sch_generic.c 15.0 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181
/*
 * net/sched/sch_generic.c	Generic packet scheduler routines.
 *
 *		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.
 *
 * Authors:	Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru>
 *              Jamal Hadi Salim, <hadi@cyberus.ca> 990601
 *              - Ingress support
 */

#include <asm/uaccess.h>
#include <asm/system.h>
#include <linux/bitops.h>
#include <linux/config.h>
#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.h>
#include <linux/sched.h>
#include <linux/string.h>
#include <linux/mm.h>
#include <linux/socket.h>
#include <linux/sockios.h>
#include <linux/in.h>
#include <linux/errno.h>
#include <linux/interrupt.h>
#include <linux/netdevice.h>
#include <linux/skbuff.h>
#include <linux/rtnetlink.h>
#include <linux/init.h>
#include <linux/rcupdate.h>
#include <linux/list.h>
#include <net/sock.h>
#include <net/pkt_sched.h>

/* Main transmission queue. */

/* Main qdisc structure lock. 

   However, modifications
   to data, participating in scheduling must be additionally
   protected with dev->queue_lock spinlock.

   The idea is the following:
   - enqueue, dequeue are serialized via top level device
     spinlock dev->queue_lock.
   - tree walking is protected by read_lock_bh(qdisc_tree_lock)
     and this lock is used only in process context.
   - updates to tree are made under rtnl semaphore or
     from softirq context (__qdisc_destroy rcu-callback)
     hence this lock needs local bh disabling.

   qdisc_tree_lock must be grabbed BEFORE dev->queue_lock!
 */
DEFINE_RWLOCK(qdisc_tree_lock);

void qdisc_lock_tree(struct net_device *dev)
{
	write_lock_bh(&qdisc_tree_lock);
	spin_lock_bh(&dev->queue_lock);
}

void qdisc_unlock_tree(struct net_device *dev)
{
	spin_unlock_bh(&dev->queue_lock);
	write_unlock_bh(&qdisc_tree_lock);
}

/* 
   dev->queue_lock serializes queue accesses for this device
   AND dev->qdisc pointer itself.

   dev->xmit_lock serializes accesses to device driver.

   dev->queue_lock and dev->xmit_lock are mutually exclusive,
   if one is grabbed, another must be free.
 */


/* Kick device.
   Note, that this procedure can be called by a watchdog timer, so that
   we do not check dev->tbusy flag here.

   Returns:  0  - queue is empty.
            >0  - queue is not empty, but throttled.
	    <0  - queue is not empty. Device is throttled, if dev->tbusy != 0.

   NOTE: Called under dev->queue_lock with locally disabled BH.
*/

int qdisc_restart(struct net_device *dev)
{
	struct Qdisc *q = dev->qdisc;
	struct sk_buff *skb;

	/* Dequeue packet */
	if ((skb = q->dequeue(q)) != NULL) {
		unsigned nolock = (dev->features & NETIF_F_LLTX);
		/*
		 * When the driver has LLTX set it does its own locking
		 * in start_xmit. No need to add additional overhead by
		 * locking again. These checks are worth it because
		 * even uncongested locks can be quite expensive.
		 * The driver can do trylock like here too, in case
		 * of lock congestion it should return -1 and the packet
		 * will be requeued.
		 */
		if (!nolock) {
			if (!spin_trylock(&dev->xmit_lock)) {
			collision:
				/* So, someone grabbed the driver. */
				
				/* It may be transient configuration error,
				   when hard_start_xmit() recurses. We detect
				   it by checking xmit owner and drop the
				   packet when deadloop is detected.
				*/
				if (dev->xmit_lock_owner == smp_processor_id()) {
					kfree_skb(skb);
					if (net_ratelimit())
						printk(KERN_DEBUG "Dead loop on netdevice %s, fix it urgently!\n", dev->name);
					return -1;
				}
				__get_cpu_var(netdev_rx_stat).cpu_collision++;
				goto requeue;
			}
			/* Remember that the driver is grabbed by us. */
			dev->xmit_lock_owner = smp_processor_id();
		}
		
		{
			/* And release queue */
			spin_unlock(&dev->queue_lock);

			if (!netif_queue_stopped(dev)) {
				int ret;
				if (netdev_nit)
					dev_queue_xmit_nit(skb, dev);

				ret = dev->hard_start_xmit(skb, dev);
				if (ret == NETDEV_TX_OK) { 
					if (!nolock) {
						dev->xmit_lock_owner = -1;
						spin_unlock(&dev->xmit_lock);
					}
					spin_lock(&dev->queue_lock);
					return -1;
				}
				if (ret == NETDEV_TX_LOCKED && nolock) {
					spin_lock(&dev->queue_lock);
					goto collision; 
				}
			}

