1 /* 2 * net/sched/sch_sfq.c Stochastic Fairness Queueing discipline. 3 * 4 * This program is free software; you can redistribute it and/or 5 * modify it under the terms of the GNU General Public License 6 * as published by the Free Software Foundation; either version 7 * 2 of the License, or (at your option) any later version. 8 * 9 * Authors: Alexey Kuznetsov, <kuznet@ms2.inr.ac.ru> 10 */ 11 12 #include <linux/module.h> 13 #include <linux/types.h> 14 #include <linux/kernel.h> 15 #include <linux/jiffies.h> 16 #include <linux/string.h> 17 #include <linux/in.h> 18 #include <linux/errno.h> 19 #include <linux/init.h> 20 #include <linux/skbuff.h> 21 #include <linux/jhash.h> 22 #include <linux/slab.h> 23 #include <linux/vmalloc.h> 24 #include <net/netlink.h> 25 #include <net/pkt_sched.h> 26 #include <net/pkt_cls.h> 27 #include <net/red.h> 28 29 30 /* Stochastic Fairness Queuing algorithm. 31 ======================================= 32 33 Source: 34 Paul E. McKenney "Stochastic Fairness Queuing", 35 IEEE INFOCOMM'90 Proceedings, San Francisco, 1990. 36 37 Paul E. McKenney "Stochastic Fairness Queuing", 38 "Interworking: Research and Experience", v.2, 1991, p.113-131. 39 40 41 See also: 42 M. Shreedhar and George Varghese "Efficient Fair 43 Queuing using Deficit Round Robin", Proc. SIGCOMM 95. 44 45 46 This is not the thing that is usually called (W)FQ nowadays. 47 It does not use any timestamp mechanism, but instead 48 processes queues in round-robin order. 49 50 ADVANTAGE: 51 52 - It is very cheap. Both CPU and memory requirements are minimal. 53 54 DRAWBACKS: 55 56 - "Stochastic" -> It is not 100% fair. 57 When hash collisions occur, several flows are considered as one. 58 59 - "Round-robin" -> It introduces larger delays than virtual clock 60 based schemes, and should not be used for isolating interactive 61 traffic from non-interactive. It means, that this scheduler 62 should be used as leaf of CBQ or P3, which put interactive traffic 63 to higher priority band. 64 65 We still need true WFQ for top level CSZ, but using WFQ 66 for the best effort traffic is absolutely pointless: 67 SFQ is superior for this purpose. 68 69 IMPLEMENTATION: 70 This implementation limits : 71 - maximal queue length per flow to 127 packets. 72 - max mtu to 2^18-1; 73 - max 65408 flows, 74 - number of hash buckets to 65536. 75 76 It is easy to increase these values, but not in flight. */ 77 78 #define SFQ_MAX_DEPTH 127 /* max number of packets per flow */ 79 #define SFQ_DEFAULT_FLOWS 128 80 #define SFQ_MAX_FLOWS (0x10000 - SFQ_MAX_DEPTH - 1) /* max number of flows */ 81 #define SFQ_EMPTY_SLOT 0xffff 82 #define SFQ_DEFAULT_HASH_DIVISOR 1024 83 84 /* We use 16 bits to store allot, and want to handle packets up to 64K 85 * Scale allot by 8 (1<<3) so that no overflow occurs. 86 */ 87 #define SFQ_ALLOT_SHIFT 3 88 #define SFQ_ALLOT_SIZE(X) DIV_ROUND_UP(X, 1 << SFQ_ALLOT_SHIFT) 89 90 /* This type should contain at least SFQ_MAX_DEPTH + 1 + SFQ_MAX_FLOWS values */ 91 typedef u16 sfq_index; 92 93 /* 94 * We dont use pointers to save space. 