Commit | Line | Data |
---|---|---|
1da177e4 LT |
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 | ||
1da177e4 | 12 | #include <linux/module.h> |
1da177e4 LT |
13 | #include <linux/types.h> |
14 | #include <linux/kernel.h> | |
15 | #include <linux/jiffies.h> | |
16 | #include <linux/string.h> | |
1da177e4 LT |
17 | #include <linux/in.h> |
18 | #include <linux/errno.h> | |
1da177e4 | 19 | #include <linux/init.h> |
1da177e4 | 20 | #include <linux/skbuff.h> |
32740ddc | 21 | #include <linux/jhash.h> |
5a0e3ad6 | 22 | #include <linux/slab.h> |
817fb15d | 23 | #include <linux/vmalloc.h> |
0ba48053 | 24 | #include <net/netlink.h> |
1da177e4 | 25 | #include <net/pkt_sched.h> |
11fca931 | 26 | #include <net/flow_keys.h> |
ddecf0f4 | 27 | #include <net/red.h> |
1da177e4 LT |
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 | ||
10297b99 | 46 | This is not the thing that is usually called (W)FQ nowadays. |
1da177e4 LT |
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 | ||
10297b99 | 56 | - "Stochastic" -> It is not 100% fair. |
1da177e4 LT |
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: | |
18cb8098 ED |
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. | |
1da177e4 LT |
75 | |
76 | It is easy to increase these values, but not in flight. */ | |
77 | ||
18cb8098 ED |
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 | |
817fb15d ED |
82 | #define SFQ_DEFAULT_HASH_DIVISOR 1024 |
83 | ||
eeaeb068 ED |
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) | |
1da177e4 | 89 | |
18cb8098 ED |
90 | /* This type should contain at least SFQ_MAX_DEPTH + 1 + SFQ_MAX_FLOWS values */ |
91 | typedef u16 sfq_index; | |
1da177e4 | 92 | |
eda83e3b ED |
93 | /* |
94 | * We dont use pointers to save space. | |
18cb8098 ED |
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] | |
eda83e3b ED |
97 | * are 'pointers' to dep[] array |
98 | */ | |
cc7ec456 | 99 | struct sfq_head { |
1da177e4 LT |
100 | sfq_index next; |
101 | sfq_index prev; | |
102 | }; | |
103 | ||
eda83e3b ED |
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 */ | |
18cb8098 | 108 | sfq_index next; /* next slot in sfq RR chain */ |
eda83e3b ED |
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 */ | |
ddecf0f4 ED |
112 | |
113 | unsigned int backlog; | |
114 | struct red_vars vars; | |
eda83e3b ED |
115 | }; |
116 | ||
cc7ec456 | 117 | struct sfq_sched_data { |
18cb8098 ED |
118 | /* frequently used fields */ |
119 | int limit; /* limit of total number of packets in this qdisc */ | |
817fb15d | 120 | unsigned int divisor; /* number of slots in hash table */ |
ddecf0f4 ED |
121 | u8 headdrop; |
122 | u8 maxdepth; /* limit of packets per flow */ | |
18cb8098 | 123 | |
32740ddc | 124 | u32 perturbation; |
ddecf0f4 ED |
125 | u8 cur_depth; /* depth of longest slot */ |
126 | u8 flags; | |
eeaeb068 | 127 | unsigned short scaled_quantum; /* SFQ_ALLOT_SIZE(quantum) */ |
25d8c0d5 | 128 | struct tcf_proto __rcu *filter_list; |
18cb8098 ED |
129 | sfq_index *ht; /* Hash table ('divisor' slots) */ |
130 | struct sfq_slot *slots; /* Flows table ('maxflows' entries) */ | |
131 | ||
ddecf0f4 ED |
132 | struct red_parms *red_parms; |
133 | struct tc_sfqred_stats stats; | |
134 | struct sfq_slot *tail; /* current slot in round */ | |
135 | ||
18cb8098 ED |
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 | ||
ddecf0f4 | 143 | unsigned int maxflows; /* number of flows in flows array */ |
18cb8098 ED |
144 | int perturb_period; |
145 | unsigned int quantum; /* Allotment per round: MUST BE >= MTU */ | |
146 | struct timer_list perturb_timer; | |
1da177e4 LT |
147 | }; |
148 | ||
eda83e3b ED |
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 | { | |
18cb8098 | 154 | if (val < SFQ_MAX_FLOWS) |
eda83e3b | 155 | return &q->slots[val].dep; |
18cb8098 | 156 | return &q->dep[val - SFQ_MAX_FLOWS]; |
eda83e3b ED |
157 | } |
158 | ||
225d9b89 ED |
159 | /* |
160 | * In order to be able to quickly rehash our queue when timer changes | |
