Blame view

block/blk-core.c 87.9 KB
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1
  /*
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
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   Jens Axboe   block: make core ...
6
7
   * kernel-doc documentation started by NeilBrown <neilb@cse.unsw.edu.au>
   *	-  July2000
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
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   Linus Torvalds   Linux-2.6.12-rc2
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   Jens Axboe   blk-mq: new multi...
19
  #include <linux/blk-mq.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
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   Linus Torvalds   Linux-2.6.12-rc2
25
26
27
28
  #include <linux/completion.h>
  #include <linux/slab.h>
  #include <linux/swap.h>
  #include <linux/writeback.h>
faccbd4b2   Andrew Morton   [PATCH] io-accoun...
29
  #include <linux/task_io_accounting_ops.h>
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
30
  #include <linux/fault-inject.h>
73c101011   Jens Axboe   block: initial pa...
31
  #include <linux/list_sort.h>
e3c78ca52   Tejun Heo   block: reorganize...
32
  #include <linux/delay.h>
aaf7c6806   Tejun Heo   block: fix elvpri...
33
  #include <linux/ratelimit.h>
6c9546675   Lin Ming   block: add runtim...
34
  #include <linux/pm_runtime.h>
55782138e   Li Zefan   tracing/events: c...
35
36
37
  
  #define CREATE_TRACE_POINTS
  #include <trace/events/block.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
38

8324aa91d   Jens Axboe   block: split tag ...
39
  #include "blk.h"
5efd61135   Tejun Heo   blkcg: add blkcg_...
40
  #include "blk-cgroup.h"
43a5e4e21   Ming Lei   block: blk-mq: su...
41
  #include "blk-mq.h"
8324aa91d   Jens Axboe   block: split tag ...
42

d07335e51   Mike Snitzer   block: Rename "bl...
43
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_remap);
b0da3f0da   Jun'ichi Nomura   Add a tracepoint ...
44
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_rq_remap);
0a82a8d13   Linus Torvalds   Revert "block: ad...
45
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_bio_complete);
cbae8d45d   NeilBrown   block: export blo...
46
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_unplug);
0bfc24559   Ingo Molnar   blktrace: port to...
47

a73f730d0   Tejun Heo   block, cfq: move ...
48
  DEFINE_IDA(blk_queue_ida);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
49
50
51
  /*
   * For the allocated request tables
   */
320ae51fe   Jens Axboe   blk-mq: new multi...
52
  struct kmem_cache *request_cachep = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
53
54
55
56
  
  /*
   * For queue allocation
   */
6728cb0e6   Jens Axboe   block: make core ...
57
  struct kmem_cache *blk_requestq_cachep;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
58
59
  
  /*
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
60
61
   * Controlling structure to kblockd
   */
ff856bad6   Jens Axboe   [BLOCK] ll_rw_blk...
62
  static struct workqueue_struct *kblockd_workqueue;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
63

8324aa91d   Jens Axboe   block: split tag ...
64
  void blk_queue_congestion_threshold(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
65
66
67
68
69
70
71
72
73
74
75
76
77
  {
  	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;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
78
79
80
81
82
83
84
85
86
87
88
89
  /**
   * blk_get_backing_dev_info - get the address of a queue's backing_dev_info
   * @bdev:	device
   *
   * Locates the passed device's request queue and returns the address of its
   * backing_dev_info
   *
   * Will return NULL if the request queue cannot be located.
   */
  struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev)
  {
  	struct backing_dev_info *ret = NULL;
165125e1e   Jens Axboe   [BLOCK] Get rid o...
90
  	struct request_queue *q = bdev_get_queue(bdev);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
91
92
93
94
95
  
  	if (q)
  		ret = &q->backing_dev_info;
  	return ret;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
96
  EXPORT_SYMBOL(blk_get_backing_dev_info);
2a4aa30c5   FUJITA Tomonori   block: rename and...
97
  void blk_rq_init(struct request_queue *q, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
98
  {
1afb20f30   FUJITA Tomonori   block: make rq_in...
99
  	memset(rq, 0, sizeof(*rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
100
  	INIT_LIST_HEAD(&rq->queuelist);
242f9dcb8   Jens Axboe   block: unify requ...
101
  	INIT_LIST_HEAD(&rq->timeout_list);
c7c22e4d5   Jens Axboe   block: add suppor...
102
  	rq->cpu = -1;
63a713867   Jens Axboe   block: fixup rq_i...
103
  	rq->q = q;
a2dec7b36   Tejun Heo   block: hide reque...
104
  	rq->__sector = (sector_t) -1;
2e662b65f   Jens Axboe   [PATCH] elevator:...
105
106
  	INIT_HLIST_NODE(&rq->hash);
  	RB_CLEAR_NODE(&rq->rb_node);
d7e3c3249   FUJITA Tomonori   block: add large ...
107
  	rq->cmd = rq->__cmd;
e2494e1b4   Li Zefan   blktrace: fix pdu...
108
  	rq->cmd_len = BLK_MAX_CDB;
63a713867   Jens Axboe   block: fixup rq_i...
109
  	rq->tag = -1;
b243ddcbe   Tejun Heo   block: move rq->s...
110
  	rq->start_time = jiffies;
9195291e5   Divyesh Shah   blkio: Increment ...
111
  	set_start_time_ns(rq);
09e099d4b   Jerome Marchand   block: fix accoun...
112
  	rq->part = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
113
  }
2a4aa30c5   FUJITA Tomonori   block: rename and...
114
  EXPORT_SYMBOL(blk_rq_init);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
115

5bb23a688   NeilBrown   Don't decrement b...
116
117
  static void req_bio_endio(struct request *rq, struct bio *bio,
  			  unsigned int nbytes, int error)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
118
  {
143a87f4c   Tejun Heo   block: improve fl...
119
120
121
122
  	if (error)
  		clear_bit(BIO_UPTODATE, &bio->bi_flags);
  	else if (!test_bit(BIO_UPTODATE, &bio->bi_flags))
  		error = -EIO;
797e7dbbe   Tejun Heo   [BLOCK] reimpleme...
123

143a87f4c   Tejun Heo   block: improve fl...
124
125
  	if (unlikely(rq->cmd_flags & REQ_QUIET))
  		set_bit(BIO_QUIET, &bio->bi_flags);
08bafc034   Keith Mannthey   block: Supress Bu...
126

f79ea4161   Kent Overstreet   block: Refactor b...
127
  	bio_advance(bio, nbytes);
7ba1ba12e   Martin K. Petersen   block: Block laye...
128

143a87f4c   Tejun Heo   block: improve fl...
129
  	/* don't actually finish bio if it's part of flush sequence */
4f024f379   Kent Overstreet   block: Abstract o...
130
  	if (bio->bi_iter.bi_size == 0 && !(rq->cmd_flags & REQ_FLUSH_SEQ))
143a87f4c   Tejun Heo   block: improve fl...
131
  		bio_endio(bio, error);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
132
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
133

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
134
135
136
  void blk_dump_rq_flags(struct request *rq, char *msg)
  {
  	int bit;
5953316db   Jens Axboe   block: make rq->c...
137
138
  	printk(KERN_INFO "%s: dev %s: type=%x, flags=%llx
  ", msg,
4aff5e233   Jens Axboe   [PATCH] Split str...
139
  		rq->rq_disk ? rq->rq_disk->disk_name : "?", rq->cmd_type,
5953316db   Jens Axboe   block: make rq->c...
140
  		(unsigned long long) rq->cmd_flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
141

83096ebf1   Tejun Heo   block: convert to...
142
143
144
145
  	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));
731ec497e   Tejun Heo   block: kill rq->data
146
147
  	printk(KERN_INFO "  bio %p, biotail %p, buffer %p, len %u
  ",
2e46e8b27   Tejun Heo   block: drop reque...
148
  	       rq->bio, rq->biotail, rq->buffer, blk_rq_bytes(rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
149

33659ebba   Christoph Hellwig   block: remove wra...
150
  	if (rq->cmd_type == REQ_TYPE_BLOCK_PC) {
6728cb0e6   Jens Axboe   block: make core ...
151
  		printk(KERN_INFO "  cdb: ");
d34c87e4b   FUJITA Tomonori   block: replace si...
152
  		for (bit = 0; bit < BLK_MAX_CDB; bit++)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
153
154
155
156
157
  			printk("%02x ", rq->cmd[bit]);
  		printk("
  ");
  	}
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
158
  EXPORT_SYMBOL(blk_dump_rq_flags);
3cca6dc1c   Jens Axboe   block: add API fo...
159
  static void blk_delay_work(struct work_struct *work)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
160
  {
3cca6dc1c   Jens Axboe   block: add API fo...
161
  	struct request_queue *q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
162

3cca6dc1c   Jens Axboe   block: add API fo...
163
164
  	q = container_of(work, struct request_queue, delay_work.work);
  	spin_lock_irq(q->queue_lock);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
165
  	__blk_run_queue(q);
3cca6dc1c   Jens Axboe   block: add API fo...
166
  	spin_unlock_irq(q->queue_lock);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
167
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
168
169
  
  /**
3cca6dc1c   Jens Axboe   block: add API fo...
170
171
172
   * blk_delay_queue - restart queueing after defined interval
   * @q:		The &struct request_queue in question
   * @msecs:	Delay in msecs
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
173
174
   *
   * Description:
3cca6dc1c   Jens Axboe   block: add API fo...
175
176
   *   Sometimes queueing needs to be postponed for a little while, to allow
   *   resources to come back. This function will make sure that queueing is
704605711   Bart Van Assche   block: Avoid sche...
177
   *   restarted around the specified time. Queue lock must be held.
3cca6dc1c   Jens Axboe   block: add API fo...
178
179
   */
  void blk_delay_queue(struct request_queue *q, unsigned long msecs)
2ad8b1ef1   Alan D. Brunelle   Add UNPLUG traces...
180
  {
704605711   Bart Van Assche   block: Avoid sche...
181
182
183
  	if (likely(!blk_queue_dead(q)))
  		queue_delayed_work(kblockd_workqueue, &q->delay_work,
  				   msecs_to_jiffies(msecs));
2ad8b1ef1   Alan D. Brunelle   Add UNPLUG traces...
184
  }
3cca6dc1c   Jens Axboe   block: add API fo...
185
  EXPORT_SYMBOL(blk_delay_queue);
2ad8b1ef1   Alan D. Brunelle   Add UNPLUG traces...
186

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
187
188
  /**
   * blk_start_queue - restart a previously stopped queue
165125e1e   Jens Axboe   [BLOCK] Get rid o...
189
   * @q:    The &struct request_queue in question
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
190
191
192
193
194
195
   *
   * 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
   *   entered. Also see blk_stop_queue(). Queue lock must be held.
   **/
165125e1e   Jens Axboe   [BLOCK] Get rid o...
196
  void blk_start_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
197
  {
a038e2536   Paolo 'Blaisorblade' Giarrusso   [PATCH] blk_start...
198
  	WARN_ON(!irqs_disabled());
75ad23bc0   Nick Piggin   block: make queue...
199
  	queue_flag_clear(QUEUE_FLAG_STOPPED, q);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
200
  	__blk_run_queue(q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
201
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
202
203
204
205
  EXPORT_SYMBOL(blk_start_queue);
  
  /**
   * blk_stop_queue - stop a queue
165125e1e   Jens Axboe   [BLOCK] Get rid o...
206
   * @q:    The &struct request_queue in question
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
207
208
209
210
211
212
213
214
215
216
217
   *
   * 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
   *   blk_start_queue() to restart queue operations. Queue lock must be held.
   **/
165125e1e   Jens Axboe   [BLOCK] Get rid o...
218
  void blk_stop_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
219
  {
136b5721d   Tejun Heo   workqueue: deprec...
220
  	cancel_delayed_work(&q->delay_work);
75ad23bc0   Nick Piggin   block: make queue...
221
  	queue_flag_set(QUEUE_FLAG_STOPPED, q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
222
223
224
225
226
227
228
229
230
231
232
233
  }
  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   Michael Opdenacker   Fix occurrences o...
234
   *     that the callbacks might use. The caller must already have made sure
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
235
236
237
   *     that its ->make_request_fn will not re-add plugging prior to calling
   *     this function.
   *
da5277700   Vivek Goyal   block: Move blk_t...
238
239
   *     This function does not cancel any asynchronous activity arising
   *     out of elevator or throttling code. That would require elevaotor_exit()
5efd61135   Tejun Heo   blkcg: add blkcg_...
240
   *     and blkcg_exit_queue() to be called with queue lock initialized.
da5277700   Vivek Goyal   block: Move blk_t...
241
   *
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
242
243
244
   */
  void blk_sync_queue(struct request_queue *q)
  {
70ed28b92   Jens Axboe   block: leave the ...
245
  	del_timer_sync(&q->timeout);
f04c1fe76   Ming Lei   block: blk-mq: ma...
246
247
248
249
250
251
252
253
254
255
  
  	if (q->mq_ops) {
  		struct blk_mq_hw_ctx *hctx;
  		int i;
  
  		queue_for_each_hw_ctx(q, hctx, i)
  			cancel_delayed_work_sync(&hctx->delayed_work);
  	} else {
  		cancel_delayed_work_sync(&q->delay_work);
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
256
257
258
259
  }
  EXPORT_SYMBOL(blk_sync_queue);
  
  /**
c246e80d8   Bart Van Assche   block: Avoid that...
260
261
262
263
264
265
266
267
268
269
270
271
272
273
   * __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)
  {
  	if (unlikely(blk_queue_dead(q)))
  		return;
24faf6f60   Bart Van Assche   block: Make blk_c...
274
275
276
277
278
279
280
281
  	/*
  	 * 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   Bart Van Assche   block: Avoid that...
282
  	q->request_fn(q);
24faf6f60   Bart Van Assche   block: Make blk_c...
283
  	q->request_fn_active--;
c246e80d8   Bart Van Assche   block: Avoid that...
284
285
286
  }
  
  /**
80a4b58e3   Jens Axboe   block: only call ...
287
   * __blk_run_queue - run a single device queue
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
288
   * @q:	The queue to run
80a4b58e3   Jens Axboe   block: only call ...
289
290
291
   *
   * Description:
   *    See @blk_run_queue. This variant must be called with the queue lock
24ecfbe27   Christoph Hellwig   block: add blk_ru...
292
   *    held and interrupts disabled.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
293
   */
24ecfbe27   Christoph Hellwig   block: add blk_ru...
294
  void __blk_run_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
295
  {
a538cd03b   Tejun Heo   block: merge blk_...
296
297
  	if (unlikely(blk_queue_stopped(q)))
  		return;
c246e80d8   Bart Van Assche   block: Avoid that...
298
  	__blk_run_queue_uncond(q);
75ad23bc0   Nick Piggin   block: make queue...
299
300
  }
  EXPORT_SYMBOL(__blk_run_queue);
dac07ec12   Jens Axboe   [BLOCK] limit req...
301

75ad23bc0   Nick Piggin   block: make queue...
302
  /**
24ecfbe27   Christoph Hellwig   block: add blk_ru...
303
304
305
306
307
   * 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
704605711   Bart Van Assche   block: Avoid sche...
308
   *    of us. The caller must hold the queue lock.
24ecfbe27   Christoph Hellwig   block: add blk_ru...
309
310
311
   */
  void blk_run_queue_async(struct request_queue *q)
  {
704605711   Bart Van Assche   block: Avoid sche...
312
  	if (likely(!blk_queue_stopped(q) && !blk_queue_dead(q)))
e7c2f9674   Tejun Heo   workqueue: use mo...
313
  		mod_delayed_work(kblockd_workqueue, &q->delay_work, 0);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
314
  }
c21e6beba   Jens Axboe   block: get rid of...
315
  EXPORT_SYMBOL(blk_run_queue_async);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
316
317
  
  /**
75ad23bc0   Nick Piggin   block: make queue...
318
319
   * blk_run_queue - run a single device queue
   * @q: The queue to run
80a4b58e3   Jens Axboe   block: only call ...
320
321
322
   *
   * Description:
   *    Invoke request handling on this queue, if it has pending work to do.
a7f557923   Tejun Heo   block: kill blk_s...
323
   *    May be used to restart queueing when a request has completed.
75ad23bc0   Nick Piggin   block: make queue...
324
325
326
327
328
329
   */
  void blk_run_queue(struct request_queue *q)
  {
  	unsigned long flags;
  
  	spin_lock_irqsave(q->queue_lock, flags);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
330
  	__blk_run_queue(q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
331
332
333
  	spin_unlock_irqrestore(q->queue_lock, flags);
  }
  EXPORT_SYMBOL(blk_run_queue);
165125e1e   Jens Axboe   [BLOCK] Get rid o...
334
  void blk_put_queue(struct request_queue *q)
483f4afc4   Al Viro   [PATCH] fix sysfs...
335
336
337
  {
  	kobject_put(&q->kobj);
  }
d86e0e83b   Jens Axboe   block: export blk...
338
  EXPORT_SYMBOL(blk_put_queue);
483f4afc4   Al Viro   [PATCH] fix sysfs...
339

e3c78ca52   Tejun Heo   block: reorganize...
340
  /**
807592a4f   Bart Van Assche   block: Let blk_dr...
341
   * __blk_drain_queue - drain requests from request_queue
e3c78ca52   Tejun Heo   block: reorganize...
342
   * @q: queue to drain
c9a929dde   Tejun Heo   block: fix reques...
343
   * @drain_all: whether to drain all requests or only the ones w/ ELVPRIV
e3c78ca52   Tejun Heo   block: reorganize...
344
   *
c9a929dde   Tejun Heo   block: fix reques...
345
346
347
   * 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   Tejun Heo   block: reorganize...
348
   */
807592a4f   Bart Van Assche   block: Let blk_dr...
349
350
351
  static void __blk_drain_queue(struct request_queue *q, bool drain_all)
  	__releases(q->queue_lock)
  	__acquires(q->queue_lock)
e3c78ca52   Tejun Heo   block: reorganize...
352
  {
458f27a98   Asias He   block: Avoid miss...
353
  	int i;
807592a4f   Bart Van Assche   block: Let blk_dr...
354
  	lockdep_assert_held(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
355
  	while (true) {
481a7d647   Tejun Heo   block: fix drain_...
356
  		bool drain = false;
e3c78ca52   Tejun Heo   block: reorganize...
357

b855b04a0   Tejun Heo   block: blk-thrott...
358
359
360
361
362
363
  		/*
  		 * The caller might be trying to drain @q before its
  		 * elevator is initialized.
  		 */
  		if (q->elevator)
  			elv_drain_elevator(q);
5efd61135   Tejun Heo   blkcg: add blkcg_...
364
  		blkcg_drain_queue(q);
e3c78ca52   Tejun Heo   block: reorganize...
365

