sch_api.c 29.4 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
/*
 * net/sched/sch_api.c	Packet scheduler API.
 *
 *		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>
 *
 * Fixes:
 *
 * Rani Assaf <rani@magic.metawire.com> :980802: JIFFIES and CPU clock sources are repaired.
 * Eduardo J. Blanco <ejbs@netlabs.com.uy> :990222: kmod support
 * Jamal Hadi Salim <hadi@nortelnetworks.com>: 990601: ingress support
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/kernel.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/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/kmod.h>
#include <linux/list.h>
#include <linux/bitops.h>
37
#include <linux/hrtimer.h>
L
Linus Torvalds 已提交
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193

#include <net/sock.h>
#include <net/pkt_sched.h>

#include <asm/processor.h>
#include <asm/uaccess.h>
#include <asm/system.h>

static int qdisc_notify(struct sk_buff *oskb, struct nlmsghdr *n, u32 clid,
			struct Qdisc *old, struct Qdisc *new);
static int tclass_notify(struct sk_buff *oskb, struct nlmsghdr *n,
			 struct Qdisc *q, unsigned long cl, int event);

/*

   Short review.
   -------------

   This file consists of two interrelated parts:

   1. queueing disciplines manager frontend.
   2. traffic classes manager frontend.

   Generally, queueing discipline ("qdisc") is a black box,
   which is able to enqueue packets and to dequeue them (when
   device is ready to send something) in order and at times
   determined by algorithm hidden in it.

   qdisc's are divided to two categories:
   - "queues", which have no internal structure visible from outside.
   - "schedulers", which split all the packets to "traffic classes",
     using "packet classifiers" (look at cls_api.c)

   In turn, classes may have child qdiscs (as rule, queues)
   attached to them etc. etc. etc.

   The goal of the routines in this file is to translate
   information supplied by user in the form of handles
   to more intelligible for kernel form, to make some sanity
   checks and part of work, which is common to all qdiscs
   and to provide rtnetlink notifications.

   All real intelligent work is done inside qdisc modules.



   Every discipline has two major routines: enqueue and dequeue.

   ---dequeue

   dequeue usually returns a skb to send. It is allowed to return NULL,
   but it does not mean that queue is empty, it just means that
   discipline does not want to send anything this time.
   Queue is really empty if q->q.qlen == 0.
   For complicated disciplines with multiple queues q->q is not
   real packet queue, but however q->q.qlen must be valid.

   ---enqueue

   enqueue returns 0, if packet was enqueued successfully.
   If packet (this one or another one) was dropped, it returns
   not zero error code.
   NET_XMIT_DROP 	- this packet dropped
     Expected action: do not backoff, but wait until queue will clear.
   NET_XMIT_CN	 	- probably this packet enqueued, but another one dropped.
     Expected action: backoff or ignore
   NET_XMIT_POLICED	- dropped by police.
     Expected action: backoff or error to real-time apps.

   Auxiliary routines:

   ---requeue

   requeues once dequeued packet. It is used for non-standard or
   just buggy devices, which can defer output even if dev->tbusy=0.

   ---reset

   returns qdisc to initial state: purge all buffers, clear all
   timers, counters (except for statistics) etc.

   ---init

   initializes newly created qdisc.

   ---destroy

   destroys resources allocated by init and during lifetime of qdisc.

   ---change

   changes qdisc parameters.
 */

/* Protects list of registered TC modules. It is pure SMP lock. */
static DEFINE_RWLOCK(qdisc_mod_lock);


/************************************************
 *	Queueing disciplines manipulation.	*
 ************************************************/


/* The list of all installed queueing disciplines. */

static struct Qdisc_ops *qdisc_base;

/* Register/uregister queueing discipline */

int register_qdisc(struct Qdisc_ops *qops)
{
	struct Qdisc_ops *q, **qp;
	int rc = -EEXIST;

	write_lock(&qdisc_mod_lock);
	for (qp = &qdisc_base; (q = *qp) != NULL; qp = &q->next)
		if (!strcmp(qops->id, q->id))
			goto out;

	if (qops->enqueue == NULL)
		qops->enqueue = noop_qdisc_ops.enqueue;
	if (qops->requeue == NULL)
		qops->requeue = noop_qdisc_ops.requeue;
	if (qops->dequeue == NULL)
		qops->dequeue = noop_qdisc_ops.dequeue;

	qops->next = NULL;
	*qp = qops;
	rc = 0;
out:
	write_unlock(&qdisc_mod_lock);
	return rc;
}

int unregister_qdisc(struct Qdisc_ops *qops)
{
	struct Qdisc_ops *q, **qp;
	int err = -ENOENT;

	write_lock(&qdisc_mod_lock);
	for (qp = &qdisc_base; (q=*qp)!=NULL; qp = &q->next)
		if (q == qops)
			break;
	if (q) {
		*qp = q->next;
		q->next = NULL;
		err = 0;
	}
	write_unlock(&qdisc_mod_lock);
	return err;
}

/* We know handle. Find qdisc among all qdisc's attached to device
   (root qdisc, all its children, children of children etc.)
 */

194
static struct Qdisc *__qdisc_lookup(struct net_device *dev, u32 handle)
L
Linus Torvalds 已提交
195 196 197 198
{
	struct Qdisc *q;

	list_for_each_entry(q, &dev->qdisc_list, list) {
199
		if (q->handle == handle)
L
Linus Torvalds 已提交
200 201 202 203 204
			return q;
	}
	return NULL;
}