161 | * q->perturbation, we store flow_keys in skb->cb[] | |
162 | */ | |
163 | struct sfq_skb_cb { | |
164 | struct flow_keys keys; | |
165 | }; | |
166 | ||
167 | static inline struct sfq_skb_cb *sfq_skb_cb(const struct sk_buff *skb) | |
168 | { | |
16bda13d DM |
169 | qdisc_cb_private_validate(skb, sizeof(struct sfq_skb_cb)); |
170 | return (struct sfq_skb_cb *)qdisc_skb_cb(skb)->data; | |
225d9b89 ED |
171 | } |
172 | ||
11fca931 ED |
173 | static unsigned int sfq_hash(const struct sfq_sched_data *q, |
174 | const struct sk_buff *skb) | |
1da177e4 | 175 | { |
225d9b89 | 176 | const struct flow_keys *keys = &sfq_skb_cb(skb)->keys; |
11fca931 ED |
177 | unsigned int hash; |
178 | ||
225d9b89 ED |
179 | hash = jhash_3words((__force u32)keys->dst, |
180 | (__force u32)keys->src ^ keys->ip_proto, | |
181 | (__force u32)keys->ports, q->perturbation); | |
11fca931 | 182 | return hash & (q->divisor - 1); |
1da177e4 LT |
183 | } |
184 | ||
7d2681a6 PM |
185 | static unsigned int sfq_classify(struct sk_buff *skb, struct Qdisc *sch, |
186 | int *qerr) | |
187 | { | |
188 | struct sfq_sched_data *q = qdisc_priv(sch); | |
189 | struct tcf_result res; | |
25d8c0d5 | 190 | struct tcf_proto *fl; |
7d2681a6 PM |
191 | int result; |
192 | ||
193 | if (TC_H_MAJ(skb->priority) == sch->handle && | |
194 | TC_H_MIN(skb->priority) > 0 && | |
817fb15d | 195 | TC_H_MIN(skb->priority) <= q->divisor) |
7d2681a6 PM |
196 | return TC_H_MIN(skb->priority); |
197 | ||
25d8c0d5 JF |
198 | fl = rcu_dereference_bh(q->filter_list); |
199 | if (!fl) { | |
225d9b89 | 200 | skb_flow_dissect(skb, &sfq_skb_cb(skb)->keys); |
7d2681a6 | 201 | return sfq_hash(q, skb) + 1; |
225d9b89 | 202 | } |
7d2681a6 | 203 | |
c27f339a | 204 | *qerr = NET_XMIT_SUCCESS | __NET_XMIT_BYPASS; |
25d8c0d5 | 205 | result = tc_classify(skb, fl, &res); |
7d2681a6 PM |
206 | if (result >= 0) { |
207 | #ifdef CONFIG_NET_CLS_ACT | |
208 | switch (result) { | |
209 | case TC_ACT_STOLEN: | |
210 | case TC_ACT_QUEUED: | |
378a2f09 | 211 | *qerr = NET_XMIT_SUCCESS | __NET_XMIT_STOLEN; |
7d2681a6 PM |
212 | case TC_ACT_SHOT: |
213 | return 0; | |
214 | } | |
215 | #endif | |
817fb15d | 216 | if (TC_H_MIN(res.classid) <= q->divisor) |
7d2681a6 PM |
217 | return TC_H_MIN(res.classid); |
218 | } | |
219 | return 0; | |
220 | } | |
221 | ||
eda83e3b | 222 | /* |
18cb8098 | 223 | * x : slot number [0 .. SFQ_MAX_FLOWS - 1] |
eda83e3b | 224 | */ |
1da177e4 LT |
225 | static inline void sfq_link(struct sfq_sched_data *q, sfq_index x) |
226 | { | |
227 | sfq_index p, n; | |
18cb8098 ED |
228 | struct sfq_slot *slot = &q->slots[x]; |
229 | int qlen = slot->qlen; | |
eda83e3b | 230 | |
18cb8098 | 231 | p = qlen + SFQ_MAX_FLOWS; |
eda83e3b | 232 | n = q->dep[qlen].next; |
1da177e4 | 233 | |
18cb8098 ED |
234 | slot->dep.next = n; |
235 | slot->dep.prev = p; | |
eda83e3b ED |
236 | |
237 | q->dep[qlen].next = x; /* sfq_dep_head(q, p)->next = x */ | |
238 | sfq_dep_head(q, n)->prev = x; | |
1da177e4 LT |
239 | } |
240 | ||
eda83e3b | 241 | #define sfq_unlink(q, x, n, p) \ |
fa08943b YY |
242 | do { \ |
243 | n = q->slots[x].dep.next; \ | |
244 | p = q->slots[x].dep.prev; \ | |
245 | sfq_dep_head(q, p)->next = n; \ | |
246 | sfq_dep_head(q, n)->prev = p; \ | |
247 | } while (0) | |
eda83e3b ED |
248 | |
249 | ||
1da177e4 LT |
250 | static inline void sfq_dec(struct sfq_sched_data *q, sfq_index x) |
251 | { | |
252 | sfq_index p, n; | |
eda83e3b | 253 | int d; |
1da177e4 | 254 | |
eda83e3b | 255 | sfq_unlink(q, x, n, p); |
1da177e4 | 256 | |
eda83e3b ED |
257 | d = q->slots[x].qlen--; |
258 | if (n == p && q->cur_depth == d) | |
259 | q->cur_depth--; | |
1da177e4 LT |
260 | sfq_link(q, x); |
261 | } | |
262 | ||
263 | static inline void sfq_inc(struct sfq_sched_data *q, sfq_index x) | |
264 | { | |
265 | sfq_index p, n; | |
266 | int d; | |
267 | ||
eda83e3b | 268 | sfq_unlink(q, x, n, p); |
1da177e4 | 269 | |
eda83e3b ED |
270 | d = ++q->slots[x].qlen; |
271 | if (q->cur_depth < d) | |
