Merge tag 'ep93xx-fixes-for-3.9-rc5' of git://github.com/RyanMallon/linux-ep93xx...
[platform/adaptation/renesas_rcar/renesas_kernel.git] / net / sched / sch_red.c
1 /*
2  * net/sched/sch_red.c  Random Early Detection queue.
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  * Changes:
12  * J Hadi Salim 980914: computation fixes
13  * Alexey Makarenko <makar@phoenix.kharkov.ua> 990814: qave on idle link was calculated incorrectly.
14  * J Hadi Salim 980816:  ECN support
15  */
16
17 #include <linux/module.h>
18 #include <linux/types.h>
19 #include <linux/kernel.h>
20 #include <linux/skbuff.h>
21 #include <net/pkt_sched.h>
22 #include <net/inet_ecn.h>
23 #include <net/red.h>
24
25
26 /*      Parameters, settable by user:
27         -----------------------------
28
29         limit           - bytes (must be > qth_max + burst)
30
31         Hard limit on queue length, should be chosen >qth_max
32         to allow packet bursts. This parameter does not
33         affect the algorithms behaviour and can be chosen
34         arbitrarily high (well, less than ram size)
35         Really, this limit will never be reached
36         if RED works correctly.
37  */
38
39 struct red_sched_data {
40         u32                     limit;          /* HARD maximal queue length */
41         unsigned char           flags;
42         struct timer_list       adapt_timer;
43         struct red_parms        parms;
44         struct red_vars         vars;
45         struct red_stats        stats;
46         struct Qdisc            *qdisc;
47 };
48
49 static inline int red_use_ecn(struct red_sched_data *q)
50 {
51         return q->flags & TC_RED_ECN;
52 }
53
54 static inline int red_use_harddrop(struct red_sched_data *q)
55 {
56         return q->flags & TC_RED_HARDDROP;
57 }
58
59 static int red_enqueue(struct sk_buff *skb, struct Qdisc *sch)
60 {
61         struct red_sched_data *q = qdisc_priv(sch);
62         struct Qdisc *child = q->qdisc;
63         int ret;
64
65         q->vars.qavg = red_calc_qavg(&q->parms,
66                                      &q->vars,
67                                      child->qstats.backlog);
68
69         if (red_is_idling(&q->vars))
70                 red_end_of_idle_period(&q->vars);
71
72         switch (red_action(&q->parms, &q->vars, q->vars.qavg)) {
73         case RED_DONT_MARK:
74                 break;
75
76         case RED_PROB_MARK:
77                 sch->qstats.overlimits++;
78                 if (!red_use_ecn(q) || !INET_ECN_set_ce(skb)) {
79                         q->stats.prob_drop++;
80                         goto congestion_drop;
81                 }
82
83                 q->stats.prob_mark++;
84                 break;
85
86         case RED_HARD_MARK:
87                 sch->qstats.overlimits++;
88                 if (red_use_harddrop(q) || !red_use_ecn(q) ||
89                     !INET_ECN_set_ce(skb)) {
90                         q->stats.forced_drop++;
91                         goto congestion_drop;
92                 }
93
94                 q->stats.forced_mark++;
95                 break;
96         }
97
98         ret = qdisc_enqueue(skb, child);
99         if (likely(ret == NET_XMIT_SUCCESS)) {
100                 sch->q.qlen++;
101         } else if (net_xmit_drop_count(ret)) {
102                 q->stats.pdrop++;
103                 sch->qstats.drops++;
104         }
105         return ret;
106
107 congestion_drop:
108         qdisc_drop(skb, sch);
109         return NET_XMIT_CN;
110 }
111
112 static struct sk_buff *red_dequeue(struct Qdisc *sch)
113 {
114         struct sk_buff *skb;
115         struct red_sched_data *q = qdisc_priv(sch);
116         struct Qdisc *child = q->qdisc;
117
118         skb = child->dequeue(child);
119         if (skb) {
120                 qdisc_bstats_update(sch, skb);
121                 sch->q.qlen--;
122         } else {
123                 if (!red_is_idling(&q->vars))
124                         red_start_of_idle_period(&q->vars);
125         }
126         return skb;
127 }
128
129 static struct sk_buff *red_peek(struct Qdisc *sch)
130 {
131         struct red_sched_data *q = qdisc_priv(sch);
132         struct Qdisc *child = q->qdisc;
133
134         return child->ops->peek(child);
135 }
136