205 206 207 208 209 210 211 212 213 214
struct Qdisc *qdisc_lookup(struct net_device *dev, u32 handle)
{
	struct Qdisc *q;

	read_lock(&qdisc_tree_lock);
	q = __qdisc_lookup(dev, handle);
	read_unlock(&qdisc_tree_lock);
	return q;
}

L
Linus Torvalds 已提交
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294
static struct Qdisc *qdisc_leaf(struct Qdisc *p, u32 classid)
{
	unsigned long cl;
	struct Qdisc *leaf;
	struct Qdisc_class_ops *cops = p->ops->cl_ops;

	if (cops == NULL)
		return NULL;
	cl = cops->get(p, classid);

	if (cl == 0)
		return NULL;
	leaf = cops->leaf(p, cl);
	cops->put(p, cl);
	return leaf;
}

/* Find queueing discipline by name */

static struct Qdisc_ops *qdisc_lookup_ops(struct rtattr *kind)
{
	struct Qdisc_ops *q = NULL;

	if (kind) {
		read_lock(&qdisc_mod_lock);
		for (q = qdisc_base; q; q = q->next) {
			if (rtattr_strcmp(kind, q->id) == 0) {
				if (!try_module_get(q->owner))
					q = NULL;
				break;
			}
		}
		read_unlock(&qdisc_mod_lock);
	}
	return q;
}

static struct qdisc_rate_table *qdisc_rtab_list;

struct qdisc_rate_table *qdisc_get_rtab(struct tc_ratespec *r, struct rtattr *tab)
{
	struct qdisc_rate_table *rtab;

	for (rtab = qdisc_rtab_list; rtab; rtab = rtab->next) {
		if (memcmp(&rtab->rate, r, sizeof(struct tc_ratespec)) == 0) {
			rtab->refcnt++;
			return rtab;
		}
	}

	if (tab == NULL || r->rate == 0 || r->cell_log == 0 || RTA_PAYLOAD(tab) != 1024)
		return NULL;

	rtab = kmalloc(sizeof(*rtab), GFP_KERNEL);
	if (rtab) {
		rtab->rate = *r;
		rtab->refcnt = 1;
		memcpy(rtab->data, RTA_DATA(tab), 1024);
		rtab->next = qdisc_rtab_list;
		qdisc_rtab_list = rtab;
	}
	return rtab;
}

void qdisc_put_rtab(struct qdisc_rate_table *tab)
{
	struct qdisc_rate_table *rtab, **rtabp;

	if (!tab || --tab->refcnt)
		return;

	for (rtabp = &qdisc_rtab_list; (rtab=*rtabp) != NULL; rtabp = &rtab->next) {
		if (rtab == tab) {
			*rtabp = rtab->next;
			kfree(rtab);
			return;
		}
	}
}

295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
static enum hrtimer_restart qdisc_watchdog(struct hrtimer *timer)
{
	struct qdisc_watchdog *wd = container_of(timer, struct qdisc_watchdog,
						 timer);

	wd->qdisc->flags &= ~TCQ_F_THROTTLED;
	netif_schedule(wd->qdisc->dev);
	return HRTIMER_NORESTART;
}

void qdisc_watchdog_init(struct qdisc_watchdog *wd, struct Qdisc *qdisc)
{
	hrtimer_init(&wd->timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
	wd->timer.function = qdisc_watchdog;
	wd->qdisc = qdisc;
}
EXPORT_SYMBOL(qdisc_watchdog_init);

void qdisc_watchdog_schedule(struct qdisc_watchdog *wd, psched_time_t expires)
{
	ktime_t time;

	wd->qdisc->flags |= TCQ_F_THROTTLED;
	time = ktime_set(0, 0);
	time = ktime_add_ns(time, PSCHED_US2NS(expires));
	hrtimer_start(&wd->timer, time, HRTIMER_MODE_ABS);
}
EXPORT_SYMBOL(qdisc_watchdog_schedule);

void qdisc_watchdog_cancel(struct qdisc_watchdog *wd)
{
	hrtimer_cancel(&wd->timer);
	wd->qdisc->flags &= ~TCQ_F_THROTTLED;
}
EXPORT_SYMBOL(qdisc_watchdog_cancel);
L
Linus Torvalds 已提交
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391

/* Allocate an unique handle from space managed by kernel */

static u32 qdisc_alloc_handle(struct net_device *dev)
{
	int i = 0x10000;
	static u32 autohandle = TC_H_MAKE(0x80000000U, 0);

	do {
		autohandle += TC_H_MAKE(0x10000U, 0);
		if (autohandle == TC_H_MAKE(TC_H_ROOT, 0))
			autohandle = TC_H_MAKE(0x80000000U, 0);
	} while	(qdisc_lookup(dev, autohandle) && --i > 0);

	return i>0 ? autohandle : 0;
}

/* Attach toplevel qdisc to device dev */

static struct Qdisc *
dev_graft_qdisc(struct net_device *dev, struct Qdisc *qdisc)
{
	struct Qdisc *oqdisc;

	if (dev->flags & IFF_UP)
		dev_deactivate(dev);

	qdisc_lock_tree(dev);
	if (qdisc && qdisc->flags&TCQ_F_INGRESS) {
		oqdisc = dev->qdisc_ingress;
		/* Prune old scheduler */
		if (oqdisc && atomic_read(&oqdisc->refcnt) <= 1) {
			/* delete */
			qdisc_reset(oqdisc);
			dev->qdisc_ingress = NULL;
		} else {  /* new */
			dev->qdisc_ingress = qdisc;
		}

