Blame view

block/blk-core.c 93 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>
eea8f41cc   Tejun Heo   blkcg: move block...
35
  #include <linux/blk-cgroup.h>
55782138e   Li Zefan   tracing/events: c...
36
37
38
  
  #define CREATE_TRACE_POINTS
  #include <trace/events/block.h>
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
39

8324aa91d   Jens Axboe   block: split tag ...
40
  #include "blk.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);
3291fa57c   Keith Busch   NVMe: Add tracepo...
46
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_split);
cbae8d45d   NeilBrown   block: export blo...
47
  EXPORT_TRACEPOINT_SYMBOL_GPL(block_unplug);
0bfc24559   Ingo Molnar   blktrace: port to...
48

a73f730d0   Tejun Heo   block, cfq: move ...
49
  DEFINE_IDA(blk_queue_ida);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
50
51
52
  /*
   * For the allocated request tables
   */
d674d4145   Wei Tang   block: do not ini...
53
  struct kmem_cache *request_cachep;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
54
55
56
57
  
  /*
   * For queue allocation
   */
6728cb0e6   Jens Axboe   block: make core ...
58
  struct kmem_cache *blk_requestq_cachep;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
59
60
  
  /*
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
61
62
   * Controlling structure to kblockd
   */
ff856bad6   Jens Axboe   [BLOCK] ll_rw_blk...
63
  static struct workqueue_struct *kblockd_workqueue;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
64

d40f75a06   Tejun Heo   writeback, blkcg:...
65
66
  static void blk_clear_congested(struct request_list *rl, int sync)
  {
d40f75a06   Tejun Heo   writeback, blkcg:...
67
68
69
  #ifdef CONFIG_CGROUP_WRITEBACK
  	clear_wb_congested(rl->blkg->wb_congested, sync);
  #else
482cf79cd   Tejun Heo   writeback, blkcg:...
70
71
72
73
74
75
  	/*
  	 * If !CGROUP_WRITEBACK, all blkg's map to bdi->wb and we shouldn't
  	 * flip its congestion state for events on other blkcgs.
  	 */
  	if (rl == &rl->q->root_rl)
  		clear_wb_congested(rl->q->backing_dev_info.wb.congested, sync);
d40f75a06   Tejun Heo   writeback, blkcg:...
76
77
78
79
80
  #endif
  }
  
  static void blk_set_congested(struct request_list *rl, int sync)
  {
d40f75a06   Tejun Heo   writeback, blkcg:...
81
82
83
  #ifdef CONFIG_CGROUP_WRITEBACK
  	set_wb_congested(rl->blkg->wb_congested, sync);
  #else
482cf79cd   Tejun Heo   writeback, blkcg:...
84
85
86
  	/* see blk_clear_congested() */
  	if (rl == &rl->q->root_rl)
  		set_wb_congested(rl->q->backing_dev_info.wb.congested, sync);
d40f75a06   Tejun Heo   writeback, blkcg:...
87
88
  #endif
  }
8324aa91d   Jens Axboe   block: split tag ...
89
  void blk_queue_congestion_threshold(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
90
91
92
93
94
95
96
97
98
99
100
101
102
  {
  	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
103
104
105
106
107
  /**
   * 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
ff9ea3238   Tejun Heo   block, bdi: an ac...
108
109
   * backing_dev_info.  This function can only be called if @bdev is opened
   * and the return value is never NULL.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
110
111
112
   */
  struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev)
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
113
  	struct request_queue *q = bdev_get_queue(bdev);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
114

ff9ea3238   Tejun Heo   block, bdi: an ac...
115
  	return &q->backing_dev_info;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
116
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
117
  EXPORT_SYMBOL(blk_get_backing_dev_info);
2a4aa30c5   FUJITA Tomonori   block: rename and...
118
  void blk_rq_init(struct request_queue *q, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
119
  {
1afb20f30   FUJITA Tomonori   block: make rq_in...
120
  	memset(rq, 0, sizeof(*rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
121
  	INIT_LIST_HEAD(&rq->queuelist);
242f9dcb8   Jens Axboe   block: unify requ...
122
  	INIT_LIST_HEAD(&rq->timeout_list);
c7c22e4d5   Jens Axboe   block: add suppor...
123
  	rq->cpu = -1;
63a713867   Jens Axboe   block: fixup rq_i...
124
  	rq->q = q;
a2dec7b36   Tejun Heo   block: hide reque...
125
  	rq->__sector = (sector_t) -1;
2e662b65f   Jens Axboe   [PATCH] elevator:...
126
127
  	INIT_HLIST_NODE(&rq->hash);
  	RB_CLEAR_NODE(&rq->rb_node);
d7e3c3249   FUJITA Tomonori   block: add large ...
128
  	rq->cmd = rq->__cmd;
e2494e1b4   Li Zefan   blktrace: fix pdu...
129
  	rq->cmd_len = BLK_MAX_CDB;
63a713867   Jens Axboe   block: fixup rq_i...
130
  	rq->tag = -1;
b243ddcbe   Tejun Heo   block: move rq->s...
131
  	rq->start_time = jiffies;
9195291e5   Divyesh Shah   blkio: Increment ...
132
  	set_start_time_ns(rq);
09e099d4b   Jerome Marchand   block: fix accoun...
133
  	rq->part = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
134
  }
2a4aa30c5   FUJITA Tomonori   block: rename and...
135
  EXPORT_SYMBOL(blk_rq_init);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
136

5bb23a688   NeilBrown   Don't decrement b...
137
138
  static void req_bio_endio(struct request *rq, struct bio *bio,
  			  unsigned int nbytes, int error)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
139
  {
78d8e58a0   Mike Snitzer   Revert "block, dm...
140
  	if (error)
4246a0b63   Christoph Hellwig   block: add a bi_e...
141
  		bio->bi_error = error;
797e7dbbe   Tejun Heo   [BLOCK] reimpleme...
142

143a87f4c   Tejun Heo   block: improve fl...
143
  	if (unlikely(rq->cmd_flags & REQ_QUIET))
b7c44ed9d   Jens Axboe   block: manipulate...
144
  		bio_set_flag(bio, BIO_QUIET);
08bafc034   Keith Mannthey   block: Supress Bu...
145

f79ea4161   Kent Overstreet   block: Refactor b...
146
  	bio_advance(bio, nbytes);
7ba1ba12e   Martin K. Petersen   block: Block laye...
147

143a87f4c   Tejun Heo   block: improve fl...
148
  	/* don't actually finish bio if it's part of flush sequence */
78d8e58a0   Mike Snitzer   Revert "block, dm...
149
  	if (bio->bi_iter.bi_size == 0 && !(rq->cmd_flags & REQ_FLUSH_SEQ))
4246a0b63   Christoph Hellwig   block: add a bi_e...
150
  		bio_endio(bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
151
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
152

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
153
154
155
  void blk_dump_rq_flags(struct request *rq, char *msg)
  {
  	int bit;
5953316db   Jens Axboe   block: make rq->c...
156
157
  	printk(KERN_INFO "%s: dev %s: type=%x, flags=%llx
  ", msg,
4aff5e233   Jens Axboe   [PATCH] Split str...
158
  		rq->rq_disk ? rq->rq_disk->disk_name : "?", rq->cmd_type,
5953316db   Jens Axboe   block: make rq->c...
159
  		(unsigned long long) rq->cmd_flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
160

83096ebf1   Tejun Heo   block: convert to...
161
162
163
164
  	printk(KERN_INFO "  sector %llu, nr/cnr %u/%u
  ",
  	       (unsigned long long)blk_rq_pos(rq),
  	       blk_rq_sectors(rq), blk_rq_cur_sectors(rq));
b4f42e283   Jens Axboe   block: remove str...
165
166
167
  	printk(KERN_INFO "  bio %p, biotail %p, len %u
  ",
  	       rq->bio, rq->biotail, blk_rq_bytes(rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
168

33659ebba   Christoph Hellwig   block: remove wra...
169
  	if (rq->cmd_type == REQ_TYPE_BLOCK_PC) {
6728cb0e6   Jens Axboe   block: make core ...
170
  		printk(KERN_INFO "  cdb: ");
d34c87e4b   FUJITA Tomonori   block: replace si...
171
  		for (bit = 0; bit < BLK_MAX_CDB; bit++)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
172
173
174
175
176
  			printk("%02x ", rq->cmd[bit]);
  		printk("
  ");
  	}
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
177
  EXPORT_SYMBOL(blk_dump_rq_flags);
3cca6dc1c   Jens Axboe   block: add API fo...
178
  static void blk_delay_work(struct work_struct *work)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
179
  {
3cca6dc1c   Jens Axboe   block: add API fo...
180
  	struct request_queue *q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
181

3cca6dc1c   Jens Axboe   block: add API fo...
182
183
  	q = container_of(work, struct request_queue, delay_work.work);
  	spin_lock_irq(q->queue_lock);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
184
  	__blk_run_queue(q);
3cca6dc1c   Jens Axboe   block: add API fo...
185
  	spin_unlock_irq(q->queue_lock);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
186
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
187
188
  
  /**
3cca6dc1c   Jens Axboe   block: add API fo...
189
190
191
   * 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
192
193
   *
   * Description:
3cca6dc1c   Jens Axboe   block: add API fo...
194
195
   *   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...
196
   *   restarted around the specified time. Queue lock must be held.
3cca6dc1c   Jens Axboe   block: add API fo...
197
198
   */
  void blk_delay_queue(struct request_queue *q, unsigned long msecs)
2ad8b1ef1   Alan D. Brunelle   Add UNPLUG traces...
199
  {
704605711   Bart Van Assche   block: Avoid sche...
200
201
202
  	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...
203
  }
3cca6dc1c   Jens Axboe   block: add API fo...
204
  EXPORT_SYMBOL(blk_delay_queue);
2ad8b1ef1   Alan D. Brunelle   Add UNPLUG traces...
205

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
206
  /**
21491412f   Jens Axboe   block: add blk_st...
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
   * blk_start_queue_async - asynchronously restart a previously stopped queue
   * @q:    The &struct request_queue in question
   *
   * Description:
   *   blk_start_queue_async() will clear the stop flag on the queue, and
   *   ensure that the request_fn for the queue is run from an async
   *   context.
   **/
  void blk_start_queue_async(struct request_queue *q)
  {
  	queue_flag_clear(QUEUE_FLAG_STOPPED, q);
  	blk_run_queue_async(q);
  }
  EXPORT_SYMBOL(blk_start_queue_async);
  
  /**
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
223
   * blk_start_queue - restart a previously stopped queue
165125e1e   Jens Axboe   [BLOCK] Get rid o...
224
   * @q:    The &struct request_queue in question
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
225
226
227
228
229
230
   *
   * 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...
231
  void blk_start_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
232
  {
a038e2536   Paolo 'Blaisorblade' Giarrusso   [PATCH] blk_start...
233
  	WARN_ON(!irqs_disabled());
75ad23bc0   Nick Piggin   block: make queue...
234
  	queue_flag_clear(QUEUE_FLAG_STOPPED, q);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
235
  	__blk_run_queue(q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
236
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
237
238
239
240
  EXPORT_SYMBOL(blk_start_queue);
  
  /**
   * blk_stop_queue - stop a queue
165125e1e   Jens Axboe   [BLOCK] Get rid o...
241
   * @q:    The &struct request_queue in question
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
242
243
244
245
246
247
248
249
250
251
252
   *
   * 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...
253
  void blk_stop_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
254
  {
136b5721d   Tejun Heo   workqueue: deprec...
255
  	cancel_delayed_work(&q->delay_work);
75ad23bc0   Nick Piggin   block: make queue...
256
  	queue_flag_set(QUEUE_FLAG_STOPPED, q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
257
258
259
260
261
262
263
264
265
266
267
268
  }
  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...
269
   *     that the callbacks might use. The caller must already have made sure
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
270
271
272
   *     that its ->make_request_fn will not re-add plugging prior to calling
   *     this function.
   *
da5277700   Vivek Goyal   block: Move blk_t...
273
   *     This function does not cancel any asynchronous activity arising
da3dae54e   Masanari Iida   Documentation: Do...
274
   *     out of elevator or throttling code. That would require elevator_exit()
5efd61135   Tejun Heo   blkcg: add blkcg_...
275
   *     and blkcg_exit_queue() to be called with queue lock initialized.
da5277700   Vivek Goyal   block: Move blk_t...
276
   *
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
277
278
279
   */
  void blk_sync_queue(struct request_queue *q)
  {
70ed28b92   Jens Axboe   block: leave the ...
280
  	del_timer_sync(&q->timeout);
f04c1fe76   Ming Lei   block: blk-mq: ma...
281
282
283
284
  
  	if (q->mq_ops) {
  		struct blk_mq_hw_ctx *hctx;
  		int i;
70f4db639   Christoph Hellwig   blk-mq: add blk_m...
285
286
287
288
  		queue_for_each_hw_ctx(q, hctx, i) {
  			cancel_delayed_work_sync(&hctx->run_work);
  			cancel_delayed_work_sync(&hctx->delay_work);
  		}
f04c1fe76   Ming Lei   block: blk-mq: ma...
289
290
291
  	} else {
  		cancel_delayed_work_sync(&q->delay_work);
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
292
293
294
295
  }
  EXPORT_SYMBOL(blk_sync_queue);
  
  /**
c246e80d8   Bart Van Assche   block: Avoid that...
296
297
298
299
300
301
302
303
304
305
306
307
308
309
   * __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...
310
311
312
313
314
315
316
317
  	/*
  	 * 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...
318
  	q->request_fn(q);
24faf6f60   Bart Van Assche   block: Make blk_c...
319
  	q->request_fn_active--;
c246e80d8   Bart Van Assche   block: Avoid that...
320
  }
a7928c157   Christoph Hellwig   block: move PM re...
321
  EXPORT_SYMBOL_GPL(__blk_run_queue_uncond);
c246e80d8   Bart Van Assche   block: Avoid that...
322
323
  
  /**
80a4b58e3   Jens Axboe   block: only call ...
324
   * __blk_run_queue - run a single device queue
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
325
   * @q:	The queue to run
80a4b58e3   Jens Axboe   block: only call ...
326
327
328
   *
   * Description:
   *    See @blk_run_queue. This variant must be called with the queue lock
24ecfbe27   Christoph Hellwig   block: add blk_ru...
329
   *    held and interrupts disabled.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
330
   */
24ecfbe27   Christoph Hellwig   block: add blk_ru...
331
  void __blk_run_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
332
  {
a538cd03b   Tejun Heo   block: merge blk_...
333
334
  	if (unlikely(blk_queue_stopped(q)))
  		return;
c246e80d8   Bart Van Assche   block: Avoid that...
335
  	__blk_run_queue_uncond(q);
75ad23bc0   Nick Piggin   block: make queue...
336
337
  }
  EXPORT_SYMBOL(__blk_run_queue);
dac07ec12   Jens Axboe   [BLOCK] limit req...
338

75ad23bc0   Nick Piggin   block: make queue...
339
  /**
24ecfbe27   Christoph Hellwig   block: add blk_ru...
340
341
342
343
344
   * 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...
345
   *    of us. The caller must hold the queue lock.
24ecfbe27   Christoph Hellwig   block: add blk_ru...
346
347
348
   */
  void blk_run_queue_async(struct request_queue *q)
  {
704605711   Bart Van Assche   block: Avoid sche...
349
  	if (likely(!blk_queue_stopped(q) && !blk_queue_dead(q)))
e7c2f9674   Tejun Heo   workqueue: use mo...
350
  		mod_delayed_work(kblockd_workqueue, &q->delay_work, 0);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
351
  }
c21e6beba   Jens Axboe   block: get rid of...
352
  EXPORT_SYMBOL(blk_run_queue_async);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
353
354
  
  /**
75ad23bc0   Nick Piggin   block: make queue...
355
356
   * blk_run_queue - run a single device queue
   * @q: The queue to run
80a4b58e3   Jens Axboe   block: only call ...
357
358
359
   *
   * Description:
   *    Invoke request handling on this queue, if it has pending work to do.
a7f557923   Tejun Heo   block: kill blk_s...
360
   *    May be used to restart queueing when a request has completed.
75ad23bc0   Nick Piggin   block: make queue...
361
362
363
364
365
366
   */
  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...
367
  	__blk_run_queue(q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
368
369
370
  	spin_unlock_irqrestore(q->queue_lock, flags);
  }
  EXPORT_SYMBOL(blk_run_queue);
165125e1e   Jens Axboe   [BLOCK] Get rid o...
371
  void blk_put_queue(struct request_queue *q)
483f4afc4   Al Viro   [PATCH] fix sysfs...
372
373
374
  {
  	kobject_put(&q->kobj);
  }
d86e0e83b   Jens Axboe   block: export blk...
375
  EXPORT_SYMBOL(blk_put_queue);
483f4afc4   Al Viro   [PATCH] fix sysfs...
376

e3c78ca52   Tejun Heo   block: reorganize...
377
  /**
807592a4f   Bart Van Assche   block: Let blk_dr...
378
   * __blk_drain_queue - drain requests from request_queue
e3c78ca52   Tejun Heo   block: reorganize...
379
   * @q: queue to drain
c9a929dde   Tejun Heo   block: fix reques...
380
   * @drain_all: whether to drain all requests or only the ones w/ ELVPRIV
e3c78ca52   Tejun Heo   block: reorganize...
381
   *
c9a929dde   Tejun Heo   block: fix reques...
382
383
384
   * 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...
385
   */
807592a4f   Bart Van Assche   block: Let blk_dr...
386
387
388
  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...
389
  {
458f27a98   Asias He   block: Avoid miss...
390
  	int i;
807592a4f   Bart Van Assche   block: Let blk_dr...
391
  	lockdep_assert_held(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
392
  	while (true) {
481a7d647   Tejun Heo   block: fix drain_...
393
  		bool drain = false;
e3c78ca52   Tejun Heo   block: reorganize...
394

b855b04a0   Tejun Heo   block: blk-thrott...
395
396
397
398
399
400
  		/*
  		 * 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_...
401
  		blkcg_drain_queue(q);
e3c78ca52   Tejun Heo   block: reorganize...
402

4eabc9412   Tejun Heo   block: don't kick...
403
404
  		/*
  		 * This function might be called on a queue which failed
b855b04a0   Tejun Heo   block: blk-thrott...
405
406
407
408
  		 * 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...
409
  		 */
b855b04a0   Tejun Heo   block: blk-thrott...
410
  		if (!list_empty(&q->queue_head) && q->request_fn)
4eabc9412   Tejun Heo   block: don't kick...
411
  			__blk_run_queue(q);
c9a929dde   Tejun Heo   block: fix reques...
412

8a5ecdd42   Tejun Heo   block: add q->nr_...
413
  		drain |= q->nr_rqs_elvpriv;
24faf6f60   Bart Van Assche   block: Make blk_c...
414
  		drain |= q->request_fn_active;
481a7d647   Tejun Heo   block: fix drain_...
415
416
417
418
419
420
421
  
  		/*
  		 * Unfortunately, requests are queued at and tracked from
  		 * multiple places and there's no single counter which can
  		 * be drained.  Check all the queues and counters.
  		 */
  		if (drain_all) {
e97c293cd   Ming Lei   block: introduce ...
422
  			struct blk_flush_queue *fq = blk_get_flush_queue(q, NULL);
481a7d647   Tejun Heo   block: fix drain_...
423
424
  			drain |= !list_empty(&q->queue_head);
  			for (i = 0; i < 2; i++) {
8a5ecdd42   Tejun Heo   block: add q->nr_...
425
  				drain |= q->nr_rqs[i];
481a7d647   Tejun Heo   block: fix drain_...
426
  				drain |= q->in_flight[i];
7c94e1c15   Ming Lei   block: introduce ...
427
428
  				if (fq)
  				    drain |= !list_empty(&fq->flush_queue[i]);
481a7d647   Tejun Heo   block: fix drain_...
429
430
  			}
  		}
e3c78ca52   Tejun Heo   block: reorganize...
431

