Blame view
block/blk-core.c
105 KB
1da177e4c
|
1 |
/* |
1da177e4c
|
2 3 4 5 |
* Copyright (C) 1991, 1992 Linus Torvalds * Copyright (C) 1994, Karl Keyte: Added support for disk statistics * Elevator latency, (C) 2000 Andrea Arcangeli <andrea@suse.de> SuSE * Queue request tables / lock, selectable elevator, Jens Axboe <axboe@suse.de> |
6728cb0e6
|
6 7 |
* kernel-doc documentation started by NeilBrown <neilb@cse.unsw.edu.au> * - July2000 |
1da177e4c
|
8 9 10 11 12 13 |
* bio rewrite, highmem i/o, etc, Jens Axboe <axboe@suse.de> - may 2001 */ /* * This handles all read/write requests to block devices */ |
1da177e4c
|
14 15 16 17 18 |
#include <linux/kernel.h> #include <linux/module.h> #include <linux/backing-dev.h> #include <linux/bio.h> #include <linux/blkdev.h> |
320ae51fe
|
19 |
#include <linux/blk-mq.h> |
1da177e4c
|
20 21 22 23 24 |
#include <linux/highmem.h> #include <linux/mm.h> #include <linux/kernel_stat.h> #include <linux/string.h> #include <linux/init.h> |
1da177e4c
|
25 26 27 28 |
#include <linux/completion.h> #include <linux/slab.h> #include <linux/swap.h> #include <linux/writeback.h> |
faccbd4b2
|
29 |
#include <linux/task_io_accounting_ops.h> |
c17bb4951
|
30 |
#include <linux/fault-inject.h> |
73c101011
|
31 |
#include <linux/list_sort.h> |
e3c78ca52
|
32 |
#include <linux/delay.h> |
aaf7c6806
|
33 |
#include <linux/ratelimit.h> |
6c9546675
|
34 |
#include <linux/pm_runtime.h> |
eea8f41cc
|
35 |
#include <linux/blk-cgroup.h> |
18fbda91c
|
36 |
#include <linux/debugfs.h> |
30abb3a67
|
37 |
#include <linux/bpf.h> |
55782138e
|
38 39 40 |
#define CREATE_TRACE_POINTS #include <trace/events/block.h> |
1da177e4c
|
41 |
|
8324aa91d
|
42 |
#include "blk.h" |
43a5e4e21
|
43 |
#include "blk-mq.h" |
bd166ef18
|
44 |
#include "blk-mq-sched.h" |
c1c80384c
|
45 |
#include "blk-rq-qos.h" |
8324aa91d
|
46 |
|
18fbda91c
|
47 48 49 |
#ifdef CONFIG_DEBUG_FS struct dentry *blk_debugfs_root; #endif |
d07335e51
|
50 |
EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_remap); |
b0da3f0da
|
51 |
EXPORT_TRACEPOINT_SYMBOL_GPL(block_rq_remap); |
0a82a8d13
|
52 |
EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_complete); |
3291fa57c
|
53 |
EXPORT_TRACEPOINT_SYMBOL_GPL(block_split); |
cbae8d45d
|
54 |
EXPORT_TRACEPOINT_SYMBOL_GPL(block_unplug); |
0bfc24559
|
55 |
|
a73f730d0
|
56 |
DEFINE_IDA(blk_queue_ida); |
1da177e4c
|
57 58 59 |
/* * For the allocated request tables */ |
d674d4145
|
60 |
struct kmem_cache *request_cachep; |
1da177e4c
|
61 62 63 64 |
/* * For queue allocation */ |
6728cb0e6
|
65 |
struct kmem_cache *blk_requestq_cachep; |
1da177e4c
|
66 67 |
/* |
1da177e4c
|
68 69 |
* Controlling structure to kblockd */ |
ff856bad6
|
70 |
static struct workqueue_struct *kblockd_workqueue; |
1da177e4c
|
71 |
|
8814ce8a0
|
72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 |
/** * blk_queue_flag_set - atomically set a queue flag * @flag: flag to be set * @q: request queue */ void blk_queue_flag_set(unsigned int flag, struct request_queue *q) { unsigned long flags; spin_lock_irqsave(q->queue_lock, flags); queue_flag_set(flag, q); spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL(blk_queue_flag_set); /** * blk_queue_flag_clear - atomically clear a queue flag * @flag: flag to be cleared * @q: request queue */ void blk_queue_flag_clear(unsigned int flag, struct request_queue *q) { unsigned long flags; spin_lock_irqsave(q->queue_lock, flags); queue_flag_clear(flag, q); spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL(blk_queue_flag_clear); /** * blk_queue_flag_test_and_set - atomically test and set a queue flag * @flag: flag to be set * @q: request queue * * Returns the previous value of @flag - 0 if the flag was not set and 1 if * the flag was already set. */ bool blk_queue_flag_test_and_set(unsigned int flag, struct request_queue *q) { unsigned long flags; bool res; spin_lock_irqsave(q->queue_lock, flags); res = queue_flag_test_and_set(flag, q); spin_unlock_irqrestore(q->queue_lock, flags); return res; } EXPORT_SYMBOL_GPL(blk_queue_flag_test_and_set); /** * blk_queue_flag_test_and_clear - atomically test and clear a queue flag * @flag: flag to be cleared * @q: request queue * * Returns the previous value of @flag - 0 if the flag was not set and 1 if * the flag was set. */ bool blk_queue_flag_test_and_clear(unsigned int flag, struct request_queue *q) { unsigned long flags; bool res; spin_lock_irqsave(q->queue_lock, flags); res = queue_flag_test_and_clear(flag, q); spin_unlock_irqrestore(q->queue_lock, flags); return res; } EXPORT_SYMBOL_GPL(blk_queue_flag_test_and_clear); |
d40f75a06
|
143 144 |
static void blk_clear_congested(struct request_list *rl, int sync) { |
d40f75a06
|
145 146 147 |
#ifdef CONFIG_CGROUP_WRITEBACK clear_wb_congested(rl->blkg->wb_congested, sync); #else |
482cf79cd
|
148 149 150 151 152 |
/* * If !CGROUP_WRITEBACK, all blkg's map to bdi->wb and we shouldn't * flip its congestion state for events on other blkcgs. */ if (rl == &rl->q->root_rl) |
dc3b17cc8
|
153 |
clear_wb_congested(rl->q->backing_dev_info->wb.congested, sync); |
d40f75a06
|
154 155 156 157 158 |
#endif } static void blk_set_congested(struct request_list *rl, int sync) { |
d40f75a06
|
159 160 161 |
#ifdef CONFIG_CGROUP_WRITEBACK set_wb_congested(rl->blkg->wb_congested, sync); #else |
482cf79cd
|
162 163 |
/* see blk_clear_congested() */ if (rl == &rl->q->root_rl) |
dc3b17cc8
|
164 |
set_wb_congested(rl->q->backing_dev_info->wb.congested, sync); |
d40f75a06
|
165 166 |
#endif } |
8324aa91d
|
167 |
void blk_queue_congestion_threshold(struct request_queue *q) |
1da177e4c
|
168 169 170 171 172 173 174 175 176 177 178 179 180 |
{ int nr; nr = q->nr_requests - (q->nr_requests / 8) + 1; if (nr > q->nr_requests) nr = q->nr_requests; q->nr_congestion_on = nr; nr = q->nr_requests - (q->nr_requests / 8) - (q->nr_requests / 16) - 1; if (nr < 1) nr = 1; q->nr_congestion_off = nr; } |
2a4aa30c5
|
181 |
void blk_rq_init(struct request_queue *q, struct request *rq) |
1da177e4c
|
182 |
{ |
1afb20f30
|
183 |
memset(rq, 0, sizeof(*rq)); |
1da177e4c
|
184 |
INIT_LIST_HEAD(&rq->queuelist); |
242f9dcb8
|
185 |
INIT_LIST_HEAD(&rq->timeout_list); |
c7c22e4d5
|
186 |
rq->cpu = -1; |
63a713867
|
187 |
rq->q = q; |
a2dec7b36
|
188 |
rq->__sector = (sector_t) -1; |
2e662b65f
|
189 190 |
INIT_HLIST_NODE(&rq->hash); RB_CLEAR_NODE(&rq->rb_node); |
63a713867
|
191 |
rq->tag = -1; |
bd166ef18
|
192 |
rq->internal_tag = -1; |
522a77756
|
193 |
rq->start_time_ns = ktime_get_ns(); |
09e099d4b
|
194 |
rq->part = NULL; |
1da177e4c
|
195 |
} |
2a4aa30c5
|
196 |
EXPORT_SYMBOL(blk_rq_init); |
1da177e4c
|
197 |
|
2a842acab
|
198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
static const struct { int errno; const char *name; } blk_errors[] = { [BLK_STS_OK] = { 0, "" }, [BLK_STS_NOTSUPP] = { -EOPNOTSUPP, "operation not supported" }, [BLK_STS_TIMEOUT] = { -ETIMEDOUT, "timeout" }, [BLK_STS_NOSPC] = { -ENOSPC, "critical space allocation" }, [BLK_STS_TRANSPORT] = { -ENOLINK, "recoverable transport" }, [BLK_STS_TARGET] = { -EREMOTEIO, "critical target" }, [BLK_STS_NEXUS] = { -EBADE, "critical nexus" }, [BLK_STS_MEDIUM] = { -ENODATA, "critical medium" }, [BLK_STS_PROTECTION] = { -EILSEQ, "protection" }, [BLK_STS_RESOURCE] = { -ENOMEM, "kernel resource" }, |
86ff7c2a8
|
212 |
[BLK_STS_DEV_RESOURCE] = { -EBUSY, "device resource" }, |
03a07c92a
|
213 |
[BLK_STS_AGAIN] = { -EAGAIN, "nonblocking retry" }, |
2a842acab
|
214 |
|
4e4cbee93
|
215 216 |
/* device mapper special case, should not leak out: */ [BLK_STS_DM_REQUEUE] = { -EREMCHG, "dm internal retry" }, |
2a842acab
|
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 |
/* everything else not covered above: */ [BLK_STS_IOERR] = { -EIO, "I/O" }, }; blk_status_t errno_to_blk_status(int errno) { int i; for (i = 0; i < ARRAY_SIZE(blk_errors); i++) { if (blk_errors[i].errno == errno) return (__force blk_status_t)i; } return BLK_STS_IOERR; } EXPORT_SYMBOL_GPL(errno_to_blk_status); int blk_status_to_errno(blk_status_t status) { int idx = (__force int)status; |
34bd9c1c4
|
237 |
if (WARN_ON_ONCE(idx >= ARRAY_SIZE(blk_errors))) |
2a842acab
|
238 239 240 241 242 243 244 245 |
return -EIO; return blk_errors[idx].errno; } EXPORT_SYMBOL_GPL(blk_status_to_errno); static void print_req_error(struct request *req, blk_status_t status) { int idx = (__force int)status; |
34bd9c1c4
|
246 |
if (WARN_ON_ONCE(idx >= ARRAY_SIZE(blk_errors))) |
2a842acab
|
247 248 249 250 251 252 253 254 |
return; printk_ratelimited(KERN_ERR "%s: %s error, dev %s, sector %llu ", __func__, blk_errors[idx].name, req->rq_disk ? req->rq_disk->disk_name : "?", (unsigned long long)blk_rq_pos(req)); } |
5bb23a688
|
255 |
static void req_bio_endio(struct request *rq, struct bio *bio, |
2a842acab
|
256 |
unsigned int nbytes, blk_status_t error) |
1da177e4c
|
257 |
{ |
78d8e58a0
|
258 |
if (error) |
4e4cbee93
|
259 |
bio->bi_status = error; |
797e7dbbe
|
260 |
|
e80640213
|
261 |
if (unlikely(rq->rq_flags & RQF_QUIET)) |
b7c44ed9d
|
262 |
bio_set_flag(bio, BIO_QUIET); |
08bafc034
|
263 |
|
f79ea4161
|
264 |
bio_advance(bio, nbytes); |
7ba1ba12e
|
265 |
|
143a87f4c
|
266 |
/* don't actually finish bio if it's part of flush sequence */ |
e80640213
|
267 |
if (bio->bi_iter.bi_size == 0 && !(rq->rq_flags & RQF_FLUSH_SEQ)) |
4246a0b63
|
268 |
bio_endio(bio); |
1da177e4c
|
269 |
} |
1da177e4c
|
270 |
|
1da177e4c
|
271 272 |
void blk_dump_rq_flags(struct request *rq, char *msg) { |
aebf526b5
|
273 274 275 |
printk(KERN_INFO "%s: dev %s: flags=%llx ", msg, rq->rq_disk ? rq->rq_disk->disk_name : "?", |
5953316db
|
276 |
(unsigned long long) rq->cmd_flags); |
1da177e4c
|
277 |
|
83096ebf1
|
278 279 280 281 |
printk(KERN_INFO " sector %llu, nr/cnr %u/%u ", (unsigned long long)blk_rq_pos(rq), blk_rq_sectors(rq), blk_rq_cur_sectors(rq)); |
b4f42e283
|
282 283 284 |
printk(KERN_INFO " bio %p, biotail %p, len %u ", rq->bio, rq->biotail, blk_rq_bytes(rq)); |
1da177e4c
|
285 |
} |
1da177e4c
|
286 |
EXPORT_SYMBOL(blk_dump_rq_flags); |
3cca6dc1c
|
287 |
static void blk_delay_work(struct work_struct *work) |
1da177e4c
|
288 |
{ |
3cca6dc1c
|
289 |
struct request_queue *q; |
1da177e4c
|
290 |
|
3cca6dc1c
|
291 292 |
q = container_of(work, struct request_queue, delay_work.work); spin_lock_irq(q->queue_lock); |
24ecfbe27
|
293 |
__blk_run_queue(q); |
3cca6dc1c
|
294 |
spin_unlock_irq(q->queue_lock); |
1da177e4c
|
295 |
} |
1da177e4c
|
296 297 |
/** |
3cca6dc1c
|
298 299 300 |
* blk_delay_queue - restart queueing after defined interval * @q: The &struct request_queue in question * @msecs: Delay in msecs |
1da177e4c
|
301 302 |
* * Description: |
3cca6dc1c
|
303 304 |
* Sometimes queueing needs to be postponed for a little while, to allow * resources to come back. This function will make sure that queueing is |
2fff8a924
|
305 |
* restarted around the specified time. |
3cca6dc1c
|
306 307 |
*/ void blk_delay_queue(struct request_queue *q, unsigned long msecs) |
2ad8b1ef1
|
308 |
{ |
2fff8a924
|
309 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
310 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
311 |
|
704605711
|
312 313 314 |
if (likely(!blk_queue_dead(q))) queue_delayed_work(kblockd_workqueue, &q->delay_work, msecs_to_jiffies(msecs)); |
2ad8b1ef1
|
315 |
} |
3cca6dc1c
|
316 |
EXPORT_SYMBOL(blk_delay_queue); |
2ad8b1ef1
|
317 |
|
1da177e4c
|
318 |
/** |
21491412f
|
319 320 321 322 323 324 325 326 327 328 |
* blk_start_queue_async - asynchronously restart a previously stopped queue * @q: The &struct request_queue in question * * Description: * blk_start_queue_async() will clear the stop flag on the queue, and * ensure that the request_fn for the queue is run from an async * context. **/ void blk_start_queue_async(struct request_queue *q) { |
2fff8a924
|
329 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
330 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
331 |
|
21491412f
|
332 333 334 335 336 337 |
queue_flag_clear(QUEUE_FLAG_STOPPED, q); blk_run_queue_async(q); } EXPORT_SYMBOL(blk_start_queue_async); /** |
1da177e4c
|
338 |
* blk_start_queue - restart a previously stopped queue |
165125e1e
|
339 |
* @q: The &struct request_queue in question |
1da177e4c
|
340 341 342 343 |
* * Description: * blk_start_queue() will clear the stop flag on the queue, and call * the request_fn for the queue if it was in a stopped state when |
2fff8a924
|
344 |
* entered. Also see blk_stop_queue(). |
1da177e4c
|
345 |
**/ |
165125e1e
|
346 |
void blk_start_queue(struct request_queue *q) |
1da177e4c
|
347 |
{ |
2fff8a924
|
348 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
349 |
WARN_ON_ONCE(q->mq_ops); |
a038e2536
|
350 |
|
75ad23bc0
|
351 |
queue_flag_clear(QUEUE_FLAG_STOPPED, q); |
24ecfbe27
|
352 |
__blk_run_queue(q); |
1da177e4c
|
353 |
} |
1da177e4c
|
354 355 356 357 |
EXPORT_SYMBOL(blk_start_queue); /** * blk_stop_queue - stop a queue |
165125e1e
|
358 |
* @q: The &struct request_queue in question |
1da177e4c
|
359 360 361 362 363 364 365 366 367 |
* * Description: * The Linux block layer assumes that a block driver will consume all * entries on the request queue when the request_fn strategy is called. * Often this will not happen, because of hardware limitations (queue * depth settings). If a device driver gets a 'queue full' response, * or if it simply chooses not to queue more I/O at one point, it can * call this function to prevent the request_fn from being called until * the driver has signalled it's ready to go again. This happens by calling |
2fff8a924
|
368 |
* blk_start_queue() to restart queue operations. |
1da177e4c
|
369 |
**/ |
165125e1e
|
370 |
void blk_stop_queue(struct request_queue *q) |
1da177e4c
|
371 |
{ |
2fff8a924
|
372 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
373 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
374 |
|
136b5721d
|
375 |
cancel_delayed_work(&q->delay_work); |
75ad23bc0
|
376 |
queue_flag_set(QUEUE_FLAG_STOPPED, q); |
1da177e4c
|
377 378 379 380 381 382 383 384 385 386 387 388 |
} EXPORT_SYMBOL(blk_stop_queue); /** * blk_sync_queue - cancel any pending callbacks on a queue * @q: the queue * * Description: * The block layer may perform asynchronous callback activity * on a queue, such as calling the unplug function after a timeout. * A block device may call blk_sync_queue to ensure that any * such activity is cancelled, thus allowing it to release resources |
59c51591a
|
389 |
* that the callbacks might use. The caller must already have made sure |
1da177e4c
|
390 391 392 |
* that its ->make_request_fn will not re-add plugging prior to calling * this function. * |
da5277700
|
393 |
* This function does not cancel any asynchronous activity arising |
da3dae54e
|
394 |
* out of elevator or throttling code. That would require elevator_exit() |
5efd61135
|
395 |
* and blkcg_exit_queue() to be called with queue lock initialized. |
da5277700
|
396 |
* |
1da177e4c
|
397 398 399 |
*/ void blk_sync_queue(struct request_queue *q) { |
70ed28b92
|
400 |
del_timer_sync(&q->timeout); |
4e9b6f208
|
401 |
cancel_work_sync(&q->timeout_work); |
f04c1fe76
|
402 403 404 405 |
if (q->mq_ops) { struct blk_mq_hw_ctx *hctx; int i; |
aba7afc56
|
406 |
cancel_delayed_work_sync(&q->requeue_work); |
21c6e939a
|
407 |
queue_for_each_hw_ctx(q, hctx, i) |
9f9937379
|
408 |
cancel_delayed_work_sync(&hctx->run_work); |
f04c1fe76
|
409 410 411 |
} else { cancel_delayed_work_sync(&q->delay_work); } |
1da177e4c
|
412 413 414 415 |
} EXPORT_SYMBOL(blk_sync_queue); /** |
c9254f2dd
|
416 417 418 419 420 421 422 423 |
* blk_set_preempt_only - set QUEUE_FLAG_PREEMPT_ONLY * @q: request queue pointer * * Returns the previous value of the PREEMPT_ONLY flag - 0 if the flag was not * set and 1 if the flag was already set. */ int blk_set_preempt_only(struct request_queue *q) { |
8814ce8a0
|
424 |
return blk_queue_flag_test_and_set(QUEUE_FLAG_PREEMPT_ONLY, q); |
c9254f2dd
|
425 426 427 428 429 |
} EXPORT_SYMBOL_GPL(blk_set_preempt_only); void blk_clear_preempt_only(struct request_queue *q) { |
8814ce8a0
|
430 |
blk_queue_flag_clear(QUEUE_FLAG_PREEMPT_ONLY, q); |
3a0a52997
|
431 |
wake_up_all(&q->mq_freeze_wq); |
c9254f2dd
|
432 433 434 435 |
} EXPORT_SYMBOL_GPL(blk_clear_preempt_only); /** |
c246e80d8
|
436 437 438 439 440 441 442 443 444 445 446 447 |
* __blk_run_queue_uncond - run a queue whether or not it has been stopped * @q: The queue to run * * Description: * Invoke request handling on a queue if there are any pending requests. * May be used to restart request handling after a request has completed. * This variant runs the queue whether or not the queue has been * stopped. Must be called with the queue lock held and interrupts * disabled. See also @blk_run_queue. */ inline void __blk_run_queue_uncond(struct request_queue *q) { |
2fff8a924
|
448 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
449 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
450 |
|
c246e80d8
|
451 452 |
if (unlikely(blk_queue_dead(q))) return; |
24faf6f60
|
453 454 455 456 457 458 459 460 |
/* * Some request_fn implementations, e.g. scsi_request_fn(), unlock * the queue lock internally. As a result multiple threads may be * running such a request function concurrently. Keep track of the * number of active request_fn invocations such that blk_drain_queue() * can wait until all these request_fn calls have finished. */ q->request_fn_active++; |
c246e80d8
|
461 |
q->request_fn(q); |
24faf6f60
|
462 |
q->request_fn_active--; |
c246e80d8
|
463 |
} |
a7928c157
|
464 |
EXPORT_SYMBOL_GPL(__blk_run_queue_uncond); |
c246e80d8
|
465 466 |
/** |
80a4b58e3
|
467 |
* __blk_run_queue - run a single device queue |
1da177e4c
|
468 |
* @q: The queue to run |
80a4b58e3
|
469 470 |
* * Description: |
2fff8a924
|
471 |
* See @blk_run_queue. |
1da177e4c
|
472 |
*/ |
24ecfbe27
|
473 |
void __blk_run_queue(struct request_queue *q) |
1da177e4c
|
474 |
{ |
2fff8a924
|
475 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
476 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
477 |
|
a538cd03b
|
478 479 |
if (unlikely(blk_queue_stopped(q))) return; |
c246e80d8
|
480 |
__blk_run_queue_uncond(q); |
75ad23bc0
|