	} else {

		oqdisc = dev->qdisc_sleeping;

		/* Prune old scheduler */
		if (oqdisc && atomic_read(&oqdisc->refcnt) <= 1)
			qdisc_reset(oqdisc);

		/* ... and graft new one */
		if (qdisc == NULL)
			qdisc = &noop_qdisc;
		dev->qdisc_sleeping = qdisc;
		dev->qdisc = &noop_qdisc;
	}

	qdisc_unlock_tree(dev);

	if (dev->flags & IFF_UP)
		dev_activate(dev);

	return oqdisc;
}

392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
void qdisc_tree_decrease_qlen(struct Qdisc *sch, unsigned int n)
{
	struct Qdisc_class_ops *cops;
	unsigned long cl;
	u32 parentid;

	if (n == 0)
		return;
	while ((parentid = sch->parent)) {
		sch = __qdisc_lookup(sch->dev, TC_H_MAJ(parentid));
		cops = sch->ops->cl_ops;
		if (cops->qlen_notify) {
			cl = cops->get(sch, parentid);
			cops->qlen_notify(sch, cl);
			cops->put(sch, cl);
		}
		sch->q.qlen -= n;
	}
}
EXPORT_SYMBOL(qdisc_tree_decrease_qlen);
L
Linus Torvalds 已提交
412 413 414 415 416 417 418 419 420 421 422 423 424 425 426

/* Graft qdisc "new" to class "classid" of qdisc "parent" or
   to device "dev".

   Old qdisc is not destroyed but returned in *old.
 */

static int qdisc_graft(struct net_device *dev, struct Qdisc *parent,
		       u32 classid,
		       struct Qdisc *new, struct Qdisc **old)
{
	int err = 0;
	struct Qdisc *q = *old;


427
	if (parent == NULL) {
L
Linus Torvalds 已提交
428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493
		if (q && q->flags&TCQ_F_INGRESS) {
			*old = dev_graft_qdisc(dev, q);
		} else {
			*old = dev_graft_qdisc(dev, new);
		}
	} else {
		struct Qdisc_class_ops *cops = parent->ops->cl_ops;

		err = -EINVAL;

		if (cops) {
			unsigned long cl = cops->get(parent, classid);
			if (cl) {
				err = cops->graft(parent, cl, new, old);
				if (new)
					new->parent = classid;
				cops->put(parent, cl);
			}
		}
	}
	return err;
}

/*
   Allocate and initialize new qdisc.

   Parameters are passed via opt.
 */

static struct Qdisc *
qdisc_create(struct net_device *dev, u32 handle, struct rtattr **tca, int *errp)
{
	int err;
	struct rtattr *kind = tca[TCA_KIND-1];
	struct Qdisc *sch;
	struct Qdisc_ops *ops;

	ops = qdisc_lookup_ops(kind);
#ifdef CONFIG_KMOD
	if (ops == NULL && kind != NULL) {
		char name[IFNAMSIZ];
		if (rtattr_strlcpy(name, kind, IFNAMSIZ) < IFNAMSIZ) {
			/* We dropped the RTNL semaphore in order to
			 * perform the module load.  So, even if we
			 * succeeded in loading the module we have to
			 * tell the caller to replay the request.  We
			 * indicate this using -EAGAIN.
			 * We replay the request because the device may
			 * go away in the mean time.
			 */
			rtnl_unlock();
			request_module("sch_%s", name);
			rtnl_lock();
			ops = qdisc_lookup_ops(kind);
			if (ops != NULL) {
				/* We will try again qdisc_lookup_ops,
				 * so don't keep a reference.
				 */
				module_put(ops->owner);
				err = -EAGAIN;
				goto err_out;
			}
		}
	}
#endif

494
	err = -ENOENT;
L
Linus Torvalds 已提交
495 496 497
	if (ops == NULL)
		goto err_out;

498 499 500
	sch = qdisc_alloc(dev, ops);
	if (IS_ERR(sch)) {
		err = PTR_ERR(sch);
L
Linus Torvalds 已提交
501
		goto err_out2;
502
	}
L
Linus Torvalds 已提交
503

504
	if (handle == TC_H_INGRESS) {
L
Linus Torvalds 已提交
505
		sch->flags |= TCQ_F_INGRESS;
506 507
		handle = TC_H_MAKE(TC_H_INGRESS, 0);
	} else if (handle == 0) {
L
Linus Torvalds 已提交
508 509 510 511 512 513
		handle = qdisc_alloc_handle(dev);
		err = -ENOMEM;
		if (handle == 0)
			goto err_out3;
	}