481a7d647   Tejun Heo   block: fix drain_...
432
  		if (!drain)
e3c78ca52   Tejun Heo   block: reorganize...
433
  			break;
807592a4f   Bart Van Assche   block: Let blk_dr...
434
435
  
  		spin_unlock_irq(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
436
  		msleep(10);
807592a4f   Bart Van Assche   block: Let blk_dr...
437
438
  
  		spin_lock_irq(q->queue_lock);
e3c78ca52   Tejun Heo   block: reorganize...
439
  	}
458f27a98   Asias He   block: Avoid miss...
440
441
442
443
444
445
446
  
  	/*
  	 * 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 ...
447
  		struct request_list *rl;
a051661ca   Tejun Heo   blkcg: implement ...
448
449
450
  		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...
451
  	}
e3c78ca52   Tejun Heo   block: reorganize...
452
  }
c9a929dde   Tejun Heo   block: fix reques...
453
  /**
d732580b4   Tejun Heo   block: implement ...
454
455
456
457
458
   * 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...
459
   * throttled or issued before.  On return, it's guaranteed that no request
80fd99792   Tejun Heo   blkcg: make sure ...
460
461
   * is being throttled or has ELVPRIV set and blk_queue_bypass() %true
   * inside queue or RCU read lock.
d732580b4   Tejun Heo   block: implement ...
462
463
464
465
   */
  void blk_queue_bypass_start(struct request_queue *q)
  {
  	spin_lock_irq(q->queue_lock);
776687bce   Tejun Heo   block, blk-mq: dr...
466
  	q->bypass_depth++;
d732580b4   Tejun Heo   block: implement ...
467
468
  	queue_flag_set(QUEUE_FLAG_BYPASS, q);
  	spin_unlock_irq(q->queue_lock);
776687bce   Tejun Heo   block, blk-mq: dr...
469
470
471
472
473
474
  	/*
  	 * Queues start drained.  Skip actual draining till init is
  	 * complete.  This avoids lenghty delays during queue init which
  	 * can happen many times during boot.
  	 */
  	if (blk_queue_init_done(q)) {
807592a4f   Bart Van Assche   block: Let blk_dr...
475
476
477
  		spin_lock_irq(q->queue_lock);
  		__blk_drain_queue(q, false);
  		spin_unlock_irq(q->queue_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
478
479
480
  		/* ensure blk_queue_bypass() is %true inside RCU read lock */
  		synchronize_rcu();
  	}
d732580b4   Tejun Heo   block: implement ...
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
  }
  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);
aed3ea94b   Jens Axboe   block: wake up wa...
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
  void blk_set_queue_dying(struct request_queue *q)
  {
  	queue_flag_set_unlocked(QUEUE_FLAG_DYING, q);
  
  	if (q->mq_ops)
  		blk_mq_wake_waiters(q);
  	else {
  		struct request_list *rl;
  
  		blk_queue_for_each_rl(rl, q) {
  			if (rl->rq_pool) {
  				wake_up(&rl->wait[BLK_RW_SYNC]);
  				wake_up(&rl->wait[BLK_RW_ASYNC]);
  			}
  		}
  	}
  }
  EXPORT_SYMBOL_GPL(blk_set_queue_dying);
d732580b4   Tejun Heo   block: implement ...
517
  /**
c9a929dde   Tejun Heo   block: fix reques...
518
519
520
   * blk_cleanup_queue - shutdown a request queue
   * @q: request queue to shutdown
   *
c246e80d8   Bart Van Assche   block: Avoid that...
521
522
   * 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...
523
   */
6728cb0e6   Jens Axboe   block: make core ...
524
  void blk_cleanup_queue(struct request_queue *q)
483f4afc4   Al Viro   [PATCH] fix sysfs...
525
  {
c9a929dde   Tejun Heo   block: fix reques...
526
  	spinlock_t *lock = q->queue_lock;
e3335de94   Jens Axboe   block: blk_cleanu...
527

3f3299d5c   Bart Van Assche   block: Rename que...
528
  	/* mark @q DYING, no new request or merges will be allowed afterwards */
483f4afc4   Al Viro   [PATCH] fix sysfs...
529
  	mutex_lock(&q->sysfs_lock);
aed3ea94b   Jens Axboe   block: wake up wa...
530
  	blk_set_queue_dying(q);
c9a929dde   Tejun Heo   block: fix reques...
531
  	spin_lock_irq(lock);
6ecf23afa   Tejun Heo   block: extend que...
532

80fd99792   Tejun Heo   blkcg: make sure ...
533
  	/*
3f3299d5c   Bart Van Assche   block: Rename que...
534
  	 * A dying queue is permanently in bypass mode till released.  Note
80fd99792   Tejun Heo   blkcg: make sure ...
535
536
537
538
539
540
541
  	 * 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...
542
543
  	q->bypass_depth++;
  	queue_flag_set(QUEUE_FLAG_BYPASS, q);
c9a929dde   Tejun Heo   block: fix reques...
544
545
  	queue_flag_set(QUEUE_FLAG_NOMERGES, q);
  	queue_flag_set(QUEUE_FLAG_NOXMERGES, q);
3f3299d5c   Bart Van Assche   block: Rename que...
546
  	queue_flag_set(QUEUE_FLAG_DYING, q);
c9a929dde   Tejun Heo   block: fix reques...
547
548
  	spin_unlock_irq(lock);
  	mutex_unlock(&q->sysfs_lock);
c246e80d8   Bart Van Assche   block: Avoid that...
549
550
551
552
  	/*
  	 * Drain all requests queued before DYING marking. Set DEAD flag to
  	 * prevent that q->request_fn() gets invoked after draining finished.
  	 */
3ef28e83a   Dan Williams   block: generic re...
553
554
555
  	blk_freeze_queue(q);
  	spin_lock_irq(lock);
  	if (!q->mq_ops)
43a5e4e21   Ming Lei   block: blk-mq: su...
556
  		__blk_drain_queue(q, true);
c246e80d8   Bart Van Assche   block: Avoid that...
557
  	queue_flag_set(QUEUE_FLAG_DEAD, q);
807592a4f   Bart Van Assche   block: Let blk_dr...
558
  	spin_unlock_irq(lock);
c9a929dde   Tejun Heo   block: fix reques...
559

5a48fc147   Dan Williams   block: blk_flush_...
560
561
  	/* for synchronous bio-based driver finish in-flight integrity i/o */
  	blk_flush_integrity();
c9a929dde   Tejun Heo   block: fix reques...
562
563
564
  	/* @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);
45a9c9d90   Bart Van Assche   blk-mq: Fix a use...
565
566
  	if (q->mq_ops)
  		blk_mq_free_queue(q);
3ef28e83a   Dan Williams   block: generic re...
567
  	percpu_ref_exit(&q->q_usage_counter);
45a9c9d90   Bart Van Assche   blk-mq: Fix a use...
568

5e5cfac0c   Asias He   block: Mitigate l...
569
570
571
572
  	spin_lock_irq(lock);
  	if (q->queue_lock != &q->__queue_lock)
  		q->queue_lock = &q->__queue_lock;
  	spin_unlock_irq(lock);
b02176f30   Tejun Heo   block: don't rele...
573
  	bdi_unregister(&q->backing_dev_info);
6cd18e711   NeilBrown   block: destroy bd...
574

c9a929dde   Tejun Heo   block: fix reques...
575
  	/* @q is and will stay empty, shutdown and put */
483f4afc4   Al Viro   [PATCH] fix sysfs...
576
577
  	blk_put_queue(q);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
578
  EXPORT_SYMBOL(blk_cleanup_queue);
271508dba   David Rientjes   block: allocate r...
579
580
581
582
583
584
585
586
587
588
589
  /* Allocate memory local to the request queue */
  static void *alloc_request_struct(gfp_t gfp_mask, void *data)
  {
  	int nid = (int)(long)data;
  	return kmem_cache_alloc_node(request_cachep, gfp_mask, nid);
  }
  
  static void free_request_struct(void *element, void *unused)
  {
  	kmem_cache_free(request_cachep, element);
  }
5b788ce3e   Tejun Heo   block: prepare fo...
590
591
  int blk_init_rl(struct request_list *rl, struct request_queue *q,
  		gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
592
  {
1abec4fdb   Mike Snitzer   block: make blk_i...
593
594
  	if (unlikely(rl->rq_pool))
  		return 0;
5b788ce3e   Tejun Heo   block: prepare fo...
595
  	rl->q = q;
1faa16d22   Jens Axboe   block: change the...
596
597
  	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...
598
599
  	init_waitqueue_head(&rl->wait[BLK_RW_SYNC]);
  	init_waitqueue_head(&rl->wait[BLK_RW_ASYNC]);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
600

271508dba   David Rientjes   block: allocate r...
601
602
603
604
  	rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, alloc_request_struct,
  					  free_request_struct,
  					  (void *)(long)q->node, gfp_mask,
  					  q->node);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
605
606
607
608
609
  	if (!rl->rq_pool)
  		return -ENOMEM;
  
  	return 0;
  }
5b788ce3e   Tejun Heo   block: prepare fo...
610
611
612
613
614
  void blk_exit_rl(struct request_list *rl)
  {
  	if (rl->rq_pool)
  		mempool_destroy(rl->rq_pool);
  }
165125e1e   Jens Axboe   [BLOCK] Get rid o...
615
  struct request_queue *blk_alloc_queue(gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
616
  {
c304a51bf   Ezequiel Garcia   block: use NUMA_N...
617
  	return blk_alloc_queue_node(gfp_mask, NUMA_NO_NODE);
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
618
619
  }
  EXPORT_SYMBOL(blk_alloc_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
620

6f3b0e8bc   Christoph Hellwig   blk-mq: add a fla...
621
  int blk_queue_enter(struct request_queue *q, bool nowait)
3ef28e83a   Dan Williams   block: generic re...
622
623
624
625
626
627
  {
  	while (true) {
  		int ret;
  
  		if (percpu_ref_tryget_live(&q->q_usage_counter))
  			return 0;
6f3b0e8bc   Christoph Hellwig   blk-mq: add a fla...
628
  		if (nowait)
3ef28e83a   Dan Williams   block: generic re...
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
  			return -EBUSY;
  
  		ret = wait_event_interruptible(q->mq_freeze_wq,
  				!atomic_read(&q->mq_freeze_depth) ||
  				blk_queue_dying(q));
  		if (blk_queue_dying(q))
  			return -ENODEV;
  		if (ret)
  			return ret;
  	}
  }
  
  void blk_queue_exit(struct request_queue *q)
  {
  	percpu_ref_put(&q->q_usage_counter);
  }
  
  static void blk_queue_usage_counter_release(struct percpu_ref *ref)
  {
  	struct request_queue *q =
  		container_of(ref, struct request_queue, q_usage_counter);
  
  	wake_up_all(&q->mq_freeze_wq);
  }
287922eb0   Christoph Hellwig   block: defer time...
653
654
655
656
657
658
  static void blk_rq_timed_out_timer(unsigned long data)
  {
  	struct request_queue *q = (struct request_queue *)data;
  
  	kblockd_schedule_work(&q->timeout_work);
  }
165125e1e   Jens Axboe   [BLOCK] Get rid o...
659
  struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id)
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
660
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
661
  	struct request_queue *q;
e0bf68dde   Peter Zijlstra   mm: bdi init hooks
662
  	int err;
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
663

8324aa91d   Jens Axboe   block: split tag ...
664
  	q = kmem_cache_alloc_node(blk_requestq_cachep,
94f6030ca   Christoph Lameter   Slab allocators: ...
665
  				gfp_mask | __GFP_ZERO, node_id);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
666
667
  	if (!q)
  		return NULL;
00380a404   Dan Carpenter   block: blk_alloc_...
668
  	q->id = ida_simple_get(&blk_queue_ida, 0, 0, gfp_mask);
a73f730d0   Tejun Heo   block, cfq: move ...
669
  	if (q->id < 0)
3d2936f45   Ming Lei   block: only alloc...
670
  		goto fail_q;
a73f730d0   Tejun Heo   block, cfq: move ...
671

54efd50bf   Kent Overstreet   block: make gener...
672
673
674
  	q->bio_split = bioset_create(BIO_POOL_SIZE, 0);
  	if (!q->bio_split)
  		goto fail_id;
0989a025d   Jens Axboe   block: don't over...
675
  	q->backing_dev_info.ra_pages =
09cbfeaf1   Kirill A. Shutemov   mm, fs: get rid o...
676
  			(VM_MAX_READAHEAD * 1024) / PAGE_SIZE;
89e9b9e07   Tejun Heo   writeback: add {C...
677
  	q->backing_dev_info.capabilities = BDI_CAP_CGROUP_WRITEBACK;
d993831fa   Jens Axboe   writeback: add na...
678
  	q->backing_dev_info.name = "block";
5151412dd   Mike Snitzer   block: initialize...
679
  	q->node = node_id;
0989a025d   Jens Axboe   block: don't over...
680

e0bf68dde   Peter Zijlstra   mm: bdi init hooks
681
  	err = bdi_init(&q->backing_dev_info);
a73f730d0   Tejun Heo   block, cfq: move ...
682
  	if (err)
54efd50bf   Kent Overstreet   block: make gener...
683
  		goto fail_split;
e0bf68dde   Peter Zijlstra   mm: bdi init hooks
684

31373d09d   Matthew Garrett   laptop-mode: Make...
685
686
  	setup_timer(&q->backing_dev_info.laptop_mode_wb_timer,
  		    laptop_mode_timer_fn, (unsigned long) q);
242f9dcb8   Jens Axboe   block: unify requ...
687
  	setup_timer(&q->timeout, blk_rq_timed_out_timer, (unsigned long) q);
b855b04a0   Tejun Heo   block: blk-thrott...
688
  	INIT_LIST_HEAD(&q->queue_head);
242f9dcb8   Jens Axboe   block: unify requ...
689
  	INIT_LIST_HEAD(&q->timeout_list);
a612fddf0   Tejun Heo   block, cfq: move ...
690
  	INIT_LIST_HEAD(&q->icq_list);
4eef30499   Tejun Heo   blkcg: move per-q...
691
  #ifdef CONFIG_BLK_CGROUP
e8989fae3   Tejun Heo   blkcg: unify blkg...
692
  	INIT_LIST_HEAD(&q->blkg_list);
4eef30499   Tejun Heo   blkcg: move per-q...
693
  #endif
3cca6dc1c   Jens Axboe   block: add API fo...
694
  	INIT_DELAYED_WORK(&q->delay_work, blk_delay_work);
483f4afc4   Al Viro   [PATCH] fix sysfs...
695

8324aa91d   Jens Axboe   block: split tag ...
696
  	kobject_init(&q->kobj, &blk_queue_ktype);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
697

483f4afc4   Al Viro   [PATCH] fix sysfs...
698
  	mutex_init(&q->sysfs_lock);
e7e72bf64   Neil Brown   Remove blkdev war...
699
  	spin_lock_init(&q->__queue_lock);
483f4afc4   Al Viro   [PATCH] fix sysfs...
700

c94a96ac9   Vivek Goyal   block: Initialize...
701
702
703
704
705
  	/*
  	 * 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...
706
707
708
  	/*
  	 * 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...
709
710
  	 * init.  The initial bypass will be finished when the queue is
  	 * registered by blk_register_queue().
b82d4b197   Tejun Heo   blkcg: make reque...
711
712
713
  	 */
  	q->bypass_depth = 1;
  	__set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags);
320ae51fe   Jens Axboe   blk-mq: new multi...
714
  	init_waitqueue_head(&q->mq_freeze_wq);
3ef28e83a   Dan Williams   block: generic re...
715
716
717
718
719
720
721
  	/*
  	 * Init percpu_ref in atomic mode so that it's faster to shutdown.
  	 * See blk_register_queue() for details.
  	 */
  	if (percpu_ref_init(&q->q_usage_counter,
  				blk_queue_usage_counter_release,
  				PERCPU_REF_INIT_ATOMIC, GFP_KERNEL))
fff4996b7   Mikulas Patocka   blk-core: Fix mem...
722
  		goto fail_bdi;
f51b802c1   Tejun Heo   blkcg: use the us...
723

3ef28e83a   Dan Williams   block: generic re...
724
725
  	if (blkcg_init_queue(q))
  		goto fail_ref;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
726
  	return q;
a73f730d0   Tejun Heo   block, cfq: move ...
727

3ef28e83a   Dan Williams   block: generic re...
728
729
  fail_ref:
  	percpu_ref_exit(&q->q_usage_counter);
fff4996b7   Mikulas Patocka   blk-core: Fix mem...
730
731
  fail_bdi:
  	bdi_destroy(&q->backing_dev_info);
54efd50bf   Kent Overstreet   block: make gener...
732
733
  fail_split:
  	bioset_free(q->bio_split);
a73f730d0   Tejun Heo   block, cfq: move ...
734
735
736
737
738
  fail_id:
  	ida_simple_remove(&blk_queue_ida, q->id);
  fail_q:
  	kmem_cache_free(blk_requestq_cachep, q);
  	return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
739
  }
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
740
  EXPORT_SYMBOL(blk_alloc_queue_node);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
  
  /**
   * 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...
764
765
   *    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
766
   *
710027a48   Randy Dunlap   Add some block/ s...
767
   *    Function returns a pointer to the initialized request queue, or %NULL if
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
768
769
770
771
772
773
   *    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...
774

165125e1e   Jens Axboe   [BLOCK] Get rid o...
775
  struct request_queue *blk_init_queue(request_fn_proc *rfn, spinlock_t *lock)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
776
  {
c304a51bf   Ezequiel Garcia   block: use NUMA_N...
777
  	return blk_init_queue_node(rfn, lock, NUMA_NO_NODE);
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
778
779
  }
  EXPORT_SYMBOL(blk_init_queue);
165125e1e   Jens Axboe   [BLOCK] Get rid o...
780
  struct request_queue *
1946089a1   Christoph Lameter   [PATCH] NUMA awar...
781
782
  blk_init_queue_node(request_fn_proc *rfn, spinlock_t *lock, int node_id)
  {
c86d1b8ae   Mike Snitzer   block: avoid unco...
783
  	struct request_queue *uninit_q, *q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
784

c86d1b8ae   Mike Snitzer   block: avoid unco...
785
786
787
  	uninit_q = blk_alloc_queue_node(GFP_KERNEL, node_id);
  	if (!uninit_q)
  		return NULL;
5151412dd   Mike Snitzer   block: initialize...
788
  	q = blk_init_allocated_queue(uninit_q, rfn, lock);
c86d1b8ae   Mike Snitzer   block: avoid unco...
789
  	if (!q)
7982e90c3   Mike Snitzer   block: fix q->flu...
790
  		blk_cleanup_queue(uninit_q);
18741986a   Christoph Hellwig   blk-mq: rework fl...
791

7982e90c3   Mike Snitzer   block: fix q->flu...
792
  	return q;
01effb0dc   Mike Snitzer   block: allow init...
793
794
  }
  EXPORT_SYMBOL(blk_init_queue_node);
dece16353   Jens Axboe   block: change ->m...
795
  static blk_qc_t blk_queue_bio(struct request_queue *q, struct bio *bio);
336b7e1f2   Mike Snitzer   block: remove exp...
796

01effb0dc   Mike Snitzer   block: allow init...
797
798
799
800
  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
801
802
  	if (!q)
  		return NULL;
f70ced091   Ming Lei   blk-mq: support p...
803
  	q->fq = blk_alloc_flush_queue(q, NUMA_NO_NODE, 0);
ba483388e   Ming Lei   block: remove blk...
804
  	if (!q->fq)
7982e90c3   Mike Snitzer   block: fix q->flu...
805
  		return NULL;
a051661ca   Tejun Heo   blkcg: implement ...
806
  	if (blk_init_rl(&q->root_rl, q, GFP_KERNEL))
708f04d2a   Dave Jones   block: free q->fl...
807
  		goto fail;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
808