			/* NETDEV_TX_BUSY - we need to requeue */
			/* Release the driver */
			if (!nolock) { 
				dev->xmit_lock_owner = -1;
				spin_unlock(&dev->xmit_lock);
			} 
			spin_lock(&dev->queue_lock);
			q = dev->qdisc;
		}

		/* Device kicked us out :(
		   This is possible in three cases:

		   0. driver is locked
		   1. fastroute is enabled
		   2. device cannot determine busy state
		      before start of transmission (f.e. dialout)
		   3. device is buggy (ppp)
		 */

requeue:
		q->ops->requeue(skb, q);
		netif_schedule(dev);
		return 1;
	}
182
	BUG_ON((int) q->q.qlen < 0);
L
Linus Torvalds 已提交
183 184 185 186 187 188 189 190 191 192 193 194 195
	return q->q.qlen;
}

static void dev_watchdog(unsigned long arg)
{
	struct net_device *dev = (struct net_device *)arg;

	spin_lock(&dev->xmit_lock);
	if (dev->qdisc != &noop_qdisc) {
		if (netif_device_present(dev) &&
		    netif_running(dev) &&
		    netif_carrier_ok(dev)) {
			if (netif_queue_stopped(dev) &&
196 197 198 199
			    time_after(jiffies, dev->trans_start + dev->watchdog_timeo)) {

				printk(KERN_INFO "NETDEV WATCHDOG: %s: transmit timed out\n",
				       dev->name);
L
Linus Torvalds 已提交
200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
				dev->tx_timeout(dev);
			}
			if (!mod_timer(&dev->watchdog_timer, jiffies + dev->watchdog_timeo))
				dev_hold(dev);
		}
	}
	spin_unlock(&dev->xmit_lock);

	dev_put(dev);
}

static void dev_watchdog_init(struct net_device *dev)
{
	init_timer(&dev->watchdog_timer);
	dev->watchdog_timer.data = (unsigned long)dev;
	dev->watchdog_timer.function = dev_watchdog;
}

void __netdev_watchdog_up(struct net_device *dev)
{
	if (dev->tx_timeout) {
		if (dev->watchdog_timeo <= 0)
			dev->watchdog_timeo = 5*HZ;
		if (!mod_timer(&dev->watchdog_timer, jiffies + dev->watchdog_timeo))
			dev_hold(dev);
	}
}

static void dev_watchdog_up(struct net_device *dev)
{
	spin_lock_bh(&dev->xmit_lock);
	__netdev_watchdog_up(dev);
	spin_unlock_bh(&dev->xmit_lock);
}

static void dev_watchdog_down(struct net_device *dev)
{
	spin_lock_bh(&dev->xmit_lock);
	if (del_timer(&dev->watchdog_timer))
239
		dev_put(dev);
L
Linus Torvalds 已提交
240 241 242
	spin_unlock_bh(&dev->xmit_lock);
}

243 244 245 246 247 248 249 250 251 252 253 254 255 256
void netif_carrier_on(struct net_device *dev)
{
	if (test_and_clear_bit(__LINK_STATE_NOCARRIER, &dev->state))
		linkwatch_fire_event(dev);
	if (netif_running(dev))
		__netdev_watchdog_up(dev);
}

void netif_carrier_off(struct net_device *dev)
{
	if (!test_and_set_bit(__LINK_STATE_NOCARRIER, &dev->state))
		linkwatch_fire_event(dev);
}

L
Linus Torvalds 已提交
257 258 259 260 261
/* "NOOP" scheduler: the best scheduler, recommended for all interfaces
   under all circumstances. It is difficult to invent anything faster or
   cheaper.
 */

262
static int noop_enqueue(struct sk_buff *skb, struct Qdisc * qdisc)
L
Linus Torvalds 已提交
263 264 265 266 267
{
	kfree_skb(skb);
	return NET_XMIT_CN;
}

268
static struct sk_buff *noop_dequeue(struct Qdisc * qdisc)
L
Linus Torvalds 已提交
269 270 271 272
{
	return NULL;
}

