Loading...
1/*
2 * The Kyber I/O scheduler. Controls latency by throttling queue depths using
3 * scalable techniques.
4 *
5 * Copyright (C) 2017 Facebook
6 *
7 * This program is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU General Public
9 * License v2 as published by the Free Software Foundation.
10 *
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * General Public License for more details.
15 *
16 * You should have received a copy of the GNU General Public License
17 * along with this program. If not, see <https://www.gnu.org/licenses/>.
18 */
19
20#include <linux/kernel.h>
21#include <linux/blkdev.h>
22#include <linux/blk-mq.h>
23#include <linux/elevator.h>
24#include <linux/module.h>
25#include <linux/sbitmap.h>
26
27#include "blk.h"
28#include "blk-mq.h"
29#include "blk-mq-debugfs.h"
30#include "blk-mq-sched.h"
31#include "blk-mq-tag.h"
32#include "blk-stat.h"
33
34/* Scheduling domains. */
35enum {
36 KYBER_READ,
37 KYBER_SYNC_WRITE,
38 KYBER_OTHER, /* Async writes, discard, etc. */
39 KYBER_NUM_DOMAINS,
40};
41
42enum {
43 KYBER_MIN_DEPTH = 256,
44
45 /*
46 * In order to prevent starvation of synchronous requests by a flood of
47 * asynchronous requests, we reserve 25% of requests for synchronous
48 * operations.
49 */
50 KYBER_ASYNC_PERCENT = 75,
51};
52
53/*
54 * Initial device-wide depths for each scheduling domain.
55 *
56 * Even for fast devices with lots of tags like NVMe, you can saturate
57 * the device with only a fraction of the maximum possible queue depth.
58 * So, we cap these to a reasonable value.
59 */
60static const unsigned int kyber_depth[] = {
61 [KYBER_READ] = 256,
62 [KYBER_SYNC_WRITE] = 128,
63 [KYBER_OTHER] = 64,
64};
65
66/*
67 * Scheduling domain batch sizes. We favor reads.
68 */
69static const unsigned int kyber_batch_size[] = {
70 [KYBER_READ] = 16,
71 [KYBER_SYNC_WRITE] = 8,
72 [KYBER_OTHER] = 8,
73};
74
75struct kyber_queue_data {
76 struct request_queue *q;
77
78 struct blk_stat_callback *cb;
79
80 /*
81 * The device is divided into multiple scheduling domains based on the
82 * request type. Each domain has a fixed number of in-flight requests of
83 * that type device-wide, limited by these tokens.
84 */
85 struct sbitmap_queue domain_tokens[KYBER_NUM_DOMAINS];
86
87 /*
88 * Async request percentage, converted to per-word depth for
89 * sbitmap_get_shallow().
90 */
91 unsigned int async_depth;
92
93 /* Target latencies in nanoseconds. */
94 u64 read_lat_nsec, write_lat_nsec;
95};
96
97struct kyber_hctx_data {
98 spinlock_t lock;
99 struct list_head rqs[KYBER_NUM_DOMAINS];
100 unsigned int cur_domain;
101 unsigned int batching;
102 wait_queue_entry_t domain_wait[KYBER_NUM_DOMAINS];
103 struct sbq_wait_state *domain_ws[KYBER_NUM_DOMAINS];
104 atomic_t wait_index[KYBER_NUM_DOMAINS];
105};
106
107static int kyber_domain_wake(wait_queue_entry_t *wait, unsigned mode, int flags,
108 void *key);
109
110static int rq_sched_domain(const struct request *rq)
111{
112 unsigned int op = rq->cmd_flags;
113
114 if ((op & REQ_OP_MASK) == REQ_OP_READ)
115 return KYBER_READ;
116 else if ((op & REQ_OP_MASK) == REQ_OP_WRITE && op_is_sync(op))
117 return KYBER_SYNC_WRITE;
118 else
119 return KYBER_OTHER;
120}
121
122enum {
123 NONE = 0,
124 GOOD = 1,
125 GREAT = 2,
126 BAD = -1,
127 AWFUL = -2,
128};
129
130#define IS_GOOD(status) ((status) > 0)
131#define IS_BAD(status) ((status) < 0)
132
133static int kyber_lat_status(struct blk_stat_callback *cb,
134 unsigned int sched_domain, u64 target)
135{
136 u64 latency;
137
138 if (!cb->stat[sched_domain].nr_samples)
139 return NONE;
140
141 latency = cb->stat[sched_domain].mean;
142 if (latency >= 2 * target)
143 return AWFUL;
144 else if (latency > target)
145 return BAD;
146 else if (latency <= target / 2)
147 return GREAT;
148 else /* (latency <= target) */
149 return GOOD;
150}
151
152/*
153 * Adjust the read or synchronous write depth given the status of reads and
154 * writes. The goal is that the latencies of the two domains are fair (i.e., if
155 * one is good, then the other is good).
156 */
157static void kyber_adjust_rw_depth(struct kyber_queue_data *kqd,
158 unsigned int sched_domain, int this_status,
159 int other_status)
160{
161 unsigned int orig_depth, depth;
162
163 /*
164 * If this domain had no samples, or reads and writes are both good or
165 * both bad, don't adjust the depth.
166 */
167 if (this_status == NONE ||
168 (IS_GOOD(this_status) && IS_GOOD(other_status)) ||
169 (IS_BAD(this_status) && IS_BAD(other_status)))
170 return;
171
172 orig_depth = depth = kqd->domain_tokens[sched_domain].sb.depth;
173
174 if (other_status == NONE) {
175 depth++;
176 } else {
177 switch (this_status) {
178 case GOOD:
179 if (other_status == AWFUL)
180 depth -= max(depth / 4, 1U);
181 else
182 depth -= max(depth / 8, 1U);
183 break;
184 case GREAT:
185 if (other_status == AWFUL)
186 depth /= 2;
187 else
188 depth -= max(depth / 4, 1U);
189 break;
190 case BAD:
191 depth++;
192 break;
193 case AWFUL:
194 if (other_status == GREAT)
195 depth += 2;
196 else
197 depth++;
198 break;
199 }
200 }
201
202 depth = clamp(depth, 1U, kyber_depth[sched_domain]);
203 if (depth != orig_depth)
204 sbitmap_queue_resize(&kqd->domain_tokens[sched_domain], depth);
205}
206
207/*
208 * Adjust the depth of other requests given the status of reads and synchronous
209 * writes. As long as either domain is doing fine, we don't throttle, but if
210 * both domains are doing badly, we throttle heavily.
211 */
212static void kyber_adjust_other_depth(struct kyber_queue_data *kqd,
213 int read_status, int write_status,
214 bool have_samples)
215{
216 unsigned int orig_depth, depth;
217 int status;
218
219 orig_depth = depth = kqd->domain_tokens[KYBER_OTHER].sb.depth;
220
221 if (read_status == NONE && write_status == NONE) {
222 depth += 2;
223 } else if (have_samples) {
224 if (read_status == NONE)
225 status = write_status;
226 else if (write_status == NONE)
227 status = read_status;
228 else
229 status = max(read_status, write_status);
230 switch (status) {
231 case GREAT:
232 depth += 2;
233 break;
234 case GOOD:
235 depth++;
236 break;
237 case BAD:
238 depth -= max(depth / 4, 1U);
239 break;
240 case AWFUL:
241 depth /= 2;
242 break;
243 }
244 }
245
246 depth = clamp(depth, 1U, kyber_depth[KYBER_OTHER]);
247 if (depth != orig_depth)
248 sbitmap_queue_resize(&kqd->domain_tokens[KYBER_OTHER], depth);
249}
250
251/*
252 * Apply heuristics for limiting queue depths based on gathered latency
253 * statistics.
254 */
255static void kyber_stat_timer_fn(struct blk_stat_callback *cb)
256{
257 struct kyber_queue_data *kqd = cb->data;
258 int read_status, write_status;
259
260 read_status = kyber_lat_status(cb, KYBER_READ, kqd->read_lat_nsec);
261 write_status = kyber_lat_status(cb, KYBER_SYNC_WRITE, kqd->write_lat_nsec);
262
263 kyber_adjust_rw_depth(kqd, KYBER_READ, read_status, write_status);
264 kyber_adjust_rw_depth(kqd, KYBER_SYNC_WRITE, write_status, read_status);
265 kyber_adjust_other_depth(kqd, read_status, write_status,
266 cb->stat[KYBER_OTHER].nr_samples != 0);
267
268 /*
269 * Continue monitoring latencies if we aren't hitting the targets or
270 * we're still throttling other requests.