287922eb0   Christoph Hellwig   block: defer time...
809
  	INIT_WORK(&q->timeout_work, blk_timeout_work);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
810
  	q->request_fn		= rfn;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
811
  	q->prep_rq_fn		= NULL;
28018c242   James Bottomley   block: implement ...
812
  	q->unprep_rq_fn		= NULL;
60ea8226c   Tejun Heo   block: fix reques...
813
  	q->queue_flags		|= QUEUE_FLAG_DEFAULT;
c94a96ac9   Vivek Goyal   block: Initialize...
814
815
816
817
  
  	/* Override internal queue lock with supplied lock pointer */
  	if (lock)
  		q->queue_lock		= lock;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
818

f3b144aa7   Jens Axboe   block: remove var...
819
820
821
  	/*
  	 * This also sets hw/phys segments, boundary and size
  	 */
c20e8de27   Jens Axboe   block: rename __m...
822
  	blk_queue_make_request(q, blk_queue_bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
823

44ec95425   Alan Stern   [SCSI] sg: cap re...
824
  	q->sg_reserved_size = INT_MAX;
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
825
826
  	/* Protect q->elevator from elevator_change */
  	mutex_lock(&q->sysfs_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
827
  	/* init elevator */
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
828
829
  	if (elevator_init(q, NULL)) {
  		mutex_unlock(&q->sysfs_lock);
708f04d2a   Dave Jones   block: free q->fl...
830
  		goto fail;
eb1c160b2   Tomoki Sekiyama   elevator: Fix a r...
831
832
833
  	}
  
  	mutex_unlock(&q->sysfs_lock);
b82d4b197   Tejun Heo   blkcg: make reque...
834
  	return q;
708f04d2a   Dave Jones   block: free q->fl...
835
836
  
  fail:
ba483388e   Ming Lei   block: remove blk...
837
  	blk_free_flush_queue(q->fq);
708f04d2a   Dave Jones   block: free q->fl...
838
  	return NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
839
  }
5151412dd   Mike Snitzer   block: initialize...
840
  EXPORT_SYMBOL(blk_init_allocated_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
841

09ac46c42   Tejun Heo   block: misc updat...
842
  bool blk_get_queue(struct request_queue *q)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
843
  {
3f3299d5c   Bart Van Assche   block: Rename que...
844
  	if (likely(!blk_queue_dying(q))) {
09ac46c42   Tejun Heo   block: misc updat...
845
846
  		__blk_get_queue(q);
  		return true;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
847
  	}
09ac46c42   Tejun Heo   block: misc updat...
848
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
849
  }
d86e0e83b   Jens Axboe   block: export blk...
850
  EXPORT_SYMBOL(blk_get_queue);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
851

5b788ce3e   Tejun Heo   block: prepare fo...
852
  static inline void blk_free_request(struct request_list *rl, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
853
  {
f1f8cc946   Tejun Heo   block, cfq: move ...
854
  	if (rq->cmd_flags & REQ_ELVPRIV) {
5b788ce3e   Tejun Heo   block: prepare fo...
855
  		elv_put_request(rl->q, rq);
f1f8cc946   Tejun Heo   block, cfq: move ...
856
  		if (rq->elv.icq)
11a3122f6   Tejun Heo   block: strip out ...
857
  			put_io_context(rq->elv.icq->ioc);
f1f8cc946   Tejun Heo   block, cfq: move ...
858
  	}
5b788ce3e   Tejun Heo   block: prepare fo...
859
  	mempool_free(rq, rl->rq_pool);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
860
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
861
862
863
864
  /*
   * 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...
865
  static inline int ioc_batching(struct request_queue *q, struct io_context *ioc)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
  {
  	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...
886
  static void ioc_set_batching(struct request_queue *q, struct io_context *ioc)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
887
888
889
890
891
892
893
  {
  	if (!ioc || ioc_batching(q, ioc))
  		return;
  
  	ioc->nr_batch_requests = q->nr_batching;
  	ioc->last_waited = jiffies;
  }
5b788ce3e   Tejun Heo   block: prepare fo...
894
  static void __freed_request(struct request_list *rl, int sync)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
895
  {
5b788ce3e   Tejun Heo   block: prepare fo...
896
  	struct request_queue *q = rl->q;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
897

d40f75a06   Tejun Heo   writeback, blkcg:...
898
899
  	if (rl->count[sync] < queue_congestion_off_threshold(q))
  		blk_clear_congested(rl, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
900

1faa16d22   Jens Axboe   block: change the...
901
902
903
  	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
904

5b788ce3e   Tejun Heo   block: prepare fo...
905
  		blk_clear_rl_full(rl, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
906
907
908
909
910
911
912
  	}
  }
  
  /*
   * A request has just been released.  Account for it, update the full and
   * congestion status, wake up any waiters.   Called under q->queue_lock.
   */
e6a40b096   Mike Christie   block: prepare re...
913
  static void freed_request(struct request_list *rl, int op, unsigned int flags)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
914
  {
5b788ce3e   Tejun Heo   block: prepare fo...
915
  	struct request_queue *q = rl->q;
d9d8c5c48   Mike Christie   block: convert is...
916
  	int sync = rw_is_sync(op, flags);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
917

8a5ecdd42   Tejun Heo   block: add q->nr_...
918
  	q->nr_rqs[sync]--;
1faa16d22   Jens Axboe   block: change the...
919
  	rl->count[sync]--;
75eb6c372   Tejun Heo   block: pass aroun...
920
  	if (flags & REQ_ELVPRIV)
8a5ecdd42   Tejun Heo   block: add q->nr_...
921
  		q->nr_rqs_elvpriv--;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
922

5b788ce3e   Tejun Heo   block: prepare fo...
923
  	__freed_request(rl, sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
924

1faa16d22   Jens Axboe   block: change the...
925
  	if (unlikely(rl->starved[sync ^ 1]))
5b788ce3e   Tejun Heo   block: prepare fo...
926
  		__freed_request(rl, sync ^ 1);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
927
  }
e3a2b3f93   Jens Axboe   blk-mq: allow cha...
928
929
930
  int blk_update_nr_requests(struct request_queue *q, unsigned int nr)
  {
  	struct request_list *rl;
d40f75a06   Tejun Heo   writeback, blkcg:...
931
  	int on_thresh, off_thresh;
e3a2b3f93   Jens Axboe   blk-mq: allow cha...
932
933
934
935
  
  	spin_lock_irq(q->queue_lock);
  	q->nr_requests = nr;
  	blk_queue_congestion_threshold(q);
d40f75a06   Tejun Heo   writeback, blkcg:...
936
937
  	on_thresh = queue_congestion_on_threshold(q);
  	off_thresh = queue_congestion_off_threshold(q);
e3a2b3f93   Jens Axboe   blk-mq: allow cha...
938

d40f75a06   Tejun Heo   writeback, blkcg:...
939
940
941
942
943
  	blk_queue_for_each_rl(rl, q) {
  		if (rl->count[BLK_RW_SYNC] >= on_thresh)
  			blk_set_congested(rl, BLK_RW_SYNC);
  		else if (rl->count[BLK_RW_SYNC] < off_thresh)
  			blk_clear_congested(rl, BLK_RW_SYNC);
e3a2b3f93   Jens Axboe   blk-mq: allow cha...
944

d40f75a06   Tejun Heo   writeback, blkcg:...
945
946
947
948
  		if (rl->count[BLK_RW_ASYNC] >= on_thresh)
  			blk_set_congested(rl, BLK_RW_ASYNC);
  		else if (rl->count[BLK_RW_ASYNC] < off_thresh)
  			blk_clear_congested(rl, BLK_RW_ASYNC);
e3a2b3f93   Jens Axboe   blk-mq: allow cha...
949

e3a2b3f93   Jens Axboe   blk-mq: allow cha...
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
  		if (rl->count[BLK_RW_SYNC] >= q->nr_requests) {
  			blk_set_rl_full(rl, BLK_RW_SYNC);
  		} else {
  			blk_clear_rl_full(rl, BLK_RW_SYNC);
  			wake_up(&rl->wait[BLK_RW_SYNC]);
  		}
  
  		if (rl->count[BLK_RW_ASYNC] >= q->nr_requests) {
  			blk_set_rl_full(rl, BLK_RW_ASYNC);
  		} else {
  			blk_clear_rl_full(rl, BLK_RW_ASYNC);
  			wake_up(&rl->wait[BLK_RW_ASYNC]);
  		}
  	}
  
  	spin_unlock_irq(q->queue_lock);
  	return 0;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
968
  /*
9d5a4e946   Mike Snitzer   block: skip eleva...
969
970
971
972
973
974
975
976
977
978
979
980
   * 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.
  	 */
1eff9d322   Jens Axboe   block: rename bio...
981
  	if (bio->bi_opf & (REQ_PREFLUSH | REQ_FUA))
9d5a4e946   Mike Snitzer   block: skip eleva...
982
983
984
985
  		return false;
  
  	return true;
  }
da8303c63   Tejun Heo   block: make get_r...
986
  /**
852c788f8   Tejun Heo   block: implement ...
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
   * 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...
1003
   * __get_request - get a free request
5b788ce3e   Tejun Heo   block: prepare fo...
1004
   * @rl: request list to allocate from
e6a40b096   Mike Christie   block: prepare re...
1005
1006
   * @op: REQ_OP_READ/REQ_OP_WRITE
   * @op_flags: rq_flag_bits
da8303c63   Tejun Heo   block: make get_r...
1007
1008
1009
1010
1011
1012
   * @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.
   *
da3dae54e   Masanari Iida   Documentation: Do...
1013
   * Must be called with @q->queue_lock held and,
a492f0754   Joe Lawrence   block,scsi: fixup...
1014
1015
   * Returns ERR_PTR on failure, with @q->queue_lock held.
   * Returns request pointer on success, with @q->queue_lock *not held*.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1016
   */
e6a40b096   Mike Christie   block: prepare re...
1017
1018
1019
  static struct request *__get_request(struct request_list *rl, int op,
  				     int op_flags, struct bio *bio,
  				     gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1020
  {
5b788ce3e   Tejun Heo   block: prepare fo...
1021
  	struct request_queue *q = rl->q;
b679281a6   Tejun Heo   block: restructur...
1022
  	struct request *rq;
7f4b35d15   Tejun Heo   block: allocate i...
1023
1024
  	struct elevator_type *et = q->elevator->type;
  	struct io_context *ioc = rq_ioc(bio);
f1f8cc946   Tejun Heo   block, cfq: move ...
1025
  	struct io_cq *icq = NULL;
d9d8c5c48   Mike Christie   block: convert is...
1026
  	const bool is_sync = rw_is_sync(op, op_flags) != 0;
75eb6c372   Tejun Heo   block: pass aroun...
1027
  	int may_queue;
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1028

3f3299d5c   Bart Van Assche   block: Rename que...
1029
  	if (unlikely(blk_queue_dying(q)))
a492f0754   Joe Lawrence   block,scsi: fixup...
1030
  		return ERR_PTR(-ENODEV);
da8303c63   Tejun Heo   block: make get_r...
1031

ba568ea0a   Mike Christie   block: prepare el...
1032
  	may_queue = elv_may_queue(q, op, op_flags);
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1033
1034
  	if (may_queue == ELV_MQUEUE_NO)
  		goto rq_starved;
1faa16d22   Jens Axboe   block: change the...
1035
1036
  	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...
1037
  			/*
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1038
1039
1040
1041
1042
  			 * 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...
1043
  			if (!blk_rl_full(rl, is_sync)) {
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1044
  				ioc_set_batching(q, ioc);
5b788ce3e   Tejun Heo   block: prepare fo...
1045
  				blk_set_rl_full(rl, is_sync);
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1046
1047
1048
1049
1050
1051
1052
1053
  			} else {
  				if (may_queue != ELV_MQUEUE_MUST
  						&& !ioc_batching(q, ioc)) {
  					/*
  					 * The queue is full and the allocating
  					 * process is not a "batcher", and not
  					 * exempted by the IO scheduler
  					 */
a492f0754   Joe Lawrence   block,scsi: fixup...
1054
  					return ERR_PTR(-ENOMEM);
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1055
1056
  				}
  			}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1057
  		}
d40f75a06   Tejun Heo   writeback, blkcg:...
1058
  		blk_set_congested(rl, is_sync);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1059
  	}
082cf69eb   Jens Axboe   [PATCH] ll_rw_blk...
1060
1061
1062
1063
1064
  	/*
  	 * 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...
1065
  	if (rl->count[is_sync] >= (3 * q->nr_requests / 2))
a492f0754   Joe Lawrence   block,scsi: fixup...
1066
  		return ERR_PTR(-ENOMEM);
fd782a4a9   Hugh Dickins   [PATCH] Fix get_r...
1067

8a5ecdd42   Tejun Heo   block: add q->nr_...
1068
  	q->nr_rqs[is_sync]++;
1faa16d22   Jens Axboe   block: change the...
1069
1070
  	rl->count[is_sync]++;
  	rl->starved[is_sync] = 0;
cb98fc8bb   Tejun Heo   [BLOCK] Reimpleme...
1071

f1f8cc946   Tejun Heo   block, cfq: move ...
1072
1073
  	/*
  	 * Decide whether the new request will be managed by elevator.  If
e6a40b096   Mike Christie   block: prepare re...
1074
  	 * so, mark @op_flags and increment elvpriv.  Non-zero elvpriv will
f1f8cc946   Tejun Heo   block, cfq: move ...
1075
1076
1077
1078
1079
1080
1081
  	 * 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 ...
1082
  	if (blk_rq_should_init_elevator(bio) && !blk_queue_bypass(q)) {
e6a40b096   Mike Christie   block: prepare re...
1083
  		op_flags |= REQ_ELVPRIV;
8a5ecdd42   Tejun Heo   block: add q->nr_...
1084
  		q->nr_rqs_elvpriv++;
f1f8cc946   Tejun Heo   block, cfq: move ...
1085
1086
  		if (et->icq_cache && ioc)
  			icq = ioc_lookup_icq(ioc, q);
9d5a4e946   Mike Snitzer   block: skip eleva...
1087
  	}
cb98fc8bb   Tejun Heo   [BLOCK] Reimpleme...
1088

f253b86b4   Jens Axboe   Revert "block: fi...
1089
  	if (blk_queue_io_stat(q))
e6a40b096   Mike Christie   block: prepare re...
1090
  		op_flags |= REQ_IO_STAT;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1091
  	spin_unlock_irq(q->queue_lock);
29e2b09ab   Tejun Heo   block: collapse b...
1092
  	/* allocate and init request */
5b788ce3e   Tejun Heo   block: prepare fo...
1093
  	rq = mempool_alloc(rl->rq_pool, gfp_mask);
29e2b09ab   Tejun Heo   block: collapse b...
1094
  	if (!rq)
b679281a6   Tejun Heo   block: restructur...
1095
  		goto fail_alloc;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1096

29e2b09ab   Tejun Heo   block: collapse b...
1097
  	blk_rq_init(q, rq);
a051661ca   Tejun Heo   blkcg: implement ...
1098
  	blk_rq_set_rl(rq, rl);
e6a40b096   Mike Christie   block: prepare re...
1099
  	req_set_op_attrs(rq, op, op_flags | REQ_ALLOCED);
29e2b09ab   Tejun Heo   block: collapse b...
1100

aaf7c6806   Tejun Heo   block: fix elvpri...
1101
  	/* init elvpriv */
e6a40b096   Mike Christie   block: prepare re...
1102
  	if (op_flags & REQ_ELVPRIV) {
aaf7c6806   Tejun Heo   block: fix elvpri...
1103
  		if (unlikely(et->icq_cache && !icq)) {
7f4b35d15   Tejun Heo   block: allocate i...
1104
1105
  			if (ioc)
  				icq = ioc_create_icq(ioc, q, gfp_mask);
aaf7c6806   Tejun Heo   block: fix elvpri...
1106
1107
  			if (!icq)
  				goto fail_elvpriv;
29e2b09ab   Tejun Heo   block: collapse b...
1108
  		}
aaf7c6806   Tejun Heo   block: fix elvpri...
1109
1110
1111
1112
1113
1114
  
  		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...
1115
1116
1117
  		if (icq)
  			get_io_context(icq->ioc);
  	}
aaf7c6806   Tejun Heo   block: fix elvpri...
1118
  out:
88ee5ef15   Jens Axboe   [BLOCK] ll_rw_blk...
1119
1120
1121
1122
1123
1124
  	/*
  	 * 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
1125
1126
  	if (ioc_batching(q, ioc))
  		ioc->nr_batch_requests--;
6728cb0e6   Jens Axboe   block: make core ...
1127

e6a40b096   Mike Christie   block: prepare re...
1128
  	trace_block_getrq(q, bio, op);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1129
  	return rq;
b679281a6   Tejun Heo   block: restructur...
1130

aaf7c6806   Tejun Heo   block: fix elvpri...
1131
1132
1133
1134
1135
1136
1137
  fail_elvpriv:
  	/*
  	 * elvpriv init failed.  ioc, icq and elvpriv aren't mempool backed
  	 * and may fail indefinitely under memory pressure and thus
  	 * shouldn't stall IO.  Treat this request as !elvpriv.  This will
  	 * disturb iosched and blkcg but weird is bettern than dead.
  	 */
7b2b10e0e   Robert Elliott   block: include fu...
1138
1139
1140
  	printk_ratelimited(KERN_WARNING "%s: dev %s: request aux data allocation failed, iosched may be disturbed
  ",
  			   __func__, dev_name(q->backing_dev_info.dev));
aaf7c6806   Tejun Heo   block: fix elvpri...
1141
1142
1143
1144
1145
  
  	rq->cmd_flags &= ~REQ_ELVPRIV;
  	rq->elv.icq = NULL;
  
  	spin_lock_irq(q->queue_lock);
8a5ecdd42   Tejun Heo   block: add q->nr_...
1146
  	q->nr_rqs_elvpriv--;
aaf7c6806   Tejun Heo   block: fix elvpri...
1147
1148
  	spin_unlock_irq(q->queue_lock);
  	goto out;
b679281a6   Tejun Heo   block: restructur...
1149
1150
1151
1152
1153
1154
1155
1156
1157
  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);
e6a40b096   Mike Christie   block: prepare re...
1158
  	freed_request(rl, op, op_flags);
b679281a6   Tejun Heo   block: restructur...
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
  