273
static int noop_requeue(struct sk_buff *skb, struct Qdisc* qdisc)
L
Linus Torvalds 已提交
274 275
{
	if (net_ratelimit())
276 277
		printk(KERN_DEBUG "%s deferred output. It is buggy.\n",
		       skb->dev->name);
L
Linus Torvalds 已提交
278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
	kfree_skb(skb);
	return NET_XMIT_CN;
}

struct Qdisc_ops noop_qdisc_ops = {
	.id		=	"noop",
	.priv_size	=	0,
	.enqueue	=	noop_enqueue,
	.dequeue	=	noop_dequeue,
	.requeue	=	noop_requeue,
	.owner		=	THIS_MODULE,
};

struct Qdisc noop_qdisc = {
	.enqueue	=	noop_enqueue,
	.dequeue	=	noop_dequeue,
	.flags		=	TCQ_F_BUILTIN,
	.ops		=	&noop_qdisc_ops,	
	.list		=	LIST_HEAD_INIT(noop_qdisc.list),
};

static struct Qdisc_ops noqueue_qdisc_ops = {
	.id		=	"noqueue",
	.priv_size	=	0,
	.enqueue	=	noop_enqueue,
	.dequeue	=	noop_dequeue,
	.requeue	=	noop_requeue,
	.owner		=	THIS_MODULE,
};

static struct Qdisc noqueue_qdisc = {
	.enqueue	=	NULL,
	.dequeue	=	noop_dequeue,
	.flags		=	TCQ_F_BUILTIN,
	.ops		=	&noqueue_qdisc_ops,
	.list		=	LIST_HEAD_INIT(noqueue_qdisc.list),
};


static const u8 prio2band[TC_PRIO_MAX+1] =
	{ 1, 2, 2, 2, 1, 2, 0, 0 , 1, 1, 1, 1, 1, 1, 1, 1 };

/* 3-band FIFO queue: old style, but should be a bit faster than
   generic prio+fifo combination.
 */

324 325
#define PFIFO_FAST_BANDS 3

326 327
static inline struct sk_buff_head *prio2list(struct sk_buff *skb,
					     struct Qdisc *qdisc)
L
Linus Torvalds 已提交
328 329
{
	struct sk_buff_head *list = qdisc_priv(qdisc);
330 331
	return list + prio2band[skb->priority & TC_PRIO_MAX];
}
L
Linus Torvalds 已提交
332

333
static int pfifo_fast_enqueue(struct sk_buff *skb, struct Qdisc* qdisc)
334 335
{
	struct sk_buff_head *list = prio2list(skb, qdisc);
L
Linus Torvalds 已提交
336

337
	if (skb_queue_len(list) < qdisc->dev->tx_queue_len) {
L
Linus Torvalds 已提交
338
		qdisc->q.qlen++;
339
		return __qdisc_enqueue_tail(skb, qdisc, list);
L
Linus Torvalds 已提交
340
	}
341 342

	return qdisc_drop(skb, qdisc);
L
Linus Torvalds 已提交
343 344
}

345
static struct sk_buff *pfifo_fast_dequeue(struct Qdisc* qdisc)
L
Linus Torvalds 已提交
346 347 348 349
{
	int prio;
	struct sk_buff_head *list = qdisc_priv(qdisc);

350 351
	for (prio = 0; prio < PFIFO_FAST_BANDS; prio++) {
		if (!skb_queue_empty(list + prio)) {
L
Linus Torvalds 已提交
352
			qdisc->q.qlen--;
353
			return __qdisc_dequeue_head(qdisc, list + prio);
L
Linus Torvalds 已提交
354 355
		}
	}
356

L
Linus Torvalds 已提交
357 358 359
	return NULL;
}

360
static int pfifo_fast_requeue(struct sk_buff *skb, struct Qdisc* qdisc)
L
Linus Torvalds 已提交
361 362
{
	qdisc->q.qlen++;
363
	return __qdisc_requeue(skb, qdisc, prio2list(skb, qdisc));
L
Linus Torvalds 已提交
364 365
}

366
static void pfifo_fast_reset(struct Qdisc* qdisc)
L
Linus Torvalds 已提交
367 368 369 370
{
	int prio;
	struct sk_buff_head *list = qdisc_priv(qdisc);

371
	for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
372 373 374
		__qdisc_reset_queue(qdisc, list + prio);

	qdisc->qstats.backlog = 0;
L
Linus Torvalds 已提交
375 376 377 378 379
	qdisc->q.qlen = 0;
}

static int pfifo_fast_dump(struct Qdisc *qdisc, struct sk_buff *skb)
{
380
	struct tc_prio_qopt opt = { .bands = PFIFO_FAST_BANDS };
L
Linus Torvalds 已提交
381 382 383 384 385 386 387 388 389 390 391

	memcpy(&opt.priomap, prio2band, TC_PRIO_MAX+1);
	RTA_PUT(skb, TCA_OPTIONS, sizeof(opt), &opt);
	return skb->len;

rtattr_failure:
	return -1;
}

static int pfifo_fast_init(struct Qdisc *qdisc, struct rtattr *opt)
{
392
	int prio;
L
Linus Torvalds 已提交
393 394
	struct sk_buff_head *list = qdisc_priv(qdisc);