271 */
272 if (!blk_stat_is_active(kqd->cb) &&
273 ((IS_BAD(read_status) || IS_BAD(write_status) ||
274 kqd->domain_tokens[KYBER_OTHER].sb.depth < kyber_depth[KYBER_OTHER])))
275 blk_stat_activate_msecs(kqd->cb, 100);
276}
277
278static unsigned int kyber_sched_tags_shift(struct kyber_queue_data *kqd)
279{
280 /*
281 * All of the hardware queues have the same depth, so we can just grab
282 * the shift of the first one.
283 */
284 return kqd->q->queue_hw_ctx[0]->sched_tags->bitmap_tags.sb.shift;
285}
286
287static struct kyber_queue_data *kyber_queue_data_alloc(struct request_queue *q)
288{
289 struct kyber_queue_data *kqd;
290 unsigned int max_tokens;
291 unsigned int shift;
292 int ret = -ENOMEM;
293 int i;
294
295 kqd = kmalloc_node(sizeof(*kqd), GFP_KERNEL, q->node);
296 if (!kqd)
297 goto err;
298 kqd->q = q;
299
300 kqd->cb = blk_stat_alloc_callback(kyber_stat_timer_fn, rq_sched_domain,
301 KYBER_NUM_DOMAINS, kqd);
302 if (!kqd->cb)
303 goto err_kqd;
304
305 /*
306 * The maximum number of tokens for any scheduling domain is at least
307 * the queue depth of a single hardware queue. If the hardware doesn't
308 * have many tags, still provide a reasonable number.
309 */
310 max_tokens = max_t(unsigned int, q->tag_set->queue_depth,
311 KYBER_MIN_DEPTH);
312 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
313 WARN_ON(!kyber_depth[i]);
314 WARN_ON(!kyber_batch_size[i]);
315 ret = sbitmap_queue_init_node(&kqd->domain_tokens[i],
316 max_tokens, -1, false, GFP_KERNEL,
317 q->node);
318 if (ret) {
319 while (--i >= 0)
320 sbitmap_queue_free(&kqd->domain_tokens[i]);
321 goto err_cb;
322 }
323 sbitmap_queue_resize(&kqd->domain_tokens[i], kyber_depth[i]);
324 }
325
326 shift = kyber_sched_tags_shift(kqd);
327 kqd->async_depth = (1U << shift) * KYBER_ASYNC_PERCENT / 100U;
328
329 kqd->read_lat_nsec = 2000000ULL;
330 kqd->write_lat_nsec = 10000000ULL;
331
332 return kqd;
333
334err_cb:
335 blk_stat_free_callback(kqd->cb);
336err_kqd:
337 kfree(kqd);
338err:
339 return ERR_PTR(ret);
340}
341
342static int kyber_init_sched(struct request_queue *q, struct elevator_type *e)
343{
344 struct kyber_queue_data *kqd;
345 struct elevator_queue *eq;
346
347 eq = elevator_alloc(q, e);
348 if (!eq)
349 return -ENOMEM;
350
351 kqd = kyber_queue_data_alloc(q);
352 if (IS_ERR(kqd)) {
353 kobject_put(&eq->kobj);
354 return PTR_ERR(kqd);
355 }
356
357 eq->elevator_data = kqd;
358 q->elevator = eq;
359
360 blk_stat_add_callback(q, kqd->cb);
361
362 return 0;
363}
364
365static void kyber_exit_sched(struct elevator_queue *e)
366{
367 struct kyber_queue_data *kqd = e->elevator_data;
368 struct request_queue *q = kqd->q;
369 int i;
370
371 blk_stat_remove_callback(q, kqd->cb);
372
373 for (i = 0; i < KYBER_NUM_DOMAINS; i++)
374 sbitmap_queue_free(&kqd->domain_tokens[i]);
375 blk_stat_free_callback(kqd->cb);
376 kfree(kqd);
377}
378
379static int kyber_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
380{
381 struct kyber_hctx_data *khd;
382 int i;
383
384 khd = kmalloc_node(sizeof(*khd), GFP_KERNEL, hctx->numa_node);
385 if (!khd)
386 return -ENOMEM;
387
388 spin_lock_init(&khd->lock);
389
390 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
391 INIT_LIST_HEAD(&khd->rqs[i]);
392 init_waitqueue_func_entry(&khd->domain_wait[i],
393 kyber_domain_wake);
394 khd->domain_wait[i].private = hctx;
395 INIT_LIST_HEAD(&khd->domain_wait[i].entry);
396 atomic_set(&khd->wait_index[i], 0);
397 }
398
399 khd->cur_domain = 0;
400 khd->batching = 0;
401
402 hctx->sched_data = khd;
403
404 return 0;
405}
406
407static void kyber_exit_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
408{
409 kfree(hctx->sched_data);
410}
411
412static int rq_get_domain_token(struct request *rq)
413{
414 return (long)rq->elv.priv[0];
415}
416
417static void rq_set_domain_token(struct request *rq, int token)
418{
419 rq->elv.priv[0] = (void *)(long)token;
420}
421
422static void rq_clear_domain_token(struct kyber_queue_data *kqd,
423 struct request *rq)
424{
425 unsigned int sched_domain;
426 int nr;
427
428 nr = rq_get_domain_token(rq);
429 if (nr != -1) {
430 sched_domain = rq_sched_domain(rq);
431 sbitmap_queue_clear(&kqd->domain_tokens[sched_domain], nr,
432 rq->mq_ctx->cpu);
433 }
434}
435
436static void kyber_limit_depth(unsigned int op, struct blk_mq_alloc_data *data)
437{
438 /*
439 * We use the scheduler tags as per-hardware queue queueing tokens.
440 * Async requests can be limited at this stage.
441 */
442 if (!op_is_sync(op)) {
443 struct kyber_queue_data *kqd = data->q->elevator->elevator_data;
444
445 data->shallow_depth = kqd->async_depth;
446 }
447}
448
449static void kyber_prepare_request(struct request *rq, struct bio *bio)
450{
451 rq_set_domain_token(rq, -1);
452}
453
454static void kyber_finish_request(struct request *rq)
455{
456 struct kyber_queue_data *kqd = rq->q->elevator->elevator_data;
457
458 rq_clear_domain_token(kqd, rq);
459}
460
461static void kyber_completed_request(struct request *rq)
462{
463 struct request_queue *q = rq->q;
464 struct kyber_queue_data *kqd = q->elevator->elevator_data;
465 unsigned int sched_domain;
466 u64 now, latency, target;
467
468 /*
469 * Check if this request met our latency goal. If not, quickly gather
470 * some statistics and start throttling.
471 */
472 sched_domain = rq_sched_domain(rq);
473 switch (sched_domain) {
474 case KYBER_READ:
475 target = kqd->read_lat_nsec;
476 break;
477 case KYBER_SYNC_WRITE:
478 target = kqd->write_lat_nsec;
479 break;
480 default:
481 return;
482 }
483
484 /* If we are already monitoring latencies, don't check again. */
485 if (blk_stat_is_active(kqd->cb))
486 return;
487
488 now = __blk_stat_time(ktime_to_ns(ktime_get()));
489 if (now < blk_stat_time(&rq->issue_stat))
490 return;
491
492 latency = now - blk_stat_time(&rq->issue_stat);
493
494 if (latency > target)
495 blk_stat_activate_msecs(kqd->cb, 10);
496}
497
498static void kyber_flush_busy_ctxs(struct kyber_hctx_data *khd,
499 struct blk_mq_hw_ctx *hctx)
500{
501 LIST_HEAD(rq_list);
502 struct request *rq, *next;
503
504 blk_mq_flush_busy_ctxs(hctx, &rq_list);
505 list_for_each_entry_safe(rq, next, &rq_list, queuelist) {
506 unsigned int sched_domain;
507
508 sched_domain = rq_sched_domain(rq);
509 list_move_tail(&rq->queuelist, &khd->rqs[sched_domain]);
510 }
511}
512
513static int kyber_domain_wake(wait_queue_entry_t *wait, unsigned mode, int flags,
514 void *key)
515{
516 struct blk_mq_hw_ctx *hctx = READ_ONCE(wait->private);
517
518 list_del_init(&wait->entry);
519 blk_mq_run_hw_queue(hctx, true);
520 return 1;
521}
522
523static int kyber_get_domain_token(struct kyber_queue_data *kqd,
524 struct kyber_hctx_data *khd,
525 struct blk_mq_hw_ctx *hctx)
526{
527 unsigned int sched_domain = khd->cur_domain;
528 struct sbitmap_queue *domain_tokens = &kqd->domain_tokens[sched_domain];
529 wait_queue_entry_t *wait = &khd->domain_wait[sched_domain];
530 struct sbq_wait_state *ws;
531 int nr;
532
533 nr = __sbitmap_queue_get(domain_tokens);
534
535 /*
536 * If we failed to get a domain token, make sure the hardware queue is
537 * run when one becomes available. Note that this is serialized on
538 * khd->lock, but we still need to be careful about the waker.