  	/*
  	 * in the very unlikely event that allocation failed and no
  	 * requests for this direction was pending, mark us starved so that
  	 * freeing of a request in the other direction will notice
  	 * us. another possible fix would be to split the rq mempool into
  	 * READ and WRITE
  	 */
  rq_starved:
  	if (unlikely(rl->count[is_sync] == 0))
  		rl->starved[is_sync] = 1;
a492f0754   Joe Lawrence   block,scsi: fixup...
1170
  	return ERR_PTR(-ENOMEM);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1171
  }
da8303c63   Tejun Heo   block: make get_r...
1172
  /**
a06e05e6a   Tejun Heo   block: refactor g...
1173
   * get_request - get a free request
da8303c63   Tejun Heo   block: make get_r...
1174
   * @q: request_queue to allocate request from
e6a40b096   Mike Christie   block: prepare re...
1175
1176
   * @op: REQ_OP_READ/REQ_OP_WRITE
   * @op_flags: rq_flag_bits
da8303c63   Tejun Heo   block: make get_r...
1177
   * @bio: bio to allocate request for (can be %NULL)
a06e05e6a   Tejun Heo   block: refactor g...
1178
   * @gfp_mask: allocation mask
da8303c63   Tejun Heo   block: make get_r...
1179
   *
d0164adc8   Mel Gorman   mm, page_alloc: d...
1180
1181
   * Get a free request from @q.  If %__GFP_DIRECT_RECLAIM is set in @gfp_mask,
   * this function keeps retrying under memory pressure and fails iff @q is dead.
d6344532a   Nick Piggin   [PATCH] blk: redu...
1182
   *
da3dae54e   Masanari Iida   Documentation: Do...
1183
   * Must be called with @q->queue_lock held and,
a492f0754   Joe Lawrence   block,scsi: fixup...
1184
1185
   * Returns ERR_PTR on failure, with @q->queue_lock held.
   * Returns request pointer on success, with @q->queue_lock *not held*.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1186
   */
e6a40b096   Mike Christie   block: prepare re...
1187
1188
1189
  static struct request *get_request(struct request_queue *q, int op,
  				   int op_flags, struct bio *bio,
  				   gfp_t gfp_mask)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1190
  {
d9d8c5c48   Mike Christie   block: convert is...
1191
  	const bool is_sync = rw_is_sync(op, op_flags) != 0;
a06e05e6a   Tejun Heo   block: refactor g...
1192
  	DEFINE_WAIT(wait);
a051661ca   Tejun Heo   blkcg: implement ...
1193
  	struct request_list *rl;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1194
  	struct request *rq;
a051661ca   Tejun Heo   blkcg: implement ...
1195
1196
  
  	rl = blk_get_rl(q, bio);	/* transferred to @rq on success */
a06e05e6a   Tejun Heo   block: refactor g...
1197
  retry:
e6a40b096   Mike Christie   block: prepare re...
1198
  	rq = __get_request(rl, op, op_flags, bio, gfp_mask);
a492f0754   Joe Lawrence   block,scsi: fixup...
1199
  	if (!IS_ERR(rq))
a06e05e6a   Tejun Heo   block: refactor g...
1200
  		return rq;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1201

d0164adc8   Mel Gorman   mm, page_alloc: d...
1202
  	if (!gfpflags_allow_blocking(gfp_mask) || unlikely(blk_queue_dying(q))) {
a051661ca   Tejun Heo   blkcg: implement ...
1203
  		blk_put_rl(rl);
a492f0754   Joe Lawrence   block,scsi: fixup...
1204
  		return rq;
a051661ca   Tejun Heo   blkcg: implement ...
1205
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1206

a06e05e6a   Tejun Heo   block: refactor g...
1207
1208
1209
  	/* wait on @rl and retry */
  	prepare_to_wait_exclusive(&rl->wait[is_sync], &wait,
  				  TASK_UNINTERRUPTIBLE);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1210

e6a40b096   Mike Christie   block: prepare re...
1211
  	trace_block_sleeprq(q, bio, op);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1212

a06e05e6a   Tejun Heo   block: refactor g...
1213
1214
  	spin_unlock_irq(q->queue_lock);
  	io_schedule();
d6344532a   Nick Piggin   [PATCH] blk: redu...
1215

a06e05e6a   Tejun Heo   block: refactor g...
1216
1217
1218
1219
1220
  	/*
  	 * 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...
1221
  	ioc_set_batching(q, current->io_context);
05caf8dbc   Zhang, Yanmin   block: Move the s...
1222

a06e05e6a   Tejun Heo   block: refactor g...
1223
1224
  	spin_lock_irq(q->queue_lock);
  	finish_wait(&rl->wait[is_sync], &wait);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1225

a06e05e6a   Tejun Heo   block: refactor g...
1226
  	goto retry;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1227
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1228
1229
  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
1230
1231
1232
1233
  {
  	struct request *rq;
  
  	BUG_ON(rw != READ && rw != WRITE);
7f4b35d15   Tejun Heo   block: allocate i...
1234
1235
  	/* create ioc upfront */
  	create_io_context(gfp_mask, q->node);
d6344532a   Nick Piggin   [PATCH] blk: redu...
1236
  	spin_lock_irq(q->queue_lock);
e6a40b096   Mike Christie   block: prepare re...
1237
  	rq = get_request(q, rw, 0, NULL, gfp_mask);
0c4de0f33   Christoph Hellwig   block: ensure bio...
1238
  	if (IS_ERR(rq)) {
da8303c63   Tejun Heo   block: make get_r...
1239
  		spin_unlock_irq(q->queue_lock);
0c4de0f33   Christoph Hellwig   block: ensure bio...
1240
1241
  		return rq;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1242

0c4de0f33   Christoph Hellwig   block: ensure bio...
1243
1244
1245
1246
  	/* q->queue_lock is unlocked at this point */
  	rq->__data_len = 0;
  	rq->__sector = (sector_t) -1;
  	rq->bio = rq->biotail = NULL;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1247
1248
  	return rq;
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1249
1250
1251
1252
  
  struct request *blk_get_request(struct request_queue *q, int rw, gfp_t gfp_mask)
  {
  	if (q->mq_ops)
6f3b0e8bc   Christoph Hellwig   blk-mq: add a fla...
1253
1254
1255
  		return blk_mq_alloc_request(q, rw,
  			(gfp_mask & __GFP_DIRECT_RECLAIM) ?
  				0 : BLK_MQ_REQ_NOWAIT);
320ae51fe   Jens Axboe   blk-mq: new multi...
1256
1257
1258
  	else
  		return blk_old_get_request(q, rw, gfp_mask);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1259
1260
1261
  EXPORT_SYMBOL(blk_get_request);
  
  /**
da3dae54e   Masanari Iida   Documentation: Do...
1262
   * blk_rq_set_block_pc - initialize a request to type BLOCK_PC
f27b087b8   Jens Axboe   block: add blk_rq...
1263
1264
1265
1266
1267
1268
   * @rq:		request to be initialized
   *
   */
  void blk_rq_set_block_pc(struct request *rq)
  {
  	rq->cmd_type = REQ_TYPE_BLOCK_PC;
f27b087b8   Jens Axboe   block: add blk_rq...
1269
  	memset(rq->__cmd, 0, sizeof(rq->__cmd));
f27b087b8   Jens Axboe   block: add blk_rq...
1270
1271
1272
1273
  }
  EXPORT_SYMBOL(blk_rq_set_block_pc);
  
  /**
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1274
1275
1276
1277
1278
1279
1280
1281
1282
   * 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...
1283
  void blk_requeue_request(struct request_queue *q, struct request *rq)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1284
  {
242f9dcb8   Jens Axboe   block: unify requ...
1285
1286
  	blk_delete_timer(rq);
  	blk_clear_rq_complete(rq);
5f3ea37c7   Arnaldo Carvalho de Melo   blktrace: port to...
1287
  	trace_block_rq_requeue(q, rq);
2056a782f   Jens Axboe   [PATCH] Block que...
1288

125c99bc8   Christoph Hellwig   scsi: add new scs...
1289
  	if (rq->cmd_flags & REQ_QUEUED)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1290
  		blk_queue_end_tag(q, rq);
ba396a6c1   James Bottomley   block: fix oops w...
1291
  	BUG_ON(blk_queued_rq(rq));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1292
1293
  	elv_requeue_request(q, rq);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1294
  EXPORT_SYMBOL(blk_requeue_request);
73c101011   Jens Axboe   block: initial pa...
1295
1296
1297
  static void add_acct_request(struct request_queue *q, struct request *rq,
  			     int where)
  {
320ae51fe   Jens Axboe   blk-mq: new multi...
1298
  	blk_account_io_start(rq, true);
7eaceacca   Jens Axboe   block: remove per...
1299
  	__elv_add_request(q, rq, where);
73c101011   Jens Axboe   block: initial pa...
1300
  }
074a7aca7   Tejun Heo   block: move stats...
1301
1302
1303
  static void part_round_stats_single(int cpu, struct hd_struct *part,
  				    unsigned long now)
  {
7276d02e2   Jens Axboe   block: only calcu...
1304
  	int inflight;
074a7aca7   Tejun Heo   block: move stats...
1305
1306
  	if (now == part->stamp)
  		return;
7276d02e2   Jens Axboe   block: only calcu...
1307
1308
  	inflight = part_in_flight(part);
  	if (inflight) {
074a7aca7   Tejun Heo   block: move stats...
1309
  		__part_stat_add(cpu, part, time_in_queue,
7276d02e2   Jens Axboe   block: only calcu...
1310
  				inflight * (now - part->stamp));
074a7aca7   Tejun Heo   block: move stats...
1311
1312
1313
1314
1315
1316
  		__part_stat_add(cpu, part, io_ticks, (now - part->stamp));
  	}
  	part->stamp = now;
  }
  
  /**
496aa8a98   Randy Dunlap   block: fix curren...
1317
1318
1319
   * 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
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
   *
   * 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...
1332
  void part_round_stats(int cpu, struct hd_struct *part)
6f2576af5   Jerome Marchand   Enhanced partitio...
1333
1334
  {
  	unsigned long now = jiffies;
074a7aca7   Tejun Heo   block: move stats...
1335
1336
1337
  	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...
1338
  }
074a7aca7   Tejun Heo   block: move stats...
1339
  EXPORT_SYMBOL_GPL(part_round_stats);
6f2576af5   Jerome Marchand   Enhanced partitio...
1340

47fafbc70   Rafael J. Wysocki   block / PM: Repla...
1341
  #ifdef CONFIG_PM
c8158819d   Lin Ming   block: implement ...
1342
1343
1344
1345
1346
1347
1348
1349
  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
1350
1351
1352
  /*
   * queue lock must be held
   */
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1353
  void __blk_put_request(struct request_queue *q, struct request *req)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1354
  {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1355
1356
  	if (unlikely(!q))
  		return;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1357

6f5ba581c   Christoph Hellwig   blk-mq: divert __...
1358
1359
1360
1361
  	if (q->mq_ops) {
  		blk_mq_free_request(req);
  		return;
  	}
c8158819d   Lin Ming   block: implement ...
1362
  	blk_pm_put_request(req);
8922e16cf   Tejun Heo   [PATCH] 01/05 Imp...
1363
  	elv_completed_request(q, req);
1cd96c242   Boaz Harrosh   block: WARN in __...
1364
1365
  	/* this is a bio leak */
  	WARN_ON(req->bio != NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1366
1367
1368
1369
  	/*
  	 * 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...
1370
  	if (req->cmd_flags & REQ_ALLOCED) {
75eb6c372   Tejun Heo   block: pass aroun...
1371
  		unsigned int flags = req->cmd_flags;
e6a40b096   Mike Christie   block: prepare re...
1372
  		int op = req_op(req);
a051661ca   Tejun Heo   blkcg: implement ...
1373
  		struct request_list *rl = blk_rq_rl(req);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1374

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1375
  		BUG_ON(!list_empty(&req->queuelist));
360f92c24   Jens Axboe   block: fix regres...
1376
  		BUG_ON(ELV_ON_HASH(req));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1377

a051661ca   Tejun Heo   blkcg: implement ...
1378
  		blk_free_request(rl, req);
e6a40b096   Mike Christie   block: prepare re...
1379
  		freed_request(rl, op, flags);
a051661ca   Tejun Heo   blkcg: implement ...
1380
  		blk_put_rl(rl);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1381
1382
  	}
  }
6e39b69e7   Mike Christie   [SCSI] export blk...
1383
  EXPORT_SYMBOL_GPL(__blk_put_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1384
1385
  void blk_put_request(struct request *req)
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1386
  	struct request_queue *q = req->q;
8922e16cf   Tejun Heo   [PATCH] 01/05 Imp...
1387

320ae51fe   Jens Axboe   blk-mq: new multi...
1388
1389
1390
1391
1392
1393
1394
1395
1396
  	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
1397
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1398
  EXPORT_SYMBOL(blk_put_request);
66ac02801   Christoph Hellwig   block: don't allo...
1399
1400
1401
1402
  /**
   * blk_add_request_payload - add a payload to a request
   * @rq: request to update
   * @page: page backing the payload
37e58237a   Ming Lin   block: add offset...
1403
   * @offset: offset in page
66ac02801   Christoph Hellwig   block: don't allo...
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
   * @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,
37e58237a   Ming Lin   block: add offset...
1414
  		int offset, unsigned int len)
66ac02801   Christoph Hellwig   block: don't allo...
1415
1416
1417
1418
  {
  	struct bio *bio = rq->bio;
  
  	bio->bi_io_vec->bv_page = page;
37e58237a   Ming Lin   block: add offset...
1419
  	bio->bi_io_vec->bv_offset = offset;
66ac02801   Christoph Hellwig   block: don't allo...
1420
  	bio->bi_io_vec->bv_len = len;
4f024f379   Kent Overstreet   block: Abstract o...
1421
  	bio->bi_iter.bi_size = len;
66ac02801   Christoph Hellwig   block: don't allo...
1422
1423
1424
1425
1426
  	bio->bi_vcnt = 1;
  	bio->bi_phys_segments = 1;
  
  	rq->__data_len = rq->resid_len = len;
  	rq->nr_phys_segments = 1;
66ac02801   Christoph Hellwig   block: don't allo...
1427
1428
  }
  EXPORT_SYMBOL_GPL(blk_add_request_payload);
320ae51fe   Jens Axboe   blk-mq: new multi...
1429
1430
  bool bio_attempt_back_merge(struct request_queue *q, struct request *req,
  			    struct bio *bio)
73c101011   Jens Axboe   block: initial pa...
1431
  {
1eff9d322   Jens Axboe   block: rename bio...
1432
  	const int ff = bio->bi_opf & REQ_FAILFAST_MASK;
73c101011   Jens Axboe   block: initial pa...
1433

73c101011   Jens Axboe   block: initial pa...
1434
1435
  	if (!ll_back_merge_fn(q, req, bio))
  		return false;
8c1cf6bb0   Tejun Heo   block: add @req t...
1436
  	trace_block_bio_backmerge(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1437
1438
1439
1440
1441
1442
  
  	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...
1443
  	req->__data_len += bio->bi_iter.bi_size;
73c101011   Jens Axboe   block: initial pa...
1444
  	req->ioprio = ioprio_best(req->ioprio, bio_prio(bio));
320ae51fe   Jens Axboe   blk-mq: new multi...
1445
  	blk_account_io_start(req, false);
73c101011   Jens Axboe   block: initial pa...
1446
1447
  	return true;
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
1448
1449
  bool bio_attempt_front_merge(struct request_queue *q, struct request *req,
  			     struct bio *bio)
73c101011   Jens Axboe   block: initial pa...
1450
  {
1eff9d322   Jens Axboe   block: rename bio...
1451
  	const int ff = bio->bi_opf & REQ_FAILFAST_MASK;
73c101011   Jens Axboe   block: initial pa...
1452

73c101011   Jens Axboe   block: initial pa...
1453
1454
  	if (!ll_front_merge_fn(q, req, bio))
  		return false;
8c1cf6bb0   Tejun Heo   block: add @req t...
1455
  	trace_block_bio_frontmerge(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1456
1457
1458
  
  	if ((req->cmd_flags & REQ_FAILFAST_MASK) != ff)
  		blk_rq_set_mixed_merge(req);
73c101011   Jens Axboe   block: initial pa...
1459
1460
  	bio->bi_next = req->bio;
  	req->bio = bio;
4f024f379   Kent Overstreet   block: Abstract o...
1461
1462
  	req->__sector = bio->bi_iter.bi_sector;
  	req->__data_len += bio->bi_iter.bi_size;
73c101011   Jens Axboe   block: initial pa...
1463
  	req->ioprio = ioprio_best(req->ioprio, bio_prio(bio));
320ae51fe   Jens Axboe   blk-mq: new multi...
1464
  	blk_account_io_start(req, false);
73c101011   Jens Axboe   block: initial pa...
1465
1466
  	return true;
  }
bd87b5898   Tejun Heo   block: drop @tsk ...
1467
  /**
320ae51fe   Jens Axboe   blk-mq: new multi...
1468
   * blk_attempt_plug_merge - try to merge with %current's plugged list
bd87b5898   Tejun Heo   block: drop @tsk ...
1469
1470
1471
   * @q: request_queue new bio is being queued at
   * @bio: new bio being queued
   * @request_count: out parameter for number of traversed plugged requests
ccc2600b8   Randy Dunlap   block: fix blk-co...
1472
1473
1474
   * @same_queue_rq: pointer to &struct request that gets filled in when
   * another request associated with @q is found on the plug list
   * (optional, may be %NULL)
bd87b5898   Tejun Heo   block: drop @tsk ...
1475
1476
1477
1478
1479
   *
   * 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...
1480
1481
1482
1483
1484
1485
   * Plugging coalesces IOs from the same issuer for the same purpose without
   * going through @q->queue_lock.  As such it's more of an issuing mechanism
   * than scheduling, and the request, while may have elvpriv data, is not
   * added on the elevator at this point.  In addition, we don't have
   * reliable access to the elevator outside queue lock.  Only check basic
   * merging parameters without querying the elevator.
da41a589f   Robert Elliott   blk-mq: Micro-opt...
1486
1487
   *
   * Caller must ensure !blk_queue_nomerges(q) beforehand.
73c101011   Jens Axboe   block: initial pa...
1488
   */
320ae51fe   Jens Axboe   blk-mq: new multi...
1489
  bool blk_attempt_plug_merge(struct request_queue *q, struct bio *bio,
5b3f341f0   Shaohua Li   blk-mq: make plug...
1490
1491
  			    unsigned int *request_count,
  			    struct request **same_queue_rq)
73c101011   Jens Axboe   block: initial pa...
1492
1493
1494
1495
  {
  	struct blk_plug *plug;
  	struct request *rq;
  	bool ret = false;
92f399c72   Shaohua Li   blk-mq: mq plug l...
1496
  	struct list_head *plug_list;
73c101011   Jens Axboe   block: initial pa...
1497

bd87b5898   Tejun Heo   block: drop @tsk ...
1498
  	plug = current->plug;
73c101011   Jens Axboe   block: initial pa...
1499
1500
  	if (!plug)
  		goto out;
56ebdaf2f   Shaohua Li   block: simplify f...
1501
  	*request_count = 0;
73c101011   Jens Axboe   block: initial pa...
1502

92f399c72   Shaohua Li   blk-mq: mq plug l...
1503
1504
1505
1506
1507
1508
  	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...
1509
  		int el_ret;
5b3f341f0   Shaohua Li   blk-mq: make plug...
1510
  		if (rq->q == q) {
1b2e19f17   Shaohua Li   block: make auto ...
1511
  			(*request_count)++;
5b3f341f0   Shaohua Li   blk-mq: make plug...
1512
1513
1514
1515
1516
1517
1518
1519
  			/*
  			 * Only blk-mq multiple hardware queues case checks the
  			 * rq in the same queue, there should be only one such
  			 * rq in a queue
  			 **/
  			if (same_queue_rq)
  				*same_queue_rq = rq;
  		}
56ebdaf2f   Shaohua Li   block: simplify f...
1520

07c2bd373   Tejun Heo   block: don't call...
1521
  		if (rq->q != q || !blk_rq_merge_ok(rq, bio))
73c101011   Jens Axboe   block: initial pa...
1522
  			continue;
050c8ea80   Tejun Heo   block: separate o...
1523
  		el_ret = blk_try_merge(rq, bio);
73c101011   Jens Axboe   block: initial pa...
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
  		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;
  }
0809e3ac6   Jeff Moyer   block: fix plug l...
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
  unsigned int blk_plug_queued_count(struct request_queue *q)
  {
  	struct blk_plug *plug;
  	struct request *rq;
  	struct list_head *plug_list;
  	unsigned int ret = 0;
  
  	plug = current->plug;
  	if (!plug)
  		goto out;
  
  	if (q->mq_ops)
  		plug_list = &plug->mq_list;
  	else
  		plug_list = &plug->list;
  
  	list_for_each_entry(rq, plug_list, queuelist) {
  		if (rq->q == q)
  			ret++;
  	}
  out:
  	return ret;
  }
86db1e297   Jens Axboe   block: continue l...
1560
  void init_request_from_bio(struct request *req, struct bio *bio)
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1561
  {
4aff5e233   Jens Axboe   [PATCH] Split str...
1562
  	req->cmd_type = REQ_TYPE_FS;
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1563

