blk-merge.c 11 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 = test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags);
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 (!test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags))
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
105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
	 * these two to be merged into one
	 */
	if (BIO_SEG_BOUNDARY(q, bio, nxt))
		return 1;

	return 0;
}

/*
 * 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;
126
	cluster = test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags);
127 128 129 130 131 132 133 134 135 136

	/*
	 * for each bio in rq
	 */
	bvprv = NULL;
	sg = NULL;
	rq_for_each_segment(bvec, rq, iter) {
		int nbytes = bvec->bv_len;

		if (bvprv && cluster) {
137
			if (sg->length + nbytes > queue_max_segment_size(q))
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
				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.
				 */
				sg->page_link &= ~0x02;
				sg = sg_next(sg);
			}

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

171 172

	if (unlikely(rq->cmd_flags & REQ_COPY_USER) &&
173 174 175
	    (blk_rq_bytes(rq) & q->dma_pad_mask)) {
		unsigned int pad_len =
			(q->dma_pad_mask & ~blk_rq_bytes(rq)) + 1;
176 177 178 179 180

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

181
	if (q->dma_drain_size && q->dma_drain_needed(rq)) {
182
		if (rq->cmd_flags & REQ_WRITE)
183 184
			memset(q->dma_drain_buffer, 0, q->dma_drain_size);

185 186 187 188 189 190 191
		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++;
192
		rq->extra_len += q->dma_drain_size;
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207
	}

	if (sg)
		sg_mark_end(sg);

	return nsegs;
}
EXPORT_SYMBOL(blk_rq_map_sg);

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);

208 209 210 211 212
	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;
213 214 215 216 217 218 219

	/*
	 * This will form the start of a new hw segment.  Bump both
	 * counters.
	 */
	req->nr_phys_segments += nr_phys_segs;
	return 1;
220 221 222 223 224 225

no_merge:
	req->cmd_flags |= REQ_NOMERGE;
	if (req == q->last_merge)
		q->last_merge = NULL;
	return 0;
226 227 228 229 230 231 232
}

int ll_back_merge_fn(struct request_queue *q, struct request *req,
		     struct bio *bio)
{
	unsigned short max_sectors;

233
	if (unlikely(req->cmd_type == REQ_TYPE_BLOCK_PC))
234
		max_sectors = queue_max_hw_sectors(q);
235
	else
236
		max_sectors = queue_max_sectors(q);
237

238
	if (blk_rq_sectors(req) + bio_sectors(bio) > max_sectors) {
239 240 241 242 243
		req->cmd_flags |= REQ_NOMERGE;
		if (req == q->last_merge)
			q->last_merge = NULL;
		return 0;
	}
244
	if (!bio_flagged(req->biotail, BIO_SEG_VALID))
245
		blk_recount_segments(q, req->biotail);
246
	if (!bio_flagged(bio, BIO_SEG_VALID))
247 248 249 250 251
		blk_recount_segments(q, bio);

	return ll_new_hw_segment(q, req, bio);
}

252
int ll_front_merge_fn(struct request_queue *q, struct request *req,
253 254 255 256
		      struct bio *bio)
{
	unsigned short max_sectors;

257
	if (unlikely(req->cmd_type == REQ_TYPE_BLOCK_PC))
258
		max_sectors = queue_max_hw_sectors(q);
259
	else
260
		max_sectors = queue_max_sectors(q);
261 262


263
	if (blk_rq_sectors(req) + bio_sectors(bio) > max_sectors) {
264 265 266 267 268
		req->cmd_flags |= REQ_NOMERGE;
		if (req == q->last_merge)
			q->last_merge = NULL;
		return 0;
	}
269
	if (!bio_flagged(bio, BIO_SEG_VALID))
270
		blk_recount_segments(q, bio);
271
	if (!bio_flagged(req->bio, BIO_SEG_VALID))
272 273 274 275 276 277 278 279 280
		blk_recount_segments(q, req->bio);

	return ll_new_hw_segment(q, req, bio);
}

static int ll_merge_requests_fn(struct request_queue *q, struct request *req,
				struct request *next)
{
	int total_phys_segments;
281 282
	unsigned int seg_size =
		req->biotail->bi_seg_back_size + next->bio->bi_seg_front_size;
283 284 285 286 287 288 289 290 291 292 293

	/*
	 * First check if the either of the requests are re-queued
	 * requests.  Can't merge them if they are.
	 */
	if (req->special || next->special)
		return 0;

	/*
	 * Will it become too large?
	 */
294
	if ((blk_rq_sectors(req) + blk_rq_sectors(next)) > queue_max_sectors(q))
295 296 297
		return 0;

	total_phys_segments = req->nr_phys_segments + next->nr_phys_segments;
298 299 300 301 302
	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;
303
		total_phys_segments--;
304
	}
305

306
	if (total_phys_segments > queue_max_segments(q))
307 308
		return 0;

309 310 311
	if (blk_integrity_rq(req) && blk_integrity_merge_rq(q, req, next))
		return 0;

312 313 314 315 316
	/* Merge is OK... */
	req->nr_phys_segments = total_phys_segments;
	return 1;
}

317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
/**
 * 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;
}

347 348 349 350 351 352 353
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();
354
		part = disk_map_sector_rcu(req->rq_disk, blk_rq_pos(req));
355 356

		part_round_stats(cpu, part);
357
		part_dec_in_flight(part, rq_data_dir(req));
358 359 360 361 362

		part_stat_unlock();
	}
}

363 364 365 366 367 368 369 370 371
/*
 * 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;

372 373 374 375 376 377 378 379 380 381 382 383
	/*
	 * Don't merge file system requests and discard requests
	 */
	if ((req->cmd_flags & REQ_DISCARD) != (next->cmd_flags & REQ_DISCARD))
		return 0;

	/*
	 * Don't merge discard requests and secure discard requests
	 */
	if ((req->cmd_flags & REQ_SECURE) != (next->cmd_flags & REQ_SECURE))
		return 0;

384 385 386
	/*
	 * not contiguous
	 */
387
	if (blk_rq_pos(req) + blk_rq_sectors(req) != blk_rq_pos(next))
388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
		return 0;

	if (rq_data_dir(req) != rq_data_dir(next)
	    || req->rq_disk != next->rq_disk
	    || next->special)
		return 0;

	/*
	 * 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;

404 405 406 407 408 409 410 411 412 413 414 415 416
	/*
	 * 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);
	}

417 418 419 420 421 422 423 424 425 426 427 428
	/*
	 * 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;

429
	req->__data_len += blk_rq_bytes(next);
430 431 432

	elv_merge_requests(q, req, next);

433 434 435 436
	/*
	 * 'next' is going away, so update stats accordingly
	 */
	blk_account_io_merge(next);
437 438

	req->ioprio = ioprio_best(req->ioprio, next->ioprio);
439 440
	if (blk_rq_cpu_valid(next))
		req->cpu = next->cpu;
441

442 443
	/* owner-ship of bio passed from next to req */
	next->bio = NULL;
444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466
	__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;
}