539 */
540 if (nr < 0 && list_empty_careful(&wait->entry)) {
541 ws = sbq_wait_ptr(domain_tokens,
542 &khd->wait_index[sched_domain]);
543 khd->domain_ws[sched_domain] = ws;
544 add_wait_queue(&ws->wait, wait);
545
546 /*
547 * Try again in case a token was freed before we got on the wait
548 * queue.
549 */
550 nr = __sbitmap_queue_get(domain_tokens);
551 }
552
553 /*
554 * If we got a token while we were on the wait queue, remove ourselves
555 * from the wait queue to ensure that all wake ups make forward
556 * progress. It's possible that the waker already deleted the entry
557 * between the !list_empty_careful() check and us grabbing the lock, but
558 * list_del_init() is okay with that.
559 */
560 if (nr >= 0 && !list_empty_careful(&wait->entry)) {
561 ws = khd->domain_ws[sched_domain];
562 spin_lock_irq(&ws->wait.lock);
563 list_del_init(&wait->entry);
564 spin_unlock_irq(&ws->wait.lock);
565 }
566
567 return nr;
568}
569
570static struct request *
571kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
572 struct kyber_hctx_data *khd,
573 struct blk_mq_hw_ctx *hctx,
574 bool *flushed)
575{
576 struct list_head *rqs;
577 struct request *rq;
578 int nr;
579
580 rqs = &khd->rqs[khd->cur_domain];
581 rq = list_first_entry_or_null(rqs, struct request, queuelist);
582
583 /*
584 * If there wasn't already a pending request and we haven't flushed the
585 * software queues yet, flush the software queues and check again.
586 */
587 if (!rq && !*flushed) {
588 kyber_flush_busy_ctxs(khd, hctx);
589 *flushed = true;
590 rq = list_first_entry_or_null(rqs, struct request, queuelist);
591 }
592
593 if (rq) {
594 nr = kyber_get_domain_token(kqd, khd, hctx);
595 if (nr >= 0) {
596 khd->batching++;
597 rq_set_domain_token(rq, nr);
598 list_del_init(&rq->queuelist);
599 return rq;
600 }
601 }
602
603 /* There were either no pending requests or no tokens. */
604 return NULL;
605}
606
607static struct request *kyber_dispatch_request(struct blk_mq_hw_ctx *hctx)
608{
609 struct kyber_queue_data *kqd = hctx->queue->elevator->elevator_data;
610 struct kyber_hctx_data *khd = hctx->sched_data;
611 bool flushed = false;
612 struct request *rq;
613 int i;
614
615 spin_lock(&khd->lock);
616
617 /*
618 * First, if we are still entitled to batch, try to dispatch a request
619 * from the batch.
620 */
621 if (khd->batching < kyber_batch_size[khd->cur_domain]) {
622 rq = kyber_dispatch_cur_domain(kqd, khd, hctx, &flushed);
623 if (rq)
624 goto out;
625 }
626
627 /*
628 * Either,
629 * 1. We were no longer entitled to a batch.
630 * 2. The domain we were batching didn't have any requests.
631 * 3. The domain we were batching was out of tokens.
632 *
633 * Start another batch. Note that this wraps back around to the original
634 * domain if no other domains have requests or tokens.
635 */
636 khd->batching = 0;
637 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
638 if (khd->cur_domain == KYBER_NUM_DOMAINS - 1)
639 khd->cur_domain = 0;
640 else
641 khd->cur_domain++;
642
643 rq = kyber_dispatch_cur_domain(kqd, khd, hctx, &flushed);
644 if (rq)
645 goto out;
646 }
647
648 rq = NULL;
649out:
650 spin_unlock(&khd->lock);
651 return rq;
652}
653
654static bool kyber_has_work(struct blk_mq_hw_ctx *hctx)
655{
656 struct kyber_hctx_data *khd = hctx->sched_data;
657 int i;
658
659 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
660 if (!list_empty_careful(&khd->rqs[i]))
661 return true;
662 }
663 return sbitmap_any_bit_set(&hctx->ctx_map);
664}
665
666#define KYBER_LAT_SHOW_STORE(op) \
667static ssize_t kyber_##op##_lat_show(struct elevator_queue *e, \
668 char *page) \
669{ \
670 struct kyber_queue_data *kqd = e->elevator_data; \
671 \
672 return sprintf(page, "%llu\n", kqd->op##_lat_nsec); \
673} \
674 \
675static ssize_t kyber_##op##_lat_store(struct elevator_queue *e, \
676 const char *page, size_t count) \
677{ \
678 struct kyber_queue_data *kqd = e->elevator_data; \
679 unsigned long long nsec; \
680 int ret; \
681 \
682 ret = kstrtoull(page, 10, &nsec); \
683 if (ret) \
684 return ret; \
685 \
686 kqd->op##_lat_nsec = nsec; \
687 \
688 return count; \
689}
690KYBER_LAT_SHOW_STORE(read);
691KYBER_LAT_SHOW_STORE(write);
692#undef KYBER_LAT_SHOW_STORE
693
694#define KYBER_LAT_ATTR(op) __ATTR(op##_lat_nsec, 0644, kyber_##op##_lat_show, kyber_##op##_lat_store)
695static struct elv_fs_entry kyber_sched_attrs[] = {
696 KYBER_LAT_ATTR(read),
697 KYBER_LAT_ATTR(write),
698 __ATTR_NULL
699};
700#undef KYBER_LAT_ATTR
701
702#ifdef CONFIG_BLK_DEBUG_FS
703#define KYBER_DEBUGFS_DOMAIN_ATTRS(domain, name) \
704static int kyber_##name##_tokens_show(void *data, struct seq_file *m) \
705{ \
706 struct request_queue *q = data; \
707 struct kyber_queue_data *kqd = q->elevator->elevator_data; \
708 \
709 sbitmap_queue_show(&kqd->domain_tokens[domain], m); \
710 return 0; \
711} \
712 \
713static void *kyber_##name##_rqs_start(struct seq_file *m, loff_t *pos) \
714 __acquires(&khd->lock) \
715{ \
716 struct blk_mq_hw_ctx *hctx = m->private; \
717 struct kyber_hctx_data *khd = hctx->sched_data; \
718 \
719 spin_lock(&khd->lock); \
720 return seq_list_start(&khd->rqs[domain], *pos); \
721} \
722 \
723static void *kyber_##name##_rqs_next(struct seq_file *m, void *v, \
724 loff_t *pos) \
725{ \
726 struct blk_mq_hw_ctx *hctx = m->private; \
727 struct kyber_hctx_data *khd = hctx->sched_data; \
728 \
729 return seq_list_next(v, &khd->rqs[domain], pos); \
730} \
731 \
732static void kyber_##name##_rqs_stop(struct seq_file *m, void *v) \
733 __releases(&khd->lock) \
734{ \
735 struct blk_mq_hw_ctx *hctx = m->private; \
736 struct kyber_hctx_data *khd = hctx->sched_data; \
737 \
738 spin_unlock(&khd->lock); \
739} \
740 \
741static const struct seq_operations kyber_##name##_rqs_seq_ops = { \
742 .start = kyber_##name##_rqs_start, \
743 .next = kyber_##name##_rqs_next, \
744 .stop = kyber_##name##_rqs_stop, \
745 .show = blk_mq_debugfs_rq_show, \
746}; \
747 \
748static int kyber_##name##_waiting_show(void *data, struct seq_file *m) \
749{ \
750 struct blk_mq_hw_ctx *hctx = data; \
751 struct kyber_hctx_data *khd = hctx->sched_data; \
752 wait_queue_entry_t *wait = &khd->domain_wait[domain]; \
753 \
754 seq_printf(m, "%d\n", !list_empty_careful(&wait->entry)); \
755 return 0; \
756}
757KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_READ, read)
758KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_SYNC_WRITE, sync_write)
759KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_OTHER, other)
760#undef KYBER_DEBUGFS_DOMAIN_ATTRS
761
762static int kyber_async_depth_show(void *data, struct seq_file *m)
763{
764 struct request_queue *q = data;