272 | q->cur_depth = d; | |
1da177e4 LT |
273 | sfq_link(q, x); |
274 | } | |
275 | ||
eda83e3b ED |
276 | /* helper functions : might be changed when/if skb use a standard list_head */ |
277 | ||
278 | /* remove one skb from tail of slot queue */ | |
279 | static inline struct sk_buff *slot_dequeue_tail(struct sfq_slot *slot) | |
280 | { | |
281 | struct sk_buff *skb = slot->skblist_prev; | |
282 | ||
283 | slot->skblist_prev = skb->prev; | |
ee09b3c1 | 284 | skb->prev->next = (struct sk_buff *)slot; |
eda83e3b ED |
285 | skb->next = skb->prev = NULL; |
286 | return skb; | |
287 | } | |
288 | ||
289 | /* remove one skb from head of slot queue */ | |
290 | static inline struct sk_buff *slot_dequeue_head(struct sfq_slot *slot) | |
291 | { | |
292 | struct sk_buff *skb = slot->skblist_next; | |
293 | ||
294 | slot->skblist_next = skb->next; | |
18c8d82a | 295 | skb->next->prev = (struct sk_buff *)slot; |
eda83e3b ED |
296 | skb->next = skb->prev = NULL; |
297 | return skb; | |
298 | } | |
299 | ||
300 | static inline void slot_queue_init(struct sfq_slot *slot) | |
301 | { | |
18cb8098 | 302 | memset(slot, 0, sizeof(*slot)); |
eda83e3b ED |
303 | slot->skblist_prev = slot->skblist_next = (struct sk_buff *)slot; |
304 | } | |
305 | ||
306 | /* add skb to slot queue (tail add) */ | |
307 | static inline void slot_queue_add(struct sfq_slot *slot, struct sk_buff *skb) | |
308 | { | |
309 | skb->prev = slot->skblist_prev; | |
310 | skb->next = (struct sk_buff *)slot; | |
311 | slot->skblist_prev->next = skb; | |
312 | slot->skblist_prev = skb; | |
313 | } | |
314 | ||
1da177e4 LT |
315 | static unsigned int sfq_drop(struct Qdisc *sch) |
316 | { | |
317 | struct sfq_sched_data *q = qdisc_priv(sch); | |
eda83e3b | 318 | sfq_index x, d = q->cur_depth; |
1da177e4 LT |
319 | struct sk_buff *skb; |
320 | unsigned int len; | |
eda83e3b | 321 | struct sfq_slot *slot; |
1da177e4 | 322 | |
eda83e3b | 323 | /* Queue is full! Find the longest slot and drop tail packet from it */ |
1da177e4 | 324 | if (d > 1) { |
eda83e3b ED |
325 | x = q->dep[d].next; |
326 | slot = &q->slots[x]; | |
327 | drop: | |
18cb8098 | 328 | skb = q->headdrop ? slot_dequeue_head(slot) : slot_dequeue_tail(slot); |
0abf77e5 | 329 | len = qdisc_pkt_len(skb); |
ddecf0f4 | 330 | slot->backlog -= len; |
1da177e4 | 331 | sfq_dec(q, x); |
eda83e3b | 332 | kfree_skb(skb); |
1da177e4 | 333 | sch->q.qlen--; |
25331d6c JF |
334 | qdisc_qstats_drop(sch); |
335 | qdisc_qstats_backlog_dec(sch, skb); | |
1da177e4 LT |
336 | return len; |
337 | } | |
338 | ||
339 | if (d == 1) { | |
340 | /* It is difficult to believe, but ALL THE SLOTS HAVE LENGTH 1. */ | |
eda83e3b ED |
341 | x = q->tail->next; |
342 | slot = &q->slots[x]; | |
343 | q->tail->next = slot->next; | |
344 | q->ht[slot->hash] = SFQ_EMPTY_SLOT; | |
345 | goto drop; | |
1da177e4 LT |
346 | } |
347 | ||
348 | return 0; | |
349 | } | |
350 | ||
ddecf0f4 ED |
351 | /* Is ECN parameter configured */ |
352 | static int sfq_prob_mark(const struct sfq_sched_data *q) | |
353 | { | |
354 | return q->flags & TC_RED_ECN; | |
355 | } | |
356 | ||
357 | /* Should packets over max threshold just be marked */ | |
358 | static int sfq_hard_mark(const struct sfq_sched_data *q) | |
359 | { | |
360 | return (q->flags & (TC_RED_ECN | TC_RED_HARDDROP)) == TC_RED_ECN; | |
361 | } | |
362 | ||
363 | static int sfq_headdrop(const struct sfq_sched_data *q) | |
364 | { | |
365 | return q->headdrop; | |
366 | } | |
367 | ||
1da177e4 | 368 | static int |
6f9e98f7 | 369 | sfq_enqueue(struct sk_buff *skb, struct Qdisc *sch) |
1da177e4 LT |
370 | { |
371 | struct sfq_sched_data *q = qdisc_priv(sch); | |
7d2681a6 | 372 | unsigned int hash; |
8efa8854 | 373 | sfq_index x, qlen; |
eda83e3b | 374 | struct sfq_slot *slot; |
7f3ff4f6 | 375 | int uninitialized_var(ret); |
ddecf0f4 ED |
376 | struct sk_buff *head; |
377 | int delta; | |
7d2681a6 PM |
378 | |
379 | hash = sfq_classify(skb, sch, &ret); | |
380 | if (hash == 0) { | |