4eabc9412   Tejun Heo   block: don't kick...
366
367
  		/*
  		 * This function might be called on a queue which failed
b855b04a0   Tejun Heo   block: blk-thrott...
368
369
370
371
  		 * 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   Tejun Heo   block: don't kick...
372
  		 */
b855b04a0   Tejun Heo   block: blk-thrott...
373
  		if (!list_empty(&q->queue_head) && q->request_fn)
4eabc9412   Tejun Heo   block: don't kick...
374
  			__blk_run_queue(q);
c9a929dde   Tejun Heo   block: fix reques...
375

8a5ecdd42   Tejun Heo   block: add q->nr_...
376
  		drain |= q->nr_rqs_elvpriv;
24faf6f60   Bart Van Assche   block: Make blk_c...
377
  		drain |= q->request_fn_active;
481a7d647   Tejun Heo   block: fix drain_...
378
379
380
381
382
383
384
385
386
  
  		/*
  		 * 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) {
  			drain |= !list_empty(&q->queue_head);
  			for (i = 0; i < 2; i++) {
8a5ecdd42   Tejun Heo   block: add q->nr_...
387
  				drain |= q->nr_rqs[i];
481a7d647   Tejun Heo   block: fix drain_...
388
389
390
391
  				drain |= q->in_flight[i];
  				drain |= !list_empty(&q->flush_queue[i]);
  			}
  		}
e3c78ca52   Tejun Heo   block: reorganize...
392

481a7d647   Tejun Heo   block: fix drain_...
393
  		if (!drain)
e3c78ca52   Tejun Heo   block: reorganize...
394
  			break;
807592a4f   Bart Van Assche   block: Let blk_dr...
395
396
  
  		spin_unlock_irq(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
397
  		msleep(10);
807592a4f   Bart Van Assche   block: Let blk_dr...
398
399
  
  		spin_lock_irq(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
400
  	}
458f27a98   Asias He   block: Avoid miss...
401
402
403
404
405
406
407
  
  	/*
  	 * 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   Tejun Heo   blkcg: implement ...
408
  		struct request_list *rl;
a051661ca   Tejun Heo   blkcg: implement ...
409
410
411
  		blk_queue_for_each_rl(rl, q)
  			for (i = 0; i < ARRAY_SIZE(rl->wait); i++)
  				wake_up_all(&rl->wait[i]);
458f27a98   Asias He   block: Avoid miss...
412
  	}
e3c78ca52   Tejun Heo   block: reorganize...
413
  }
c9a929dde   Tejun Heo   block: fix reques...
414
  /**
d732580b4   Tejun Heo   block: implement ...
415
416
417
418
419
   * 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   Tejun Heo   block: extend que...
420
   * throttled or issued before.  On return, it's guaranteed that no request
80fd99792   Tejun Heo   blkcg: make sure ...
421
422
   * is being throttled or has ELVPRIV set and blk_queue_bypass() %true
   * inside queue or RCU read lock.
d732580b4   Tejun Heo   block: implement ...
423
424
425
   */
  void blk_queue_bypass_start(struct request_queue *q)
  {
b82d4b197   Tejun Heo   blkcg: make reque...
426
  	bool drain;
d732580b4   Tejun Heo   block: implement ...
427
  	spin_lock_irq(q->queue_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
428
  	drain = !q->bypass_depth++;
d732580b4   Tejun Heo   block: implement ...
429
430
  	queue_flag_set(QUEUE_FLAG_BYPASS, q);
  	spin_unlock_irq(q->queue_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
431
  	if (drain) {
807592a4f   Bart Van Assche   block: Let blk_dr...
432
433
434
  		spin_lock_irq(q->queue_lock);
  		__blk_drain_queue(q, false);
  		spin_unlock_irq(q->queue_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
435
436
437
  		/* ensure blk_queue_bypass() is %true inside RCU read lock */
  		synchronize_rcu();
  	}
d732580b4   Tejun Heo   block: implement ...
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
  }
  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.
   */
  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);
  
  /**
c9a929dde   Tejun Heo   block: fix reques...
458
459
460
   * blk_cleanup_queue - shutdown a request queue
   * @q: request queue to shutdown
   *
c246e80d8   Bart Van Assche   block: Avoid that...
461
462
   * Mark @q DYING, drain all pending requests, mark @q DEAD, destroy and
   * put it.  All future requests will be failed immediately with -ENODEV.
c94a96ac9   Vivek Goyal   block: Initialize...
463
   */
6728cb0e6   Jens Axboe   block: make core ...
464
  void blk_cleanup_queue(struct request_queue *q)
483f4afc4   Al Viro   [PATCH] fix sysfs...
465
  {
c9a929dde   Tejun Heo   block: fix reques...
466
  	spinlock_t *lock = q->queue_lock;
e3335de94   Jens Axboe   block: blk_cleanu...
467

3f3299d5c   Bart Van Assche   block: Rename que...
468
  	/* mark @q DYING, no new request or merges will be allowed afterwards */
483f4afc4   Al Viro   [PATCH] fix sysfs...
469
  	mutex_lock(&q->sysfs_lock);
3f3299d5c   Bart Van Assche   block: Rename que...
470
  	queue_flag_set_unlocked(QUEUE_FLAG_DYING, q);
c9a929dde   Tejun Heo   block: fix reques...
471
  	spin_lock_irq(lock);
6ecf23afa   Tejun Heo   block: extend que...
472

80fd99792   Tejun Heo   blkcg: make sure ...
473
  	/*
3f3299d5c   Bart Van Assche   block: Rename que...
474
  	 * A dying queue is permanently in bypass mode till released.  Note
80fd99792   Tejun Heo   blkcg: make sure ...
475
476
477
478
479
480
481
  	 * 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   Tejun Heo   block: extend que...
482
483
  	q->bypass_depth++;
  	queue_flag_set(QUEUE_FLAG_BYPASS, q);
c9a929dde   Tejun Heo   block: fix reques...
484
485
  	queue_flag_set(QUEUE_FLAG_NOMERGES, q);
  	queue_flag_set(QUEUE_FLAG_NOXMERGES, q);
3f3299d5c   Bart Van Assche   block: Rename que...
486
  	queue_flag_set(QUEUE_FLAG_DYING, q);
c9a929dde   Tejun Heo   block: fix reques...
487
488
  	spin_unlock_irq(lock);
  	mutex_unlock(&q->sysfs_lock);
c246e80d8   Bart Van Assche   block: Avoid that...
489
490
491
492
  	/*
  	 * Drain all requests queued before DYING marking. Set DEAD flag to
  	 * prevent that q->request_fn() gets invoked after draining finished.
  	 */
43a5e4e21   Ming Lei   block: blk-mq: su...
493
494
495
496
497
498
499
  	if (q->mq_ops) {
  		blk_mq_drain_queue(q);
  		spin_lock_irq(lock);
  	} else {
  		spin_lock_irq(lock);
  		__blk_drain_queue(q, true);
  	}
c246e80d8   Bart Van Assche   block: Avoid that...
500
  	queue_flag_set(QUEUE_FLAG_DEAD, q);
807592a4f   Bart Van Assche   block: Let blk_dr...
501
  	spin_unlock_irq(lock);
c9a929dde   Tejun Heo   block: fix reques...
502
503
504
505
  
  	/* @q won't process any more request, flush async actions */
  	del_timer_sync(&q->backing_dev_info.laptop_mode_wb_timer);
  	blk_sync_queue(q);
5e5cfac0c   Asias He   block: Mitigate l...
506
507
508
509
  	spin_lock_irq(lock);
  	if (q->queue_lock != &q->__queue_lock)
  		q->queue_lock = &q->__queue_lock;
  	spin_unlock_irq(lock);
c9a929dde   Tejun Heo   block: fix reques...
510
  	/* @q is and will stay empty, shutdown and put */
483f4afc4   Al Viro   [PATCH] fix sysfs...
511
512
  	blk_put_queue(q);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
513
  EXPORT_SYMBOL(blk_cleanup_queue);
5b788ce3e   Tejun Heo   block: prepare fo...
514
515
  int blk_init_rl(struct request_list *rl, struct request_queue *q,
  		gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
516
  {
1abec4fdb   Mike Snitzer   block: make blk_i...
517
518
  	if (unlikely(rl->rq_pool))
  		return 0;
5b788ce3e   Tejun Heo   block: prepare fo...
519
  	rl->q = q;
1faa16d22   Jens Axboe   block: change the...
520
521
  	rl->count[BLK_RW_SYNC] = rl->count[BLK_RW_ASYNC] = 0;
  	rl->starved[BLK_RW_SYNC] = rl->starved[BLK_RW_ASYNC] = 0;
1faa16d22   Jens Axboe   block: change the...
522
523
  	init_waitqueue_head(&rl->wait[BLK_RW_SYNC]);
  	init_waitqueue_head(&rl->wait[BLK_RW_ASYNC]);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
524

1946089a1   Christoph Lameter   [PATCH] NUMA awar...
525
  	rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, mempool_alloc_slab,
a91a5ac68   Tejun Heo   mempool: add @gfp...
526
  					  mempool_free_slab, request_cachep,
5b788ce3e   Tejun Heo   block: prepare fo...
527
  					  gfp_mask, q->node);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
528
529
530
531
532
  	if (!rl->rq_pool)
  		return -ENOMEM;
  
  	return 0;
  }
5b788ce3e   Tejun Heo   block: prepare fo...
533
534
535
536
537
  void blk_exit_rl(struct request_list *rl)
  {
  	if (rl->rq_pool)
  		mempool_destroy(rl->rq_pool);
  }
165125e1e   Jens Axboe   [BLOCK] Get rid o...
538
  struct request_queue *blk_alloc_queue(gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
539
  {
c304a51bf   Ezequiel Garcia   block: use NUMA_N...
540
  	return blk_alloc_queue_node(gfp_mask, NUMA_NO_NODE);
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
541
542
  }
  EXPORT_SYMBOL(blk_alloc_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
543

165125e1e   Jens Axboe   [BLOCK] Get rid o...
544
  struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id)
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
545
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
546
  	struct request_queue *q;
e0bf68dde   Peter Zijlstra   mm: bdi init hooks
547
  	int err;
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
548

8324aa91d   Jens Axboe   block: split tag ...
549
  	q = kmem_cache_alloc_node(blk_requestq_cachep,
94f6030ca   Christoph Lameter   Slab allocators: ...
550
  				gfp_mask | __GFP_ZERO, node_id);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
551
552
  	if (!q)
  		return NULL;
320ae51fe   Jens Axboe   blk-mq: new multi...
553
554
  	if (percpu_counter_init(&q->mq_usage_counter, 0))
  		goto fail_q;
00380a404   Dan Carpenter   block: blk_alloc_...
555
  	q->id = ida_simple_get(&blk_queue_ida, 0, 0, gfp_mask);
a73f730d0   Tejun Heo   block, cfq: move ...
556
  	if (q->id < 0)
320ae51fe   Jens Axboe   blk-mq: new multi...
557
  		goto fail_c;
a73f730d0   Tejun Heo   block, cfq: move ...
558

0989a025d   Jens Axboe   block: don't over...
559
560
561
562
  	q->backing_dev_info.ra_pages =
  			(VM_MAX_READAHEAD * 1024) / PAGE_CACHE_SIZE;
  	q->backing_dev_info.state = 0;
  	q->backing_dev_info.capabilities = BDI_CAP_MAP_COPY;
d993831fa   Jens Axboe   writeback: add na...
563
  	q->backing_dev_info.name = "block";
5151412dd   Mike Snitzer   block: initialize...
564
  	q->node = node_id;
0989a025d   Jens Axboe   block: don't over...
565

e0bf68dde   Peter Zijlstra   mm: bdi init hooks
566
  	err = bdi_init(&q->backing_dev_info);
a73f730d0   Tejun Heo   block, cfq: move ...
567
568
  	if (err)
  		goto fail_id;
e0bf68dde   Peter Zijlstra   mm: bdi init hooks
569

31373d09d   Matthew Garrett   laptop-mode: Make...
570
571
  	setup_timer(&q->backing_dev_info.laptop_mode_wb_timer,
  		    laptop_mode_timer_fn, (unsigned long) q);
242f9dcb8   Jens Axboe   block: unify requ...
572
  	setup_timer(&q->timeout, blk_rq_timed_out_timer, (unsigned long) q);
b855b04a0   Tejun Heo   block: blk-thrott...
573
  	INIT_LIST_HEAD(&q->queue_head);
242f9dcb8   Jens Axboe   block: unify requ...
574
  	INIT_LIST_HEAD(&q->timeout_list);
a612fddf0   Tejun Heo   block, cfq: move ...
575
  	INIT_LIST_HEAD(&q->icq_list);
4eef30499   Tejun Heo   blkcg: move per-q...
576
  #ifdef CONFIG_BLK_CGROUP
e8989fae3   Tejun Heo   blkcg: unify blkg...
577
  	INIT_LIST_HEAD(&q->blkg_list);
4eef30499   Tejun Heo   blkcg: move per-q...
578
  #endif
ae1b15396   Tejun Heo   block: reimplemen...
579
580
581
  	INIT_LIST_HEAD(&q->flush_queue[0]);
  	INIT_LIST_HEAD(&q->flush_queue[1]);
  	INIT_LIST_HEAD(&q->flush_data_in_flight);
3cca6dc1c   Jens Axboe   block: add API fo...
582
  	INIT_DELAYED_WORK(&q->delay_work, blk_delay_work);
483f4afc4   Al Viro   [PATCH] fix sysfs...
583

8324aa91d   Jens Axboe   block: split tag ...
584
  	kobject_init(&q->kobj, &blk_queue_ktype);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
585

483f4afc4   Al Viro   [PATCH] fix sysfs...
586
  	mutex_init(&q->sysfs_lock);
e7e72bf64   Neil Brown   Remove blkdev war...
587
  	spin_lock_init(&q->__queue_lock);
483f4afc4   Al Viro   [PATCH] fix sysfs...
588

c94a96ac9   Vivek Goyal   block: Initialize...
589
590
591
592
593
  	/*
  	 * By default initialize queue_lock to internal lock and driver can
  	 * override it later if need be.
  	 */
  	q->queue_lock = &q->__queue_lock;
b82d4b197   Tejun Heo   blkcg: make reque...
594
595
596
  	/*
  	 * A queue starts its life with bypass turned on to avoid
  	 * unnecessary bypass on/off overhead and nasty surprises during
749fefe67   Tejun Heo   block: lift the i...
597
598
  	 * init.  The initial bypass will be finished when the queue is
  	 * registered by blk_register_queue().
b82d4b197   Tejun Heo   blkcg: make reque...
599
600
601
  	 */
  	q->bypass_depth = 1;
  	__set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags);
320ae51fe   Jens Axboe   blk-mq: new multi...
602
  	init_waitqueue_head(&q->mq_freeze_wq);
5efd61135   Tejun Heo   blkcg: add blkcg_...
603
  	if (blkcg_init_queue(q))
fff4996b7   Mikulas Patocka   blk-core: Fix mem...
604
  		goto fail_bdi;
f51b802c1   Tejun Heo   blkcg: use the us...
605

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
606
  	return q;
a73f730d0   Tejun Heo   block, cfq: move ...
607

fff4996b7   Mikulas Patocka   blk-core: Fix mem...
608
609
  fail_bdi:
  	bdi_destroy(&q->backing_dev_info);
a73f730d0   Tejun Heo   block, cfq: move ...
610
611
  fail_id:
  	ida_simple_remove(&blk_queue_ida, q->id);
320ae51fe   Jens Axboe   blk-mq: new multi...
612
613
  fail_c:
  	percpu_counter_destroy(&q->mq_usage_counter);
a73f730d0   Tejun Heo   block, cfq: move ...
614
615
616
  fail_q:
  	kmem_cache_free(blk_requestq_cachep, q);
  	return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
617
  }
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
618
  EXPORT_SYMBOL(blk_alloc_queue_node);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
  
  /**
   * 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   Paolo 'Blaisorblade' Giarrusso   [PATCH] blk_start...
642
643
   *    request queue; this lock will be taken also from interrupt context, so irq
   *    disabling is needed for it.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
644
   *
710027a48   Randy Dunlap   Add some block/ s...
645
   *    Function returns a pointer to the initialized request queue, or %NULL if
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
646
647
648
649
650
651
   *    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   Christoph Lameter   [PATCH] NUMA awar...
652

165125e1e   Jens Axboe   [BLOCK] Get rid o...
653
  struct request_queue *blk_init_queue(request_fn_proc *rfn, spinlock_t *lock)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
654
  {
c304a51bf   Ezequiel Garcia   block: use NUMA_N...
655
  	return blk_init_queue_node(rfn, lock, NUMA_NO_NODE);
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
656
657
  }
  EXPORT_SYMBOL(blk_init_queue);
165125e1e   Jens Axboe   [BLOCK] Get rid o...
658
  struct request_queue *
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
659
660
  blk_init_queue_node(request_fn_proc *rfn, spinlock_t *lock, int node_id)
  {
c86d1b8ae   Mike Snitzer   block: avoid unco...
661
  	struct request_queue *uninit_q, *q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
662

c86d1b8ae   Mike Snitzer   block: avoid unco...
663
664
665
  	uninit_q = blk_alloc_queue_node(GFP_KERNEL, node_id);
  	if (!uninit_q)
  		return NULL;
18741986a   Christoph Hellwig   blk-mq: rework fl...
666
667
668
  	uninit_q->flush_rq = kzalloc(sizeof(struct request), GFP_KERNEL);
  	if (!uninit_q->flush_rq)
  		goto out_cleanup_queue;
5151412dd   Mike Snitzer   block: initialize...
669
  	q = blk_init_allocated_queue(uninit_q, rfn, lock);
c86d1b8ae   Mike Snitzer   block: avoid unco...
670
  	if (!q)
18741986a   Christoph Hellwig   blk-mq: rework fl...
671
  		goto out_free_flush_rq;
c86d1b8ae   Mike Snitzer   block: avoid unco...
672
  	return q;
18741986a   Christoph Hellwig   blk-mq: rework fl...
673
674
675
676
677
678
  
  out_free_flush_rq:
  	kfree(uninit_q->flush_rq);
  out_cleanup_queue:
  	blk_cleanup_queue(uninit_q);
  	return NULL;
01effb0dc   Mike Snitzer   block: allow init...
679
680
681
682
683
684
685
  }
  EXPORT_SYMBOL(blk_init_queue_node);
  
  struct request_queue *
  blk_init_allocated_queue(struct request_queue *q, request_fn_proc *rfn,
  			 spinlock_t *lock)
  {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
686
687
  	if (!q)
  		return NULL;
a051661ca   Tejun Heo   blkcg: implement ...
688
  	if (blk_init_rl(&q->root_rl, q, GFP_KERNEL))
8669aafdb   Al Viro   [PATCH] fix doubl...
689
  		return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
690
691
  
  	q->request_fn		= rfn;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
692
  	q->prep_rq_fn		= NULL;
28018c242   James Bottomley   block: implement ...
693
  	q->unprep_rq_fn		= NULL;
60ea8226c   Tejun Heo   block: fix reques...
694
  	q->queue_flags		|= QUEUE_FLAG_DEFAULT;
c94a96ac9   Vivek Goyal   block: Initialize...
695
696
697
698
  