765 struct kyber_queue_data *kqd = q->elevator->elevator_data;
766
767 seq_printf(m, "%u\n", kqd->async_depth);
768 return 0;
769}
770
771static int kyber_cur_domain_show(void *data, struct seq_file *m)
772{
773 struct blk_mq_hw_ctx *hctx = data;
774 struct kyber_hctx_data *khd = hctx->sched_data;
775
776 switch (khd->cur_domain) {
777 case KYBER_READ:
778 seq_puts(m, "READ\n");
779 break;
780 case KYBER_SYNC_WRITE:
781 seq_puts(m, "SYNC_WRITE\n");
782 break;
783 case KYBER_OTHER:
784 seq_puts(m, "OTHER\n");
785 break;
786 default:
787 seq_printf(m, "%u\n", khd->cur_domain);
788 break;
789 }
790 return 0;
791}
792
793static int kyber_batching_show(void *data, struct seq_file *m)
794{
795 struct blk_mq_hw_ctx *hctx = data;
796 struct kyber_hctx_data *khd = hctx->sched_data;
797
798 seq_printf(m, "%u\n", khd->batching);
799 return 0;
800}
801
802#define KYBER_QUEUE_DOMAIN_ATTRS(name) \
803 {#name "_tokens", 0400, kyber_##name##_tokens_show}
804static const struct blk_mq_debugfs_attr kyber_queue_debugfs_attrs[] = {
805 KYBER_QUEUE_DOMAIN_ATTRS(read),
806 KYBER_QUEUE_DOMAIN_ATTRS(sync_write),
807 KYBER_QUEUE_DOMAIN_ATTRS(other),
808 {"async_depth", 0400, kyber_async_depth_show},
809 {},
810};
811#undef KYBER_QUEUE_DOMAIN_ATTRS
812
813#define KYBER_HCTX_DOMAIN_ATTRS(name) \
814 {#name "_rqs", 0400, .seq_ops = &kyber_##name##_rqs_seq_ops}, \
815 {#name "_waiting", 0400, kyber_##name##_waiting_show}
816static const struct blk_mq_debugfs_attr kyber_hctx_debugfs_attrs[] = {
817 KYBER_HCTX_DOMAIN_ATTRS(read),
818 KYBER_HCTX_DOMAIN_ATTRS(sync_write),
819 KYBER_HCTX_DOMAIN_ATTRS(other),
820 {"cur_domain", 0400, kyber_cur_domain_show},
821 {"batching", 0400, kyber_batching_show},
822 {},
823};
824#undef KYBER_HCTX_DOMAIN_ATTRS
825#endif
826
827static struct elevator_type kyber_sched = {
828 .ops.mq = {
829 .init_sched = kyber_init_sched,
830 .exit_sched = kyber_exit_sched,
831 .init_hctx = kyber_init_hctx,
832 .exit_hctx = kyber_exit_hctx,
833 .limit_depth = kyber_limit_depth,
834 .prepare_request = kyber_prepare_request,
835 .finish_request = kyber_finish_request,
836 .requeue_request = kyber_finish_request,
837 .completed_request = kyber_completed_request,
838 .dispatch_request = kyber_dispatch_request,
839 .has_work = kyber_has_work,
840 },
841 .uses_mq = true,
842#ifdef CONFIG_BLK_DEBUG_FS
843 .queue_debugfs_attrs = kyber_queue_debugfs_attrs,
844 .hctx_debugfs_attrs = kyber_hctx_debugfs_attrs,
845#endif
846 .elevator_attrs = kyber_sched_attrs,
847 .elevator_name = "kyber",
848 .elevator_owner = THIS_MODULE,
849};
850
851static int __init kyber_init(void)
852{
853 return elv_register(&kyber_sched);
854}
855
856static void __exit kyber_exit(void)
857{
858 elv_unregister(&kyber_sched);
859}
860
861module_init(kyber_init);
862module_exit(kyber_exit);
863
864MODULE_AUTHOR("Omar Sandoval");
865MODULE_LICENSE("GPL");
866MODULE_DESCRIPTION("Kyber I/O scheduler");
1// SPDX-License-Identifier: GPL-2.0
2/*
3 * The Kyber I/O scheduler. Controls latency by throttling queue depths using
4 * scalable techniques.
5 *
6 * Copyright (C) 2017 Facebook
7 */
8
9#include <linux/kernel.h>
10#include <linux/blkdev.h>
11#include <linux/module.h>
12#include <linux/sbitmap.h>
13
14#include <trace/events/block.h>
15
16#include "elevator.h"
17#include "blk.h"
18#include "blk-mq.h"
19#include "blk-mq-debugfs.h"
20#include "blk-mq-sched.h"
21
22#define CREATE_TRACE_POINTS
23#include <trace/events/kyber.h>
24
25/*
26 * Scheduling domains: the device is divided into multiple domains based on the
27 * request type.
28 */
29enum {
30 KYBER_READ,
31 KYBER_WRITE,
32 KYBER_DISCARD,
33 KYBER_OTHER,
34 KYBER_NUM_DOMAINS,
35};
36
37static const char *kyber_domain_names[] = {
38 [KYBER_READ] = "READ",
39 [KYBER_WRITE] = "WRITE",
40 [KYBER_DISCARD] = "DISCARD",
41 [KYBER_OTHER] = "OTHER",
42};
43
44enum {
45 /*
46 * In order to prevent starvation of synchronous requests by a flood of
47 * asynchronous requests, we reserve 25% of requests for synchronous
48 * operations.
49 */
50 KYBER_ASYNC_PERCENT = 75,
51};
52
53/*
54 * Maximum device-wide depth for each scheduling domain.
55 *
56 * Even for fast devices with lots of tags like NVMe, you can saturate the
57 * device with only a fraction of the maximum possible queue depth. So, we cap
58 * these to a reasonable value.
59 */
60static const unsigned int kyber_depth[] = {
61 [KYBER_READ] = 256,
62 [KYBER_WRITE] = 128,
63 [KYBER_DISCARD] = 64,
64 [KYBER_OTHER] = 16,
65};
66
67/*
68 * Default latency targets for each scheduling domain.
69 */
70static const u64 kyber_latency_targets[] = {
71 [KYBER_READ] = 2ULL * NSEC_PER_MSEC,
72 [KYBER_WRITE] = 10ULL * NSEC_PER_MSEC,
73 [KYBER_DISCARD] = 5ULL * NSEC_PER_SEC,
74};
75
76/*
77 * Batch size (number of requests we'll dispatch in a row) for each scheduling
78 * domain.
79 */
80static const unsigned int kyber_batch_size[] = {
81 [KYBER_READ] = 16,
82 [KYBER_WRITE] = 8,
83 [KYBER_DISCARD] = 1,
84 [KYBER_OTHER] = 1,
85};
86
87/*
88 * Requests latencies are recorded in a histogram with buckets defined relative
89 * to the target latency:
90 *
91 * <= 1/4 * target latency
92 * <= 1/2 * target latency
93 * <= 3/4 * target latency
94 * <= target latency
95 * <= 1 1/4 * target latency
96 * <= 1 1/2 * target latency
97 * <= 1 3/4 * target latency
98 * > 1 3/4 * target latency
99 */
100enum {
101 /*
102 * The width of the latency histogram buckets is
103 * 1 / (1 << KYBER_LATENCY_SHIFT) * target latency.
104 */
105 KYBER_LATENCY_SHIFT = 2,
106 /*
107 * The first (1 << KYBER_LATENCY_SHIFT) buckets are <= target latency,
108 * thus, "good".
109 */
110 KYBER_GOOD_BUCKETS = 1 << KYBER_LATENCY_SHIFT,
111 /* There are also (1 << KYBER_LATENCY_SHIFT) "bad" buckets. */
112 KYBER_LATENCY_BUCKETS = 2 << KYBER_LATENCY_SHIFT,
113};
114
115/*
116 * We measure both the total latency and the I/O latency (i.e., latency after
117 * submitting to the device).
118 */
119enum {
120 KYBER_TOTAL_LATENCY,
121 KYBER_IO_LATENCY,
122};
123
124static const char *kyber_latency_type_names[] = {
125 [KYBER_TOTAL_LATENCY] = "total",
126 [KYBER_IO_LATENCY] = "I/O",
127};
128
129/*
130 * Per-cpu latency histograms: total latency and I/O latency for each scheduling
131 * domain except for KYBER_OTHER.
132 */
133struct kyber_cpu_latency {
134 atomic_t buckets[KYBER_OTHER][2][KYBER_LATENCY_BUCKETS];
135};
136
137/*
138 * There is a same mapping between ctx & hctx and kcq & khd,
139 * we use request->mq_ctx->index_hw to index the kcq in khd.
140 */
141struct kyber_ctx_queue {
142 /*
143 * Used to ensure operations on rq_list and kcq_map to be an atmoic one.
144 * Also protect the rqs on rq_list when merge.
