blk-merge.c 13.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11
/*
 * Functions related to segment and merge handling
 */
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
#include <linux/scatterlist.h>

#include "blk.h"

12
static unsigned int __blk_recalc_rq_segments(struct request_queue *q,
13
					     struct bio *bio)
14 15
{
	struct bio_vec *bv, *bvprv = NULL;
16 17
	int cluster, i, high, highprv = 1;
	unsigned int seg_size, nr_phys_segs;
18
	struct bio *fbio, *bbio;
19

20 21
	if (!bio)
		return 0;
22

23
	fbio = bio;
24
	cluster = blk_queue_cluster(q);
Mikulas Patocka's avatar
Mikulas Patocka committed
25
	seg_size = 0;
26
	nr_phys_segs = 0;
27 28 29 30 31 32 33
	for_each_bio(bio) {
		bio_for_each_segment(bv, bio, i) {
			/*
			 * the trick here is making sure that a high page is
			 * never considered part of another segment, since that
			 * might change with the bounce page.
			 */
34
			high = page_to_pfn(bv->bv_page) > queue_bounce_pfn(q);
35
			if (high || highprv)
36
				goto new_segment;
37
			if (cluster) {
38 39
				if (seg_size + bv->bv_len
				    > queue_max_segment_size(q))
40 41 42 43 44
					goto new_segment;
				if (!BIOVEC_PHYS_MERGEABLE(bvprv, bv))
					goto new_segment;
				if (!BIOVEC_SEG_BOUNDARY(q, bvprv, bv))
					goto new_segment;
45

46 47 48 49
				seg_size += bv->bv_len;
				bvprv = bv;
				continue;
			}
50
new_segment:
51 52 53
			if (nr_phys_segs == 1 && seg_size >
			    fbio->bi_seg_front_size)
				fbio->bi_seg_front_size = seg_size;
54

55 56 57 58 59
			nr_phys_segs++;
			bvprv = bv;
			seg_size = bv->bv_len;
			highprv = high;
		}
60
		bbio = bio;
61 62
	}

63 64 65 66
	if (nr_phys_segs == 1 && seg_size > fbio->bi_seg_front_size)
		fbio->bi_seg_front_size = seg_size;
	if (seg_size > bbio->bi_seg_back_size)
		bbio->bi_seg_back_size = seg_size;
67 68 69 70 71 72

	return nr_phys_segs;
}

void blk_recalc_rq_segments(struct request *rq)
{
73
	rq->nr_phys_segments = __blk_recalc_rq_segments(rq->q, rq->bio);
74 75 76 77 78
}

void blk_recount_segments(struct request_queue *q, struct bio *bio)
{
	struct bio *nxt = bio->bi_next;
79

80
	bio->bi_next = NULL;
81
	bio->bi_phys_segments = __blk_recalc_rq_segments(q, bio);
82 83 84 85 86 87 88 89
	bio->bi_next = nxt;
	bio->bi_flags |= (1 << BIO_SEG_VALID);
}
EXPORT_SYMBOL(blk_recount_segments);

static int blk_phys_contig_segment(struct request_queue *q, struct bio *bio,
				   struct bio *nxt)
{
90
	if (!blk_queue_cluster(q))
91 92
		return 0;

93
	if (bio->bi_seg_back_size + nxt->bi_seg_front_size >
94
	    queue_max_segment_size(q))
95 96
		return 0;

97 98 99 100 101 102
	if (!bio_has_data(bio))
		return 1;

	if (!BIOVEC_PHYS_MERGEABLE(__BVEC_END(bio), __BVEC_START(nxt)))
		return 0;

103
	/*
104
	 * bio and nxt are contiguous in memory; check if the queue allows
105 106 107 108 109 110 111 112
	 * these two to be merged into one
	 */
	if (BIO_SEG_BOUNDARY(q, bio, nxt))
		return 1;

	return 0;
}

113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145
static void
__blk_segment_map_sg(struct request_queue *q, struct bio_vec *bvec,
		     struct scatterlist *sglist, struct bio_vec **bvprv,
		     struct scatterlist **sg, int *nsegs, int *cluster)
{

	int nbytes = bvec->bv_len;

	if (*bvprv && *cluster) {
		if ((*sg)->length + nbytes > queue_max_segment_size(q))
			goto new_segment;

		if (!BIOVEC_PHYS_MERGEABLE(*bvprv, bvec))
			goto new_segment;
		if (!BIOVEC_SEG_BOUNDARY(q, *bvprv, bvec))
			goto new_segment;