  	/* Override internal queue lock with supplied lock pointer */
  	if (lock)
  		q->queue_lock		= lock;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
699

f3b144aa7   Jens Axboe   block: remove var...
700
701
702
  	/*
  	 * This also sets hw/phys segments, boundary and size
  	 */
c20e8de27   Jens Axboe   block: rename __m...
703
  	blk_queue_make_request(q, blk_queue_bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
704

44ec95425   Alan Stern   [SCSI] sg: cap re...
705
  	q->sg_reserved_size = INT_MAX;
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
706
707
  	/* Protect q->elevator from elevator_change */
  	mutex_lock(&q->sysfs_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
708
  	/* init elevator */
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
709
710
  	if (elevator_init(q, NULL)) {
  		mutex_unlock(&q->sysfs_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
711
  		return NULL;
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
712
713
714
  	}
  
  	mutex_unlock(&q->sysfs_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
715
  	return q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
716
  }
5151412dd   Mike Snitzer   block: initialize...
717
  EXPORT_SYMBOL(blk_init_allocated_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
718

09ac46c42   Tejun Heo   block: misc updat...
719
  bool blk_get_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
720
  {
3f3299d5c   Bart Van Assche   block: Rename que...
721
  	if (likely(!blk_queue_dying(q))) {
09ac46c42   Tejun Heo   block: misc updat...
722
723
  		__blk_get_queue(q);
  		return true;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
724
  	}
09ac46c42   Tejun Heo   block: misc updat...
725
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
726
  }
d86e0e83b   Jens Axboe   block: export blk...
727
  EXPORT_SYMBOL(blk_get_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
728

5b788ce3e   Tejun Heo   block: prepare fo...
729
  static inline void blk_free_request(struct request_list *rl, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
730
  {
f1f8cc946   Tejun Heo   block, cfq: move ...
731
  	if (rq->cmd_flags & REQ_ELVPRIV) {
5b788ce3e   Tejun Heo   block: prepare fo...
732
  		elv_put_request(rl->q, rq);
f1f8cc946   Tejun Heo   block, cfq: move ...
733
  		if (rq->elv.icq)
11a3122f6   Tejun Heo   block: strip out ...
734
  			put_io_context(rq->elv.icq->ioc);
f1f8cc946   Tejun Heo   block, cfq: move ...
735
  	}
5b788ce3e   Tejun Heo   block: prepare fo...
736
  	mempool_free(rq, rl->rq_pool);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
737
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
738
739
740
741
  /*
   * ioc_batching returns true if the ioc is a valid batching request and
   * should be given priority access to a request.
   */
165125e1e   Jens Axboe   [BLOCK] Get rid o...
742
  static inline int ioc_batching(struct request_queue *q, struct io_context *ioc)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
  {
  	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   Jens Axboe   [BLOCK] Get rid o...
763
  static void ioc_set_batching(struct request_queue *q, struct io_context *ioc)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
764
765
766
767
768
769
770
  {
  	if (!ioc || ioc_batching(q, ioc))
  		return;
  
  	ioc->nr_batch_requests = q->nr_batching;
  	ioc->last_waited = jiffies;
  }
5b788ce3e   Tejun Heo   block: prepare fo...
771
  static void __freed_request(struct request_list *rl, int sync)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
772
  {
5b788ce3e   Tejun Heo   block: prepare fo...
773
  	struct request_queue *q = rl->q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
774

a051661ca   Tejun Heo   blkcg: implement ...
775
776
777
778
779
780
  	/*
  	 * bdi isn't aware of blkcg yet.  As all async IOs end up root
  	 * blkcg anyway, just use root blkcg state.
  	 */
  	if (rl == &q->root_rl &&
  	    rl->count[sync] < queue_congestion_off_threshold(q))
1faa16d22   Jens Axboe   block: change the...
781
  		blk_clear_queue_congested(q, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
782

1faa16d22   Jens Axboe   block: change the...
783
784
785
  	if (rl->count[sync] + 1 <= q->nr_requests) {
  		if (waitqueue_active(&rl->wait[sync]))
  			wake_up(&rl->wait[sync]);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
786

5b788ce3e   Tejun Heo   block: prepare fo...
787
  		blk_clear_rl_full(rl, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
788
789
790
791
792
793
794
  	}
  }
  
  /*
   * A request has just been released.  Account for it, update the full and
   * congestion status, wake up any waiters.   Called under q->queue_lock.
   */
5b788ce3e   Tejun Heo   block: prepare fo...
795
  static void freed_request(struct request_list *rl, unsigned int flags)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
796
  {
5b788ce3e   Tejun Heo   block: prepare fo...
797
  	struct request_queue *q = rl->q;
75eb6c372   Tejun Heo   block: pass aroun...
798
  	int sync = rw_is_sync(flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
799

8a5ecdd42   Tejun Heo   block: add q->nr_...
800
  	q->nr_rqs[sync]--;
1faa16d22   Jens Axboe   block: change the...
801
  	rl->count[sync]--;
75eb6c372   Tejun Heo   block: pass aroun...
802
  	if (flags & REQ_ELVPRIV)
8a5ecdd42   Tejun Heo   block: add q->nr_...
803
  		q->nr_rqs_elvpriv--;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
804

5b788ce3e   Tejun Heo   block: prepare fo...
805
  	__freed_request(rl, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
806

1faa16d22   Jens Axboe   block: change the...
807
  	if (unlikely(rl->starved[sync ^ 1]))
5b788ce3e   Tejun Heo   block: prepare fo...
808
  		__freed_request(rl, sync ^ 1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
809
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
810
  /*
9d5a4e946   Mike Snitzer   block: skip eleva...
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
   * Determine if elevator data should be initialized when allocating the
   * request associated with @bio.
   */
  static bool blk_rq_should_init_elevator(struct bio *bio)
  {
  	if (!bio)
  		return true;
  
  	/*
  	 * Flush requests do not use the elevator so skip initialization.
  	 * This allows a request to share the flush and elevator data.
  	 */
  	if (bio->bi_rw & (REQ_FLUSH | REQ_FUA))
  		return false;
  
  	return true;
  }
da8303c63   Tejun Heo   block: make get_r...
828
  /**
852c788f8   Tejun Heo   block: implement ...
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
   * rq_ioc - determine io_context for request allocation
   * @bio: request being allocated is for this bio (can be %NULL)
   *
   * Determine io_context to use for request allocation for @bio.  May return
   * %NULL if %current->io_context doesn't exist.
   */
  static struct io_context *rq_ioc(struct bio *bio)
  {
  #ifdef CONFIG_BLK_CGROUP
  	if (bio && bio->bi_ioc)
  		return bio->bi_ioc;
  #endif
  	return current->io_context;
  }
  
  /**
a06e05e6a   Tejun Heo   block: refactor g...
845
   * __get_request - get a free request
5b788ce3e   Tejun Heo   block: prepare fo...
846
   * @rl: request list to allocate from
da8303c63   Tejun Heo   block: make get_r...
847
848
849
850
851
852
853
854
855
856
   * @rw_flags: RW and SYNC flags
   * @bio: bio to allocate request for (can be %NULL)
   * @gfp_mask: allocation mask
   *
   * Get a free request from @q.  This function may fail under memory
   * pressure or if @q is dead.
   *
   * Must be callled with @q->queue_lock held and,
   * Returns %NULL on failure, with @q->queue_lock held.
   * Returns !%NULL on success, with @q->queue_lock *not held*.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
857
   */
5b788ce3e   Tejun Heo   block: prepare fo...
858
  static struct request *__get_request(struct request_list *rl, int rw_flags,
a06e05e6a   Tejun Heo   block: refactor g...
859
  				     struct bio *bio, gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
860
  {
5b788ce3e   Tejun Heo   block: prepare fo...
861
  	struct request_queue *q = rl->q;
b679281a6   Tejun Heo   block: restructur...
862
  	struct request *rq;
7f4b35d15   Tejun Heo   block: allocate i...
863
864
  	struct elevator_type *et = q->elevator->type;
  	struct io_context *ioc = rq_ioc(bio);
f1f8cc946   Tejun Heo   block, cfq: move ...
865
  	struct io_cq *icq = NULL;
1faa16d22   Jens Axboe   block: change the...
866
  	const bool is_sync = rw_is_sync(rw_flags) != 0;
75eb6c372   Tejun Heo   block: pass aroun...
867
  	int may_queue;
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
868

3f3299d5c   Bart Van Assche   block: Rename que...
869
  	if (unlikely(blk_queue_dying(q)))
da8303c63   Tejun Heo   block: make get_r...
870
  		return NULL;
7749a8d42   Jens Axboe   [PATCH] Propagate...
871
  	may_queue = elv_may_queue(q, rw_flags);
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
872
873
  	if (may_queue == ELV_MQUEUE_NO)
  		goto rq_starved;
1faa16d22   Jens Axboe   block: change the...
874
875
  	if (rl->count[is_sync]+1 >= queue_congestion_on_threshold(q)) {
  		if (rl->count[is_sync]+1 >= q->nr_requests) {
f2dbd76a0   Tejun Heo   block, cfq: repla...
876
  			/*
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
877
878
879
880
881
  			 * 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   Tejun Heo   block: prepare fo...
882
  			if (!blk_rl_full(rl, is_sync)) {
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
883
  				ioc_set_batching(q, ioc);
5b788ce3e   Tejun Heo   block: prepare fo...
884
  				blk_set_rl_full(rl, is_sync);
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
885
886
887
888
889
890
891
892
  			} 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
  					 */
b679281a6   Tejun Heo   block: restructur...
893
  					return NULL;
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
894
895
  				}
  			}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
896
  		}
a051661ca   Tejun Heo   blkcg: implement ...
897
898
899
900
901
902
  		/*
  		 * bdi isn't aware of blkcg yet.  As all async IOs end up
  		 * root blkcg anyway, just use root blkcg state.
  		 */
  		if (rl == &q->root_rl)
  			blk_set_queue_congested(q, is_sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
903
  	}
082cf69eb   Jens Axboe   [PATCH] ll_rw_blk...
904
905
906
907
908
  	/*
  	 * 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   Jens Axboe   block: change the...
909
  	if (rl->count[is_sync] >= (3 * q->nr_requests / 2))
b679281a6   Tejun Heo   block: restructur...
910
  		return NULL;
fd782a4a9   Hugh Dickins   [PATCH] Fix get_r...
911

8a5ecdd42   Tejun Heo   block: add q->nr_...
912
  	q->nr_rqs[is_sync]++;
1faa16d22   Jens Axboe   block: change the...
913
914
  	rl->count[is_sync]++;
  	rl->starved[is_sync] = 0;
cb98fc8bb   Tejun Heo   [BLOCK] Reimpleme...
915

f1f8cc946   Tejun Heo   block, cfq: move ...
916
917
918
919
920
921
922
923
924
925
  	/*
  	 * Decide whether the new request will be managed by elevator.  If
  	 * so, mark @rw_flags and increment elvpriv.  Non-zero elvpriv will
  	 * 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.
  	 *
  	 * Also, lookup icq while holding queue_lock.  If it doesn't exist,
  	 * it will be created after releasing queue_lock.
  	 */
d732580b4   Tejun Heo   block: implement ...
926
  	if (blk_rq_should_init_elevator(bio) && !blk_queue_bypass(q)) {
75eb6c372   Tejun Heo   block: pass aroun...
927
  		rw_flags |= REQ_ELVPRIV;
8a5ecdd42   Tejun Heo   block: add q->nr_...
928
  		q->nr_rqs_elvpriv++;
f1f8cc946   Tejun Heo   block, cfq: move ...
929
930
  		if (et->icq_cache && ioc)
  			icq = ioc_lookup_icq(ioc, q);
9d5a4e946   Mike Snitzer   block: skip eleva...
931
  	}
cb98fc8bb   Tejun Heo   [BLOCK] Reimpleme...
932

f253b86b4   Jens Axboe   Revert "block: fi...
933
934
  	if (blk_queue_io_stat(q))
  		rw_flags |= REQ_IO_STAT;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
935
  	spin_unlock_irq(q->queue_lock);
29e2b09ab   Tejun Heo   block: collapse b...
936
  	/* allocate and init request */
5b788ce3e   Tejun Heo   block: prepare fo...
937
  	rq = mempool_alloc(rl->rq_pool, gfp_mask);
29e2b09ab   Tejun Heo   block: collapse b...
938
  	if (!rq)
b679281a6   Tejun Heo   block: restructur...
939
  		goto fail_alloc;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
940

29e2b09ab   Tejun Heo   block: collapse b...
941
  	blk_rq_init(q, rq);
a051661ca   Tejun Heo   blkcg: implement ...
942
  	blk_rq_set_rl(rq, rl);
29e2b09ab   Tejun Heo   block: collapse b...
943
  	rq->cmd_flags = rw_flags | REQ_ALLOCED;
aaf7c6806   Tejun Heo   block: fix elvpri...
944
  	/* init elvpriv */
29e2b09ab   Tejun Heo   block: collapse b...
945
  	if (rw_flags & REQ_ELVPRIV) {
aaf7c6806   Tejun Heo   block: fix elvpri...
946
  		if (unlikely(et->icq_cache && !icq)) {
7f4b35d15   Tejun Heo   block: allocate i...
947
948
  			if (ioc)
  				icq = ioc_create_icq(ioc, q, gfp_mask);
aaf7c6806   Tejun Heo   block: fix elvpri...
949
950
  			if (!icq)
  				goto fail_elvpriv;
29e2b09ab   Tejun Heo   block: collapse b...
951
  		}
aaf7c6806   Tejun Heo   block: fix elvpri...
952
953
954
955
956
957
  
  		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   Tejun Heo   block: collapse b...
958
959
960
  		if (icq)
  			get_io_context(icq->ioc);
  	}
aaf7c6806   Tejun Heo   block: fix elvpri...
961
  out:
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
962
963
964
965
966
967
  	/*
  	 * 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   Linus Torvalds   Linux-2.6.12-rc2
968
969
  	if (ioc_batching(q, ioc))
  		ioc->nr_batch_requests--;
6728cb0e6   Jens Axboe   block: make core ...
970

1faa16d22   Jens Axboe   block: change the...
971
  	trace_block_getrq(q, bio, rw_flags & 1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
972
  	return rq;
b679281a6   Tejun Heo   block: restructur...
973

aaf7c6806   Tejun Heo   block: fix elvpri...
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
  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.
  	 */
  	printk_ratelimited(KERN_WARNING "%s: request aux data allocation failed, iosched may be disturbed
  ",
  			   dev_name(q->backing_dev_info.dev));
  
  	rq->cmd_flags &= ~REQ_ELVPRIV;
  	rq->elv.icq = NULL;
  
  	spin_lock_irq(q->queue_lock);
8a5ecdd42   Tejun Heo   block: add q->nr_...
989
  	q->nr_rqs_elvpriv--;
aaf7c6806   Tejun Heo   block: fix elvpri...
990
991
  	spin_unlock_irq(q->queue_lock);
  	goto out;
b679281a6   Tejun Heo   block: restructur...
992
993
994
995
996
997
998
999
1000
  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);
5b788ce3e   Tejun Heo   block: prepare fo...
1001
  	freed_request(rl, rw_flags);
b679281a6   Tejun Heo   block: restructur...
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
  
  	/*
  	 * 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;
  	return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1014
  }
da8303c63   Tejun Heo   block: make get_r...
1015
  /**
a06e05e6a   Tejun Heo   block: refactor g...
1016
   * get_request - get a free request
da8303c63   Tejun Heo   block: make get_r...
1017
1018
1019
   * @q: request_queue to allocate request from
   * @rw_flags: RW and SYNC flags
   * @bio: bio to allocate request for (can be %NULL)
a06e05e6a   Tejun Heo   block: refactor g...
1020
   * @gfp_mask: allocation mask
da8303c63   Tejun Heo   block: make get_r...
1021
   *
a06e05e6a   Tejun Heo   block: refactor g...
1022
1023
   * Get a free request from @q.  If %__GFP_WAIT is set in @gfp_mask, this
   * function keeps retrying under memory pressure and fails iff @q is dead.
d6344532a   Nick Piggin   [PATCH] blk: redu...
1024
   *
da8303c63   Tejun Heo   block: make get_r...
1025
1026
1027
   * Must be callled with @q->queue_lock held and,
   * Returns %NULL on failure, with @q->queue_lock held.
   * Returns !%NULL on success, with @q->queue_lock *not held*.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1028
   */
a06e05e6a   Tejun Heo   block: refactor g...
1029
1030
  static struct request *get_request(struct request_queue *q, int rw_flags,
  				   struct bio *bio, gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1031
  {
1faa16d22   Jens Axboe   block: change the...
1032
  	const bool is_sync = rw_is_sync(rw_flags) != 0;
a06e05e6a   Tejun Heo   block: refactor g...
1033
  	DEFINE_WAIT(wait);
a051661ca   Tejun Heo   blkcg: implement ...
1034
  	struct request_list *rl;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1035
  	struct request *rq;
a051661ca   Tejun Heo   blkcg: implement ...
1036
1037
  
  	rl = blk_get_rl(q, bio);	/* transferred to @rq on success */
a06e05e6a   Tejun Heo   block: refactor g...
1038
  retry:
a051661ca   Tejun Heo   blkcg: implement ...
1039
  	rq = __get_request(rl, rw_flags, bio, gfp_mask);
a06e05e6a   Tejun Heo   block: refactor g...
1040
1041
  	if (rq)
  		return rq;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1042