395 396
	for (prio = 0; prio < PFIFO_FAST_BANDS; prio++)
		skb_queue_head_init(list + prio);
L
Linus Torvalds 已提交
397 398 399 400 401 402

	return 0;
}

static struct Qdisc_ops pfifo_fast_ops = {
	.id		=	"pfifo_fast",
403
	.priv_size	=	PFIFO_FAST_BANDS * sizeof(struct sk_buff_head),
L
Linus Torvalds 已提交
404 405 406 407 408 409 410 411 412
	.enqueue	=	pfifo_fast_enqueue,
	.dequeue	=	pfifo_fast_dequeue,
	.requeue	=	pfifo_fast_requeue,
	.init		=	pfifo_fast_init,
	.reset		=	pfifo_fast_reset,
	.dump		=	pfifo_fast_dump,
	.owner		=	THIS_MODULE,
};

413
struct Qdisc *qdisc_alloc(struct net_device *dev, struct Qdisc_ops *ops)
L
Linus Torvalds 已提交
414 415 416
{
	void *p;
	struct Qdisc *sch;
417 418
	unsigned int size;
	int err = -ENOBUFS;
L
Linus Torvalds 已提交
419 420

	/* ensure that the Qdisc and the private data are 32-byte aligned */
421 422
	size = QDISC_ALIGN(sizeof(*sch));
	size += ops->priv_size + (QDISC_ALIGNTO - 1);
L
Linus Torvalds 已提交
423 424 425

	p = kmalloc(size, GFP_KERNEL);
	if (!p)
426
		goto errout;
L
Linus Torvalds 已提交
427
	memset(p, 0, size);
428 429
	sch = (struct Qdisc *) QDISC_ALIGN((unsigned long) p);
	sch->padded = (char *) sch - (char *) p;
L
Linus Torvalds 已提交
430 431 432 433 434 435 436 437 438 439

	INIT_LIST_HEAD(&sch->list);
	skb_queue_head_init(&sch->q);
	sch->ops = ops;
	sch->enqueue = ops->enqueue;
	sch->dequeue = ops->dequeue;
	sch->dev = dev;
	dev_hold(dev);
	sch->stats_lock = &dev->queue_lock;
	atomic_set(&sch->refcnt, 1);
440 441 442 443 444 445 446 447 448 449 450 451 452 453

	return sch;
errout:
	return ERR_PTR(-err);
}

struct Qdisc * qdisc_create_dflt(struct net_device *dev, struct Qdisc_ops *ops)
{
	struct Qdisc *sch;
	
	sch = qdisc_alloc(dev, ops);
	if (IS_ERR(sch))
		goto errout;

L
Linus Torvalds 已提交
454 455 456
	if (!ops->init || ops->init(sch, NULL) == 0)
		return sch;

457
	qdisc_destroy(sch);
458
errout:
L
Linus Torvalds 已提交
459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554
	return NULL;
}

/* Under dev->queue_lock and BH! */

void qdisc_reset(struct Qdisc *qdisc)
{
	struct Qdisc_ops *ops = qdisc->ops;

	if (ops->reset)
		ops->reset(qdisc);
}

/* this is the rcu callback function to clean up a qdisc when there 
 * are no further references to it */

static void __qdisc_destroy(struct rcu_head *head)
{
	struct Qdisc *qdisc = container_of(head, struct Qdisc, q_rcu);
	struct Qdisc_ops  *ops = qdisc->ops;

#ifdef CONFIG_NET_ESTIMATOR
	gen_kill_estimator(&qdisc->bstats, &qdisc->rate_est);
#endif
	write_lock(&qdisc_tree_lock);
	if (ops->reset)
		ops->reset(qdisc);
	if (ops->destroy)
		ops->destroy(qdisc);
	write_unlock(&qdisc_tree_lock);
	module_put(ops->owner);

	dev_put(qdisc->dev);
	kfree((char *) qdisc - qdisc->padded);
}

/* Under dev->queue_lock and BH! */

void qdisc_destroy(struct Qdisc *qdisc)
{
	struct list_head cql = LIST_HEAD_INIT(cql);
	struct Qdisc *cq, *q, *n;

	if (qdisc->flags & TCQ_F_BUILTIN ||
		!atomic_dec_and_test(&qdisc->refcnt))
		return;

	if (!list_empty(&qdisc->list)) {
		if (qdisc->ops->cl_ops == NULL)
			list_del(&qdisc->list);
		else
			list_move(&qdisc->list, &cql);
	}