137 static unsigned int red_drop(struct Qdisc *sch)
138 {
139         struct red_sched_data *q = qdisc_priv(sch);
140         struct Qdisc *child = q->qdisc;
141         unsigned int len;
142
143         if (child->ops->drop && (len = child->ops->drop(child)) > 0) {
144                 q->stats.other++;
145                 sch->qstats.drops++;
146                 sch->q.qlen--;
147                 return len;
148         }
149
150         if (!red_is_idling(&q->vars))
151                 red_start_of_idle_period(&q->vars);
152
153         return 0;
154 }
155
156 static void red_reset(struct Qdisc *sch)
157 {
158         struct red_sched_data *q = qdisc_priv(sch);
159
160         qdisc_reset(q->qdisc);
161         sch->q.qlen = 0;
162         red_restart(&q->vars);
163 }
164
165 static void red_destroy(struct Qdisc *sch)
166 {
167         struct red_sched_data *q = qdisc_priv(sch);
168
169         del_timer_sync(&q->adapt_timer);
170         qdisc_destroy(q->qdisc);
171 }
172
173 static const struct nla_policy red_policy[TCA_RED_MAX + 1] = {
174         [TCA_RED_PARMS] = { .len = sizeof(struct tc_red_qopt) },
175         [TCA_RED_STAB]  = { .len = RED_STAB_SIZE },
176         [TCA_RED_MAX_P] = { .type = NLA_U32 },
177 };
178
179 static int red_change(struct Qdisc *sch, struct nlattr *opt)
180 {
181         struct red_sched_data *q = qdisc_priv(sch);
182         struct nlattr *tb[TCA_RED_MAX + 1];
183         struct tc_red_qopt *ctl;
184         struct Qdisc *child = NULL;
185         int err;
186         u32 max_P;
187
188         if (opt == NULL)
189                 return -EINVAL;
190
191         err = nla_parse_nested(tb, TCA_RED_MAX, opt, red_policy);
192         if (err < 0)
193                 return err;
194
195         if (tb[TCA_RED_PARMS] == NULL ||
196             tb[TCA_RED_STAB] == NULL)
197                 return -EINVAL;
198
199         max_P = tb[TCA_RED_MAX_P] ? nla_get_u32(tb[TCA_RED_MAX_P]) : 0;
200
201         ctl = nla_data(tb[TCA_RED_PARMS]);
202
203         if (ctl->limit > 0) {
204                 child = fifo_create_dflt(sch, &bfifo_qdisc_ops, ctl->limit);
205                 if (IS_ERR(child))
206                         return PTR_ERR(child);
207         }
208
209         sch_tree_lock(sch);
210         q->flags = ctl->flags;
211         q->limit = ctl->limit;
212         if (child) {
213                 qdisc_tree_decrease_qlen(q->qdisc, q->qdisc->q.qlen);
214                 qdisc_destroy(q->qdisc);
215                 q->qdisc = child;
216         }
217
218         red_set_parms(&q->parms,
219                       ctl->qth_min, ctl->qth_max, ctl->Wlog,
220                       ctl->Plog, ctl->Scell_log,
221                       nla_data(tb[TCA_RED_STAB]),
222                       max_P);
223         red_set_vars(&q->vars);
224
225         del_timer(&q->adapt_timer);
226         if (ctl->flags & TC_RED_ADAPTATIVE)
227                 mod_timer(&q->adapt_timer, jiffies + HZ/2);
228
229         if (!q->qdisc->q.qlen)
230                 red_start_of_idle_period(&q->vars);
231
232         sch_tree_unlock(sch);
233         return 0;
234 }
235
236 static inline void red_adaptative_timer(unsigned long arg)
237 {
238         struct Qdisc *sch = (struct Qdisc *)arg;
239         struct red_sched_data *q = qdisc_priv(sch);
240         spinlock_t *root_lock = qdisc_lock(qdisc_root_sleeping(sch));
241
242         spin_lock(root_lock);
243         red_adaptative_algo(&q->parms, &q->vars);
244         mod_timer(&q->adapt_timer, jiffies + HZ/2);
245         spin_unlock(root_lock);
246 }
247
248 static int red_init(struct Qdisc *sch, struct nlattr *opt)
249 {
250         struct red_sched_data *q = qdisc_priv(sch);
251
252         q->qdisc = &noop_qdisc;
253         setup_timer(&q->adapt_timer, red_adaptative_timer, (unsigned long)sch);
254         return red_change(sch, opt);
255 }
256
257 static int red_dump(struct Qdisc *sch, struct sk_buff *skb)
258 {
259         struct red_sched_data *q = qdisc_priv(sch);
260         struct nlattr *opts = NULL;
261         struct tc_red_qopt opt = {
262                 .limit          = q->limit,
263                 .flags          = q->flags,
264                 .qth_min        = q->parms.qth_min >> q->parms.Wlog,