3f3299d5c   Bart Van Assche   block: Rename que...
1043
  	if (!(gfp_mask & __GFP_WAIT) || unlikely(blk_queue_dying(q))) {
a051661ca   Tejun Heo   blkcg: implement ...
1044
  		blk_put_rl(rl);
a06e05e6a   Tejun Heo   block: refactor g...
1045
  		return NULL;
a051661ca   Tejun Heo   blkcg: implement ...
1046
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1047

a06e05e6a   Tejun Heo   block: refactor g...
1048
1049
1050
  	/* wait on @rl and retry */
  	prepare_to_wait_exclusive(&rl->wait[is_sync], &wait,
  				  TASK_UNINTERRUPTIBLE);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1051

a06e05e6a   Tejun Heo   block: refactor g...
1052
  	trace_block_sleeprq(q, bio, rw_flags & 1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1053

a06e05e6a   Tejun Heo   block: refactor g...
1054
1055
  	spin_unlock_irq(q->queue_lock);
  	io_schedule();
d6344532a   Nick Piggin   [PATCH] blk: redu...
1056

a06e05e6a   Tejun Heo   block: refactor g...
1057
1058
1059
1060
1061
  	/*
  	 * 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   Tejun Heo   block: refactor g...
1062
  	ioc_set_batching(q, current->io_context);
05caf8dbc   Zhang, Yanmin   block: Move the s...
1063

a06e05e6a   Tejun Heo   block: refactor g...
1064
1065
  	spin_lock_irq(q->queue_lock);
  	finish_wait(&rl->wait[is_sync], &wait);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1066

a06e05e6a   Tejun Heo   block: refactor g...
1067
  	goto retry;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1068
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1069
1070
  static struct request *blk_old_get_request(struct request_queue *q, int rw,
  		gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1071
1072
1073
1074
  {
  	struct request *rq;
  
  	BUG_ON(rw != READ && rw != WRITE);
7f4b35d15   Tejun Heo   block: allocate i...
1075
1076
  	/* create ioc upfront */
  	create_io_context(gfp_mask, q->node);
d6344532a   Nick Piggin   [PATCH] blk: redu...
1077
  	spin_lock_irq(q->queue_lock);
a06e05e6a   Tejun Heo   block: refactor g...
1078
  	rq = get_request(q, rw, NULL, gfp_mask);
da8303c63   Tejun Heo   block: make get_r...
1079
1080
  	if (!rq)
  		spin_unlock_irq(q->queue_lock);
d6344532a   Nick Piggin   [PATCH] blk: redu...
1081
  	/* q->queue_lock is unlocked at this point */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1082
1083
1084
  
  	return rq;
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1085
1086
1087
1088
  
  struct request *blk_get_request(struct request_queue *q, int rw, gfp_t gfp_mask)
  {
  	if (q->mq_ops)
18741986a   Christoph Hellwig   blk-mq: rework fl...
1089
  		return blk_mq_alloc_request(q, rw, gfp_mask);
320ae51fe   Jens Axboe   blk-mq: new multi...
1090
1091
1092
  	else
  		return blk_old_get_request(q, rw, gfp_mask);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1093
1094
1095
  EXPORT_SYMBOL(blk_get_request);
  
  /**
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1096
   * blk_make_request - given a bio, allocate a corresponding struct request.
8ebf97560   Randy Dunlap   block: fix kernel...
1097
   * @q: target request queue
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1098
1099
   * @bio:  The bio describing the memory mappings that will be submitted for IO.
   *        It may be a chained-bio properly constructed by block/bio layer.
8ebf97560   Randy Dunlap   block: fix kernel...
1100
   * @gfp_mask: gfp flags to be used for memory allocation
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1101
   *
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1102
1103
1104
1105
   * blk_make_request is the parallel of generic_make_request for BLOCK_PC
   * type commands. Where the struct request needs to be farther initialized by
   * the caller. It is passed a &struct bio, which describes the memory info of
   * the I/O transfer.
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1106
   *
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1107
1108
1109
1110
1111
1112
1113
1114
1115
   * The caller of blk_make_request must make sure that bi_io_vec
   * are set to describe the memory buffers. That bio_data_dir() will return
   * the needed direction of the request. (And all bio's in the passed bio-chain
   * are properly set accordingly)
   *
   * If called under none-sleepable conditions, mapped bio buffers must not
   * need bouncing, by calling the appropriate masked or flagged allocator,
   * suitable for the target device. Otherwise the call to blk_queue_bounce will
   * BUG.
53674ac5a   Jens Axboe   block: add warnin...
1116
1117
1118
1119
1120
1121
1122
1123
1124
   *
   * WARNING: When allocating/cloning a bio-chain, careful consideration should be
   * given to how you allocate bios. In particular, you cannot use __GFP_WAIT for
   * anything but the first bio in the chain. Otherwise you risk waiting for IO
   * completion of a bio that hasn't been submitted yet, thus resulting in a
   * deadlock. Alternatively bios should be allocated using bio_kmalloc() instead
   * of bio_alloc(), as that avoids the mempool deadlock.
   * If possible a big IO should be split into smaller parts when allocation
   * fails. Partial allocation should not be an error, or you risk a live-lock.
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1125
   */
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1126
1127
  struct request *blk_make_request(struct request_queue *q, struct bio *bio,
  				 gfp_t gfp_mask)
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1128
  {
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
  	struct request *rq = blk_get_request(q, bio_data_dir(bio), gfp_mask);
  
  	if (unlikely(!rq))
  		return ERR_PTR(-ENOMEM);
  
  	for_each_bio(bio) {
  		struct bio *bounce_bio = bio;
  		int ret;
  
  		blk_queue_bounce(q, &bounce_bio);
  		ret = blk_rq_append_bio(q, rq, bounce_bio);
  		if (unlikely(ret)) {
  			blk_put_request(rq);
  			return ERR_PTR(ret);
  		}
  	}
  
  	return rq;
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1147
  }
79eb63e9e   Boaz Harrosh   block: Add blk_ma...
1148
  EXPORT_SYMBOL(blk_make_request);
dc72ef4ae   Jens Axboe   [PATCH] Add blk_s...
1149
1150
  
  /**
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1151
1152
1153
1154
1155
1156
1157
1158
1159
   * 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   Jens Axboe   [BLOCK] Get rid o...
1160
  void blk_requeue_request(struct request_queue *q, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1161
  {
242f9dcb8   Jens Axboe   block: unify requ...
1162
1163
  	blk_delete_timer(rq);
  	blk_clear_rq_complete(rq);
5f3ea37c7   Arnaldo Carvalho de Melo   blktrace: port to...
1164
  	trace_block_rq_requeue(q, rq);
2056a782f   Jens Axboe   [PATCH] Block que...
1165

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1166
1167
  	if (blk_rq_tagged(rq))
  		blk_queue_end_tag(q, rq);
ba396a6c1   James Bottomley   block: fix oops w...
1168
  	BUG_ON(blk_queued_rq(rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1169
1170
  	elv_requeue_request(q, rq);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1171
  EXPORT_SYMBOL(blk_requeue_request);
73c101011   Jens Axboe   block: initial pa...
1172
1173
1174
  static void add_acct_request(struct request_queue *q, struct request *rq,
  			     int where)
  {
320ae51fe   Jens Axboe   blk-mq: new multi...
1175
  	blk_account_io_start(rq, true);
7eaceacca   Jens Axboe   block: remove per...
1176
  	__elv_add_request(q, rq, where);
73c101011   Jens Axboe   block: initial pa...
1177
  }
074a7aca7   Tejun Heo   block: move stats...
1178
1179
1180
1181
1182
  static void part_round_stats_single(int cpu, struct hd_struct *part,
  				    unsigned long now)
  {
  	if (now == part->stamp)
  		return;
316d315bf   Nikanth Karthikesan   block: Seperate r...
1183
  	if (part_in_flight(part)) {
074a7aca7   Tejun Heo   block: move stats...
1184
  		__part_stat_add(cpu, part, time_in_queue,
316d315bf   Nikanth Karthikesan   block: Seperate r...
1185
  				part_in_flight(part) * (now - part->stamp));
074a7aca7   Tejun Heo   block: move stats...
1186
1187
1188
1189
1190
1191
  		__part_stat_add(cpu, part, io_ticks, (now - part->stamp));
  	}
  	part->stamp = now;
  }
  
  /**
496aa8a98   Randy Dunlap   block: fix curren...
1192
1193
1194
   * part_round_stats() - Round off the performance stats on a struct disk_stats.
   * @cpu: cpu number for stats access
   * @part: target partition
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
   *
   * 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.
   */
c99590591   Tejun Heo   block: fix diskst...
1207
  void part_round_stats(int cpu, struct hd_struct *part)
6f2576af5   Jerome Marchand   Enhanced partitio...
1208
1209
  {
  	unsigned long now = jiffies;
074a7aca7   Tejun Heo   block: move stats...
1210
1211
1212
  	if (part->partno)
  		part_round_stats_single(cpu, &part_to_disk(part)->part0, now);
  	part_round_stats_single(cpu, part, now);
6f2576af5   Jerome Marchand   Enhanced partitio...
1213
  }
074a7aca7   Tejun Heo   block: move stats...
1214
  EXPORT_SYMBOL_GPL(part_round_stats);
6f2576af5   Jerome Marchand   Enhanced partitio...
1215

c8158819d   Lin Ming   block: implement ...
1216
1217
1218
1219
1220
1221
1222
1223
1224
  #ifdef CONFIG_PM_RUNTIME
  static void blk_pm_put_request(struct request *rq)
  {
  	if (rq->q->dev && !(rq->cmd_flags & REQ_PM) && !--rq->q->nr_pending)
  		pm_runtime_mark_last_busy(rq->q->dev);
  }
  #else
  static inline void blk_pm_put_request(struct request *rq) {}
  #endif
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1225
1226
1227
  /*
   * queue lock must be held
   */
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1228
  void __blk_put_request(struct request_queue *q, struct request *req)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1229
  {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1230
1231
  	if (unlikely(!q))
  		return;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1232

6f5ba581c   Christoph Hellwig   blk-mq: divert __...
1233
1234
1235
1236
  	if (q->mq_ops) {
  		blk_mq_free_request(req);
  		return;
  	}
c8158819d   Lin Ming   block: implement ...
1237
  	blk_pm_put_request(req);
8922e16cf   Tejun Heo   [PATCH] 01/05 Imp...
1238
  	elv_completed_request(q, req);
1cd96c242   Boaz Harrosh   block: WARN in __...
1239
1240
  	/* this is a bio leak */
  	WARN_ON(req->bio != NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1241
1242
1243
1244
  	/*
  	 * Request may not have originated from ll_rw_blk. if not,
  	 * it didn't come out of our reserved rq pools
  	 */
49171e5c6   Jens Axboe   [PATCH] Remove st...
1245
  	if (req->cmd_flags & REQ_ALLOCED) {
75eb6c372   Tejun Heo   block: pass aroun...
1246
  		unsigned int flags = req->cmd_flags;
a051661ca   Tejun Heo   blkcg: implement ...
1247
  		struct request_list *rl = blk_rq_rl(req);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1248

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1249
  		BUG_ON(!list_empty(&req->queuelist));
9817064b6   Jens Axboe   [PATCH] elevator:...
1250
  		BUG_ON(!hlist_unhashed(&req->hash));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1251

a051661ca   Tejun Heo   blkcg: implement ...
1252
1253
1254
  		blk_free_request(rl, req);
  		freed_request(rl, flags);
  		blk_put_rl(rl);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1255
1256
  	}
  }
6e39b69e7   Mike Christie   [SCSI] export blk...
1257
  EXPORT_SYMBOL_GPL(__blk_put_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1258
1259
  void blk_put_request(struct request *req)
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1260
  	struct request_queue *q = req->q;
8922e16cf   Tejun Heo   [PATCH] 01/05 Imp...
1261

320ae51fe   Jens Axboe   blk-mq: new multi...
1262
1263
1264
1265
1266
1267
1268
1269
1270
  	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   Linus Torvalds   Linux-2.6.12-rc2
1271
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1272
  EXPORT_SYMBOL(blk_put_request);
66ac02801   Christoph Hellwig   block: don't allo...
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
  /**
   * blk_add_request_payload - add a payload to a request
   * @rq: request to update
   * @page: page backing the payload
   * @len: length of the payload.
   *
   * This allows to later add a payload to an already submitted request by
   * a block driver.  The driver needs to take care of freeing the payload
   * itself.
   *
   * Note that this is a quite horrible hack and nothing but handling of
   * discard requests should ever use it.
   */
  void blk_add_request_payload(struct request *rq, struct page *page,
  		unsigned int len)
  {
  	struct bio *bio = rq->bio;
  
  	bio->bi_io_vec->bv_page = page;
  	bio->bi_io_vec->bv_offset = 0;
  	bio->bi_io_vec->bv_len = len;
4f024f379   Kent Overstreet   block: Abstract o...
1294
  	bio->bi_iter.bi_size = len;
66ac02801   Christoph Hellwig   block: don't allo...
1295
1296
1297
1298
1299
1300
1301
1302
  	bio->bi_vcnt = 1;
  	bio->bi_phys_segments = 1;
  
  	rq->__data_len = rq->resid_len = len;
  	rq->nr_phys_segments = 1;
  	rq->buffer = bio_data(bio);
  }
  EXPORT_SYMBOL_GPL(blk_add_request_payload);
320ae51fe   Jens Axboe   blk-mq: new multi...
1303
1304
  bool bio_attempt_back_merge(struct request_queue *q, struct request *req,
  			    struct bio *bio)
73c101011   Jens Axboe   block: initial pa...
1305
1306
  {
  	const int ff = bio->bi_rw & REQ_FAILFAST_MASK;
73c101011   Jens Axboe   block: initial pa...
1307
1308
  	if (!ll_back_merge_fn(q, req, bio))
  		return false;
8c1cf6bb0   Tejun Heo   block: add @req t...
1309
  	trace_block_bio_backmerge(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1310
1311
1312
1313
1314
1315
  
  	if ((req->cmd_flags & REQ_FAILFAST_MASK) != ff)
  		blk_rq_set_mixed_merge(req);
  
  	req->biotail->bi_next = bio;
  	req->biotail = bio;
4f024f379   Kent Overstreet   block: Abstract o...
1316
  	req->__data_len += bio->bi_iter.bi_size;
73c101011   Jens Axboe   block: initial pa...
1317
  	req->ioprio = ioprio_best(req->ioprio, bio_prio(bio));
320ae51fe   Jens Axboe   blk-mq: new multi...
1318
  	blk_account_io_start(req, false);
73c101011   Jens Axboe   block: initial pa...
1319
1320
  	return true;
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1321
1322
  bool bio_attempt_front_merge(struct request_queue *q, struct request *req,
  			     struct bio *bio)
73c101011   Jens Axboe   block: initial pa...
1323
1324
  {
  	const int ff = bio->bi_rw & REQ_FAILFAST_MASK;
73c101011   Jens Axboe   block: initial pa...
1325

73c101011   Jens Axboe   block: initial pa...
1326
1327
  	if (!ll_front_merge_fn(q, req, bio))
  		return false;
8c1cf6bb0   Tejun Heo   block: add @req t...
1328
  	trace_block_bio_frontmerge(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1329
1330
1331
  
  	if ((req->cmd_flags & REQ_FAILFAST_MASK) != ff)
  		blk_rq_set_mixed_merge(req);
73c101011   Jens Axboe   block: initial pa...
1332
1333
1334
1335
1336
1337
1338
1339
1340
  	bio->bi_next = req->bio;
  	req->bio = bio;
  
  	/*
  	 * may not be valid. if the low level driver said
  	 * it didn't need a bounce buffer then it better
  	 * not touch req->buffer either...
  	 */
  	req->buffer = bio_data(bio);
4f024f379   Kent Overstreet   block: Abstract o...
1341
1342
  	req->__sector = bio->bi_iter.bi_sector;
  	req->__data_len += bio->bi_iter.bi_size;
73c101011   Jens Axboe   block: initial pa...
1343
  	req->ioprio = ioprio_best(req->ioprio, bio_prio(bio));
320ae51fe   Jens Axboe   blk-mq: new multi...
1344
  	blk_account_io_start(req, false);
73c101011   Jens Axboe   block: initial pa...
1345
1346
  	return true;
  }
bd87b5898   Tejun Heo   block: drop @tsk ...
1347
  /**
320ae51fe   Jens Axboe   blk-mq: new multi...
1348
   * blk_attempt_plug_merge - try to merge with %current's plugged list
bd87b5898   Tejun Heo   block: drop @tsk ...
1349
1350
1351
1352
1353
1354
1355
1356
   * @q: request_queue new bio is being queued at
   * @bio: new bio being queued
   * @request_count: out parameter for number of traversed plugged requests
   *
   * 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   Tejun Heo   block: don't call...
1357
1358
1359
1360
1361
1362
   * 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.
73c101011   Jens Axboe   block: initial pa...
1363
   */
320ae51fe   Jens Axboe   blk-mq: new multi...
1364
1365
  bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
  			    unsigned int *request_count)
73c101011   Jens Axboe   block: initial pa...
1366
1367
1368
1369
  {
  	struct blk_plug *plug;
  	struct request *rq;
  	bool ret = false;
92f399c72   Shaohua Li   blk-mq: mq plug l...
1370
  	struct list_head *plug_list;
73c101011   Jens Axboe   block: initial pa...
1371

23779fbc9   Alireza Haghdoost   block: Enable sys...
1372
1373
  	if (blk_queue_nomerges(q))
  		goto out;
bd87b5898   Tejun Heo   block: drop @tsk ...
1374
  	plug = current->plug;
73c101011   Jens Axboe   block: initial pa...
1375
1376
  	if (!plug)
  		goto out;
56ebdaf2f   Shaohua Li   block: simplify f...
1377
  	*request_count = 0;
73c101011   Jens Axboe   block: initial pa...
1378

92f399c72   Shaohua Li   blk-mq: mq plug l...
1379
1380
1381
1382
1383
1384
  	if (q->mq_ops)
  		plug_list = &plug->mq_list;
  	else
  		plug_list = &plug->list;
  
  	list_for_each_entry_reverse(rq, plug_list, queuelist) {
73c101011   Jens Axboe   block: initial pa...
1385
  		int el_ret;
1b2e19f17   Shaohua Li   block: make auto ...
1386
1387
  		if (rq->q == q)
  			(*request_count)++;
56ebdaf2f   Shaohua Li   block: simplify f...
1388