1eff9d322   Jens Axboe   block: rename bio...
1564
1565
  	req->cmd_flags |= bio->bi_opf & REQ_COMMON_MASK;
  	if (bio->bi_opf & REQ_RAHEAD)
a82afdfcb   Tejun Heo   block: use the sa...
1566
  		req->cmd_flags |= REQ_FAILFAST_MASK;
b31dc66a5   Jens Axboe   [PATCH] Kill PF_S...
1567

52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1568
  	req->errors = 0;
4f024f379   Kent Overstreet   block: Abstract o...
1569
  	req->__sector = bio->bi_iter.bi_sector;
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1570
  	req->ioprio = bio_prio(bio);
bc1c56fde   NeilBrown   Share code betwee...
1571
  	blk_rq_bio_prep(req->q, req, bio);
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1572
  }
dece16353   Jens Axboe   block: change ->m...
1573
  static blk_qc_t blk_queue_bio(struct request_queue *q, struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1574
  {
1eff9d322   Jens Axboe   block: rename bio...
1575
  	const bool sync = !!(bio->bi_opf & REQ_SYNC);
73c101011   Jens Axboe   block: initial pa...
1576
  	struct blk_plug *plug;
e6a40b096   Mike Christie   block: prepare re...
1577
  	int el_ret, rw_flags = 0, where = ELEVATOR_INSERT_SORT;
73c101011   Jens Axboe   block: initial pa...
1578
  	struct request *req;
56ebdaf2f   Shaohua Li   block: simplify f...
1579
  	unsigned int request_count = 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1580

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1581
1582
1583
1584
1585
1586
  	/*
  	 * 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);
23688bf4f   Junichi Nomura   block: ensure to ...
1587
  	blk_queue_split(q, &bio, q->bio_split);
ffecfd1a7   Darrick J. Wong   block: optionally...
1588
  	if (bio_integrity_enabled(bio) && bio_integrity_prep(bio)) {
4246a0b63   Christoph Hellwig   block: add a bi_e...
1589
1590
  		bio->bi_error = -EIO;
  		bio_endio(bio);
dece16353   Jens Axboe   block: change ->m...
1591
  		return BLK_QC_T_NONE;
ffecfd1a7   Darrick J. Wong   block: optionally...
1592
  	}
1eff9d322   Jens Axboe   block: rename bio...
1593
  	if (bio->bi_opf & (REQ_PREFLUSH | REQ_FUA)) {
73c101011   Jens Axboe   block: initial pa...
1594
  		spin_lock_irq(q->queue_lock);
ae1b15396   Tejun Heo   block: reimplemen...
1595
  		where = ELEVATOR_INSERT_FLUSH;
28e7d1845   Tejun Heo   block: drop barri...
1596
1597
  		goto get_rq;
  	}
73c101011   Jens Axboe   block: initial pa...
1598
1599
1600
1601
  	/*
  	 * Check if we can merge with the plugged list before grabbing
  	 * any locks.
  	 */
0809e3ac6   Jeff Moyer   block: fix plug l...
1602
1603
  	if (!blk_queue_nomerges(q)) {
  		if (blk_attempt_plug_merge(q, bio, &request_count, NULL))
dece16353   Jens Axboe   block: change ->m...
1604
  			return BLK_QC_T_NONE;
0809e3ac6   Jeff Moyer   block: fix plug l...
1605
1606
  	} else
  		request_count = blk_plug_queued_count(q);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1607

73c101011   Jens Axboe   block: initial pa...
1608
  	spin_lock_irq(q->queue_lock);
2056a782f   Jens Axboe   [PATCH] Block que...
1609

73c101011   Jens Axboe   block: initial pa...
1610
1611
  	el_ret = elv_merge(q, &req, bio);
  	if (el_ret == ELEVATOR_BACK_MERGE) {
73c101011   Jens Axboe   block: initial pa...
1612
  		if (bio_attempt_back_merge(q, req, bio)) {
07c2bd373   Tejun Heo   block: don't call...
1613
  			elv_bio_merged(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1614
1615
1616
1617
1618
  			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...
1619
  		if (bio_attempt_front_merge(q, req, bio)) {
07c2bd373   Tejun Heo   block: don't call...
1620
  			elv_bio_merged(q, req, bio);
73c101011   Jens Axboe   block: initial pa...
1621
1622
1623
  			if (!attempt_front_merge(q, req))
  				elv_merged_request(q, req, el_ret);
  			goto out_unlock;
80a761fd3   Tejun Heo   block: implement ...
1624
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1625
  	}
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1626
  get_rq:
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1627
  	/*
7749a8d42   Jens Axboe   [PATCH] Propagate...
1628
1629
1630
1631
  	 * 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.
  	 */
7749a8d42   Jens Axboe   [PATCH] Propagate...
1632
  	if (sync)
7b6d91dae   Christoph Hellwig   block: unify flag...
1633
  		rw_flags |= REQ_SYNC;
7749a8d42   Jens Axboe   [PATCH] Propagate...
1634
1635
  
  	/*
b8269db45   Jens Axboe   cfq-iosched: temp...
1636
1637
  	 * Add in META/PRIO flags, if set, before we get to the IO scheduler
  	 */
1eff9d322   Jens Axboe   block: rename bio...
1638
  	rw_flags |= (bio->bi_opf & (REQ_META | REQ_PRIO));
b8269db45   Jens Axboe   cfq-iosched: temp...
1639
1640
  
  	/*
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1641
  	 * Grab a free request. This is might sleep but can not fail.
d6344532a   Nick Piggin   [PATCH] blk: redu...
1642
  	 * Returns with the queue unlocked.
450991bc1   Nick Piggin   [PATCH] blk: __ma...
1643
  	 */
e6a40b096   Mike Christie   block: prepare re...
1644
  	req = get_request(q, bio_data_dir(bio), rw_flags, bio, GFP_NOIO);
a492f0754   Joe Lawrence   block,scsi: fixup...
1645
  	if (IS_ERR(req)) {
4246a0b63   Christoph Hellwig   block: add a bi_e...
1646
1647
  		bio->bi_error = PTR_ERR(req);
  		bio_endio(bio);
da8303c63   Tejun Heo   block: make get_r...
1648
1649
  		goto out_unlock;
  	}
d6344532a   Nick Piggin   [PATCH] blk: redu...
1650

450991bc1   Nick Piggin   [PATCH] blk: __ma...
1651
1652
1653
1654
1655
  	/*
  	 * 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
1656
  	 */
52d9e6753   Tejun Heo   [BLOCK] ll_rw_blk...
1657
  	init_request_from_bio(req, bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1658

9562ad9ab   Tao Ma   block: Remove the...
1659
  	if (test_bit(QUEUE_FLAG_SAME_COMP, &q->queue_flags))
11ccf116d   Jens Axboe   block: fix warnin...
1660
  		req->cpu = raw_smp_processor_id();
73c101011   Jens Axboe   block: initial pa...
1661
1662
  
  	plug = current->plug;
721a9602e   Jens Axboe   block: kill off R...
1663
  	if (plug) {
dc6d36c97   Jens Axboe   block: readd plug...
1664
1665
  		/*
  		 * If this is the first request added after a plug, fire
7aef2e780   Jianpeng Ma   block: trace all ...
1666
  		 * of a plug trace.
dc6d36c97   Jens Axboe   block: readd plug...
1667
  		 */
7aef2e780   Jianpeng Ma   block: trace all ...
1668
  		if (!request_count)
dc6d36c97   Jens Axboe   block: readd plug...
1669
  			trace_block_plug(q);
3540d5e89   Shaohua Li   block: avoid unne...
1670
  		else {
019ceb7d5   Shaohua Li   block: add missed...
1671
  			if (request_count >= BLK_MAX_REQUEST_COUNT) {
3540d5e89   Shaohua Li   block: avoid unne...
1672
  				blk_flush_plug_list(plug, false);
019ceb7d5   Shaohua Li   block: add missed...
1673
1674
  				trace_block_plug(q);
  			}
73c101011   Jens Axboe   block: initial pa...
1675
  		}
73c101011   Jens Axboe   block: initial pa...
1676
  		list_add_tail(&req->queuelist, &plug->list);
320ae51fe   Jens Axboe   blk-mq: new multi...
1677
  		blk_account_io_start(req, true);
73c101011   Jens Axboe   block: initial pa...
1678
1679
1680
  	} else {
  		spin_lock_irq(q->queue_lock);
  		add_acct_request(q, req, where);
24ecfbe27   Christoph Hellwig   block: add blk_ru...
1681
  		__blk_run_queue(q);
73c101011   Jens Axboe   block: initial pa...
1682
1683
1684
  out_unlock:
  		spin_unlock_irq(q->queue_lock);
  	}
dece16353   Jens Axboe   block: change ->m...
1685
1686
  
  	return BLK_QC_T_NONE;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1687
1688
1689
1690
1691
1692
1693
1694
  }
  
  /*
   * 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...
1695
  	if (bio_sectors(bio) && bdev != bdev->bd_contains) {
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1696
  		struct hd_struct *p = bdev->bd_part;
4f024f379   Kent Overstreet   block: Abstract o...
1697
  		bio->bi_iter.bi_sector += p->start_sect;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1698
  		bio->bi_bdev = bdev->bd_contains;
c7149d6bc   Alan D. Brunelle   Fix remap handlin...
1699

d07335e51   Mike Snitzer   block: Rename "bl...
1700
1701
  		trace_block_bio_remap(bdev_get_queue(bio->bi_bdev), bio,
  				      bdev->bd_dev,
4f024f379   Kent Overstreet   block: Abstract o...
1702
  				      bio->bi_iter.bi_sector - p->start_sect);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1703
1704
  	}
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1705
1706
1707
1708
1709
1710
  static void handle_bad_sector(struct bio *bio)
  {
  	char b[BDEVNAME_SIZE];
  
  	printk(KERN_INFO "attempt to access beyond end of device
  ");
6296b9604   Mike Christie   block, drivers, f...
1711
1712
  	printk(KERN_INFO "%s: rw=%d, want=%Lu, limit=%Lu
  ",
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1713
  			bdevname(bio->bi_bdev, b),
1eff9d322   Jens Axboe   block: rename bio...
1714
  			bio->bi_opf,
f73a1c7d1   Kent Overstreet   block: Add bio_en...
1715
  			(unsigned long long)bio_end_sector(bio),
77304d2ab   Mike Snitzer   block: read i_siz...
1716
  			(long long)(i_size_read(bio->bi_bdev->bd_inode) >> 9));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1717
  }
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1718
1719
1720
1721
1722
1723
1724
1725
1726
  #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...
1727
  static bool should_fail_request(struct hd_struct *part, unsigned int bytes)
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1728
  {
b2c9cd379   Akinobu Mita   fail_make_request...
1729
  	return part->make_it_fail && should_fail(&fail_make_request, bytes);
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1730
1731
1732
1733
  }
  
  static int __init fail_make_request_debugfs(void)
  {
dd48c085c   Akinobu Mita   fault-injection: ...
1734
1735
  	struct dentry *dir = fault_create_debugfs_attr("fail_make_request",
  						NULL, &fail_make_request);
21f9fcd81   Duan Jiong   block: replace IS...
1736
  	return PTR_ERR_OR_ZERO(dir);
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1737
1738
1739
1740
1741
  }
  
  late_initcall(fail_make_request_debugfs);
  
  #else /* CONFIG_FAIL_MAKE_REQUEST */
b2c9cd379   Akinobu Mita   fail_make_request...
1742
1743
  static inline bool should_fail_request(struct hd_struct *part,
  					unsigned int bytes)
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1744
  {
b2c9cd379   Akinobu Mita   fail_make_request...
1745
  	return false;
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1746
1747
1748
  }
  
  #endif /* CONFIG_FAIL_MAKE_REQUEST */
c07e2b412   Jens Axboe   block: factor our...
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
  /*
   * 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...
1760
  	maxsector = i_size_read(bio->bi_bdev->bd_inode) >> 9;
c07e2b412   Jens Axboe   block: factor our...
1761
  	if (maxsector) {
4f024f379   Kent Overstreet   block: Abstract o...
1762
  		sector_t sector = bio->bi_iter.bi_sector;
c07e2b412   Jens Axboe   block: factor our...
1763
1764
1765
1766
1767
1768
1769
1770
1771
1772
1773
1774
1775
1776
  
  		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...
1777
1778
  static noinline_for_stack bool
  generic_make_request_checks(struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1779
  {
165125e1e   Jens Axboe   [BLOCK] Get rid o...
1780
  	struct request_queue *q;
5a7bbad27   Christoph Hellwig   block: remove sup...
1781
  	int nr_sectors = bio_sectors(bio);
51fd77bd9   Jens Axboe   [BLOCK] Don't all...
1782
  	int err = -EIO;
5a7bbad27   Christoph Hellwig   block: remove sup...
1783
1784
  	char b[BDEVNAME_SIZE];
  	struct hd_struct *part;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1785
1786
  
  	might_sleep();
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1787

c07e2b412   Jens Axboe   block: factor our...
1788
1789
  	if (bio_check_eod(bio, nr_sectors))
  		goto end_io;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1790

5a7bbad27   Christoph Hellwig   block: remove sup...
1791
1792
1793
1794
1795
1796
1797
  	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...
1798
  			(long long) bio->bi_iter.bi_sector);
5a7bbad27   Christoph Hellwig   block: remove sup...
1799
1800
  		goto end_io;
  	}
c17bb4951   Akinobu Mita   [PATCH] fault-inj...
1801

5a7bbad27   Christoph Hellwig   block: remove sup...
1802
  	part = bio->bi_bdev->bd_part;
4f024f379   Kent Overstreet   block: Abstract o...
1803
  	if (should_fail_request(part, bio->bi_iter.bi_size) ||
5a7bbad27   Christoph Hellwig   block: remove sup...
1804
  	    should_fail_request(&part_to_disk(part)->part0,
4f024f379   Kent Overstreet   block: Abstract o...
1805
  				bio->bi_iter.bi_size))
5a7bbad27   Christoph Hellwig   block: remove sup...
1806
  		goto end_io;
2056a782f   Jens Axboe   [PATCH] Block que...
1807

5a7bbad27   Christoph Hellwig   block: remove sup...
1808
1809
1810
1811
1812
  	/*
  	 * 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...
1813

5a7bbad27   Christoph Hellwig   block: remove sup...
1814
1815
  	if (bio_check_eod(bio, nr_sectors))
  		goto end_io;
1e87901e1   Tejun Heo   block: filter flu...
1816

5a7bbad27   Christoph Hellwig   block: remove sup...
1817
1818
1819
1820
1821
  	/*
  	 * Filter flush bio's early so that make_request based
  	 * drivers without flush support don't have to worry
  	 * about them.
  	 */
1eff9d322   Jens Axboe   block: rename bio...
1822
  	if ((bio->bi_opf & (REQ_PREFLUSH | REQ_FUA)) &&
c888a8f95   Jens Axboe   block: kill off q...
1823
  	    !test_bit(QUEUE_FLAG_WC, &q->queue_flags)) {
1eff9d322   Jens Axboe   block: rename bio...
1824
  		bio->bi_opf &= ~(REQ_PREFLUSH | REQ_FUA);
5a7bbad27   Christoph Hellwig   block: remove sup...
1825
1826
  		if (!nr_sectors) {
  			err = 0;
51fd77bd9   Jens Axboe   [BLOCK] Don't all...
1827
1828
  			goto end_io;
  		}
5a7bbad27   Christoph Hellwig   block: remove sup...
1829
  	}
5ddfe9691   NeilBrown   [PATCH] md: check...
1830

288dab8a3   Christoph Hellwig   block: add a sepa...
1831
1832
1833
1834
1835
1836
1837
1838
1839
1840
1841
1842
1843
1844
1845
  	switch (bio_op(bio)) {
  	case REQ_OP_DISCARD:
  		if (!blk_queue_discard(q))
  			goto not_supported;
  		break;
  	case REQ_OP_SECURE_ERASE:
  		if (!blk_queue_secure_erase(q))
  			goto not_supported;
  		break;
  	case REQ_OP_WRITE_SAME:
  		if (!bdev_write_same(bio->bi_bdev))
  			goto not_supported;
  		break;
  	default:
  		break;
5a7bbad27   Christoph Hellwig   block: remove sup...
1846
  	}
01edede41   Minchan Kim   block: trace bio ...
1847

7f4b35d15   Tejun Heo   block: allocate i...
1848
1849
1850
1851
1852
1853
1854
  	/*
  	 * Various block parts want %current->io_context and lazy ioc
  	 * allocation ends up trading a lot of pain for a small amount of
  	 * memory.  Just allocate it upfront.  This may fail and block
  	 * layer knows how to live with it.
  	 */
  	create_io_context(GFP_ATOMIC, q->node);
ae1188963   Tejun Heo   blkcg: consolidat...
1855
1856
  	if (!blkcg_bio_issue_check(q, bio))
  		return false;
27a84d54c   Christoph Hellwig   block: refactor g...
1857

5a7bbad27   Christoph Hellwig   block: remove sup...
1858
  	trace_block_bio_queue(q, bio);
27a84d54c   Christoph Hellwig   block: refactor g...
1859
  	return true;
a7384677b   Tejun Heo   block: remove dup...
1860

288dab8a3   Christoph Hellwig   block: add a sepa...
1861
1862
  not_supported:
  	err = -EOPNOTSUPP;
a7384677b   Tejun Heo   block: remove dup...
1863
  end_io:
4246a0b63   Christoph Hellwig   block: add a bi_e...
1864
1865
  	bio->bi_error = err;
  	bio_endio(bio);
27a84d54c   Christoph Hellwig   block: refactor g...
1866
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1867
  }
27a84d54c   Christoph Hellwig   block: refactor g...
1868
1869
1870
1871
1872
1873
1874
1875
1876
1877
1878
1879
1880
1881
1882
1883
1884
1885
1886
1887
1888
1889
1890
  /**
   * 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...
1891
   */
dece16353   Jens Axboe   block: change ->m...
1892
  blk_qc_t generic_make_request(struct bio *bio)
d89d87965   Neil Brown   When stacked bloc...
1893
  {
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1894
  	struct bio_list bio_list_on_stack;
dece16353   Jens Axboe   block: change ->m...
1895
  	blk_qc_t ret = BLK_QC_T_NONE;
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1896

27a84d54c   Christoph Hellwig   block: refactor g...
1897
  	if (!generic_make_request_checks(bio))
dece16353   Jens Axboe   block: change ->m...
1898
  		goto out;
27a84d54c   Christoph Hellwig   block: refactor g...
1899
1900
1901
1902
1903
1904
1905
1906
1907
1908
1909
  
  	/*
  	 * 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...
1910
  	if (current->bio_list) {
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1911
  		bio_list_add(current->bio_list, bio);
dece16353   Jens Axboe   block: change ->m...
1912
  		goto out;
d89d87965   Neil Brown   When stacked bloc...
1913
  	}
27a84d54c   Christoph Hellwig   block: refactor g...
1914

d89d87965   Neil Brown   When stacked bloc...
1915
1916
1917
1918
1919
  	/* 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...
1920
1921
  	 * 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...
1922
  	 * added.  ->make_request() may indeed add some more bios
d89d87965   Neil Brown   When stacked bloc...
1923
1924
1925
  	 * 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...
1926
  	 * of the top of the list (no pretending) and so remove it from
27a84d54c   Christoph Hellwig   block: refactor g...
1927
  	 * bio_list, and call into ->make_request() again.
d89d87965   Neil Brown   When stacked bloc...
1928
1929
  	 */
  	BUG_ON(bio->bi_next);
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1930
1931
  	bio_list_init(&bio_list_on_stack);
  	current->bio_list = &bio_list_on_stack;
d89d87965   Neil Brown   When stacked bloc...
1932
  	do {
27a84d54c   Christoph Hellwig   block: refactor g...
1933
  		struct request_queue *q = bdev_get_queue(bio->bi_bdev);
6f3b0e8bc   Christoph Hellwig   blk-mq: add a fla...
1934
  		if (likely(blk_queue_enter(q, false) == 0)) {
dece16353   Jens Axboe   block: change ->m...
1935
  			ret = q->make_request_fn(q, bio);
3ef28e83a   Dan Williams   block: generic re...
1936
1937
  
  			blk_queue_exit(q);
27a84d54c   Christoph Hellwig   block: refactor g...
1938

3ef28e83a   Dan Williams   block: generic re...
1939
1940
1941
1942
1943
1944
1945
  			bio = bio_list_pop(current->bio_list);
  		} else {
  			struct bio *bio_next = bio_list_pop(current->bio_list);
  
  			bio_io_error(bio);
  			bio = bio_next;
  		}
d89d87965   Neil Brown   When stacked bloc...
1946
  	} while (bio);
bddd87c7e   Akinobu Mita   blk-core: use BIO...
1947
  	current->bio_list = NULL; /* deactivate */
dece16353   Jens Axboe   block: change ->m...
1948
1949
1950
  
  out:
  	return ret;
d89d87965   Neil Brown   When stacked bloc...
1951
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1952
1953
1954
  EXPORT_SYMBOL(generic_make_request);
  