		(*sg)->length += nbytes;
	} else {
new_segment:
		if (!*sg)
			*sg = sglist;
		else {
			/*
			 * If the driver previously mapped a shorter
			 * list, we could see a termination bit
			 * prematurely unless it fully inits the sg
			 * table on each mapping. We KNOW that there
			 * must be more entries here or the driver
			 * would be buggy, so force clear the
			 * termination bit to avoid doing a full
			 * sg_init_table() in drivers for each command.
			 */
146
			sg_unmark_end(*sg);
147 148 149 150 151 152 153 154 155
			*sg = sg_next(*sg);
		}

		sg_set_page(*sg, bvec->bv_page, nbytes, bvec->bv_offset);
		(*nsegs)++;
	}
	*bvprv = bvec;
}

156 157 158 159 160 161 162 163 164 165 166 167 168
/*
 * map a request to scatterlist, return number of sg entries setup. Caller
 * must make sure sg can hold rq->nr_phys_segments entries
 */
int blk_rq_map_sg(struct request_queue *q, struct request *rq,
		  struct scatterlist *sglist)
{
	struct bio_vec *bvec, *bvprv;
	struct req_iterator iter;
	struct scatterlist *sg;
	int nsegs, cluster;

	nsegs = 0;
169
	cluster = blk_queue_cluster(q);
170 171 172 173 174 175 176

	/*
	 * for each bio in rq
	 */
	bvprv = NULL;
	sg = NULL;
	rq_for_each_segment(bvec, rq, iter) {
177 178
		__blk_segment_map_sg(q, bvec, sglist, &bvprv, &sg,
				     &nsegs, &cluster);
179 180
	} /* segments in rq */

181 182

	if (unlikely(rq->cmd_flags & REQ_COPY_USER) &&
183 184 185
	    (blk_rq_bytes(rq) & q->dma_pad_mask)) {
		unsigned int pad_len =
			(q->dma_pad_mask & ~blk_rq_bytes(rq)) + 1;
186 187 188 189 190

		sg->length += pad_len;
		rq->extra_len += pad_len;
	}

191
	if (q->dma_drain_size && q->dma_drain_needed(rq)) {
192
		if (rq->cmd_flags & REQ_WRITE)
193 194
			memset(q->dma_drain_buffer, 0, q->dma_drain_size);

195 196 197 198 199 200 201
		sg->page_link &= ~0x02;
		sg = sg_next(sg);
		sg_set_page(sg, virt_to_page(q->dma_drain_buffer),
			    q->dma_drain_size,
			    ((unsigned long)q->dma_drain_buffer) &
			    (PAGE_SIZE - 1));
		nsegs++;
202
		rq->extra_len += q->dma_drain_size;
203 204 205 206 207 208 209 210 211
	}

	if (sg)
		sg_mark_end(sg);

	return nsegs;
}
EXPORT_SYMBOL(blk_rq_map_sg);

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248
/**
 * blk_bio_map_sg - map a bio to a scatterlist
 * @q: request_queue in question
 * @bio: bio being mapped
 * @sglist: scatterlist being mapped
 *
 * Note:
 *    Caller must make sure sg can hold bio->bi_phys_segments entries
 *
 * Will return the number of sg entries setup
 */
int blk_bio_map_sg(struct request_queue *q, struct bio *bio,
		   struct scatterlist *sglist)
{
	struct bio_vec *bvec, *bvprv;
	struct scatterlist *sg;
	int nsegs, cluster;
	unsigned long i;

	nsegs = 0;
	cluster = blk_queue_cluster(q);

	bvprv = NULL;
	sg = NULL;
	bio_for_each_segment(bvec, bio, i) {
		__blk_segment_map_sg(q, bvec, sglist, &bvprv, &sg,
				     &nsegs, &cluster);
	} /* segments in bio */

	if (sg)
		sg_mark_end(sg);

	BUG_ON(bio->bi_phys_segments && nsegs > bio->bi_phys_segments);
	return nsegs;
}
EXPORT_SYMBOL(blk_bio_map_sg);

249 250 251 252 253 254
static inline int ll_new_hw_segment(struct request_queue *q,
				    struct request *req,
				    struct bio *bio)
{
	int nr_phys_segs = bio_phys_segments(q, bio);

255 256 257 258 259
	if (req->nr_phys_segments + nr_phys_segs > queue_max_segments(q))
		goto no_merge;

	if (bio_integrity(bio) && blk_integrity_merge_bio(q, req, bio))
		goto no_merge;
260 261 262 263 264 265 266