c27f339a | 381 | if (ret & __NET_XMIT_BYPASS) |
25331d6c | 382 | qdisc_qstats_drop(sch); |
7d2681a6 PM |
383 | kfree_skb(skb); |
384 | return ret; | |
385 | } | |
386 | hash--; | |
1da177e4 LT |
387 | |
388 | x = q->ht[hash]; | |
eda83e3b ED |
389 | slot = &q->slots[x]; |
390 | if (x == SFQ_EMPTY_SLOT) { | |
391 | x = q->dep[0].next; /* get a free slot */ | |
18cb8098 ED |
392 | if (x >= SFQ_MAX_FLOWS) |
393 | return qdisc_drop(skb, sch); | |
eda83e3b ED |
394 | q->ht[hash] = x; |
395 | slot = &q->slots[x]; | |
396 | slot->hash = hash; | |
ddecf0f4 ED |
397 | slot->backlog = 0; /* should already be 0 anyway... */ |
398 | red_set_vars(&slot->vars); | |
399 | goto enqueue; | |
400 | } | |
401 | if (q->red_parms) { | |
402 | slot->vars.qavg = red_calc_qavg_no_idle_time(q->red_parms, | |
403 | &slot->vars, | |
404 | slot->backlog); | |
405 | switch (red_action(q->red_parms, | |
406 | &slot->vars, | |
407 | slot->vars.qavg)) { | |
408 | case RED_DONT_MARK: | |
409 | break; | |
410 | ||
411 | case RED_PROB_MARK: | |
25331d6c | 412 | qdisc_qstats_overlimit(sch); |
ddecf0f4 ED |
413 | if (sfq_prob_mark(q)) { |
414 | /* We know we have at least one packet in queue */ | |
415 | if (sfq_headdrop(q) && | |
416 | INET_ECN_set_ce(slot->skblist_next)) { | |
417 | q->stats.prob_mark_head++; | |
418 | break; | |
419 | } | |
420 | if (INET_ECN_set_ce(skb)) { | |
421 | q->stats.prob_mark++; | |
422 | break; | |
423 | } | |
424 | } | |
425 | q->stats.prob_drop++; | |
426 | goto congestion_drop; | |
427 | ||
428 | case RED_HARD_MARK: | |
25331d6c | 429 | qdisc_qstats_overlimit(sch); |
ddecf0f4 ED |
430 | if (sfq_hard_mark(q)) { |
431 | /* We know we have at least one packet in queue */ | |
432 | if (sfq_headdrop(q) && | |
433 | INET_ECN_set_ce(slot->skblist_next)) { | |
434 | q->stats.forced_mark_head++; | |
435 | break; | |
436 | } | |
437 | if (INET_ECN_set_ce(skb)) { | |
438 | q->stats.forced_mark++; | |
439 | break; | |
440 | } | |
441 | } | |
442 | q->stats.forced_drop++; | |
443 | goto congestion_drop; | |
444 | } | |
1da177e4 | 445 | } |
6f9e98f7 | 446 | |
18cb8098 | 447 | if (slot->qlen >= q->maxdepth) { |
ddecf0f4 ED |
448 | congestion_drop: |
449 | if (!sfq_headdrop(q)) | |
18cb8098 ED |
450 | return qdisc_drop(skb, sch); |
451 | ||
ddecf0f4 | 452 | /* We know we have at least one packet in queue */ |
18cb8098 | 453 | head = slot_dequeue_head(slot); |
ddecf0f4 ED |
454 | delta = qdisc_pkt_len(head) - qdisc_pkt_len(skb); |
455 | sch->qstats.backlog -= delta; | |
456 | slot->backlog -= delta; | |
18cb8098 ED |
457 | qdisc_drop(head, sch); |
458 | ||
18cb8098 ED |
459 | slot_queue_add(slot, skb); |
460 | return NET_XMIT_CN; | |
461 | } | |
32740ddc | 462 | |
ddecf0f4 | 463 | enqueue: |
25331d6c | 464 | qdisc_qstats_backlog_inc(sch, skb); |
ddecf0f4 | 465 | slot->backlog += qdisc_pkt_len(skb); |
eda83e3b | 466 | slot_queue_add(slot, skb); |
1da177e4 | 467 | sfq_inc(q, x); |
eda83e3b ED |
468 | if (slot->qlen == 1) { /* The flow is new */ |
469 | if (q->tail == NULL) { /* It is the first flow */ | |
470 | slot->next = x; | |
1da177e4 | 471 | } else { |
eda83e3b ED |
472 | slot->next = q->tail->next; |
473 | q->tail->next = x; | |
1da177e4 | 474 | } |
cc34eb67 ED |
475 | /* We put this flow at the end of our flow list. |
476 | * This might sound unfair for a new flow to wait after old ones, | |
477 | * but we could endup servicing new flows only, and freeze old ones. | |
478 | */ | |
479 | q->tail = slot; | |
ddecf0f4 | 480 | /* We could use a bigger initial quantum for new flows */ |
eeaeb068 | 481 | slot->allot = q->scaled_quantum; |
1da177e4 | 482 | } |
9190b3b3 | 483 | if (++sch->q.qlen <= q->limit) |
9871e50e | 484 | return NET_XMIT_SUCCESS; |
1da177e4 | 485 | |
8efa8854 | 486 | qlen = slot->qlen; |
1da177e4 | 487 | sfq_drop(sch); |
8efa8854 ED |
488 | /* Return Congestion Notification only if we dropped a packet |
489 | * from this flow. | |
490 | */ | |
e1738bd9 ED |
491 | if (qlen != slot->qlen) |