  /**
710027a48   Randy Dunlap   Add some block/ s...
1955
   * submit_bio - submit a bio to the block device layer for I/O
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1956
1957
1958
1959
   * @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...
1960
   * interfaces; @bio must be presetup and ready for I/O.
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1961
1962
   *
   */
4e49ea4a3   Mike Christie   block/fs/drivers:...
1963
  blk_qc_t submit_bio(struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1964
  {
bf2de6f5a   Jens Axboe   block: Initial su...
1965
1966
1967
1968
  	/*
  	 * 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...
1969
  	if (bio_has_data(bio)) {
4363ac7c1   Martin K. Petersen   block: Implement ...
1970
  		unsigned int count;
95fe6c1a2   Mike Christie   block, fs, mm, dr...
1971
  		if (unlikely(bio_op(bio) == REQ_OP_WRITE_SAME))
4363ac7c1   Martin K. Petersen   block: Implement ...
1972
1973
1974
  			count = bdev_logical_block_size(bio->bi_bdev) >> 9;
  		else
  			count = bio_sectors(bio);
a8ebb056a   Mike Christie   block, drivers, c...
1975
  		if (op_is_write(bio_op(bio))) {
bf2de6f5a   Jens Axboe   block: Initial su...
1976
1977
  			count_vm_events(PGPGOUT, count);
  		} else {
4f024f379   Kent Overstreet   block: Abstract o...
1978
  			task_io_account_read(bio->bi_iter.bi_size);
bf2de6f5a   Jens Axboe   block: Initial su...
1979
1980
1981
1982
1983
  			count_vm_events(PGPGIN, count);
  		}
  
  		if (unlikely(block_dump)) {
  			char b[BDEVNAME_SIZE];
8dcbdc742   San Mehat   block: block_dump...
1984
1985
  			printk(KERN_DEBUG "%s(%d): %s block %Lu on %s (%u sectors)
  ",
ba25f9dcc   Pavel Emelyanov   Use helpers to ob...
1986
  			current->comm, task_pid_nr(current),
a8ebb056a   Mike Christie   block, drivers, c...
1987
  				op_is_write(bio_op(bio)) ? "WRITE" : "READ",
4f024f379   Kent Overstreet   block: Abstract o...
1988
  				(unsigned long long)bio->bi_iter.bi_sector,
8dcbdc742   San Mehat   block: block_dump...
1989
1990
  				bdevname(bio->bi_bdev, b),
  				count);
bf2de6f5a   Jens Axboe   block: Initial su...
1991
  		}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1992
  	}
dece16353   Jens Axboe   block: change ->m...
1993
  	return generic_make_request(bio);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1994
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
1995
  EXPORT_SYMBOL(submit_bio);
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
1996
  /**
bf4e6b4e7   Hannes Reinecke   block: Always che...
1997
1998
   * blk_cloned_rq_check_limits - Helper function to check a cloned request
   *                              for new the queue limits
82124d603   Kiyoshi Ueda   block: add reques...
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
   * @q:  the queue
   * @rq: the request being checked
   *
   * Description:
   *    @rq may have been made based on weaker limitations of upper-level queues
   *    in request stacking drivers, and it may violate the limitation of @q.
   *    Since the block layer and the underlying device driver trust @rq
   *    after it is inserted to @q, it should be checked against @q before
   *    the insertion using this generic function.
   *
82124d603   Kiyoshi Ueda   block: add reques...
2009
   *    Request stacking drivers like request-based dm may change the queue
bf4e6b4e7   Hannes Reinecke   block: Always che...
2010
2011
   *    limits when retrying requests on other queues. Those requests need
   *    to be checked against the new queue limits again during dispatch.
82124d603   Kiyoshi Ueda   block: add reques...
2012
   */
bf4e6b4e7   Hannes Reinecke   block: Always che...
2013
2014
  static int blk_cloned_rq_check_limits(struct request_queue *q,
  				      struct request *rq)
82124d603   Kiyoshi Ueda   block: add reques...
2015
  {
8fe0d473f   Mike Christie   block: convert me...
2016
  	if (blk_rq_sectors(rq) > blk_queue_get_max_sectors(q, req_op(rq))) {
82124d603   Kiyoshi Ueda   block: add reques...
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
  		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...
2029
  	if (rq->nr_phys_segments > queue_max_segments(q)) {
82124d603   Kiyoshi Ueda   block: add reques...
2030
2031
2032
2033
2034
2035
2036
  		printk(KERN_ERR "%s: over max segments limit.
  ", __func__);
  		return -EIO;
  	}
  
  	return 0;
  }
82124d603   Kiyoshi Ueda   block: add reques...
2037
2038
2039
2040
2041
2042
2043
2044
2045
  
  /**
   * 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 ...
2046
  	int where = ELEVATOR_INSERT_BACK;
82124d603   Kiyoshi Ueda   block: add reques...
2047

bf4e6b4e7   Hannes Reinecke   block: Always che...
2048
  	if (blk_cloned_rq_check_limits(q, rq))
82124d603   Kiyoshi Ueda   block: add reques...
2049
  		return -EIO;
b2c9cd379   Akinobu Mita   fail_make_request...
2050
2051
  	if (rq->rq_disk &&
  	    should_fail_request(&rq->rq_disk->part0, blk_rq_bytes(rq)))
82124d603   Kiyoshi Ueda   block: add reques...
2052
  		return -EIO;
82124d603   Kiyoshi Ueda   block: add reques...
2053

7fb4898e0   Keith Busch   block: add blk-mq...
2054
2055
2056
  	if (q->mq_ops) {
  		if (blk_queue_io_stat(q))
  			blk_account_io_start(rq, true);
6acfe68ba   Mike Snitzer   dm: fix excessive...
2057
  		blk_mq_insert_request(rq, false, true, false);
7fb4898e0   Keith Busch   block: add blk-mq...
2058
2059
  		return 0;
  	}
82124d603   Kiyoshi Ueda   block: add reques...
2060
  	spin_lock_irqsave(q->queue_lock, flags);
3f3299d5c   Bart Van Assche   block: Rename que...
2061
  	if (unlikely(blk_queue_dying(q))) {
8ba61435d   Tejun Heo   block: add missin...
2062
2063
2064
  		spin_unlock_irqrestore(q->queue_lock, flags);
  		return -ENODEV;
  	}
82124d603   Kiyoshi Ueda   block: add reques...
2065
2066
2067
2068
2069
2070
  
  	/*
  	 * Submitting request must be dequeued before calling this function
  	 * because it will be linked to another request_queue
  	 */
  	BUG_ON(blk_queued_rq(rq));
28a8f0d31   Mike Christie   block, drivers, f...
2071
  	if (rq->cmd_flags & (REQ_PREFLUSH | REQ_FUA))
4853abaae   Jeff Moyer   block: fix flush ...
2072
2073
2074
  		where = ELEVATOR_INSERT_FLUSH;
  
  	add_acct_request(q, rq, where);
e67b77c79   Jeff Moyer   blk-flush: move t...
2075
2076
  	if (where == ELEVATOR_INSERT_FLUSH)
  		__blk_run_queue(q);
82124d603   Kiyoshi Ueda   block: add reques...
2077
2078
2079
2080
2081
  	spin_unlock_irqrestore(q->queue_lock, flags);
  
  	return 0;
  }
  EXPORT_SYMBOL_GPL(blk_insert_cloned_request);
80a761fd3   Tejun Heo   block: implement ...
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
  /**
   * 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) {
1eff9d322   Jens Axboe   block: rename bio...
2115
  		if ((bio->bi_opf & ff) != ff)
80a761fd3   Tejun Heo   block: implement ...
2116
  			break;
4f024f379   Kent Overstreet   block: Abstract o...
2117
  		bytes += bio->bi_iter.bi_size;
80a761fd3   Tejun Heo   block: implement ...
2118
2119
2120
2121
2122
2123
2124
  	}
  
  	/* 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...
2125
  void blk_account_io_completion(struct request *req, unsigned int bytes)
bc58ba946   Jens Axboe   block: add sysfs ...
2126
  {
c2553b584   Jens Axboe   block: make blk_d...
2127
  	if (blk_do_io_stat(req)) {
bc58ba946   Jens Axboe   block: add sysfs ...
2128
2129
2130
2131
2132
  		const int rw = rq_data_dir(req);
  		struct hd_struct *part;
  		int cpu;
  
  		cpu = part_stat_lock();
09e099d4b   Jerome Marchand   block: fix accoun...
2133
  		part = req->part;
bc58ba946   Jens Axboe   block: add sysfs ...
2134
2135
2136
2137
  		part_stat_add(cpu, part, sectors[rw], bytes >> 9);
  		part_stat_unlock();
  	}
  }
320ae51fe   Jens Axboe   blk-mq: new multi...
2138
  void blk_account_io_done(struct request *req)
bc58ba946   Jens Axboe   block: add sysfs ...
2139
  {
bc58ba946   Jens Axboe   block: add sysfs ...
2140
  	/*
dd4c133f3   Tejun Heo   block: rename bar...
2141
2142
2143
  	 * 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 ...
2144
  	 */
414b4ff5e   Tejun Heo   block: add REQ_FL...
2145
  	if (blk_do_io_stat(req) && !(req->cmd_flags & REQ_FLUSH_SEQ)) {
bc58ba946   Jens Axboe   block: add sysfs ...
2146
2147
2148
2149
2150
2151
  		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...
2152
  		part = req->part;
bc58ba946   Jens Axboe   block: add sysfs ...
2153
2154
2155
2156
  
  		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...
2157
  		part_dec_in_flight(part, rw);
bc58ba946   Jens Axboe   block: add sysfs ...
2158

6c23a9681   Jens Axboe   block: add intern...
2159
  		hd_struct_put(part);
bc58ba946   Jens Axboe   block: add sysfs ...
2160
2161
2162
  		part_stat_unlock();
  	}
  }
47fafbc70   Rafael J. Wysocki   block / PM: Repla...
2163
  #ifdef CONFIG_PM
c8158819d   Lin Ming   block: implement ...
2164
2165
2166
2167
2168
2169
2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
  /*
   * 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...
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
2198
2199
2200
2201
2202
2203
2204
2205
2206
2207
2208
2209
2210
2211
2212
2213
2214
2215
2216
2217
2218
  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...
2219
  /**
9934c8c04   Tejun Heo   block: implement ...
2220
2221
2222
2223
2224
2225
2226
2227
2228
2229
2230
2231
2232
2233
2234
2235
   * 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...
2236
2237
2238
2239
2240
  {
  	struct request *rq;
  	int ret;
  
  	while ((rq = __elv_next_request(q)) != NULL) {
c8158819d   Lin Ming   block: implement ...
2241
2242
2243
2244
  
  		rq = blk_pm_peek_request(q, rq);
  		if (!rq)
  			break;
158dbda00   Tejun Heo   block: reorganize...
2245
2246
2247
2248
2249
2250
  		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...
2251
  			if (rq->cmd_flags & REQ_SORTED)
158dbda00   Tejun Heo   block: reorganize...
2252
2253
2254
2255
2256
2257
2258
2259
2260
2261
2262
2263
2264
2265
2266
2267
2268
2269
  				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...
2270
  		if (q->dma_drain_size && blk_rq_bytes(rq)) {
158dbda00   Tejun Heo   block: reorganize...
2271
2272
2273
2274
2275
2276
2277
2278
2279
2280
2281
2282
2283
2284
2285
2286
2287
2288
2289
2290
2291
2292
  			/*
  			 * 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...
2293
  			if (q->dma_drain_size && blk_rq_bytes(rq) &&
158dbda00   Tejun Heo   block: reorganize...
2294
2295
2296
2297
2298
2299
2300
2301
2302
2303
  			    !(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;
0fb5b1fb3   Martin K. Petersen   block/sd: Return ...
2304
2305
  		} else if (ret == BLKPREP_KILL || ret == BLKPREP_INVALID) {
  			int err = (ret == BLKPREP_INVALID) ? -EREMOTEIO : -EIO;
158dbda00   Tejun Heo   block: reorganize...
2306
  			rq->cmd_flags |= REQ_QUIET;
c143dc903   James Bottomley   block: fix an oop...
2307
2308
2309
2310
2311
  			/*
  			 * Mark this request as started so we don't trigger
  			 * any debug logic in the end I/O path.
  			 */
  			blk_start_request(rq);
0fb5b1fb3   Martin K. Petersen   block/sd: Return ...
2312
  			__blk_end_request_all(rq, err);
158dbda00   Tejun Heo   block: reorganize...
2313
2314
2315
2316
2317
2318
2319
2320
2321
  		} else {
  			printk(KERN_ERR "%s: bad return=%d
  ", __func__, ret);
  			break;
  		}
  	}
  
  	return rq;
  }
9934c8c04   Tejun Heo   block: implement ...
2322
  EXPORT_SYMBOL(blk_peek_request);
158dbda00   Tejun Heo   block: reorganize...
2323

9934c8c04   Tejun Heo   block: implement ...
2324
  void blk_dequeue_request(struct request *rq)
158dbda00   Tejun Heo   block: reorganize...
2325
  {
9934c8c04   Tejun Heo   block: implement ...
2326
  	struct request_queue *q = rq->q;
158dbda00   Tejun Heo   block: reorganize...
2327
2328
2329
2330
2331
2332
2333
2334
2335
2336
  	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 ...
2337
  	if (blk_account_rq(rq)) {
0a7ae2ff0   Jens Axboe   block: change the...
2338
  		q->in_flight[rq_is_sync(rq)]++;
9195291e5   Divyesh Shah   blkio: Increment ...
2339
2340
  		set_io_start_time_ns(rq);
  	}
158dbda00   Tejun Heo   block: reorganize...
2341
  }
5efccd17c   Tejun Heo   block: reorder re...
2342
  /**
9934c8c04   Tejun Heo   block: implement ...
2343
2344
2345
2346
2347
2348
2349
2350
2351
2352
2353
2354
2355
2356
2357
2358
2359
2360
   * 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...
2361
2362
  	 * 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 ...
2363
  	 */
5f49f6317   Tejun Heo   block: set rq->re...
2364
  	req->resid_len = blk_rq_bytes(req);
dbb66c4be   FUJITA Tomonori   block: needs to s...
2365
2366
  	if (unlikely(blk_bidi_rq(req)))
  		req->next_rq->resid_len = blk_rq_bytes(req->next_rq);
4912aa6c1   Jeff Moyer   block: fix race b...
2367
  	BUG_ON(test_bit(REQ_ATOM_COMPLETE, &req->atomic_flags));
9934c8c04   Tejun Heo   block: implement ...
2368
2369
2370
2371
2372
2373
2374
2375
2376
2377
2378
2379
2380
2381
2382
2383
2384
2385
2386
2387
2388
2389
2390
2391
2392
2393
2394
2395
2396
2397
2398
  	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...
2399
   * blk_update_request - Special helper function for request stacking drivers
8ebf97560   Randy Dunlap   block: fix kernel...
2400
   * @req:      the request being processed
710027a48   Randy Dunlap   Add some block/ s...
2401
   * @error:    %0 for success, < %0 for error
8ebf97560   Randy Dunlap   block: fix kernel...
2402
   * @nr_bytes: number of bytes to complete @req
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
2403
2404
   *
   * Description:
8ebf97560   Randy Dunlap   block: fix kernel...
2405
2406
2407
   *     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...
2408
2409
2410
2411
2412
2413
2414
   *
   *     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: ...
2415
2416
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2417
2418
   *     %false - this request doesn't have any more data
   *     %true  - this request has more data
3bcddeac1   Kiyoshi Ueda   blk_end_request: ...
2419
   **/
2e60e0229   Tejun Heo   block: clean up r...
2420
  bool blk_update_request(struct request *req, int error, unsigned int nr_bytes)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2421
  {
f79ea4161   Kent Overstreet   block: Refactor b...
2422
  	int total_bytes;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2423

4a0efdc93   Hannes Reinecke   block: misplaced ...
2424
  	trace_block_rq_complete(req->q, req, nr_bytes);
2e60e0229   Tejun Heo   block: clean up r...
2425
2426
  	if (!req->bio)
  		return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2427
  	/*
6f41469c6   Tejun Heo   block: clear req-...
2428
2429
2430
2431
2432
2433
  	 * 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
2434
  	 */
33659ebba   Christoph Hellwig   block: remove wra...
2435
  	if (req->cmd_type == REQ_TYPE_FS)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2436
  		req->errors = 0;
33659ebba   Christoph Hellwig   block: remove wra...
2437
2438
  	if (error && req->cmd_type == REQ_TYPE_FS &&
  	    !(req->cmd_flags & REQ_QUIET)) {
79775567e   Hannes Reinecke   [SCSI] block: imp...
2439
2440
2441
2442
2443
2444
2445
2446
2447
2448
2449
2450
  		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...
2451
2452
2453
  		case -ETIMEDOUT:
  			error_type = "timeout";
  			break;
a9d6ceb83   Hannes Reinecke   [SCSI] return ENO...
2454
2455
2456
  		case -ENOSPC:
  			error_type = "critical space allocation";
  			break;
7e782af57   Hannes Reinecke   [SCSI] Return ENO...
2457
2458
2459
  		case -ENODATA:
  			error_type = "critical medium";
  			break;
79775567e   Hannes Reinecke   [SCSI] block: imp...
2460
2461
2462
2463
2464
  		case -EIO:
  		default:
  			error_type = "I/O";
  			break;
  		}
ef3ecb66b   Robert Elliott   block: make blk_u...
2465
2466
2467
  		printk_ratelimited(KERN_ERR "%s: %s error, dev %s, sector %llu
  ",
  				   __func__, error_type, req->rq_disk ?
37d7b34f0   Yi Zou   block: rate-limit...
2468
2469
  				   req->rq_disk->disk_name : "?",
  				   (unsigned long long)blk_rq_pos(req));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2470
  	}
bc58ba946   Jens Axboe   block: add sysfs ...
2471
  	blk_account_io_completion(req, nr_bytes);
d72d904a5   Jens Axboe   [BLOCK] Update re...
2472

f79ea4161   Kent Overstreet   block: Refactor b...
2473
2474
2475
  	total_bytes = 0;
  	while (req->bio) {
  		struct bio *bio = req->bio;
4f024f379   Kent Overstreet   block: Abstract o...
2476
  		unsigned bio_bytes = min(bio->bi_iter.bi_size, nr_bytes);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2477

4f024f379   Kent Overstreet   block: Abstract o...
2478
  		if (bio_bytes == bio->bi_iter.bi_size)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2479
  			req->bio = bio->bi_next;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2480

f79ea4161   Kent Overstreet   block: Refactor b...
2481
  		req_bio_endio(req, bio, bio_bytes, error);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2482

f79ea4161   Kent Overstreet   block: Refactor b...
2483
2484
  		total_bytes += bio_bytes;
  		nr_bytes -= bio_bytes;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2485

f79ea4161   Kent Overstreet   block: Refactor b...
2486
2487
  		if (!nr_bytes)
  			break;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2488
2489
2490
2491
2492
  	}
  
