bio.c 52.6 KB
Newer Older
L
Linus Torvalds 已提交
1
/*
2
 * Copyright (C) 2001 Jens Axboe <axboe@kernel.dk>
L
Linus Torvalds 已提交
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public Licens
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-
 *
 */
#include <linux/mm.h>
#include <linux/swap.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
22
#include <linux/uio.h>
23
#include <linux/iocontext.h>
L
Linus Torvalds 已提交
24 25 26
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/kernel.h>
27
#include <linux/export.h>
L
Linus Torvalds 已提交
28 29
#include <linux/mempool.h>
#include <linux/workqueue.h>
30
#include <linux/cgroup.h>
31
#include <linux/blk-cgroup.h>
L
Linus Torvalds 已提交
32

33
#include <trace/events/block.h>
34
#include "blk.h"
J
Josef Bacik 已提交
35
#include "blk-rq-qos.h"
36

37 38 39 40 41 42
/*
 * Test patch to inline a certain number of bi_io_vec's inside the bio
 * itself, to shrink a bio data allocation from two mempool calls to one
 */
#define BIO_INLINE_VECS		4

L
Linus Torvalds 已提交
43 44 45 46 47
/*
 * if you change this list, also change bvec_alloc or things will
 * break badly! cannot be bigger than what you can fit into an
 * unsigned short
 */
48
#define BV(x, n) { .nr_vecs = x, .name = "biovec-"#n }
49
static struct biovec_slab bvec_slabs[BVEC_POOL_NR] __read_mostly = {
50
	BV(1, 1), BV(4, 4), BV(16, 16), BV(64, 64), BV(128, 128), BV(BIO_MAX_PAGES, max),
L
Linus Torvalds 已提交
51 52 53 54 55 56 57
};
#undef BV

/*
 * fs_bio_set is the bio_set containing bio and iovec memory pools used by
 * IO code that does not need private memory pools.
 */
58
struct bio_set fs_bio_set;
59
EXPORT_SYMBOL(fs_bio_set);
L
Linus Torvalds 已提交
60

61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77
/*
 * Our slab pool management
 */
struct bio_slab {
	struct kmem_cache *slab;
	unsigned int slab_ref;
	unsigned int slab_size;
	char name[8];
};
static DEFINE_MUTEX(bio_slab_lock);
static struct bio_slab *bio_slabs;
static unsigned int bio_slab_nr, bio_slab_max;

static struct kmem_cache *bio_find_or_create_slab(unsigned int extra_size)
{
	unsigned int sz = sizeof(struct bio) + extra_size;
	struct kmem_cache *slab = NULL;
78
	struct bio_slab *bslab, *new_bio_slabs;
79
	unsigned int new_bio_slab_max;
80 81 82 83 84 85
	unsigned int i, entry = -1;

	mutex_lock(&bio_slab_lock);

	i = 0;
	while (i < bio_slab_nr) {
86
		bslab = &bio_slabs[i];
87 88 89 90 91 92 93 94 95 96 97 98 99 100 101

		if (!bslab->slab && entry == -1)
			entry = i;
		else if (bslab->slab_size == sz) {
			slab = bslab->slab;
			bslab->slab_ref++;
			break;
		}
		i++;
	}

	if (slab)
		goto out_unlock;

	if (bio_slab_nr == bio_slab_max && entry == -1) {
102
		new_bio_slab_max = bio_slab_max << 1;
103
		new_bio_slabs = krealloc(bio_slabs,
104
					 new_bio_slab_max * sizeof(struct bio_slab),
105 106
					 GFP_KERNEL);
		if (!new_bio_slabs)
107
			goto out_unlock;
108
		bio_slab_max = new_bio_slab_max;
109
		bio_slabs = new_bio_slabs;
110 111 112 113 114 115 116
	}
	if (entry == -1)
		entry = bio_slab_nr++;

	bslab = &bio_slabs[entry];

	snprintf(bslab->name, sizeof(bslab->name), "bio-%d", entry);
117 118
	slab = kmem_cache_create(bslab->name, sz, ARCH_KMALLOC_MINALIGN,
				 SLAB_HWCACHE_ALIGN, NULL);
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158
	if (!slab)
		goto out_unlock;

	bslab->slab = slab;
	bslab->slab_ref = 1;
	bslab->slab_size = sz;
out_unlock:
	mutex_unlock(&bio_slab_lock);
	return slab;
}

static void bio_put_slab(struct bio_set *bs)
{
	struct bio_slab *bslab = NULL;
	unsigned int i;

	mutex_lock(&bio_slab_lock);

	for (i = 0; i < bio_slab_nr; i++) {
		if (bs->bio_slab == bio_slabs[i].slab) {
			bslab = &bio_slabs[i];
			break;
		}
	}

	if (WARN(!bslab, KERN_ERR "bio: unable to find slab!\n"))
		goto out;

	WARN_ON(!bslab->slab_ref);

	if (--bslab->slab_ref)
		goto out;

	kmem_cache_destroy(bslab->slab);
	bslab->slab = NULL;

out:
	mutex_unlock(&bio_slab_lock);
}

159 160 161 162 163
unsigned int bvec_nr_vecs(unsigned short idx)
{
	return bvec_slabs[idx].nr_vecs;
}

164
void bvec_free(mempool_t *pool, struct bio_vec *bv, unsigned int idx)
165
{
166 167 168 169 170
	if (!idx)
		return;
	idx--;

	BIO_BUG_ON(idx >= BVEC_POOL_NR);
171

172
	if (idx == BVEC_POOL_MAX) {
173
		mempool_free(bv, pool);
174
	} else {
175 176 177 178 179 180
		struct biovec_slab *bvs = bvec_slabs + idx;

		kmem_cache_free(bvs->slab, bv);
	}
}

181 182
struct bio_vec *bvec_alloc(gfp_t gfp_mask, int nr, unsigned long *idx,
			   mempool_t *pool)
L
Linus Torvalds 已提交
183 184 185
{
	struct bio_vec *bvl;

186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215
	/*
	 * see comment near bvec_array define!
	 */
	switch (nr) {
	case 1:
		*idx = 0;
		break;
	case 2 ... 4:
		*idx = 1;
		break;
	case 5 ... 16:
		*idx = 2;
		break;
	case 17 ... 64:
		*idx = 3;
		break;
	case 65 ... 128:
		*idx = 4;
		break;
	case 129 ... BIO_MAX_PAGES:
		*idx = 5;
		break;
	default:
		return NULL;
	}