492 | return NET_XMIT_CN; | |
493 | ||
494 | /* As we dropped a packet, better let upper stack know this */ | |
495 | qdisc_tree_decrease_qlen(sch, 1); | |
496 | return NET_XMIT_SUCCESS; | |
1da177e4 LT |
497 | } |
498 | ||
1da177e4 | 499 | static struct sk_buff * |
6f9e98f7 | 500 | sfq_dequeue(struct Qdisc *sch) |
1da177e4 LT |
501 | { |
502 | struct sfq_sched_data *q = qdisc_priv(sch); | |
503 | struct sk_buff *skb; | |
aa3e2199 | 504 | sfq_index a, next_a; |
eda83e3b | 505 | struct sfq_slot *slot; |
1da177e4 LT |
506 | |
507 | /* No active slots */ | |
eda83e3b | 508 | if (q->tail == NULL) |
1da177e4 LT |
509 | return NULL; |
510 | ||
eeaeb068 | 511 | next_slot: |
eda83e3b ED |
512 | a = q->tail->next; |
513 | slot = &q->slots[a]; | |
eeaeb068 ED |
514 | if (slot->allot <= 0) { |
515 | q->tail = slot; | |
516 | slot->allot += q->scaled_quantum; | |
517 | goto next_slot; | |
518 | } | |
eda83e3b | 519 | skb = slot_dequeue_head(slot); |
1da177e4 | 520 | sfq_dec(q, a); |
9190b3b3 | 521 | qdisc_bstats_update(sch, skb); |
1da177e4 | 522 | sch->q.qlen--; |
25331d6c | 523 | qdisc_qstats_backlog_dec(sch, skb); |
ddecf0f4 | 524 | slot->backlog -= qdisc_pkt_len(skb); |
1da177e4 | 525 | /* Is the slot empty? */ |
eda83e3b ED |
526 | if (slot->qlen == 0) { |
527 | q->ht[slot->hash] = SFQ_EMPTY_SLOT; | |
528 | next_a = slot->next; | |
aa3e2199 | 529 | if (a == next_a) { |
eda83e3b | 530 | q->tail = NULL; /* no more active slots */ |
1da177e4 LT |
531 | return skb; |
532 | } | |
eda83e3b | 533 | q->tail->next = next_a; |
eeaeb068 ED |
534 | } else { |
535 | slot->allot -= SFQ_ALLOT_SIZE(qdisc_pkt_len(skb)); | |
1da177e4 LT |
536 | } |
537 | return skb; | |
538 | } | |
539 | ||
540 | static void | |
6f9e98f7 | 541 | sfq_reset(struct Qdisc *sch) |
1da177e4 LT |
542 | { |
543 | struct sk_buff *skb; | |
544 | ||
545 | while ((skb = sfq_dequeue(sch)) != NULL) | |
546 | kfree_skb(skb); | |
547 | } | |
548 | ||
225d9b89 ED |
549 | /* |
550 | * When q->perturbation is changed, we rehash all queued skbs | |
551 | * to avoid OOO (Out Of Order) effects. | |
552 | * We dont use sfq_dequeue()/sfq_enqueue() because we dont want to change | |
553 | * counters. | |
554 | */ | |
18cb8098 | 555 | static void sfq_rehash(struct Qdisc *sch) |
225d9b89 | 556 | { |
18cb8098 | 557 | struct sfq_sched_data *q = qdisc_priv(sch); |
225d9b89 ED |
558 | struct sk_buff *skb; |
559 | int i; | |
560 | struct sfq_slot *slot; | |
561 | struct sk_buff_head list; | |
18cb8098 | 562 | int dropped = 0; |
225d9b89 ED |
563 | |
564 | __skb_queue_head_init(&list); | |
565 | ||
18cb8098 | 566 | for (i = 0; i < q->maxflows; i++) { |
225d9b89 ED |
567 | slot = &q->slots[i]; |
568 | if (!slot->qlen) | |
569 | continue; | |
570 | while (slot->qlen) { | |
571 | skb = slot_dequeue_head(slot); | |
572 | sfq_dec(q, i); | |
573 | __skb_queue_tail(&list, skb); | |
574 | } | |
ddecf0f4 ED |
575 | slot->backlog = 0; |
576 | red_set_vars(&slot->vars); | |
225d9b89 ED |
577 | q->ht[slot->hash] = SFQ_EMPTY_SLOT; |
578 | } | |
579 | q->tail = NULL; | |
580 | ||
581 | while ((skb = __skb_dequeue(&list)) != NULL) { | |
582 | unsigned int hash = sfq_hash(q, skb); | |
583 | sfq_index x = q->ht[hash]; | |
584 | ||
585 | slot = &q->slots[x]; | |
586 | if (x == SFQ_EMPTY_SLOT) { | |
587 | x = q->dep[0].next; /* get a free slot */ | |
18cb8098 | 588 | if (x >= SFQ_MAX_FLOWS) { |
25331d6c JF |
589 | drop: |
590 | qdisc_qstats_backlog_dec(sch, skb); | |
18cb8098 ED |
591 | kfree_skb(skb); |
592 | dropped++; | |
593 | continue; | |
594 | } | |
225d9b89 ED |
595 | q->ht[hash] = x; |
596 | slot = &q->slots[x]; | |
597 | slot->hash = hash; | |
598 | } | |
18cb8098 ED |
599 | if (slot->qlen >= q->maxdepth) |
600 | goto drop; | |
225d9b89 | 601 | slot_queue_add(slot, skb); |
ddecf0f4 ED |
602 | if (q->red_parms) |
603 | slot->vars.qavg = red_calc_qavg(q->red_parms, | |
604 | &slot->vars, | |
605 | slot->backlog); | |
606 | slot->backlog += qdisc_pkt_len(skb); | |
225d9b89 ED |