145 */
146 spinlock_t lock;
147 struct list_head rq_list[KYBER_NUM_DOMAINS];
148} ____cacheline_aligned_in_smp;
149
150struct kyber_queue_data {
151 struct request_queue *q;
152 dev_t dev;
153
154 /*
155 * Each scheduling domain has a limited number of in-flight requests
156 * device-wide, limited by these tokens.
157 */
158 struct sbitmap_queue domain_tokens[KYBER_NUM_DOMAINS];
159
160 /*
161 * Async request percentage, converted to per-word depth for
162 * sbitmap_get_shallow().
163 */
164 unsigned int async_depth;
165
166 struct kyber_cpu_latency __percpu *cpu_latency;
167
168 /* Timer for stats aggregation and adjusting domain tokens. */
169 struct timer_list timer;
170
171 unsigned int latency_buckets[KYBER_OTHER][2][KYBER_LATENCY_BUCKETS];
172
173 unsigned long latency_timeout[KYBER_OTHER];
174
175 int domain_p99[KYBER_OTHER];
176
177 /* Target latencies in nanoseconds. */
178 u64 latency_targets[KYBER_OTHER];
179};
180
181struct kyber_hctx_data {
182 spinlock_t lock;
183 struct list_head rqs[KYBER_NUM_DOMAINS];
184 unsigned int cur_domain;
185 unsigned int batching;
186 struct kyber_ctx_queue *kcqs;
187 struct sbitmap kcq_map[KYBER_NUM_DOMAINS];
188 struct sbq_wait domain_wait[KYBER_NUM_DOMAINS];
189 struct sbq_wait_state *domain_ws[KYBER_NUM_DOMAINS];
190 atomic_t wait_index[KYBER_NUM_DOMAINS];
191};
192
193static int kyber_domain_wake(wait_queue_entry_t *wait, unsigned mode, int flags,
194 void *key);
195
196static unsigned int kyber_sched_domain(blk_opf_t opf)
197{
198 switch (opf & REQ_OP_MASK) {
199 case REQ_OP_READ:
200 return KYBER_READ;
201 case REQ_OP_WRITE:
202 return KYBER_WRITE;
203 case REQ_OP_DISCARD:
204 return KYBER_DISCARD;
205 default:
206 return KYBER_OTHER;
207 }
208}
209
210static void flush_latency_buckets(struct kyber_queue_data *kqd,
211 struct kyber_cpu_latency *cpu_latency,
212 unsigned int sched_domain, unsigned int type)
213{
214 unsigned int *buckets = kqd->latency_buckets[sched_domain][type];
215 atomic_t *cpu_buckets = cpu_latency->buckets[sched_domain][type];
216 unsigned int bucket;
217
218 for (bucket = 0; bucket < KYBER_LATENCY_BUCKETS; bucket++)
219 buckets[bucket] += atomic_xchg(&cpu_buckets[bucket], 0);
220}
221
222/*
223 * Calculate the histogram bucket with the given percentile rank, or -1 if there
224 * aren't enough samples yet.
225 */
226static int calculate_percentile(struct kyber_queue_data *kqd,
227 unsigned int sched_domain, unsigned int type,
228 unsigned int percentile)
229{
230 unsigned int *buckets = kqd->latency_buckets[sched_domain][type];
231 unsigned int bucket, samples = 0, percentile_samples;
232
233 for (bucket = 0; bucket < KYBER_LATENCY_BUCKETS; bucket++)
234 samples += buckets[bucket];
235
236 if (!samples)
237 return -1;
238
239 /*
240 * We do the calculation once we have 500 samples or one second passes
241 * since the first sample was recorded, whichever comes first.
242 */
243 if (!kqd->latency_timeout[sched_domain])
244 kqd->latency_timeout[sched_domain] = max(jiffies + HZ, 1UL);
245 if (samples < 500 &&
246 time_is_after_jiffies(kqd->latency_timeout[sched_domain])) {
247 return -1;
248 }
249 kqd->latency_timeout[sched_domain] = 0;
250
251 percentile_samples = DIV_ROUND_UP(samples * percentile, 100);
252 for (bucket = 0; bucket < KYBER_LATENCY_BUCKETS - 1; bucket++) {
253 if (buckets[bucket] >= percentile_samples)
254 break;
255 percentile_samples -= buckets[bucket];
256 }
257 memset(buckets, 0, sizeof(kqd->latency_buckets[sched_domain][type]));
258
259 trace_kyber_latency(kqd->dev, kyber_domain_names[sched_domain],
260 kyber_latency_type_names[type], percentile,
261 bucket + 1, 1 << KYBER_LATENCY_SHIFT, samples);
262
263 return bucket;
264}
265
266static void kyber_resize_domain(struct kyber_queue_data *kqd,
267 unsigned int sched_domain, unsigned int depth)
268{
269 depth = clamp(depth, 1U, kyber_depth[sched_domain]);
270 if (depth != kqd->domain_tokens[sched_domain].sb.depth) {
271 sbitmap_queue_resize(&kqd->domain_tokens[sched_domain], depth);
272 trace_kyber_adjust(kqd->dev, kyber_domain_names[sched_domain],
273 depth);
274 }
275}
276
277static void kyber_timer_fn(struct timer_list *t)
278{
279 struct kyber_queue_data *kqd = from_timer(kqd, t, timer);
280 unsigned int sched_domain;
281 int cpu;
282 bool bad = false;
283
284 /* Sum all of the per-cpu latency histograms. */
285 for_each_online_cpu(cpu) {
286 struct kyber_cpu_latency *cpu_latency;
287
288 cpu_latency = per_cpu_ptr(kqd->cpu_latency, cpu);
289 for (sched_domain = 0; sched_domain < KYBER_OTHER; sched_domain++) {
290 flush_latency_buckets(kqd, cpu_latency, sched_domain,
291 KYBER_TOTAL_LATENCY);
292 flush_latency_buckets(kqd, cpu_latency, sched_domain,
293 KYBER_IO_LATENCY);
294 }
295 }
296
297 /*
298 * Check if any domains have a high I/O latency, which might indicate
299 * congestion in the device. Note that we use the p90; we don't want to
300 * be too sensitive to outliers here.
301 */
302 for (sched_domain = 0; sched_domain < KYBER_OTHER; sched_domain++) {
303 int p90;
304
305 p90 = calculate_percentile(kqd, sched_domain, KYBER_IO_LATENCY,
306 90);
307 if (p90 >= KYBER_GOOD_BUCKETS)
308 bad = true;
309 }
310
311 /*
312 * Adjust the scheduling domain depths. If we determined that there was
313 * congestion, we throttle all domains with good latencies. Either way,
314 * we ease up on throttling domains with bad latencies.
315 */
316 for (sched_domain = 0; sched_domain < KYBER_OTHER; sched_domain++) {
317 unsigned int orig_depth, depth;
318 int p99;
319
320 p99 = calculate_percentile(kqd, sched_domain,
321 KYBER_TOTAL_LATENCY, 99);
322 /*
323 * This is kind of subtle: different domains will not
324 * necessarily have enough samples to calculate the latency
325 * percentiles during the same window, so we have to remember
326 * the p99 for the next time we observe congestion; once we do,
327 * we don't want to throttle again until we get more data, so we
328 * reset it to -1.
329 */
330 if (bad) {
331 if (p99 < 0)
332 p99 = kqd->domain_p99[sched_domain];
333 kqd->domain_p99[sched_domain] = -1;
334 } else if (p99 >= 0) {
335 kqd->domain_p99[sched_domain] = p99;
336 }
337 if (p99 < 0)
338 continue;
339
340 /*
341 * If this domain has bad latency, throttle less. Otherwise,
342 * throttle more iff we determined that there is congestion.
343 *
344 * The new depth is scaled linearly with the p99 latency vs the
345 * latency target. E.g., if the p99 is 3/4 of the target, then
346 * we throttle down to 3/4 of the current depth, and if the p99
347 * is 2x the target, then we double the depth.