	/*
	 * idx now points to the pool we want to allocate from. only the
	 * 1-vec entry pool is mempool backed.
	 */
216
	if (*idx == BVEC_POOL_MAX) {
217
fallback:
218
		bvl = mempool_alloc(pool, gfp_mask);
219 220
	} else {
		struct biovec_slab *bvs = bvec_slabs + *idx;
221
		gfp_t __gfp_mask = gfp_mask & ~(__GFP_DIRECT_RECLAIM | __GFP_IO);
222

J
Jens Axboe 已提交
223
		/*
224 225 226
		 * Make this allocation restricted and don't dump info on
		 * allocation failures, since we'll fallback to the mempool
		 * in case of failure.
J
Jens Axboe 已提交
227
		 */
228
		__gfp_mask |= __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN;
L
Linus Torvalds 已提交
229

J
Jens Axboe 已提交
230
		/*
231
		 * Try a slab allocation. If this fails and __GFP_DIRECT_RECLAIM
232
		 * is set, retry with the 1-entry mempool
J
Jens Axboe 已提交
233
		 */
234
		bvl = kmem_cache_alloc(bvs->slab, __gfp_mask);
235
		if (unlikely(!bvl && (gfp_mask & __GFP_DIRECT_RECLAIM))) {
236
			*idx = BVEC_POOL_MAX;
237 238 239 240
			goto fallback;
		}
	}

241
	(*idx)++;
L
Linus Torvalds 已提交
242 243 244
	return bvl;
}

245
void bio_uninit(struct bio *bio)
L
Linus Torvalds 已提交
246
{
K
Kent Overstreet 已提交
247 248
	bio_disassociate_task(bio);
}
249
EXPORT_SYMBOL(bio_uninit);
250

K
Kent Overstreet 已提交
251 252 253 254 255
static void bio_free(struct bio *bio)
{
	struct bio_set *bs = bio->bi_pool;
	void *p;

256
	bio_uninit(bio);
K
Kent Overstreet 已提交
257 258

	if (bs) {
259
		bvec_free(&bs->bvec_pool, bio->bi_io_vec, BVEC_POOL_IDX(bio));
K
Kent Overstreet 已提交
260 261 262 263 264

		/*
		 * If we have front padding, adjust the bio pointer before freeing
		 */
		p = bio;
265 266
		p -= bs->front_pad;

267
		mempool_free(p, &bs->bio_pool);
K
Kent Overstreet 已提交
268 269 270 271
	} else {
		/* Bio was allocated by bio_kmalloc() */
		kfree(bio);
	}
P
Peter Osterlund 已提交
272 273
}

274 275 276 277 278
/*
 * Users of this function have their own bio allocation. Subsequently,
 * they must remember to pair any call to bio_init() with bio_uninit()
 * when IO has completed, or when the bio is released.
 */
279 280
void bio_init(struct bio *bio, struct bio_vec *table,
	      unsigned short max_vecs)
L
Linus Torvalds 已提交
281
{
J
Jens Axboe 已提交
282
	memset(bio, 0, sizeof(*bio));
283
	atomic_set(&bio->__bi_remaining, 1);
284
	atomic_set(&bio->__bi_cnt, 1);
285 286 287

	bio->bi_io_vec = table;
	bio->bi_max_vecs = max_vecs;
L
Linus Torvalds 已提交
288
}
289
EXPORT_SYMBOL(bio_init);
L
Linus Torvalds 已提交
290

K
Kent Overstreet 已提交
291 292 293 294 295 296 297 298 299 300 301 302 303 304
/**
 * bio_reset - reinitialize a bio
 * @bio:	bio to reset
 *
 * Description:
 *   After calling bio_reset(), @bio will be in the same state as a freshly
 *   allocated bio returned bio bio_alloc_bioset() - the only fields that are
 *   preserved are the ones that are initialized by bio_alloc_bioset(). See
 *   comment in struct bio.
 */
void bio_reset(struct bio *bio)
{
	unsigned long flags = bio->bi_flags & (~0UL << BIO_RESET_BITS);

305
	bio_uninit(bio);
K
Kent Overstreet 已提交
306 307

	memset(bio, 0, BIO_RESET_BYTES);
308
	bio->bi_flags = flags;
309
	atomic_set(&bio->__bi_remaining, 1);
K
Kent Overstreet 已提交
310 311 312
}
EXPORT_SYMBOL(bio_reset);

313
static struct bio *__bio_chain_endio(struct bio *bio)
K
Kent Overstreet 已提交
314
{
315 316
	struct bio *parent = bio->bi_private;

317 318
	if (!parent->bi_status)
		parent->bi_status = bio->bi_status;
K
Kent Overstreet 已提交
319
	bio_put(bio);
320 321 322 323 324 325
	return parent;
}

static void bio_chain_endio(struct bio *bio)
{
	bio_endio(__bio_chain_endio(bio));
K
Kent Overstreet 已提交
326 327 328 329
}

/**
 * bio_chain - chain bio completions
330 331
 * @bio: the target bio
 * @parent: the @bio's parent bio
K
Kent Overstreet 已提交
332 333 334 335 336 337 338 339 340 341 342 343 344
 *
 * The caller won't have a bi_end_io called when @bio completes - instead,
 * @parent's bi_end_io won't be called until both @parent and @bio have
 * completed; the chained bio will also be freed when it completes.
 *
 * The caller must not set bi_private or bi_end_io in @bio.
 */
void bio_chain(struct bio *bio, struct bio *parent)
{
	BUG_ON(bio->bi_private || bio->bi_end_io);

	bio->bi_private = parent;
	bio->bi_end_io	= bio_chain_endio;
345
	bio_inc_remaining(parent);
K
Kent Overstreet 已提交
346 347 348
}
EXPORT_SYMBOL(bio_chain);

349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370
static void bio_alloc_rescue(struct work_struct *work)
{
	struct bio_set *bs = container_of(work, struct bio_set, rescue_work);
	struct bio *bio;

	while (1) {
		spin_lock(&bs->rescue_lock);
		bio = bio_list_pop(&bs->rescue_list);
		spin_unlock(&bs->rescue_lock);

		if (!bio)
			break;

		generic_make_request(bio);
	}
}

static void punt_bios_to_rescuer(struct bio_set *bs)
{
	struct bio_list punt, nopunt;
	struct bio *bio;