481 482 |
} EXPORT_SYMBOL(__blk_run_queue); |
dac07ec12
|
483 |
|
75ad23bc0
|
484 |
/** |
24ecfbe27
|
485 486 487 488 489 |
* blk_run_queue_async - run a single device queue in workqueue context * @q: The queue to run * * Description: * Tells kblockd to perform the equivalent of @blk_run_queue on behalf |
2fff8a924
|
490 491 492 493 494 495 |
* of us. * * Note: * Since it is not allowed to run q->delay_work after blk_cleanup_queue() * has canceled q->delay_work, callers must hold the queue lock to avoid * race conditions between blk_cleanup_queue() and blk_run_queue_async(). |
24ecfbe27
|
496 497 498 |
*/ void blk_run_queue_async(struct request_queue *q) { |
2fff8a924
|
499 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
500 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
501 |
|
704605711
|
502 |
if (likely(!blk_queue_stopped(q) && !blk_queue_dead(q))) |
e7c2f9674
|
503 |
mod_delayed_work(kblockd_workqueue, &q->delay_work, 0); |
24ecfbe27
|
504 |
} |
c21e6beba
|
505 |
EXPORT_SYMBOL(blk_run_queue_async); |
24ecfbe27
|
506 507 |
/** |
75ad23bc0
|
508 509 |
* blk_run_queue - run a single device queue * @q: The queue to run |
80a4b58e3
|
510 511 512 |
* * Description: * Invoke request handling on this queue, if it has pending work to do. |
a7f557923
|
513 |
* May be used to restart queueing when a request has completed. |
75ad23bc0
|
514 515 516 517 |
*/ void blk_run_queue(struct request_queue *q) { unsigned long flags; |
332ebbf7f
|
518 |
WARN_ON_ONCE(q->mq_ops); |
75ad23bc0
|
519 |
spin_lock_irqsave(q->queue_lock, flags); |
24ecfbe27
|
520 |
__blk_run_queue(q); |
1da177e4c
|
521 522 523 |
spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL(blk_run_queue); |
165125e1e
|
524 |
void blk_put_queue(struct request_queue *q) |
483f4afc4
|
525 526 527 |
{ kobject_put(&q->kobj); } |
d86e0e83b
|
528 |
EXPORT_SYMBOL(blk_put_queue); |
483f4afc4
|
529 |
|
e3c78ca52
|
530 |
/** |
807592a4f
|
531 |
* __blk_drain_queue - drain requests from request_queue |
e3c78ca52
|
532 |
* @q: queue to drain |
c9a929dde
|
533 |
* @drain_all: whether to drain all requests or only the ones w/ ELVPRIV |
e3c78ca52
|
534 |
* |
c9a929dde
|
535 536 537 |
* Drain requests from @q. If @drain_all is set, all requests are drained. * If not, only ELVPRIV requests are drained. The caller is responsible * for ensuring that no new requests which need to be drained are queued. |
e3c78ca52
|
538 |
*/ |
807592a4f
|
539 540 541 |
static void __blk_drain_queue(struct request_queue *q, bool drain_all) __releases(q->queue_lock) __acquires(q->queue_lock) |
e3c78ca52
|
542 |
{ |
458f27a98
|
543 |
int i; |
807592a4f
|
544 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
545 |
WARN_ON_ONCE(q->mq_ops); |
807592a4f
|
546 |
|
e3c78ca52
|
547 |
while (true) { |
481a7d647
|
548 |
bool drain = false; |
e3c78ca52
|
549 |
|
b855b04a0
|
550 551 552 553 554 555 |
/* * The caller might be trying to drain @q before its * elevator is initialized. */ if (q->elevator) elv_drain_elevator(q); |
5efd61135
|
556 |
blkcg_drain_queue(q); |
e3c78ca52
|
557 |
|
4eabc9412
|
558 559 |
/* * This function might be called on a queue which failed |
b855b04a0
|
560 561 562 563 |
* driver init after queue creation or is not yet fully * active yet. Some drivers (e.g. fd and loop) get unhappy * in such cases. Kick queue iff dispatch queue has * something on it and @q has request_fn set. |
4eabc9412
|
564 |
*/ |
b855b04a0
|
565 |
if (!list_empty(&q->queue_head) && q->request_fn) |
4eabc9412
|
566 |
__blk_run_queue(q); |
c9a929dde
|
567 |
|
8a5ecdd42
|
568 |
drain |= q->nr_rqs_elvpriv; |
24faf6f60
|
569 |
drain |= q->request_fn_active; |
481a7d647
|
570 571 572 573 574 575 576 |
/* * Unfortunately, requests are queued at and tracked from * multiple places and there's no single counter which can * be drained. Check all the queues and counters. */ if (drain_all) { |
e97c293cd
|
577 |
struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL); |
481a7d647
|
578 579 |
drain |= !list_empty(&q->queue_head); for (i = 0; i < 2; i++) { |
8a5ecdd42
|
580 |
drain |= q->nr_rqs[i]; |
481a7d647
|
581 |
drain |= q->in_flight[i]; |
7c94e1c15
|
582 583 |
if (fq) drain |= !list_empty(&fq->flush_queue[i]); |
481a7d647
|
584 585 |
} } |
e3c78ca52
|
586 |
|
481a7d647
|
587 |
if (!drain) |
e3c78ca52
|
588 |
break; |
807592a4f
|
589 590 |
spin_unlock_irq(q->queue_lock); |
e3c78ca52
|
591 |
msleep(10); |
807592a4f
|
592 593 |
spin_lock_irq(q->queue_lock); |
e3c78ca52
|
594 |
} |
458f27a98
|
595 596 597 598 599 600 601 |
/* * With queue marked dead, any woken up waiter will fail the * allocation path, so the wakeup chaining is lost and we're * left with hung waiters. We need to wake up those waiters. */ if (q->request_fn) { |
a051661ca
|
602 |
struct request_list *rl; |
a051661ca
|
603 604 605 |
blk_queue_for_each_rl(rl, q) for (i = 0; i < ARRAY_SIZE(rl->wait); i++) wake_up_all(&rl->wait[i]); |
458f27a98
|
606 |
} |
e3c78ca52
|
607 |
} |
454be724f
|
608 609 610 611 612 613 |
void blk_drain_queue(struct request_queue *q) { spin_lock_irq(q->queue_lock); __blk_drain_queue(q, true); spin_unlock_irq(q->queue_lock); } |
c9a929dde
|
614 |
/** |
d732580b4
|
615 616 617 618 619 |
* blk_queue_bypass_start - enter queue bypass mode * @q: queue of interest * * In bypass mode, only the dispatch FIFO queue of @q is used. This * function makes @q enter bypass mode and drains all requests which were |
6ecf23afa
|
620 |
* throttled or issued before. On return, it's guaranteed that no request |
80fd99792
|
621 622 |
* is being throttled or has ELVPRIV set and blk_queue_bypass() %true * inside queue or RCU read lock. |
d732580b4
|
623 624 625 |
*/ void blk_queue_bypass_start(struct request_queue *q) { |
332ebbf7f
|
626 |
WARN_ON_ONCE(q->mq_ops); |
d732580b4
|
627 |
spin_lock_irq(q->queue_lock); |
776687bce
|
628 |
q->bypass_depth++; |
d732580b4
|
629 630 |
queue_flag_set(QUEUE_FLAG_BYPASS, q); spin_unlock_irq(q->queue_lock); |
776687bce
|
631 632 633 634 635 636 |
/* * Queues start drained. Skip actual draining till init is * complete. This avoids lenghty delays during queue init which * can happen many times during boot. */ if (blk_queue_init_done(q)) { |
807592a4f
|
637 638 639 |
spin_lock_irq(q->queue_lock); __blk_drain_queue(q, false); spin_unlock_irq(q->queue_lock); |
b82d4b197
|
640 641 642 |
/* ensure blk_queue_bypass() is %true inside RCU read lock */ synchronize_rcu(); } |
d732580b4
|
643 644 645 646 647 648 649 650 |
} EXPORT_SYMBOL_GPL(blk_queue_bypass_start); /** * blk_queue_bypass_end - leave queue bypass mode * @q: queue of interest * * Leave bypass mode and restore the normal queueing behavior. |
332ebbf7f
|
651 652 653 |
* * Note: although blk_queue_bypass_start() is only called for blk-sq queues, * this function is called for both blk-sq and blk-mq queues. |
d732580b4
|
654 655 656 657 658 659 660 661 662 663 |
*/ void blk_queue_bypass_end(struct request_queue *q) { spin_lock_irq(q->queue_lock); if (!--q->bypass_depth) queue_flag_clear(QUEUE_FLAG_BYPASS, q); WARN_ON_ONCE(q->bypass_depth < 0); spin_unlock_irq(q->queue_lock); } EXPORT_SYMBOL_GPL(blk_queue_bypass_end); |
aed3ea94b
|
664 665 |
void blk_set_queue_dying(struct request_queue *q) { |
8814ce8a0
|
666 |
blk_queue_flag_set(QUEUE_FLAG_DYING, q); |
aed3ea94b
|
667 |
|
d3cfb2a0a
|
668 669 670 671 672 673 |
/* * When queue DYING flag is set, we need to block new req * entering queue, so we call blk_freeze_queue_start() to * prevent I/O from crossing blk_queue_enter(). */ blk_freeze_queue_start(q); |
aed3ea94b
|
674 675 676 677 |
if (q->mq_ops) blk_mq_wake_waiters(q); else { struct request_list *rl; |
bbfc3c5d6
|
678 |
spin_lock_irq(q->queue_lock); |
aed3ea94b
|
679 680 |
blk_queue_for_each_rl(rl, q) { if (rl->rq_pool) { |
34d9715ac
|
681 682 |
wake_up_all(&rl->wait[BLK_RW_SYNC]); wake_up_all(&rl->wait[BLK_RW_ASYNC]); |
aed3ea94b
|
683 684 |
} } |
bbfc3c5d6
|
685 |
spin_unlock_irq(q->queue_lock); |
aed3ea94b
|
686 |
} |
055f6e18e
|
687 688 689 |
/* Make blk_queue_enter() reexamine the DYING flag. */ wake_up_all(&q->mq_freeze_wq); |
aed3ea94b
|
690 691 |
} EXPORT_SYMBOL_GPL(blk_set_queue_dying); |
4cf6324b1
|
692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 |
/* Unconfigure the I/O scheduler and dissociate from the cgroup controller. */ void blk_exit_queue(struct request_queue *q) { /* * Since the I/O scheduler exit code may access cgroup information, * perform I/O scheduler exit before disassociating from the block * cgroup controller. */ if (q->elevator) { ioc_clear_queue(q); elevator_exit(q, q->elevator); q->elevator = NULL; } /* * Remove all references to @q from the block cgroup controller before * restoring @q->queue_lock to avoid that restoring this pointer causes * e.g. blkcg_print_blkgs() to crash. */ blkcg_exit_queue(q); /* * Since the cgroup code may dereference the @q->backing_dev_info * pointer, only decrease its reference count after having removed the * association with the block cgroup controller. */ bdi_put(q->backing_dev_info); } |
d732580b4
|
720 |
/** |
c9a929dde
|
721 722 723 |
* blk_cleanup_queue - shutdown a request queue * @q: request queue to shutdown * |
c246e80d8
|
724 725 |
* Mark @q DYING, drain all pending requests, mark @q DEAD, destroy and * put it. All future requests will be failed immediately with -ENODEV. |
c94a96ac9
|
726 |
*/ |
6728cb0e6
|
727 |
void blk_cleanup_queue(struct request_queue *q) |
483f4afc4
|
728 |
{ |
c9a929dde
|
729 |
spinlock_t *lock = q->queue_lock; |
e3335de94
|
730 |
|
3f3299d5c
|
731 |
/* mark @q DYING, no new request or merges will be allowed afterwards */ |
483f4afc4
|
732 |
mutex_lock(&q->sysfs_lock); |
aed3ea94b
|
733 |
blk_set_queue_dying(q); |
c9a929dde
|
734 |
spin_lock_irq(lock); |
6ecf23afa
|
735 |
|
80fd99792
|
736 |
/* |
3f3299d5c
|
737 |
* A dying queue is permanently in bypass mode till released. Note |
80fd99792
|
738 739 740 741 742 743 744 |
* that, unlike blk_queue_bypass_start(), we aren't performing * synchronize_rcu() after entering bypass mode to avoid the delay * as some drivers create and destroy a lot of queues while * probing. This is still safe because blk_release_queue() will be * called only after the queue refcnt drops to zero and nothing, * RCU or not, would be traversing the queue by then. */ |
6ecf23afa
|
745 746 |
q->bypass_depth++; queue_flag_set(QUEUE_FLAG_BYPASS, q); |
c9a929dde
|
747 748 |
queue_flag_set(QUEUE_FLAG_NOMERGES, q); queue_flag_set(QUEUE_FLAG_NOXMERGES, q); |
3f3299d5c
|
749 |
queue_flag_set(QUEUE_FLAG_DYING, q); |
c9a929dde
|
750 751 |
spin_unlock_irq(lock); mutex_unlock(&q->sysfs_lock); |
c246e80d8
|
752 753 754 755 |
/* * Drain all requests queued before DYING marking. Set DEAD flag to * prevent that q->request_fn() gets invoked after draining finished. */ |
3ef28e83a
|
756 |
blk_freeze_queue(q); |
9c1051aac
|
757 |
spin_lock_irq(lock); |
c246e80d8
|
758 |
queue_flag_set(QUEUE_FLAG_DEAD, q); |
807592a4f
|
759 |
spin_unlock_irq(lock); |
c9a929dde
|
760 |
|
c2856ae2f
|
761 762 763 764 |
/* * make sure all in-progress dispatch are completed because * blk_freeze_queue() can only complete all requests, and * dispatch may still be in-progress since we dispatch requests |
1311326cf
|
765 766 |
* from more than one contexts. * |
410306a0f
|
767 768 |
* We rely on driver to deal with the race in case that queue * initialization isn't done. |
c2856ae2f
|
769 |
*/ |
1311326cf
|
770 |
if (q->mq_ops && blk_queue_init_done(q)) |
c2856ae2f
|
771 |
blk_mq_quiesce_queue(q); |
5a48fc147
|
772 773 |
/* for synchronous bio-based driver finish in-flight integrity i/o */ blk_flush_integrity(); |
c9a929dde
|
774 |
/* @q won't process any more request, flush async actions */ |
dc3b17cc8
|
775 |
del_timer_sync(&q->backing_dev_info->laptop_mode_wb_timer); |
c9a929dde
|
776 |
blk_sync_queue(q); |
a063057d7
|
777 778 779 780 781 |
/* * I/O scheduler exit is only safe after the sysfs scheduler attribute * has been removed. */ WARN_ON_ONCE(q->kobj.state_in_sysfs); |
4cf6324b1
|
782 |
blk_exit_queue(q); |
a063057d7
|
783 |
|
45a9c9d90
|
784 785 |
if (q->mq_ops) blk_mq_free_queue(q); |
3ef28e83a
|
786 |
percpu_ref_exit(&q->q_usage_counter); |
45a9c9d90
|
787 |
|
5e5cfac0c
|
788 789 790 791 |
spin_lock_irq(lock); if (q->queue_lock != &q->__queue_lock) q->queue_lock = &q->__queue_lock; spin_unlock_irq(lock); |
c9a929dde
|
792 |
/* @q is and will stay empty, shutdown and put */ |
483f4afc4
|
793 794 |
blk_put_queue(q); } |
1da177e4c
|
795 |
EXPORT_SYMBOL(blk_cleanup_queue); |
271508dba
|
796 |
/* Allocate memory local to the request queue */ |
6d247d7f7
|
797 |
static void *alloc_request_simple(gfp_t gfp_mask, void *data) |
271508dba
|
798 |
{ |
6d247d7f7
|
799 800 801 |
struct request_queue *q = data; return kmem_cache_alloc_node(request_cachep, gfp_mask, q->node); |
271508dba
|
802 |
} |
6d247d7f7
|
803 |
static void free_request_simple(void *element, void *data) |
271508dba
|
804 805 806 |
{ kmem_cache_free(request_cachep, element); } |
6d247d7f7
|
807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 |
static void *alloc_request_size(gfp_t gfp_mask, void *data) { struct request_queue *q = data; struct request *rq; rq = kmalloc_node(sizeof(struct request) + q->cmd_size, gfp_mask, q->node); if (rq && q->init_rq_fn && q->init_rq_fn(q, rq, gfp_mask) < 0) { kfree(rq); rq = NULL; } return rq; } static void free_request_size(void *element, void *data) { struct request_queue *q = data; if (q->exit_rq_fn) q->exit_rq_fn(q, element); kfree(element); } |
5b788ce3e
|
829 830 |
int blk_init_rl(struct request_list *rl, struct request_queue *q, gfp_t gfp_mask) |
1da177e4c
|
831 |
{ |
85acb3ba2
|
832 |
if (unlikely(rl->rq_pool) || q->mq_ops) |
1abec4fdb
|
833 |
return 0; |
5b788ce3e
|
834 |
rl->q = q; |
1faa16d22
|
835 836 |
rl->count[BLK_RW_SYNC] = rl->count[BLK_RW_ASYNC] = 0; rl->starved[BLK_RW_SYNC] = rl->starved[BLK_RW_ASYNC] = 0; |
1faa16d22
|
837 838 |
init_waitqueue_head(&rl->wait[BLK_RW_SYNC]); init_waitqueue_head(&rl->wait[BLK_RW_ASYNC]); |
1da177e4c
|
839 |
|
6d247d7f7
|
840 841 842 843 844 845 846 847 848 |
if (q->cmd_size) { rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, alloc_request_size, free_request_size, q, gfp_mask, q->node); } else { rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, alloc_request_simple, free_request_simple, q, gfp_mask, q->node); } |
1da177e4c
|
849 850 |
if (!rl->rq_pool) return -ENOMEM; |
b425e5049
|
851 852 |
if (rl != &q->root_rl) WARN_ON_ONCE(!blk_get_queue(q)); |
1da177e4c
|
853 854 |
return 0; } |
b425e5049
|
855 |
void blk_exit_rl(struct request_queue *q, struct request_list *rl) |
5b788ce3e
|
856 |
{ |
b425e5049
|
857 |
if (rl->rq_pool) { |
5b788ce3e
|
858 |
mempool_destroy(rl->rq_pool); |
b425e5049
|
859 860 861 |
if (rl != &q->root_rl) blk_put_queue(q); } |
5b788ce3e
|
862 |
} |
165125e1e
|
863 |
struct request_queue *blk_alloc_queue(gfp_t gfp_mask) |
1da177e4c
|
864 |
{ |
5ee0524ba
|
865 |
return blk_alloc_queue_node(gfp_mask, NUMA_NO_NODE, NULL); |
1946089a1
|
866 867 |
} EXPORT_SYMBOL(blk_alloc_queue); |
1da177e4c
|
868 |
|
3a0a52997
|
869 870 871 872 873 |
/** * blk_queue_enter() - try to increase q->q_usage_counter * @q: request queue pointer * @flags: BLK_MQ_REQ_NOWAIT and/or BLK_MQ_REQ_PREEMPT */ |
9a95e4ef7
|
874 |
int blk_queue_enter(struct request_queue *q, blk_mq_req_flags_t flags) |
3ef28e83a
|
875 |
{ |
3a0a52997
|
876 |
const bool preempt = flags & BLK_MQ_REQ_PREEMPT; |
3ef28e83a
|
877 |
while (true) { |
3a0a52997
|
878 |
bool success = false; |
3ef28e83a
|
879 |
|
818e0fa29
|
880 |
rcu_read_lock(); |
3a0a52997
|
881 882 883 884 885 886 887 888 889 890 891 892 |
if (percpu_ref_tryget_live(&q->q_usage_counter)) { /* * The code that sets the PREEMPT_ONLY flag is * responsible for ensuring that that flag is globally * visible before the queue is unfrozen. */ if (preempt || !blk_queue_preempt_only(q)) { success = true; } else { percpu_ref_put(&q->q_usage_counter); } } |
818e0fa29
|
893 |
rcu_read_unlock(); |
3a0a52997
|
894 895 |
if (success) |
3ef28e83a
|
896 |
return 0; |
3a0a52997
|
897 |
if (flags & BLK_MQ_REQ_NOWAIT) |
3ef28e83a
|
898 |
return -EBUSY; |
5ed61d3f0
|
899 |
/* |
1671d522c
|
900 |
* read pair of barrier in blk_freeze_queue_start(), |
5ed61d3f0
|
901 |
* we need to order reading __PERCPU_REF_DEAD flag of |
d3cfb2a0a
|
902 903 904 |
* .q_usage_counter and reading .mq_freeze_depth or * queue dying flag, otherwise the following wait may * never return if the two reads are reordered. |
5ed61d3f0
|
905 906 |
*/ smp_rmb(); |
1dc3039bc
|
907 908 909 910 |
wait_event(q->mq_freeze_wq, (atomic_read(&q->mq_freeze_depth) == 0 && (preempt || !blk_queue_preempt_only(q))) || blk_queue_dying(q)); |
3ef28e83a
|
911 912 |
if (blk_queue_dying(q)) return -ENODEV; |
3ef28e83a
|
913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 |
} } void blk_queue_exit(struct request_queue *q) { percpu_ref_put(&q->q_usage_counter); } static void blk_queue_usage_counter_release(struct percpu_ref *ref) { struct request_queue *q = container_of(ref, struct request_queue, q_usage_counter); wake_up_all(&q->mq_freeze_wq); } |
bca237a52
|
928 |
static void blk_rq_timed_out_timer(struct timer_list *t) |
287922eb0
|
929 |
{ |
bca237a52
|
930 |
struct request_queue *q = from_timer(q, t, timeout); |
287922eb0
|
931 932 933 |
kblockd_schedule_work(&q->timeout_work); } |
498f6650a
|
934 935 936 937 938 939 940 941 942 943 944 945 946 |
/** * blk_alloc_queue_node - allocate a request queue * @gfp_mask: memory allocation flags * @node_id: NUMA node to allocate memory from * @lock: For legacy queues, pointer to a spinlock that will be used to e.g. * serialize calls to the legacy .request_fn() callback. Ignored for * blk-mq request queues. * * Note: pass the queue lock as the third argument to this function instead of * setting the queue lock pointer explicitly to avoid triggering a sporadic * crash in the blkcg code. This function namely calls blkcg_init_queue() and * the queue lock pointer must be set before blkcg_init_queue() is called. */ |
5ee0524ba
|
947 948 |
struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id, spinlock_t *lock) |
1946089a1
|
949 |
{ |
165125e1e