07c2bd373   Tejun Heo   block: don't call...
1389
  		if (rq->q != q || !blk_rq_merge_ok(rq, bio))
73c101011   Jens Axboe   block: initial pa...
1390
  			continue;
050c8ea80   Tejun Heo   block: separate o...
1391
  		el_ret = blk_try_merge(rq, bio);
73c101011   Jens Axboe   block: initial pa...
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
  		if (el_ret == ELEVATOR_BACK_MERGE) {
  			ret = bio_attempt_back_merge(q, rq, bio);
  			if (ret)
  				break;
  		} else if (el_ret == ELEVATOR_FRONT_MERGE) {
  			ret = bio_attempt_front_merge(q, rq, bio);
  			if (ret)
  				break;
  		}
  	}
  out:
  	return ret;
  }
86db1e297   Jens Axboe   block: continue l...
1405
  void init_request_from_bio(struct request *req, struct bio *bio)
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1406
  {
4aff5e233   Jens Axboe   [PATCH] Split str...
1407
  	req->cmd_type = REQ_TYPE_FS;
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1408

7b6d91dae   Christoph Hellwig   block: unify flag...
1409
1410
  	req->cmd_flags |= bio->bi_rw & REQ_COMMON_MASK;
  	if (bio->bi_rw & REQ_RAHEAD)
a82afdfcb   Tejun Heo   block: use the sa...
1411
  		req->cmd_flags |= REQ_FAILFAST_MASK;
b31dc66a5   Jens Axboe   [PATCH] Kill PF_S...
1412

52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1413
  	req->errors = 0;
4f024f379   Kent Overstreet   block: Abstract o...
1414
  	req->__sector = bio->bi_iter.bi_sector;
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1415
  	req->ioprio = bio_prio(bio);
bc1c56fde   NeilBrown   Share code betwee...
1416
  	blk_rq_bio_prep(req->q, req, bio);
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1417
  }
5a7bbad27   Christoph Hellwig   block: remove sup...
1418
  void blk_queue_bio(struct request_queue *q, struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1419
  {
5e00d1b5b   Jiri Slaby   BLOCK: fix bio.bi...
1420
  	const bool sync = !!(bio->bi_rw & REQ_SYNC);
73c101011   Jens Axboe   block: initial pa...
1421
1422
1423
  	struct blk_plug *plug;
  	int el_ret, rw_flags, where = ELEVATOR_INSERT_SORT;
  	struct request *req;
56ebdaf2f   Shaohua Li   block: simplify f...
1424
  	unsigned int request_count = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1425

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1426
1427
1428
1429
1430
1431
  	/*
  	 * 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);
ffecfd1a7   Darrick J. Wong   block: optionally...
1432
1433
1434
1435
  	if (bio_integrity_enabled(bio) && bio_integrity_prep(bio)) {
  		bio_endio(bio, -EIO);
  		return;
  	}
4fed947cb   Tejun Heo   block: implement ...
1436
  	if (bio->bi_rw & (REQ_FLUSH | REQ_FUA)) {
73c101011   Jens Axboe   block: initial pa...
1437
  		spin_lock_irq(q->queue_lock);
ae1b15396   Tejun Heo   block: reimplemen...
1438
  		where = ELEVATOR_INSERT_FLUSH;
28e7d1845   Tejun Heo   block: drop barri...
1439
1440
  		goto get_rq;
  	}
73c101011   Jens Axboe   block: initial pa...
1441
1442
1443
1444
  	/*
  	 * Check if we can merge with the plugged list before grabbing
  	 * any locks.
  	 */
320ae51fe   Jens Axboe   blk-mq: new multi...
1445
  	if (blk_attempt_plug_merge(q, bio, &request_count))
5a7bbad27   Christoph Hellwig   block: remove sup...
1446
  		return;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1447

73c101011   Jens Axboe   block: initial pa...
1448
  	spin_lock_irq(q->queue_lock);
2056a782f   Jens Axboe   [PATCH] Block que...
1449

73c101011   Jens Axboe   block: initial pa...
1450
1451
  	el_ret = elv_merge(q, &req, bio);
  	if (el_ret == ELEVATOR_BACK_MERGE) {
73c101011   Jens Axboe   block: initial pa...
1452
  		if (bio_attempt_back_merge(q, req, bio)) {
07c2bd373   Tejun Heo   block: don't call...
1453
  			elv_bio_merged(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1454
1455
1456
1457
1458
  			if (!attempt_back_merge(q, req))
  				elv_merged_request(q, req, el_ret);
  			goto out_unlock;
  		}
  	} else if (el_ret == ELEVATOR_FRONT_MERGE) {
73c101011   Jens Axboe   block: initial pa...
1459
  		if (bio_attempt_front_merge(q, req, bio)) {
07c2bd373   Tejun Heo   block: don't call...
1460
  			elv_bio_merged(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1461
1462
1463
  			if (!attempt_front_merge(q, req))
  				elv_merged_request(q, req, el_ret);
  			goto out_unlock;
80a761fd3   Tejun Heo   block: implement ...
1464
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1465
  	}
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1466
  get_rq:
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1467
  	/*
7749a8d42   Jens Axboe   [PATCH] Propagate...
1468
1469
1470
1471
1472
1473
  	 * This sync check and mask will be re-done in init_request_from_bio(),
  	 * but we need to set it earlier to expose the sync flag to the
  	 * rq allocator and io schedulers.
  	 */
  	rw_flags = bio_data_dir(bio);
  	if (sync)
7b6d91dae   Christoph Hellwig   block: unify flag...
1474
  		rw_flags |= REQ_SYNC;
7749a8d42   Jens Axboe   [PATCH] Propagate...
1475
1476
  
  	/*
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1477
  	 * Grab a free request. This is might sleep but can not fail.
d6344532a   Nick Piggin   [PATCH] blk: redu...
1478
  	 * Returns with the queue unlocked.
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1479
  	 */
a06e05e6a   Tejun Heo   block: refactor g...
1480
  	req = get_request(q, rw_flags, bio, GFP_NOIO);
da8303c63   Tejun Heo   block: make get_r...
1481
1482
1483
1484
  	if (unlikely(!req)) {
  		bio_endio(bio, -ENODEV);	/* @q is dead */
  		goto out_unlock;
  	}
d6344532a   Nick Piggin   [PATCH] blk: redu...
1485

450991bc1   Nick Piggin   [PATCH] blk: __ma...
1486
1487
1488
1489
1490
  	/*
  	 * 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   Linus Torvalds   Linux-2.6.12-rc2
1491
  	 */
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1492
  	init_request_from_bio(req, bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1493

9562ad9ab   Tao Ma   block: Remove the...
1494
  	if (test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags))
11ccf116d   Jens Axboe   block: fix warnin...
1495
  		req->cpu = raw_smp_processor_id();
73c101011   Jens Axboe   block: initial pa...
1496
1497
  
  	plug = current->plug;
721a9602e   Jens Axboe   block: kill off R...
1498
  	if (plug) {
dc6d36c97   Jens Axboe   block: readd plug...
1499
1500
  		/*
  		 * If this is the first request added after a plug, fire
7aef2e780   Jianpeng Ma   block: trace all ...
1501
  		 * of a plug trace.
dc6d36c97   Jens Axboe   block: readd plug...
1502
  		 */
7aef2e780   Jianpeng Ma   block: trace all ...
1503
  		if (!request_count)
dc6d36c97   Jens Axboe   block: readd plug...
1504
  			trace_block_plug(q);
3540d5e89   Shaohua Li   block: avoid unne...
1505
  		else {
019ceb7d5   Shaohua Li   block: add missed...
1506
  			if (request_count >= BLK_MAX_REQUEST_COUNT) {
3540d5e89   Shaohua Li   block: avoid unne...
1507
  				blk_flush_plug_list(plug, false);
019ceb7d5   Shaohua Li   block: add missed...
1508
1509
  				trace_block_plug(q);
  			}
73c101011   Jens Axboe   block: initial pa...
1510
  		}
73c101011   Jens Axboe   block: initial pa...
1511
  		list_add_tail(&req->queuelist, &plug->list);
320ae51fe   Jens Axboe   blk-mq: new multi...
1512
  		blk_account_io_start(req, true);
73c101011   Jens Axboe   block: initial pa...
1513
1514
1515
  	} else {
  		spin_lock_irq(q->queue_lock);
  		add_acct_request(q, req, where);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
1516
  		__blk_run_queue(q);
73c101011   Jens Axboe   block: initial pa...
1517
1518
1519
  out_unlock:
  		spin_unlock_irq(q->queue_lock);
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1520
  }
c20e8de27   Jens Axboe   block: rename __m...
1521
  EXPORT_SYMBOL_GPL(blk_queue_bio);	/* for device mapper only */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1522
1523
1524
1525
1526
1527
1528
  
  /*
   * If bio->bi_dev is a partition, remap the location
   */
  static inline void blk_partition_remap(struct bio *bio)
  {
  	struct block_device *bdev = bio->bi_bdev;
bf2de6f5a   Jens Axboe   block: Initial su...
1529
  	if (bio_sectors(bio) && bdev != bdev->bd_contains) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1530
  		struct hd_struct *p = bdev->bd_part;
4f024f379   Kent Overstreet   block: Abstract o...
1531
  		bio->bi_iter.bi_sector += p->start_sect;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1532
  		bio->bi_bdev = bdev->bd_contains;
c7149d6bc   Alan D. Brunelle   Fix remap handlin...
1533

d07335e51   Mike Snitzer   block: Rename "bl...
1534
1535
  		trace_block_bio_remap(bdev_get_queue(bio->bi_bdev), bio,
  				      bdev->bd_dev,
4f024f379   Kent Overstreet   block: Abstract o...
1536
  				      bio->bi_iter.bi_sector - p->start_sect);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1537
1538
  	}
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
  static void handle_bad_sector(struct bio *bio)
  {
  	char b[BDEVNAME_SIZE];
  
  	printk(KERN_INFO "attempt to access beyond end of device
  ");
  	printk(KERN_INFO "%s: rw=%ld, want=%Lu, limit=%Lu
  ",
  			bdevname(bio->bi_bdev, b),
  			bio->bi_rw,
f73a1c7d1   Kent Overstreet   block: Add bio_en...
1549
  			(unsigned long long)bio_end_sector(bio),
77304d2ab   Mike Snitzer   block: read i_siz...
1550
  			(long long)(i_size_read(bio->bi_bdev->bd_inode) >> 9));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1551
1552
1553
  
  	set_bit(BIO_EOF, &bio->bi_flags);
  }
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1554
1555
1556
1557
1558
1559
1560
1561
1562
  #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   Akinobu Mita   fail_make_request...
1563
  static bool should_fail_request(struct hd_struct *part, unsigned int bytes)
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1564
  {
b2c9cd379   Akinobu Mita   fail_make_request...
1565
  	return part->make_it_fail && should_fail(&fail_make_request, bytes);
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1566
1567
1568
1569
  }
  
  static int __init fail_make_request_debugfs(void)
  {
dd48c085c   Akinobu Mita   fault-injection: ...
1570
1571
1572
1573
  	struct dentry *dir = fault_create_debugfs_attr("fail_make_request",
  						NULL, &fail_make_request);
  
  	return IS_ERR(dir) ? PTR_ERR(dir) : 0;
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1574
1575
1576
1577
1578
  }
  
  late_initcall(fail_make_request_debugfs);
  
  #else /* CONFIG_FAIL_MAKE_REQUEST */
b2c9cd379   Akinobu Mita   fail_make_request...
1579
1580
  static inline bool should_fail_request(struct hd_struct *part,
  					unsigned int bytes)
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1581
  {
b2c9cd379   Akinobu Mita   fail_make_request...
1582
  	return false;
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1583
1584
1585
  }
  
  #endif /* CONFIG_FAIL_MAKE_REQUEST */
c07e2b412   Jens Axboe   block: factor our...
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
  /*
   * Check whether this bio extends beyond the end of the device.
   */
  static inline int bio_check_eod(struct bio *bio, unsigned int nr_sectors)
  {
  	sector_t maxsector;
  
  	if (!nr_sectors)
  		return 0;
  
  	/* Test device or partition size, when known. */
77304d2ab   Mike Snitzer   block: read i_siz...
1597
  	maxsector = i_size_read(bio->bi_bdev->bd_inode) >> 9;
c07e2b412   Jens Axboe   block: factor our...
1598
  	if (maxsector) {
4f024f379   Kent Overstreet   block: Abstract o...
1599
  		sector_t sector = bio->bi_iter.bi_sector;
c07e2b412   Jens Axboe   block: factor our...
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
  
  		if (maxsector < nr_sectors || maxsector - nr_sectors < sector) {
  			/*
  			 * This may well happen - the kernel calls bread()
  			 * without checking the size of the device, e.g., when
  			 * mounting a device.
  			 */
  			handle_bad_sector(bio);
  			return 1;
  		}
  	}
  
  	return 0;
  }
27a84d54c   Christoph Hellwig   block: refactor g...
1614
1615
  static noinline_for_stack bool
  generic_make_request_checks(struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1616
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1617
  	struct request_queue *q;
5a7bbad27   Christoph Hellwig   block: remove sup...
1618
  	int nr_sectors = bio_sectors(bio);
51fd77bd9   Jens Axboe   [BLOCK] Don't all...
1619
  	int err = -EIO;
5a7bbad27   Christoph Hellwig   block: remove sup...
1620
1621
  	char b[BDEVNAME_SIZE];
  	struct hd_struct *part;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1622
1623
  
  	might_sleep();
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1624

c07e2b412   Jens Axboe   block: factor our...
1625
1626
  	if (bio_check_eod(bio, nr_sectors))
  		goto end_io;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1627

5a7bbad27   Christoph Hellwig   block: remove sup...
1628
1629
1630
1631
1632
1633
1634
  	q = bdev_get_queue(bio->bi_bdev);
  	if (unlikely(!q)) {
  		printk(KERN_ERR
  		       "generic_make_request: Trying to access "
  			"nonexistent block-device %s (%Lu)
  ",
  			bdevname(bio->bi_bdev, b),
4f024f379   Kent Overstreet   block: Abstract o...
1635
  			(long long) bio->bi_iter.bi_sector);
5a7bbad27   Christoph Hellwig   block: remove sup...
1636
1637
  		goto end_io;
  	}
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1638

e2a60da74   Martin K. Petersen   block: Clean up s...
1639
1640
  	if (likely(bio_is_rw(bio) &&
  		   nr_sectors > queue_max_hw_sectors(q))) {
5a7bbad27   Christoph Hellwig   block: remove sup...
1641
1642
1643
1644
1645
1646
1647
  		printk(KERN_ERR "bio too big device %s (%u > %u)
  ",
  		       bdevname(bio->bi_bdev, b),
  		       bio_sectors(bio),
  		       queue_max_hw_sectors(q));
  		goto end_io;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1648

5a7bbad27   Christoph Hellwig   block: remove sup...
1649
  	part = bio->bi_bdev->bd_part;
4f024f379   Kent Overstreet   block: Abstract o...
1650
  	if (should_fail_request(part, bio->bi_iter.bi_size) ||
5a7bbad27   Christoph Hellwig   block: remove sup...
1651
  	    should_fail_request(&part_to_disk(part)->part0,
4f024f379   Kent Overstreet   block: Abstract o...
1652
  				bio->bi_iter.bi_size))
5a7bbad27   Christoph Hellwig   block: remove sup...
1653
  		goto end_io;
2056a782f   Jens Axboe   [PATCH] Block que...
1654

5a7bbad27   Christoph Hellwig   block: remove sup...
1655
1656
1657
1658
1659
  	/*
  	 * If this device has partitions, remap block n
  	 * of partition p to block n+start(p) of the disk.
  	 */
  	blk_partition_remap(bio);
2056a782f   Jens Axboe   [PATCH] Block que...
1660

5a7bbad27   Christoph Hellwig   block: remove sup...
1661
1662
  	if (bio_check_eod(bio, nr_sectors))
  		goto end_io;
1e87901e1   Tejun Heo   block: filter flu...
1663

5a7bbad27   Christoph Hellwig   block: remove sup...
1664
1665
1666
1667
1668
1669
1670
1671
1672
  	/*
  	 * Filter flush bio's early so that make_request based
  	 * drivers without flush support don't have to worry
  	 * about them.
  	 */
  	if ((bio->bi_rw & (REQ_FLUSH | REQ_FUA)) && !q->flush_flags) {
  		bio->bi_rw &= ~(REQ_FLUSH | REQ_FUA);
  		if (!nr_sectors) {
  			err = 0;
51fd77bd9   Jens Axboe   [BLOCK] Don't all...
1673
1674
  			goto end_io;
  		}
5a7bbad27   Christoph Hellwig   block: remove sup...
1675
  	}
5ddfe9691   NeilBrown   [PATCH] md: check...
1676

5a7bbad27   Christoph Hellwig   block: remove sup...
1677
1678
  	if ((bio->bi_rw & REQ_DISCARD) &&
  	    (!blk_queue_discard(q) ||
e2a60da74   Martin K. Petersen   block: Clean up s...
1679
  	     ((bio->bi_rw & REQ_SECURE) && !blk_queue_secdiscard(q)))) {
5a7bbad27   Christoph Hellwig   block: remove sup...
1680
1681
1682
  		err = -EOPNOTSUPP;
  		goto end_io;
  	}
01edede41   Minchan Kim   block: trace bio ...
1683

4363ac7c1   Martin K. Petersen   block: Implement ...
1684
  	if (bio->bi_rw & REQ_WRITE_SAME && !bdev_write_same(bio->bi_bdev)) {
5a7bbad27   Christoph Hellwig   block: remove sup...
1685
1686
1687
  		err = -EOPNOTSUPP;
  		goto end_io;
  	}
01edede41   Minchan Kim   block: trace bio ...
1688

7f4b35d15   Tejun Heo   block: allocate i...
1689
1690
1691
1692
1693
1694
1695
  	/*
  	 * 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);
bc16a4f93   Tejun Heo   block: reorganize...
1696
1697
  	if (blk_throtl_bio(q, bio))
  		return false;	/* throttled, will be resubmitted later */
27a84d54c   Christoph Hellwig   block: refactor g...
1698

5a7bbad27   Christoph Hellwig   block: remove sup...
1699
  	trace_block_bio_queue(q, bio);
27a84d54c   Christoph Hellwig   block: refactor g...
1700
  	return true;
a7384677b   Tejun Heo   block: remove dup...
1701
1702
1703
  
  end_io:
  	bio_endio(bio, err);
27a84d54c   Christoph Hellwig   block: refactor g...
1704
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1705
  }
27a84d54c   Christoph Hellwig   block: refactor g...
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
  /**
   * 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   Neil Brown   When stacked bloc...
1729
1730
1731
   */
  void generic_make_request(struct bio *bio)
  {
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1732
  	struct bio_list bio_list_on_stack;
27a84d54c   Christoph Hellwig   block: refactor g...
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
  	if (!generic_make_request_checks(bio))
  		return;
  