371 372
	if (WARN_ON_ONCE(!bs->rescue_workqueue))
		return;
373 374 375 376 377 378 379 380 381 382 383 384 385 386
	/*
	 * In order to guarantee forward progress we must punt only bios that
	 * were allocated from this bio_set; otherwise, if there was a bio on
	 * there for a stacking driver higher up in the stack, processing it
	 * could require allocating bios from this bio_set, and doing that from
	 * our own rescuer would be bad.
	 *
	 * Since bio lists are singly linked, pop them all instead of trying to
	 * remove from the middle of the list:
	 */

	bio_list_init(&punt);
	bio_list_init(&nopunt);

387
	while ((bio = bio_list_pop(&current->bio_list[0])))
388
		bio_list_add(bio->bi_pool == bs ? &punt : &nopunt, bio);
389
	current->bio_list[0] = nopunt;
390

391 392 393 394
	bio_list_init(&nopunt);
	while ((bio = bio_list_pop(&current->bio_list[1])))
		bio_list_add(bio->bi_pool == bs ? &punt : &nopunt, bio);
	current->bio_list[1] = nopunt;
395 396 397 398 399 400 401 402

	spin_lock(&bs->rescue_lock);
	bio_list_merge(&bs->rescue_list, &punt);
	spin_unlock(&bs->rescue_lock);

	queue_work(bs->rescue_workqueue, &bs->rescue_work);
}

L
Linus Torvalds 已提交
403 404
/**
 * bio_alloc_bioset - allocate a bio for I/O
405
 * @gfp_mask:   the GFP_* mask given to the slab allocator
L
Linus Torvalds 已提交
406
 * @nr_iovecs:	number of iovecs to pre-allocate
407
 * @bs:		the bio_set to allocate from.
L
Linus Torvalds 已提交
408 409
 *
 * Description:
410 411 412
 *   If @bs is NULL, uses kmalloc() to allocate the bio; else the allocation is
 *   backed by the @bs's mempool.
 *
413 414 415 416 417 418
 *   When @bs is not NULL, if %__GFP_DIRECT_RECLAIM is set then bio_alloc will
 *   always be able to allocate a bio. This is due to the mempool guarantees.
 *   To make this work, callers must never allocate more than 1 bio at a time
 *   from this pool. Callers that need to allocate more than 1 bio must always
 *   submit the previously allocated bio for IO before attempting to allocate
 *   a new one. Failure to do so can cause deadlocks under memory pressure.
419
 *
420 421 422 423 424 425 426 427 428 429 430 431 432 433 434
 *   Note that when running under generic_make_request() (i.e. any block
 *   driver), bios are not submitted until after you return - see the code in
 *   generic_make_request() that converts recursion into iteration, to prevent
 *   stack overflows.
 *
 *   This would normally mean allocating multiple bios under
 *   generic_make_request() would be susceptible to deadlocks, but we have
 *   deadlock avoidance code that resubmits any blocked bios from a rescuer
 *   thread.
 *
 *   However, we do not guarantee forward progress for allocations from other
 *   mempools. Doing multiple allocations from the same mempool under
 *   generic_make_request() should be avoided - instead, use bio_set's front_pad
 *   for per bio allocations.
 *
435 436 437
 *   RETURNS:
 *   Pointer to new bio on success, NULL on failure.
 */
438 439
struct bio *bio_alloc_bioset(gfp_t gfp_mask, unsigned int nr_iovecs,
			     struct bio_set *bs)
L
Linus Torvalds 已提交
440
{
441
	gfp_t saved_gfp = gfp_mask;
442 443
	unsigned front_pad;
	unsigned inline_vecs;
I
Ingo Molnar 已提交
444
	struct bio_vec *bvl = NULL;
T
Tejun Heo 已提交
445 446 447
	struct bio *bio;
	void *p;

448 449 450 451 452 453 454 455 456 457
	if (!bs) {
		if (nr_iovecs > UIO_MAXIOV)
			return NULL;

		p = kmalloc(sizeof(struct bio) +
			    nr_iovecs * sizeof(struct bio_vec),
			    gfp_mask);
		front_pad = 0;
		inline_vecs = nr_iovecs;
	} else {
J
Junichi Nomura 已提交
458
		/* should not use nobvec bioset for nr_iovecs > 0 */
459 460
		if (WARN_ON_ONCE(!mempool_initialized(&bs->bvec_pool) &&
				 nr_iovecs > 0))
J
Junichi Nomura 已提交
461
			return NULL;
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477
		/*
		 * generic_make_request() converts recursion to iteration; this
		 * means if we're running beneath it, any bios we allocate and
		 * submit will not be submitted (and thus freed) until after we
		 * return.
		 *
		 * This exposes us to a potential deadlock if we allocate
		 * multiple bios from the same bio_set() while running
		 * underneath generic_make_request(). If we were to allocate
		 * multiple bios (say a stacking block driver that was splitting
		 * bios), we would deadlock if we exhausted the mempool's
		 * reserve.
		 *
		 * We solve this, and guarantee forward progress, with a rescuer
		 * workqueue per bio_set. If we go to allocate and there are
		 * bios on current->bio_list, we first try the allocation
478 479 480
		 * without __GFP_DIRECT_RECLAIM; if that fails, we punt those
		 * bios we would be blocking to the rescuer workqueue before
		 * we retry with the original gfp_flags.
481 482
		 */

483 484
		if (current->bio_list &&
		    (!bio_list_empty(&current->bio_list[0]) ||
485 486
		     !bio_list_empty(&current->bio_list[1])) &&
		    bs->rescue_workqueue)
487
			gfp_mask &= ~__GFP_DIRECT_RECLAIM;
488

489
		p = mempool_alloc(&bs->bio_pool, gfp_mask);
490 491 492
		if (!p && gfp_mask != saved_gfp) {
			punt_bios_to_rescuer(bs);
			gfp_mask = saved_gfp;
493
			p = mempool_alloc(&bs->bio_pool, gfp_mask);
494 495
		}