348 */
349 if (bad || p99 >= KYBER_GOOD_BUCKETS) {
350 orig_depth = kqd->domain_tokens[sched_domain].sb.depth;
351 depth = (orig_depth * (p99 + 1)) >> KYBER_LATENCY_SHIFT;
352 kyber_resize_domain(kqd, sched_domain, depth);
353 }
354 }
355}
356
357static struct kyber_queue_data *kyber_queue_data_alloc(struct request_queue *q)
358{
359 struct kyber_queue_data *kqd;
360 int ret = -ENOMEM;
361 int i;
362
363 kqd = kzalloc_node(sizeof(*kqd), GFP_KERNEL, q->node);
364 if (!kqd)
365 goto err;
366
367 kqd->q = q;
368 kqd->dev = disk_devt(q->disk);
369
370 kqd->cpu_latency = alloc_percpu_gfp(struct kyber_cpu_latency,
371 GFP_KERNEL | __GFP_ZERO);
372 if (!kqd->cpu_latency)
373 goto err_kqd;
374
375 timer_setup(&kqd->timer, kyber_timer_fn, 0);
376
377 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
378 WARN_ON(!kyber_depth[i]);
379 WARN_ON(!kyber_batch_size[i]);
380 ret = sbitmap_queue_init_node(&kqd->domain_tokens[i],
381 kyber_depth[i], -1, false,
382 GFP_KERNEL, q->node);
383 if (ret) {
384 while (--i >= 0)
385 sbitmap_queue_free(&kqd->domain_tokens[i]);
386 goto err_buckets;
387 }
388 }
389
390 for (i = 0; i < KYBER_OTHER; i++) {
391 kqd->domain_p99[i] = -1;
392 kqd->latency_targets[i] = kyber_latency_targets[i];
393 }
394
395 return kqd;
396
397err_buckets:
398 free_percpu(kqd->cpu_latency);
399err_kqd:
400 kfree(kqd);
401err:
402 return ERR_PTR(ret);
403}
404
405static int kyber_init_sched(struct request_queue *q, struct elevator_type *e)
406{
407 struct kyber_queue_data *kqd;
408 struct elevator_queue *eq;
409
410 eq = elevator_alloc(q, e);
411 if (!eq)
412 return -ENOMEM;
413
414 kqd = kyber_queue_data_alloc(q);
415 if (IS_ERR(kqd)) {
416 kobject_put(&eq->kobj);
417 return PTR_ERR(kqd);
418 }
419
420 blk_stat_enable_accounting(q);
421
422 blk_queue_flag_clear(QUEUE_FLAG_SQ_SCHED, q);
423
424 eq->elevator_data = kqd;
425 q->elevator = eq;
426
427 return 0;
428}
429
430static void kyber_exit_sched(struct elevator_queue *e)
431{
432 struct kyber_queue_data *kqd = e->elevator_data;
433 int i;
434
435 timer_shutdown_sync(&kqd->timer);
436 blk_stat_disable_accounting(kqd->q);
437
438 for (i = 0; i < KYBER_NUM_DOMAINS; i++)
439 sbitmap_queue_free(&kqd->domain_tokens[i]);
440 free_percpu(kqd->cpu_latency);
441 kfree(kqd);
442}
443
444static void kyber_ctx_queue_init(struct kyber_ctx_queue *kcq)
445{
446 unsigned int i;
447
448 spin_lock_init(&kcq->lock);
449 for (i = 0; i < KYBER_NUM_DOMAINS; i++)
450 INIT_LIST_HEAD(&kcq->rq_list[i]);
451}
452
453static void kyber_depth_updated(struct blk_mq_hw_ctx *hctx)
454{
455 struct kyber_queue_data *kqd = hctx->queue->elevator->elevator_data;
456 struct blk_mq_tags *tags = hctx->sched_tags;
457 unsigned int shift = tags->bitmap_tags.sb.shift;
458
459 kqd->async_depth = (1U << shift) * KYBER_ASYNC_PERCENT / 100U;
460
461 sbitmap_queue_min_shallow_depth(&tags->bitmap_tags, kqd->async_depth);
462}
463
464static int kyber_init_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
465{
466 struct kyber_hctx_data *khd;
467 int i;
468
469 khd = kmalloc_node(sizeof(*khd), GFP_KERNEL, hctx->numa_node);
470 if (!khd)
471 return -ENOMEM;
472
473 khd->kcqs = kmalloc_array_node(hctx->nr_ctx,
474 sizeof(struct kyber_ctx_queue),
475 GFP_KERNEL, hctx->numa_node);
476 if (!khd->kcqs)
477 goto err_khd;
478
479 for (i = 0; i < hctx->nr_ctx; i++)
480 kyber_ctx_queue_init(&khd->kcqs[i]);
481
482 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
483 if (sbitmap_init_node(&khd->kcq_map[i], hctx->nr_ctx,
484 ilog2(8), GFP_KERNEL, hctx->numa_node,
485 false, false)) {
486 while (--i >= 0)
487 sbitmap_free(&khd->kcq_map[i]);
488 goto err_kcqs;
489 }
490 }
491
492 spin_lock_init(&khd->lock);
493
494 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
495 INIT_LIST_HEAD(&khd->rqs[i]);
496 khd->domain_wait[i].sbq = NULL;
497 init_waitqueue_func_entry(&khd->domain_wait[i].wait,
498 kyber_domain_wake);
499 khd->domain_wait[i].wait.private = hctx;
500 INIT_LIST_HEAD(&khd->domain_wait[i].wait.entry);
501 atomic_set(&khd->wait_index[i], 0);
502 }
503
504 khd->cur_domain = 0;
505 khd->batching = 0;
506
507 hctx->sched_data = khd;
508 kyber_depth_updated(hctx);
509
510 return 0;
511
512err_kcqs:
513 kfree(khd->kcqs);
514err_khd:
515 kfree(khd);
516 return -ENOMEM;
517}
518
519static void kyber_exit_hctx(struct blk_mq_hw_ctx *hctx, unsigned int hctx_idx)
520{
521 struct kyber_hctx_data *khd = hctx->sched_data;
522 int i;
523
524 for (i = 0; i < KYBER_NUM_DOMAINS; i++)
525 sbitmap_free(&khd->kcq_map[i]);
526 kfree(khd->kcqs);
527 kfree(hctx->sched_data);
528}
529
530static int rq_get_domain_token(struct request *rq)
531{
532 return (long)rq->elv.priv[0];
533}
534
535static void rq_set_domain_token(struct request *rq, int token)
536{
537 rq->elv.priv[0] = (void *)(long)token;
538}
539
540static void rq_clear_domain_token(struct kyber_queue_data *kqd,
541 struct request *rq)
542{
543 unsigned int sched_domain;
544 int nr;
545
546 nr = rq_get_domain_token(rq);
547 if (nr != -1) {
548 sched_domain = kyber_sched_domain(rq->cmd_flags);
549 sbitmap_queue_clear(&kqd->domain_tokens[sched_domain], nr,
550 rq->mq_ctx->cpu);
551 }
552}
553
554static void kyber_limit_depth(blk_opf_t opf, struct blk_mq_alloc_data *data)
555{
556 /*
557 * We use the scheduler tags as per-hardware queue queueing tokens.
558 * Async requests can be limited at this stage.