  	/*
  	 * 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   Akinobu Mita   blk-core: use BIO...
1746
  	if (current->bio_list) {
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1747
  		bio_list_add(current->bio_list, bio);
d89d87965   Neil Brown   When stacked bloc...
1748
1749
  		return;
  	}
27a84d54c   Christoph Hellwig   block: refactor g...
1750

d89d87965   Neil Brown   When stacked bloc...
1751
1752
1753
1754
1755
  	/* 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   Akinobu Mita   blk-core: use BIO...
1756
1757
  	 * we assign bio_list to a pointer to the bio_list_on_stack,
  	 * thus initialising the bio_list of new bios to be
27a84d54c   Christoph Hellwig   block: refactor g...
1758
  	 * added.  ->make_request() may indeed add some more bios
d89d87965   Neil Brown   When stacked bloc...
1759
1760
1761
  	 * 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   Akinobu Mita   blk-core: use BIO...
1762
  	 * of the top of the list (no pretending) and so remove it from
27a84d54c   Christoph Hellwig   block: refactor g...
1763
  	 * bio_list, and call into ->make_request() again.
d89d87965   Neil Brown   When stacked bloc...
1764
1765
  	 */
  	BUG_ON(bio->bi_next);
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1766
1767
  	bio_list_init(&bio_list_on_stack);
  	current->bio_list = &bio_list_on_stack;
d89d87965   Neil Brown   When stacked bloc...
1768
  	do {
27a84d54c   Christoph Hellwig   block: refactor g...
1769
1770
1771
  		struct request_queue *q = bdev_get_queue(bio->bi_bdev);
  
  		q->make_request_fn(q, bio);
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1772
  		bio = bio_list_pop(current->bio_list);
d89d87965   Neil Brown   When stacked bloc...
1773
  	} while (bio);
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1774
  	current->bio_list = NULL; /* deactivate */
d89d87965   Neil Brown   When stacked bloc...
1775
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1776
1777
1778
  EXPORT_SYMBOL(generic_make_request);
  
  /**
710027a48   Randy Dunlap   Add some block/ s...
1779
   * submit_bio - submit a bio to the block device layer for I/O
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1780
1781
1782
1783
1784
   * @rw: whether to %READ or %WRITE, or maybe to %READA (read ahead)
   * @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   Randy Dunlap   Add some block/ s...
1785
   * interfaces; @bio must be presetup and ready for I/O.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1786
1787
1788
1789
   *
   */
  void submit_bio(int rw, struct bio *bio)
  {
22e2c507c   Jens Axboe   [PATCH] Update cf...
1790
  	bio->bi_rw |= rw;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1791

bf2de6f5a   Jens Axboe   block: Initial su...
1792
1793
1794
1795
  	/*
  	 * If it's a regular read/write or a barrier with data attached,
  	 * go through the normal accounting stuff before submission.
  	 */
e2a60da74   Martin K. Petersen   block: Clean up s...
1796
  	if (bio_has_data(bio)) {
4363ac7c1   Martin K. Petersen   block: Implement ...
1797
1798
1799
1800
1801
1802
  		unsigned int count;
  
  		if (unlikely(rw & REQ_WRITE_SAME))
  			count = bdev_logical_block_size(bio->bi_bdev) >> 9;
  		else
  			count = bio_sectors(bio);
bf2de6f5a   Jens Axboe   block: Initial su...
1803
1804
1805
  		if (rw & WRITE) {
  			count_vm_events(PGPGOUT, count);
  		} else {
4f024f379   Kent Overstreet   block: Abstract o...
1806
  			task_io_account_read(bio->bi_iter.bi_size);
bf2de6f5a   Jens Axboe   block: Initial su...
1807
1808
1809
1810
1811
  			count_vm_events(PGPGIN, count);
  		}
  
  		if (unlikely(block_dump)) {
  			char b[BDEVNAME_SIZE];
8dcbdc742   San Mehat   block: block_dump...
1812
1813
  			printk(KERN_DEBUG "%s(%d): %s block %Lu on %s (%u sectors)
  ",
ba25f9dcc   Pavel Emelyanov   Use helpers to ob...
1814
  			current->comm, task_pid_nr(current),
bf2de6f5a   Jens Axboe   block: Initial su...
1815
  				(rw & WRITE) ? "WRITE" : "READ",
4f024f379   Kent Overstreet   block: Abstract o...
1816
  				(unsigned long long)bio->bi_iter.bi_sector,
8dcbdc742   San Mehat   block: block_dump...
1817
1818
  				bdevname(bio->bi_bdev, b),
  				count);
bf2de6f5a   Jens Axboe   block: Initial su...
1819
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1820
1821
1822
1823
  	}
  
  	generic_make_request(bio);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1824
  EXPORT_SYMBOL(submit_bio);
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
1825
  /**
82124d603   Kiyoshi Ueda   block: add reques...
1826
1827
1828
1829
1830
1831
1832
1833
1834
1835
1836
1837
   * blk_rq_check_limits - Helper function to check a request for the queue limit
   * @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.
   *
   *    This function should also be useful for request stacking drivers
eef35c2d4   Stefan Weil   Fix spelling fuct...
1838
   *    in some cases below, so export this function.
82124d603   Kiyoshi Ueda   block: add reques...
1839
1840
1841
1842
1843
1844
1845
1846
1847
   *    Request stacking drivers like request-based dm may change the queue
   *    limits while requests are in the queue (e.g. dm's table swapping).
   *    Such request stacking drivers should check those requests agaist
   *    the new queue limits again when they dispatch those requests,
   *    although such checkings are also done against the old queue limits
   *    when submitting requests.
   */
  int blk_rq_check_limits(struct request_queue *q, struct request *rq)
  {
e2a60da74   Martin K. Petersen   block: Clean up s...
1848
  	if (!rq_mergeable(rq))
3383977fa   ike Snitzer   block: update req...
1849
  		return 0;
f31dc1cd4   Martin K. Petersen   block: Consolidat...
1850
  	if (blk_rq_sectors(rq) > blk_queue_get_max_sectors(q, rq->cmd_flags)) {
82124d603   Kiyoshi Ueda   block: add reques...
1851
1852
1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
  		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   Martin K. Petersen   block: Consolidat...
1863
  	if (rq->nr_phys_segments > queue_max_segments(q)) {
82124d603   Kiyoshi Ueda   block: add reques...
1864
1865
1866
1867
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
  		printk(KERN_ERR "%s: over max segments limit.
  ", __func__);
  		return -EIO;
  	}
  
  	return 0;
  }
  EXPORT_SYMBOL_GPL(blk_rq_check_limits);
  
  /**
   * blk_insert_cloned_request - Helper for stacking drivers to submit a request
   * @q:  the queue to submit the request
   * @rq: the request being queued
   */
  int blk_insert_cloned_request(struct request_queue *q, struct request *rq)
  {
  	unsigned long flags;
4853abaae   Jeff Moyer   block: fix flush ...
1881
  	int where = ELEVATOR_INSERT_BACK;
82124d603   Kiyoshi Ueda   block: add reques...
1882
1883
1884
  
  	if (blk_rq_check_limits(q, rq))
  		return -EIO;
b2c9cd379   Akinobu Mita   fail_make_request...
1885
1886
  	if (rq->rq_disk &&
  	    should_fail_request(&rq->rq_disk->part0, blk_rq_bytes(rq)))
82124d603   Kiyoshi Ueda   block: add reques...
1887
  		return -EIO;
82124d603   Kiyoshi Ueda   block: add reques...
1888
1889
  
  	spin_lock_irqsave(q->queue_lock, flags);
3f3299d5c   Bart Van Assche   block: Rename que...
1890
  	if (unlikely(blk_queue_dying(q))) {
8ba61435d   Tejun Heo   block: add missin...
1891
1892
1893
  		spin_unlock_irqrestore(q->queue_lock, flags);
  		return -ENODEV;
  	}
82124d603   Kiyoshi Ueda   block: add reques...
1894
1895
1896
1897
1898
1899
  
  	/*
  	 * Submitting request must be dequeued before calling this function
  	 * because it will be linked to another request_queue
  	 */
  	BUG_ON(blk_queued_rq(rq));
4853abaae   Jeff Moyer   block: fix flush ...
1900
1901
1902
1903
  	if (rq->cmd_flags & (REQ_FLUSH|REQ_FUA))
  		where = ELEVATOR_INSERT_FLUSH;
  
  	add_acct_request(q, rq, where);
e67b77c79   Jeff Moyer   blk-flush: move t...
1904
1905
  	if (where == ELEVATOR_INSERT_FLUSH)
  		__blk_run_queue(q);
82124d603   Kiyoshi Ueda   block: add reques...
1906
1907
1908
1909
1910
  	spin_unlock_irqrestore(q->queue_lock, flags);
  
  	return 0;
  }
  EXPORT_SYMBOL_GPL(blk_insert_cloned_request);
80a761fd3   Tejun Heo   block: implement ...
1911
1912
1913
1914
1915
1916
1917
1918
1919
1920
1921
1922
1923
1924
1925
1926
1927
1928
1929
1930
1931
1932
1933
1934
1935
1936
1937
1938
1939
1940
1941
1942
1943
1944
1945
  /**
   * 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.
   *
   * Context:
   *     queue_lock must be held.
   */
  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;
  
  	if (!(rq->cmd_flags & REQ_MIXED_MERGE))
  		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) {
  		if ((bio->bi_rw & ff) != ff)
  			break;
4f024f379   Kent Overstreet   block: Abstract o...
1946
  		bytes += bio->bi_iter.bi_size;
80a761fd3   Tejun Heo   block: implement ...
1947
1948
1949
1950
1951
1952
1953
  	}
  
  	/* this could lead to infinite loop */
  	BUG_ON(blk_rq_bytes(rq) && !bytes);
  	return bytes;
  }
  EXPORT_SYMBOL_GPL(blk_rq_err_bytes);
320ae51fe   Jens Axboe   blk-mq: new multi...
1954
  void blk_account_io_completion(struct request *req, unsigned int bytes)
bc58ba946   Jens Axboe   block: add sysfs ...
1955
  {
c2553b584   Jens Axboe   block: make blk_d...
1956
  	if (blk_do_io_stat(req)) {
bc58ba946   Jens Axboe   block: add sysfs ...
1957
1958
1959
1960
1961
  		const int rw = rq_data_dir(req);
  		struct hd_struct *part;
  		int cpu;
  
  		cpu = part_stat_lock();
09e099d4b   Jerome Marchand   block: fix accoun...
1962
  		part = req->part;
bc58ba946   Jens Axboe   block: add sysfs ...
1963
1964
1965
1966
  		part_stat_add(cpu, part, sectors[rw], bytes >> 9);
  		part_stat_unlock();
  	}
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1967
  void blk_account_io_done(struct request *req)
bc58ba946   Jens Axboe   block: add sysfs ...
1968
  {
bc58ba946   Jens Axboe   block: add sysfs ...
1969
  	/*
dd4c133f3   Tejun Heo   block: rename bar...
1970
1971
1972
  	 * Account IO completion.  flush_rq isn't accounted as a
  	 * normal IO on queueing nor completion.  Accounting the
  	 * containing request is enough.
bc58ba946   Jens Axboe   block: add sysfs ...
1973
  	 */
414b4ff5e   Tejun Heo   block: add REQ_FL...
1974
  	if (blk_do_io_stat(req) && !(req->cmd_flags & REQ_FLUSH_SEQ)) {
bc58ba946   Jens Axboe   block: add sysfs ...
1975
1976
1977
1978
1979
1980
  		unsigned long duration = jiffies - req->start_time;
  		const int rw = rq_data_dir(req);
  		struct hd_struct *part;
  		int cpu;
  
  		cpu = part_stat_lock();
09e099d4b   Jerome Marchand   block: fix accoun...
1981
  		part = req->part;
bc58ba946   Jens Axboe   block: add sysfs ...
1982
1983
1984
1985
  
  		part_stat_inc(cpu, part, ios[rw]);
  		part_stat_add(cpu, part, ticks[rw], duration);
  		part_round_stats(cpu, part);
316d315bf   Nikanth Karthikesan   block: Seperate r...
1986
  		part_dec_in_flight(part, rw);
bc58ba946   Jens Axboe   block: add sysfs ...
1987

6c23a9681   Jens Axboe   block: add intern...
1988
  		hd_struct_put(part);
bc58ba946   Jens Axboe   block: add sysfs ...
1989
1990
1991
  		part_stat_unlock();
  	}
  }
c8158819d   Lin Ming   block: implement ...
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
  #ifdef CONFIG_PM_RUNTIME
  /*
   * Don't process normal requests when queue is suspended
   * or in the process of suspending/resuming
   */
  static struct request *blk_pm_peek_request(struct request_queue *q,
  					   struct request *rq)
  {
  	if (q->dev && (q->rpm_status == RPM_SUSPENDED ||
  	    (q->rpm_status != RPM_ACTIVE && !(rq->cmd_flags & REQ_PM))))
  		return NULL;
  	else
  		return rq;
  }
  #else
  static inline struct request *blk_pm_peek_request(struct request_queue *q,
  						  struct request *rq)
  {
  	return rq;
  }
  #endif
320ae51fe   Jens Axboe   blk-mq: new multi...
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
  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);
  		}
  		part_round_stats(cpu, part);
  		part_inc_in_flight(part, rw);
  		rq->part = part;
  	}
  
  	part_stat_unlock();
  }
53a08807c   Tejun Heo   block: internal d...
2048
  /**
9934c8c04   Tejun Heo   block: implement ...
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
   * 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.
   *
   * Context:
   *     queue_lock must be held.
   */
  struct request *blk_peek_request(struct request_queue *q)
158dbda00   Tejun Heo   block: reorganize...
2065
2066
2067
2068
2069
  {
  	struct request *rq;
  	int ret;
  
  	while ((rq = __elv_next_request(q)) != NULL) {
c8158819d   Lin Ming   block: implement ...
2070
2071
2072
2073
  
  		rq = blk_pm_peek_request(q, rq);
  		if (!rq)
  			break;
158dbda00   Tejun Heo   block: reorganize...
2074
2075
2076
2077
2078
2079
  		if (!(rq->cmd_flags & REQ_STARTED)) {
  			/*
  			 * This is the first time the device driver
  			 * sees this request (possibly after
  			 * requeueing).  Notify IO scheduler.
  			 */
33659ebba   Christoph Hellwig   block: remove wra...
2080
  			if (rq->cmd_flags & REQ_SORTED)
158dbda00   Tejun Heo   block: reorganize...
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
  				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
  			 */
  			rq->cmd_flags |= REQ_STARTED;
  			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;
  		}
  
  		if (rq->cmd_flags & REQ_DONTPREP)
  			break;
2e46e8b27   Tejun Heo   block: drop reque...
2099
  		if (q->dma_drain_size && blk_rq_bytes(rq)) {
158dbda00   Tejun Heo   block: reorganize...
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
2118
2119
2120
2121
  			/*
  			 * 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
  			 * avoid resource deadlock.  REQ_STARTED will
  			 * prevent other fs requests from passing this one.
  			 */
2e46e8b27   Tejun Heo   block: drop reque...
2122
  			if (q->dma_drain_size && blk_rq_bytes(rq) &&
158dbda00   Tejun Heo   block: reorganize...
2123
2124
2125
2126
2127
2128
2129
2130
2131
2132
2133
2134
  			    !(rq->cmd_flags & REQ_DONTPREP)) {
  				/*
  				 * remove the space for the drain we added
  				 * so that we don't add it again
  				 */
  				--rq->nr_phys_segments;
  			}
  
  			rq = NULL;
  			break;
  		} else if (ret == BLKPREP_KILL) {
  			rq->cmd_flags |= REQ_QUIET;
c143dc903   James Bottomley   block: fix an oop...
2135
2136
2137
2138
2139
  			/*
  			 * Mark this request as started so we don't trigger
  			 * any debug logic in the end I/O path.
  			 */
  			blk_start_request(rq);
40cbbb781   Tejun Heo   block: implement ...
2140
  			__blk_end_request_all(rq, -EIO);
158dbda00   Tejun Heo   block: reorganize...
2141
2142
2143
2144
2145
2146
2147
2148
2149
  		} else {
  			printk(KERN_ERR "%s: bad return=%d
  ", __func__, ret);
  			break;
  		}
  	}
  
  	return rq;
  }
9934c8c04   Tejun Heo   block: implement ...
2150
  EXPORT_SYMBOL(blk_peek_request);
158dbda00   Tejun Heo   block: reorganize...
2151

9934c8c04   Tejun Heo   block: implement ...
2152
  void blk_dequeue_request(struct request *rq)
158dbda00   Tejun Heo   block: reorganize...
2153
  {
9934c8c04   Tejun Heo   block: implement ...
2154
  	struct request_queue *q = rq->q;
158dbda00   Tejun Heo   block: reorganize...
2155
2156
2157
2158
2159
2160
2161
2162
2163
2164
  	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.
  	 */
9195291e5   Divyesh Shah   blkio: Increment ...
2165
  	if (blk_account_rq(rq)) {
0a7ae2ff0   Jens Axboe   block: change the...
2166
  		q->in_flight[rq_is_sync(rq)]++;
9195291e5   Divyesh Shah   blkio: Increment ...
2167
2168
  		set_io_start_time_ns(rq);
  	}
158dbda00   Tejun Heo   block: reorganize...
2169
  }
5efccd17c   Tejun Heo   block: reorder re...
2170
  /**
9934c8c04   Tejun Heo   block: implement ...
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
   * 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.
   *
   *     Block internal functions which don't want to start timer should
   *     call blk_dequeue_request().
   *
   * Context:
   *     queue_lock must be held.
   */
  void blk_start_request(struct request *req)
  {
  	blk_dequeue_request(req);
  
  	/*
5f49f6317   Tejun Heo   block: set rq->re...
2189
2190
  	 * We are now handing the request to the hardware, initialize
  	 * resid_len to full count and add the timeout handler.
9934c8c04   Tejun Heo   block: implement ...
2191
  	 */
5f49f6317   Tejun Heo   block: set rq->re...
2192
  	req->resid_len = blk_rq_bytes(req);
dbb66c4be   FUJITA Tomonori   block: needs to s...
2193
2194
  	if (unlikely(blk_bidi_rq(req)))
  		req->next_rq->resid_len = blk_rq_bytes(req->next_rq);
4912aa6c1   Jeff Moyer   block: fix race b...
2195
  	BUG_ON(test_bit(REQ_ATOM_COMPLETE, &req->atomic_flags));
9934c8c04   Tejun Heo   block: implement ...
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
2219
2220
2221
2222
2223
2224
2225
2226
  	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.
   *
   * Context:
   *     queue_lock must be held.
   */
  struct request *blk_fetch_request(struct request_queue *q)
  {
  	struct request *rq;
  
  	rq = blk_peek_request(q);
  	if (rq)
  		blk_start_request(rq);
  	return rq;
  }
  EXPORT_SYMBOL(blk_fetch_request);
  