514
	sch->handle = handle;
L
Linus Torvalds 已提交
515 516

	if (!ops->init || (err = ops->init(sch, tca[TCA_OPTIONS-1])) == 0) {
517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533
#ifdef CONFIG_NET_ESTIMATOR
		if (tca[TCA_RATE-1]) {
			err = gen_new_estimator(&sch->bstats, &sch->rate_est,
						sch->stats_lock,
						tca[TCA_RATE-1]);
			if (err) {
				/*
				 * Any broken qdiscs that would require
				 * a ops->reset() here? The qdisc was never
				 * in action so it shouldn't be necessary.
				 */
				if (ops->destroy)
					ops->destroy(sch);
				goto err_out3;
			}
		}
#endif
L
Linus Torvalds 已提交
534 535 536 537 538 539 540 541
		qdisc_lock_tree(dev);
		list_add_tail(&sch->list, &dev->qdisc_list);
		qdisc_unlock_tree(dev);

		return sch;
	}
err_out3:
	dev_put(dev);
542
	kfree((char *) sch - sch->padded);
L
Linus Torvalds 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633
err_out2:
	module_put(ops->owner);
err_out:
	*errp = err;
	return NULL;
}

static int qdisc_change(struct Qdisc *sch, struct rtattr **tca)
{
	if (tca[TCA_OPTIONS-1]) {
		int err;

		if (sch->ops->change == NULL)
			return -EINVAL;
		err = sch->ops->change(sch, tca[TCA_OPTIONS-1]);
		if (err)
			return err;
	}
#ifdef CONFIG_NET_ESTIMATOR
	if (tca[TCA_RATE-1])
		gen_replace_estimator(&sch->bstats, &sch->rate_est,
			sch->stats_lock, tca[TCA_RATE-1]);
#endif
	return 0;
}

struct check_loop_arg
{
	struct qdisc_walker 	w;
	struct Qdisc		*p;
	int			depth;
};

static int check_loop_fn(struct Qdisc *q, unsigned long cl, struct qdisc_walker *w);

static int check_loop(struct Qdisc *q, struct Qdisc *p, int depth)
{
	struct check_loop_arg	arg;

	if (q->ops->cl_ops == NULL)
		return 0;

	arg.w.stop = arg.w.skip = arg.w.count = 0;
	arg.w.fn = check_loop_fn;
	arg.depth = depth;
	arg.p = p;
	q->ops->cl_ops->walk(q, &arg.w);
	return arg.w.stop ? -ELOOP : 0;
}

static int
check_loop_fn(struct Qdisc *q, unsigned long cl, struct qdisc_walker *w)
{
	struct Qdisc *leaf;
	struct Qdisc_class_ops *cops = q->ops->cl_ops;
	struct check_loop_arg *arg = (struct check_loop_arg *)w;

	leaf = cops->leaf(q, cl);
	if (leaf) {
		if (leaf == arg->p || arg->depth > 7)
			return -ELOOP;
		return check_loop(leaf, arg->p, arg->depth + 1);
	}
	return 0;
}

/*
 * Delete/get qdisc.
 */

static int tc_get_qdisc(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
{
	struct tcmsg *tcm = NLMSG_DATA(n);
	struct rtattr **tca = arg;
	struct net_device *dev;
	u32 clid = tcm->tcm_parent;
	struct Qdisc *q = NULL;
	struct Qdisc *p = NULL;
	int err;

	if ((dev = __dev_get_by_index(tcm->tcm_ifindex)) == NULL)
		return -ENODEV;

	if (clid) {
		if (clid != TC_H_ROOT) {
			if (TC_H_MAJ(clid) != TC_H_MAJ(TC_H_INGRESS)) {
				if ((p = qdisc_lookup(dev, TC_H_MAJ(clid))) == NULL)
					return -ENOENT;
				q = qdisc_leaf(p, clid);
			} else { /* ingress */
				q = dev->qdisc_ingress;
634
			}
L
Linus Torvalds 已提交
635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780
		} else {
			q = dev->qdisc_sleeping;
		}
		if (!q)
			return -ENOENT;

		if (tcm->tcm_handle && q->handle != tcm->tcm_handle)
			return -EINVAL;
	} else {
		if ((q = qdisc_lookup(dev, tcm->tcm_handle)) == NULL)
			return -ENOENT;
	}

	if (tca[TCA_KIND-1] && rtattr_strcmp(tca[TCA_KIND-1], q->ops->id))
		return -EINVAL;

	if (n->nlmsg_type == RTM_DELQDISC) {
		if (!clid)
			return -EINVAL;
		if (q->handle == 0)
			return -ENOENT;
		if ((err = qdisc_graft(dev, p, clid, NULL, &q)) != 0)
			return err;
		if (q) {
			qdisc_notify(skb, n, clid, q, NULL);
			spin_lock_bh(&dev->queue_lock);
			qdisc_destroy(q);
			spin_unlock_bh(&dev->queue_lock);
		}
	} else {
		qdisc_notify(skb, n, clid, NULL, q);
	}
	return 0;
}

/*
   Create/change qdisc.
 */

static int tc_modify_qdisc(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
{
	struct tcmsg *tcm;
	struct rtattr **tca;
	struct net_device *dev;
	u32 clid;
	struct Qdisc *q, *p;
	int err;

replay:
	/* Reinit, just in case something touches this. */
	tcm = NLMSG_DATA(n);
	tca = arg;
	clid = tcm->tcm_parent;
	q = p = NULL;

	if ((dev = __dev_get_by_index(tcm->tcm_ifindex)) == NULL)
		return -ENODEV;

	if (clid) {
		if (clid != TC_H_ROOT) {
			if (clid != TC_H_INGRESS) {
				if ((p = qdisc_lookup(dev, TC_H_MAJ(clid))) == NULL)
					return -ENOENT;
				q = qdisc_leaf(p, clid);
			} else { /*ingress */
				q = dev->qdisc_ingress;
			}
		} else {
			q = dev->qdisc_sleeping;
		}