|
950 |
struct request_queue *q; |
338aa96d5
|
951 |
int ret; |
1946089a1
|
952 |
|
8324aa91d
|
953 |
q = kmem_cache_alloc_node(blk_requestq_cachep, |
94f6030ca
|
954 |
gfp_mask | __GFP_ZERO, node_id); |
1da177e4c
|
955 956 |
if (!q) return NULL; |
cbf62af35
|
957 958 959 960 |
INIT_LIST_HEAD(&q->queue_head); q->last_merge = NULL; q->end_sector = 0; q->boundary_rq = NULL; |
00380a404
|
961 |
q->id = ida_simple_get(&blk_queue_ida, 0, 0, gfp_mask); |
a73f730d0
|
962 |
if (q->id < 0) |
3d2936f45
|
963 |
goto fail_q; |
a73f730d0
|
964 |
|
338aa96d5
|
965 966 |
ret = bioset_init(&q->bio_split, BIO_POOL_SIZE, 0, BIOSET_NEED_BVECS); if (ret) |
54efd50bf
|
967 |
goto fail_id; |
d03f6cdc1
|
968 969 970 |
q->backing_dev_info = bdi_alloc_node(gfp_mask, node_id); if (!q->backing_dev_info) goto fail_split; |
a83b576c9
|
971 972 973 |
q->stats = blk_alloc_queue_stats(); if (!q->stats) goto fail_stats; |
dc3b17cc8
|
974 |
q->backing_dev_info->ra_pages = |
09cbfeaf1
|
975 |
(VM_MAX_READAHEAD * 1024) / PAGE_SIZE; |
dc3b17cc8
|
976 977 |
q->backing_dev_info->capabilities = BDI_CAP_CGROUP_WRITEBACK; q->backing_dev_info->name = "block"; |
5151412dd
|
978 |
q->node = node_id; |
0989a025d
|
979 |
|
bca237a52
|
980 981 982 |
timer_setup(&q->backing_dev_info->laptop_mode_wb_timer, laptop_mode_timer_fn, 0); timer_setup(&q->timeout, blk_rq_timed_out_timer, 0); |
4e9b6f208
|
983 |
INIT_WORK(&q->timeout_work, NULL); |
242f9dcb8
|
984 |
INIT_LIST_HEAD(&q->timeout_list); |
a612fddf0
|
985 |
INIT_LIST_HEAD(&q->icq_list); |
4eef30499
|
986 |
#ifdef CONFIG_BLK_CGROUP |
e8989fae3
|
987 |
INIT_LIST_HEAD(&q->blkg_list); |
4eef30499
|
988 |
#endif |
3cca6dc1c
|
989 |
INIT_DELAYED_WORK(&q->delay_work, blk_delay_work); |
483f4afc4
|
990 |
|
8324aa91d
|
991 |
kobject_init(&q->kobj, &blk_queue_ktype); |
1da177e4c
|
992 |
|
5acb3cc2c
|
993 994 995 |
#ifdef CONFIG_BLK_DEV_IO_TRACE mutex_init(&q->blk_trace_mutex); #endif |
483f4afc4
|
996 |
mutex_init(&q->sysfs_lock); |
e7e72bf64
|
997 |
spin_lock_init(&q->__queue_lock); |
483f4afc4
|
998 |
|
498f6650a
|
999 1000 |
if (!q->mq_ops) q->queue_lock = lock ? : &q->__queue_lock; |
c94a96ac9
|
1001 |
|
b82d4b197
|
1002 1003 1004 |
/* * A queue starts its life with bypass turned on to avoid * unnecessary bypass on/off overhead and nasty surprises during |
749fefe67
|
1005 1006 |
* init. The initial bypass will be finished when the queue is * registered by blk_register_queue(). |
b82d4b197
|
1007 1008 |
*/ q->bypass_depth = 1; |
f78bac2c8
|
1009 |
queue_flag_set_unlocked(QUEUE_FLAG_BYPASS, q); |
b82d4b197
|
1010 |
|
320ae51fe
|
1011 |
init_waitqueue_head(&q->mq_freeze_wq); |
3ef28e83a
|
1012 1013 1014 1015 1016 1017 1018 |
/* * Init percpu_ref in atomic mode so that it's faster to shutdown. * See blk_register_queue() for details. */ if (percpu_ref_init(&q->q_usage_counter, blk_queue_usage_counter_release, PERCPU_REF_INIT_ATOMIC, GFP_KERNEL)) |
fff4996b7
|
1019 |
goto fail_bdi; |
f51b802c1
|
1020 |
|
3ef28e83a
|
1021 1022 |
if (blkcg_init_queue(q)) goto fail_ref; |
1da177e4c
|
1023 |
return q; |
a73f730d0
|
1024 |
|
3ef28e83a
|
1025 1026 |
fail_ref: percpu_ref_exit(&q->q_usage_counter); |
fff4996b7
|
1027 |
fail_bdi: |
a83b576c9
|
1028 1029 |
blk_free_queue_stats(q->stats); fail_stats: |
d03f6cdc1
|
1030 |
bdi_put(q->backing_dev_info); |
54efd50bf
|
1031 |
fail_split: |
338aa96d5
|
1032 |
bioset_exit(&q->bio_split); |
a73f730d0
|
1033 1034 1035 1036 1037 |
fail_id: ida_simple_remove(&blk_queue_ida, q->id); fail_q: kmem_cache_free(blk_requestq_cachep, q); return NULL; |
1da177e4c
|
1038 |
} |
1946089a1
|
1039 |
EXPORT_SYMBOL(blk_alloc_queue_node); |
1da177e4c
|
1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 |
/** * blk_init_queue - prepare a request queue for use with a block device * @rfn: The function to be called to process requests that have been * placed on the queue. * @lock: Request queue spin lock * * Description: * If a block device wishes to use the standard request handling procedures, * which sorts requests and coalesces adjacent requests, then it must * call blk_init_queue(). The function @rfn will be called when there * are requests on the queue that need to be processed. If the device * supports plugging, then @rfn may not be called immediately when requests * are available on the queue, but may be called at some time later instead. * Plugged queues are generally unplugged when a buffer belonging to one * of the requests on the queue is needed, or due to memory pressure. * * @rfn is not required, or even expected, to remove all requests off the * queue, but only as many as it can handle at a time. If it does leave * requests on the queue, it is responsible for arranging that the requests * get dealt with eventually. * * The queue spin lock must be held while manipulating the requests on the |
a038e2536
|
1063 1064 |
* request queue; this lock will be taken also from interrupt context, so irq * disabling is needed for it. |
1da177e4c
|
1065 |
* |
710027a48
|
1066 |
* Function returns a pointer to the initialized request queue, or %NULL if |
1da177e4c
|
1067 1068 1069 1070 1071 1072 |
* it didn't succeed. * * Note: * blk_init_queue() must be paired with a blk_cleanup_queue() call * when the block device is deactivated (such as at module unload). **/ |
1946089a1
|
1073 |
|
165125e1e
|
1074 |
struct request_queue *blk_init_queue(request_fn_proc *rfn, spinlock_t *lock) |
1da177e4c
|
1075 |
{ |
c304a51bf
|
1076 |
return blk_init_queue_node(rfn, lock, NUMA_NO_NODE); |
1946089a1
|
1077 1078 |
} EXPORT_SYMBOL(blk_init_queue); |
165125e1e
|
1079 |
struct request_queue * |
1946089a1
|
1080 1081 |
blk_init_queue_node(request_fn_proc *rfn, spinlock_t *lock, int node_id) { |
5ea708d15
|
1082 |
struct request_queue *q; |
1da177e4c
|
1083 |
|
498f6650a
|
1084 |
q = blk_alloc_queue_node(GFP_KERNEL, node_id, lock); |
5ea708d15
|
1085 |
if (!q) |
c86d1b8ae
|
1086 |
return NULL; |
5ea708d15
|
1087 |
q->request_fn = rfn; |
5ea708d15
|
1088 1089 1090 1091 |
if (blk_init_allocated_queue(q) < 0) { blk_cleanup_queue(q); return NULL; } |
18741986a
|
1092 |
|
7982e90c3
|
1093 |
return q; |
01effb0dc
|
1094 1095 |
} EXPORT_SYMBOL(blk_init_queue_node); |
dece16353
|
1096 |
static blk_qc_t blk_queue_bio(struct request_queue *q, struct bio *bio); |
336b7e1f2
|
1097 |
|
1da177e4c
|
1098 |
|
5ea708d15
|
1099 1100 |
int blk_init_allocated_queue(struct request_queue *q) { |
332ebbf7f
|
1101 |
WARN_ON_ONCE(q->mq_ops); |
6d247d7f7
|
1102 |
q->fq = blk_alloc_flush_queue(q, NUMA_NO_NODE, q->cmd_size); |
ba483388e
|
1103 |
if (!q->fq) |
5ea708d15
|
1104 |
return -ENOMEM; |
7982e90c3
|
1105 |
|
6d247d7f7
|
1106 1107 |
if (q->init_rq_fn && q->init_rq_fn(q, q->fq->flush_rq, GFP_KERNEL)) goto out_free_flush_queue; |
7982e90c3
|
1108 |
|
a051661ca
|
1109 |
if (blk_init_rl(&q->root_rl, q, GFP_KERNEL)) |
6d247d7f7
|
1110 |
goto out_exit_flush_rq; |
1da177e4c
|
1111 |
|
287922eb0
|
1112 |
INIT_WORK(&q->timeout_work, blk_timeout_work); |
60ea8226c
|
1113 |
q->queue_flags |= QUEUE_FLAG_DEFAULT; |
c94a96ac9
|
1114 |
|
f3b144aa7
|
1115 1116 1117 |
/* * This also sets hw/phys segments, boundary and size */ |
c20e8de27
|
1118 |
blk_queue_make_request(q, blk_queue_bio); |
1da177e4c
|
1119 |
|
44ec95425
|
1120 |
q->sg_reserved_size = INT_MAX; |
acddf3b30
|
1121 |
if (elevator_init(q)) |
6d247d7f7
|
1122 |
goto out_exit_flush_rq; |
5ea708d15
|
1123 |
return 0; |
eb1c160b2
|
1124 |
|
6d247d7f7
|
1125 1126 1127 1128 |
out_exit_flush_rq: if (q->exit_rq_fn) q->exit_rq_fn(q, q->fq->flush_rq); out_free_flush_queue: |
ba483388e
|
1129 |
blk_free_flush_queue(q->fq); |
54648cf1e
|
1130 |
q->fq = NULL; |
5ea708d15
|
1131 |
return -ENOMEM; |
1da177e4c
|
1132 |
} |
5151412dd
|
1133 |
EXPORT_SYMBOL(blk_init_allocated_queue); |
1da177e4c
|
1134 |
|
09ac46c42
|
1135 |
bool blk_get_queue(struct request_queue *q) |
1da177e4c
|
1136 |
{ |
3f3299d5c
|
1137 |
if (likely(!blk_queue_dying(q))) { |
09ac46c42
|
1138 1139 |
__blk_get_queue(q); return true; |
1da177e4c
|
1140 |
} |
09ac46c42
|
1141 |
return false; |
1da177e4c
|
1142 |
} |
d86e0e83b
|
1143 |
EXPORT_SYMBOL(blk_get_queue); |
1da177e4c
|
1144 |
|
5b788ce3e
|
1145 |
static inline void blk_free_request(struct request_list *rl, struct request *rq) |
1da177e4c
|
1146 |
{ |
e80640213
|
1147 |
if (rq->rq_flags & RQF_ELVPRIV) { |
5b788ce3e
|
1148 |
elv_put_request(rl->q, rq); |
f1f8cc946
|
1149 |
if (rq->elv.icq) |
11a3122f6
|
1150 |
put_io_context(rq->elv.icq->ioc); |
f1f8cc946
|
1151 |
} |
5b788ce3e
|
1152 |
mempool_free(rq, rl->rq_pool); |
1da177e4c
|
1153 |
} |
1da177e4c
|
1154 1155 1156 1157 |
/* * ioc_batching returns true if the ioc is a valid batching request and * should be given priority access to a request. */ |
165125e1e
|
1158 |
static inline int ioc_batching(struct request_queue *q, struct io_context *ioc) |
1da177e4c
|
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 |
{ if (!ioc) return 0; /* * Make sure the process is able to allocate at least 1 request * even if the batch times out, otherwise we could theoretically * lose wakeups. */ return ioc->nr_batch_requests == q->nr_batching || (ioc->nr_batch_requests > 0 && time_before(jiffies, ioc->last_waited + BLK_BATCH_TIME)); } /* * ioc_set_batching sets ioc to be a new "batcher" if it is not one. This * will cause the process to be a "batcher" on all queues in the system. This * is the behaviour we want though - once it gets a wakeup it should be given * a nice run. */ |
165125e1e
|
1179 |
static void ioc_set_batching(struct request_queue *q, struct io_context *ioc) |
1da177e4c
|
1180 1181 1182 1183 1184 1185 1186 |
{ if (!ioc || ioc_batching(q, ioc)) return; ioc->nr_batch_requests = q->nr_batching; ioc->last_waited = jiffies; } |
5b788ce3e
|
1187 |
static void __freed_request(struct request_list *rl, int sync) |
1da177e4c
|
1188 |
{ |
5b788ce3e
|
1189 |
struct request_queue *q = rl->q; |
1da177e4c
|
1190 |
|
d40f75a06
|
1191 1192 |
if (rl->count[sync] < queue_congestion_off_threshold(q)) blk_clear_congested(rl, sync); |
1da177e4c
|
1193 |
|
1faa16d22
|
1194 1195 1196 |
if (rl->count[sync] + 1 <= q->nr_requests) { if (waitqueue_active(&rl->wait[sync])) wake_up(&rl->wait[sync]); |
1da177e4c
|
1197 |
|
5b788ce3e
|
1198 |
blk_clear_rl_full(rl, sync); |
1da177e4c
|
1199 1200 1201 1202 1203 1204 1205 |
} } /* * A request has just been released. Account for it, update the full and * congestion status, wake up any waiters. Called under q->queue_lock. */ |
e80640213
|
1206 1207 |
static void freed_request(struct request_list *rl, bool sync, req_flags_t rq_flags) |
1da177e4c
|
1208 |
{ |
5b788ce3e
|
1209 |
struct request_queue *q = rl->q; |
1da177e4c
|
1210 |
|
8a5ecdd42
|
1211 |
q->nr_rqs[sync]--; |
1faa16d22
|
1212 |
rl->count[sync]--; |
e80640213
|
1213 |
if (rq_flags & RQF_ELVPRIV) |
8a5ecdd42
|
1214 |
q->nr_rqs_elvpriv--; |
1da177e4c
|
1215 |
|
5b788ce3e
|
1216 |
__freed_request(rl, sync); |
1da177e4c
|
1217 |
|
1faa16d22
|
1218 |
if (unlikely(rl->starved[sync ^ 1])) |
5b788ce3e
|
1219 |
__freed_request(rl, sync ^ 1); |
1da177e4c
|
1220 |
} |
e3a2b3f93
|
1221 1222 1223 |
int blk_update_nr_requests(struct request_queue *q, unsigned int nr) { struct request_list *rl; |
d40f75a06
|
1224 |
int on_thresh, off_thresh; |
e3a2b3f93
|
1225 |
|
332ebbf7f
|
1226 |
WARN_ON_ONCE(q->mq_ops); |
e3a2b3f93
|
1227 1228 1229 |
spin_lock_irq(q->queue_lock); q->nr_requests = nr; blk_queue_congestion_threshold(q); |
d40f75a06
|
1230 1231 |
on_thresh = queue_congestion_on_threshold(q); off_thresh = queue_congestion_off_threshold(q); |
e3a2b3f93
|
1232 |
|
d40f75a06
|
1233 1234 1235 1236 1237 |
blk_queue_for_each_rl(rl, q) { if (rl->count[BLK_RW_SYNC] >= on_thresh) blk_set_congested(rl, BLK_RW_SYNC); else if (rl->count[BLK_RW_SYNC] < off_thresh) blk_clear_congested(rl, BLK_RW_SYNC); |
e3a2b3f93
|
1238 |
|
d40f75a06
|
1239 1240 1241 1242 |
if (rl->count[BLK_RW_ASYNC] >= on_thresh) blk_set_congested(rl, BLK_RW_ASYNC); else if (rl->count[BLK_RW_ASYNC] < off_thresh) blk_clear_congested(rl, BLK_RW_ASYNC); |
e3a2b3f93
|
1243 |
|
e3a2b3f93
|
1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 |
if (rl->count[BLK_RW_SYNC] >= q->nr_requests) { blk_set_rl_full(rl, BLK_RW_SYNC); } else { blk_clear_rl_full(rl, BLK_RW_SYNC); wake_up(&rl->wait[BLK_RW_SYNC]); } if (rl->count[BLK_RW_ASYNC] >= q->nr_requests) { blk_set_rl_full(rl, BLK_RW_ASYNC); } else { blk_clear_rl_full(rl, BLK_RW_ASYNC); wake_up(&rl->wait[BLK_RW_ASYNC]); } } spin_unlock_irq(q->queue_lock); return 0; } |
da8303c63
|
1262 |
/** |
a06e05e6a
|
1263 |
* __get_request - get a free request |
5b788ce3e
|
1264 |
* @rl: request list to allocate from |
ef295ecf0
|
1265 |
* @op: operation and flags |
da8303c63
|
1266 |
* @bio: bio to allocate request for (can be %NULL) |
6a15674d1
|
1267 |
* @flags: BLQ_MQ_REQ_* flags |
4accf5fc7
|
1268 |
* @gfp_mask: allocator flags |
da8303c63
|
1269 1270 1271 1272 |
* * Get a free request from @q. This function may fail under memory * pressure or if @q is dead. * |
da3dae54e
|
1273 |
* Must be called with @q->queue_lock held and, |
a492f0754
|
1274 1275 |
* Returns ERR_PTR on failure, with @q->queue_lock held. * Returns request pointer on success, with @q->queue_lock *not held*. |
1da177e4c
|
1276 |
*/ |
ef295ecf0
|
1277 |
static struct request *__get_request(struct request_list *rl, unsigned int op, |
4accf5fc7
|
1278 |
struct bio *bio, blk_mq_req_flags_t flags, gfp_t gfp_mask) |
1da177e4c
|
1279 |
{ |
5b788ce3e
|
1280 |
struct request_queue *q = rl->q; |
b679281a6
|
1281 |
struct request *rq; |
7f4b35d15
|
1282 1283 |
struct elevator_type *et = q->elevator->type; struct io_context *ioc = rq_ioc(bio); |
f1f8cc946
|
1284 |
struct io_cq *icq = NULL; |
ef295ecf0
|
1285 |
const bool is_sync = op_is_sync(op); |
75eb6c372
|
1286 |
int may_queue; |
e80640213
|
1287 |
req_flags_t rq_flags = RQF_ALLOCED; |
88ee5ef15
|
1288 |
|
2fff8a924
|
1289 |
lockdep_assert_held(q->queue_lock); |
3f3299d5c
|
1290 |
if (unlikely(blk_queue_dying(q))) |
a492f0754
|
1291 |
return ERR_PTR(-ENODEV); |
da8303c63
|
1292 |
|
ef295ecf0
|
1293 |
may_queue = elv_may_queue(q, op); |
88ee5ef15
|
1294 1295 |
if (may_queue == ELV_MQUEUE_NO) goto rq_starved; |
1faa16d22
|
1296 1297 |
if (rl->count[is_sync]+1 >= queue_congestion_on_threshold(q)) { if (rl->count[is_sync]+1 >= q->nr_requests) { |
f2dbd76a0
|
1298 |
/* |
88ee5ef15
|
1299 1300 1301 1302 1303 |
* The queue will fill after this allocation, so set * it as full, and mark this process as "batching". * This process will be allowed to complete a batch of * requests, others will be blocked. */ |
5b788ce3e
|
1304 |
if (!blk_rl_full(rl, is_sync)) { |
88ee5ef15
|
1305 |
ioc_set_batching(q, ioc); |
5b788ce3e
|
1306 |
blk_set_rl_full(rl, is_sync); |
88ee5ef15
|
1307 1308 1309 1310 1311 1312 1313 1314 |
} else { if (may_queue != ELV_MQUEUE_MUST && !ioc_batching(q, ioc)) { /* * The queue is full and the allocating * process is not a "batcher", and not * exempted by the IO scheduler */ |
a492f0754
|
1315 |
return ERR_PTR(-ENOMEM); |
88ee5ef15
|
1316 1317 |
} } |
1da177e4c
|
1318 |
} |
d40f75a06
|
1319 |
blk_set_congested(rl, is_sync); |
1da177e4c
|
1320 |
} |
082cf69eb
|
1321 1322 1323 1324 1325 |
/* * Only allow batching queuers to allocate up to 50% over the defined * limit of requests, otherwise we could have thousands of requests * allocated with any setting of ->nr_requests */ |
1faa16d22
|
1326 |
if (rl->count[is_sync] >= (3 * q->nr_requests / 2)) |
a492f0754
|
1327 |
return ERR_PTR(-ENOMEM); |
fd782a4a9
|
1328 |
|
8a5ecdd42
|
1329 |
q->nr_rqs[is_sync]++; |
1faa16d22
|
1330 1331 |
rl->count[is_sync]++; rl->starved[is_sync] = 0; |
cb98fc8bb
|
1332 |
|
f1f8cc946
|
1333 1334 |
/* * Decide whether the new request will be managed by elevator. If |
e80640213
|
1335 |
* so, mark @rq_flags and increment elvpriv. Non-zero elvpriv will |
f1f8cc946
|
1336 1337 1338 1339 |
* prevent the current elevator from being destroyed until the new * request is freed. This guarantees icq's won't be destroyed and * makes creating new ones safe. * |
e6f7f93d5
|
1340 1341 1342 |
* Flush requests do not use the elevator so skip initialization. * This allows a request to share the flush and elevator data. * |
f1f8cc946
|
1343 1344 1345 |
* Also, lookup icq while holding queue_lock. If it doesn't exist, * it will be created after releasing queue_lock. */ |
e6f7f93d5
|
1346 |
if (!op_is_flush(op) && !blk_queue_bypass(q)) { |
e80640213
|
1347 |
rq_flags |= RQF_ELVPRIV; |
8a5ecdd42
|
1348 |
q->nr_rqs_elvpriv++; |
f1f8cc946
|
1349 1350 |
if (et->icq_cache && ioc) icq = ioc_lookup_icq(ioc, q); |
9d5a4e946
|
1351 |
} |
cb98fc8bb
|
1352 |
|
f253b86b4
|
1353 |
if (blk_queue_io_stat(q)) |
e80640213
|
1354 |
rq_flags |= RQF_IO_STAT; |
1da177e4c
|
1355 |
spin_unlock_irq(q->queue_lock); |
29e2b09ab
|
1356 |
/* allocate and init request */ |
5b788ce3e
|
1357 |
rq = mempool_alloc(rl->rq_pool, gfp_mask); |
29e2b09ab
|
1358 |
if (!rq) |
b679281a6
|
1359 |
goto fail_alloc; |
1da177e4c
|
1360 |
|
29e2b09ab
|
1361 |
blk_rq_init(q, rq); |
a051661ca
|
1362 |
blk_rq_set_rl(rq, rl); |
ef295ecf0
|
1363 |
rq->cmd_flags = op; |
e80640213
|
1364 |
rq->rq_flags = rq_flags; |
1b6d65a0b
|
1365 1366 |
if (flags & BLK_MQ_REQ_PREEMPT) rq->rq_flags |= RQF_PREEMPT; |
29e2b09ab
|
1367 |
|
aaf7c6806
|
1368 |
/* init elvpriv */ |
e80640213
|
1369 |
if (rq_flags & RQF_ELVPRIV) { |
aaf7c6806
|
1370 |
if (unlikely(et->icq_cache && !icq)) { |
7f4b35d15
|
1371 1372 |
if (ioc) icq = ioc_create_icq(ioc, q, gfp_mask); |
aaf7c6806
|
1373 1374 |
if (!icq) goto fail_elvpriv; |
29e2b09ab
|
1375 |
} |
aaf7c6806
|
1376 1377 1378 1379 1380 1381 |
rq->elv.icq = icq; if (unlikely(elv_set_request(q, rq, bio, gfp_mask))) goto fail_elvpriv; /* @rq->elv.icq holds io_context until @rq is freed */ |
29e2b09ab
|
1382 1383 1384 |
if (icq) get_io_context(icq->ioc); } |
aaf7c6806
|
1385 |
out: |
88ee5ef15
|
1386 1387 1388 1389 1390 1391 |
/* * ioc may be NULL here, and ioc_batching will be false. That's * OK, if the queue is under the request limit then requests need * not count toward the nr_batch_requests limit. There will always * be some limit enforced by BLK_BATCH_TIME. */ |
1da177e4c
|
1392 1393 |