496 497 498 499
		front_pad = bs->front_pad;
		inline_vecs = BIO_INLINE_VECS;
	}

T
Tejun Heo 已提交
500 501
	if (unlikely(!p))
		return NULL;
L
Linus Torvalds 已提交
502

503
	bio = p + front_pad;
504
	bio_init(bio, NULL, 0);
I
Ingo Molnar 已提交
505

506
	if (nr_iovecs > inline_vecs) {
507 508
		unsigned long idx = 0;

509
		bvl = bvec_alloc(gfp_mask, nr_iovecs, &idx, &bs->bvec_pool);
510 511 512
		if (!bvl && gfp_mask != saved_gfp) {
			punt_bios_to_rescuer(bs);
			gfp_mask = saved_gfp;
513
			bvl = bvec_alloc(gfp_mask, nr_iovecs, &idx, &bs->bvec_pool);
514 515
		}

I
Ingo Molnar 已提交
516 517
		if (unlikely(!bvl))
			goto err_free;
518

519
		bio->bi_flags |= idx << BVEC_POOL_OFFSET;
520 521
	} else if (nr_iovecs) {
		bvl = bio->bi_inline_vecs;
L
Linus Torvalds 已提交
522
	}
523 524

	bio->bi_pool = bs;
I
Ingo Molnar 已提交
525 526
	bio->bi_max_vecs = nr_iovecs;
	bio->bi_io_vec = bvl;
L
Linus Torvalds 已提交
527
	return bio;
I
Ingo Molnar 已提交
528 529

err_free:
530
	mempool_free(p, &bs->bio_pool);
I
Ingo Molnar 已提交
531
	return NULL;
L
Linus Torvalds 已提交
532
}
533
EXPORT_SYMBOL(bio_alloc_bioset);
L
Linus Torvalds 已提交
534

535
void zero_fill_bio_iter(struct bio *bio, struct bvec_iter start)
L
Linus Torvalds 已提交
536 537
{
	unsigned long flags;
538 539
	struct bio_vec bv;
	struct bvec_iter iter;
L
Linus Torvalds 已提交
540

541
	__bio_for_each_segment(bv, bio, iter, start) {
542 543 544
		char *data = bvec_kmap_irq(&bv, &flags);
		memset(data, 0, bv.bv_len);
		flush_dcache_page(bv.bv_page);
L
Linus Torvalds 已提交
545 546 547
		bvec_kunmap_irq(data, &flags);
	}
}
548
EXPORT_SYMBOL(zero_fill_bio_iter);
L
Linus Torvalds 已提交
549 550 551 552 553 554 555

/**
 * bio_put - release a reference to a bio
 * @bio:   bio to release reference to
 *
 * Description:
 *   Put a reference to a &struct bio, either one you have gotten with
556
 *   bio_alloc, bio_get or bio_clone_*. The last put of a bio will free it.
L
Linus Torvalds 已提交
557 558 559
 **/
void bio_put(struct bio *bio)
{
560
	if (!bio_flagged(bio, BIO_REFFED))
K
Kent Overstreet 已提交
561
		bio_free(bio);
562 563 564 565 566 567 568 569 570
	else {
		BIO_BUG_ON(!atomic_read(&bio->__bi_cnt));

		/*
		 * last put frees it
		 */
		if (atomic_dec_and_test(&bio->__bi_cnt))
			bio_free(bio);
	}
L
Linus Torvalds 已提交
571
}
572
EXPORT_SYMBOL(bio_put);
L
Linus Torvalds 已提交
573

574
inline int bio_phys_segments(struct request_queue *q, struct bio *bio)
L
Linus Torvalds 已提交
575 576 577 578 579 580
{
	if (unlikely(!bio_flagged(bio, BIO_SEG_VALID)))
		blk_recount_segments(q, bio);

	return bio->bi_phys_segments;
}
581
EXPORT_SYMBOL(bio_phys_segments);
L
Linus Torvalds 已提交
582

K
Kent Overstreet 已提交
583 584 585 586 587 588 589 590 591 592 593 594 595
/**
 * 	__bio_clone_fast - clone a bio that shares the original bio's biovec
 * 	@bio: destination bio
 * 	@bio_src: bio to clone
 *
 *	Clone a &bio. Caller will own the returned bio, but not
 *	the actual data it points to. Reference count of returned
 * 	bio will be one.
 *
 * 	Caller must ensure that @bio_src is not freed before @bio.
 */
void __bio_clone_fast(struct bio *bio, struct bio *bio_src)
{
596
	BUG_ON(bio->bi_pool && BVEC_POOL_IDX(bio));
K
Kent Overstreet 已提交
597 598

	/*
599
	 * most users will be overriding ->bi_disk with a new target,
K
Kent Overstreet 已提交
600 601
	 * so we don't set nor calculate new physical/hw segment counts here
	 */
602
	bio->bi_disk = bio_src->bi_disk;
603
	bio->bi_partno = bio_src->bi_partno;
604
	bio_set_flag(bio, BIO_CLONED);
S
Shaohua Li 已提交
605 606
	if (bio_flagged(bio_src, BIO_THROTTLED))
		bio_set_flag(bio, BIO_THROTTLED);
J
Jens Axboe 已提交
607
	bio->bi_opf = bio_src->bi_opf;
608
	bio->bi_write_hint = bio_src->bi_write_hint;
K
Kent Overstreet 已提交
609 610
	bio->bi_iter = bio_src->bi_iter;
	bio->bi_io_vec = bio_src->bi_io_vec;
611 612

	bio_clone_blkcg_association(bio, bio_src);
K
Kent Overstreet 已提交
613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648
}
EXPORT_SYMBOL(__bio_clone_fast);

/**
 *	bio_clone_fast - clone a bio that shares the original bio's biovec
 *	@bio: bio to clone
 *	@gfp_mask: allocation priority
 *	@bs: bio_set to allocate from
 *
 * 	Like __bio_clone_fast, only also allocates the returned bio
 */
struct bio *bio_clone_fast(struct bio *bio, gfp_t gfp_mask, struct bio_set *bs)
{
	struct bio *b;

	b = bio_alloc_bioset(gfp_mask, 0, bs);
	if (!b)
		return NULL;

	__bio_clone_fast(b, bio);

	if (bio_integrity(bio)) {
		int ret;

		ret = bio_integrity_clone(b, bio, gfp_mask);

		if (ret < 0) {
			bio_put(b);
			return NULL;
		}
	}

	return b;
}
EXPORT_SYMBOL(bio_clone_fast);