95 * Small indexes [0 ... SFQ_MAX_FLOWS - 1] are 'pointers' to slots[] array 96 * while following values [SFQ_MAX_FLOWS ... SFQ_MAX_FLOWS + SFQ_MAX_DEPTH] 97 * are 'pointers' to dep[] array 98 */ 99 struct sfq_head { 100 sfq_index next; 101 sfq_index prev; 102 }; 103 104 struct sfq_slot { 105 struct sk_buff *skblist_next; 106 struct sk_buff *skblist_prev; 107 sfq_index qlen; /* number of skbs in skblist */ 108 sfq_index next; /* next slot in sfq RR chain */ 109 struct sfq_head dep; /* anchor in dep[] chains */ 110 unsigned short hash; /* hash value (index in ht[]) */ 111 short allot; /* credit for this slot */ 112 113 unsigned int backlog; 114 struct red_vars vars; 115 }; 116 117 struct sfq_sched_data { 118 /* frequently used fields */ 119 int limit; /* limit of total number of packets in this qdisc */ 120 unsigned int divisor; /* number of slots in hash table */ 121 u8 headdrop; 122 u8 maxdepth; /* limit of packets per flow */ 123 124 u32 perturbation; 125 u8 cur_depth; /* depth of longest slot */ 126 u8 flags; 127 unsigned short scaled_quantum; /* SFQ_ALLOT_SIZE(quantum) */ 128 struct tcf_proto __rcu *filter_list; 129 sfq_index *ht; /* Hash table ('divisor' slots) */ 130 struct sfq_slot *slots; /* Flows table ('maxflows' entries) */ 131 132 struct red_parms *red_parms; 133 struct tc_sfqred_stats stats; 134 struct sfq_slot *tail; /* current slot in round */ 135 136 struct sfq_head dep[SFQ_MAX_DEPTH + 1]; 137 /* Linked lists of slots, indexed by depth 138 * dep[0] : list of unused flows 139 * dep[1] : list of flows with 1 packet 140 * dep[X] : list of flows with X packets 141 */ 142 143 unsigned int maxflows; /* number of flows in flows array */ 144 int perturb_period; 145 unsigned int quantum; /* Allotment per round: MUST BE >= MTU */ 146 struct timer_list perturb_timer; 147 }; 148 149 /* 150 * sfq_head are either in a sfq_slot or in dep[] array 151 */ 152 static inline struct sfq_head *sfq_dep_head(struct sfq_sched_data *q, sfq_index val) 153 { 154 if (val < SFQ_MAX_FLOWS) 155 return &q->slots[val].dep; 156 return &q->dep[val - SFQ_MAX_FLOWS]; 157 } 158 159 static unsigned int sfq_hash(const struct sfq_sched_data *q, 160 const struct sk_buff *skb) 161 { 162 return skb_get_hash_perturb(skb, q->perturbation) & (q->divisor - 1); 163 } 164 165 static unsigned int sfq_classify(struct sk_buff *skb, struct Qdisc *sch, 166 int *qerr) 167 { 168 struct sfq_sched_data *q = qdisc_priv(sch); 169 struct tcf_result res; 170 struct tcf_proto *fl; 171 int result; 172 173 if (TC_H_MAJ(skb->priority) == sch->handle && 174 TC_H_MIN(skb->priority) > 0 && 175 TC_H_MIN(skb->priority) <= q->divisor) 176 return TC_H_MIN(skb->priority); 177 178 fl = rcu_dereference_bh(q->filter_list); 179 if (!fl) 180 return sfq_hash(q, skb) + 1; 181 182 *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; 183 result = tc_classify(skb, fl, &res, false); 184 if (result >= 0) { 185 #ifdef CONFIG_NET_CLS_ACT 186 switch (result) { 187 case TC_ACT_STOLEN: 188 case TC_ACT_QUEUED: 189 *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; 190 case TC_ACT_SHOT: 191 return 0; 192 } 193 #endif 194 if (TC_H_MIN(res.classid) <= q->divisor) 