if (ioc_batching(q, ioc)) ioc->nr_batch_requests--; |
6728cb0e6
|
1394 |
|
e6a40b096
|
1395 |
trace_block_getrq(q, bio, op); |
1da177e4c
|
1396 |
return rq; |
b679281a6
|
1397 |
|
aaf7c6806
|
1398 1399 1400 1401 1402 1403 1404 |
fail_elvpriv: /* * elvpriv init failed. ioc, icq and elvpriv aren't mempool backed * and may fail indefinitely under memory pressure and thus * shouldn't stall IO. Treat this request as !elvpriv. This will * disturb iosched and blkcg but weird is bettern than dead. */ |
7b2b10e0e
|
1405 1406 |
printk_ratelimited(KERN_WARNING "%s: dev %s: request aux data allocation failed, iosched may be disturbed ", |
dc3b17cc8
|
1407 |
__func__, dev_name(q->backing_dev_info->dev)); |
aaf7c6806
|
1408 |
|
e80640213
|
1409 |
rq->rq_flags &= ~RQF_ELVPRIV; |
aaf7c6806
|
1410 1411 1412 |
rq->elv.icq = NULL; spin_lock_irq(q->queue_lock); |
8a5ecdd42
|
1413 |
q->nr_rqs_elvpriv--; |
aaf7c6806
|
1414 1415 |
spin_unlock_irq(q->queue_lock); goto out; |
b679281a6
|
1416 1417 1418 1419 1420 1421 1422 1423 1424 |
fail_alloc: /* * Allocation failed presumably due to memory. Undo anything we * might have messed up. * * Allocating task should really be put onto the front of the wait * queue, but this is pretty rare. */ spin_lock_irq(q->queue_lock); |
e80640213
|
1425 |
freed_request(rl, is_sync, rq_flags); |
b679281a6
|
1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 |
/* * in the very unlikely event that allocation failed and no * requests for this direction was pending, mark us starved so that * freeing of a request in the other direction will notice * us. another possible fix would be to split the rq mempool into * READ and WRITE */ rq_starved: if (unlikely(rl->count[is_sync] == 0)) rl->starved[is_sync] = 1; |
a492f0754
|
1437 |
return ERR_PTR(-ENOMEM); |
1da177e4c
|
1438 |
} |
da8303c63
|
1439 |
/** |
a06e05e6a
|
1440 |
* get_request - get a free request |
da8303c63
|
1441 |
* @q: request_queue to allocate request from |
ef295ecf0
|
1442 |
* @op: operation and flags |
da8303c63
|
1443 |
* @bio: bio to allocate request for (can be %NULL) |
6a15674d1
|
1444 |
* @flags: BLK_MQ_REQ_* flags. |
4accf5fc7
|
1445 |
* @gfp: allocator flags |
da8303c63
|
1446 |
* |
a9a14d367
|
1447 |
* Get a free request from @q. If %BLK_MQ_REQ_NOWAIT is set in @flags, |
d0164adc8
|
1448 |
* this function keeps retrying under memory pressure and fails iff @q is dead. |
d6344532a
|
1449 |
* |
da3dae54e
|
1450 |
* Must be called with @q->queue_lock held and, |
a492f0754
|
1451 1452 |
* Returns ERR_PTR on failure, with @q->queue_lock held. * Returns request pointer on success, with @q->queue_lock *not held*. |
1da177e4c
|
1453 |
*/ |
ef295ecf0
|
1454 |
static struct request *get_request(struct request_queue *q, unsigned int op, |
4accf5fc7
|
1455 |
struct bio *bio, blk_mq_req_flags_t flags, gfp_t gfp) |
1da177e4c
|
1456 |
{ |
ef295ecf0
|
1457 |
const bool is_sync = op_is_sync(op); |
a06e05e6a
|
1458 |
DEFINE_WAIT(wait); |
a051661ca
|
1459 |
struct request_list *rl; |
1da177e4c
|
1460 |
struct request *rq; |
a051661ca
|
1461 |
|
2fff8a924
|
1462 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
1463 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
1464 |
|
a051661ca
|
1465 |
rl = blk_get_rl(q, bio); /* transferred to @rq on success */ |
a06e05e6a
|
1466 |
retry: |
4accf5fc7
|
1467 |
rq = __get_request(rl, op, bio, flags, gfp); |
a492f0754
|
1468 |
if (!IS_ERR(rq)) |
a06e05e6a
|
1469 |
return rq; |
1da177e4c
|
1470 |
|
03a07c92a
|
1471 1472 1473 1474 |
if (op & REQ_NOWAIT) { blk_put_rl(rl); return ERR_PTR(-EAGAIN); } |
6a15674d1
|
1475 |
if ((flags & BLK_MQ_REQ_NOWAIT) || unlikely(blk_queue_dying(q))) { |
a051661ca
|
1476 |
blk_put_rl(rl); |
a492f0754
|
1477 |
return rq; |
a051661ca
|
1478 |
} |
1da177e4c
|
1479 |
|
a06e05e6a
|
1480 1481 1482 |
/* wait on @rl and retry */ prepare_to_wait_exclusive(&rl->wait[is_sync], &wait, TASK_UNINTERRUPTIBLE); |
1da177e4c
|
1483 |
|
e6a40b096
|
1484 |
trace_block_sleeprq(q, bio, op); |
1da177e4c
|
1485 |
|
a06e05e6a
|
1486 1487 |
spin_unlock_irq(q->queue_lock); io_schedule(); |
d6344532a
|
1488 |
|
a06e05e6a
|
1489 1490 1491 1492 1493 |
/* * After sleeping, we become a "batching" process and will be able * to allocate at least one request, and up to a big batch of them * for a small period time. See ioc_batching, ioc_set_batching */ |
a06e05e6a
|
1494 |
ioc_set_batching(q, current->io_context); |
05caf8dbc
|
1495 |
|
a06e05e6a
|
1496 1497 |
spin_lock_irq(q->queue_lock); finish_wait(&rl->wait[is_sync], &wait); |
1da177e4c
|
1498 |
|
a06e05e6a
|
1499 |
goto retry; |
1da177e4c
|
1500 |
} |
6a15674d1
|
1501 |
/* flags: BLK_MQ_REQ_PREEMPT and/or BLK_MQ_REQ_NOWAIT. */ |
cd6ce1482
|
1502 |
static struct request *blk_old_get_request(struct request_queue *q, |
9a95e4ef7
|
1503 |
unsigned int op, blk_mq_req_flags_t flags) |
1da177e4c
|
1504 1505 |
{ struct request *rq; |
c3036021c
|
1506 |
gfp_t gfp_mask = flags & BLK_MQ_REQ_NOWAIT ? GFP_ATOMIC : GFP_NOIO; |
055f6e18e
|
1507 |
int ret = 0; |
1da177e4c
|
1508 |
|
332ebbf7f
|
1509 |
WARN_ON_ONCE(q->mq_ops); |
7f4b35d15
|
1510 1511 |
/* create ioc upfront */ create_io_context(gfp_mask, q->node); |
3a0a52997
|
1512 |
ret = blk_queue_enter(q, flags); |
055f6e18e
|
1513 1514 |
if (ret) return ERR_PTR(ret); |
d6344532a
|
1515 |
spin_lock_irq(q->queue_lock); |
4accf5fc7
|
1516 |
rq = get_request(q, op, NULL, flags, gfp_mask); |
0c4de0f33
|
1517 |
if (IS_ERR(rq)) { |
da8303c63
|
1518 |
spin_unlock_irq(q->queue_lock); |
055f6e18e
|
1519 |
blk_queue_exit(q); |
0c4de0f33
|
1520 1521 |
return rq; } |
1da177e4c
|
1522 |
|
0c4de0f33
|
1523 1524 1525 1526 |
/* q->queue_lock is unlocked at this point */ rq->__data_len = 0; rq->__sector = (sector_t) -1; rq->bio = rq->biotail = NULL; |
1da177e4c
|
1527 1528 |
return rq; } |
320ae51fe
|
1529 |
|
6a15674d1
|
1530 |
/** |
ff005a066
|
1531 |
* blk_get_request - allocate a request |
6a15674d1
|
1532 1533 1534 1535 |
* @q: request queue to allocate a request for * @op: operation (REQ_OP_*) and REQ_* flags, e.g. REQ_SYNC. * @flags: BLK_MQ_REQ_* flags, e.g. BLK_MQ_REQ_NOWAIT. */ |
ff005a066
|
1536 1537 |
struct request *blk_get_request(struct request_queue *q, unsigned int op, blk_mq_req_flags_t flags) |
320ae51fe
|
1538 |
{ |
d280bab30
|
1539 |
struct request *req; |
6a15674d1
|
1540 |
WARN_ON_ONCE(op & REQ_NOWAIT); |
1b6d65a0b
|
1541 |
WARN_ON_ONCE(flags & ~(BLK_MQ_REQ_NOWAIT | BLK_MQ_REQ_PREEMPT)); |
6a15674d1
|
1542 |
|
d280bab30
|
1543 |
if (q->mq_ops) { |
6a15674d1
|
1544 |
req = blk_mq_alloc_request(q, op, flags); |
d280bab30
|
1545 1546 1547 |
if (!IS_ERR(req) && q->mq_ops->initialize_rq_fn) q->mq_ops->initialize_rq_fn(req); } else { |
6a15674d1
|
1548 |
req = blk_old_get_request(q, op, flags); |
d280bab30
|
1549 1550 1551 1552 1553 |
if (!IS_ERR(req) && q->initialize_rq_fn) q->initialize_rq_fn(req); } return req; |
320ae51fe
|
1554 |
} |
1da177e4c
|
1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 |
EXPORT_SYMBOL(blk_get_request); /** * blk_requeue_request - put a request back on queue * @q: request queue where request should be inserted * @rq: request to be inserted * * Description: * Drivers often keep queueing requests until the hardware cannot accept * more, when that condition happens we need to put the request back * on the queue. Must be called with queue lock held. */ |
165125e1e
|
1567 |
void blk_requeue_request(struct request_queue *q, struct request *rq) |
1da177e4c
|
1568 |
{ |
2fff8a924
|
1569 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
1570 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
1571 |
|
242f9dcb8
|
1572 1573 |
blk_delete_timer(rq); blk_clear_rq_complete(rq); |
5f3ea37c7
|
1574 |
trace_block_rq_requeue(q, rq); |
a79050434
|
1575 |
rq_qos_requeue(q, rq); |
2056a782f
|
1576 |
|
e80640213
|
1577 |
if (rq->rq_flags & RQF_QUEUED) |
1da177e4c
|
1578 |
blk_queue_end_tag(q, rq); |
ba396a6c1
|
1579 |
BUG_ON(blk_queued_rq(rq)); |
1da177e4c
|
1580 1581 |
elv_requeue_request(q, rq); } |
1da177e4c
|
1582 |
EXPORT_SYMBOL(blk_requeue_request); |
73c101011
|
1583 1584 1585 |
static void add_acct_request(struct request_queue *q, struct request *rq, int where) { |
320ae51fe
|
1586 |
blk_account_io_start(rq, true); |
7eaceacca
|
1587 |
__elv_add_request(q, rq, where); |
73c101011
|
1588 |
} |
d62e26b3f
|
1589 |
static void part_round_stats_single(struct request_queue *q, int cpu, |
b8d62b3a9
|
1590 1591 |
struct hd_struct *part, unsigned long now, unsigned int inflight) |
074a7aca7
|
1592 |
{ |
b8d62b3a9
|
1593 |
if (inflight) { |
074a7aca7
|
1594 |
__part_stat_add(cpu, part, time_in_queue, |
b8d62b3a9
|
1595 |
inflight * (now - part->stamp)); |
074a7aca7
|
1596 1597 1598 1599 1600 1601 |
__part_stat_add(cpu, part, io_ticks, (now - part->stamp)); } part->stamp = now; } /** |
496aa8a98
|
1602 |
* part_round_stats() - Round off the performance stats on a struct disk_stats. |
d62e26b3f
|
1603 |
* @q: target block queue |
496aa8a98
|
1604 1605 |
* @cpu: cpu number for stats access * @part: target partition |
1da177e4c
|
1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 |
* * The average IO queue length and utilisation statistics are maintained * by observing the current state of the queue length and the amount of * time it has been in this state for. * * Normally, that accounting is done on IO completion, but that can result * in more than a second's worth of IO being accounted for within any one * second, leading to >100% utilisation. To deal with that, we call this * function to do a round-off before returning the results when reading * /proc/diskstats. This accounts immediately for all queue usage up to * the current jiffies and restarts the counters again. */ |
d62e26b3f
|
1618 |
void part_round_stats(struct request_queue *q, int cpu, struct hd_struct *part) |
6f2576af5
|
1619 |
{ |
b8d62b3a9
|
1620 |
struct hd_struct *part2 = NULL; |
6f2576af5
|
1621 |
unsigned long now = jiffies; |
b8d62b3a9
|
1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 |
unsigned int inflight[2]; int stats = 0; if (part->stamp != now) stats |= 1; if (part->partno) { part2 = &part_to_disk(part)->part0; if (part2->stamp != now) stats |= 2; } if (!stats) return; part_in_flight(q, part, inflight); |
6f2576af5
|
1638 |
|
b8d62b3a9
|
1639 1640 1641 1642 |
if (stats & 2) part_round_stats_single(q, cpu, part2, now, inflight[1]); if (stats & 1) part_round_stats_single(q, cpu, part, now, inflight[0]); |
6f2576af5
|
1643 |
} |
074a7aca7
|
1644 |
EXPORT_SYMBOL_GPL(part_round_stats); |
6f2576af5
|
1645 |
|
47fafbc70
|
1646 |
#ifdef CONFIG_PM |
c8158819d
|
1647 1648 |
static void blk_pm_put_request(struct request *rq) { |
e80640213
|
1649 |
if (rq->q->dev && !(rq->rq_flags & RQF_PM) && !--rq->q->nr_pending) |
c8158819d
|
1650 1651 1652 1653 1654 |
pm_runtime_mark_last_busy(rq->q->dev); } #else static inline void blk_pm_put_request(struct request *rq) {} #endif |
165125e1e
|
1655 |
void __blk_put_request(struct request_queue *q, struct request *req) |
1da177e4c
|
1656 |
{ |
e80640213
|
1657 |
req_flags_t rq_flags = req->rq_flags; |
1da177e4c
|
1658 1659 |
if (unlikely(!q)) return; |
1da177e4c
|
1660 |
|
6f5ba581c
|
1661 1662 1663 1664 |
if (q->mq_ops) { blk_mq_free_request(req); return; } |
2fff8a924
|
1665 |
lockdep_assert_held(q->queue_lock); |
6cc77e9cb
|
1666 |
blk_req_zone_write_unlock(req); |
c8158819d
|
1667 |
blk_pm_put_request(req); |
8922e16cf
|
1668 |
elv_completed_request(q, req); |
1cd96c242
|
1669 1670 |
/* this is a bio leak */ WARN_ON(req->bio != NULL); |
a79050434
|
1671 |
rq_qos_done(q, req); |
87760e5ee
|
1672 |
|
1da177e4c
|
1673 1674 1675 1676 |
/* * Request may not have originated from ll_rw_blk. if not, * it didn't come out of our reserved rq pools */ |
e80640213
|
1677 |
if (rq_flags & RQF_ALLOCED) { |
a051661ca
|
1678 |
struct request_list *rl = blk_rq_rl(req); |
ef295ecf0
|
1679 |
bool sync = op_is_sync(req->cmd_flags); |
1da177e4c
|
1680 |
|
1da177e4c
|
1681 |
BUG_ON(!list_empty(&req->queuelist)); |
360f92c24
|
1682 |
BUG_ON(ELV_ON_HASH(req)); |
1da177e4c
|
1683 |
|
a051661ca
|
1684 |
blk_free_request(rl, req); |
e80640213
|
1685 |
freed_request(rl, sync, rq_flags); |
a051661ca
|
1686 |
blk_put_rl(rl); |
055f6e18e
|
1687 |
blk_queue_exit(q); |
1da177e4c
|
1688 1689 |
} } |
6e39b69e7
|
1690 |
EXPORT_SYMBOL_GPL(__blk_put_request); |
1da177e4c
|
1691 1692 |
void blk_put_request(struct request *req) { |
165125e1e
|
1693 |
struct request_queue *q = req->q; |
8922e16cf
|
1694 |
|
320ae51fe
|
1695 1696 1697 1698 1699 1700 1701 1702 1703 |
if (q->mq_ops) blk_mq_free_request(req); else { unsigned long flags; spin_lock_irqsave(q->queue_lock, flags); __blk_put_request(q, req); spin_unlock_irqrestore(q->queue_lock, flags); } |
1da177e4c
|
1704 |
} |
1da177e4c
|
1705 |
EXPORT_SYMBOL(blk_put_request); |
320ae51fe
|
1706 1707 |
bool bio_attempt_back_merge(struct request_queue *q, struct request *req, struct bio *bio) |
73c101011
|
1708 |
{ |
1eff9d322
|
1709 |
const int ff = bio->bi_opf & REQ_FAILFAST_MASK; |
73c101011
|
1710 |
|
73c101011
|
1711 1712 |
if (!ll_back_merge_fn(q, req, bio)) return false; |
8c1cf6bb0
|
1713 |
trace_block_bio_backmerge(q, req, bio); |
73c101011
|
1714 1715 1716 1717 1718 1719 |
if ((req->cmd_flags & REQ_FAILFAST_MASK) != ff) blk_rq_set_mixed_merge(req); req->biotail->bi_next = bio; req->biotail = bio; |
4f024f379
|
1720 |
req->__data_len += bio->bi_iter.bi_size; |
73c101011
|
1721 |
req->ioprio = ioprio_best(req->ioprio, bio_prio(bio)); |
320ae51fe
|
1722 |
blk_account_io_start(req, false); |
73c101011
|
1723 1724 |
return true; } |
320ae51fe
|
1725 1726 |
bool bio_attempt_front_merge(struct request_queue *q, struct request *req, struct bio *bio) |
73c101011
|
1727 |
{ |
1eff9d322
|
1728 |
const int ff = bio->bi_opf & REQ_FAILFAST_MASK; |
73c101011
|
1729 |
|
73c101011
|
1730 1731 |
if (!ll_front_merge_fn(q, req, bio)) return false; |
8c1cf6bb0
|
1732 |
trace_block_bio_frontmerge(q, req, bio); |
73c101011
|
1733 1734 1735 |
if ((req->cmd_flags & REQ_FAILFAST_MASK) != ff) blk_rq_set_mixed_merge(req); |
73c101011
|
1736 1737 |
bio->bi_next = req->bio; req->bio = bio; |
4f024f379
|
1738 1739 |
req->__sector = bio->bi_iter.bi_sector; req->__data_len += bio->bi_iter.bi_size; |
73c101011
|
1740 |
req->ioprio = ioprio_best(req->ioprio, bio_prio(bio)); |
320ae51fe
|
1741 |
blk_account_io_start(req, false); |
73c101011
|
1742 1743 |
return true; } |
1e739730c
|
1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 |
bool bio_attempt_discard_merge(struct request_queue *q, struct request *req, struct bio *bio) { unsigned short segments = blk_rq_nr_discard_segments(req); if (segments >= queue_max_discard_segments(q)) goto no_merge; if (blk_rq_sectors(req) + bio_sectors(bio) > blk_rq_get_max_sectors(req, blk_rq_pos(req))) goto no_merge; req->biotail->bi_next = bio; req->biotail = bio; req->__data_len += bio->bi_iter.bi_size; req->ioprio = ioprio_best(req->ioprio, bio_prio(bio)); req->nr_phys_segments = segments + 1; blk_account_io_start(req, false); return true; no_merge: req_set_nomerge(q, req); return false; } |
bd87b5898
|
1767 |
/** |
320ae51fe
|
1768 |
* blk_attempt_plug_merge - try to merge with %current's plugged list |
bd87b5898
|
1769 1770 1771 |
* @q: request_queue new bio is being queued at * @bio: new bio being queued * @request_count: out parameter for number of traversed plugged requests |
ccc2600b8
|
1772 1773 1774 |
* @same_queue_rq: pointer to &struct request that gets filled in when * another request associated with @q is found on the plug list * (optional, may be %NULL) |
bd87b5898
|
1775 1776 1777 1778 1779 |
* * Determine whether @bio being queued on @q can be merged with a request * on %current's plugged list. Returns %true if merge was successful, * otherwise %false. * |
07c2bd373
|
1780 1781 1782 1783 1784 1785 |
* Plugging coalesces IOs from the same issuer for the same purpose without * going through @q->queue_lock. As such it's more of an issuing mechanism * than scheduling, and the request, while may have elvpriv data, is not * added on the elevator at this point. In addition, we don't have * reliable access to the elevator outside queue lock. Only check basic * merging parameters without querying the elevator. |
da41a589f
|
1786 1787 |
* * Caller must ensure !blk_queue_nomerges(q) beforehand. |
73c101011
|
1788 |
*/ |
320ae51fe
|
1789 |
bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio, |
5b3f341f0
|
1790 1791 |
unsigned int *request_count, struct request **same_queue_rq) |
73c101011
|
1792 1793 1794 |
{ struct blk_plug *plug; struct request *rq; |
92f399c72
|
1795 |
struct list_head *plug_list; |
73c101011
|
1796 |
|
bd87b5898
|
1797 |
plug = current->plug; |
73c101011
|
1798 |
if (!plug) |
34fe7c054
|
1799 |
return false; |
56ebdaf2f
|
1800 |
*request_count = 0; |
73c101011
|
1801 |
|
92f399c72
|
1802 1803 1804 1805 1806 1807 |
if (q->mq_ops) plug_list = &plug->mq_list; else plug_list = &plug->list; list_for_each_entry_reverse(rq, plug_list, queuelist) { |
34fe7c054
|
1808 |
bool merged = false; |
73c101011
|
1809 |
|
5b3f341f0
|
1810 |
if (rq->q == q) { |
1b2e19f17
|
1811 |
(*request_count)++; |
5b3f341f0
|
1812 1813 1814 1815 1816 1817 1818 1819 |
/* * Only blk-mq multiple hardware queues case checks the * rq in the same queue, there should be only one such * rq in a queue **/ if (same_queue_rq) *same_queue_rq = rq; } |
56ebdaf2f
|
1820 |
|
07c2bd373
|
1821 |
if (rq->q != q || !blk_rq_merge_ok(rq, bio)) |
73c101011
|
1822 |
continue; |
34fe7c054
|
1823 1824 1825 1826 1827 1828 1829 |
switch (blk_try_merge(rq, bio)) { case ELEVATOR_BACK_MERGE: merged = bio_attempt_back_merge(q, rq, bio); break; case ELEVATOR_FRONT_MERGE: merged = bio_attempt_front_merge(q, rq, bio); break; |
1e739730c
|
1830 1831 1832 |
case ELEVATOR_DISCARD_MERGE: merged = bio_attempt_discard_merge(q, rq, bio); break; |
34fe7c054
|
1833 1834 |
default: break; |
73c101011
|
1835 |
} |
34fe7c054
|
1836 1837 1838 |
if (merged) return true; |
73c101011
|
1839 |
} |
34fe7c054
|
1840 1841 |
return false; |
73c101011
|
1842 |
} |
0809e3ac6
|
1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 |