		/* It may be default qdisc, ignore it */
		if (q && q->handle == 0)
			q = NULL;

		if (!q || !tcm->tcm_handle || q->handle != tcm->tcm_handle) {
			if (tcm->tcm_handle) {
				if (q && !(n->nlmsg_flags&NLM_F_REPLACE))
					return -EEXIST;
				if (TC_H_MIN(tcm->tcm_handle))
					return -EINVAL;
				if ((q = qdisc_lookup(dev, tcm->tcm_handle)) == NULL)
					goto create_n_graft;
				if (n->nlmsg_flags&NLM_F_EXCL)
					return -EEXIST;
				if (tca[TCA_KIND-1] && rtattr_strcmp(tca[TCA_KIND-1], q->ops->id))
					return -EINVAL;
				if (q == p ||
				    (p && check_loop(q, p, 0)))
					return -ELOOP;
				atomic_inc(&q->refcnt);
				goto graft;
			} else {
				if (q == NULL)
					goto create_n_graft;

				/* This magic test requires explanation.
				 *
				 *   We know, that some child q is already
				 *   attached to this parent and have choice:
				 *   either to change it or to create/graft new one.
				 *
				 *   1. We are allowed to create/graft only
				 *   if CREATE and REPLACE flags are set.
				 *
				 *   2. If EXCL is set, requestor wanted to say,
				 *   that qdisc tcm_handle is not expected
				 *   to exist, so that we choose create/graft too.
				 *
				 *   3. The last case is when no flags are set.
				 *   Alas, it is sort of hole in API, we
				 *   cannot decide what to do unambiguously.
				 *   For now we select create/graft, if
				 *   user gave KIND, which does not match existing.
				 */
				if ((n->nlmsg_flags&NLM_F_CREATE) &&
				    (n->nlmsg_flags&NLM_F_REPLACE) &&
				    ((n->nlmsg_flags&NLM_F_EXCL) ||
				     (tca[TCA_KIND-1] &&
				      rtattr_strcmp(tca[TCA_KIND-1], q->ops->id))))
					goto create_n_graft;
			}
		}
	} else {
		if (!tcm->tcm_handle)
			return -EINVAL;
		q = qdisc_lookup(dev, tcm->tcm_handle);
	}

	/* Change qdisc parameters */
	if (q == NULL)
		return -ENOENT;
	if (n->nlmsg_flags&NLM_F_EXCL)
		return -EEXIST;
	if (tca[TCA_KIND-1] && rtattr_strcmp(tca[TCA_KIND-1], q->ops->id))
		return -EINVAL;
	err = qdisc_change(q, tca);
	if (err == 0)
		qdisc_notify(skb, n, clid, NULL, q);
	return err;

create_n_graft:
	if (!(n->nlmsg_flags&NLM_F_CREATE))
		return -ENOENT;
	if (clid == TC_H_INGRESS)
		q = qdisc_create(dev, tcm->tcm_parent, tca, &err);
781
	else
L
Linus Torvalds 已提交
782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811
		q = qdisc_create(dev, tcm->tcm_handle, tca, &err);
	if (q == NULL) {
		if (err == -EAGAIN)
			goto replay;
		return err;
	}

graft:
	if (1) {
		struct Qdisc *old_q = NULL;
		err = qdisc_graft(dev, p, clid, q, &old_q);
		if (err) {
			if (q) {
				spin_lock_bh(&dev->queue_lock);
				qdisc_destroy(q);
				spin_unlock_bh(&dev->queue_lock);
			}
			return err;
		}
		qdisc_notify(skb, n, clid, old_q, q);
		if (old_q) {
			spin_lock_bh(&dev->queue_lock);
			qdisc_destroy(old_q);
			spin_unlock_bh(&dev->queue_lock);
		}
	}
	return 0;
}

static int tc_fill_qdisc(struct sk_buff *skb, struct Qdisc *q, u32 clid,
J
Jamal Hadi Salim 已提交
812
			 u32 pid, u32 seq, u16 flags, int event)
L
Linus Torvalds 已提交
813 814 815 816 817 818
{
	struct tcmsg *tcm;
	struct nlmsghdr  *nlh;
	unsigned char	 *b = skb->tail;
	struct gnet_dump d;

J
Jamal Hadi Salim 已提交
819
	nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*tcm), flags);
L
Linus Torvalds 已提交
820 821
	tcm = NLMSG_DATA(nlh);
	tcm->tcm_family = AF_UNSPEC;
822 823
	tcm->tcm__pad1 = 0;
	tcm->tcm__pad2 = 0;
L
Linus Torvalds 已提交
824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845
	tcm->tcm_ifindex = q->dev->ifindex;
	tcm->tcm_parent = clid;
	tcm->tcm_handle = q->handle;
	tcm->tcm_info = atomic_read(&q->refcnt);
	RTA_PUT(skb, TCA_KIND, IFNAMSIZ, q->ops->id);
	if (q->ops->dump && q->ops->dump(q, skb) < 0)
		goto rtattr_failure;
	q->qstats.qlen = q->q.qlen;

	if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS,
			TCA_XSTATS, q->stats_lock, &d) < 0)
		goto rtattr_failure;

	if (q->ops->dump_stats && q->ops->dump_stats(q, &d) < 0)
		goto rtattr_failure;

	if (gnet_stats_copy_basic(&d, &q->bstats) < 0 ||
#ifdef CONFIG_NET_ESTIMATOR
	    gnet_stats_copy_rate_est(&d, &q->rate_est) < 0 ||
#endif
	    gnet_stats_copy_queue(&d, &q->qstats) < 0)
		goto rtattr_failure;
846