  /**
2e60e0229   Tejun Heo   block: clean up r...
2227
   * blk_update_request - Special helper function for request stacking drivers
8ebf97560   Randy Dunlap   block: fix kernel...
2228
   * @req:      the request being processed
710027a48   Randy Dunlap   Add some block/ s...
2229
   * @error:    %0 for success, < %0 for error
8ebf97560   Randy Dunlap   block: fix kernel...
2230
   * @nr_bytes: number of bytes to complete @req
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
2231
2232
   *
   * Description:
8ebf97560   Randy Dunlap   block: fix kernel...
2233
2234
2235
   *     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   Tejun Heo   block: clean up r...
2236
2237
2238
2239
2240
2241
2242
   *
   *     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   Kiyoshi Ueda   blk_end_request: ...
2243
2244
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2245
2246
   *     %false - this request doesn't have any more data
   *     %true  - this request has more data
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
2247
   **/
2e60e0229   Tejun Heo   block: clean up r...
2248
  bool blk_update_request(struct request *req, int error, unsigned int nr_bytes)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2249
  {
f79ea4161   Kent Overstreet   block: Refactor b...
2250
  	int total_bytes;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2251

2e60e0229   Tejun Heo   block: clean up r...
2252
2253
  	if (!req->bio)
  		return false;
5f3ea37c7   Arnaldo Carvalho de Melo   blktrace: port to...
2254
  	trace_block_rq_complete(req->q, req);
2056a782f   Jens Axboe   [PATCH] Block que...
2255

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2256
  	/*
6f41469c6   Tejun Heo   block: clear req-...
2257
2258
2259
2260
2261
2262
  	 * For fs requests, rq is just carrier of independent bio's
  	 * and each partial completion should be handled separately.
  	 * Reset per-request error on each partial completion.
  	 *
  	 * TODO: tj: This is too subtle.  It would be better to let
  	 * low level drivers do what they see fit.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2263
  	 */
33659ebba   Christoph Hellwig   block: remove wra...
2264
  	if (req->cmd_type == REQ_TYPE_FS)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2265
  		req->errors = 0;
33659ebba   Christoph Hellwig   block: remove wra...
2266
2267
  	if (error && req->cmd_type == REQ_TYPE_FS &&
  	    !(req->cmd_flags & REQ_QUIET)) {
79775567e   Hannes Reinecke   [SCSI] block: imp...
2268
2269
2270
2271
2272
2273
2274
2275
2276
2277
2278
2279
  		char *error_type;
  
  		switch (error) {
  		case -ENOLINK:
  			error_type = "recoverable transport";
  			break;
  		case -EREMOTEIO:
  			error_type = "critical target";
  			break;
  		case -EBADE:
  			error_type = "critical nexus";
  			break;
d1ffc1f86   Hannes Reinecke   block/dasd: detai...
2280
2281
2282
  		case -ETIMEDOUT:
  			error_type = "timeout";
  			break;
a9d6ceb83   Hannes Reinecke   [SCSI] return ENO...
2283
2284
2285
  		case -ENOSPC:
  			error_type = "critical space allocation";
  			break;
7e782af57   Hannes Reinecke   [SCSI] Return ENO...
2286
2287
2288
  		case -ENODATA:
  			error_type = "critical medium";
  			break;
79775567e   Hannes Reinecke   [SCSI] block: imp...
2289
2290
2291
2292
2293
  		case -EIO:
  		default:
  			error_type = "I/O";
  			break;
  		}
37d7b34f0   Yi Zou   block: rate-limit...
2294
2295
2296
2297
2298
  		printk_ratelimited(KERN_ERR "end_request: %s error, dev %s, sector %llu
  ",
  				   error_type, req->rq_disk ?
  				   req->rq_disk->disk_name : "?",
  				   (unsigned long long)blk_rq_pos(req));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2299
  	}
bc58ba946   Jens Axboe   block: add sysfs ...
2300
  	blk_account_io_completion(req, nr_bytes);
d72d904a5   Jens Axboe   [BLOCK] Update re...
2301

f79ea4161   Kent Overstreet   block: Refactor b...
2302
2303
2304
  	total_bytes = 0;
  	while (req->bio) {
  		struct bio *bio = req->bio;
4f024f379   Kent Overstreet   block: Abstract o...
2305
  		unsigned bio_bytes = min(bio->bi_iter.bi_size, nr_bytes);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2306

4f024f379   Kent Overstreet   block: Abstract o...
2307
  		if (bio_bytes == bio->bi_iter.bi_size)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2308
  			req->bio = bio->bi_next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2309

f79ea4161   Kent Overstreet   block: Refactor b...
2310
  		req_bio_endio(req, bio, bio_bytes, error);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2311

f79ea4161   Kent Overstreet   block: Refactor b...
2312
2313
  		total_bytes += bio_bytes;
  		nr_bytes -= bio_bytes;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2314

f79ea4161   Kent Overstreet   block: Refactor b...
2315
2316
  		if (!nr_bytes)
  			break;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2317
2318
2319
2320
2321
  	}
  
  	/*
  	 * completely done
  	 */
2e60e0229   Tejun Heo   block: clean up r...
2322
2323
2324
2325
2326
2327
  	if (!req->bio) {
  		/*
  		 * Reset counters so that the request stacking driver
  		 * can find how many bytes remain in the request
  		 * later.
  		 */
a2dec7b36   Tejun Heo   block: hide reque...
2328
  		req->__data_len = 0;
2e60e0229   Tejun Heo   block: clean up r...
2329
2330
  		return false;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2331

a2dec7b36   Tejun Heo   block: hide reque...
2332
  	req->__data_len -= total_bytes;
2e46e8b27   Tejun Heo   block: drop reque...
2333
2334
2335
  	req->buffer = bio_data(req->bio);
  
  	/* update sector only for requests with clear definition of sector */
e2a60da74   Martin K. Petersen   block: Clean up s...
2336
  	if (req->cmd_type == REQ_TYPE_FS)
a2dec7b36   Tejun Heo   block: hide reque...
2337
  		req->__sector += total_bytes >> 9;
2e46e8b27   Tejun Heo   block: drop reque...
2338

80a761fd3   Tejun Heo   block: implement ...
2339
2340
2341
2342
2343
  	/* mixed attributes always follow the first bio */
  	if (req->cmd_flags & REQ_MIXED_MERGE) {
  		req->cmd_flags &= ~REQ_FAILFAST_MASK;
  		req->cmd_flags |= req->bio->bi_rw & REQ_FAILFAST_MASK;
  	}
2e46e8b27   Tejun Heo   block: drop reque...
2344
2345
2346
2347
2348
  	/*
  	 * 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)) {
8182924bc   Jens Axboe   block: dump reque...
2349
  		blk_dump_rq_flags(req, "request botched");
a2dec7b36   Tejun Heo   block: hide reque...
2350
  		req->__data_len = blk_rq_cur_bytes(req);
2e46e8b27   Tejun Heo   block: drop reque...
2351
2352
2353
  	}
  
  	/* recalculate the number of segments */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2354
  	blk_recalc_rq_segments(req);
2e46e8b27   Tejun Heo   block: drop reque...
2355

2e60e0229   Tejun Heo   block: clean up r...
2356
  	return true;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2357
  }
2e60e0229   Tejun Heo   block: clean up r...
2358
  EXPORT_SYMBOL_GPL(blk_update_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2359

2e60e0229   Tejun Heo   block: clean up r...
2360
2361
2362
  static bool blk_update_bidi_request(struct request *rq, int error,
  				    unsigned int nr_bytes,
  				    unsigned int bidi_bytes)
5efccd17c   Tejun Heo   block: reorder re...
2363
  {
2e60e0229   Tejun Heo   block: clean up r...
2364
2365
  	if (blk_update_request(rq, error, nr_bytes))
  		return true;
5efccd17c   Tejun Heo   block: reorder re...
2366

2e60e0229   Tejun Heo   block: clean up r...
2367
2368
2369
2370
  	/* 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   Tejun Heo   block: reorder re...
2371

e2e1a148b   Jens Axboe   block: add sysfs ...
2372
2373
  	if (blk_queue_add_random(rq->q))
  		add_disk_randomness(rq->rq_disk);
2e60e0229   Tejun Heo   block: clean up r...
2374
2375
  
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2376
  }
28018c242   James Bottomley   block: implement ...
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
  /**
   * 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;
  
  	req->cmd_flags &= ~REQ_DONTPREP;
  	if (q->unprep_rq_fn)
  		q->unprep_rq_fn(q, req);
  }
  EXPORT_SYMBOL_GPL(blk_unprep_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2396
2397
2398
  /*
   * queue lock must be held
   */
2e60e0229   Tejun Heo   block: clean up r...
2399
  static void blk_finish_request(struct request *req, int error)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2400
  {
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2401
2402
  	if (blk_rq_tagged(req))
  		blk_queue_end_tag(req->q, req);
ba396a6c1   James Bottomley   block: fix oops w...
2403
  	BUG_ON(blk_queued_rq(req));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2404

33659ebba   Christoph Hellwig   block: remove wra...
2405
  	if (unlikely(laptop_mode) && req->cmd_type == REQ_TYPE_FS)
31373d09d   Matthew Garrett   laptop-mode: Make...
2406
  		laptop_io_completion(&req->q->backing_dev_info);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2407

e78042e5b   Mike Anderson   blk: move blk_del...
2408
  	blk_delete_timer(req);
28018c242   James Bottomley   block: implement ...
2409
2410
  	if (req->cmd_flags & REQ_DONTPREP)
  		blk_unprep_request(req);
bc58ba946   Jens Axboe   block: add sysfs ...
2411
  	blk_account_io_done(req);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2412

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2413
  	if (req->end_io)
8ffdc6550   Tejun Heo   [BLOCK] add @upto...
2414
  		req->end_io(req, error);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2415
2416
2417
  	else {
  		if (blk_bidi_rq(req))
  			__blk_put_request(req->next_rq->q, req->next_rq);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2418
  		__blk_put_request(req->q, req);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2419
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2420
  }
3b11313a6   Kiyoshi Ueda   blk_end_request: ...
2421
  /**
2e60e0229   Tejun Heo   block: clean up r...
2422
2423
2424
2425
2426
   * blk_end_bidi_request - Complete a bidi request
   * @rq:         the request to complete
   * @error:      %0 for success, < %0 for error
   * @nr_bytes:   number of bytes to complete @rq
   * @bidi_bytes: number of bytes to complete @rq->next_rq
a0cd12854   Jens Axboe   block: add end_qu...
2427
2428
   *
   * Description:
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2429
   *     Ends I/O on a number of bytes attached to @rq and @rq->next_rq.
2e60e0229   Tejun Heo   block: clean up r...
2430
2431
2432
   *     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   Kiyoshi Ueda   blk_end_request: ...
2433
2434
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2435
2436
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
a0cd12854   Jens Axboe   block: add end_qu...
2437
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2438
  static bool blk_end_bidi_request(struct request *rq, int error,
32fab448e   Kiyoshi Ueda   block: add reques...
2439
2440
  				 unsigned int nr_bytes, unsigned int bidi_bytes)
  {
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2441
  	struct request_queue *q = rq->q;
2e60e0229   Tejun Heo   block: clean up r...
2442
  	unsigned long flags;
32fab448e   Kiyoshi Ueda   block: add reques...
2443

2e60e0229   Tejun Heo   block: clean up r...
2444
2445
  	if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes))
  		return true;
32fab448e   Kiyoshi Ueda   block: add reques...
2446

336cdb400   Kiyoshi Ueda   blk_end_request: ...
2447
  	spin_lock_irqsave(q->queue_lock, flags);
2e60e0229   Tejun Heo   block: clean up r...
2448
  	blk_finish_request(rq, error);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2449
  	spin_unlock_irqrestore(q->queue_lock, flags);
2e60e0229   Tejun Heo   block: clean up r...
2450
  	return false;
32fab448e   Kiyoshi Ueda   block: add reques...
2451
  }
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2452
  /**
2e60e0229   Tejun Heo   block: clean up r...
2453
2454
   * __blk_end_bidi_request - Complete a bidi request with queue lock held
   * @rq:         the request to complete
710027a48   Randy Dunlap   Add some block/ s...
2455
   * @error:      %0 for success, < %0 for error
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2456
2457
   * @nr_bytes:   number of bytes to complete @rq
   * @bidi_bytes: number of bytes to complete @rq->next_rq
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2458
2459
   *
   * Description:
2e60e0229   Tejun Heo   block: clean up r...
2460
2461
   *     Identical to blk_end_bidi_request() except that queue lock is
   *     assumed to be locked on entry and remains so on return.
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2462
2463
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2464
2465
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2466
   **/
4853abaae   Jeff Moyer   block: fix flush ...
2467
  bool __blk_end_bidi_request(struct request *rq, int error,
b1f744937   FUJITA Tomonori   block: move compl...
2468
  				   unsigned int nr_bytes, unsigned int bidi_bytes)
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2469
  {
2e60e0229   Tejun Heo   block: clean up r...
2470
2471
  	if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes))
  		return true;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2472

2e60e0229   Tejun Heo   block: clean up r...
2473
  	blk_finish_request(rq, error);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2474

2e60e0229   Tejun Heo   block: clean up r...
2475
  	return false;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2476
  }
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2477
2478
2479
2480
  
  /**
   * blk_end_request - Helper function for drivers to complete the request.
   * @rq:       the request being processed
710027a48   Randy Dunlap   Add some block/ s...
2481
   * @error:    %0 for success, < %0 for error
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2482
2483
2484
2485
2486
2487
2488
   * @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   FUJITA Tomonori   block: move compl...
2489
2490
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2491
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2492
  bool blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2493
  {
b1f744937   FUJITA Tomonori   block: move compl...
2494
  	return blk_end_bidi_request(rq, error, nr_bytes, 0);
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2495
  }
56ad1740d   Jens Axboe   block: make the e...
2496
  EXPORT_SYMBOL(blk_end_request);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2497
2498
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2499
2500
   * blk_end_request_all - Helper function for drives to finish the request.
   * @rq: the request to finish
8ebf97560   Randy Dunlap   block: fix kernel...
2501
   * @error: %0 for success, < %0 for error
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2502
2503
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2504
2505
2506
   *     Completely finish @rq.
   */
  void blk_end_request_all(struct request *rq, int error)
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2507
  {
b1f744937   FUJITA Tomonori   block: move compl...
2508
2509
  	bool pending;
  	unsigned int bidi_bytes = 0;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2510

b1f744937   FUJITA Tomonori   block: move compl...
2511
2512
  	if (unlikely(blk_bidi_rq(rq)))
  		bidi_bytes = blk_rq_bytes(rq->next_rq);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2513

b1f744937   FUJITA Tomonori   block: move compl...
2514
2515
2516
  	pending = blk_end_bidi_request(rq, error, blk_rq_bytes(rq), bidi_bytes);
  	BUG_ON(pending);
  }
56ad1740d   Jens Axboe   block: make the e...
2517
  EXPORT_SYMBOL(blk_end_request_all);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2518

b1f744937   FUJITA Tomonori   block: move compl...
2519
2520
2521
  /**
   * blk_end_request_cur - Helper function to finish the current request chunk.
   * @rq: the request to finish the current chunk for
8ebf97560   Randy Dunlap   block: fix kernel...
2522
   * @error: %0 for success, < %0 for error
b1f744937   FUJITA Tomonori   block: move compl...
2523
2524
2525
2526
2527
2528
2529
2530
2531
2532
2533
   *
   * Description:
   *     Complete the current consecutively mapped chunk from @rq.
   *
   * Return:
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
   */
  bool blk_end_request_cur(struct request *rq, int error)
  {
  	return blk_end_request(rq, error, blk_rq_cur_bytes(rq));
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2534
  }
56ad1740d   Jens Axboe   block: make the e...
2535
  EXPORT_SYMBOL(blk_end_request_cur);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2536

e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2537
  /**
80a761fd3   Tejun Heo   block: implement ...
2538
2539
2540
2541
2542
2543
2544
2545
2546
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
   * blk_end_request_err - Finish a request till the next failure boundary.
   * @rq: the request to finish till the next failure boundary for
   * @error: must be negative errno
   *
   * Description:
   *     Complete @rq till the next failure boundary.
   *
   * Return:
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
   */
  bool blk_end_request_err(struct request *rq, int error)
  {
  	WARN_ON(error >= 0);
  	return blk_end_request(rq, error, blk_rq_err_bytes(rq));
  }
  EXPORT_SYMBOL_GPL(blk_end_request_err);
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2557
2558
2559
2560
   * __blk_end_request - Helper function for drivers to complete the request.
   * @rq:       the request being processed
   * @error:    %0 for success, < %0 for error
   * @nr_bytes: number of bytes to complete
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2561
2562
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2563
   *     Must be called with queue lock held unlike blk_end_request().
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2564
2565
   *
   * Return:
b1f744937   FUJITA Tomonori   block: move compl...
2566
2567
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2568
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2569
  bool __blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2570
  {
b1f744937   FUJITA Tomonori   block: move compl...
2571
  	return __blk_end_bidi_request(rq, error, nr_bytes, 0);
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2572
  }
56ad1740d   Jens Axboe   block: make the e...
2573
  EXPORT_SYMBOL(__blk_end_request);
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2574
2575
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2576
2577
   * __blk_end_request_all - Helper function for drives to finish the request.
   * @rq: the request to finish
8ebf97560   Randy Dunlap   block: fix kernel...
2578
   * @error: %0 for success, < %0 for error
32fab448e   Kiyoshi Ueda   block: add reques...
2579
2580
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2581
   *     Completely finish @rq.  Must be called with queue lock held.
32fab448e   Kiyoshi Ueda   block: add reques...
2582
   */
b1f744937   FUJITA Tomonori   block: move compl...
2583
  void __blk_end_request_all(struct request *rq, int error)
32fab448e   Kiyoshi Ueda   block: add reques...
2584
  {
b1f744937   FUJITA Tomonori   block: move compl...
2585
2586
2587
2588
2589
2590
2591
2592
  	bool pending;
  	unsigned int bidi_bytes = 0;
  