unsigned int blk_plug_queued_count(struct request_queue *q) { struct blk_plug *plug; struct request *rq; struct list_head *plug_list; unsigned int ret = 0; plug = current->plug; if (!plug) goto out; if (q->mq_ops) plug_list = &plug->mq_list; else plug_list = &plug->list; list_for_each_entry(rq, plug_list, queuelist) { if (rq->q == q) ret++; } out: return ret; } |
da8d7f079
|
1866 |
void blk_init_request_from_bio(struct request *req, struct bio *bio) |
52d9e6753
|
1867 |
{ |
0be0dee64
|
1868 |
struct io_context *ioc = rq_ioc(bio); |
1eff9d322
|
1869 |
if (bio->bi_opf & REQ_RAHEAD) |
a82afdfcb
|
1870 |
req->cmd_flags |= REQ_FAILFAST_MASK; |
b31dc66a5
|
1871 |
|
4f024f379
|
1872 |
req->__sector = bio->bi_iter.bi_sector; |
5dc8b362a
|
1873 1874 |
if (ioprio_valid(bio_prio(bio))) req->ioprio = bio_prio(bio); |
0be0dee64
|
1875 1876 1877 1878 |
else if (ioc) req->ioprio = ioc->ioprio; else req->ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_NONE, 0); |
cb6934f8e
|
1879 |
req->write_hint = bio->bi_write_hint; |
bc1c56fde
|
1880 |
blk_rq_bio_prep(req->q, req, bio); |
52d9e6753
|
1881 |
} |
da8d7f079
|
1882 |
EXPORT_SYMBOL_GPL(blk_init_request_from_bio); |
52d9e6753
|
1883 |
|
dece16353
|
1884 |
static blk_qc_t blk_queue_bio(struct request_queue *q, struct bio *bio) |
1da177e4c
|
1885 |
{ |
73c101011
|
1886 |
struct blk_plug *plug; |
34fe7c054
|
1887 |
int where = ELEVATOR_INSERT_SORT; |
e4d750c97
|
1888 |
struct request *req, *free; |
56ebdaf2f
|
1889 |
unsigned int request_count = 0; |
1da177e4c
|
1890 |
|
1da177e4c
|
1891 1892 1893 1894 1895 1896 |
/* * low level driver can indicate that it wants pages above a * certain limit bounced to low memory (ie for highmem, or even * ISA dma in theory) */ blk_queue_bounce(q, &bio); |
af67c31fb
|
1897 |
blk_queue_split(q, &bio); |
23688bf4f
|
1898 |
|
e23947bd7
|
1899 |
if (!bio_integrity_prep(bio)) |
dece16353
|
1900 |
return BLK_QC_T_NONE; |
ffecfd1a7
|
1901 |
|
f73f44eb0
|
1902 |
if (op_is_flush(bio->bi_opf)) { |
73c101011
|
1903 |
spin_lock_irq(q->queue_lock); |
ae1b15396
|
1904 |
where = ELEVATOR_INSERT_FLUSH; |
28e7d1845
|
1905 1906 |
goto get_rq; } |
73c101011
|
1907 1908 1909 1910 |
/* * Check if we can merge with the plugged list before grabbing * any locks. */ |
0809e3ac6
|
1911 1912 |
if (!blk_queue_nomerges(q)) { if (blk_attempt_plug_merge(q, bio, &request_count, NULL)) |
dece16353
|
1913 |
return BLK_QC_T_NONE; |
0809e3ac6
|
1914 1915 |
} else request_count = blk_plug_queued_count(q); |
1da177e4c
|
1916 |
|
73c101011
|
1917 |
spin_lock_irq(q->queue_lock); |
2056a782f
|
1918 |
|
34fe7c054
|
1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 |
switch (elv_merge(q, &req, bio)) { case ELEVATOR_BACK_MERGE: if (!bio_attempt_back_merge(q, req, bio)) break; elv_bio_merged(q, req, bio); free = attempt_back_merge(q, req); if (free) __blk_put_request(q, free); else elv_merged_request(q, req, ELEVATOR_BACK_MERGE); goto out_unlock; case ELEVATOR_FRONT_MERGE: if (!bio_attempt_front_merge(q, req, bio)) break; elv_bio_merged(q, req, bio); free = attempt_front_merge(q, req); if (free) __blk_put_request(q, free); else elv_merged_request(q, req, ELEVATOR_FRONT_MERGE); goto out_unlock; default: break; |
1da177e4c
|
1942 |
} |
450991bc1
|
1943 |
get_rq: |
c1c80384c
|
1944 |
rq_qos_throttle(q, bio, q->queue_lock); |
87760e5ee
|
1945 |
|
1da177e4c
|
1946 |
/* |
450991bc1
|
1947 |
* Grab a free request. This is might sleep but can not fail. |
d6344532a
|
1948 |
* Returns with the queue unlocked. |
450991bc1
|
1949 |
*/ |
055f6e18e
|
1950 |
blk_queue_enter_live(q); |
c3036021c
|
1951 |
req = get_request(q, bio->bi_opf, bio, 0, GFP_NOIO); |
a492f0754
|
1952 |
if (IS_ERR(req)) { |
055f6e18e
|
1953 |
blk_queue_exit(q); |
c1c80384c
|
1954 |
rq_qos_cleanup(q, bio); |
4e4cbee93
|
1955 1956 1957 1958 |
if (PTR_ERR(req) == -ENOMEM) bio->bi_status = BLK_STS_RESOURCE; else bio->bi_status = BLK_STS_IOERR; |
4246a0b63
|
1959 |
bio_endio(bio); |
da8303c63
|
1960 1961 |
goto out_unlock; } |
d6344532a
|
1962 |
|
c1c80384c
|
1963 |
rq_qos_track(q, req, bio); |
87760e5ee
|
1964 |
|
450991bc1
|
1965 1966 1967 1968 1969 |
/* * After dropping the lock and possibly sleeping here, our request * may now be mergeable after it had proven unmergeable (above). * We don't worry about that case for efficiency. It won't happen * often, and the elevators are able to handle it. |
1da177e4c
|
1970 |
*/ |
da8d7f079
|
1971 |
blk_init_request_from_bio(req, bio); |
1da177e4c
|
1972 |
|
9562ad9ab
|
1973 |
if (test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags)) |
11ccf116d
|
1974 |
req->cpu = raw_smp_processor_id(); |
73c101011
|
1975 1976 |
plug = current->plug; |
721a9602e
|
1977 |
if (plug) { |
dc6d36c97
|
1978 1979 |
/* * If this is the first request added after a plug, fire |
7aef2e780
|
1980 |
* of a plug trace. |
0a6219a95
|
1981 1982 1983 |
* * @request_count may become stale because of schedule * out, so check plug list again. |
dc6d36c97
|
1984 |
*/ |
0a6219a95
|
1985 |
if (!request_count || list_empty(&plug->list)) |
dc6d36c97
|
1986 |
trace_block_plug(q); |
3540d5e89
|
1987 |
else { |
50d24c344
|
1988 1989 1990 |
struct request *last = list_entry_rq(plug->list.prev); if (request_count >= BLK_MAX_REQUEST_COUNT || blk_rq_bytes(last) >= BLK_PLUG_FLUSH_SIZE) { |
3540d5e89
|
1991 |
blk_flush_plug_list(plug, false); |
019ceb7d5
|
1992 1993 |
trace_block_plug(q); } |
73c101011
|
1994 |
} |
73c101011
|
1995 |
list_add_tail(&req->queuelist, &plug->list); |
320ae51fe
|
1996 |
blk_account_io_start(req, true); |
73c101011
|
1997 1998 1999 |
} else { spin_lock_irq(q->queue_lock); add_acct_request(q, req, where); |
24ecfbe27
|
2000 |
__blk_run_queue(q); |
73c101011
|
2001 2002 2003 |
out_unlock: spin_unlock_irq(q->queue_lock); } |
dece16353
|
2004 2005 |
return BLK_QC_T_NONE; |
1da177e4c
|
2006 |
} |
52c5e62d4
|
2007 |
static void handle_bad_sector(struct bio *bio, sector_t maxsector) |
1da177e4c
|
2008 2009 2010 2011 2012 |
{ char b[BDEVNAME_SIZE]; printk(KERN_INFO "attempt to access beyond end of device "); |
6296b9604
|
2013 2014 |
printk(KERN_INFO "%s: rw=%d, want=%Lu, limit=%Lu ", |
74d46992e
|
2015 |
bio_devname(bio, b), bio->bi_opf, |
f73a1c7d1
|
2016 |
(unsigned long long)bio_end_sector(bio), |
52c5e62d4
|
2017 |
(long long)maxsector); |
1da177e4c
|
2018 |
} |
c17bb4951
|
2019 2020 2021 2022 2023 2024 2025 2026 2027 |
#ifdef CONFIG_FAIL_MAKE_REQUEST static DECLARE_FAULT_ATTR(fail_make_request); static int __init setup_fail_make_request(char *str) { return setup_fault_attr(&fail_make_request, str); } __setup("fail_make_request=", setup_fail_make_request); |
b2c9cd379
|
2028 |
static bool should_fail_request(struct hd_struct *part, unsigned int bytes) |
c17bb4951
|
2029 |
{ |
b2c9cd379
|
2030 |
return part->make_it_fail && should_fail(&fail_make_request, bytes); |
c17bb4951
|
2031 2032 2033 2034 |
} static int __init fail_make_request_debugfs(void) { |
dd48c085c
|
2035 2036 |
struct dentry *dir = fault_create_debugfs_attr("fail_make_request", NULL, &fail_make_request); |
21f9fcd81
|
2037 |
return PTR_ERR_OR_ZERO(dir); |
c17bb4951
|
2038 2039 2040 2041 2042 |
} late_initcall(fail_make_request_debugfs); #else /* CONFIG_FAIL_MAKE_REQUEST */ |
b2c9cd379
|
2043 2044 |
static inline bool should_fail_request(struct hd_struct *part, unsigned int bytes) |
c17bb4951
|
2045 |
{ |
b2c9cd379
|
2046 |
return false; |
c17bb4951
|
2047 2048 2049 |
} #endif /* CONFIG_FAIL_MAKE_REQUEST */ |
721c7fc70
|
2050 2051 |
static inline bool bio_check_ro(struct bio *bio, struct hd_struct *part) { |
b089cfd95
|
2052 |
const int op = bio_op(bio); |
8b2ded1c9
|
2053 |
if (part->policy && op_is_write(op)) { |
721c7fc70
|
2054 |
char b[BDEVNAME_SIZE]; |
8b2ded1c9
|
2055 2056 |
if (op_is_flush(bio->bi_opf) && !bio_sectors(bio)) return false; |
a32e236eb
|
2057 |
WARN_ONCE(1, |
721c7fc70
|
2058 2059 2060 2061 |
"generic_make_request: Trying to write " "to read-only block-device %s (partno %d) ", bio_devname(bio, b), part->partno); |
a32e236eb
|
2062 2063 |
/* Older lvm-tools actually trigger this */ return false; |
721c7fc70
|
2064 2065 2066 2067 |
} return false; } |
30abb3a67
|
2068 2069 2070 2071 2072 2073 2074 |
static noinline int should_fail_bio(struct bio *bio) { if (should_fail_request(&bio->bi_disk->part0, bio->bi_iter.bi_size)) return -EIO; return 0; } ALLOW_ERROR_INJECTION(should_fail_bio, ERRNO); |
c07e2b412
|
2075 |
/* |
52c5e62d4
|
2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 |
* Check whether this bio extends beyond the end of the device or partition. * This may well happen - the kernel calls bread() without checking the size of * the device, e.g., when mounting a file system. */ static inline int bio_check_eod(struct bio *bio, sector_t maxsector) { unsigned int nr_sectors = bio_sectors(bio); if (nr_sectors && maxsector && (nr_sectors > maxsector || bio->bi_iter.bi_sector > maxsector - nr_sectors)) { handle_bad_sector(bio, maxsector); return -EIO; } return 0; } /* |
74d46992e
|
2094 2095 2096 2097 2098 |
* Remap block n of partition p to block n+start(p) of the disk. */ static inline int blk_partition_remap(struct bio *bio) { struct hd_struct *p; |
52c5e62d4
|
2099 |
int ret = -EIO; |
74d46992e
|
2100 |
|
721c7fc70
|
2101 2102 |
rcu_read_lock(); p = __disk_get_part(bio->bi_disk, bio->bi_partno); |
52c5e62d4
|
2103 2104 2105 2106 2107 |
if (unlikely(!p)) goto out; if (unlikely(should_fail_request(p, bio->bi_iter.bi_size))) goto out; if (unlikely(bio_check_ro(bio, p))) |
721c7fc70
|
2108 |
goto out; |
721c7fc70
|
2109 |
|
74d46992e
|
2110 2111 2112 2113 |
/* * Zone reset does not include bi_size so bio_sectors() is always 0. * Include a test for the reset op code and perform the remap if needed. */ |
52c5e62d4
|
2114 2115 2116 2117 |
if (bio_sectors(bio) || bio_op(bio) == REQ_OP_ZONE_RESET) { if (bio_check_eod(bio, part_nr_sects_read(p))) goto out; bio->bi_iter.bi_sector += p->start_sect; |
52c5e62d4
|
2118 2119 2120 |
trace_block_bio_remap(bio->bi_disk->queue, bio, part_devt(p), bio->bi_iter.bi_sector - p->start_sect); } |
c04fa44b7
|
2121 |
bio->bi_partno = 0; |
52c5e62d4
|
2122 |
ret = 0; |
721c7fc70
|
2123 2124 |
out: rcu_read_unlock(); |
74d46992e
|
2125 2126 |
return ret; } |
27a84d54c
|
2127 2128 |
static noinline_for_stack bool generic_make_request_checks(struct bio *bio) |
1da177e4c
|
2129 |
{ |
165125e1e
|
2130 |
struct request_queue *q; |
5a7bbad27
|
2131 |
int nr_sectors = bio_sectors(bio); |
4e4cbee93
|
2132 |
blk_status_t status = BLK_STS_IOERR; |
5a7bbad27
|
2133 |
char b[BDEVNAME_SIZE]; |
1da177e4c
|
2134 2135 |
might_sleep(); |
1da177e4c
|
2136 |
|
74d46992e
|
2137 |
q = bio->bi_disk->queue; |
5a7bbad27
|
2138 2139 2140 2141 2142 |
if (unlikely(!q)) { printk(KERN_ERR "generic_make_request: Trying to access " "nonexistent block-device %s (%Lu) ", |
74d46992e
|
2143 |
bio_devname(bio, b), (long long)bio->bi_iter.bi_sector); |
5a7bbad27
|
2144 2145 |
goto end_io; } |
c17bb4951
|
2146 |
|
03a07c92a
|
2147 2148 2149 2150 |
/* * For a REQ_NOWAIT based request, return -EOPNOTSUPP * if queue is not a request based queue. */ |
03a07c92a
|
2151 2152 |
if ((bio->bi_opf & REQ_NOWAIT) && !queue_is_rq_based(q)) goto not_supported; |
30abb3a67
|
2153 |
if (should_fail_bio(bio)) |
5a7bbad27
|
2154 |
goto end_io; |
2056a782f
|
2155 |
|
52c5e62d4
|
2156 2157 |
if (bio->bi_partno) { if (unlikely(blk_partition_remap(bio))) |
721c7fc70
|
2158 2159 |
goto end_io; } else { |
52c5e62d4
|
2160 2161 2162 |
if (unlikely(bio_check_ro(bio, &bio->bi_disk->part0))) goto end_io; if (unlikely(bio_check_eod(bio, get_capacity(bio->bi_disk)))) |
721c7fc70
|
2163 2164 |
goto end_io; } |
2056a782f
|
2165 |
|
5a7bbad27
|
2166 2167 2168 2169 2170 |
/* * Filter flush bio's early so that make_request based * drivers without flush support don't have to worry * about them. */ |
f3a8ab7d5
|
2171 |
if (op_is_flush(bio->bi_opf) && |
c888a8f95
|
2172 |
!test_bit(QUEUE_FLAG_WC, &q->queue_flags)) { |
1eff9d322
|
2173 |
bio->bi_opf &= ~(REQ_PREFLUSH | REQ_FUA); |
5a7bbad27
|
2174 |
if (!nr_sectors) { |
4e4cbee93
|
2175 |
status = BLK_STS_OK; |
51fd77bd9
|
2176 2177 |
goto end_io; } |
5a7bbad27
|
2178 |
} |
5ddfe9691
|
2179 |
|
288dab8a3
|
2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 |
switch (bio_op(bio)) { case REQ_OP_DISCARD: if (!blk_queue_discard(q)) goto not_supported; break; case REQ_OP_SECURE_ERASE: if (!blk_queue_secure_erase(q)) goto not_supported; break; case REQ_OP_WRITE_SAME: |
74d46992e
|
2190 |
if (!q->limits.max_write_same_sectors) |
288dab8a3
|
2191 |
goto not_supported; |
58886785d
|
2192 |
break; |
2d253440b
|
2193 2194 |
case REQ_OP_ZONE_REPORT: case REQ_OP_ZONE_RESET: |
74d46992e
|
2195 |
if (!blk_queue_is_zoned(q)) |
2d253440b
|
2196 |
goto not_supported; |
288dab8a3
|
2197 |
break; |
a6f0788ec
|
2198 |
case REQ_OP_WRITE_ZEROES: |
74d46992e
|
2199 |
if (!q->limits.max_write_zeroes_sectors) |
a6f0788ec
|
2200 2201 |
goto not_supported; break; |
288dab8a3
|
2202 2203 |
default: break; |
5a7bbad27
|
2204 |
} |
01edede41
|
2205 |
|
7f4b35d15
|
2206 2207 2208 2209 2210 2211 2212 |
/* * Various block parts want %current->io_context and lazy ioc * allocation ends up trading a lot of pain for a small amount of * memory. Just allocate it upfront. This may fail and block * layer knows how to live with it. */ create_io_context(GFP_ATOMIC, q->node); |
ae1188963
|
2213 2214 |
if (!blkcg_bio_issue_check(q, bio)) return false; |
27a84d54c
|
2215 |
|
fbbaf700e
|
2216 2217 2218 2219 2220 2221 2222 |
if (!bio_flagged(bio, BIO_TRACE_COMPLETION)) { trace_block_bio_queue(q, bio); /* Now that enqueuing has been traced, we need to trace * completion as well. */ bio_set_flag(bio, BIO_TRACE_COMPLETION); } |
27a84d54c
|
2223 |
return true; |
a7384677b
|
2224 |
|
288dab8a3
|
2225 |
not_supported: |
4e4cbee93
|
2226 |
status = BLK_STS_NOTSUPP; |
a7384677b
|
2227 |
end_io: |
4e4cbee93
|
2228 |
bio->bi_status = status; |
4246a0b63
|
2229 |
bio_endio(bio); |
27a84d54c
|
2230 |
return false; |
1da177e4c
|
2231 |
} |
27a84d54c
|
2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 |
/** * generic_make_request - hand a buffer to its device driver for I/O * @bio: The bio describing the location in memory and on the device. * * generic_make_request() is used to make I/O requests of block * devices. It is passed a &struct bio, which describes the I/O that needs * to be done. * * generic_make_request() does not return any status. The * success/failure status of the request, along with notification of * completion, is delivered asynchronously through the bio->bi_end_io * function described (one day) else where. * * The caller of generic_make_request must make sure that bi_io_vec * are set to describe the memory buffer, and that bi_dev and bi_sector are * set to describe the device address, and the * bi_end_io and optionally bi_private are set to describe how * completion notification should be signaled. * * generic_make_request and the drivers it calls may use bi_next if this * bio happens to be merged with someone else, and may resubmit the bio to * a lower device by calling into generic_make_request recursively, which * means the bio should NOT be touched after the call to ->make_request_fn. |
d89d87965
|
2255 |
*/ |
dece16353
|
2256 |
blk_qc_t generic_make_request(struct bio *bio) |
d89d87965
|
2257 |
{ |
f5fe1b519
|
2258 2259 2260 2261 2262 2263 2264 2265 |
/* * bio_list_on_stack[0] contains bios submitted by the current * make_request_fn. * bio_list_on_stack[1] contains bios that were submitted before * the current make_request_fn, but that haven't been processed * yet. */ struct bio_list bio_list_on_stack[2]; |
37f9579f4
|
2266 2267 |
blk_mq_req_flags_t flags = 0; struct request_queue *q = bio->bi_disk->queue; |
dece16353
|
2268 |
blk_qc_t ret = BLK_QC_T_NONE; |
bddd87c7e
|
2269 |
|
37f9579f4
|
2270 2271 |
if (bio->bi_opf & REQ_NOWAIT) flags = BLK_MQ_REQ_NOWAIT; |
cd4a4ae46
|
2272 2273 2274 |
if (bio_flagged(bio, BIO_QUEUE_ENTERED)) blk_queue_enter_live(q); else if (blk_queue_enter(q, flags) < 0) { |
37f9579f4
|
2275 2276 2277 2278 2279 2280 |
if (!blk_queue_dying(q) && (bio->bi_opf & REQ_NOWAIT)) bio_wouldblock_error(bio); else bio_io_error(bio); return ret; } |
27a84d54c
|
2281 |
if (!generic_make_request_checks(bio)) |
dece16353
|
2282 |
goto out; |
27a84d54c
|
2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 |
/* * We only want one ->make_request_fn to be active at a time, else * stack usage with stacked devices could be a problem. So use * current->bio_list to keep a list of requests submited by a * make_request_fn function. current->bio_list is also used as a * flag to say if generic_make_request is currently active in this * task or not. If it is NULL, then no make_request is active. If * it is non-NULL, then a make_request is active, and new requests * should be added at the tail */ |
bddd87c7e
|
2294 |
if (current->bio_list) { |
f5fe1b519
|
2295 |
bio_list_add(¤t->bio_list[0], bio); |
dece16353
|
2296 |
goto out; |
d89d87965
|
2297 |
} |
27a84d54c
|
2298 |
|
d89d87965
|
2299 2300 2301 2302 2303 |
/* following loop may be a bit non-obvious, and so deserves some * explanation. * Before entering the loop, bio->bi_next is NULL (as all callers * ensure that) so we have a list with a single bio. * We pretend that we have just taken it off a longer list, so |
bddd87c7e
|
2304 2305 |
* we assign bio_list to a pointer to the bio_list_on_stack, * thus initialising the bio_list of new bios to be |
27a84d54c
|
2306 |
* added. ->make_request() may indeed add some more bios |
d89d87965
|
2307 2308 2309 |
* through a recursive call to generic_make_request. If it * did, we find a non-NULL value in bio_list and re-enter the loop * from the top. In this case we really did just take the bio |
bddd87c7e
|
2310 |
* of the top of the list (no pretending) and so remove it from |
27a84d54c
|
2311 |
* bio_list, and call into ->make_request() again. |
d89d87965
|
2312 2313 |
*/ BUG_ON(bio->bi_next); |
f5fe1b519
|
2314 2315 |
bio_list_init(&bio_list_on_stack[0]); current->bio_list = bio_list_on_stack; |
d89d87965
|
2316 |
do { |
37f9579f4
|
2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 |