L
Linus Torvalds 已提交
847 848
	if (gnet_stats_finish_copy(&d) < 0)
		goto rtattr_failure;
849

L
Linus Torvalds 已提交
850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878
	nlh->nlmsg_len = skb->tail - b;
	return skb->len;

nlmsg_failure:
rtattr_failure:
	skb_trim(skb, b - skb->data);
	return -1;
}

static int qdisc_notify(struct sk_buff *oskb, struct nlmsghdr *n,
			u32 clid, struct Qdisc *old, struct Qdisc *new)
{
	struct sk_buff *skb;
	u32 pid = oskb ? NETLINK_CB(oskb).pid : 0;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb)
		return -ENOBUFS;

	if (old && old->handle) {
		if (tc_fill_qdisc(skb, old, clid, pid, n->nlmsg_seq, 0, RTM_DELQDISC) < 0)
			goto err_out;
	}
	if (new) {
		if (tc_fill_qdisc(skb, new, clid, pid, n->nlmsg_seq, old ? NLM_F_REPLACE : 0, RTM_NEWQDISC) < 0)
			goto err_out;
	}

	if (skb->len)
879
		return rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO);
L
Linus Torvalds 已提交
880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900

err_out:
	kfree_skb(skb);
	return -EINVAL;
}

static int tc_dump_qdisc(struct sk_buff *skb, struct netlink_callback *cb)
{
	int idx, q_idx;
	int s_idx, s_q_idx;
	struct net_device *dev;
	struct Qdisc *q;

	s_idx = cb->args[0];
	s_q_idx = q_idx = cb->args[1];
	read_lock(&dev_base_lock);
	for (dev=dev_base, idx=0; dev; dev = dev->next, idx++) {
		if (idx < s_idx)
			continue;
		if (idx > s_idx)
			s_q_idx = 0;
901
		read_lock(&qdisc_tree_lock);
L
Linus Torvalds 已提交
902 903 904 905 906 907 908 909
		q_idx = 0;
		list_for_each_entry(q, &dev->qdisc_list, list) {
			if (q_idx < s_q_idx) {
				q_idx++;
				continue;
			}
			if (tc_fill_qdisc(skb, q, q->parent, NETLINK_CB(cb->skb).pid,
					  cb->nlh->nlmsg_seq, NLM_F_MULTI, RTM_NEWQDISC) <= 0) {
910
				read_unlock(&qdisc_tree_lock);
L
Linus Torvalds 已提交
911 912 913 914
				goto done;
			}
			q_idx++;
		}
915
		read_unlock(&qdisc_tree_lock);
L
Linus Torvalds 已提交
916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991
	}

done:
	read_unlock(&dev_base_lock);

	cb->args[0] = idx;
	cb->args[1] = q_idx;

	return skb->len;
}



/************************************************
 *	Traffic classes manipulation.		*
 ************************************************/



static int tc_ctl_tclass(struct sk_buff *skb, struct nlmsghdr *n, void *arg)
{
	struct tcmsg *tcm = NLMSG_DATA(n);
	struct rtattr **tca = arg;
	struct net_device *dev;
	struct Qdisc *q = NULL;
	struct Qdisc_class_ops *cops;
	unsigned long cl = 0;
	unsigned long new_cl;
	u32 pid = tcm->tcm_parent;
	u32 clid = tcm->tcm_handle;
	u32 qid = TC_H_MAJ(clid);
	int err;

	if ((dev = __dev_get_by_index(tcm->tcm_ifindex)) == NULL)
		return -ENODEV;

	/*
	   parent == TC_H_UNSPEC - unspecified parent.
	   parent == TC_H_ROOT   - class is root, which has no parent.
	   parent == X:0	 - parent is root class.
	   parent == X:Y	 - parent is a node in hierarchy.
	   parent == 0:Y	 - parent is X:Y, where X:0 is qdisc.

	   handle == 0:0	 - generate handle from kernel pool.
	   handle == 0:Y	 - class is X:Y, where X:0 is qdisc.
	   handle == X:Y	 - clear.
	   handle == X:0	 - root class.
	 */

	/* Step 1. Determine qdisc handle X:0 */

	if (pid != TC_H_ROOT) {
		u32 qid1 = TC_H_MAJ(pid);

		if (qid && qid1) {
			/* If both majors are known, they must be identical. */
			if (qid != qid1)
				return -EINVAL;
		} else if (qid1) {
			qid = qid1;
		} else if (qid == 0)
			qid = dev->qdisc_sleeping->handle;

		/* Now qid is genuine qdisc handle consistent
		   both with parent and child.