195 return TC_H_MIN(res.classid); 196 } 197 return 0; 198 } 199 200 /* 201 * x : slot number [0 .. SFQ_MAX_FLOWS - 1] 202 */ 203 static inline void sfq_link(struct sfq_sched_data *q, sfq_index x) 204 { 205 sfq_index p, n; 206 struct sfq_slot *slot = &q->slots[x]; 207 int qlen = slot->qlen; 208 209 p = qlen + SFQ_MAX_FLOWS; 210 n = q->dep[qlen].next; 211 212 slot->dep.next = n; 213 slot->dep.prev = p; 214 215 q->dep[qlen].next = x; /* sfq_dep_head(q, p)->next = x */ 216 sfq_dep_head(q, n)->prev = x; 217 } 218 219 #define sfq_unlink(q, x, n, p) \ 220 do { \ 221 n = q->slots[x].dep.next; \ 222 p = q->slots[x].dep.prev; \ 223 sfq_dep_head(q, p)->next = n; \ 224 sfq_dep_head(q, n)->prev = p; \ 225 } while (0) 226 227 228 static inline void sfq_dec(struct sfq_sched_data *q, sfq_index x) 229 { 230 sfq_index p, n; 231 int d; 232 233 sfq_unlink(q, x, n, p); 234 235 d = q->slots[x].qlen--; 236 if (n == p && q->cur_depth == d) 237 q->cur_depth--; 238 sfq_link(q, x); 239 } 240 241 static inline void sfq_inc(struct sfq_sched_data *q, sfq_index x) 242 { 243 sfq_index p, n; 244 int d; 245 246 sfq_unlink(q, x, n, p); 247 248 d = ++q->slots[x].qlen; 249 if (q->cur_depth < d) 250 q->cur_depth = d; 251 sfq_link(q, x); 252 } 253 254 /* helper functions : might be changed when/if skb use a standard list_head */ 255 256 /* remove one skb from tail of slot queue */ 257 static inline struct sk_buff *slot_dequeue_tail(struct sfq_slot *slot) 258 { 259 struct sk_buff *skb = slot->skblist_prev; 260 261 slot->skblist_prev = skb->prev; 262 skb->prev->next = (struct sk_buff *)slot; 263 skb->next = skb->prev = NULL; 264 return skb; 265 } 266 267 /* remove one skb from head of slot queue */ 268 static inline struct sk_buff *slot_dequeue_head(struct sfq_slot *slot) 269 { 270 struct sk_buff *skb = slot->skblist_next; 271 272 slot->skblist_next = skb->next; 273 skb->next->prev = (struct sk_buff *)slot; 274 skb->next = skb->prev = NULL; 275 return skb; 276 } 277 278 static inline void slot_queue_init(struct sfq_slot *slot) 279 { 280 memset(slot, 0, sizeof(*slot)); 281 slot->skblist_prev = slot->skblist_next = (struct sk_buff *)slot; 282 } 283 284 /* add skb to slot queue (tail add) */ 285 static inline void slot_queue_add(struct sfq_slot *slot, struct sk_buff *skb) 286 { 287 skb->prev = slot->skblist_prev; 288 skb->next = (struct sk_buff *)slot; 289 slot->skblist_prev->next = skb; 290 slot->skblist_prev = skb; 291 } 292 293 static unsigned int sfq_drop(struct Qdisc *sch) 294 { 295 struct sfq_sched_data *q = qdisc_priv(sch); 296 sfq_index x, d = q->cur_depth; 297 struct sk_buff *skb; 298 unsigned int len; 299 struct sfq_slot *slot; 300 301 /* Queue is full! Find the longest slot and drop tail packet from it */ 302 if (d > 1) { 303 x = q->dep[d].next; 304 slot = &q->slots[x]; 305 drop: 306 skb = q->headdrop ? slot_dequeue_head(slot) : slot_dequeue_tail(slot); 307 len = qdisc_pkt_len(skb); 308 slot->backlog -= len; 309 sfq_dec(q, x); 310 sch->q.qlen--; 311 qdisc_qstats_drop(sch); 312 qdisc_qstats_backlog_dec(sch, skb); 313 kfree_skb(skb); 314 return len; 315 } 316 317 if (d == 1) { 318 /* It is difficult to believe, but ALL THE SLOTS HAVE LENGTH 1. */ 319 x = q->tail->next; 320 slot = &q->slots[x]; 321 q->tail->next = slot->next; 322 q->ht[slot->hash] = SFQ_EMPTY_SLOT; 323 goto drop; 324 } 325 326 return 0; 327 } 328 329 /* Is ECN parameter configured */ 330 static int sfq_prob_mark(const struct sfq_sched_data *q) 331 { 332 return q->flags & TC_RED_ECN; 333 } 334 335 /* Should packets over max threshold just be marked */ 336 static int sfq_hard_mark(const struct sfq_sched_data *q) 337 { 338 return (q->flags & (TC_RED_ECN | TC_RED_HARDDROP)) == TC_RED_ECN; 339 } 340 341 static int sfq_headdrop(const struct sfq_sched_data *q) 342 { 343 return q->headdrop; 344 } 345 346 static int 347 sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch, struct sk_buff **to_free) 348 { 349 struct sfq_sched_data *q = qdisc_priv(sch); 350 unsigned int hash, dropped; 351 sfq_index x, qlen; 352 struct sfq_slot *slot; 353 int uninitialized_var(ret); 354 struct sk_buff *head; 355 int delta; 356 357 hash = sfq_classify(skb, sch, &ret); 358 if (hash == 0) { 359 if (ret & __NET_XMIT_BYPASS) 360 qdisc_qstats_drop(sch); 361 kfree_skb(skb); 362 return ret; 363 } 364 hash--; 365 366 x = q->ht[hash]; 367 slot = &q->slots[x]; 368 if (x == SFQ_EMPTY_SLOT) { 369 x = q->dep[0].next; /* get a free slot */ 370 if (x >= SFQ_MAX_FLOWS) 371 return qdisc_drop(skb, sch, to_free); 372 q->ht[hash] = x; 373 slot = &q->slots[x]; 374 slot->hash = hash; 375 slot->backlog = 0; /* should already be 0 anyway... */ 376 red_set_vars(&slot->vars); 377 goto enqueue; 378 } 379 if (q->red_parms) { 380 slot->vars.qavg = red_calc_qavg_no_idle_time(q->red_parms, 381 &slot->vars, 382 slot->backlog); 383 switch (red_action(q->red_parms, 384 &slot->vars, 385 slot->vars.qavg)) { 386 case RED_DONT_MARK: 387 break; 388 389 case RED_PROB_MARK: 390 qdisc_qstats_overlimit(sch); 391 if (sfq_prob_mark(q)) { 392 /* We know we have at least one packet in queue */ 393 if (sfq_headdrop(q) && 394 INET_ECN_set_ce(slot->skblist_next)) { 395 q->stats.prob_mark_head++; 396 break; 397 } 398 if (INET_ECN_set_ce(skb)) { 399 q->stats.prob_mark++; 400 break; 401 } 402 } 403 q->stats.prob_drop++; 404 goto congestion_drop; 405 406 case RED_HARD_MARK: 407 qdisc_qstats_overlimit(sch); 408 if (sfq_hard_mark(q)) { 409 /* We know we have at least one packet in queue */ 410 if (sfq_headdrop(q) && 411 INET_ECN_set_ce(slot->skblist_next)) { 412 q->stats.forced_mark_head++; 413 break; 414 } 415 if (INET_ECN_set_ce(skb)) { 416 q->stats.forced_mark++; 417 break; 418 } 419 } 420 q->stats.forced_drop++; 421 goto congestion_drop; 422 } 423 } 424 425 if (slot->qlen >= q->maxdepth) { 426 congestion_drop: 427 if (!sfq_headdrop(q)) 428 return qdisc_drop(skb, sch, to_free); 429 430 /* We know we have at least one packet in queue */ 431 head = slot_dequeue_head(slot); 432 delta = qdisc_pkt_len(head) - qdisc_pkt_len(skb); 433 sch->qstats.backlog -= delta; 434 slot->backlog -= delta; 435 qdisc_drop(head, sch, to_free); 436 437 slot_queue_add(slot, skb); 438 return NET_XMIT_CN; 439 } 440 441 enqueue: 442 