	/* unlink inner qdiscs from dev->qdisc_list immediately */
	list_for_each_entry(cq, &cql, list)
		list_for_each_entry_safe(q, n, &qdisc->dev->qdisc_list, list)
			if (TC_H_MAJ(q->parent) == TC_H_MAJ(cq->handle)) {
				if (q->ops->cl_ops == NULL)
					list_del_init(&q->list);
				else
					list_move_tail(&q->list, &cql);
			}
	list_for_each_entry_safe(cq, n, &cql, list)
		list_del_init(&cq->list);

	call_rcu(&qdisc->q_rcu, __qdisc_destroy);
}

void dev_activate(struct net_device *dev)
{
	/* No queueing discipline is attached to device;
	   create default one i.e. pfifo_fast for devices,
	   which need queueing and noqueue_qdisc for
	   virtual interfaces
	 */

	if (dev->qdisc_sleeping == &noop_qdisc) {
		struct Qdisc *qdisc;
		if (dev->tx_queue_len) {
			qdisc = qdisc_create_dflt(dev, &pfifo_fast_ops);
			if (qdisc == NULL) {
				printk(KERN_INFO "%s: activation failed\n", dev->name);
				return;
			}
			write_lock_bh(&qdisc_tree_lock);
			list_add_tail(&qdisc->list, &dev->qdisc_list);
			write_unlock_bh(&qdisc_tree_lock);
		} else {
			qdisc =  &noqueue_qdisc;
		}
		write_lock_bh(&qdisc_tree_lock);
		dev->qdisc_sleeping = qdisc;
		write_unlock_bh(&qdisc_tree_lock);
	}

555 556 557 558
	if (!netif_carrier_ok(dev))
		/* Delay activation until next carrier-on event */
		return;

L
Linus Torvalds 已提交
559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618
	spin_lock_bh(&dev->queue_lock);
	rcu_assign_pointer(dev->qdisc, dev->qdisc_sleeping);
	if (dev->qdisc != &noqueue_qdisc) {
		dev->trans_start = jiffies;
		dev_watchdog_up(dev);
	}
	spin_unlock_bh(&dev->queue_lock);
}

void dev_deactivate(struct net_device *dev)
{
	struct Qdisc *qdisc;

	spin_lock_bh(&dev->queue_lock);
	qdisc = dev->qdisc;
	dev->qdisc = &noop_qdisc;

	qdisc_reset(qdisc);

	spin_unlock_bh(&dev->queue_lock);

	dev_watchdog_down(dev);

	while (test_bit(__LINK_STATE_SCHED, &dev->state))
		yield();

	spin_unlock_wait(&dev->xmit_lock);
}

void dev_init_scheduler(struct net_device *dev)
{
	qdisc_lock_tree(dev);
	dev->qdisc = &noop_qdisc;
	dev->qdisc_sleeping = &noop_qdisc;
	INIT_LIST_HEAD(&dev->qdisc_list);
	qdisc_unlock_tree(dev);

	dev_watchdog_init(dev);
}

void dev_shutdown(struct net_device *dev)
{
	struct Qdisc *qdisc;

	qdisc_lock_tree(dev);
	qdisc = dev->qdisc_sleeping;
	dev->qdisc = &noop_qdisc;
	dev->qdisc_sleeping = &noop_qdisc;
	qdisc_destroy(qdisc);
#if defined(CONFIG_NET_SCH_INGRESS) || defined(CONFIG_NET_SCH_INGRESS_MODULE)
        if ((qdisc = dev->qdisc_ingress) != NULL) {
		dev->qdisc_ingress = NULL;
		qdisc_destroy(qdisc);
        }
#endif
	BUG_TRAP(!timer_pending(&dev->watchdog_timer));
	qdisc_unlock_tree(dev);
}

EXPORT_SYMBOL(__netdev_watchdog_up);
619 620
EXPORT_SYMBOL(netif_carrier_on);
EXPORT_SYMBOL(netif_carrier_off);
L
Linus Torvalds 已提交
621 622 623
EXPORT_SYMBOL(noop_qdisc);
EXPORT_SYMBOL(noop_qdisc_ops);
EXPORT_SYMBOL(qdisc_create_dflt);
624
EXPORT_SYMBOL(qdisc_alloc);
L
Linus Torvalds 已提交
625 626 627 628 629
EXPORT_SYMBOL(qdisc_destroy);
EXPORT_SYMBOL(qdisc_reset);
EXPORT_SYMBOL(qdisc_restart);
EXPORT_SYMBOL(qdisc_lock_tree);
EXPORT_SYMBOL(qdisc_unlock_tree);