649 650 651 652 653 654 655 656 657 658 659
/**
 * 	bio_clone_bioset - clone a bio
 * 	@bio_src: bio to clone
 *	@gfp_mask: allocation priority
 *	@bs: bio_set to allocate from
 *
 *	Clone bio. Caller will own the returned bio, but not the actual data it
 *	points to. Reference count of returned bio will be one.
 */
struct bio *bio_clone_bioset(struct bio *bio_src, gfp_t gfp_mask,
			     struct bio_set *bs)
L
Linus Torvalds 已提交
660
{
661 662 663
	struct bvec_iter iter;
	struct bio_vec bv;
	struct bio *bio;
L
Linus Torvalds 已提交
664

665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686
	/*
	 * Pre immutable biovecs, __bio_clone() used to just do a memcpy from
	 * bio_src->bi_io_vec to bio->bi_io_vec.
	 *
	 * We can't do that anymore, because:
	 *
	 *  - The point of cloning the biovec is to produce a bio with a biovec
	 *    the caller can modify: bi_idx and bi_bvec_done should be 0.
	 *
	 *  - The original bio could've had more than BIO_MAX_PAGES biovecs; if
	 *    we tried to clone the whole thing bio_alloc_bioset() would fail.
	 *    But the clone should succeed as long as the number of biovecs we
	 *    actually need to allocate is fewer than BIO_MAX_PAGES.
	 *
	 *  - Lastly, bi_vcnt should not be looked at or relied upon by code
	 *    that does not own the bio - reason being drivers don't use it for
	 *    iterating over the biovec anymore, so expecting it to be kept up
	 *    to date (i.e. for clones that share the parent biovec) is just
	 *    asking for trouble and would force extra work on
	 *    __bio_clone_fast() anyways.
	 */

687
	bio = bio_alloc_bioset(gfp_mask, bio_segments(bio_src), bs);
688
	if (!bio)
689
		return NULL;
690
	bio->bi_disk		= bio_src->bi_disk;
J
Jens Axboe 已提交
691
	bio->bi_opf		= bio_src->bi_opf;
692
	bio->bi_write_hint	= bio_src->bi_write_hint;
693 694
	bio->bi_iter.bi_sector	= bio_src->bi_iter.bi_sector;
	bio->bi_iter.bi_size	= bio_src->bi_iter.bi_size;
695

A
Adrian Hunter 已提交
696 697 698
	switch (bio_op(bio)) {
	case REQ_OP_DISCARD:
	case REQ_OP_SECURE_ERASE:
699
	case REQ_OP_WRITE_ZEROES:
A
Adrian Hunter 已提交
700 701
		break;
	case REQ_OP_WRITE_SAME:
702
		bio->bi_io_vec[bio->bi_vcnt++] = bio_src->bi_io_vec[0];
A
Adrian Hunter 已提交
703 704
		break;
	default:
705
		bio_for_each_segment(bv, bio_src, iter)
A
Adrian Hunter 已提交
706 707
			bio->bi_io_vec[bio->bi_vcnt++] = bv;
		break;
708 709
	}

710 711
	if (bio_integrity(bio_src)) {
		int ret;
712

713
		ret = bio_integrity_clone(bio, bio_src, gfp_mask);
L
Li Zefan 已提交
714
		if (ret < 0) {
715
			bio_put(bio);
716
			return NULL;
L
Li Zefan 已提交
717
		}
P
Peter Osterlund 已提交
718
	}
L
Linus Torvalds 已提交
719

720 721
	bio_clone_blkcg_association(bio, bio_src);

722
	return bio;
L
Linus Torvalds 已提交
723
}
724
EXPORT_SYMBOL(bio_clone_bioset);
L
Linus Torvalds 已提交
725 726

/**
K
Kent Overstreet 已提交
727 728 729 730 731 732
 *	bio_add_pc_page	-	attempt to add page to bio
 *	@q: the target queue
 *	@bio: destination bio
 *	@page: page to add
 *	@len: vec entry length
 *	@offset: vec entry offset
L
Linus Torvalds 已提交
733
 *
K
Kent Overstreet 已提交
734 735 736 737 738 739
 *	Attempt to add a page to the bio_vec maplist. This can fail for a
 *	number of reasons, such as the bio being full or target block device
 *	limitations. The target block device must allow bio's up to PAGE_SIZE,
 *	so it is always possible to add a single page to an empty bio.
 *
 *	This should only be used by REQ_PC bios.
L
Linus Torvalds 已提交
740
 */
K
Kent Overstreet 已提交
741 742
int bio_add_pc_page(struct request_queue *q, struct bio *bio, struct page
		    *page, unsigned int len, unsigned int offset)
L
Linus Torvalds 已提交
743 744 745 746 747 748 749 750 751 752
{
	int retried_segments = 0;
	struct bio_vec *bvec;

	/*
	 * cloned bio must not modify vec list
	 */
	if (unlikely(bio_flagged(bio, BIO_CLONED)))
		return 0;

K
Kent Overstreet 已提交
753
	if (((bio->bi_iter.bi_size + len) >> 9) > queue_max_hw_sectors(q))
L
Linus Torvalds 已提交
754 755
		return 0;

756 757 758 759 760 761 762 763 764 765 766
	/*
	 * For filesystems with a blocksize smaller than the pagesize
	 * we will often be called with the same page as last time and
	 * a consecutive offset.  Optimize this special case.
	 */
	if (bio->bi_vcnt > 0) {
		struct bio_vec *prev = &bio->bi_io_vec[bio->bi_vcnt - 1];

		if (page == prev->bv_page &&
		    offset == prev->bv_offset + prev->bv_len) {
			prev->bv_len += len;
767
			bio->bi_iter.bi_size += len;
768 769
			goto done;
		}
770 771 772 773 774

		/*
		 * If the queue doesn't support SG gaps and adding this
		 * offset would create a gap, disallow it.
		 */
775
		if (bvec_gap_to_prev(q, prev, offset))
776
			return 0;
777 778
	}

779
	if (bio_full(bio))
L
Linus Torvalds 已提交
780 781 782
		return 0;