	/*
	 * This will form the start of a new hw segment.  Bump both
	 * counters.
	 */
	req->nr_phys_segments += nr_phys_segs;
	return 1;
267 268 269 270 271 272

no_merge:
	req->cmd_flags |= REQ_NOMERGE;
	if (req == q->last_merge)
		q->last_merge = NULL;
	return 0;
273 274 275 276 277
}

int ll_back_merge_fn(struct request_queue *q, struct request *req,
		     struct bio *bio)
{
278 279
	if (blk_rq_sectors(req) + bio_sectors(bio) >
	    blk_rq_get_max_sectors(req)) {
280 281 282 283 284
		req->cmd_flags |= REQ_NOMERGE;
		if (req == q->last_merge)
			q->last_merge = NULL;
		return 0;
	}
285
	if (!bio_flagged(req->biotail, BIO_SEG_VALID))
286
		blk_recount_segments(q, req->biotail);
287
	if (!bio_flagged(bio, BIO_SEG_VALID))
288 289 290 291 292
		blk_recount_segments(q, bio);

	return ll_new_hw_segment(q, req, bio);
}

293
int ll_front_merge_fn(struct request_queue *q, struct request *req,
294 295
		      struct bio *bio)
{
296 297
	if (blk_rq_sectors(req) + bio_sectors(bio) >
	    blk_rq_get_max_sectors(req)) {
298 299 300 301 302
		req->cmd_flags |= REQ_NOMERGE;
		if (req == q->last_merge)
			q->last_merge = NULL;
		return 0;
	}
303
	if (!bio_flagged(bio, BIO_SEG_VALID))
304
		blk_recount_segments(q, bio);
305
	if (!bio_flagged(req->bio, BIO_SEG_VALID))
306 307 308 309 310
		blk_recount_segments(q, req->bio);

	return ll_new_hw_segment(q, req, bio);
}

311 312 313 314 315 316 317 318 319 320 321
/*
 * blk-mq uses req->special to carry normal driver per-request payload, it
 * does not indicate a prepared command that we cannot merge with.
 */
static bool req_no_special_merge(struct request *req)
{
	struct request_queue *q = req->q;

	return !q->mq_ops && req->special;
}

322 323 324 325
static int ll_merge_requests_fn(struct request_queue *q, struct request *req,
				struct request *next)
{
	int total_phys_segments;
326 327
	unsigned int seg_size =
		req->biotail->bi_seg_back_size + next->bio->bi_seg_front_size;
328 329 330 331 332

	/*
	 * First check if the either of the requests are re-queued
	 * requests.  Can't merge them if they are.
	 */
333
	if (req_no_special_merge(req) || req_no_special_merge(next))
334 335 336 337 338
		return 0;

	/*
	 * Will it become too large?
	 */
339 340
	if ((blk_rq_sectors(req) + blk_rq_sectors(next)) >
	    blk_rq_get_max_sectors(req))
341 342 343
		return 0;

	total_phys_segments = req->nr_phys_segments + next->nr_phys_segments;
344 345 346 347 348
	if (blk_phys_contig_segment(q, req->biotail, next->bio)) {
		if (req->nr_phys_segments == 1)
			req->bio->bi_seg_front_size = seg_size;
		if (next->nr_phys_segments == 1)
			next->biotail->bi_seg_back_size = seg_size;
349
		total_phys_segments--;
350
	}
351

352
	if (total_phys_segments > queue_max_segments(q))
353 354
		return 0;

355 356 357
	if (blk_integrity_rq(req) && blk_integrity_merge_rq(q, req, next))
		return 0;

358 359 360 361 362
	/* Merge is OK... */
	req->nr_phys_segments = total_phys_segments;
	return 1;
}

363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392
/**
 * blk_rq_set_mixed_merge - mark a request as mixed merge
 * @rq: request to mark as mixed merge
 *
 * Description:
 *     @rq is about to be mixed merged.  Make sure the attributes
 *     which can be mixed are set in each bio and mark @rq as mixed
 *     merged.
 */
void blk_rq_set_mixed_merge(struct request *rq)
{
	unsigned int ff = rq->cmd_flags & REQ_FAILFAST_MASK;
	struct bio *bio;

	if (rq->cmd_flags & REQ_MIXED_MERGE)
		return;

	/*
	 * @rq will no longer represent mixable attributes for all the
	 * contained bios.  It will just track those of the first one.
	 * Distributes the attributs to each bio.
	 */
	for (bio = rq->bio; bio; bio = bio->bi_next) {
		WARN_ON_ONCE((bio->bi_rw & REQ_FAILFAST_MASK) &&
			     (bio->bi_rw & REQ_FAILFAST_MASK) != ff);
		bio->bi_rw |= ff;
	}
	rq->cmd_flags |= REQ_MIXED_MERGE;
}

393 394 395 396 397 398 399
static void blk_account_io_merge(struct request *req)
{
	if (blk_do_io_stat(req)) {
		struct hd_struct *part;
		int cpu;

		cpu = part_stat_lock();
400
		part = req->part;
401 402

		part_round_stats(cpu, part);
403
		part_dec_in_flight(part, rq_data_dir(req));
404

405
		hd_struct_put(part);
406 407 408 409
		part_stat_unlock();
	}
}

410 411 412 413 414 415 416 417 418
/*
 * Has to be called with the request spinlock acquired
 */
static int attempt_merge(struct request_queue *q, struct request *req,
			  struct request *next)
{
	if (!rq_mergeable(req) || !rq_mergeable(next))
		return 0;