bool enter_succeeded = true; if (unlikely(q != bio->bi_disk->queue)) { if (q) blk_queue_exit(q); q = bio->bi_disk->queue; flags = 0; if (bio->bi_opf & REQ_NOWAIT) flags = BLK_MQ_REQ_NOWAIT; if (blk_queue_enter(q, flags) < 0) { enter_succeeded = false; q = NULL; } } |
27a84d54c
|
2331 |
|
37f9579f4
|
2332 |
if (enter_succeeded) { |
79bd99596
|
2333 2334 2335 |
struct bio_list lower, same; /* Create a fresh bio_list for all subordinate requests */ |
f5fe1b519
|
2336 2337 |
bio_list_on_stack[1] = bio_list_on_stack[0]; bio_list_init(&bio_list_on_stack[0]); |
dece16353
|
2338 |
ret = q->make_request_fn(q, bio); |
3ef28e83a
|
2339 |
|
79bd99596
|
2340 2341 2342 2343 2344 |
/* sort new bios into those for a lower level * and those for the same level */ bio_list_init(&lower); bio_list_init(&same); |
f5fe1b519
|
2345 |
while ((bio = bio_list_pop(&bio_list_on_stack[0])) != NULL) |
74d46992e
|
2346 |
if (q == bio->bi_disk->queue) |
79bd99596
|
2347 2348 2349 2350 |
bio_list_add(&same, bio); else bio_list_add(&lower, bio); /* now assemble so we handle the lowest level first */ |
f5fe1b519
|
2351 2352 2353 |
bio_list_merge(&bio_list_on_stack[0], &lower); bio_list_merge(&bio_list_on_stack[0], &same); bio_list_merge(&bio_list_on_stack[0], &bio_list_on_stack[1]); |
3ef28e83a
|
2354 |
} else { |
03a07c92a
|
2355 2356 2357 2358 2359 |
if (unlikely(!blk_queue_dying(q) && (bio->bi_opf & REQ_NOWAIT))) bio_wouldblock_error(bio); else bio_io_error(bio); |
3ef28e83a
|
2360 |
} |
f5fe1b519
|
2361 |
bio = bio_list_pop(&bio_list_on_stack[0]); |
d89d87965
|
2362 |
} while (bio); |
bddd87c7e
|
2363 |
current->bio_list = NULL; /* deactivate */ |
dece16353
|
2364 2365 |
out: |
37f9579f4
|
2366 2367 |
if (q) blk_queue_exit(q); |
dece16353
|
2368 |
return ret; |
d89d87965
|
2369 |
} |
1da177e4c
|
2370 2371 2372 |
EXPORT_SYMBOL(generic_make_request); /** |
f421e1d9a
|
2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 |
* direct_make_request - hand a buffer directly to its device driver for I/O * @bio: The bio describing the location in memory and on the device. * * This function behaves like generic_make_request(), but does not protect * against recursion. Must only be used if the called driver is known * to not call generic_make_request (or direct_make_request) again from * its make_request function. (Calling direct_make_request again from * a workqueue is perfectly fine as that doesn't recurse). */ blk_qc_t direct_make_request(struct bio *bio) { struct request_queue *q = bio->bi_disk->queue; bool nowait = bio->bi_opf & REQ_NOWAIT; blk_qc_t ret; if (!generic_make_request_checks(bio)) return BLK_QC_T_NONE; |
3a0a52997
|
2390 |
if (unlikely(blk_queue_enter(q, nowait ? BLK_MQ_REQ_NOWAIT : 0))) { |
f421e1d9a
|
2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 |
if (nowait && !blk_queue_dying(q)) bio->bi_status = BLK_STS_AGAIN; else bio->bi_status = BLK_STS_IOERR; bio_endio(bio); return BLK_QC_T_NONE; } ret = q->make_request_fn(q, bio); blk_queue_exit(q); return ret; } EXPORT_SYMBOL_GPL(direct_make_request); /** |
710027a48
|
2406 |
* submit_bio - submit a bio to the block device layer for I/O |
1da177e4c
|
2407 2408 2409 2410 |
* @bio: The &struct bio which describes the I/O * * submit_bio() is very similar in purpose to generic_make_request(), and * uses that function to do most of the work. Both are fairly rough |
710027a48
|
2411 |
* interfaces; @bio must be presetup and ready for I/O. |
1da177e4c
|
2412 2413 |
* */ |
4e49ea4a3
|
2414 |
blk_qc_t submit_bio(struct bio *bio) |
1da177e4c
|
2415 |
{ |
bf2de6f5a
|
2416 2417 2418 2419 |
/* * If it's a regular read/write or a barrier with data attached, * go through the normal accounting stuff before submission. */ |
e2a60da74
|
2420 |
if (bio_has_data(bio)) { |
4363ac7c1
|
2421 |
unsigned int count; |
95fe6c1a2
|
2422 |
if (unlikely(bio_op(bio) == REQ_OP_WRITE_SAME)) |
7c5a0dcf5
|
2423 |
count = queue_logical_block_size(bio->bi_disk->queue) >> 9; |
4363ac7c1
|
2424 2425 |
else count = bio_sectors(bio); |
a8ebb056a
|
2426 |
if (op_is_write(bio_op(bio))) { |
bf2de6f5a
|
2427 2428 |
count_vm_events(PGPGOUT, count); } else { |
4f024f379
|
2429 |
task_io_account_read(bio->bi_iter.bi_size); |
bf2de6f5a
|
2430 2431 2432 2433 2434 |
count_vm_events(PGPGIN, count); } if (unlikely(block_dump)) { char b[BDEVNAME_SIZE]; |
8dcbdc742
|
2435 2436 |
printk(KERN_DEBUG "%s(%d): %s block %Lu on %s (%u sectors) ", |
ba25f9dcc
|
2437 |
current->comm, task_pid_nr(current), |
a8ebb056a
|
2438 |
op_is_write(bio_op(bio)) ? "WRITE" : "READ", |
4f024f379
|
2439 |
(unsigned long long)bio->bi_iter.bi_sector, |
74d46992e
|
2440 |
bio_devname(bio, b), count); |
bf2de6f5a
|
2441 |
} |
1da177e4c
|
2442 |
} |
dece16353
|
2443 |
return generic_make_request(bio); |
1da177e4c
|
2444 |
} |
1da177e4c
|
2445 |
EXPORT_SYMBOL(submit_bio); |
ea435e1b9
|
2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 |
bool blk_poll(struct request_queue *q, blk_qc_t cookie) { if (!q->poll_fn || !blk_qc_t_valid(cookie)) return false; if (current->plug) blk_flush_plug_list(current->plug, false); return q->poll_fn(q, cookie); } EXPORT_SYMBOL_GPL(blk_poll); |
3bcddeac1
|
2456 |
/** |
bf4e6b4e7
|
2457 2458 |
* blk_cloned_rq_check_limits - Helper function to check a cloned request * for new the queue limits |
82124d603
|
2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 |
* @q: the queue * @rq: the request being checked * * Description: * @rq may have been made based on weaker limitations of upper-level queues * in request stacking drivers, and it may violate the limitation of @q. * Since the block layer and the underlying device driver trust @rq * after it is inserted to @q, it should be checked against @q before * the insertion using this generic function. * |
82124d603
|
2469 |
* Request stacking drivers like request-based dm may change the queue |
bf4e6b4e7
|
2470 2471 |
* limits when retrying requests on other queues. Those requests need * to be checked against the new queue limits again during dispatch. |
82124d603
|
2472 |
*/ |
bf4e6b4e7
|
2473 2474 |
static int blk_cloned_rq_check_limits(struct request_queue *q, struct request *rq) |
82124d603
|
2475 |
{ |
8fe0d473f
|
2476 |
if (blk_rq_sectors(rq) > blk_queue_get_max_sectors(q, req_op(rq))) { |
82124d603
|
2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 |
printk(KERN_ERR "%s: over max size limit. ", __func__); return -EIO; } /* * queue's settings related to segment counting like q->bounce_pfn * may differ from that of other stacking queues. * Recalculate it to check the request correctly on this queue's * limitation. */ blk_recalc_rq_segments(rq); |
8a78362c4
|
2489 |
if (rq->nr_phys_segments > queue_max_segments(q)) { |
82124d603
|
2490 2491 2492 2493 2494 2495 2496 |
printk(KERN_ERR "%s: over max segments limit. ", __func__); return -EIO; } return 0; } |
82124d603
|
2497 2498 2499 2500 2501 2502 |
/** * blk_insert_cloned_request - Helper for stacking drivers to submit a request * @q: the queue to submit the request * @rq: the request being queued */ |
2a842acab
|
2503 |
blk_status_t blk_insert_cloned_request(struct request_queue *q, struct request *rq) |
82124d603
|
2504 2505 |
{ unsigned long flags; |
4853abaae
|
2506 |
int where = ELEVATOR_INSERT_BACK; |
82124d603
|
2507 |
|
bf4e6b4e7
|
2508 |
if (blk_cloned_rq_check_limits(q, rq)) |
2a842acab
|
2509 |
return BLK_STS_IOERR; |
82124d603
|
2510 |
|
b2c9cd379
|
2511 2512 |
if (rq->rq_disk && should_fail_request(&rq->rq_disk->part0, blk_rq_bytes(rq))) |
2a842acab
|
2513 |
return BLK_STS_IOERR; |
82124d603
|
2514 |
|
7fb4898e0
|
2515 2516 2517 |
if (q->mq_ops) { if (blk_queue_io_stat(q)) blk_account_io_start(rq, true); |
157f377be
|
2518 2519 2520 2521 2522 |
/* * Since we have a scheduler attached on the top device, * bypass a potential scheduler on the bottom device for * insert. */ |
c77ff7fd0
|
2523 |
return blk_mq_request_issue_directly(rq); |
7fb4898e0
|
2524 |
} |
82124d603
|
2525 |
spin_lock_irqsave(q->queue_lock, flags); |
3f3299d5c
|
2526 |
if (unlikely(blk_queue_dying(q))) { |
8ba61435d
|
2527 |
spin_unlock_irqrestore(q->queue_lock, flags); |
2a842acab
|
2528 |
return BLK_STS_IOERR; |
8ba61435d
|
2529 |
} |
82124d603
|
2530 2531 2532 2533 2534 2535 |
/* * Submitting request must be dequeued before calling this function * because it will be linked to another request_queue */ BUG_ON(blk_queued_rq(rq)); |
f73f44eb0
|
2536 |
if (op_is_flush(rq->cmd_flags)) |
4853abaae
|
2537 2538 2539 |
where = ELEVATOR_INSERT_FLUSH; add_acct_request(q, rq, where); |
e67b77c79
|
2540 2541 |
if (where == ELEVATOR_INSERT_FLUSH) __blk_run_queue(q); |
82124d603
|
2542 |
spin_unlock_irqrestore(q->queue_lock, flags); |
2a842acab
|
2543 |
return BLK_STS_OK; |
82124d603
|
2544 2545 |
} EXPORT_SYMBOL_GPL(blk_insert_cloned_request); |
80a761fd3
|
2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 |
/** * blk_rq_err_bytes - determine number of bytes till the next failure boundary * @rq: request to examine * * Description: * A request could be merge of IOs which require different failure * handling. This function determines the number of bytes which * can be failed from the beginning of the request without * crossing into area which need to be retried further. * * Return: * The number of bytes to fail. |
80a761fd3
|
2558 2559 2560 2561 2562 2563 |
*/ unsigned int blk_rq_err_bytes(const struct request *rq) { unsigned int ff = rq->cmd_flags & REQ_FAILFAST_MASK; unsigned int bytes = 0; struct bio *bio; |
e80640213
|
2564 |
if (!(rq->rq_flags & RQF_MIXED_MERGE)) |
80a761fd3
|
2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 |
return blk_rq_bytes(rq); /* * Currently the only 'mixing' which can happen is between * different fastfail types. We can safely fail portions * which have all the failfast bits that the first one has - * the ones which are at least as eager to fail as the first * one. */ for (bio = rq->bio; bio; bio = bio->bi_next) { |
1eff9d322
|
2575 |
if ((bio->bi_opf & ff) != ff) |
80a761fd3
|
2576 |
break; |
4f024f379
|
2577 |
bytes += bio->bi_iter.bi_size; |
80a761fd3
|
2578 2579 2580 2581 2582 2583 2584 |
} /* this could lead to infinite loop */ BUG_ON(blk_rq_bytes(rq) && !bytes); return bytes; } EXPORT_SYMBOL_GPL(blk_rq_err_bytes); |
320ae51fe
|
2585 |
void blk_account_io_completion(struct request *req, unsigned int bytes) |
bc58ba946
|
2586 |
{ |
c2553b584
|
2587 |
if (blk_do_io_stat(req)) { |
ddcf35d39
|
2588 |
const int sgrp = op_stat_group(req_op(req)); |
bc58ba946
|
2589 2590 2591 2592 |
struct hd_struct *part; int cpu; cpu = part_stat_lock(); |
09e099d4b
|
2593 |
part = req->part; |
ddcf35d39
|
2594 |
part_stat_add(cpu, part, sectors[sgrp], bytes >> 9); |
bc58ba946
|
2595 2596 2597 |
part_stat_unlock(); } } |
522a77756
|
2598 |
void blk_account_io_done(struct request *req, u64 now) |
bc58ba946
|
2599 |
{ |
bc58ba946
|
2600 |
/* |
dd4c133f3
|
2601 2602 2603 |
* Account IO completion. flush_rq isn't accounted as a * normal IO on queueing nor completion. Accounting the * containing request is enough. |
bc58ba946
|
2604 |
*/ |
e80640213
|
2605 |
if (blk_do_io_stat(req) && !(req->rq_flags & RQF_FLUSH_SEQ)) { |
ddcf35d39
|
2606 |
const int sgrp = op_stat_group(req_op(req)); |
bc58ba946
|
2607 2608 2609 2610 |
struct hd_struct *part; int cpu; cpu = part_stat_lock(); |
09e099d4b
|
2611 |
part = req->part; |
bc58ba946
|
2612 |
|
ddcf35d39
|
2613 |
part_stat_inc(cpu, part, ios[sgrp]); |
b57e99b4b
|
2614 |
part_stat_add(cpu, part, nsecs[sgrp], now - req->start_time_ns); |
d62e26b3f
|
2615 |
part_round_stats(req->q, cpu, part); |
ddcf35d39
|
2616 |
part_dec_in_flight(req->q, part, rq_data_dir(req)); |
bc58ba946
|
2617 |
|
6c23a9681
|
2618 |
hd_struct_put(part); |
bc58ba946
|
2619 2620 2621 |
part_stat_unlock(); } } |
47fafbc70
|
2622 |
#ifdef CONFIG_PM |
c8158819d
|
2623 2624 2625 2626 |
/* * Don't process normal requests when queue is suspended * or in the process of suspending/resuming */ |
e4f36b249
|
2627 |
static bool blk_pm_allow_request(struct request *rq) |
c8158819d
|
2628 |
{ |
e4f36b249
|
2629 2630 2631 2632 2633 2634 |
switch (rq->q->rpm_status) { case RPM_RESUMING: case RPM_SUSPENDING: return rq->rq_flags & RQF_PM; case RPM_SUSPENDED: return false; |
e9a838533
|
2635 2636 |
default: return true; |
e4f36b249
|
2637 |
} |
c8158819d
|
2638 2639 |
} #else |
e4f36b249
|
2640 |
static bool blk_pm_allow_request(struct request *rq) |
c8158819d
|
2641 |
{ |
e4f36b249
|
2642 |
return true; |
c8158819d
|
2643 2644 |
} #endif |
320ae51fe
|
2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 |
void blk_account_io_start(struct request *rq, bool new_io) { struct hd_struct *part; int rw = rq_data_dir(rq); int cpu; if (!blk_do_io_stat(rq)) return; cpu = part_stat_lock(); if (!new_io) { part = rq->part; part_stat_inc(cpu, part, merges[rw]); } else { part = disk_map_sector_rcu(rq->rq_disk, blk_rq_pos(rq)); if (!hd_struct_try_get(part)) { /* * The partition is already being removed, * the request will be accounted on the disk only * * We take a reference on disk->part0 although that * partition will never be deleted, so we can treat * it as any other partition. */ part = &rq->rq_disk->part0; hd_struct_get(part); } |
d62e26b3f
|
2673 2674 |
part_round_stats(rq->q, cpu, part); part_inc_in_flight(rq->q, part, rw); |
320ae51fe
|
2675 2676 2677 2678 2679 |
rq->part = part; } part_stat_unlock(); } |
9c9883744
|
2680 2681 2682 2683 2684 2685 2686 2687 |
static struct request *elv_next_request(struct request_queue *q) { struct request *rq; struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL); WARN_ON_ONCE(q->mq_ops); while (1) { |
e4f36b249
|
2688 2689 2690 2691 2692 2693 |
list_for_each_entry(rq, &q->queue_head, queuelist) { if (blk_pm_allow_request(rq)) return rq; if (rq->rq_flags & RQF_SOFTBARRIER) break; |
9c9883744
|
2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 |
} /* * Flush request is running and flush request isn't queueable * in the drive, we can hold the queue till flush request is * finished. Even we don't do this, driver can't dispatch next * requests and will requeue them. And this can improve * throughput too. For example, we have request flush1, write1, * flush 2. flush1 is dispatched, then queue is hold, write1 * isn't inserted to queue. After flush1 is finished, flush2 * will be dispatched. Since disk cache is already clean, * flush2 will be finished very soon, so looks like flush2 is * folded to flush1. * Since the queue is hold, a flag is set to indicate the queue * should be restarted later. Please see flush_end_io() for * details. */ if (fq->flush_pending_idx != fq->flush_running_idx && !queue_flush_queueable(q)) { fq->flush_queue_delayed = 1; return NULL; } if (unlikely(blk_queue_bypass(q)) || !q->elevator->type->ops.sq.elevator_dispatch_fn(q, 0)) return NULL; } } |
53a08807c
|
2721 |
/** |
9934c8c04
|
2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 |
* blk_peek_request - peek at the top of a request queue * @q: request queue to peek at * * Description: * Return the request at the top of @q. The returned request * should be started using blk_start_request() before LLD starts * processing it. * * Return: * Pointer to the request at the top of @q if available. Null * otherwise. |
9934c8c04
|
2733 2734 |
*/ struct request *blk_peek_request(struct request_queue *q) |
158dbda00
|
2735 2736 2737 |
{ struct request *rq; int ret; |
2fff8a924
|
2738 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
2739 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
2740 |
|
9c9883744
|
2741 |
while ((rq = elv_next_request(q)) != NULL) { |
e80640213
|
2742 |
if (!(rq->rq_flags & RQF_STARTED)) { |
158dbda00
|
2743 2744 2745 2746 2747 |
/* * This is the first time the device driver * sees this request (possibly after * requeueing). Notify IO scheduler. */ |
e80640213
|
2748 |
if (rq->rq_flags & RQF_SORTED) |
158dbda00
|
2749 2750 2751 2752 2753 2754 2755 |
elv_activate_rq(q, rq); /* * just mark as started even if we don't start * it, a request that has been delayed should * not be passed by new incoming requests */ |
e80640213
|
2756 |
rq->rq_flags |= RQF_STARTED; |
158dbda00
|
2757 2758 2759 2760 2761 2762 2763 |
trace_block_rq_issue(q, rq); } if (!q->boundary_rq || q->boundary_rq == rq) { q->end_sector = rq_end_sector(rq); q->boundary_rq = NULL; } |
e80640213
|
2764 |
if (rq->rq_flags & RQF_DONTPREP) |
158dbda00
|
2765 |
break; |
2e46e8b27
|
2766 |
if (q->dma_drain_size && blk_rq_bytes(rq)) { |
158dbda00
|
2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 |
/* * make sure space for the drain appears we * know we can do this because max_hw_segments * has been adjusted to be one fewer than the * device can handle */ rq->nr_phys_segments++; } if (!q->prep_rq_fn) break; ret = q->prep_rq_fn(q, rq); if (ret == BLKPREP_OK) { break; } else if (ret == BLKPREP_DEFER) { /* * the request may have been (partially) prepped. * we need to keep this request in the front to |
e80640213
|
2786 |
* avoid resource deadlock. RQF_STARTED will |
158dbda00
|
2787 2788 |
* prevent other fs requests from passing this one. */ |
2e46e8b27
|
2789 |
if (q->dma_drain_size && blk_rq_bytes(rq) && |
e80640213
|
2790 |
!(rq->rq_flags & RQF_DONTPREP)) { |
158dbda00
|
2791 2792 2793 2794 2795 2796 2797 2798 2799 |
/* * remove the space for the drain we added * so that we don't add it again */ --rq->nr_phys_segments; } rq = NULL; break; |
0fb5b1fb3
|
2800 |
} else if (ret == BLKPREP_KILL || ret == BLKPREP_INVALID) { |
e80640213
|
2801 |
rq->rq_flags |= RQF_QUIET; |
c143dc903
|
2802 2803 2804 2805 2806 |
/* * Mark this request as started so we don't trigger * any debug logic in the end I/O path. */ blk_start_request(rq); |
2a842acab
|
2807 2808 |
__blk_end_request_all(rq, ret == BLKPREP_INVALID ? BLK_STS_TARGET : BLK_STS_IOERR); |
158dbda00
|
2809 2810 2811 2812 2813 2814 2815 2816 2817 |
} else { printk(KERN_ERR "%s: bad return=%d ", __func__, ret); break; } } return rq; } |
9934c8c04
|
2818 |
EXPORT_SYMBOL(blk_peek_request); |
158dbda00
|
2819 |
|
5034435c8
|
2820 |
static void blk_dequeue_request(struct request *rq) |
158dbda00
|
2821 |
{ |
9934c8c04
|
2822 |
struct request_queue *q = rq->q; |