qdisc_qstats_backlog_inc(sch, skb); 443 slot->backlog += qdisc_pkt_len(skb); 444 slot_queue_add(slot, skb); 445 sfq_inc(q, x); 446 if (slot->qlen == 1) { /* The flow is new */ 447 if (q->tail == NULL) { /* It is the first flow */ 448 slot->next = x; 449 } else { 450 slot->next = q->tail->next; 451 q->tail->next = x; 452 } 453 /* We put this flow at the end of our flow list. 454 * This might sound unfair for a new flow to wait after old ones, 455 * but we could endup servicing new flows only, and freeze old ones. 456 */ 457 q->tail = slot; 458 /* We could use a bigger initial quantum for new flows */ 459 slot->allot = q->scaled_quantum; 460 } 461 if (++sch->q.qlen <= q->limit) 462 return NET_XMIT_SUCCESS; 463 464 qlen = slot->qlen; 465 dropped = sfq_drop(sch); 466 /* Return Congestion Notification only if we dropped a packet 467 * from this flow. 468 */ 469 if (qlen != slot->qlen) 470 return NET_XMIT_CN; 471 472 /* As we dropped a packet, better let upper stack know this */ 473 qdisc_tree_reduce_backlog(sch, 1, dropped); 474 return NET_XMIT_SUCCESS; 475 } 476 477 static struct sk_buff * 478 sfq_dequeue(struct Qdisc *sch) 479 { 480 struct sfq_sched_data *q = qdisc_priv(sch); 481 struct sk_buff *skb; 482 sfq_index a, next_a; 483 struct sfq_slot *slot; 484 485 /* No active slots */ 486 if (q->tail == NULL) 487 return NULL; 488 489 next_slot: 490 a = q->tail->next; 491 slot = &q->slots[a]; 492 if (slot->allot <= 0) { 493 q->tail = slot; 494 slot->allot += q->scaled_quantum; 495 goto next_slot; 496 } 497 skb = slot_dequeue_head(slot); 498 sfq_dec(q, a); 499 qdisc_bstats_update(sch, skb); 500 sch->q.qlen--; 501 qdisc_qstats_backlog_dec(sch, skb); 502 slot->backlog -= qdisc_pkt_len(skb); 503 /* Is the slot empty? */ 504 if (slot->qlen == 0) { 505 q->ht[slot->hash] = SFQ_EMPTY_SLOT; 506 next_a = slot->next; 507 if (a == next_a) { 508 q->tail = NULL; /* no more active slots */ 509 return skb; 510 } 511 q->tail->next = next_a; 512 } else { 513 slot->allot -= SFQ_ALLOT_SIZE(qdisc_pkt_len(skb)); 514 } 515 return skb; 516 } 517 518 static void 519 sfq_reset(struct Qdisc *sch) 520 { 521 struct sk_buff *skb; 522 523 while ((skb = sfq_dequeue(sch)) != NULL) 524 rtnl_kfree_skbs(skb, skb); 525 } 526 527 /* 528 * When q->perturbation is changed, we rehash all queued skbs 529 * to avoid OOO (Out Of Order) effects. 530 * We dont use sfq_dequeue()/sfq_enqueue() because we dont want to change 531 * counters. 532 */ 533 static void sfq_rehash(struct Qdisc *sch) 534 { 535 struct sfq_sched_data *q = qdisc_priv(sch); 536 struct sk_buff *skb; 537 int i; 538 struct sfq_slot *slot; 539 struct sk_buff_head list; 540 int dropped = 0; 541 unsigned int drop_len = 0; 542 543 __skb_queue_head_init(&list); 544 545 for (i = 0; i < q->maxflows; i++) { 546 slot = &q->slots[i]; 547 if (!slot->qlen) 548 continue; 549 while (slot->qlen) { 550 skb = slot_dequeue_head(slot); 551 sfq_dec(q, i); 552 __skb_queue_tail(&list, skb); 553 } 554 slot->backlog = 0; 555 red_set_vars(&slot->vars); 556 q->ht[slot->hash] = SFQ_EMPTY_SLOT; 557 } 558 q->tail = NULL; 559 560 while ((skb = __skb_dequeue(&list)) != NULL) { 