607 | sfq_inc(q, x); |
608 | if (slot->qlen == 1) { /* The flow is new */ | |
609 | if (q->tail == NULL) { /* It is the first flow */ | |
610 | slot->next = x; | |
611 | } else { | |
612 | slot->next = q->tail->next; | |
613 | q->tail->next = x; | |
614 | } | |
615 | q->tail = slot; | |
616 | slot->allot = q->scaled_quantum; | |
617 | } | |
618 | } | |
18cb8098 ED |
619 | sch->q.qlen -= dropped; |
620 | qdisc_tree_decrease_qlen(sch, dropped); | |
225d9b89 ED |
621 | } |
622 | ||
1da177e4 LT |
623 | static void sfq_perturbation(unsigned long arg) |
624 | { | |
6f9e98f7 | 625 | struct Qdisc *sch = (struct Qdisc *)arg; |
1da177e4 | 626 | struct sfq_sched_data *q = qdisc_priv(sch); |
225d9b89 | 627 | spinlock_t *root_lock = qdisc_lock(qdisc_root_sleeping(sch)); |
1da177e4 | 628 | |
225d9b89 | 629 | spin_lock(root_lock); |
63862b5b | 630 | q->perturbation = prandom_u32(); |
225d9b89 | 631 | if (!q->filter_list && q->tail) |
18cb8098 | 632 | sfq_rehash(sch); |
225d9b89 | 633 | spin_unlock(root_lock); |
1da177e4 | 634 | |
32740ddc AK |
635 | if (q->perturb_period) |
636 | mod_timer(&q->perturb_timer, jiffies + q->perturb_period); | |
1da177e4 LT |
637 | } |
638 | ||
1e90474c | 639 | static int sfq_change(struct Qdisc *sch, struct nlattr *opt) |
1da177e4 LT |
640 | { |
641 | struct sfq_sched_data *q = qdisc_priv(sch); | |
1e90474c | 642 | struct tc_sfq_qopt *ctl = nla_data(opt); |
18cb8098 | 643 | struct tc_sfq_qopt_v1 *ctl_v1 = NULL; |
5e50da01 | 644 | unsigned int qlen; |
ddecf0f4 | 645 | struct red_parms *p = NULL; |
1da177e4 | 646 | |
1e90474c | 647 | if (opt->nla_len < nla_attr_size(sizeof(*ctl))) |
1da177e4 | 648 | return -EINVAL; |
18cb8098 ED |
649 | if (opt->nla_len >= nla_attr_size(sizeof(*ctl_v1))) |
650 | ctl_v1 = nla_data(opt); | |
119b3d38 | 651 | if (ctl->divisor && |
652 | (!is_power_of_2(ctl->divisor) || ctl->divisor > 65536)) | |
653 | return -EINVAL; | |
ddecf0f4 ED |
654 | if (ctl_v1 && ctl_v1->qth_min) { |
655 | p = kmalloc(sizeof(*p), GFP_KERNEL); | |
656 | if (!p) | |
657 | return -ENOMEM; | |
658 | } | |
1da177e4 | 659 | sch_tree_lock(sch); |
18cb8098 ED |
660 | if (ctl->quantum) { |
661 | q->quantum = ctl->quantum; | |
662 | q->scaled_quantum = SFQ_ALLOT_SIZE(q->quantum); | |
663 | } | |
6f9e98f7 | 664 | q->perturb_period = ctl->perturb_period * HZ; |
18cb8098 ED |
665 | if (ctl->flows) |
666 | q->maxflows = min_t(u32, ctl->flows, SFQ_MAX_FLOWS); | |
667 | if (ctl->divisor) { | |
817fb15d | 668 | q->divisor = ctl->divisor; |
18cb8098 ED |
669 | q->maxflows = min_t(u32, q->maxflows, q->divisor); |
670 | } | |
671 | if (ctl_v1) { | |
672 | if (ctl_v1->depth) | |
673 | q->maxdepth = min_t(u32, ctl_v1->depth, SFQ_MAX_DEPTH); | |
ddecf0f4 ED |
674 | if (p) { |
675 | swap(q->red_parms, p); | |
676 | red_set_parms(q->red_parms, | |
677 | ctl_v1->qth_min, ctl_v1->qth_max, | |
678 | ctl_v1->Wlog, | |
679 | ctl_v1->Plog, ctl_v1->Scell_log, | |
680 | NULL, | |
681 | ctl_v1->max_P); | |
682 | } | |
683 | q->flags = ctl_v1->flags; | |
18cb8098 ED |
684 | q->headdrop = ctl_v1->headdrop; |
685 | } | |
686 | if (ctl->limit) { | |
687 | q->limit = min_t(u32, ctl->limit, q->maxdepth * q->maxflows); | |
688 | q->maxflows = min_t(u32, q->maxflows, q->limit); | |
689 | } | |
690 | ||
5e50da01 | 691 | qlen = sch->q.qlen; |
5588b40d | 692 | while (sch->q.qlen > q->limit) |
1da177e4 | 693 | sfq_drop(sch); |
5e50da01 | 694 | qdisc_tree_decrease_qlen(sch, qlen - sch->q.qlen); |
1da177e4 LT |
695 | |
696 | del_timer(&q->perturb_timer); | |
697 | if (q->perturb_period) { | |
32740ddc | 698 | mod_timer(&q->perturb_timer, jiffies + q->perturb_period); |
63862b5b | 699 | q->perturbation = prandom_u32(); |
1da177e4 LT |
700 | } |
701 | sch_tree_unlock(sch); | |
ddecf0f4 | 702 | kfree(p); |
1da177e4 LT |
703 | return 0; |
704 | } | |
705 | ||
bd16a6cc ED |
706 | static void *sfq_alloc(size_t sz) |
707 | { | |
708 | void *ptr = kmalloc(sz, GFP_KERNEL | __GFP_NOWARN); | |
709 | ||
710 | if (!ptr) | |