158dbda00
|
2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 |
BUG_ON(list_empty(&rq->queuelist)); BUG_ON(ELV_ON_HASH(rq)); list_del_init(&rq->queuelist); /* * the time frame between a request being removed from the lists * and to it is freed is accounted as io that is in progress at * the driver side. */ |
522a77756
|
2833 |
if (blk_account_rq(rq)) |
0a7ae2ff0
|
2834 |
q->in_flight[rq_is_sync(rq)]++; |
158dbda00
|
2835 |
} |
5efccd17c
|
2836 |
/** |
9934c8c04
|
2837 2838 2839 2840 2841 2842 |
* blk_start_request - start request processing on the driver * @req: request to dequeue * * Description: * Dequeue @req and start timeout timer on it. This hands off the * request to the driver. |
9934c8c04
|
2843 2844 2845 |
*/ void blk_start_request(struct request *req) { |
2fff8a924
|
2846 |
lockdep_assert_held(req->q->queue_lock); |
332ebbf7f
|
2847 |
WARN_ON_ONCE(req->q->mq_ops); |
2fff8a924
|
2848 |
|
9934c8c04
|
2849 |
blk_dequeue_request(req); |
cf43e6be8
|
2850 |
if (test_bit(QUEUE_FLAG_STATS, &req->q->queue_flags)) { |
544ccc8dc
|
2851 2852 2853 2854 |
req->io_start_time_ns = ktime_get_ns(); #ifdef CONFIG_BLK_DEV_THROTTLING_LOW req->throtl_size = blk_rq_sectors(req); #endif |
cf43e6be8
|
2855 |
req->rq_flags |= RQF_STATS; |
a79050434
|
2856 |
rq_qos_issue(req->q, req); |
cf43e6be8
|
2857 |
} |
e14575b3d
|
2858 |
BUG_ON(blk_rq_is_complete(req)); |
9934c8c04
|
2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 |
blk_add_timer(req); } EXPORT_SYMBOL(blk_start_request); /** * blk_fetch_request - fetch a request from a request queue * @q: request queue to fetch a request from * * Description: * Return the request at the top of @q. The request is started on * return and LLD can start processing it immediately. * * Return: * Pointer to the request at the top of @q if available. Null * otherwise. |
9934c8c04
|
2874 2875 2876 2877 |
*/ struct request *blk_fetch_request(struct request_queue *q) { struct request *rq; |
2fff8a924
|
2878 |
lockdep_assert_held(q->queue_lock); |
332ebbf7f
|
2879 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
2880 |
|
9934c8c04
|
2881 2882 2883 2884 2885 2886 |
rq = blk_peek_request(q); if (rq) blk_start_request(rq); return rq; } EXPORT_SYMBOL(blk_fetch_request); |
ef71de8b1
|
2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 |
/* * Steal bios from a request and add them to a bio list. * The request must not have been partially completed before. */ void blk_steal_bios(struct bio_list *list, struct request *rq) { if (rq->bio) { if (list->tail) list->tail->bi_next = rq->bio; else list->head = rq->bio; list->tail = rq->biotail; rq->bio = NULL; rq->biotail = NULL; } rq->__data_len = 0; } EXPORT_SYMBOL_GPL(blk_steal_bios); |
9934c8c04
|
2907 |
/** |
2e60e0229
|
2908 |
* blk_update_request - Special helper function for request stacking drivers |
8ebf97560
|
2909 |
* @req: the request being processed |
2a842acab
|
2910 |
* @error: block status code |
8ebf97560
|
2911 |
* @nr_bytes: number of bytes to complete @req |
3bcddeac1
|
2912 2913 |
* * Description: |
8ebf97560
|
2914 2915 2916 |
* Ends I/O on a number of bytes attached to @req, but doesn't complete * the request structure even if @req doesn't have leftover. * If @req has leftover, sets it up for the next range of segments. |
2e60e0229
|
2917 2918 2919 2920 2921 2922 2923 |
* * This special helper function is only for request stacking drivers * (e.g. request-based dm) so that they can handle partial completion. * Actual device drivers should use blk_end_request instead. * * Passing the result of blk_rq_bytes() as @nr_bytes guarantees * %false return from this function. |
3bcddeac1
|
2924 |
* |
1954e9a99
|
2925 2926 2927 2928 |
* Note: * The RQF_SPECIAL_PAYLOAD flag is ignored on purpose in both * blk_rq_bytes() and in blk_update_request(). * |
3bcddeac1
|
2929 |
* Return: |
2e60e0229
|
2930 2931 |
* %false - this request doesn't have any more data * %true - this request has more data |
3bcddeac1
|
2932 |
**/ |
2a842acab
|
2933 2934 |
bool blk_update_request(struct request *req, blk_status_t error, unsigned int nr_bytes) |
1da177e4c
|
2935 |
{ |
f79ea4161
|
2936 |
int total_bytes; |
1da177e4c
|
2937 |
|
2a842acab
|
2938 |
trace_block_rq_complete(req, blk_status_to_errno(error), nr_bytes); |
4a0efdc93
|
2939 |
|
2e60e0229
|
2940 2941 |
if (!req->bio) return false; |
2a842acab
|
2942 2943 2944 |
if (unlikely(error && !blk_rq_is_passthrough(req) && !(req->rq_flags & RQF_QUIET))) print_req_error(req, error); |
1da177e4c
|
2945 |
|
bc58ba946
|
2946 |
blk_account_io_completion(req, nr_bytes); |
d72d904a5
|
2947 |
|
f79ea4161
|
2948 2949 2950 |
total_bytes = 0; while (req->bio) { struct bio *bio = req->bio; |
4f024f379
|
2951 |
unsigned bio_bytes = min(bio->bi_iter.bi_size, nr_bytes); |
1da177e4c
|
2952 |
|
9c24c10a2
|
2953 |
if (bio_bytes == bio->bi_iter.bi_size) |
1da177e4c
|
2954 |
req->bio = bio->bi_next; |
1da177e4c
|
2955 |
|
fbbaf700e
|
2956 2957 |
/* Completion has already been traced */ bio_clear_flag(bio, BIO_TRACE_COMPLETION); |
f79ea4161
|
2958 |
req_bio_endio(req, bio, bio_bytes, error); |
1da177e4c
|
2959 |
|
f79ea4161
|
2960 2961 |
total_bytes += bio_bytes; nr_bytes -= bio_bytes; |
1da177e4c
|
2962 |
|
f79ea4161
|
2963 2964 |
if (!nr_bytes) break; |
1da177e4c
|
2965 2966 2967 2968 2969 |
} /* * completely done */ |
2e60e0229
|
2970 2971 2972 2973 2974 2975 |
if (!req->bio) { /* * Reset counters so that the request stacking driver * can find how many bytes remain in the request * later. */ |
a2dec7b36
|
2976 |
req->__data_len = 0; |
2e60e0229
|
2977 2978 |
return false; } |
1da177e4c
|
2979 |
|
a2dec7b36
|
2980 |
req->__data_len -= total_bytes; |
2e46e8b27
|
2981 2982 |
/* update sector only for requests with clear definition of sector */ |
57292b58d
|
2983 |
if (!blk_rq_is_passthrough(req)) |
a2dec7b36
|
2984 |
req->__sector += total_bytes >> 9; |
2e46e8b27
|
2985 |
|
80a761fd3
|
2986 |
/* mixed attributes always follow the first bio */ |
e80640213
|
2987 |
if (req->rq_flags & RQF_MIXED_MERGE) { |
80a761fd3
|
2988 |
req->cmd_flags &= ~REQ_FAILFAST_MASK; |
1eff9d322
|
2989 |
req->cmd_flags |= req->bio->bi_opf & REQ_FAILFAST_MASK; |
80a761fd3
|
2990 |
} |
ed6565e73
|
2991 2992 2993 2994 2995 2996 2997 2998 2999 |
if (!(req->rq_flags & RQF_SPECIAL_PAYLOAD)) { /* * If total number of sectors is less than the first segment * size, something has gone terribly wrong. */ if (blk_rq_bytes(req) < blk_rq_cur_bytes(req)) { blk_dump_rq_flags(req, "request botched"); req->__data_len = blk_rq_cur_bytes(req); } |
2e46e8b27
|
3000 |
|
ed6565e73
|
3001 3002 3003 |
/* recalculate the number of segments */ blk_recalc_rq_segments(req); } |
2e46e8b27
|
3004 |
|
2e60e0229
|
3005 |
return true; |
1da177e4c
|
3006 |
} |
2e60e0229
|
3007 |
EXPORT_SYMBOL_GPL(blk_update_request); |
1da177e4c
|
3008 |
|
2a842acab
|
3009 |
static bool blk_update_bidi_request(struct request *rq, blk_status_t error, |
2e60e0229
|
3010 3011 |
unsigned int nr_bytes, unsigned int bidi_bytes) |
5efccd17c
|
3012 |
{ |
2e60e0229
|
3013 3014 |
if (blk_update_request(rq, error, nr_bytes)) return true; |
5efccd17c
|
3015 |
|
2e60e0229
|
3016 3017 3018 3019 |
/* Bidi request must be completed as a whole */ if (unlikely(blk_bidi_rq(rq)) && blk_update_request(rq->next_rq, error, bidi_bytes)) return true; |
5efccd17c
|
3020 |
|
e2e1a148b
|
3021 3022 |
if (blk_queue_add_random(rq->q)) add_disk_randomness(rq->rq_disk); |
2e60e0229
|
3023 3024 |
return false; |
1da177e4c
|
3025 |
} |
28018c242
|
3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 |
/** * blk_unprep_request - unprepare a request * @req: the request * * This function makes a request ready for complete resubmission (or * completion). It happens only after all error handling is complete, * so represents the appropriate moment to deallocate any resources * that were allocated to the request in the prep_rq_fn. The queue * lock is held when calling this. */ void blk_unprep_request(struct request *req) { struct request_queue *q = req->q; |
e80640213
|
3039 |
req->rq_flags &= ~RQF_DONTPREP; |
28018c242
|
3040 3041 3042 3043 |
if (q->unprep_rq_fn) q->unprep_rq_fn(q, req); } EXPORT_SYMBOL_GPL(blk_unprep_request); |
2a842acab
|
3044 |
void blk_finish_request(struct request *req, blk_status_t error) |
1da177e4c
|
3045 |
{ |
cf43e6be8
|
3046 |
struct request_queue *q = req->q; |
522a77756
|
3047 |
u64 now = ktime_get_ns(); |
cf43e6be8
|
3048 |
|
2fff8a924
|
3049 |
lockdep_assert_held(req->q->queue_lock); |
332ebbf7f
|
3050 |
WARN_ON_ONCE(q->mq_ops); |
2fff8a924
|
3051 |
|
cf43e6be8
|
3052 |
if (req->rq_flags & RQF_STATS) |
522a77756
|
3053 |
blk_stat_add(req, now); |
cf43e6be8
|
3054 |
|
e80640213
|
3055 |
if (req->rq_flags & RQF_QUEUED) |
cf43e6be8
|
3056 |
blk_queue_end_tag(q, req); |
b8286239d
|
3057 |
|
ba396a6c1
|
3058 |
BUG_ON(blk_queued_rq(req)); |
1da177e4c
|
3059 |
|
57292b58d
|
3060 |
if (unlikely(laptop_mode) && !blk_rq_is_passthrough(req)) |
dc3b17cc8
|
3061 |
laptop_io_completion(req->q->backing_dev_info); |
1da177e4c
|
3062 |
|
e78042e5b
|
3063 |
blk_delete_timer(req); |
e80640213
|
3064 |
if (req->rq_flags & RQF_DONTPREP) |
28018c242
|
3065 |
blk_unprep_request(req); |
522a77756
|
3066 |
blk_account_io_done(req, now); |
b8286239d
|
3067 |
|
87760e5ee
|
3068 |
if (req->end_io) { |
a79050434
|
3069 |
rq_qos_done(q, req); |
8ffdc6550
|
3070 |
req->end_io(req, error); |
87760e5ee
|
3071 |
} else { |
b8286239d
|
3072 3073 |
if (blk_bidi_rq(req)) __blk_put_request(req->next_rq->q, req->next_rq); |
cf43e6be8
|
3074 |
__blk_put_request(q, req); |
b8286239d
|
3075 |
} |
1da177e4c
|
3076 |
} |
12120077b
|
3077 |
EXPORT_SYMBOL(blk_finish_request); |
1da177e4c
|
3078 |
|
3b11313a6
|
3079 |
/** |
2e60e0229
|
3080 3081 |
* blk_end_bidi_request - Complete a bidi request * @rq: the request to complete |
2a842acab
|
3082 |
* @error: block status code |
2e60e0229
|
3083 3084 |
* @nr_bytes: number of bytes to complete @rq * @bidi_bytes: number of bytes to complete @rq->next_rq |
a0cd12854
|
3085 3086 |
* * Description: |
e3a04fe34
|
3087 |
* Ends I/O on a number of bytes attached to @rq and @rq->next_rq. |
2e60e0229
|
3088 3089 3090 |
* Drivers that supports bidi can safely call this member for any * type of request, bidi or uni. In the later case @bidi_bytes is * just ignored. |
336cdb400
|
3091 3092 |
* * Return: |
2e60e0229
|
3093 3094 |
* %false - we are done with this request * %true - still buffers pending for this request |
a0cd12854
|
3095 |
**/ |
2a842acab
|
3096 |
static bool blk_end_bidi_request(struct request *rq, blk_status_t error, |
32fab448e
|
3097 3098 |
unsigned int nr_bytes, unsigned int bidi_bytes) { |
336cdb400
|
3099 |
struct request_queue *q = rq->q; |
2e60e0229
|
3100 |
unsigned long flags; |
32fab448e
|
3101 |
|
332ebbf7f
|
3102 |
WARN_ON_ONCE(q->mq_ops); |
2e60e0229
|
3103 3104 |
if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes)) return true; |
32fab448e
|
3105 |
|
336cdb400
|
3106 |
spin_lock_irqsave(q->queue_lock, flags); |
2e60e0229
|
3107 |
blk_finish_request(rq, error); |
336cdb400
|
3108 |
spin_unlock_irqrestore(q->queue_lock, flags); |
2e60e0229
|
3109 |
return false; |
32fab448e
|
3110 |
} |
336cdb400
|
3111 |
/** |
2e60e0229
|
3112 3113 |
* __blk_end_bidi_request - Complete a bidi request with queue lock held * @rq: the request to complete |
2a842acab
|
3114 |
* @error: block status code |
e3a04fe34
|
3115 3116 |
* @nr_bytes: number of bytes to complete @rq * @bidi_bytes: number of bytes to complete @rq->next_rq |
336cdb400
|
3117 3118 |
* * Description: |
2e60e0229
|
3119 3120 |
* Identical to blk_end_bidi_request() except that queue lock is * assumed to be locked on entry and remains so on return. |
336cdb400
|
3121 3122 |
* * Return: |
2e60e0229
|
3123 3124 |
* %false - we are done with this request * %true - still buffers pending for this request |
336cdb400
|
3125 |
**/ |
2a842acab
|
3126 |
static bool __blk_end_bidi_request(struct request *rq, blk_status_t error, |
b1f744937
|
3127 |
unsigned int nr_bytes, unsigned int bidi_bytes) |
336cdb400
|
3128 |
{ |
2fff8a924
|
3129 |
lockdep_assert_held(rq->q->queue_lock); |
332ebbf7f
|
3130 |
WARN_ON_ONCE(rq->q->mq_ops); |
2fff8a924
|
3131 |
|
2e60e0229
|
3132 3133 |
if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes)) return true; |
336cdb400
|
3134 |
|
2e60e0229
|
3135 |
blk_finish_request(rq, error); |
336cdb400
|
3136 |
|
2e60e0229
|
3137 |
return false; |
336cdb400
|
3138 |
} |
e19a3ab05
|
3139 3140 3141 3142 |
/** * blk_end_request - Helper function for drivers to complete the request. * @rq: the request being processed |
2a842acab
|
3143 |
* @error: block status code |
e19a3ab05
|
3144 3145 3146 3147 3148 3149 3150 |
* @nr_bytes: number of bytes to complete * * Description: * Ends I/O on a number of bytes attached to @rq. * If @rq has leftover, sets it up for the next range of segments. * * Return: |
b1f744937
|
3151 3152 |
* %false - we are done with this request * %true - still buffers pending for this request |
e19a3ab05
|
3153 |
**/ |
2a842acab
|
3154 3155 |
bool blk_end_request(struct request *rq, blk_status_t error, unsigned int nr_bytes) |
e19a3ab05
|
3156 |
{ |
332ebbf7f
|
3157 |
WARN_ON_ONCE(rq->q->mq_ops); |
b1f744937
|
3158 |
return blk_end_bidi_request(rq, error, nr_bytes, 0); |
e19a3ab05
|
3159 |
} |
56ad1740d
|
3160 |
EXPORT_SYMBOL(blk_end_request); |
336cdb400
|
3161 3162 |
/** |
b1f744937
|
3163 3164 |
* blk_end_request_all - Helper function for drives to finish the request. * @rq: the request to finish |
2a842acab
|
3165 |
* @error: block status code |
336cdb400
|
3166 3167 |
* * Description: |
b1f744937
|
3168 3169 |
* Completely finish @rq. */ |
2a842acab
|
3170 |
void blk_end_request_all(struct request *rq, blk_status_t error) |
336cdb400
|
3171 |
{ |
b1f744937
|
3172 3173 |
bool pending; unsigned int bidi_bytes = 0; |
336cdb400
|
3174 |
|
b1f744937
|
3175 3176 |
if (unlikely(blk_bidi_rq(rq))) bidi_bytes = blk_rq_bytes(rq->next_rq); |
336cdb400
|
3177 |
|
b1f744937
|
3178 3179 3180 |
pending = blk_end_bidi_request(rq, error, blk_rq_bytes(rq), bidi_bytes); BUG_ON(pending); } |
56ad1740d
|
3181 |
EXPORT_SYMBOL(blk_end_request_all); |
336cdb400
|
3182 |
|
b1f744937
|
3183 |
/** |
b1f744937
|
3184 3185 |
* __blk_end_request - Helper function for drivers to complete the request. * @rq: the request being processed |
2a842acab
|
3186 |
* @error: block status code |
b1f744937
|
3187 |
* @nr_bytes: number of bytes to complete |
e3a04fe34
|
3188 3189 |
* * Description: |
b1f744937
|
3190 |
* Must be called with queue lock held unlike blk_end_request(). |
e3a04fe34
|
3191 3192 |
* * Return: |
b1f744937
|
3193 3194 |
* %false - we are done with this request * %true - still buffers pending for this request |
e3a04fe34
|
3195 |
**/ |
2a842acab
|
3196 3197 |
bool __blk_end_request(struct request *rq, blk_status_t error, unsigned int nr_bytes) |
e3a04fe34
|
3198 |
{ |
2fff8a924
|
3199 |
lockdep_assert_held(rq->q->queue_lock); |
332ebbf7f
|
3200 |
WARN_ON_ONCE(rq->q->mq_ops); |
2fff8a924
|
3201 |
|
b1f744937
|
3202 |
return __blk_end_bidi_request(rq, error, nr_bytes, 0); |
e3a04fe34
|
3203 |
} |
56ad1740d
|
3204 |
EXPORT_SYMBOL(__blk_end_request); |
e3a04fe34
|
3205 3206 |
/** |
b1f744937
|
3207 3208 |
* __blk_end_request_all - Helper function for drives to finish the request. * @rq: the request to finish |
2a842acab
|
3209 |
* @error: block status code |
32fab448e
|
3210 3211 |
* * Description: |
b1f744937
|
3212 |
* Completely finish @rq. Must be called with queue lock held. |
32fab448e
|
3213 |
*/ |
2a842acab
|
3214 |
void __blk_end_request_all(struct request *rq, blk_status_t error) |
32fab448e
|
3215 |
{ |
b1f744937
|
3216 3217 |
bool pending; unsigned int bidi_bytes = 0; |
2fff8a924
|
3218 |
lockdep_assert_held(rq->q->queue_lock); |
332ebbf7f
|
3219 |
WARN_ON_ONCE(rq->q->mq_ops); |
2fff8a924
|
3220 |
|
b1f744937
|
3221 3222 3223 3224 3225 |
if (unlikely(blk_bidi_rq(rq))) bidi_bytes = blk_rq_bytes(rq->next_rq); pending = __blk_end_bidi_request(rq, error, blk_rq_bytes(rq), bidi_bytes); BUG_ON(pending); |
32fab448e
|
3226 |
} |
56ad1740d
|
3227 |
EXPORT_SYMBOL(__blk_end_request_all); |
32fab448e
|
3228 3229 |
/** |
b1f744937
|
3230 3231 |
* __blk_end_request_cur - Helper function to finish the current request chunk. * @rq: the request to finish the current chunk for |
2a842acab
|
3232 |
* @error: block status code |
e19a3ab05
|
3233 3234 |
* * Description: |
b1f744937
|
3235 3236 |
* Complete the current consecutively mapped chunk from @rq. Must * be called with queue lock held. |
e19a3ab05
|
3237 3238 |
* * Return: |
b1f744937
|
3239 3240 3241 |
* %false - we are done with this request * %true - still buffers pending for this request */ |
2a842acab
|
3242 |
bool __blk_end_request_cur(struct request *rq, blk_status_t error) |
e19a3ab05
|
3243 |
{ |
b1f744937
|
3244 |
return __blk_end_request(rq, error, blk_rq_cur_bytes(rq)); |
e19a3ab05
|
3245 |
} |
56ad1740d
|
3246 |
EXPORT_SYMBOL(__blk_end_request_cur); |
e19a3ab05
|
3247 |
|
86db1e297
|
3248 3249 |
void blk_rq_bio_prep(struct request_queue *q, struct request *rq, struct bio *bio) |
1da177e4c
|
3250 |
{ |
b4f42e283
|
3251 |
if (bio_has_data(bio)) |
fb2dce862
|
3252 |
rq->nr_phys_segments = bio_phys_segments(q, bio); |
445251d0f
|
3253 3254 |
else if (bio_op(bio) == REQ_OP_DISCARD) rq->nr_phys_segments = 1; |
b4f42e283
|
3255 |
|
4f024f379
|
3256 |
rq->__data_len = bio->bi_iter.bi_size; |
1da177e4c
|
3257 |
rq->bio = rq->biotail = bio; |
1da177e4c
|
3258 |
|
74d46992e
|
3259 3260 |
if (bio->bi_disk) rq->rq_disk = bio->bi_disk; |
66846572b
|
3261 |
} |
1da177e4c
|
3262 |
|
2d4dc890b
|
3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 |
#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE /** * rq_flush_dcache_pages - Helper function to flush all pages in a request * @rq: the request to be flushed * * Description: * Flush all pages in @rq. */ void rq_flush_dcache_pages(struct request *rq) { struct req_iterator iter; |
7988613b0
|
3274 |
struct bio_vec bvec; |
2d4dc890b
|
3275 3276 |
rq_for_each_segment(bvec, rq, iter) |
7988613b0
|
3277 |
flush_dcache_page(bvec.bv_page); |
2d4dc890b