		   TC_H_MAJ(pid) still may be unspecified, complete it now.
		 */
		if (pid)
			pid = TC_H_MAKE(qid, pid);
	} else {
		if (qid == 0)
			qid = dev->qdisc_sleeping->handle;
	}

	/* OK. Locate qdisc */
992
	if ((q = qdisc_lookup(dev, qid)) == NULL)
L
Linus Torvalds 已提交
993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		return -ENOENT;

	/* An check that it supports classes */
	cops = q->ops->cl_ops;
	if (cops == NULL)
		return -EINVAL;

	/* Now try to get class */
	if (clid == 0) {
		if (pid == TC_H_ROOT)
			clid = qid;
	} else
		clid = TC_H_MAKE(qid, clid);

	if (clid)
		cl = cops->get(q, clid);

	if (cl == 0) {
		err = -ENOENT;
		if (n->nlmsg_type != RTM_NEWTCLASS || !(n->nlmsg_flags&NLM_F_CREATE))
			goto out;
	} else {
		switch (n->nlmsg_type) {
1016
		case RTM_NEWTCLASS:
L
Linus Torvalds 已提交
1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049
			err = -EEXIST;
			if (n->nlmsg_flags&NLM_F_EXCL)
				goto out;
			break;
		case RTM_DELTCLASS:
			err = cops->delete(q, cl);
			if (err == 0)
				tclass_notify(skb, n, q, cl, RTM_DELTCLASS);
			goto out;
		case RTM_GETTCLASS:
			err = tclass_notify(skb, n, q, cl, RTM_NEWTCLASS);
			goto out;
		default:
			err = -EINVAL;
			goto out;
		}
	}

	new_cl = cl;
	err = cops->change(q, clid, pid, tca, &new_cl);
	if (err == 0)
		tclass_notify(skb, n, q, new_cl, RTM_NEWTCLASS);

out:
	if (cl)
		cops->put(q, cl);

	return err;
}


static int tc_fill_tclass(struct sk_buff *skb, struct Qdisc *q,
			  unsigned long cl,
J
Jamal Hadi Salim 已提交
1050
			  u32 pid, u32 seq, u16 flags, int event)
L
Linus Torvalds 已提交
1051 1052 1053 1054 1055 1056 1057
{
	struct tcmsg *tcm;
	struct nlmsghdr  *nlh;
	unsigned char	 *b = skb->tail;
	struct gnet_dump d;
	struct Qdisc_class_ops *cl_ops = q->ops->cl_ops;

J
Jamal Hadi Salim 已提交
1058
	nlh = NLMSG_NEW(skb, pid, seq, event, sizeof(*tcm), flags);
L
Linus Torvalds 已提交
1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102
	tcm = NLMSG_DATA(nlh);
	tcm->tcm_family = AF_UNSPEC;
	tcm->tcm_ifindex = q->dev->ifindex;
	tcm->tcm_parent = q->handle;
	tcm->tcm_handle = q->handle;
	tcm->tcm_info = 0;
	RTA_PUT(skb, TCA_KIND, IFNAMSIZ, q->ops->id);
	if (cl_ops->dump && cl_ops->dump(q, cl, skb, tcm) < 0)
		goto rtattr_failure;

	if (gnet_stats_start_copy_compat(skb, TCA_STATS2, TCA_STATS,
			TCA_XSTATS, q->stats_lock, &d) < 0)
		goto rtattr_failure;

	if (cl_ops->dump_stats && cl_ops->dump_stats(q, cl, &d) < 0)
		goto rtattr_failure;

	if (gnet_stats_finish_copy(&d) < 0)
		goto rtattr_failure;

	nlh->nlmsg_len = skb->tail - b;
	return skb->len;

nlmsg_failure:
rtattr_failure:
	skb_trim(skb, b - skb->data);
	return -1;
}

static int tclass_notify(struct sk_buff *oskb, struct nlmsghdr *n,
			  struct Qdisc *q, unsigned long cl, int event)
{
	struct sk_buff *skb;
	u32 pid = oskb ? NETLINK_CB(oskb).pid : 0;

	skb = alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL);
	if (!skb)
		return -ENOBUFS;

	if (tc_fill_tclass(skb, q, cl, pid, n->nlmsg_seq, 0, event) < 0) {
		kfree_skb(skb);
		return -EINVAL;
	}

1103
	return rtnetlink_send(skb, pid, RTNLGRP_TC, n->nlmsg_flags&NLM_F_ECHO);
L
Linus Torvalds 已提交
1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
}

struct qdisc_dump_args
{
	struct qdisc_walker w;
	struct sk_buff *skb;
	struct netlink_callback *cb;
};

static int qdisc_class_dump(struct Qdisc *q, unsigned long cl, struct qdisc_walker *arg)
{
	struct qdisc_dump_args *a = (struct qdisc_dump_args *)arg;

	return tc_fill_tclass(a->skb, q, cl, NETLINK_CB(a->cb->skb).pid,
			      a->cb->nlh->nlmsg_seq, NLM_F_MULTI, RTM_NEWTCLASS);
}

static int tc_dump_tclass(struct sk_buff *skb, struct netlink_callback *cb)
{
	int t;
	int s_t;
	struct net_device *dev;
	struct Qdisc *q;
	struct tcmsg *tcm = (struct tcmsg*)NLMSG_DATA(cb->nlh);
	struct qdisc_dump_args arg;

	if (cb->nlh->nlmsg_len < NLMSG_LENGTH(sizeof(*tcm)))
		return 0;
	if ((dev = dev_get_by_index(tcm->tcm_ifindex)) == NULL)
		return 0;

	s_t = cb->args[0];
	t = 0;