  	/*
  	 * completely done
  	 */
2e60e0229   Tejun Heo   block: clean up r...
2493
2494
2495
2496
2497
2498
  	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...
2499
  		req->__data_len = 0;
2e60e0229   Tejun Heo   block: clean up r...
2500
2501
  		return false;
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2502

a2dec7b36   Tejun Heo   block: hide reque...
2503
  	req->__data_len -= total_bytes;
2e46e8b27   Tejun Heo   block: drop reque...
2504
2505
  
  	/* update sector only for requests with clear definition of sector */
e2a60da74   Martin K. Petersen   block: Clean up s...
2506
  	if (req->cmd_type == REQ_TYPE_FS)
a2dec7b36   Tejun Heo   block: hide reque...
2507
  		req->__sector += total_bytes >> 9;
2e46e8b27   Tejun Heo   block: drop reque...
2508

80a761fd3   Tejun Heo   block: implement ...
2509
2510
2511
  	/* mixed attributes always follow the first bio */
  	if (req->cmd_flags & REQ_MIXED_MERGE) {
  		req->cmd_flags &= ~REQ_FAILFAST_MASK;
1eff9d322   Jens Axboe   block: rename bio...
2512
  		req->cmd_flags |= req->bio->bi_opf & REQ_FAILFAST_MASK;
80a761fd3   Tejun Heo   block: implement ...
2513
  	}
2e46e8b27   Tejun Heo   block: drop reque...
2514
2515
2516
2517
2518
  	/*
  	 * 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...
2519
  		blk_dump_rq_flags(req, "request botched");
a2dec7b36   Tejun Heo   block: hide reque...
2520
  		req->__data_len = blk_rq_cur_bytes(req);
2e46e8b27   Tejun Heo   block: drop reque...
2521
2522
2523
  	}
  
  	/* recalculate the number of segments */
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2524
  	blk_recalc_rq_segments(req);
2e46e8b27   Tejun Heo   block: drop reque...
2525

2e60e0229   Tejun Heo   block: clean up r...
2526
  	return true;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2527
  }
2e60e0229   Tejun Heo   block: clean up r...
2528
  EXPORT_SYMBOL_GPL(blk_update_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2529

2e60e0229   Tejun Heo   block: clean up r...
2530
2531
2532
  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...
2533
  {
2e60e0229   Tejun Heo   block: clean up r...
2534
2535
  	if (blk_update_request(rq, error, nr_bytes))
  		return true;
5efccd17c   Tejun Heo   block: reorder re...
2536

2e60e0229   Tejun Heo   block: clean up r...
2537
2538
2539
2540
  	/* 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...
2541

e2e1a148b   Jens Axboe   block: add sysfs ...
2542
2543
  	if (blk_queue_add_random(rq->q))
  		add_disk_randomness(rq->rq_disk);
2e60e0229   Tejun Heo   block: clean up r...
2544
2545
  
  	return false;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2546
  }
28018c242   James Bottomley   block: implement ...
2547
2548
2549
2550
2551
2552
2553
2554
2555
2556
2557
2558
2559
2560
2561
2562
2563
2564
2565
  /**
   * 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
2566
2567
2568
  /*
   * queue lock must be held
   */
12120077b   Christoph Hellwig   block: export blk...
2569
  void blk_finish_request(struct request *req, int error)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2570
  {
125c99bc8   Christoph Hellwig   scsi: add new scs...
2571
  	if (req->cmd_flags & REQ_QUEUED)
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2572
  		blk_queue_end_tag(req->q, req);
ba396a6c1   James Bottomley   block: fix oops w...
2573
  	BUG_ON(blk_queued_rq(req));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2574

33659ebba   Christoph Hellwig   block: remove wra...
2575
  	if (unlikely(laptop_mode) && req->cmd_type == REQ_TYPE_FS)
31373d09d   Matthew Garrett   laptop-mode: Make...
2576
  		laptop_io_completion(&req->q->backing_dev_info);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2577

e78042e5b   Mike Anderson   blk: move blk_del...
2578
  	blk_delete_timer(req);
28018c242   James Bottomley   block: implement ...
2579
2580
  	if (req->cmd_flags & REQ_DONTPREP)
  		blk_unprep_request(req);
bc58ba946   Jens Axboe   block: add sysfs ...
2581
  	blk_account_io_done(req);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2582

1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2583
  	if (req->end_io)
8ffdc6550   Tejun Heo   [BLOCK] add @upto...
2584
  		req->end_io(req, error);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2585
2586
2587
  	else {
  		if (blk_bidi_rq(req))
  			__blk_put_request(req->next_rq->q, req->next_rq);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2588
  		__blk_put_request(req->q, req);
b8286239d   Kiyoshi Ueda   blk_end_request: ...
2589
  	}
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2590
  }
12120077b   Christoph Hellwig   block: export blk...
2591
  EXPORT_SYMBOL(blk_finish_request);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2592

3b11313a6   Kiyoshi Ueda   blk_end_request: ...
2593
  /**
2e60e0229   Tejun Heo   block: clean up r...
2594
2595
2596
2597
2598
   * 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...
2599
2600
   *
   * Description:
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2601
   *     Ends I/O on a number of bytes attached to @rq and @rq->next_rq.
2e60e0229   Tejun Heo   block: clean up r...
2602
2603
2604
   *     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: ...
2605
2606
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2607
2608
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
a0cd12854   Jens Axboe   block: add end_qu...
2609
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2610
  static bool blk_end_bidi_request(struct request *rq, int error,
32fab448e   Kiyoshi Ueda   block: add reques...
2611
2612
  				 unsigned int nr_bytes, unsigned int bidi_bytes)
  {
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2613
  	struct request_queue *q = rq->q;
2e60e0229   Tejun Heo   block: clean up r...
2614
  	unsigned long flags;
32fab448e   Kiyoshi Ueda   block: add reques...
2615

2e60e0229   Tejun Heo   block: clean up r...
2616
2617
  	if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes))
  		return true;
32fab448e   Kiyoshi Ueda   block: add reques...
2618

336cdb400   Kiyoshi Ueda   blk_end_request: ...
2619
  	spin_lock_irqsave(q->queue_lock, flags);
2e60e0229   Tejun Heo   block: clean up r...
2620
  	blk_finish_request(rq, error);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2621
  	spin_unlock_irqrestore(q->queue_lock, flags);
2e60e0229   Tejun Heo   block: clean up r...
2622
  	return false;
32fab448e   Kiyoshi Ueda   block: add reques...
2623
  }
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2624
  /**
2e60e0229   Tejun Heo   block: clean up r...
2625
2626
   * __blk_end_bidi_request - Complete a bidi request with queue lock held
   * @rq:         the request to complete
710027a48   Randy Dunlap   Add some block/ s...
2627
   * @error:      %0 for success, < %0 for error
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2628
2629
   * @nr_bytes:   number of bytes to complete @rq
   * @bidi_bytes: number of bytes to complete @rq->next_rq
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2630
2631
   *
   * Description:
2e60e0229   Tejun Heo   block: clean up r...
2632
2633
   *     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: ...
2634
2635
   *
   * Return:
2e60e0229   Tejun Heo   block: clean up r...
2636
2637
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2638
   **/
4853abaae   Jeff Moyer   block: fix flush ...
2639
  bool __blk_end_bidi_request(struct request *rq, int error,
b1f744937   FUJITA Tomonori   block: move compl...
2640
  				   unsigned int nr_bytes, unsigned int bidi_bytes)
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2641
  {
2e60e0229   Tejun Heo   block: clean up r...
2642
2643
  	if (blk_update_bidi_request(rq, error, nr_bytes, bidi_bytes))
  		return true;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2644

2e60e0229   Tejun Heo   block: clean up r...
2645
  	blk_finish_request(rq, error);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2646

2e60e0229   Tejun Heo   block: clean up r...
2647
  	return false;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2648
  }
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2649
2650
2651
2652
  
  /**
   * blk_end_request - Helper function for drivers to complete the request.
   * @rq:       the request being processed
710027a48   Randy Dunlap   Add some block/ s...
2653
   * @error:    %0 for success, < %0 for error
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2654
2655
2656
2657
2658
2659
2660
   * @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...
2661
2662
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2663
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2664
  bool blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2665
  {
b1f744937   FUJITA Tomonori   block: move compl...
2666
  	return blk_end_bidi_request(rq, error, nr_bytes, 0);
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2667
  }
56ad1740d   Jens Axboe   block: make the e...
2668
  EXPORT_SYMBOL(blk_end_request);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2669
2670
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2671
2672
   * blk_end_request_all - Helper function for drives to finish the request.
   * @rq: the request to finish
8ebf97560   Randy Dunlap   block: fix kernel...
2673
   * @error: %0 for success, < %0 for error
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2674
2675
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2676
2677
2678
   *     Completely finish @rq.
   */
  void blk_end_request_all(struct request *rq, int error)
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2679
  {
b1f744937   FUJITA Tomonori   block: move compl...
2680
2681
  	bool pending;
  	unsigned int bidi_bytes = 0;
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2682

b1f744937   FUJITA Tomonori   block: move compl...
2683
2684
  	if (unlikely(blk_bidi_rq(rq)))
  		bidi_bytes = blk_rq_bytes(rq->next_rq);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2685

b1f744937   FUJITA Tomonori   block: move compl...
2686
2687
2688
  	pending = blk_end_bidi_request(rq, error, blk_rq_bytes(rq), bidi_bytes);
  	BUG_ON(pending);
  }
56ad1740d   Jens Axboe   block: make the e...
2689
  EXPORT_SYMBOL(blk_end_request_all);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2690

b1f744937   FUJITA Tomonori   block: move compl...
2691
2692
2693
  /**
   * 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...
2694
   * @error: %0 for success, < %0 for error
b1f744937   FUJITA Tomonori   block: move compl...
2695
2696
2697
2698
2699
2700
2701
2702
2703
2704
2705
   *
   * 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: ...
2706
  }
56ad1740d   Jens Axboe   block: make the e...
2707
  EXPORT_SYMBOL(blk_end_request_cur);
336cdb400   Kiyoshi Ueda   blk_end_request: ...
2708

e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2709
  /**
80a761fd3   Tejun Heo   block: implement ...
2710
2711
2712
2713
2714
2715
2716
2717
2718
2719
2720
2721
2722
2723
2724
2725
2726
2727
2728
   * 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...
2729
2730
2731
2732
   * __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: ...
2733
2734
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2735
   *     Must be called with queue lock held unlike blk_end_request().
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2736
2737
   *
   * Return:
b1f744937   FUJITA Tomonori   block: move compl...
2738
2739
   *     %false - we are done with this request
   *     %true  - still buffers pending for this request
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2740
   **/
b1f744937   FUJITA Tomonori   block: move compl...
2741
  bool __blk_end_request(struct request *rq, int error, unsigned int nr_bytes)
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2742
  {
b1f744937   FUJITA Tomonori   block: move compl...
2743
  	return __blk_end_bidi_request(rq, error, nr_bytes, 0);
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2744
  }
56ad1740d   Jens Axboe   block: make the e...
2745
  EXPORT_SYMBOL(__blk_end_request);
e3a04fe34   Kiyoshi Ueda   blk_end_request: ...
2746
2747
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2748
2749
   * __blk_end_request_all - Helper function for drives to finish the request.
   * @rq: the request to finish
8ebf97560   Randy Dunlap   block: fix kernel...
2750
   * @error: %0 for success, < %0 for error
32fab448e   Kiyoshi Ueda   block: add reques...
2751
2752
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2753
   *     Completely finish @rq.  Must be called with queue lock held.
32fab448e   Kiyoshi Ueda   block: add reques...
2754
   */
b1f744937   FUJITA Tomonori   block: move compl...
2755
  void __blk_end_request_all(struct request *rq, int error)
32fab448e   Kiyoshi Ueda   block: add reques...
2756
  {
b1f744937   FUJITA Tomonori   block: move compl...
2757
2758
2759
2760
2761
2762
2763
2764
  	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...
2765
  }
56ad1740d   Jens Axboe   block: make the e...
2766
  EXPORT_SYMBOL(__blk_end_request_all);
32fab448e   Kiyoshi Ueda   block: add reques...
2767
2768
  
  /**
b1f744937   FUJITA Tomonori   block: move compl...
2769
2770
   * __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...
2771
   * @error: %0 for success, < %0 for error
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2772
2773
   *
   * Description:
b1f744937   FUJITA Tomonori   block: move compl...
2774
2775
   *     Complete the current consecutively mapped chunk from @rq.  Must
   *     be called with queue lock held.
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2776
2777
   *
   * Return:
b1f744937   FUJITA Tomonori   block: move compl...
2778
2779
2780
2781
   *     %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: ...
2782
  {
b1f744937   FUJITA Tomonori   block: move compl...
2783
  	return __blk_end_request(rq, error, blk_rq_cur_bytes(rq));
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2784
  }
56ad1740d   Jens Axboe   block: make the e...
2785
  EXPORT_SYMBOL(__blk_end_request_cur);
e19a3ab05   Kiyoshi Ueda   blk_end_request: ...
2786

80a761fd3   Tejun Heo   block: implement ...
2787
2788
2789
2790
2791
2792
2793
2794
2795
2796
2797
2798
2799
2800
2801
2802
2803
2804
2805
  /**
   * __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...
2806
2807
  void blk_rq_bio_prep(struct request_queue *q, struct request *rq,
  		     struct bio *bio)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2808
  {
4993b77d3   Mike Christie   block: copy bio o...
2809
  	req_set_op(rq, bio_op(bio));
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2810

b4f42e283   Jens Axboe   block: remove str...
2811
  	if (bio_has_data(bio))
fb2dce862   David Woodhouse   Add 'discard' req...
2812
  		rq->nr_phys_segments = bio_phys_segments(q, bio);
b4f42e283   Jens Axboe   block: remove str...
2813

4f024f379   Kent Overstreet   block: Abstract o...
2814
  	rq->__data_len = bio->bi_iter.bi_size;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2815
  	rq->bio = rq->biotail = bio;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2816

66846572b   NeilBrown   Stop exporting bl...
2817
2818
2819
  	if (bio->bi_bdev)
  		rq->rq_disk = bio->bi_bdev->bd_disk;
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2820

2d4dc890b   Ilya Loginov   block: add helper...
2821
2822
2823
2824
2825
2826
2827
2828
2829
2830
2831
  #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...
2832
  	struct bio_vec bvec;
2d4dc890b   Ilya Loginov   block: add helper...
2833
2834
  
  	rq_for_each_segment(bvec, rq, iter)
7988613b0   Kent Overstreet   block: Convert bi...
2835
  		flush_dcache_page(bvec.bv_page);
2d4dc890b   Ilya Loginov   block: add helper...
2836
2837
2838
  }
  EXPORT_SYMBOL_GPL(rq_flush_dcache_pages);
  #endif
ef9e3facd   Kiyoshi Ueda   block: add lld bu...
2839
2840
2841
2842
2843
2844
2845
2846
2847
2848
2849
2850
2851
2852
2853
2854
2855
2856
2857
2858
2859
2860
2861
2862
2863
2864
2865
  /**
   * blk_lld_busy - Check if underlying low-level drivers of a device are busy
   * @q : the queue of the device being checked
   *
   * Description:
   *    Check if underlying low-level drivers of a device are busy.
   *    If the drivers want to export their busy state, they must set own
   *    exporting function using blk_queue_lld_busy() first.
   *
   *    Basically, this function is used only by request stacking drivers
   *    to stop dispatching requests to underlying devices when underlying
   *    devices are busy.  This behavior helps more I/O merging on the queue
   *    of the request stacking driver and prevents I/O throughput regression
   *    on burst I/O load.
   *
   * Return:
   *    0 - Not busy (The request stacking driver should dispatch request)
   *    1 - Busy (The request stacking driver should stop dispatching request)
   */
  int blk_lld_busy(struct request_queue *q)
  {
  	if (q->lld_busy_fn)
  		return q->lld_busy_fn(q);
  
  	return 0;
  }
  EXPORT_SYMBOL_GPL(blk_lld_busy);
78d8e58a0   Mike Snitzer   Revert "block, dm...
2866
2867
2868
2869
2870
2871
2872
2873
2874
2875
2876
2877
2878
2879
2880
2881
2882
2883
2884
2885
2886
2887
2888
2889
  /**
   * blk_rq_unprep_clone - Helper function to free all bios in a cloned request
   * @rq: the clone request to be cleaned up
   *
   * Description:
   *     Free all bios in @rq for a cloned request.
   */
  void blk_rq_unprep_clone(struct request *rq)
  {
  	struct bio *bio;
  
  	while ((bio = rq->bio) != NULL) {
  		rq->bio = bio->bi_next;
  
  		bio_put(bio);
  	}
  }
  EXPORT_SYMBOL_GPL(blk_rq_unprep_clone);
  
  /*
   * Copy attributes of the original request to the clone request.
   * The actual data parts (e.g. ->cmd, ->sense) are not copied.
   */
  static void __blk_rq_prep_clone(struct request *dst, struct request *src)
b0fd271d5   Kiyoshi Ueda   block: add reques...
2890
2891
  {
  	dst->cpu = src->cpu;
4993b77d3   Mike Christie   block: copy bio o...
2892
2893
  	req_set_op_attrs(dst, req_op(src),
  			 (src->cmd_flags & REQ_CLONE_MASK) | REQ_NOMERGE);
b0fd271d5   Kiyoshi Ueda   block: add reques...
2894
2895
2896
2897
2898
2899
  	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;
78d8e58a0   Mike Snitzer   Revert "block, dm...
2900
2901
2902
2903
2904
2905
2906
2907
2908
2909
2910
2911
2912
2913
2914
2915
2916
2917
2918
2919
2920
2921
2922
2923
2924
2925
2926
2927
2928
2929
2930
2931
2932
2933
2934
2935
2936
2937
2938
2939
2940
2941
2942
2943
2944
2945
2946
2947
2948
2949
2950
2951
2952
2953
2954
2955
  }
  
  /**
   * blk_rq_prep_clone - Helper function to setup clone request
   * @rq: the request to be setup
   * @rq_src: original request to be cloned
   * @bs: bio_set that bios for clone are allocated from
   * @gfp_mask: memory allocation mask for bio
   * @bio_ctr: setup function to be called for each clone bio.
   *           Returns %0 for success, non %0 for failure.
   * @data: private data to be passed to @bio_ctr
   *
   * Description:
   *     Clones bios in @rq_src to @rq, and copies attributes of @rq_src to @rq.
   *     The actual data parts of @rq_src (e.g. ->cmd, ->sense)
   *     are not copied, and copying such parts is the caller's responsibility.
   *     Also, pages which the original bios are pointing to are not copied
   *     and the cloned bios just point same pages.
   *     So cloned bios must be completed before original bios, which means
   *     the caller must complete @rq before @rq_src.
   */
  int blk_rq_prep_clone(struct request *rq, struct request *rq_src,
  		      struct bio_set *bs, gfp_t gfp_mask,
  		      int (*bio_ctr)(struct bio *, struct bio *, void *),
  		      void *data)
  {
  	struct bio *bio, *bio_src;
  
  	if (!bs)
  		bs = fs_bio_set;
  
  	__rq_for_each_bio(bio_src, rq_src) {
  		bio = bio_clone_fast(bio_src, gfp_mask, bs);
  		if (!bio)
  			goto free_and_out;
  
  		if (bio_ctr && bio_ctr(bio, bio_src, data))
  			goto free_and_out;
  
  		if (rq->bio) {
  			rq->biotail->bi_next = bio;
  			rq->biotail = bio;
  		} else
  			rq->bio = rq->biotail = bio;
  	}
  