561 unsigned int hash = sfq_hash(q, skb); 562 sfq_index x = q->ht[hash]; 563 564 slot = &q->slots[x]; 565 if (x == SFQ_EMPTY_SLOT) { 566 x = q->dep[0].next; /* get a free slot */ 567 if (x >= SFQ_MAX_FLOWS) { 568 drop: 569 qdisc_qstats_backlog_dec(sch, skb); 570 drop_len += qdisc_pkt_len(skb); 571 kfree_skb(skb); 572 dropped++; 573 continue; 574 } 575 q->ht[hash] = x; 576 slot = &q->slots[x]; 577 slot->hash = hash; 578 } 579 if (slot->qlen >= q->maxdepth) 580 goto drop; 581 slot_queue_add(slot, skb); 582 if (q->red_parms) 583 slot->vars.qavg = red_calc_qavg(q->red_parms, 584 &slot->vars, 585 slot->backlog); 586 slot->backlog += qdisc_pkt_len(skb); 587 sfq_inc(q, x); 588 if (slot->qlen == 1) { /* The flow is new */ 589 if (q->tail == NULL) { /* It is the first flow */ 590 slot->next = x; 591 } else { 592 slot->next = q->tail->next; 593 q->tail->next = x; 594 } 595 q->tail = slot; 596 slot->allot = q->scaled_quantum; 597 } 598 } 599 sch->q.qlen -= dropped; 600 qdisc_tree_reduce_backlog(sch, dropped, drop_len); 601 } 602 603 static void sfq_perturbation(unsigned long arg) 604 { 605 struct Qdisc *sch = (struct Qdisc *)arg; 606 struct sfq_sched_data *q = qdisc_priv(sch); 607 spinlock_t *root_lock = qdisc_lock(qdisc_root_sleeping(sch)); 608 609 spin_lock(root_lock); 610 q->perturbation = prandom_u32(); 611 if (!q->filter_list && q->tail) 612 sfq_rehash(sch); 613 spin_unlock(root_lock); 614 615 if (q->perturb_period) 616 mod_timer(&q->perturb_timer, jiffies + q->perturb_period); 617 } 618 619 static int sfq_change(struct Qdisc *sch, struct nlattr *opt) 620 { 621 struct sfq_sched_data *q = qdisc_priv(sch); 622 struct tc_sfq_qopt *ctl = nla_data(opt); 623 struct tc_sfq_qopt_v1 *ctl_v1 = NULL; 624 unsigned int qlen, dropped = 0; 625 struct red_parms *p = NULL; 626 627 if (opt->nla_len < nla_attr_size(sizeof(*ctl))) 628 return -EINVAL; 629 if (opt->nla_len >= nla_attr_size(sizeof(*ctl_v1))) 630 ctl_v1 = nla_data(opt); 631 if (ctl->divisor && 632 (!is_power_of_2(ctl->divisor) || ctl->divisor > 65536)) 633 return -EINVAL; 634 if (ctl_v1 && ctl_v1->qth_min) { 635 p = kmalloc(sizeof(*p), GFP_KERNEL); 636 if (!p) 637 return -ENOMEM; 638 } 639 sch_tree_lock(sch); 640 if (ctl->quantum) { 641 q->quantum = ctl->quantum; 642 q->scaled_quantum = SFQ_ALLOT_SIZE(q->quantum); 643 } 644 q->perturb_period = ctl->perturb_period * HZ; 645 if (ctl->flows) 646 q->maxflows = min_t(u32, ctl->flows, SFQ_MAX_FLOWS); 647 if (ctl->divisor) { 648 q->divisor = ctl->divisor; 649 q->maxflows = min_t(u32, q->maxflows, q->divisor); 650 } 651 if (ctl_v1) { 652 if (ctl_v1->depth) 653 q->maxdepth = min_t(u32, ctl_v1->depth, SFQ_MAX_DEPTH); 654 if (p) { 655 swap(q->red_parms, p); 656 red_set_parms(q->red_parms, 657 ctl_v1->qth_min, ctl_v1->qth_max, 658 ctl_v1->Wlog, 659 ctl_v1->Plog, ctl_v1->Scell_log, 660 NULL, 661 ctl_v1->max_P); 662 } 663 q->flags = ctl_v1->flags; 664 q->headdrop = ctl_v1->headdrop; 665 } 666 if (ctl->limit) { 667 q->limit = min_t(u32, ctl->limit, q->maxdepth * q->maxflows); 668 q->maxflows = min_t(u32, q->maxflows, q->limit); 669 } 670 671 qlen = sch->q.qlen; 672 