1138
	read_lock(&qdisc_tree_lock);
L
Linus Torvalds 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159
	list_for_each_entry(q, &dev->qdisc_list, list) {
		if (t < s_t || !q->ops->cl_ops ||
		    (tcm->tcm_parent &&
		     TC_H_MAJ(tcm->tcm_parent) != q->handle)) {
			t++;
			continue;
		}
		if (t > s_t)
			memset(&cb->args[1], 0, sizeof(cb->args)-sizeof(cb->args[0]));
		arg.w.fn = qdisc_class_dump;
		arg.skb = skb;
		arg.cb = cb;
		arg.w.stop  = 0;
		arg.w.skip = cb->args[1];
		arg.w.count = 0;
		q->ops->cl_ops->walk(q, &arg.w);
		cb->args[1] = arg.w.count;
		if (arg.w.stop)
			break;
		t++;
	}
1160
	read_unlock(&qdisc_tree_lock);
L
Linus Torvalds 已提交
1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175

	cb->args[0] = t;

	dev_put(dev);
	return skb->len;
}

/* Main classifier routine: scans classifier chain attached
   to this qdisc, (optionally) tests for protocol and asks
   specific classifiers.
 */
int tc_classify(struct sk_buff *skb, struct tcf_proto *tp,
	struct tcf_result *res)
{
	int err = 0;
A
Al Viro 已提交
1176
	__be16 protocol = skb->protocol;
L
Linus Torvalds 已提交
1177 1178 1179 1180 1181 1182 1183 1184
#ifdef CONFIG_NET_CLS_ACT
	struct tcf_proto *otp = tp;
reclassify:
#endif
	protocol = skb->protocol;

	for ( ; tp; tp = tp->next) {
		if ((tp->protocol == protocol ||
1185
			tp->protocol == htons(ETH_P_ALL)) &&
L
Linus Torvalds 已提交
1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199
			(err = tp->classify(skb, tp, res)) >= 0) {
#ifdef CONFIG_NET_CLS_ACT
			if ( TC_ACT_RECLASSIFY == err) {
				__u32 verd = (__u32) G_TC_VERD(skb->tc_verd);
				tp = otp;

				if (MAX_REC_LOOP < verd++) {
					printk("rule prio %d protocol %02x reclassify is buggy packet dropped\n",
						tp->prio&0xffff, ntohs(tp->protocol));
					return TC_ACT_SHOT;
				}
				skb->tc_verd = SET_TC_VERD(skb->tc_verd,verd);
				goto reclassify;
			} else {
1200
				if (skb->tc_verd)
L
Linus Torvalds 已提交
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217
					skb->tc_verd = SET_TC_VERD(skb->tc_verd,0);
				return err;
			}
#else

			return err;
#endif
		}

	}
	return -1;
}

#ifdef CONFIG_PROC_FS
static int psched_show(struct seq_file *seq, void *v)
{
	seq_printf(seq, "%08x %08x %08x %08x\n",
1218
		   (u32)NSEC_PER_USEC, (u32)PSCHED_US2NS(1),
P
Patrick McHardy 已提交
1219 1220
		   1000000,
		   (u32)NSEC_PER_SEC/(u32)ktime_to_ns(KTIME_MONOTONIC_RES));
L
Linus Torvalds 已提交
1221 1222 1223 1224 1225 1226 1227 1228 1229

	return 0;
}

static int psched_open(struct inode *inode, struct file *file)
{
	return single_open(file, psched_show, PDE(inode)->data);
}

1230
static const struct file_operations psched_fops = {
L
Linus Torvalds 已提交
1231 1232 1233 1234 1235
	.owner = THIS_MODULE,
	.open = psched_open,
	.read  = seq_read,
	.llseek = seq_lseek,
	.release = single_release,
1236
};
L
Linus Torvalds 已提交
1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273
#endif

static int __init pktsched_init(void)
{
	struct rtnetlink_link *link_p;

	link_p = rtnetlink_links[PF_UNSPEC];

	/* Setup rtnetlink links. It is made here to avoid
	   exporting large number of public symbols.
	 */

	if (link_p) {
		link_p[RTM_NEWQDISC-RTM_BASE].doit = tc_modify_qdisc;
		link_p[RTM_DELQDISC-RTM_BASE].doit = tc_get_qdisc;
		link_p[RTM_GETQDISC-RTM_BASE].doit = tc_get_qdisc;
		link_p[RTM_GETQDISC-RTM_BASE].dumpit = tc_dump_qdisc;
		link_p[RTM_NEWTCLASS-RTM_BASE].doit = tc_ctl_tclass;
		link_p[RTM_DELTCLASS-RTM_BASE].doit = tc_ctl_tclass;
		link_p[RTM_GETTCLASS-RTM_BASE].doit = tc_ctl_tclass;
		link_p[RTM_GETTCLASS-RTM_BASE].dumpit = tc_dump_tclass;
	}

	register_qdisc(&pfifo_qdisc_ops);
	register_qdisc(&bfifo_qdisc_ops);
	proc_net_fops_create("psched", 0, &psched_fops);

	return 0;
}

subsys_initcall(pktsched_init);

EXPORT_SYMBOL(qdisc_get_rtab);
EXPORT_SYMBOL(qdisc_put_rtab);
EXPORT_SYMBOL(register_qdisc);
EXPORT_SYMBOL(unregister_qdisc);
EXPORT_SYMBOL(tc_classify);