify
	 * fields of taskstats.
266 267
	 */

268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289
	return 0;
}


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

	spin_lock_irqsave(&tsk->signal->stats_lock, 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->signal->stats_lock, flags);
	return;
290 291
}

292 293 294 295 296 297
static int add_del_listener(pid_t pid, cpumask_t *maskp, int isadd)
{
	struct listener_list *listeners;
	struct listener *s, *tmp;
	unsigned int cpu;
	cpumask_t mask = *maskp;
298

299 300 301 302 303 304 305 306 307 308 309
	if (!cpus_subset(mask, cpu_possible_map))
		return -EINVAL;

	if (isadd == REGISTER) {
		for_each_cpu_mask(cpu, mask) {
			s = kmalloc_node(sizeof(struct listener), GFP_KERNEL,
					 cpu_to_node(cpu));
			if (!s)
				goto cleanup;
			s->pid = pid;
			INIT_LIST_HEAD(&s->list);
310
			s->valid = 1;
311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359

			listeners = &per_cpu(listener_array, cpu);
			down_write(&listeners->sem);
			list_add(&s->list, &listeners->list);
			up_write(&listeners->sem);
		}
		return 0;
	}

	/* Deregister or cleanup */
cleanup:
	for_each_cpu_mask(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 0;
}

static int parse(struct nlattr *na, cpumask_t *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 int taskstats_user_cmd(struct sk_buff *skb, struct genl_info *info)
360 361 362 363 364 365 366
{
	int rc = 0;
	struct sk_buff *rep_skb;
	struct taskstats stats;
	void *reply;
	size_t size;
	struct nlattr *na;
367 368 369 370 371 372 373 374 375 376 377 378 379
	cpumask_t mask;

	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_REGISTER_CPUMASK], &mask);
	if (rc < 0)
		return rc;
	if (rc == 0)
		return add_del_listener(info->snd_pid, &mask, REGISTER);

	rc = parse(info->attrs[TASKSTATS_CMD_ATTR_DEREGISTER_CPUMASK], &mask);
	if (rc < 0)
		return rc;
	if (rc == 0)
		return add_del_listener(info->snd_pid, &mask, DEREGISTER);
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418

	/*
	 * Size includes space for nested attributes
	 */
	size = nla_total_size(sizeof(u32)) +
		nla_total_size(sizeof(struct taskstats)) + nla_total_size(0);

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

	if (info->attrs[TASKSTATS_CMD_ATTR_PID]) {
		u32 pid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_PID]);
		rc = fill_pid(pid, NULL, &stats);
		if (rc < 0)
			goto err;

		na = nla_nest_start(rep_skb, TASKSTATS_TYPE_AGGR_PID);
		NLA_PUT_U32(rep_skb, TASKSTATS_TYPE_PID, pid);
		NLA_PUT_TYPE(rep_skb, struct taskstats, TASKSTATS_TYPE_STATS,
				stats);
	} else if (info->attrs[TASKSTATS_CMD_ATTR_TGID]) {
		u32 tgid = nla_get_u32(info->attrs[TASKSTATS_CMD_ATTR_TGID]);
		rc = fill_tgid(tgid, NULL, &stats);
		if (rc < 0)
			goto err;

		na = nla_nest_start(rep_skb, TASKSTATS_TYPE_AGGR_TGID);
		NLA_PUT_U32(rep_skb, TASKSTATS_TYPE_TGID, tgid);
		NLA_PUT_TYPE(rep_skb, struct taskstats, TASKSTATS_TYPE_STATS,
				stats);
	} else {
		rc = -EINVAL;
		goto err;
	}

	nla_nest_end(rep_skb, na);

419
	return send_reply(rep_skb, info->snd_pid);
420 421 422 423 424 425 426 427

nla_put_failure:
	return genlmsg_cancel(rep_skb, reply);
err:
	nlmsg_free(rep_skb);
	return rc;
}

428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453
void taskstats_exit_alloc(struct taskstats **ptidstats, unsigned int *mycpu)
{
	struct listener_list *listeners;
	struct taskstats *tmp;
	/*
	 * This is the cpu on which the task is exiting currently and will
	 * be the one for which the exit event is sent, even if the cpu
	 * on which this function is running changes later.
	 */
	*mycpu = raw_smp_processor_id();