while (sch->q.qlen > q->limit) 673 dropped += sfq_drop(sch); 674 qdisc_tree_reduce_backlog(sch, qlen - sch->q.qlen, dropped); 675 676 del_timer(&q->perturb_timer); 677 if (q->perturb_period) { 678 mod_timer(&q->perturb_timer, jiffies + q->perturb_period); 679 q->perturbation = prandom_u32(); 680 } 681 sch_tree_unlock(sch); 682 kfree(p); 683 return 0; 684 } 685 686 static void *sfq_alloc(size_t sz) 687 { 688 void *ptr = kmalloc(sz, GFP_KERNEL | __GFP_NOWARN); 689 690 if (!ptr) 691 ptr = vmalloc(sz); 692 return ptr; 693 } 694 695 static void sfq_free(void *addr) 696 { 697 kvfree(addr); 698 } 699 700 static void sfq_destroy(struct Qdisc *sch) 701 { 702 struct sfq_sched_data *q = qdisc_priv(sch); 703 704 tcf_destroy_chain(&q->filter_list); 705 q->perturb_period = 0; 706 del_timer_sync(&q->perturb_timer); 707 sfq_free(q->ht); 708 sfq_free(q->slots); 709 kfree(q->red_parms); 710 } 711 712 static int sfq_init(struct Qdisc *sch, struct nlattr *opt) 713 { 714 struct sfq_sched_data *q = qdisc_priv(sch); 715 int i; 716 717 q->perturb_timer.function = sfq_perturbation; 718 q->perturb_timer.data = (unsigned long)sch; 719 init_timer_deferrable(&q->perturb_timer); 720 721 for (i = 0; i < SFQ_MAX_DEPTH + 1; i++) { 722 q->dep[i].next = i + SFQ_MAX_FLOWS; 723 q->dep[i].prev = i + SFQ_MAX_FLOWS; 724 } 725 726 q->limit = SFQ_MAX_DEPTH; 727 q->maxdepth = SFQ_MAX_DEPTH; 728 q->cur_depth = 0; 729 q->tail = NULL; 730 q->divisor = SFQ_DEFAULT_HASH_DIVISOR; 731 q->maxflows = SFQ_DEFAULT_FLOWS; 732 q->quantum = psched_mtu(qdisc_dev(sch)); 733 q->scaled_quantum = SFQ_ALLOT_SIZE(q->quantum); 734 q->perturb_period = 0; 735 q->perturbation = prandom_u32(); 736 737 if (opt) { 738 int err = sfq_change(sch, opt); 739 if (err) 740 return err; 741 } 742 743 q->ht = sfq_alloc(sizeof(q->ht[0]) * q->divisor); 744 q->slots = sfq_alloc(sizeof(q->slots[0]) * q->maxflows); 745 if (!q->ht || !q->slots) { 746 /* Note: sfq_destroy() will be called by our caller */ 747 return -ENOMEM; 748 } 749 750 for (i = 0; i < q->divisor; i++) 751 q->ht[i] = SFQ_EMPTY_SLOT; 752 753 for (i = 0; i < q->maxflows; i++) { 754 slot_queue_init(&q->slots[i]); 755 sfq_link(q, i); 756 } 757 if (q->limit >= 1) 758 sch->flags |= TCQ_F_CAN_BYPASS; 759 else 760 sch->flags &= ~TCQ_F_CAN_BYPASS; 761 return 0; 762 } 763 764 static int sfq_dump(struct Qdisc *sch, struct sk_buff *skb) 765 { 766 struct sfq_sched_data *q = qdisc_priv(sch); 767 unsigned char *b = skb_tail_pointer(skb); 768 struct tc_sfq_qopt_v1 opt; 769 struct red_parms *p = q->red_parms; 770 771 memset(&opt, 0, sizeof(opt)); 772 opt.v0.quantum = q->quantum; 773 opt.v0.perturb_period = q->perturb_period / HZ; 774 opt.v0.limit = q->limit; 775 opt.v0.divisor = q->divisor; 776 opt.v0.flows = q->maxflows; 777 opt.depth = q->maxdepth; 778 opt.headdrop = q->headdrop; 779 780 if (p) { 781 opt.qth_min = p->qth_min >> p->Wlog; 782 opt.qth_max = p->qth_max >> p->Wlog; 783 opt.Wlog = p->Wlog; 784 opt.Plog = p->Plog; 785 opt.Scell_log = p->Scell_log; 786 opt.max_P = p->max_P; 787 } 788 memcpy(&opt.stats, &q->stats, sizeof(opt.stats)); 789 opt.flags = q->flags; 790 791 if (nla_put(skb, TCA_OPTIONS, sizeof(opt), &opt)) 792 goto nla_put_failure; 793 794 return skb->len; 795 796 nla_put_failure: 797 nlmsg_trim(skb, b); 798 return -1; 799 } 800 801 static struct Qdisc *sfq_leaf(struct Qdisc *sch, unsigned long arg) 802 { 803 return NULL; 804 } 805 806 static unsigned long sfq_get(struct Qdisc *sch, u32 classid) 807 { 808 return 0; 809 } 810 811 static unsigned long sfq_bind(struct Qdisc *sch, unsigned long parent, 812 u32 classid) 813 { 814 /* we cannot bypass queue discipline anymore */ 815 sch->flags &= ~TCQ_F_CAN_BYPASS; 816 return 0; 817 } 818 819 static void sfq_put(struct Qdisc *q, unsigned long cl) 820 { 821 } 822 823 static struct tcf_proto __rcu **sfq_find_tcf(struct Qdisc *sch, 824 unsigned long cl) 825 { 826 struct sfq_sched_data *q = qdisc_priv(sch); 827 828 if (cl) 829 return NULL; 830 return &q->filter_list; 831 } 832 833 static int sfq_dump_class(struct Qdisc *sch, unsigned long cl, 834 struct sk_buff *skb, struct tcmsg *tcm) 835 { 836 tcm->tcm_handle |= TC_H_MIN(cl); 837 return 0; 838 } 839 840 static int sfq_dump_class_stats(struct Qdisc *sch, unsigned long cl, 841 struct gnet_dump *d) 842 { 843 struct sfq_sched_data *q = qdisc_priv(sch); 844 sfq_index idx = q->ht[cl - 1]; 845 struct gnet_stats_queue qs = { 0 }; 846 struct tc_sfq_xstats xstats = { 0 }; 847 848 if (idx != SFQ_EMPTY_SLOT) { 849 const struct sfq_slot *slot = &q->slots[idx]; 850 851 xstats.allot = slot->allot << SFQ_ALLOT_SHIFT; 852 qs.qlen = slot->qlen; 853 qs.backlog = slot->backlog; 854 } 855 if (gnet_stats_copy_queue(d, NULL, &qs, qs.qlen) < 0) 856 return -1; 857 return gnet_stats_copy_app(d, &xstats, sizeof(xstats)); 858 } 859 860 static void sfq_walk(struct Qdisc *sch, struct qdisc_walker *arg) 861 { 862 struct sfq_sched_data *q = qdisc_priv(sch); 863 unsigned int i; 864 865 if (arg->stop) 866 return; 867 868 for (i = 0; i < q->divisor; i++) { 869 if (q->ht[i] == SFQ_EMPTY_SLOT || 870 arg->count < arg->skip) { 871 arg->count++; 872 continue; 873 } 874 if (arg->fn(sch, i + 1, arg) < 0) { 875 arg->stop = 1; 876 break; 877 } 878 arg->count++; 879 } 880 } 881 882 static const struct Qdisc_class_ops sfq_class_ops = { 883 .leaf = sfq_leaf, 884 .get = sfq_get, 885 .put = sfq_put, 886 .tcf_chain = sfq_find_tcf, 887 .bind_tcf = sfq_bind, 888 .unbind_tcf = sfq_put, 889 .dump = sfq_dump_class, 890 .dump_stats = sfq_dump_class_stats, 891 .walk = sfq_walk, 892 }; 893 894 static struct Qdisc_ops sfq_qdisc_ops __read_mostly = { 895 .cl_ops = &sfq_class_ops, 896 .id = "sfq", 897 .priv_size = sizeof(struct sfq_sched_data), 898 .enqueue = sfq_enqueue, 899 .dequeue = sfq_dequeue, 900 .peek = qdisc_peek_dequeued, 901 .init = sfq_init, 902 .reset = sfq_reset, 903 .destroy = sfq_destroy, 904 .change = NULL, 905 .dump = sfq_dump, 906 .owner = THIS_MODULE, 907 }; 908 909 static int __init sfq_module_init(void) 910 { 911 return register_qdisc(&sfq_qdisc_ops); 912 } 913 static void __exit sfq_module_exit(void) 914 { 915 unregister_qdisc(&sfq_qdisc_ops); 916 } 917 module_init(sfq_module_init) 918 module_exit(sfq_module_exit) 919 MODULE_LICENSE("GPL"); 920