|
3278 3279 3280 |
} EXPORT_SYMBOL_GPL(rq_flush_dcache_pages); #endif |
ef9e3facd
|
3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 |
/** * blk_lld_busy - Check if underlying low-level drivers of a device are busy * @q : the queue of the device being checked * * Description: * Check if underlying low-level drivers of a device are busy. * If the drivers want to export their busy state, they must set own * exporting function using blk_queue_lld_busy() first. * * Basically, this function is used only by request stacking drivers * to stop dispatching requests to underlying devices when underlying * devices are busy. This behavior helps more I/O merging on the queue * of the request stacking driver and prevents I/O throughput regression * on burst I/O load. * * Return: * 0 - Not busy (The request stacking driver should dispatch request) * 1 - Busy (The request stacking driver should stop dispatching request) */ int blk_lld_busy(struct request_queue *q) { if (q->lld_busy_fn) return q->lld_busy_fn(q); return 0; } EXPORT_SYMBOL_GPL(blk_lld_busy); |
78d8e58a0
|
3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 |
/** * blk_rq_unprep_clone - Helper function to free all bios in a cloned request * @rq: the clone request to be cleaned up * * Description: * Free all bios in @rq for a cloned request. */ void blk_rq_unprep_clone(struct request *rq) { struct bio *bio; while ((bio = rq->bio) != NULL) { rq->bio = bio->bi_next; bio_put(bio); } } EXPORT_SYMBOL_GPL(blk_rq_unprep_clone); /* * Copy attributes of the original request to the clone request. * The actual data parts (e.g. ->cmd, ->sense) are not copied. */ static void __blk_rq_prep_clone(struct request *dst, struct request *src) |
b0fd271d5
|
3332 3333 |
{ dst->cpu = src->cpu; |
b0fd271d5
|
3334 3335 |
dst->__sector = blk_rq_pos(src); dst->__data_len = blk_rq_bytes(src); |
297ba57dc
|
3336 3337 3338 3339 |
if (src->rq_flags & RQF_SPECIAL_PAYLOAD) { dst->rq_flags |= RQF_SPECIAL_PAYLOAD; dst->special_vec = src->special_vec; } |
b0fd271d5
|
3340 3341 3342 |
dst->nr_phys_segments = src->nr_phys_segments; dst->ioprio = src->ioprio; dst->extra_len = src->extra_len; |
78d8e58a0
|
3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 |
} /** * blk_rq_prep_clone - Helper function to setup clone request * @rq: the request to be setup * @rq_src: original request to be cloned * @bs: bio_set that bios for clone are allocated from * @gfp_mask: memory allocation mask for bio * @bio_ctr: setup function to be called for each clone bio. * Returns %0 for success, non %0 for failure. * @data: private data to be passed to @bio_ctr * * Description: * Clones bios in @rq_src to @rq, and copies attributes of @rq_src to @rq. * The actual data parts of @rq_src (e.g. ->cmd, ->sense) * are not copied, and copying such parts is the caller's responsibility. * Also, pages which the original bios are pointing to are not copied * and the cloned bios just point same pages. * So cloned bios must be completed before original bios, which means * the caller must complete @rq before @rq_src. */ int blk_rq_prep_clone(struct request *rq, struct request *rq_src, struct bio_set *bs, gfp_t gfp_mask, int (*bio_ctr)(struct bio *, struct bio *, void *), void *data) { struct bio *bio, *bio_src; if (!bs) |
f4f8154a0
|
3372 |
bs = &fs_bio_set; |
78d8e58a0
|
3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 |
__rq_for_each_bio(bio_src, rq_src) { bio = bio_clone_fast(bio_src, gfp_mask, bs); if (!bio) goto free_and_out; if (bio_ctr && bio_ctr(bio, bio_src, data)) goto free_and_out; if (rq->bio) { rq->biotail->bi_next = bio; rq->biotail = bio; } else rq->bio = rq->biotail = bio; } __blk_rq_prep_clone(rq, rq_src); return 0; free_and_out: if (bio) bio_put(bio); blk_rq_unprep_clone(rq); return -ENOMEM; |
b0fd271d5
|
3399 3400 |
} EXPORT_SYMBOL_GPL(blk_rq_prep_clone); |
59c3d45e4
|
3401 |
int kblockd_schedule_work(struct work_struct *work) |
1da177e4c
|
3402 3403 3404 |
{ return queue_work(kblockd_workqueue, work); } |
1da177e4c
|
3405 |
EXPORT_SYMBOL(kblockd_schedule_work); |
ee63cfa7f
|
3406 3407 3408 3409 3410 |
int kblockd_schedule_work_on(int cpu, struct work_struct *work) { return queue_work_on(cpu, kblockd_workqueue, work); } EXPORT_SYMBOL(kblockd_schedule_work_on); |
818cd1cba
|
3411 3412 3413 3414 3415 3416 |
int kblockd_mod_delayed_work_on(int cpu, struct delayed_work *dwork, unsigned long delay) { return mod_delayed_work_on(cpu, kblockd_workqueue, dwork, delay); } EXPORT_SYMBOL(kblockd_mod_delayed_work_on); |
75df71362
|
3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 |
/** * blk_start_plug - initialize blk_plug and track it inside the task_struct * @plug: The &struct blk_plug that needs to be initialized * * Description: * Tracking blk_plug inside the task_struct will help with auto-flushing the * pending I/O should the task end up blocking between blk_start_plug() and * blk_finish_plug(). This is important from a performance perspective, but * also ensures that we don't deadlock. For instance, if the task is blocking * for a memory allocation, memory reclaim could end up wanting to free a * page belonging to that request that is currently residing in our private * plug. By flushing the pending I/O when the process goes to sleep, we avoid * this kind of deadlock. */ |
73c101011
|
3431 3432 3433 |
void blk_start_plug(struct blk_plug *plug) { struct task_struct *tsk = current; |
dd6cf3e18
|
3434 3435 3436 3437 3438 |
/* * If this is a nested plug, don't actually assign it. */ if (tsk->plug) return; |
73c101011
|
3439 |
INIT_LIST_HEAD(&plug->list); |
320ae51fe
|
3440 |
INIT_LIST_HEAD(&plug->mq_list); |
048c9374a
|
3441 |
INIT_LIST_HEAD(&plug->cb_list); |
73c101011
|
3442 |
/* |
dd6cf3e18
|
3443 3444 |
* Store ordering should not be needed here, since a potential * preempt will imply a full memory barrier |
73c101011
|
3445 |
*/ |
dd6cf3e18
|
3446 |
tsk->plug = plug; |
73c101011
|
3447 3448 3449 3450 3451 3452 3453 |
} EXPORT_SYMBOL(blk_start_plug); static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) { struct request *rqa = container_of(a, struct request, queuelist); struct request *rqb = container_of(b, struct request, queuelist); |
975927b94
|
3454 3455 |
return !(rqa->q < rqb->q || (rqa->q == rqb->q && blk_rq_pos(rqa) < blk_rq_pos(rqb))); |
73c101011
|
3456 |
} |
49cac01e1
|
3457 3458 3459 3460 3461 3462 |
/* * If 'from_schedule' is true, then postpone the dispatch of requests * until a safe kblockd context. We due this to avoid accidental big * additional stack usage in driver dispatch, in places where the originally * plugger did not intend it. */ |
f6603783f
|
3463 |
static void queue_unplugged(struct request_queue *q, unsigned int depth, |
49cac01e1
|
3464 |
bool from_schedule) |
99e22598e
|
3465 |
__releases(q->queue_lock) |
94b5eb28b
|
3466 |
{ |
2fff8a924
|
3467 |
lockdep_assert_held(q->queue_lock); |
49cac01e1
|
3468 |
trace_block_unplug(q, depth, !from_schedule); |
99e22598e
|
3469 |
|
704605711
|
3470 |
if (from_schedule) |
24ecfbe27
|
3471 |
blk_run_queue_async(q); |
704605711
|
3472 |
else |
24ecfbe27
|
3473 |
__blk_run_queue(q); |
50864670b
|
3474 |
spin_unlock_irq(q->queue_lock); |
94b5eb28b
|
3475 |
} |
74018dc30
|
3476 |
static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule) |
048c9374a
|
3477 3478 |
{ LIST_HEAD(callbacks); |
2a7d5559b
|
3479 3480 |
while (!list_empty(&plug->cb_list)) { list_splice_init(&plug->cb_list, &callbacks); |
048c9374a
|
3481 |
|
2a7d5559b
|
3482 3483 |
while (!list_empty(&callbacks)) { struct blk_plug_cb *cb = list_first_entry(&callbacks, |
048c9374a
|
3484 3485 |
struct blk_plug_cb, list); |
2a7d5559b
|
3486 |
list_del(&cb->list); |
74018dc30
|
3487 |
cb->callback(cb, from_schedule); |
2a7d5559b
|
3488 |
} |
048c9374a
|
3489 3490 |
} } |
9cbb17508
|
3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 |
struct blk_plug_cb *blk_check_plugged(blk_plug_cb_fn unplug, void *data, int size) { struct blk_plug *plug = current->plug; struct blk_plug_cb *cb; if (!plug) return NULL; list_for_each_entry(cb, &plug->cb_list, list) if (cb->callback == unplug && cb->data == data) return cb; /* Not currently on the callback list */ BUG_ON(size < sizeof(*cb)); cb = kzalloc(size, GFP_ATOMIC); if (cb) { cb->data = data; cb->callback = unplug; list_add(&cb->list, &plug->cb_list); } return cb; } EXPORT_SYMBOL(blk_check_plugged); |
49cac01e1
|
3515 |
void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) |
73c101011
|
3516 3517 |
{ struct request_queue *q; |
73c101011
|
3518 |
struct request *rq; |
109b81296
|
3519 |
LIST_HEAD(list); |
94b5eb28b
|
3520 |
unsigned int depth; |
73c101011
|
3521 |
|
74018dc30
|
3522 |
flush_plug_callbacks(plug, from_schedule); |
320ae51fe
|
3523 3524 3525 |
if (!list_empty(&plug->mq_list)) blk_mq_flush_plug_list(plug, from_schedule); |
73c101011
|
3526 3527 |
if (list_empty(&plug->list)) return; |
109b81296
|
3528 |
list_splice_init(&plug->list, &list); |
422765c26
|
3529 |
list_sort(NULL, &list, plug_rq_cmp); |
73c101011
|
3530 3531 |
q = NULL; |
94b5eb28b
|
3532 |
depth = 0; |
188112722
|
3533 |
|
109b81296
|
3534 3535 |
while (!list_empty(&list)) { rq = list_entry_rq(list.next); |
73c101011
|
3536 |
list_del_init(&rq->queuelist); |
73c101011
|
3537 3538 |
BUG_ON(!rq->q); if (rq->q != q) { |
99e22598e
|
3539 3540 3541 3542 |
/* * This drops the queue lock */ if (q) |
49cac01e1
|
3543 |
queue_unplugged(q, depth, from_schedule); |
73c101011
|
3544 |
q = rq->q; |
94b5eb28b
|
3545 |
depth = 0; |
50864670b
|
3546 |
spin_lock_irq(q->queue_lock); |
73c101011
|
3547 |
} |
8ba61435d
|
3548 3549 3550 3551 |
/* * Short-circuit if @q is dead */ |
3f3299d5c
|
3552 |
if (unlikely(blk_queue_dying(q))) { |
2a842acab
|
3553 |
__blk_end_request_all(rq, BLK_STS_IOERR); |
8ba61435d
|
3554 3555 |
continue; } |
73c101011
|
3556 3557 3558 |
/* * rq is already accounted, so use raw insert */ |
f73f44eb0
|
3559 |
if (op_is_flush(rq->cmd_flags)) |
401a18e92
|
3560 3561 3562 |
__elv_add_request(q, rq, ELEVATOR_INSERT_FLUSH); else __elv_add_request(q, rq, ELEVATOR_INSERT_SORT_MERGE); |
94b5eb28b
|
3563 3564 |
depth++; |
73c101011
|
3565 |
} |
99e22598e
|
3566 3567 3568 3569 |
/* * This drops the queue lock */ if (q) |
49cac01e1
|
3570 |
queue_unplugged(q, depth, from_schedule); |
73c101011
|
3571 |
} |
73c101011
|
3572 3573 3574 |
void blk_finish_plug(struct blk_plug *plug) { |
dd6cf3e18
|
3575 3576 |
if (plug != current->plug) return; |
f6603783f
|
3577 |
blk_flush_plug_list(plug, false); |
73c101011
|
3578 |
|
dd6cf3e18
|
3579 |
current->plug = NULL; |
73c101011
|
3580 |
} |
88b996cd0
|
3581 |
EXPORT_SYMBOL(blk_finish_plug); |
73c101011
|
3582 |
|
47fafbc70
|
3583 |
#ifdef CONFIG_PM |
6c9546675
|
3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 |
/** * blk_pm_runtime_init - Block layer runtime PM initialization routine * @q: the queue of the device * @dev: the device the queue belongs to * * Description: * Initialize runtime-PM-related fields for @q and start auto suspend for * @dev. Drivers that want to take advantage of request-based runtime PM * should call this function after @dev has been initialized, and its * request queue @q has been allocated, and runtime PM for it can not happen * yet(either due to disabled/forbidden or its usage_count > 0). In most * cases, driver should call this function before any I/O has taken place. * * This function takes care of setting up using auto suspend for the device, * the autosuspend delay is set to -1 to make runtime suspend impossible * until an updated value is either set by user or by driver. Drivers do * not need to touch other autosuspend settings. * * The block layer runtime PM is request based, so only works for drivers * that use request as their IO unit instead of those directly use bio's. */ void blk_pm_runtime_init(struct request_queue *q, struct device *dev) { |
b233f1270
|
3607 3608 3609 |
/* Don't enable runtime PM for blk-mq until it is ready */ if (q->mq_ops) { pm_runtime_disable(dev); |
765e40b67
|
3610 |
return; |
b233f1270
|
3611 |
} |
765e40b67
|
3612 |
|
6c9546675
|
3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 |
q->dev = dev; q->rpm_status = RPM_ACTIVE; pm_runtime_set_autosuspend_delay(q->dev, -1); pm_runtime_use_autosuspend(q->dev); } EXPORT_SYMBOL(blk_pm_runtime_init); /** * blk_pre_runtime_suspend - Pre runtime suspend check * @q: the queue of the device * * Description: * This function will check if runtime suspend is allowed for the device * by examining if there are any requests pending in the queue. If there * are requests pending, the device can not be runtime suspended; otherwise, * the queue's status will be updated to SUSPENDING and the driver can * proceed to suspend the device. * * For the not allowed case, we mark last busy for the device so that * runtime PM core will try to autosuspend it some time later. * * This function should be called near the start of the device's * runtime_suspend callback. * * Return: * 0 - OK to runtime suspend the device * -EBUSY - Device should not be runtime suspended */ int blk_pre_runtime_suspend(struct request_queue *q) { int ret = 0; |
4fd41a855
|
3644 3645 |
if (!q->dev) return ret; |
6c9546675
|
3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 |
spin_lock_irq(q->queue_lock); if (q->nr_pending) { ret = -EBUSY; pm_runtime_mark_last_busy(q->dev); } else { q->rpm_status = RPM_SUSPENDING; } spin_unlock_irq(q->queue_lock); return ret; } EXPORT_SYMBOL(blk_pre_runtime_suspend); /** * blk_post_runtime_suspend - Post runtime suspend processing * @q: the queue of the device * @err: return value of the device's runtime_suspend function * * Description: * Update the queue's runtime status according to the return value of the * device's runtime suspend function and mark last busy for the device so * that PM core will try to auto suspend the device at a later time. * * This function should be called near the end of the device's * runtime_suspend callback. */ void blk_post_runtime_suspend(struct request_queue *q, int err) { |
4fd41a855
|
3673 3674 |
if (!q->dev) return; |
6c9546675
|
3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 |
spin_lock_irq(q->queue_lock); if (!err) { q->rpm_status = RPM_SUSPENDED; } else { q->rpm_status = RPM_ACTIVE; pm_runtime_mark_last_busy(q->dev); } spin_unlock_irq(q->queue_lock); } EXPORT_SYMBOL(blk_post_runtime_suspend); /** * blk_pre_runtime_resume - Pre runtime resume processing * @q: the queue of the device * * Description: * Update the queue's runtime status to RESUMING in preparation for the * runtime resume of the device. * * This function should be called near the start of the device's * runtime_resume callback. */ void blk_pre_runtime_resume(struct request_queue *q) { |
4fd41a855
|
3699 3700 |
if (!q->dev) return; |
6c9546675
|
3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 |
spin_lock_irq(q->queue_lock); q->rpm_status = RPM_RESUMING; spin_unlock_irq(q->queue_lock); } EXPORT_SYMBOL(blk_pre_runtime_resume); /** * blk_post_runtime_resume - Post runtime resume processing * @q: the queue of the device * @err: return value of the device's runtime_resume function * * Description: * Update the queue's runtime status according to the return value of the * device's runtime_resume function. If it is successfully resumed, process * the requests that are queued into the device's queue when it is resuming * and then mark last busy and initiate autosuspend for it. * * This function should be called near the end of the device's * runtime_resume callback. */ void blk_post_runtime_resume(struct request_queue *q, int err) { |
4fd41a855
|
3723 3724 |
if (!q->dev) return; |
6c9546675
|
3725 3726 3727 3728 3729 |
spin_lock_irq(q->queue_lock); if (!err) { q->rpm_status = RPM_ACTIVE; __blk_run_queue(q); pm_runtime_mark_last_busy(q->dev); |
c60855cdb
|
3730 |
pm_request_autosuspend(q->dev); |
6c9546675
|
3731 3732 3733 3734 3735 3736 |
} else { q->rpm_status = RPM_SUSPENDED; } spin_unlock_irq(q->queue_lock); } EXPORT_SYMBOL(blk_post_runtime_resume); |
d07ab6d11
|
3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 |
/** * blk_set_runtime_active - Force runtime status of the queue to be active * @q: the queue of the device * * If the device is left runtime suspended during system suspend the resume * hook typically resumes the device and corrects runtime status * accordingly. However, that does not affect the queue runtime PM status * which is still "suspended". This prevents processing requests from the * queue. * * This function can be used in driver's resume hook to correct queue * runtime PM status and re-enable peeking requests from the queue. It * should be called before first request is added to the queue. */ void blk_set_runtime_active(struct request_queue *q) { spin_lock_irq(q->queue_lock); q->rpm_status = RPM_ACTIVE; pm_runtime_mark_last_busy(q->dev); pm_request_autosuspend(q->dev); spin_unlock_irq(q->queue_lock); } EXPORT_SYMBOL(blk_set_runtime_active); |
6c9546675
|
3761 |
#endif |
1da177e4c
|
3762 3763 |
int __init blk_dev_init(void) { |
ef295ecf0
|
3764 3765 |
BUILD_BUG_ON(REQ_OP_LAST >= (1 << REQ_OP_BITS)); BUILD_BUG_ON(REQ_OP_BITS + REQ_FLAG_BITS > 8 * |
0762b23d2
|
3766 |
FIELD_SIZEOF(struct request, cmd_flags)); |
ef295ecf0
|
3767 3768 |
BUILD_BUG_ON(REQ_OP_BITS + REQ_FLAG_BITS > 8 * FIELD_SIZEOF(struct bio, bi_opf)); |
9eb55b030
|
3769 |
|
89b90be2d
|
3770 3771 |
/* used for unplugging and affects IO latency/throughput - HIGHPRI */ kblockd_workqueue = alloc_workqueue("kblockd", |
28747fcd2
|
3772 |
WQ_MEM_RECLAIM | WQ_HIGHPRI, 0); |
1da177e4c
|
3773 3774 3775 3776 3777 |
if (!kblockd_workqueue) panic("Failed to create kblockd "); request_cachep = kmem_cache_create("blkdev_requests", |
20c2df83d
|
3778 |
sizeof(struct request), 0, SLAB_PANIC, NULL); |
1da177e4c
|
3779 |
|
c2789bd40
|
3780 |
blk_requestq_cachep = kmem_cache_create("request_queue", |
165125e1e
|
3781 |
sizeof(struct request_queue), 0, SLAB_PANIC, NULL); |
1da177e4c
|
3782 |
|
18fbda91c
|
3783 3784 3785 |
#ifdef CONFIG_DEBUG_FS blk_debugfs_root = debugfs_create_dir("block", NULL); #endif |
d38ecf935
|
3786 |
return 0; |
1da177e4c
|
3787 |
} |