	/*
783 784 785 786 787 788 789 790 791 792 793 794 795 796
	 * setup the new entry, we might clear it again later if we
	 * cannot add the page
	 */
	bvec = &bio->bi_io_vec[bio->bi_vcnt];
	bvec->bv_page = page;
	bvec->bv_len = len;
	bvec->bv_offset = offset;
	bio->bi_vcnt++;
	bio->bi_phys_segments++;
	bio->bi_iter.bi_size += len;

	/*
	 * Perform a recount if the number of segments is greater
	 * than queue_max_segments(q).
L
Linus Torvalds 已提交
797 798
	 */

799
	while (bio->bi_phys_segments > queue_max_segments(q)) {
L
Linus Torvalds 已提交
800 801

		if (retried_segments)
802
			goto failed;
L
Linus Torvalds 已提交
803 804 805 806 807 808

		retried_segments = 1;
		blk_recount_segments(q, bio);
	}

	/* If we may be able to merge these biovecs, force a recount */
809
	if (bio->bi_vcnt > 1 && (BIOVEC_PHYS_MERGEABLE(bvec-1, bvec)))
810
		bio_clear_flag(bio, BIO_SEG_VALID);
L
Linus Torvalds 已提交
811

812
 done:
L
Linus Torvalds 已提交
813
	return len;
814 815 816 817 818 819 820 821 822

 failed:
	bvec->bv_page = NULL;
	bvec->bv_len = 0;
	bvec->bv_offset = 0;
	bio->bi_vcnt--;
	bio->bi_iter.bi_size -= len;
	blk_recount_segments(q, bio);
	return 0;
L
Linus Torvalds 已提交
823
}
824
EXPORT_SYMBOL(bio_add_pc_page);
825

L
Linus Torvalds 已提交
826
/**
827 828 829 830 831
 * __bio_try_merge_page - try appending data to an existing bvec.
 * @bio: destination bio
 * @page: page to add
 * @len: length of the data to add
 * @off: offset of the data in @page
L
Linus Torvalds 已提交
832
 *
833 834 835 836 837
 * Try to add the data at @page + @off to the last bvec of @bio.  This is a
 * a useful optimisation for file systems with a block size smaller than the
 * page size.
 *
 * Return %true on success or %false on failure.
L
Linus Torvalds 已提交
838
 */
839 840
bool __bio_try_merge_page(struct bio *bio, struct page *page,
		unsigned int len, unsigned int off)
L
Linus Torvalds 已提交
841
{
K
Kent Overstreet 已提交
842
	if (WARN_ON_ONCE(bio_flagged(bio, BIO_CLONED)))
843
		return false;
844

K
Kent Overstreet 已提交
845
	if (bio->bi_vcnt > 0) {
846
		struct bio_vec *bv = &bio->bi_io_vec[bio->bi_vcnt - 1];
847

848
		if (page == bv->bv_page && off == bv->bv_offset + bv->bv_len) {
K
Kent Overstreet 已提交
849
			bv->bv_len += len;
850 851
			bio->bi_iter.bi_size += len;
			return true;
K
Kent Overstreet 已提交
852 853
		}
	}
854 855 856
	return false;
}
EXPORT_SYMBOL_GPL(__bio_try_merge_page);
K
Kent Overstreet 已提交
857

858 859 860 861 862 863 864 865 866 867 868 869 870 871
/**
 * __bio_add_page - add page to a bio in a new segment
 * @bio: destination bio
 * @page: page to add
 * @len: length of the data to add
 * @off: offset of the data in @page
 *
 * Add the data at @page + @off to @bio as a new bvec.  The caller must ensure
 * that @bio has space for another bvec.
 */
void __bio_add_page(struct bio *bio, struct page *page,
		unsigned int len, unsigned int off)
{
	struct bio_vec *bv = &bio->bi_io_vec[bio->bi_vcnt];
K
Kent Overstreet 已提交
872

873 874 875 876 877 878
	WARN_ON_ONCE(bio_flagged(bio, BIO_CLONED));
	WARN_ON_ONCE(bio_full(bio));

	bv->bv_page = page;
	bv->bv_offset = off;
	bv->bv_len = len;
K
Kent Overstreet 已提交
879 880

	bio->bi_iter.bi_size += len;
881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902
	bio->bi_vcnt++;
}
EXPORT_SYMBOL_GPL(__bio_add_page);

/**
 *	bio_add_page	-	attempt to add page to bio
 *	@bio: destination bio
 *	@page: page to add
 *	@len: vec entry length
 *	@offset: vec entry offset
 *
 *	Attempt to add a page to the bio_vec maplist. This will only fail
 *	if either bio->bi_vcnt == bio->bi_max_vecs or it's a cloned bio.
 */
int bio_add_page(struct bio *bio, struct page *page,
		 unsigned int len, unsigned int offset)
{
	if (!__bio_try_merge_page(bio, page, len, offset)) {
		if (bio_full(bio))
			return 0;
		__bio_add_page(bio, page, len, offset);
	}
K
Kent Overstreet 已提交
903
	return len;
L
Linus Torvalds 已提交
904
}
905
EXPORT_SYMBOL(bio_add_page);
L
Linus Torvalds 已提交
906

907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955
/**
 * bio_iov_iter_get_pages - pin user or kernel pages and add them to a bio
 * @bio: bio to add pages to
 * @iter: iov iterator describing the region to be mapped
 *
 * Pins as many pages from *iter and appends them to @bio's bvec array. The
 * pages will have to be released using put_page() when done.
 */
int bio_iov_iter_get_pages(struct bio *bio, struct iov_iter *iter)
{
	unsigned short nr_pages = bio->bi_max_vecs - bio->bi_vcnt;
	struct bio_vec *bv = bio->bi_io_vec + bio->bi_vcnt;
	struct page **pages = (struct page **)bv;
	size_t offset, diff;
	ssize_t size;

	size = iov_iter_get_pages(iter, pages, LONG_MAX, nr_pages, &offset);
	if (unlikely(size <= 0))
		return size ? size : -EFAULT;
	nr_pages = (size + offset + PAGE_SIZE - 1) / PAGE_SIZE;