	*ptidstats = NULL;
	tmp = kmem_cache_zalloc(taskstats_cache, SLAB_KERNEL);
	if (!tmp)
		return;

	listeners = &per_cpu(listener_array, *mycpu);
	down_read(&listeners->sem);
	if (!list_empty(&listeners->list)) {
		*ptidstats = tmp;
		tmp = NULL;
	}
	up_read(&listeners->sem);
	kfree(tmp);
}

454 455
/* Send pid data out on exit */
void taskstats_exit_send(struct task_struct *tsk, struct taskstats *tidstats,
456
			int group_dead, unsigned int mycpu)
457 458 459 460 461 462 463
{
	int rc;
	struct sk_buff *rep_skb;
	void *reply;
	size_t size;
	int is_thread_group;
	struct nlattr *na;
464
	unsigned long flags;
465 466 467 468

	if (!family_registered || !tidstats)
		return;

469 470 471
	spin_lock_irqsave(&tsk->signal->stats_lock, flags);
	is_thread_group = tsk->signal->stats ? 1 : 0;
	spin_unlock_irqrestore(&tsk->signal->stats_lock, flags);
472

473
	rc = 0;
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
	/*
	 * Size includes space for nested attributes
	 */
	size = nla_total_size(sizeof(u32)) +
		nla_total_size(sizeof(struct taskstats)) + nla_total_size(0);

	if (is_thread_group)
		size = 2 * size;	/* PID + STATS + TGID + STATS */

	rc = prepare_reply(NULL, TASKSTATS_CMD_NEW, &rep_skb, &reply, size);
	if (rc < 0)
		goto ret;

	rc = fill_pid(tsk->pid, tsk, tidstats);
	if (rc < 0)
		goto err_skb;

	na = nla_nest_start(rep_skb, TASKSTATS_TYPE_AGGR_PID);
	NLA_PUT_U32(rep_skb, TASKSTATS_TYPE_PID, (u32)tsk->pid);
	NLA_PUT_TYPE(rep_skb, struct taskstats, TASKSTATS_TYPE_STATS,
			*tidstats);
	nla_nest_end(rep_skb, na);

497 498
	if (!is_thread_group)
		goto send;
499 500

	/*
501 502
	 * tsk has/had a thread group so fill the tsk->signal->stats structure
	 * Doesn't matter if tsk is the leader or the last group member leaving
503
	 */
504 505 506 507

	fill_tgid_exit(tsk);
	if (!group_dead)
		goto send;
508 509 510

	na = nla_nest_start(rep_skb, TASKSTATS_TYPE_AGGR_TGID);
	NLA_PUT_U32(rep_skb, TASKSTATS_TYPE_TGID, (u32)tsk->tgid);
511
	/* No locking needed for tsk->signal->stats since group is dead */
512
	NLA_PUT_TYPE(rep_skb, struct taskstats, TASKSTATS_TYPE_STATS,
513
			*tsk->signal->stats);
514 515
	nla_nest_end(rep_skb, na);

516
send:
517
	send_cpu_listeners(rep_skb, mycpu);
518
	return;
519 520 521 522 523 524 525 526 527 528 529 530

nla_put_failure:
	genlmsg_cancel(rep_skb, reply);
	goto ret;
err_skb:
	nlmsg_free(rep_skb);
ret:
	return;
}

static struct genl_ops taskstats_ops = {
	.cmd		= TASKSTATS_CMD_GET,
531
	.doit		= taskstats_user_cmd,
532 533 534 535 536 537
	.policy		= taskstats_cmd_get_policy,
};

/* Needed early in initialization */
void __init taskstats_init_early(void)
{
538 539
	unsigned int i;

540 541 542
	taskstats_cache = kmem_cache_create("taskstats_cache",
						sizeof(struct taskstats),
						0, SLAB_PANIC, NULL, NULL);
543 544 545 546
	for_each_possible_cpu(i) {
		INIT_LIST_HEAD(&(per_cpu(listener_array, i).list));
		init_rwsem(&(per_cpu(listener_array, i).sem));
	}
547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572
}

static int __init taskstats_init(void)
{
	int rc;

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

	rc = genl_register_ops(&family, &taskstats_ops);
	if (rc < 0)
		goto err;

	family_registered = 1;
	return 0;
err:
	genl_unregister_family(&family);
	return rc;
}

/*
 * late initcall ensures initialization of statistics collection
 * mechanisms precedes initialization of the taskstats interface
 */
late_initcall(taskstats_init);