711 | ptr = vmalloc(sz); | |
712 | return ptr; | |
713 | } | |
714 | ||
715 | static void sfq_free(void *addr) | |
716 | { | |
4cb28970 | 717 | kvfree(addr); |
bd16a6cc ED |
718 | } |
719 | ||
720 | static void sfq_destroy(struct Qdisc *sch) | |
721 | { | |
722 | struct sfq_sched_data *q = qdisc_priv(sch); | |
723 | ||
724 | tcf_destroy_chain(&q->filter_list); | |
725 | q->perturb_period = 0; | |
726 | del_timer_sync(&q->perturb_timer); | |
727 | sfq_free(q->ht); | |
18cb8098 | 728 | sfq_free(q->slots); |
ddecf0f4 | 729 | kfree(q->red_parms); |
bd16a6cc ED |
730 | } |
731 | ||
1e90474c | 732 | static int sfq_init(struct Qdisc *sch, struct nlattr *opt) |
1da177e4 LT |
733 | { |
734 | struct sfq_sched_data *q = qdisc_priv(sch); | |
735 | int i; | |
736 | ||
d3e99483 | 737 | q->perturb_timer.function = sfq_perturbation; |
c19a28e1 | 738 | q->perturb_timer.data = (unsigned long)sch; |
d3e99483 | 739 | init_timer_deferrable(&q->perturb_timer); |
1da177e4 | 740 | |
18cb8098 ED |
741 | for (i = 0; i < SFQ_MAX_DEPTH + 1; i++) { |
742 | q->dep[i].next = i + SFQ_MAX_FLOWS; | |
743 | q->dep[i].prev = i + SFQ_MAX_FLOWS; | |
1da177e4 | 744 | } |
6f9e98f7 | 745 | |
18cb8098 ED |
746 | q->limit = SFQ_MAX_DEPTH; |
747 | q->maxdepth = SFQ_MAX_DEPTH; | |
eda83e3b ED |
748 | q->cur_depth = 0; |
749 | q->tail = NULL; | |
817fb15d | 750 | q->divisor = SFQ_DEFAULT_HASH_DIVISOR; |
18cb8098 | 751 | q->maxflows = SFQ_DEFAULT_FLOWS; |
02a9098e ED |
752 | q->quantum = psched_mtu(qdisc_dev(sch)); |
753 | q->scaled_quantum = SFQ_ALLOT_SIZE(q->quantum); | |
754 | q->perturb_period = 0; | |
63862b5b | 755 | q->perturbation = prandom_u32(); |
02a9098e ED |
756 | |
757 | if (opt) { | |
1da177e4 LT |
758 | int err = sfq_change(sch, opt); |
759 | if (err) | |
760 | return err; | |
761 | } | |
6f9e98f7 | 762 | |
bd16a6cc | 763 | q->ht = sfq_alloc(sizeof(q->ht[0]) * q->divisor); |
18cb8098 ED |
764 | q->slots = sfq_alloc(sizeof(q->slots[0]) * q->maxflows); |
765 | if (!q->ht || !q->slots) { | |
bd16a6cc | 766 | sfq_destroy(sch); |
817fb15d | 767 | return -ENOMEM; |
bd16a6cc | 768 | } |
817fb15d ED |
769 | for (i = 0; i < q->divisor; i++) |
770 | q->ht[i] = SFQ_EMPTY_SLOT; | |
771 | ||
18cb8098 | 772 | for (i = 0; i < q->maxflows; i++) { |
18c8d82a | 773 | slot_queue_init(&q->slots[i]); |
1da177e4 | 774 | sfq_link(q, i); |
18c8d82a | 775 | } |
23624935 ED |
776 | if (q->limit >= 1) |
777 | sch->flags |= TCQ_F_CAN_BYPASS; | |
778 | else | |
779 | sch->flags &= ~TCQ_F_CAN_BYPASS; | |
1da177e4 LT |
780 | return 0; |
781 | } | |
782 | ||
1da177e4 LT |
783 | static int sfq_dump(struct Qdisc *sch, struct sk_buff *skb) |
784 | { | |
785 | struct sfq_sched_data *q = qdisc_priv(sch); | |
27a884dc | 786 | unsigned char *b = skb_tail_pointer(skb); |
18cb8098 | 787 | struct tc_sfq_qopt_v1 opt; |
ddecf0f4 | 788 | struct red_parms *p = q->red_parms; |
18cb8098 ED |
789 | |
790 | memset(&opt, 0, sizeof(opt)); | |
791 | opt.v0.quantum = q->quantum; | |
792 | opt.v0.perturb_period = q->perturb_period / HZ; | |
793 | opt.v0.limit = q->limit; | |
794 | opt.v0.divisor = q->divisor; | |
795 | opt.v0.flows = q->maxflows; | |
796 | opt.depth = q->maxdepth; | |
797 | opt.headdrop = q->headdrop; | |
1da177e4 | 798 | |
ddecf0f4 ED |
799 | if (p) { |
800 | opt.qth_min = p->qth_min >> p->Wlog; | |
801 | opt.qth_max = p->qth_max >> p->Wlog; | |
802 | opt.Wlog = p->Wlog; | |
803 | opt.Plog = p->Plog; | |
804 | opt.Scell_log = p->Scell_log; | |
805 | opt.max_P = p->max_P; | |
806 | } | |
807 | memcpy(&opt.stats, &q->stats, sizeof(opt.stats)); | |
808 | opt.flags = q->flags; | |
809 | ||
1b34ec43 DM |
810 | if (nla_put(skb, TCA_OPTIONS, sizeof(opt), &opt)) |
811 | goto nla_put_failure; | |
1da177e4 LT |
812 | |
813 | return skb->len; | |
814 | ||
1e90474c | 815 | nla_put_failure: |
dc5fc579 | 816 | nlmsg_trim(skb, b); |
1da177e4 LT |
817 | return -1; |
818 | } | |
819 | ||
41065fba JP |
820 | static struct Qdisc *sfq_leaf(struct Qdisc *sch, unsigned long arg) |
821 | { | |