265                 .qth_max        = q->parms.qth_max >> q->parms.Wlog,
266                 .Wlog           = q->parms.Wlog,
267                 .Plog           = q->parms.Plog,
268                 .Scell_log      = q->parms.Scell_log,
269         };
270
271         sch->qstats.backlog = q->qdisc->qstats.backlog;
272         opts = nla_nest_start(skb, TCA_OPTIONS);
273         if (opts == NULL)
274                 goto nla_put_failure;
275         if (nla_put(skb, TCA_RED_PARMS, sizeof(opt), &opt) ||
276             nla_put_u32(skb, TCA_RED_MAX_P, q->parms.max_P))
277                 goto nla_put_failure;
278         return nla_nest_end(skb, opts);
279
280 nla_put_failure:
281         nla_nest_cancel(skb, opts);
282         return -EMSGSIZE;
283 }
284
285 static int red_dump_stats(struct Qdisc *sch, struct gnet_dump *d)
286 {
287         struct red_sched_data *q = qdisc_priv(sch);
288         struct tc_red_xstats st = {
289                 .early  = q->stats.prob_drop + q->stats.forced_drop,
290                 .pdrop  = q->stats.pdrop,
291                 .other  = q->stats.other,
292                 .marked = q->stats.prob_mark + q->stats.forced_mark,
293         };
294
295         return gnet_stats_copy_app(d, &st, sizeof(st));
296 }
297
298 static int red_dump_class(struct Qdisc *sch, unsigned long cl,
299                           struct sk_buff *skb, struct tcmsg *tcm)
300 {
301         struct red_sched_data *q = qdisc_priv(sch);
302
303         tcm->tcm_handle |= TC_H_MIN(1);
304         tcm->tcm_info = q->qdisc->handle;
305         return 0;
306 }
307
308 static int red_graft(struct Qdisc *sch, unsigned long arg, struct Qdisc *new,
309                      struct Qdisc **old)
310 {
311         struct red_sched_data *q = qdisc_priv(sch);
312
313         if (new == NULL)
314                 new = &noop_qdisc;
315
316         sch_tree_lock(sch);
317         *old = q->qdisc;
318         q->qdisc = new;
319         qdisc_tree_decrease_qlen(*old, (*old)->q.qlen);
320         qdisc_reset(*old);
321         sch_tree_unlock(sch);
322         return 0;
323 }
324
325 static struct Qdisc *red_leaf(struct Qdisc *sch, unsigned long arg)
326 {
327         struct red_sched_data *q = qdisc_priv(sch);
328         return q->qdisc;
329 }
330
331 static unsigned long red_get(struct Qdisc *sch, u32 classid)
332 {
333         return 1;
334 }
335
336 static void red_put(struct Qdisc *sch, unsigned long arg)
337 {
338 }
339
340 static void red_walk(struct Qdisc *sch, struct qdisc_walker *walker)
341 {
342         if (!walker->stop) {
343                 if (walker->count >= walker->skip)
344                         if (walker->fn(sch, 1, walker) < 0) {
345                                 walker->stop = 1;
346                                 return;
347                         }
348                 walker->count++;
349         }
350 }
351
352 static const struct Qdisc_class_ops red_class_ops = {
353         .graft          =       red_graft,
354         .leaf           =       red_leaf,
355         .get            =       red_get,
356         .put            =       red_put,
357         .walk           =       red_walk,
358         .dump           =       red_dump_class,
359 };
360
361 static struct Qdisc_ops red_qdisc_ops __read_mostly = {
362         .id             =       "red",
363         .priv_size      =       sizeof(struct red_sched_data),
364         .cl_ops         =       &red_class_ops,
365         .enqueue        =       red_enqueue,
366         .dequeue        =       red_dequeue,
367         .peek           =       red_peek,
368         .drop           =       red_drop,
369         .init           =       red_init,
370         .reset          =       red_reset,
371         .destroy        =       red_destroy,
372         .change         =       red_change,
373         .dump           =       red_dump,
374         .dump_stats     =       red_dump_stats,
375         .owner          =       THIS_MODULE,
376 };
377
378 static int __init red_module_init(void)
379 {
380         return register_qdisc(&red_qdisc_ops);
381 }
382
383 static void __exit red_module_exit(void)
384 {
385         unregister_qdisc(&red_qdisc_ops);
386 }
387
388 module_init(red_module_init)
389 module_exit(red_module_exit)
390
391 MODULE_LICENSE("GPL");