559 */
560 if (!op_is_sync(opf)) {
561 struct kyber_queue_data *kqd = data->q->elevator->elevator_data;
562
563 data->shallow_depth = kqd->async_depth;
564 }
565}
566
567static bool kyber_bio_merge(struct request_queue *q, struct bio *bio,
568 unsigned int nr_segs)
569{
570 struct blk_mq_ctx *ctx = blk_mq_get_ctx(q);
571 struct blk_mq_hw_ctx *hctx = blk_mq_map_queue(q, bio->bi_opf, ctx);
572 struct kyber_hctx_data *khd = hctx->sched_data;
573 struct kyber_ctx_queue *kcq = &khd->kcqs[ctx->index_hw[hctx->type]];
574 unsigned int sched_domain = kyber_sched_domain(bio->bi_opf);
575 struct list_head *rq_list = &kcq->rq_list[sched_domain];
576 bool merged;
577
578 spin_lock(&kcq->lock);
579 merged = blk_bio_list_merge(hctx->queue, rq_list, bio, nr_segs);
580 spin_unlock(&kcq->lock);
581
582 return merged;
583}
584
585static void kyber_prepare_request(struct request *rq)
586{
587 rq_set_domain_token(rq, -1);
588}
589
590static void kyber_insert_requests(struct blk_mq_hw_ctx *hctx,
591 struct list_head *rq_list,
592 blk_insert_t flags)
593{
594 struct kyber_hctx_data *khd = hctx->sched_data;
595 struct request *rq, *next;
596
597 list_for_each_entry_safe(rq, next, rq_list, queuelist) {
598 unsigned int sched_domain = kyber_sched_domain(rq->cmd_flags);
599 struct kyber_ctx_queue *kcq = &khd->kcqs[rq->mq_ctx->index_hw[hctx->type]];
600 struct list_head *head = &kcq->rq_list[sched_domain];
601
602 spin_lock(&kcq->lock);
603 trace_block_rq_insert(rq);
604 if (flags & BLK_MQ_INSERT_AT_HEAD)
605 list_move(&rq->queuelist, head);
606 else
607 list_move_tail(&rq->queuelist, head);
608 sbitmap_set_bit(&khd->kcq_map[sched_domain],
609 rq->mq_ctx->index_hw[hctx->type]);
610 spin_unlock(&kcq->lock);
611 }
612}
613
614static void kyber_finish_request(struct request *rq)
615{
616 struct kyber_queue_data *kqd = rq->q->elevator->elevator_data;
617
618 rq_clear_domain_token(kqd, rq);
619}
620
621static void add_latency_sample(struct kyber_cpu_latency *cpu_latency,
622 unsigned int sched_domain, unsigned int type,
623 u64 target, u64 latency)
624{
625 unsigned int bucket;
626 u64 divisor;
627
628 if (latency > 0) {
629 divisor = max_t(u64, target >> KYBER_LATENCY_SHIFT, 1);
630 bucket = min_t(unsigned int, div64_u64(latency - 1, divisor),
631 KYBER_LATENCY_BUCKETS - 1);
632 } else {
633 bucket = 0;
634 }
635
636 atomic_inc(&cpu_latency->buckets[sched_domain][type][bucket]);
637}
638
639static void kyber_completed_request(struct request *rq, u64 now)
640{
641 struct kyber_queue_data *kqd = rq->q->elevator->elevator_data;
642 struct kyber_cpu_latency *cpu_latency;
643 unsigned int sched_domain;
644 u64 target;
645
646 sched_domain = kyber_sched_domain(rq->cmd_flags);
647 if (sched_domain == KYBER_OTHER)
648 return;
649
650 cpu_latency = get_cpu_ptr(kqd->cpu_latency);
651 target = kqd->latency_targets[sched_domain];
652 add_latency_sample(cpu_latency, sched_domain, KYBER_TOTAL_LATENCY,
653 target, now - rq->start_time_ns);
654 add_latency_sample(cpu_latency, sched_domain, KYBER_IO_LATENCY, target,
655 now - rq->io_start_time_ns);
656 put_cpu_ptr(kqd->cpu_latency);
657
658 timer_reduce(&kqd->timer, jiffies + HZ / 10);
659}
660
661struct flush_kcq_data {
662 struct kyber_hctx_data *khd;
663 unsigned int sched_domain;
664 struct list_head *list;
665};
666
667static bool flush_busy_kcq(struct sbitmap *sb, unsigned int bitnr, void *data)
668{
669 struct flush_kcq_data *flush_data = data;
670 struct kyber_ctx_queue *kcq = &flush_data->khd->kcqs[bitnr];
671
672 spin_lock(&kcq->lock);
673 list_splice_tail_init(&kcq->rq_list[flush_data->sched_domain],
674 flush_data->list);
675 sbitmap_clear_bit(sb, bitnr);
676 spin_unlock(&kcq->lock);
677
678 return true;
679}
680
681static void kyber_flush_busy_kcqs(struct kyber_hctx_data *khd,
682 unsigned int sched_domain,
683 struct list_head *list)
684{
685 struct flush_kcq_data data = {
686 .khd = khd,
687 .sched_domain = sched_domain,
688 .list = list,
689 };
690
691 sbitmap_for_each_set(&khd->kcq_map[sched_domain],
692 flush_busy_kcq, &data);
693}
694
695static int kyber_domain_wake(wait_queue_entry_t *wqe, unsigned mode, int flags,
696 void *key)
697{
698 struct blk_mq_hw_ctx *hctx = READ_ONCE(wqe->private);
699 struct sbq_wait *wait = container_of(wqe, struct sbq_wait, wait);
700
701 sbitmap_del_wait_queue(wait);
702 blk_mq_run_hw_queue(hctx, true);
703 return 1;
704}
705
706static int kyber_get_domain_token(struct kyber_queue_data *kqd,
707 struct kyber_hctx_data *khd,
708 struct blk_mq_hw_ctx *hctx)
709{
710 unsigned int sched_domain = khd->cur_domain;
711 struct sbitmap_queue *domain_tokens = &kqd->domain_tokens[sched_domain];
712 struct sbq_wait *wait = &khd->domain_wait[sched_domain];
713 struct sbq_wait_state *ws;
714 int nr;
715
716 nr = __sbitmap_queue_get(domain_tokens);
717
718 /*
719 * If we failed to get a domain token, make sure the hardware queue is
720 * run when one becomes available. Note that this is serialized on
721 * khd->lock, but we still need to be careful about the waker.
722 */
723 if (nr < 0 && list_empty_careful(&wait->wait.entry)) {
724 ws = sbq_wait_ptr(domain_tokens,
725 &khd->wait_index[sched_domain]);
726 khd->domain_ws[sched_domain] = ws;
727 sbitmap_add_wait_queue(domain_tokens, ws, wait);
728
729 /*
730 * Try again in case a token was freed before we got on the wait
731 * queue.
732 */
733 nr = __sbitmap_queue_get(domain_tokens);
734 }
735
736 /*
737 * If we got a token while we were on the wait queue, remove ourselves
738 * from the wait queue to ensure that all wake ups make forward
739 * progress. It's possible that the waker already deleted the entry
740 * between the !list_empty_careful() check and us grabbing the lock, but
741 * list_del_init() is okay with that.
742 */
743 if (nr >= 0 && !list_empty_careful(&wait->wait.entry)) {
744 ws = khd->domain_ws[sched_domain];
745 spin_lock_irq(&ws->wait.lock);
746 sbitmap_del_wait_queue(wait);
747 spin_unlock_irq(&ws->wait.lock);
748 }
749
750 return nr;
751}
752
753static struct request *
754kyber_dispatch_cur_domain(struct kyber_queue_data *kqd,
755 struct kyber_hctx_data *khd,
756 struct blk_mq_hw_ctx *hctx)
757{
758 struct list_head *rqs;
759 struct request *rq;
760 int nr;
761
762 rqs = &khd->rqs[khd->cur_domain];
763
764 /*
765 * If we already have a flushed request, then we just need to get a
766 * token for it. Otherwise, if there are pending requests in the kcqs,
767 * flush the kcqs, but only if we can get a token. If not, we should
768 * leave the requests in the kcqs so that they can be merged. Note that
769 * khd->lock serializes the flushes, so if we observed any bit set in
770 * the kcq_map, we will always get a request.
771 */
772 rq = list_first_entry_or_null(rqs, struct request, queuelist);
773 if (rq) {
774 nr = kyber_get_domain_token(kqd, khd, hctx);
775 if (nr >= 0) {
776 khd->batching++;
777 rq_set_domain_token(rq, nr);
778 list_del_init(&rq->queuelist);
779 return rq;
780 } else {
781 trace_kyber_throttled(kqd->dev,
782 kyber_domain_names[khd->cur_domain]);
783 }
784 } else if (sbitmap_any_bit_set(&khd->kcq_map[khd->cur_domain])) {
785 nr = kyber_get_domain_token(kqd, khd, hctx);
786 if (nr >= 0) {
787 kyber_flush_busy_kcqs(khd, khd->cur_domain, rqs);
788 rq = list_first_entry(rqs, struct request, queuelist);
789 khd->batching++;
790 rq_set_domain_token(rq, nr);
791 list_del_init(&rq->queuelist);
792 return rq;
793 } else {
794 trace_kyber_throttled(kqd->dev,
795 kyber_domain_names[khd->cur_domain]);
796 }
797 }
798
799 /* There were either no pending requests or no tokens. */
800 return NULL;
801}
802
803static struct request *kyber_dispatch_request(struct blk_mq_hw_ctx *hctx)
804{
805 struct kyber_queue_data *kqd = hctx->queue->elevator->elevator_data;
806 struct kyber_hctx_data *khd = hctx->sched_data;
807 struct request *rq;
808 int i;
809
810 spin_lock(&khd->lock);
811
812 /*
813 * First, if we are still entitled to batch, try to dispatch a request
814 * from the batch.
815 */
816 if (khd->batching < kyber_batch_size[khd->cur_domain]) {
817 rq = kyber_dispatch_cur_domain(kqd, khd, hctx);
818 if (rq)
819 goto out;
820 }
821
822 /*
823 * Either,
824 * 1. We were no longer entitled to a batch.
825 * 2. The domain we were batching didn't have any requests.
826 * 3. The domain we were batching was out of tokens.
827 *
828 * Start another batch. Note that this wraps back around to the original
829 * domain if no other domains have requests or tokens.