822 | return NULL; | |
823 | } | |
824 | ||
7d2681a6 PM |
825 | static unsigned long sfq_get(struct Qdisc *sch, u32 classid) |
826 | { | |
827 | return 0; | |
828 | } | |
829 | ||
eb4a5527 JP |
830 | static unsigned long sfq_bind(struct Qdisc *sch, unsigned long parent, |
831 | u32 classid) | |
832 | { | |
23624935 ED |
833 | /* we cannot bypass queue discipline anymore */ |
834 | sch->flags &= ~TCQ_F_CAN_BYPASS; | |
eb4a5527 JP |
835 | return 0; |
836 | } | |
837 | ||
da7115d9 JP |
838 | static void sfq_put(struct Qdisc *q, unsigned long cl) |
839 | { | |
840 | } | |
841 | ||
25d8c0d5 JF |
842 | static struct tcf_proto __rcu **sfq_find_tcf(struct Qdisc *sch, |
843 | unsigned long cl) | |
7d2681a6 PM |
844 | { |
845 | struct sfq_sched_data *q = qdisc_priv(sch); | |
846 | ||
847 | if (cl) | |
848 | return NULL; | |
849 | return &q->filter_list; | |
850 | } | |
851 | ||
94de78d1 PM |
852 | static int sfq_dump_class(struct Qdisc *sch, unsigned long cl, |
853 | struct sk_buff *skb, struct tcmsg *tcm) | |
854 | { | |
855 | tcm->tcm_handle |= TC_H_MIN(cl); | |
856 | return 0; | |
857 | } | |
858 | ||
859 | static int sfq_dump_class_stats(struct Qdisc *sch, unsigned long cl, | |
860 | struct gnet_dump *d) | |
861 | { | |
862 | struct sfq_sched_data *q = qdisc_priv(sch); | |
ee09b3c1 ED |
863 | sfq_index idx = q->ht[cl - 1]; |
864 | struct gnet_stats_queue qs = { 0 }; | |
865 | struct tc_sfq_xstats xstats = { 0 }; | |
c4266263 | 866 | |
ee09b3c1 ED |
867 | if (idx != SFQ_EMPTY_SLOT) { |
868 | const struct sfq_slot *slot = &q->slots[idx]; | |
94de78d1 | 869 | |
eeaeb068 | 870 | xstats.allot = slot->allot << SFQ_ALLOT_SHIFT; |
ee09b3c1 | 871 | qs.qlen = slot->qlen; |
ddecf0f4 | 872 | qs.backlog = slot->backlog; |
ee09b3c1 | 873 | } |
b0ab6f92 | 874 | if (gnet_stats_copy_queue(d, NULL, &qs, qs.qlen) < 0) |
94de78d1 PM |
875 | return -1; |
876 | return gnet_stats_copy_app(d, &xstats, sizeof(xstats)); | |
877 | } | |
878 | ||
7d2681a6 PM |
879 | static void sfq_walk(struct Qdisc *sch, struct qdisc_walker *arg) |
880 | { | |
94de78d1 PM |
881 | struct sfq_sched_data *q = qdisc_priv(sch); |
882 | unsigned int i; | |
883 | ||
884 | if (arg->stop) | |
885 | return; | |
886 | ||
817fb15d | 887 | for (i = 0; i < q->divisor; i++) { |
eda83e3b | 888 | if (q->ht[i] == SFQ_EMPTY_SLOT || |
94de78d1 PM |
889 | arg->count < arg->skip) { |
890 | arg->count++; | |
891 | continue; | |
892 | } | |
893 | if (arg->fn(sch, i + 1, arg) < 0) { | |
894 | arg->stop = 1; | |
895 | break; | |
896 | } | |
897 | arg->count++; | |
898 | } | |
7d2681a6 PM |
899 | } |
900 | ||
901 | static const struct Qdisc_class_ops sfq_class_ops = { | |
41065fba | 902 | .leaf = sfq_leaf, |
7d2681a6 | 903 | .get = sfq_get, |
da7115d9 | 904 | .put = sfq_put, |
7d2681a6 | 905 | .tcf_chain = sfq_find_tcf, |
eb4a5527 | 906 | .bind_tcf = sfq_bind, |
da7115d9 | 907 | .unbind_tcf = sfq_put, |
94de78d1 PM |
908 | .dump = sfq_dump_class, |
909 | .dump_stats = sfq_dump_class_stats, | |
7d2681a6 PM |
910 | .walk = sfq_walk, |
911 | }; | |
912 | ||
20fea08b | 913 | static struct Qdisc_ops sfq_qdisc_ops __read_mostly = { |
7d2681a6 | 914 | .cl_ops = &sfq_class_ops, |
1da177e4 LT |
915 | .id = "sfq", |
916 | .priv_size = sizeof(struct sfq_sched_data), | |
917 | .enqueue = sfq_enqueue, | |
918 | .dequeue = sfq_dequeue, | |
07bd8df5 | 919 | .peek = qdisc_peek_dequeued, |
1da177e4 LT |
920 | .drop = sfq_drop, |
921 | .init = sfq_init, | |
922 | .reset = sfq_reset, | |
923 | .destroy = sfq_destroy, | |
924 | .change = NULL, | |
925 | .dump = sfq_dump, | |
926 | .owner = THIS_MODULE, | |
927 | }; | |
928 | ||
929 | static int __init sfq_module_init(void) | |
930 | { | |
931 | return register_qdisc(&sfq_qdisc_ops); | |
932 | } | |
10297b99 | 933 | static void __exit sfq_module_exit(void) |
1da177e4 LT |
934 | { |
935 | unregister_qdisc(&sfq_qdisc_ops); | |
936 | } | |
937 | module_init(sfq_module_init) | |
938 | module_exit(sfq_module_exit) | |
939 | MODULE_LICENSE("GPL"); |