  	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   Kiyoshi Ueda   block: add reques...
2593
  }
56ad1740d   Jens Axboe   block: make the e...
2594
  EXPORT_SYMBOL(__blk_end_request_all);
32fab448e   Kiyoshi Ueda   block: add reques...
2595
2596
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2597
2598
   * __blk_end_request_cur - Helper function to finish the current request chunk.
   * @rq: the request to finish the current chunk for
8ebf97560   Randy Dunlap   block: fix kernel...
2599
   * @error: %0 for success, < %0 for error
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2600
2601
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2602
2603
   *     Complete the current consecutively mapped chunk from @rq.  Must
   *     be called with queue lock held.
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2604
2605
   *
   * Return:
b1f744937   FUJITA Tomonori   block: move compl...
2606
2607
2608
2609
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
   */
  bool __blk_end_request_cur(struct request *rq, int error)
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2610
  {
b1f744937   FUJITA Tomonori   block: move compl...
2611
  	return __blk_end_request(rq, error, blk_rq_cur_bytes(rq));
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2612
  }
56ad1740d   Jens Axboe   block: make the e...
2613
  EXPORT_SYMBOL(__blk_end_request_cur);
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2614

80a761fd3   Tejun Heo   block: implement ...
2615
2616
2617
2618
2619
2620
2621
2622
2623
2624
2625
2626
2627
2628
2629
2630
2631
2632
2633
  /**
   * __blk_end_request_err - Finish a request till the next failure boundary.
   * @rq: the request to finish till the next failure boundary for
   * @error: must be negative errno
   *
   * Description:
   *     Complete @rq till the next failure boundary.  Must be called
   *     with queue lock held.
   *
   * Return:
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
   */
  bool __blk_end_request_err(struct request *rq, int error)
  {
  	WARN_ON(error >= 0);
  	return __blk_end_request(rq, error, blk_rq_err_bytes(rq));
  }
  EXPORT_SYMBOL_GPL(__blk_end_request_err);
86db1e297   Jens Axboe   block: continue l...
2634
2635
  void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
  		     struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2636
  {
a82afdfcb   Tejun Heo   block: use the sa...
2637
  	/* Bit 0 (R/W) is identical in rq->cmd_flags and bio->bi_rw */
7b6d91dae   Christoph Hellwig   block: unify flag...
2638
  	rq->cmd_flags |= bio->bi_rw & REQ_WRITE;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2639

fb2dce862   David Woodhouse   Add 'discard' req...
2640
2641
  	if (bio_has_data(bio)) {
  		rq->nr_phys_segments = bio_phys_segments(q, bio);
fb2dce862   David Woodhouse   Add 'discard' req...
2642
2643
  		rq->buffer = bio_data(bio);
  	}
4f024f379   Kent Overstreet   block: Abstract o...
2644
  	rq->__data_len = bio->bi_iter.bi_size;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2645
  	rq->bio = rq->biotail = bio;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2646

66846572b   NeilBrown   Stop exporting bl...
2647
2648
2649
  	if (bio->bi_bdev)
  		rq->rq_disk = bio->bi_bdev->bd_disk;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2650

2d4dc890b   Ilya Loginov   block: add helper...
2651
2652
2653
2654
2655
2656
2657
2658
2659
2660
2661
  #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   Kent Overstreet   block: Convert bi...
2662
  	struct bio_vec bvec;
2d4dc890b   Ilya Loginov   block: add helper...
2663
2664
  
  	rq_for_each_segment(bvec, rq, iter)
7988613b0   Kent Overstreet   block: Convert bi...
2665
  		flush_dcache_page(bvec.bv_page);
2d4dc890b   Ilya Loginov   block: add helper...
2666
2667
2668
  }
  EXPORT_SYMBOL_GPL(rq_flush_dcache_pages);
  #endif
ef9e3facd   Kiyoshi Ueda   block: add lld bu...
2669
2670
2671
2672
2673
2674
2675
2676
2677
2678
2679
2680
2681
2682
2683
2684
2685
2686
2687
2688
2689
2690
2691
2692
2693
2694
2695
  /**
   * 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);
b0fd271d5   Kiyoshi Ueda   block: add reques...
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
2721
  /**
   * 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, ->buffer, ->sense) are not copied.
   */
  static void __blk_rq_prep_clone(struct request *dst, struct request *src)
  {
  	dst->cpu = src->cpu;
3a2edd0d6   Tejun Heo   block: make __blk...
2722
  	dst->cmd_flags = (src->cmd_flags & REQ_CLONE_MASK) | REQ_NOMERGE;
b0fd271d5   Kiyoshi Ueda   block: add reques...
2723
2724
2725
2726
2727
2728
2729
2730
2731
2732
2733
2734
2735
2736
2737
2738
2739
2740
2741
2742
2743
2744
2745
2746
2747
2748
2749
2750
2751
2752
2753
2754
2755
2756
2757
2758
2759
2760
2761
2762
  	dst->cmd_type = src->cmd_type;
  	dst->__sector = blk_rq_pos(src);
  	dst->__data_len = blk_rq_bytes(src);
  	dst->nr_phys_segments = src->nr_phys_segments;
  	dst->ioprio = src->ioprio;
  	dst->extra_len = src->extra_len;
  }
  
  /**
   * 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, ->buffer, ->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)
  		bs = fs_bio_set;
  
  	blk_rq_init(NULL, rq);
  
  	__rq_for_each_bio(bio_src, rq_src) {
bf800ef18   Kent Overstreet   block: Add bio_cl...
2763
  		bio = bio_clone_bioset(bio_src, gfp_mask, bs);
b0fd271d5   Kiyoshi Ueda   block: add reques...
2764
2765
  		if (!bio)
  			goto free_and_out;
b0fd271d5   Kiyoshi Ueda   block: add reques...
2766
2767
2768
2769
2770
2771
2772
2773
2774
2775
2776
2777
2778
2779
2780
2781
  		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)
4254bba17   Kent Overstreet   block: Kill bi_de...
2782
  		bio_put(bio);
b0fd271d5   Kiyoshi Ueda   block: add reques...
2783
2784
2785
2786
2787
  	blk_rq_unprep_clone(rq);
  
  	return -ENOMEM;
  }
  EXPORT_SYMBOL_GPL(blk_rq_prep_clone);
18887ad91   Jens Axboe   block: make kbloc...
2788
  int kblockd_schedule_work(struct request_queue *q, struct work_struct *work)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2789
2790
2791
  {
  	return queue_work(kblockd_workqueue, work);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2792
  EXPORT_SYMBOL(kblockd_schedule_work);
e43473b7f   Vivek Goyal   blkio: Core imple...
2793
2794
2795
2796
2797
2798
  int kblockd_schedule_delayed_work(struct request_queue *q,
  			struct delayed_work *dwork, unsigned long delay)
  {
  	return queue_delayed_work(kblockd_workqueue, dwork, delay);
  }
  EXPORT_SYMBOL(kblockd_schedule_delayed_work);
73c101011   Jens Axboe   block: initial pa...
2799
  #define PLUG_MAGIC	0x91827364
75df71362   Suresh Jayaraman   block: document b...
2800
2801
2802
2803
2804
2805
2806
2807
2808
2809
2810
2811
2812
2813
  /**
   * 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   Jens Axboe   block: initial pa...
2814
2815
2816
2817
2818
2819
  void blk_start_plug(struct blk_plug *plug)
  {
  	struct task_struct *tsk = current;
  
  	plug->magic = PLUG_MAGIC;
  	INIT_LIST_HEAD(&plug->list);
320ae51fe   Jens Axboe   blk-mq: new multi...
2820
  	INIT_LIST_HEAD(&plug->mq_list);
048c9374a   NeilBrown   block: Enhance ne...
2821
  	INIT_LIST_HEAD(&plug->cb_list);
73c101011   Jens Axboe   block: initial pa...
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
2832
2833
2834
2835
2836
2837
2838
2839
2840
  
  	/*
  	 * If this is a nested plug, don't actually assign it. It will be
  	 * flushed on its own.
  	 */
  	if (!tsk->plug) {
  		/*
  		 * Store ordering should not be needed here, since a potential
  		 * preempt will imply a full memory barrier
  		 */
  		tsk->plug = plug;
  	}
  }
  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   Jianpeng Ma   block: Add blk_rq...
2841
2842
  	return !(rqa->q < rqb->q ||
  		(rqa->q == rqb->q && blk_rq_pos(rqa) < blk_rq_pos(rqb)));
73c101011   Jens Axboe   block: initial pa...
2843
  }
49cac01e1   Jens Axboe   block: make unplu...
2844
2845
2846
2847
2848
2849
  /*
   * 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   Jens Axboe   block: only force...
2850
  static void queue_unplugged(struct request_queue *q, unsigned int depth,
49cac01e1   Jens Axboe   block: make unplu...
2851
  			    bool from_schedule)
99e22598e   Jens Axboe   block: drop queue...
2852
  	__releases(q->queue_lock)
94b5eb28b   Jens Axboe   block: fixup bloc...
2853
  {
49cac01e1   Jens Axboe   block: make unplu...
2854
  	trace_block_unplug(q, depth, !from_schedule);
99e22598e   Jens Axboe   block: drop queue...
2855

704605711   Bart Van Assche   block: Avoid sche...
2856
  	if (from_schedule)
24ecfbe27   Christoph Hellwig   block: add blk_ru...
2857
  		blk_run_queue_async(q);
704605711   Bart Van Assche   block: Avoid sche...
2858
  	else
24ecfbe27   Christoph Hellwig   block: add blk_ru...
2859
  		__blk_run_queue(q);
704605711   Bart Van Assche   block: Avoid sche...
2860
  	spin_unlock(q->queue_lock);
94b5eb28b   Jens Axboe   block: fixup bloc...
2861
  }
74018dc30   NeilBrown   blk: pass from_sc...
2862
  static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule)
048c9374a   NeilBrown   block: Enhance ne...
2863
2864
  {
  	LIST_HEAD(callbacks);
2a7d5559b   Shaohua Li   block: stack unplug
2865
2866
  	while (!list_empty(&plug->cb_list)) {
  		list_splice_init(&plug->cb_list, &callbacks);
048c9374a   NeilBrown   block: Enhance ne...
2867

2a7d5559b   Shaohua Li   block: stack unplug
2868
2869
  		while (!list_empty(&callbacks)) {
  			struct blk_plug_cb *cb = list_first_entry(&callbacks,
048c9374a   NeilBrown   block: Enhance ne...
2870
2871
  							  struct blk_plug_cb,
  							  list);
2a7d5559b   Shaohua Li   block: stack unplug
2872
  			list_del(&cb->list);
74018dc30   NeilBrown   blk: pass from_sc...
2873
  			cb->callback(cb, from_schedule);
2a7d5559b   Shaohua Li   block: stack unplug
2874
  		}
048c9374a   NeilBrown   block: Enhance ne...
2875
2876
  	}
  }
9cbb17508   NeilBrown   blk: centralize n...
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
2890
2891
2892
2893
2894
2895
2896
2897
2898
2899
2900
  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   Jens Axboe   block: make unplu...
2901
  void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
73c101011   Jens Axboe   block: initial pa...
2902
2903
2904
2905
  {
  	struct request_queue *q;
  	unsigned long flags;
  	struct request *rq;
109b81296   NeilBrown   block: splice plu...
2906
  	LIST_HEAD(list);
94b5eb28b   Jens Axboe   block: fixup bloc...
2907
  	unsigned int depth;
73c101011   Jens Axboe   block: initial pa...
2908
2909
  
  	BUG_ON(plug->magic != PLUG_MAGIC);
74018dc30   NeilBrown   blk: pass from_sc...
2910
  	flush_plug_callbacks(plug, from_schedule);
320ae51fe   Jens Axboe   blk-mq: new multi...
2911
2912
2913
  
  	if (!list_empty(&plug->mq_list))
  		blk_mq_flush_plug_list(plug, from_schedule);
73c101011   Jens Axboe   block: initial pa...
2914
2915
  	if (list_empty(&plug->list))
  		return;
109b81296   NeilBrown   block: splice plu...
2916
  	list_splice_init(&plug->list, &list);
422765c26   Jianpeng Ma   block: Remove sho...
2917
  	list_sort(NULL, &list, plug_rq_cmp);
73c101011   Jens Axboe   block: initial pa...
2918
2919
  
  	q = NULL;
94b5eb28b   Jens Axboe   block: fixup bloc...
2920
  	depth = 0;
188112722   Jens Axboe   block: add commen...
2921
2922
2923
2924
2925
  
  	/*
  	 * Save and disable interrupts here, to avoid doing it for every
  	 * queue lock we have to take.
  	 */
73c101011   Jens Axboe   block: initial pa...
2926
  	local_irq_save(flags);
109b81296   NeilBrown   block: splice plu...
2927
2928
  	while (!list_empty(&list)) {
  		rq = list_entry_rq(list.next);
73c101011   Jens Axboe   block: initial pa...
2929
  		list_del_init(&rq->queuelist);
73c101011   Jens Axboe   block: initial pa...
2930
2931
  		BUG_ON(!rq->q);
  		if (rq->q != q) {
99e22598e   Jens Axboe   block: drop queue...
2932
2933
2934
2935
  			/*
  			 * This drops the queue lock
  			 */
  			if (q)
49cac01e1   Jens Axboe   block: make unplu...
2936
  				queue_unplugged(q, depth, from_schedule);
73c101011   Jens Axboe   block: initial pa...
2937
  			q = rq->q;
94b5eb28b   Jens Axboe   block: fixup bloc...
2938
  			depth = 0;
73c101011   Jens Axboe   block: initial pa...
2939
2940
  			spin_lock(q->queue_lock);
  		}
8ba61435d   Tejun Heo   block: add missin...
2941
2942
2943
2944
  
  		/*
  		 * Short-circuit if @q is dead
  		 */
3f3299d5c   Bart Van Assche   block: Rename que...
2945
  		if (unlikely(blk_queue_dying(q))) {
8ba61435d   Tejun Heo   block: add missin...
2946
2947
2948
  			__blk_end_request_all(rq, -ENODEV);
  			continue;
  		}
73c101011   Jens Axboe   block: initial pa...
2949
2950
2951
  		/*
  		 * rq is already accounted, so use raw insert
  		 */
401a18e92   Jens Axboe   block: fix bug wi...
2952
2953
2954
2955
  		if (rq->cmd_flags & (REQ_FLUSH | REQ_FUA))
  			__elv_add_request(q, rq, ELEVATOR_INSERT_FLUSH);
  		else
  			__elv_add_request(q, rq, ELEVATOR_INSERT_SORT_MERGE);
94b5eb28b   Jens Axboe   block: fixup bloc...
2956
2957
  
  		depth++;
73c101011   Jens Axboe   block: initial pa...
2958
  	}
99e22598e   Jens Axboe   block: drop queue...
2959
2960
2961
2962
  	/*
  	 * This drops the queue lock
  	 */
  	if (q)
49cac01e1   Jens Axboe   block: make unplu...
2963
  		queue_unplugged(q, depth, from_schedule);
73c101011   Jens Axboe   block: initial pa...
2964

73c101011   Jens Axboe   block: initial pa...
2965
2966
  	local_irq_restore(flags);
  }
73c101011   Jens Axboe   block: initial pa...
2967
2968
2969
  
  void blk_finish_plug(struct blk_plug *plug)
  {
f6603783f   Jens Axboe   block: only force...
2970
  	blk_flush_plug_list(plug, false);
73c101011   Jens Axboe   block: initial pa...
2971

88b996cd0   Christoph Hellwig   block: cleanup th...
2972
2973
  	if (plug == current->plug)
  		current->plug = NULL;
73c101011   Jens Axboe   block: initial pa...
2974
  }
88b996cd0   Christoph Hellwig   block: cleanup th...
2975
  EXPORT_SYMBOL(blk_finish_plug);
73c101011   Jens Axboe   block: initial pa...
2976

6c9546675   Lin Ming   block: add runtim...
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
2989
2990
2991
2992
2993
2994
2995
2996
2997
2998
2999
3000
3001
3002
3003
3004
3005
3006
3007
3008
3009
3010
3011
3012
3013
3014
3015
3016
3017
3018
3019
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3030
3031
3032
3033
3034
3035
3036
3037
3038
3039
3040
3041
3042
3043
3044
3045
3046
3047
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
3072
3073
3074
3075
3076
3077
3078
3079
3080
3081
3082
3083
3084
3085
3086
3087
3088
3089
3090
3091
3092
3093
3094
3095
3096
3097
3098
3099
3100
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
  #ifdef CONFIG_PM_RUNTIME
  /**
   * 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)
  {
  	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;
  
  	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)
  {
  	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)
  {
  	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)
  {
  	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   Aaron Lu   blkpm: avoid slee...
3111
  		pm_request_autosuspend(q->dev);
6c9546675   Lin Ming   block: add runtim...
3112
3113
3114
3115
3116
3117
3118
  	} else {
  		q->rpm_status = RPM_SUSPENDED;
  	}
  	spin_unlock_irq(q->queue_lock);
  }
  EXPORT_SYMBOL(blk_post_runtime_resume);
  #endif
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3119
3120
  int __init blk_dev_init(void)
  {
9eb55b030   Nikanth Karthikesan   block: catch tryi...
3121
3122
  	BUILD_BUG_ON(__REQ_NR_BITS > 8 *
  			sizeof(((struct request *)0)->cmd_flags));
89b90be2d   Tejun Heo   block: make kbloc...
3123
3124
  	/* used for unplugging and affects IO latency/throughput - HIGHPRI */
  	kblockd_workqueue = alloc_workqueue("kblockd",
695588f94   Viresh Kumar   block: queue work...
3125
3126
  					    WQ_MEM_RECLAIM | WQ_HIGHPRI |
  					    WQ_POWER_EFFICIENT, 0);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3127
3128
3129
3130
3131
  	if (!kblockd_workqueue)
  		panic("Failed to create kblockd
  ");
  
  	request_cachep = kmem_cache_create("blkdev_requests",
20c2df83d   Paul Mundt   mm: Remove slab d...
3132
  			sizeof(struct request), 0, SLAB_PANIC, NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3133

8324aa91d   Jens Axboe   block: split tag ...
3134
  	blk_requestq_cachep = kmem_cache_create("blkdev_queue",
165125e1e   Jens Axboe   [BLOCK] Get rid o...
3135
  			sizeof(struct request_queue), 0, SLAB_PANIC, NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3136

d38ecf935   Jens Axboe   io context sharin...
3137
  	return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3138
  }