419 420 421
	if (!blk_check_merge_flags(req->cmd_flags, next->cmd_flags))
		return 0;

422 423 424
	/*
	 * not contiguous
	 */
425
	if (blk_rq_pos(req) + blk_rq_sectors(req) != blk_rq_pos(next))
426 427 428 429
		return 0;

	if (rq_data_dir(req) != rq_data_dir(next)
	    || req->rq_disk != next->rq_disk
430
	    || req_no_special_merge(next))
431 432
		return 0;

433 434 435 436
	if (req->cmd_flags & REQ_WRITE_SAME &&
	    !blk_write_same_mergeable(req->bio, next->bio))
		return 0;

437 438 439 440 441 442 443 444 445
	/*
	 * If we are allowed to merge, then append bio list
	 * from next to rq and release next. merge_requests_fn
	 * will have updated segment counts, update sector
	 * counts here.
	 */
	if (!ll_merge_requests_fn(q, req, next))
		return 0;

446 447 448 449 450 451 452 453 454 455 456 457 458
	/*
	 * If failfast settings disagree or any of the two is already
	 * a mixed merge, mark both as mixed before proceeding.  This
	 * makes sure that all involved bios have mixable attributes
	 * set properly.
	 */
	if ((req->cmd_flags | next->cmd_flags) & REQ_MIXED_MERGE ||
	    (req->cmd_flags & REQ_FAILFAST_MASK) !=
	    (next->cmd_flags & REQ_FAILFAST_MASK)) {
		blk_rq_set_mixed_merge(req);
		blk_rq_set_mixed_merge(next);
	}

459 460 461 462 463 464 465 466 467 468 469 470
	/*
	 * At this point we have either done a back merge
	 * or front merge. We need the smaller start_time of
	 * the merged requests to be the current request
	 * for accounting purposes.
	 */
	if (time_after(req->start_time, next->start_time))
		req->start_time = next->start_time;

	req->biotail->bi_next = next->bio;
	req->biotail = next->biotail;

471
	req->__data_len += blk_rq_bytes(next);
472 473 474

	elv_merge_requests(q, req, next);

475 476 477 478
	/*
	 * 'next' is going away, so update stats accordingly
	 */
	blk_account_io_merge(next);
479 480

	req->ioprio = ioprio_best(req->ioprio, next->ioprio);
481 482
	if (blk_rq_cpu_valid(next))
		req->cpu = next->cpu;
483

484 485
	/* owner-ship of bio passed from next to req */
	next->bio = NULL;
486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508
	__blk_put_request(q, next);
	return 1;
}

int attempt_back_merge(struct request_queue *q, struct request *rq)
{
	struct request *next = elv_latter_request(q, rq);

	if (next)
		return attempt_merge(q, rq, next);

	return 0;
}

int attempt_front_merge(struct request_queue *q, struct request *rq)
{
	struct request *prev = elv_former_request(q, rq);

	if (prev)
		return attempt_merge(q, prev, rq);

	return 0;
}
509 510 511 512 513 514

int blk_attempt_req_merge(struct request_queue *q, struct request *rq,
			  struct request *next)
{
	return attempt_merge(q, rq, next);
}
515 516 517

bool blk_rq_merge_ok(struct request *rq, struct bio *bio)
{
518
	if (!rq_mergeable(rq) || !bio_mergeable(bio))
519 520
		return false;

521 522 523
	if (!blk_check_merge_flags(rq->cmd_flags, bio->bi_rw))
		return false;

524 525 526 527 528
	/* different data direction or already started, don't merge */
	if (bio_data_dir(bio) != rq_data_dir(rq))
		return false;

	/* must be same device and not a special request */
529
	if (rq->rq_disk != bio->bi_bdev->bd_disk || req_no_special_merge(rq))
530 531 532 533 534 535
		return false;

	/* only merge integrity protected bio into ditto rq */
	if (bio_integrity(bio) != blk_integrity_rq(rq))
		return false;

536 537 538 539 540
	/* must be using the same buffer */
	if (rq->cmd_flags & REQ_WRITE_SAME &&
	    !blk_write_same_mergeable(rq->bio, bio))
		return false;

541 542 543 544 545
	return true;
}

int blk_try_merge(struct request *rq, struct bio *bio)
{
546
	if (blk_rq_pos(rq) + blk_rq_sectors(rq) == bio->bi_iter.bi_sector)
547
		return ELEVATOR_BACK_MERGE;
548
	else if (blk_rq_pos(rq) - bio_sectors(bio) == bio->bi_iter.bi_sector)
549 550 551
		return ELEVATOR_FRONT_MERGE;
	return ELEVATOR_NO_MERGE;
}