  	__blk_rq_prep_clone(rq, rq_src);
  
  	return 0;
  
  free_and_out:
  	if (bio)
  		bio_put(bio);
  	blk_rq_unprep_clone(rq);
  
  	return -ENOMEM;
b0fd271d5   Kiyoshi Ueda   block: add reques...
2956
2957
  }
  EXPORT_SYMBOL_GPL(blk_rq_prep_clone);
59c3d45e4   Jens Axboe   block: remove 'q'...
2958
  int kblockd_schedule_work(struct work_struct *work)
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2959
2960
2961
  {
  	return queue_work(kblockd_workqueue, work);
  }
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
2962
  EXPORT_SYMBOL(kblockd_schedule_work);
59c3d45e4   Jens Axboe   block: remove 'q'...
2963
2964
  int kblockd_schedule_delayed_work(struct delayed_work *dwork,
  				  unsigned long delay)
e43473b7f   Vivek Goyal   blkio: Core imple...
2965
2966
2967
2968
  {
  	return queue_delayed_work(kblockd_workqueue, dwork, delay);
  }
  EXPORT_SYMBOL(kblockd_schedule_delayed_work);
8ab14595b   Jens Axboe   block: add kblock...
2969
2970
2971
2972
2973
2974
  int kblockd_schedule_delayed_work_on(int cpu, struct delayed_work *dwork,
  				     unsigned long delay)
  {
  	return queue_delayed_work_on(cpu, kblockd_workqueue, dwork, delay);
  }
  EXPORT_SYMBOL(kblockd_schedule_delayed_work_on);
75df71362   Suresh Jayaraman   block: document b...
2975
2976
2977
2978
2979
2980
2981
2982
2983
2984
2985
2986
2987
2988
  /**
   * 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...
2989
2990
2991
  void blk_start_plug(struct blk_plug *plug)
  {
  	struct task_struct *tsk = current;
dd6cf3e18   Shaohua Li   blk: clean up plug
2992
2993
2994
2995
2996
  	/*
  	 * If this is a nested plug, don't actually assign it.
  	 */
  	if (tsk->plug)
  		return;
73c101011   Jens Axboe   block: initial pa...
2997
  	INIT_LIST_HEAD(&plug->list);
320ae51fe   Jens Axboe   blk-mq: new multi...
2998
  	INIT_LIST_HEAD(&plug->mq_list);
048c9374a   NeilBrown   block: Enhance ne...
2999
  	INIT_LIST_HEAD(&plug->cb_list);
73c101011   Jens Axboe   block: initial pa...
3000
  	/*
dd6cf3e18   Shaohua Li   blk: clean up plug
3001
3002
  	 * Store ordering should not be needed here, since a potential
  	 * preempt will imply a full memory barrier
73c101011   Jens Axboe   block: initial pa...
3003
  	 */
dd6cf3e18   Shaohua Li   blk: clean up plug
3004
  	tsk->plug = plug;
73c101011   Jens Axboe   block: initial pa...
3005
3006
3007
3008
3009
3010
3011
  }
  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...
3012
3013
  	return !(rqa->q < rqb->q ||
  		(rqa->q == rqb->q && blk_rq_pos(rqa) < blk_rq_pos(rqb)));
73c101011   Jens Axboe   block: initial pa...
3014
  }
49cac01e1   Jens Axboe   block: make unplu...
3015
3016
3017
3018
3019
3020
  /*
   * 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...
3021
  static void queue_unplugged(struct request_queue *q, unsigned int depth,
49cac01e1   Jens Axboe   block: make unplu...
3022
  			    bool from_schedule)
99e22598e   Jens Axboe   block: drop queue...
3023
  	__releases(q->queue_lock)
94b5eb28b   Jens Axboe   block: fixup bloc...
3024
  {
49cac01e1   Jens Axboe   block: make unplu...
3025
  	trace_block_unplug(q, depth, !from_schedule);
99e22598e   Jens Axboe   block: drop queue...
3026

704605711   Bart Van Assche   block: Avoid sche...
3027
  	if (from_schedule)
24ecfbe27   Christoph Hellwig   block: add blk_ru...
3028
  		blk_run_queue_async(q);
704605711   Bart Van Assche   block: Avoid sche...
3029
  	else
24ecfbe27   Christoph Hellwig   block: add blk_ru...
3030
  		__blk_run_queue(q);
704605711   Bart Van Assche   block: Avoid sche...
3031
  	spin_unlock(q->queue_lock);
94b5eb28b   Jens Axboe   block: fixup bloc...
3032
  }
74018dc30   NeilBrown   blk: pass from_sc...
3033
  static void flush_plug_callbacks(struct blk_plug *plug, bool from_schedule)
048c9374a   NeilBrown   block: Enhance ne...
3034
3035
  {
  	LIST_HEAD(callbacks);
2a7d5559b   Shaohua Li   block: stack unplug
3036
3037
  	while (!list_empty(&plug->cb_list)) {
  		list_splice_init(&plug->cb_list, &callbacks);
048c9374a   NeilBrown   block: Enhance ne...
3038

2a7d5559b   Shaohua Li   block: stack unplug
3039
3040
  		while (!list_empty(&callbacks)) {
  			struct blk_plug_cb *cb = list_first_entry(&callbacks,
048c9374a   NeilBrown   block: Enhance ne...
3041
3042
  							  struct blk_plug_cb,
  							  list);
2a7d5559b   Shaohua Li   block: stack unplug
3043
  			list_del(&cb->list);
74018dc30   NeilBrown   blk: pass from_sc...
3044
  			cb->callback(cb, from_schedule);
2a7d5559b   Shaohua Li   block: stack unplug
3045
  		}
048c9374a   NeilBrown   block: Enhance ne...
3046
3047
  	}
  }
9cbb17508   NeilBrown   blk: centralize n...
3048
3049
3050
3051
3052
3053
3054
3055
3056
3057
3058
3059
3060
3061
3062
3063
3064
3065
3066
3067
3068
3069
3070
3071
  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...
3072
  void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule)
73c101011   Jens Axboe   block: initial pa...
3073
3074
3075
3076
  {
  	struct request_queue *q;
  	unsigned long flags;
  	struct request *rq;
109b81296   NeilBrown   block: splice plu...
3077
  	LIST_HEAD(list);
94b5eb28b   Jens Axboe   block: fixup bloc...
3078
  	unsigned int depth;
73c101011   Jens Axboe   block: initial pa...
3079

74018dc30   NeilBrown   blk: pass from_sc...
3080
  	flush_plug_callbacks(plug, from_schedule);
320ae51fe   Jens Axboe   blk-mq: new multi...
3081
3082
3083
  
  	if (!list_empty(&plug->mq_list))
  		blk_mq_flush_plug_list(plug, from_schedule);
73c101011   Jens Axboe   block: initial pa...
3084
3085
  	if (list_empty(&plug->list))
  		return;
109b81296   NeilBrown   block: splice plu...
3086
  	list_splice_init(&plug->list, &list);
422765c26   Jianpeng Ma   block: Remove sho...
3087
  	list_sort(NULL, &list, plug_rq_cmp);
73c101011   Jens Axboe   block: initial pa...
3088
3089
  
  	q = NULL;
94b5eb28b   Jens Axboe   block: fixup bloc...
3090
  	depth = 0;
188112722   Jens Axboe   block: add commen...
3091
3092
3093
3094
3095
  
  	/*
  	 * Save and disable interrupts here, to avoid doing it for every
  	 * queue lock we have to take.
  	 */
73c101011   Jens Axboe   block: initial pa...
3096
  	local_irq_save(flags);
109b81296   NeilBrown   block: splice plu...
3097
3098
  	while (!list_empty(&list)) {
  		rq = list_entry_rq(list.next);
73c101011   Jens Axboe   block: initial pa...
3099
  		list_del_init(&rq->queuelist);
73c101011   Jens Axboe   block: initial pa...
3100
3101
  		BUG_ON(!rq->q);
  		if (rq->q != q) {
99e22598e   Jens Axboe   block: drop queue...
3102
3103
3104
3105
  			/*
  			 * This drops the queue lock
  			 */
  			if (q)
49cac01e1   Jens Axboe   block: make unplu...
3106
  				queue_unplugged(q, depth, from_schedule);
73c101011   Jens Axboe   block: initial pa...
3107
  			q = rq->q;
94b5eb28b   Jens Axboe   block: fixup bloc...
3108
  			depth = 0;
73c101011   Jens Axboe   block: initial pa...
3109
3110
  			spin_lock(q->queue_lock);
  		}
8ba61435d   Tejun Heo   block: add missin...
3111
3112
3113
3114
  
  		/*
  		 * Short-circuit if @q is dead
  		 */
3f3299d5c   Bart Van Assche   block: Rename que...
3115
  		if (unlikely(blk_queue_dying(q))) {
8ba61435d   Tejun Heo   block: add missin...
3116
3117
3118
  			__blk_end_request_all(rq, -ENODEV);
  			continue;
  		}
73c101011   Jens Axboe   block: initial pa...
3119
3120
3121
  		/*
  		 * rq is already accounted, so use raw insert
  		 */
28a8f0d31   Mike Christie   block, drivers, f...
3122
  		if (rq->cmd_flags & (REQ_PREFLUSH | REQ_FUA))
401a18e92   Jens Axboe   block: fix bug wi...
3123
3124
3125
  			__elv_add_request(q, rq, ELEVATOR_INSERT_FLUSH);
  		else
  			__elv_add_request(q, rq, ELEVATOR_INSERT_SORT_MERGE);
94b5eb28b   Jens Axboe   block: fixup bloc...
3126
3127
  
  		depth++;
73c101011   Jens Axboe   block: initial pa...
3128
  	}
99e22598e   Jens Axboe   block: drop queue...
3129
3130
3131
3132
  	/*
  	 * This drops the queue lock
  	 */
  	if (q)
49cac01e1   Jens Axboe   block: make unplu...
3133
  		queue_unplugged(q, depth, from_schedule);
73c101011   Jens Axboe   block: initial pa...
3134

73c101011   Jens Axboe   block: initial pa...
3135
3136
  	local_irq_restore(flags);
  }
73c101011   Jens Axboe   block: initial pa...
3137
3138
3139
  
  void blk_finish_plug(struct blk_plug *plug)
  {
dd6cf3e18   Shaohua Li   blk: clean up plug
3140
3141
  	if (plug != current->plug)
  		return;
f6603783f   Jens Axboe   block: only force...
3142
  	blk_flush_plug_list(plug, false);
73c101011   Jens Axboe   block: initial pa...
3143

dd6cf3e18   Shaohua Li   blk: clean up plug
3144
  	current->plug = NULL;
73c101011   Jens Axboe   block: initial pa...
3145
  }
88b996cd0   Christoph Hellwig   block: cleanup th...
3146
  EXPORT_SYMBOL(blk_finish_plug);
73c101011   Jens Axboe   block: initial pa...
3147

05229beed   Jens Axboe   block: add block ...
3148
3149
3150
3151
3152
3153
3154
3155
3156
3157
3158
3159
3160
3161
3162
3163
3164
3165
3166
3167
3168
3169
3170
3171
3172
3173
3174
3175
3176
3177
3178
3179
3180
3181
3182
3183
3184
3185
3186
3187
  bool blk_poll(struct request_queue *q, blk_qc_t cookie)
  {
  	struct blk_plug *plug;
  	long state;
  
  	if (!q->mq_ops || !q->mq_ops->poll || !blk_qc_t_valid(cookie) ||
  	    !test_bit(QUEUE_FLAG_POLL, &q->queue_flags))
  		return false;
  
  	plug = current->plug;
  	if (plug)
  		blk_flush_plug_list(plug, false);
  
  	state = current->state;
  	while (!need_resched()) {
  		unsigned int queue_num = blk_qc_t_to_queue_num(cookie);
  		struct blk_mq_hw_ctx *hctx = q->queue_hw_ctx[queue_num];
  		int ret;
  
  		hctx->poll_invoked++;
  
  		ret = q->mq_ops->poll(hctx, blk_qc_t_to_tag(cookie));
  		if (ret > 0) {
  			hctx->poll_success++;
  			set_current_state(TASK_RUNNING);
  			return true;
  		}
  
  		if (signal_pending_state(state, current))
  			set_current_state(TASK_RUNNING);
  
  		if (current->state == TASK_RUNNING)
  			return true;
  		if (ret < 0)
  			break;
  		cpu_relax();
  	}
  
  	return false;
  }
9645c1a23   Sagi Grimberg   block: Export blk...
3188
  EXPORT_SYMBOL_GPL(blk_poll);
05229beed   Jens Axboe   block: add block ...
3189

47fafbc70   Rafael J. Wysocki   block / PM: Repla...
3190
  #ifdef CONFIG_PM
6c9546675   Lin Ming   block: add runtim...
3191
3192
3193
3194
3195
3196
3197
3198
3199
3200
3201
3202
3203
3204
3205
3206
3207
3208
3209
3210
3211
3212
3213
3214
3215
3216
3217
3218
3219
3220
3221
3222
3223
3224
3225
3226
3227
3228
3229
3230
3231
3232
3233
3234
3235
3236
3237
3238
3239
3240
3241
3242
3243
3244
  /**
   * 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;
4fd41a855   Ken Xue   SCSI: Fix NULL po...
3245
3246
  	if (!q->dev)
  		return ret;
6c9546675   Lin Ming   block: add runtim...
3247
3248
3249
3250
3251
3252
3253
3254
3255
3256
3257
3258
3259
3260
3261
3262
3263
3264
3265
3266
3267
3268
3269
3270
3271
3272
3273
  	spin_lock_irq(q->queue_lock);
  	if (q->nr_pending) {
  		ret = -EBUSY;
  		pm_runtime_mark_last_busy(q->dev);
  	} else {
  		q->rpm_status = RPM_SUSPENDING;
  	}
  	spin_unlock_irq(q->queue_lock);
  	return ret;
  }
  EXPORT_SYMBOL(blk_pre_runtime_suspend);
  
  /**
   * blk_post_runtime_suspend - Post runtime suspend processing
   * @q: the queue of the device
   * @err: return value of the device's runtime_suspend function
   *
   * Description:
   *    Update the queue's runtime status according to the return value of the
   *    device's runtime suspend function and mark last busy for the device so
   *    that PM core will try to auto suspend the device at a later time.
   *
   *    This function should be called near the end of the device's
   *    runtime_suspend callback.
   */
  void blk_post_runtime_suspend(struct request_queue *q, int err)
  {
4fd41a855   Ken Xue   SCSI: Fix NULL po...
3274
3275
  	if (!q->dev)
  		return;
6c9546675   Lin Ming   block: add runtim...
3276
3277
3278
3279
3280
3281
3282
3283
3284
3285
3286
3287
3288
3289
3290
3291
3292
3293
3294
3295
3296
3297
3298
3299
  	spin_lock_irq(q->queue_lock);
  	if (!err) {
  		q->rpm_status = RPM_SUSPENDED;
  	} else {
  		q->rpm_status = RPM_ACTIVE;
  		pm_runtime_mark_last_busy(q->dev);
  	}
  	spin_unlock_irq(q->queue_lock);
  }
  EXPORT_SYMBOL(blk_post_runtime_suspend);
  
  /**
   * blk_pre_runtime_resume - Pre runtime resume processing
   * @q: the queue of the device
   *
   * Description:
   *    Update the queue's runtime status to RESUMING in preparation for the
   *    runtime resume of the device.
   *
   *    This function should be called near the start of the device's
   *    runtime_resume callback.
   */
  void blk_pre_runtime_resume(struct request_queue *q)
  {
4fd41a855   Ken Xue   SCSI: Fix NULL po...
3300
3301
  	if (!q->dev)
  		return;
6c9546675   Lin Ming   block: add runtim...
3302
3303
3304
3305
3306
3307
3308
3309
3310
3311
3312
3313
3314
3315
3316
3317
3318
3319
3320
3321
3322
3323
  	spin_lock_irq(q->queue_lock);
  	q->rpm_status = RPM_RESUMING;
  	spin_unlock_irq(q->queue_lock);
  }
  EXPORT_SYMBOL(blk_pre_runtime_resume);
  
  /**
   * blk_post_runtime_resume - Post runtime resume processing
   * @q: the queue of the device
   * @err: return value of the device's runtime_resume function
   *
   * Description:
   *    Update the queue's runtime status according to the return value of the
   *    device's runtime_resume function. If it is successfully resumed, process
   *    the requests that are queued into the device's queue when it is resuming
   *    and then mark last busy and initiate autosuspend for it.
   *
   *    This function should be called near the end of the device's
   *    runtime_resume callback.
   */
  void blk_post_runtime_resume(struct request_queue *q, int err)
  {
4fd41a855   Ken Xue   SCSI: Fix NULL po...
3324
3325
  	if (!q->dev)
  		return;
6c9546675   Lin Ming   block: add runtim...
3326
3327
3328
3329
3330
  	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...
3331
  		pm_request_autosuspend(q->dev);
6c9546675   Lin Ming   block: add runtim...
3332
3333
3334
3335
3336
3337
  	} else {
  		q->rpm_status = RPM_SUSPENDED;
  	}
  	spin_unlock_irq(q->queue_lock);
  }
  EXPORT_SYMBOL(blk_post_runtime_resume);
d07ab6d11   Mika Westerberg   block: Add blk_se...
3338
3339
3340
3341
3342
3343
3344
3345
3346
3347
3348
3349
3350
3351
3352
3353
3354
3355
3356
3357
3358
3359
3360
3361
  
  /**
   * blk_set_runtime_active - Force runtime status of the queue to be active
   * @q: the queue of the device
   *
   * If the device is left runtime suspended during system suspend the resume
   * hook typically resumes the device and corrects runtime status
   * accordingly. However, that does not affect the queue runtime PM status
   * which is still "suspended". This prevents processing requests from the
   * queue.
   *
   * This function can be used in driver's resume hook to correct queue
   * runtime PM status and re-enable peeking requests from the queue. It
   * should be called before first request is added to the queue.
   */
  void blk_set_runtime_active(struct request_queue *q)
  {
  	spin_lock_irq(q->queue_lock);
  	q->rpm_status = RPM_ACTIVE;
  	pm_runtime_mark_last_busy(q->dev);
  	pm_request_autosuspend(q->dev);
  	spin_unlock_irq(q->queue_lock);
  }
  EXPORT_SYMBOL(blk_set_runtime_active);
6c9546675   Lin Ming   block: add runtim...
3362
  #endif
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3363
3364
  int __init blk_dev_init(void)
  {
9eb55b030   Nikanth Karthikesan   block: catch tryi...
3365
  	BUILD_BUG_ON(__REQ_NR_BITS > 8 *
0762b23d2   Maninder Singh   block: use FIELD_...
3366
  			FIELD_SIZEOF(struct request, cmd_flags));
9eb55b030   Nikanth Karthikesan   block: catch tryi...
3367

89b90be2d   Tejun Heo   block: make kbloc...
3368
3369
  	/* used for unplugging and affects IO latency/throughput - HIGHPRI */
  	kblockd_workqueue = alloc_workqueue("kblockd",
28747fcd2   Matias Bjørling   block: remove WQ_...
3370
  					    WQ_MEM_RECLAIM | WQ_HIGHPRI, 0);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3371
3372
3373
3374
3375
  	if (!kblockd_workqueue)
  		panic("Failed to create kblockd
  ");
  
  	request_cachep = kmem_cache_create("blkdev_requests",
20c2df83d   Paul Mundt   mm: Remove slab d...
3376
  			sizeof(struct request), 0, SLAB_PANIC, NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3377

c2789bd40   Ilya Dryomov   block: rename req...
3378
  	blk_requestq_cachep = kmem_cache_create("request_queue",
165125e1e   Jens Axboe   [BLOCK] Get rid o...
3379
  			sizeof(struct request_queue), 0, SLAB_PANIC, NULL);
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3380

d38ecf935   Jens Axboe   io context sharin...
3381
  	return 0;
1da177e4c   Linus Torvalds   Linux-2.6.12-rc2
3382
  }