830 */
831 khd->batching = 0;
832 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
833 if (khd->cur_domain == KYBER_NUM_DOMAINS - 1)
834 khd->cur_domain = 0;
835 else
836 khd->cur_domain++;
837
838 rq = kyber_dispatch_cur_domain(kqd, khd, hctx);
839 if (rq)
840 goto out;
841 }
842
843 rq = NULL;
844out:
845 spin_unlock(&khd->lock);
846 return rq;
847}
848
849static bool kyber_has_work(struct blk_mq_hw_ctx *hctx)
850{
851 struct kyber_hctx_data *khd = hctx->sched_data;
852 int i;
853
854 for (i = 0; i < KYBER_NUM_DOMAINS; i++) {
855 if (!list_empty_careful(&khd->rqs[i]) ||
856 sbitmap_any_bit_set(&khd->kcq_map[i]))
857 return true;
858 }
859
860 return false;
861}
862
863#define KYBER_LAT_SHOW_STORE(domain, name) \
864static ssize_t kyber_##name##_lat_show(struct elevator_queue *e, \
865 char *page) \
866{ \
867 struct kyber_queue_data *kqd = e->elevator_data; \
868 \
869 return sprintf(page, "%llu\n", kqd->latency_targets[domain]); \
870} \
871 \
872static ssize_t kyber_##name##_lat_store(struct elevator_queue *e, \
873 const char *page, size_t count) \
874{ \
875 struct kyber_queue_data *kqd = e->elevator_data; \
876 unsigned long long nsec; \
877 int ret; \
878 \
879 ret = kstrtoull(page, 10, &nsec); \
880 if (ret) \
881 return ret; \
882 \
883 kqd->latency_targets[domain] = nsec; \
884 \
885 return count; \
886}
887KYBER_LAT_SHOW_STORE(KYBER_READ, read);
888KYBER_LAT_SHOW_STORE(KYBER_WRITE, write);
889#undef KYBER_LAT_SHOW_STORE
890
891#define KYBER_LAT_ATTR(op) __ATTR(op##_lat_nsec, 0644, kyber_##op##_lat_show, kyber_##op##_lat_store)
892static struct elv_fs_entry kyber_sched_attrs[] = {
893 KYBER_LAT_ATTR(read),
894 KYBER_LAT_ATTR(write),
895 __ATTR_NULL
896};
897#undef KYBER_LAT_ATTR
898
899#ifdef CONFIG_BLK_DEBUG_FS
900#define KYBER_DEBUGFS_DOMAIN_ATTRS(domain, name) \
901static int kyber_##name##_tokens_show(void *data, struct seq_file *m) \
902{ \
903 struct request_queue *q = data; \
904 struct kyber_queue_data *kqd = q->elevator->elevator_data; \
905 \
906 sbitmap_queue_show(&kqd->domain_tokens[domain], m); \
907 return 0; \
908} \
909 \
910static void *kyber_##name##_rqs_start(struct seq_file *m, loff_t *pos) \
911 __acquires(&khd->lock) \
912{ \
913 struct blk_mq_hw_ctx *hctx = m->private; \
914 struct kyber_hctx_data *khd = hctx->sched_data; \
915 \
916 spin_lock(&khd->lock); \
917 return seq_list_start(&khd->rqs[domain], *pos); \
918} \
919 \
920static void *kyber_##name##_rqs_next(struct seq_file *m, void *v, \
921 loff_t *pos) \
922{ \
923 struct blk_mq_hw_ctx *hctx = m->private; \
924 struct kyber_hctx_data *khd = hctx->sched_data; \
925 \
926 return seq_list_next(v, &khd->rqs[domain], pos); \
927} \
928 \
929static void kyber_##name##_rqs_stop(struct seq_file *m, void *v) \
930 __releases(&khd->lock) \
931{ \
932 struct blk_mq_hw_ctx *hctx = m->private; \
933 struct kyber_hctx_data *khd = hctx->sched_data; \
934 \
935 spin_unlock(&khd->lock); \
936} \
937 \
938static const struct seq_operations kyber_##name##_rqs_seq_ops = { \
939 .start = kyber_##name##_rqs_start, \
940 .next = kyber_##name##_rqs_next, \
941 .stop = kyber_##name##_rqs_stop, \
942 .show = blk_mq_debugfs_rq_show, \
943}; \
944 \
945static int kyber_##name##_waiting_show(void *data, struct seq_file *m) \
946{ \
947 struct blk_mq_hw_ctx *hctx = data; \
948 struct kyber_hctx_data *khd = hctx->sched_data; \
949 wait_queue_entry_t *wait = &khd->domain_wait[domain].wait; \
950 \
951 seq_printf(m, "%d\n", !list_empty_careful(&wait->entry)); \
952 return 0; \
953}
954KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_READ, read)
955KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_WRITE, write)
956KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_DISCARD, discard)
957KYBER_DEBUGFS_DOMAIN_ATTRS(KYBER_OTHER, other)
958#undef KYBER_DEBUGFS_DOMAIN_ATTRS
959
960static int kyber_async_depth_show(void *data, struct seq_file *m)
961{
962 struct request_queue *q = data;
963 struct kyber_queue_data *kqd = q->elevator->elevator_data;
964
965 seq_printf(m, "%u\n", kqd->async_depth);
966 return 0;
967}
968
969static int kyber_cur_domain_show(void *data, struct seq_file *m)
970{
971 struct blk_mq_hw_ctx *hctx = data;
972 struct kyber_hctx_data *khd = hctx->sched_data;
973
974 seq_printf(m, "%s\n", kyber_domain_names[khd->cur_domain]);
975 return 0;
976}
977
978static int kyber_batching_show(void *data, struct seq_file *m)
979{
980 struct blk_mq_hw_ctx *hctx = data;
981 struct kyber_hctx_data *khd = hctx->sched_data;
982
983 seq_printf(m, "%u\n", khd->batching);
984 return 0;
985}
986
987#define KYBER_QUEUE_DOMAIN_ATTRS(name) \
988 {#name "_tokens", 0400, kyber_##name##_tokens_show}
989static const struct blk_mq_debugfs_attr kyber_queue_debugfs_attrs[] = {
990 KYBER_QUEUE_DOMAIN_ATTRS(read),
991 KYBER_QUEUE_DOMAIN_ATTRS(write),
992 KYBER_QUEUE_DOMAIN_ATTRS(discard),
993 KYBER_QUEUE_DOMAIN_ATTRS(other),
994 {"async_depth", 0400, kyber_async_depth_show},
995 {},
996};
997#undef KYBER_QUEUE_DOMAIN_ATTRS
998
999#define KYBER_HCTX_DOMAIN_ATTRS(name) \
1000 {#name "_rqs", 0400, .seq_ops = &kyber_##name##_rqs_seq_ops}, \
1001 {#name "_waiting", 0400, kyber_##name##_waiting_show}
1002static const struct blk_mq_debugfs_attr kyber_hctx_debugfs_attrs[] = {
1003 KYBER_HCTX_DOMAIN_ATTRS(read),
1004 KYBER_HCTX_DOMAIN_ATTRS(write),
1005 KYBER_HCTX_DOMAIN_ATTRS(discard),
1006 KYBER_HCTX_DOMAIN_ATTRS(other),
1007 {"cur_domain", 0400, kyber_cur_domain_show},
1008 {"batching", 0400, kyber_batching_show},
1009 {},
1010};
1011#undef KYBER_HCTX_DOMAIN_ATTRS
1012#endif
1013
1014static struct elevator_type kyber_sched = {
1015 .ops = {
1016 .init_sched = kyber_init_sched,
1017 .exit_sched = kyber_exit_sched,
1018 .init_hctx = kyber_init_hctx,
1019 .exit_hctx = kyber_exit_hctx,
1020 .limit_depth = kyber_limit_depth,
1021 .bio_merge = kyber_bio_merge,
1022 .prepare_request = kyber_prepare_request,
1023 .insert_requests = kyber_insert_requests,
1024 .finish_request = kyber_finish_request,
1025 .requeue_request = kyber_finish_request,
1026 .completed_request = kyber_completed_request,
1027 .dispatch_request = kyber_dispatch_request,
1028 .has_work = kyber_has_work,
1029 .depth_updated = kyber_depth_updated,
1030 },
1031#ifdef CONFIG_BLK_DEBUG_FS
1032 .queue_debugfs_attrs = kyber_queue_debugfs_attrs,
1033 .hctx_debugfs_attrs = kyber_hctx_debugfs_attrs,
1034#endif
1035 .elevator_attrs = kyber_sched_attrs,
1036 .elevator_name = "kyber",
1037 .elevator_owner = THIS_MODULE,
1038};
1039
1040static int __init kyber_init(void)
1041{
1042 return elv_register(&kyber_sched);
1043}
1044
1045static void __exit kyber_exit(void)
1046{
1047 elv_unregister(&kyber_sched);
1048}
1049
1050module_init(kyber_init);
1051module_exit(kyber_exit);
1052
1053MODULE_AUTHOR("Omar Sandoval");
1054MODULE_LICENSE("GPL");
1055MODULE_DESCRIPTION("Kyber I/O scheduler");