	/*
	 * Deep magic below:  We need to walk the pinned pages backwards
	 * because we are abusing the space allocated for the bio_vecs
	 * for the page array.  Because the bio_vecs are larger than the
	 * page pointers by definition this will always work.  But it also
	 * means we can't use bio_add_page, so any changes to it's semantics
	 * need to be reflected here as well.
	 */
	bio->bi_iter.bi_size += size;
	bio->bi_vcnt += nr_pages;

	diff = (nr_pages * PAGE_SIZE - offset) - size;
	while (nr_pages--) {
		bv[nr_pages].bv_page = pages[nr_pages];
		bv[nr_pages].bv_len = PAGE_SIZE;
		bv[nr_pages].bv_offset = 0;
	}

	bv[0].bv_offset += offset;
	bv[0].bv_len -= offset;
	if (diff)
		bv[bio->bi_vcnt - 1].bv_len -= diff;

	iov_iter_advance(iter, size);
	return 0;
}
EXPORT_SYMBOL_GPL(bio_iov_iter_get_pages);

956
static void submit_bio_wait_endio(struct bio *bio)
957
{
958
	complete(bio->bi_private);
959 960 961 962 963 964 965 966
}

/**
 * submit_bio_wait - submit a bio, and wait until it completes
 * @bio: The &struct bio which describes the I/O
 *
 * Simple wrapper around submit_bio(). Returns 0 on success, or the error from
 * bio_endio() on failure.
967 968 969 970
 *
 * WARNING: Unlike to how submit_bio() is usually used, this function does not
 * result in bio reference to be consumed. The caller must drop the reference
 * on his own.
971
 */
972
int submit_bio_wait(struct bio *bio)
973
{
974
	DECLARE_COMPLETION_ONSTACK_MAP(done, bio->bi_disk->lockdep_map);
975

976
	bio->bi_private = &done;
977
	bio->bi_end_io = submit_bio_wait_endio;
J
Jens Axboe 已提交
978
	bio->bi_opf |= REQ_SYNC;
979
	submit_bio(bio);
980
	wait_for_completion_io(&done);
981

982
	return blk_status_to_errno(bio->bi_status);
983 984 985
}
EXPORT_SYMBOL(submit_bio_wait);

K
Kent Overstreet 已提交
986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
/**
 * bio_advance - increment/complete a bio by some number of bytes
 * @bio:	bio to advance
 * @bytes:	number of bytes to complete
 *
 * This updates bi_sector, bi_size and bi_idx; if the number of bytes to
 * complete doesn't align with a bvec boundary, then bv_len and bv_offset will
 * be updated on the last bvec as well.
 *
 * @bio will then represent the remaining, uncompleted portion of the io.
 */
void bio_advance(struct bio *bio, unsigned bytes)
{
	if (bio_integrity(bio))
		bio_integrity_advance(bio, bytes);

K
Kent Overstreet 已提交
1002
	bio_advance_iter(bio, &bio->bi_iter, bytes);
K
Kent Overstreet 已提交
1003 1004 1005
}
EXPORT_SYMBOL(bio_advance);

1006 1007
void bio_copy_data_iter(struct bio *dst, struct bvec_iter *dst_iter,
			struct bio *src, struct bvec_iter *src_iter)
K
Kent Overstreet 已提交
1008
{
1009
	struct bio_vec src_bv, dst_bv;
K
Kent Overstreet 已提交
1010
	void *src_p, *dst_p;
1011
	unsigned bytes;
K
Kent Overstreet 已提交
1012

1013 1014 1015
	while (src_iter->bi_size && dst_iter->bi_size) {
		src_bv = bio_iter_iovec(src, *src_iter);
		dst_bv = bio_iter_iovec(dst, *dst_iter);
1016 1017

		bytes = min(src_bv.bv_len, dst_bv.bv_len);
K
Kent Overstreet 已提交
1018

1019 1020
		src_p = kmap_atomic(src_bv.bv_page);
		dst_p = kmap_atomic(dst_bv.bv_page);
K
Kent Overstreet 已提交
1021

1022 1023
		memcpy(dst_p + dst_bv.bv_offset,
		       src_p + src_bv.bv_offset,
K
Kent Overstreet 已提交
1024 1025 1026 1027 1028
		       bytes);

		kunmap_atomic(dst_p);
		kunmap_atomic(src_p);

1029 1030
		flush_dcache_page(dst_bv.bv_page);

1031 1032
		bio_advance_iter(src, src_iter, bytes);
		bio_advance_iter(dst, dst_iter, bytes);
K
Kent Overstreet 已提交
1033 1034
	}
}
1035 1036 1037
EXPORT_SYMBOL(bio_copy_data_iter);

/**
1038 1039 1040
 * bio_copy_data - copy contents of data buffers from one bio to another
 * @src: source bio
 * @dst: destination bio
1041 1042 1043 1044 1045 1046
 *
 * Stops when it reaches the end of either @src or @dst - that is, copies
 * min(src->bi_size, dst->bi_size) bytes (or the equivalent for lists of bios).
 */
void bio_copy_data(struct bio *dst, struct bio *src)
{
1047 1048 1049 1050
	struct bvec_iter src_iter = src->bi_iter;
	struct bvec_iter dst_iter = dst->bi_iter;

	bio_copy_data_iter(dst, &dst_iter, src, &src_iter);
1051
}
K
Kent Overstreet 已提交
1052 1053
EXPORT_SYMBOL(bio_copy_data);

1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090
/**
 * bio_list_copy_data - copy contents of data buffers from one chain of bios to
 * another
 * @src: source bio list
 * @dst: destination bio list
 *
 * Stops when it reaches the end of either the @src list or @dst list - that is,
 * copies min(src->bi_size, dst->bi_size) bytes (or the equivalent for lists of
 * bios).
 */
void bio_list_copy_data(struct bio *dst, struct bio *src)
{
	struct bvec_iter src_iter = src->bi_iter;
	struct bvec_iter dst_iter = dst->bi_iter;

	while (1) {
		if (!src_iter.bi_size) {
			src = src->bi_next;
			if (!src)
				break;

			src_iter = src->bi_iter;
		}

		if (!dst_iter.bi_size) {
			dst = dst->bi_next;
			if (!dst)
				break;

			dst_iter = dst->bi_iter;
		}

		bio_copy_data_iter(dst, &dst_iter, src, &src_iter);
	}
}
EXPORT_SYMBOL(bio_list_copy_data);

L
Linus Torvalds 已提交
1091
struct bio_map_data {
1092
	int is_our_pages;
1093 1094
	struct iov_iter iter;
	struct iovec iov[];
L
Linus Torvalds 已提交
1095 1096
};

1097
static struct bio_map_data *bio_alloc_map_data(struct iov_iter *data,
1098
					       gfp_t gfp_mask)
L
Linus Torvalds 已提交
1099
{
1100 1101
	struct bio_map_data *bmd;
	if (data->nr_segs > UIO_MAXIOV)
1102
		return NULL;
L
Linus Torvalds 已提交
1103

1104 1105 1106 1107 1108 1109 1110 1111
	bmd = kmalloc(sizeof(struct bio_map_data) +
		       sizeof(struct iovec) * data->nr_segs, gfp_mask);
	if (!bmd)
		return NULL;
	memcpy(bmd->iov, data->iov, sizeof(struct iovec) * data->nr_segs);
	bmd->iter = *data;
	bmd->iter.iov = bmd->iov;
	return bmd;
L
Linus Torvalds 已提交
1112 1113
}

1114 1115 1116 1117 1118 1119 1120 1121
/**
 * bio_copy_from_iter - copy all pages from iov_iter to bio
 * @bio: The &struct bio which describes the I/O as destination
 * @iter: iov_iter as source
 *
 * Copy all pages from iov_iter to bio.
 * Returns 0 on success, or error on failure.
 */
1122
static int bio_copy_from_iter(struct bio *bio, struct iov_iter *iter)
1123
{
1124
	int i;
1125 1126
	struct bio_vec *bvec;

1127
	bio_for_each_segment_all(bvec, bio, i) {
1128
		ssize_t ret;
1129

1130 1131 1132
		ret = copy_page_from_iter(bvec->bv_page,
					  bvec->bv_offset,
					  bvec->bv_len,
1133
					  iter);
1134

1135
		if (!iov_iter_count(iter))
1136 1137 1138 1139
			break;

		if (ret < bvec->bv_len)
			return -EFAULT;
1140 1141
	}

1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173
	return 0;
}

/**
 * bio_copy_to_iter - copy all pages from bio to iov_iter
 * @bio: The &struct bio which describes the I/O as source
 * @iter: iov_iter as destination
 *
 * Copy all pages from bio to iov_iter.
 * Returns 0 on success, or error on failure.
 */
static int bio_copy_to_iter(struct bio *bio, struct iov_iter iter)
{
	int i;
	struct bio_vec *bvec;

	bio_for_each_segment_all(bvec, bio, i) {
		ssize_t ret;

		ret = copy_page_to_iter(bvec->bv_page,
					bvec->bv_offset,
					bvec->bv_len,
					&iter);

		if (!iov_iter_count(&iter))
			break;

		if (ret < bvec->bv_len)
			return -EFAULT;
	}

	return 0;
1174 1175
}

1176
void bio_free_pages(struct bio *bio)
1177 1178 1179 1180 1181 1182 1183
{
	struct bio_vec *bvec;
	int i;

	bio_for_each_segment_all(bvec, bio, i)
		__free_page(bvec->bv_page);
}
1184
EXPORT_SYMBOL(bio_free_pages);
1185

L
Linus Torvalds 已提交
1186 1187 1188 1189
/**
 *	bio_uncopy_user	-	finish previously mapped bio
 *	@bio: bio being terminated
 *
1190
 *	Free pages allocated from bio_copy_user_iov() and write back data
L
Linus Torvalds 已提交
1191 1192 1193 1194 1195
 *	to user space in case of a read.
 */
int bio_uncopy_user(struct bio *bio)
{
	struct bio_map_data *bmd = bio->bi_private;
1196
	int ret = 0;
L
Linus Torvalds 已提交
1197

1198 1199 1200
	if (!bio_flagged(bio, BIO_NULL_MAPPED)) {
		/*
		 * if we're in a workqueue, the request is orphaned, so
1201 1202
		 * don't copy into a random user address space, just free
		 * and return -EINTR so user space doesn't expect any data.
1203
		 */
1204 1205 1206
		if (!current->mm)
			ret = -EINTR;
		else if (bio_data_dir(bio) == READ)
1207
			ret = bio_copy_to_iter(bio, bmd->iter);
1208 1209
		if (bmd->is_our_pages)
			bio_free_pages(bio);
1210
	}
1211
	kfree(bmd);
L
Linus Torvalds 已提交
1212 1213 1214 1215 1216
	bio_put(bio);
	return ret;
}

/**
1217
 *	bio_copy_user_iov	-	copy user data to bio
1218 1219 1220 1221
 *	@q:		destination block queue
 *	@map_data:	pointer to the rq_map_data holding pages (if necessary)
 *	@iter:		iovec iterator
 *	@gfp_mask:	memory allocation flags
L
Linus Torvalds 已提交
1222 1223 1224 1225 1226
 *
 *	Prepares and returns a bio for indirect user io, bouncing data
 *	to/from kernel pages as necessary. Must be paired with
 *	call bio_uncopy_user() on io completion.
 */
1227 1228
struct bio *bio_copy_user_iov(struct request_queue *q,
			      struct rq_map_data *map_data,
1229
			      struct iov_iter *iter,
1230
			      gfp_t gfp_mask)
L
Linus Torvalds 已提交
1231 1232 1233 1234
{
	struct bio_map_data *bmd;
	struct page *page;
	struct bio *bio;
1235 1236
	int i = 0, ret;
	int nr_pages;
1237
	unsigned int len = iter->count;
G
Geliang Tang 已提交
1238
	unsigned int offset = map_data ? offset_in_page(map_data->offset) : 0;
L
Linus Torvalds 已提交
1239

1240
	bmd = bio_alloc_map_data(iter, gfp_mask);
L
Linus Torvalds 已提交
1241 1242 1243
	if (!bmd)
		return ERR_PTR(-ENOMEM);

1244 1245 1246 1247 1248 1249 1250
	/*
	 * We need to do a deep copy of the iov_iter including the iovecs.
	 * The caller provided iov might point to an on-stack or otherwise
	 * shortlived one.
	 */
	bmd->is_our_pages = map_data ? 0 : 1;

1251 1252 1253
	nr_pages = DIV_ROUND_UP(offset + len, PAGE_SIZE);
	if (nr_pages > BIO_MAX_PAGES)
		nr_pages = BIO_MAX_PAGES;
1254

L
Linus Torvalds 已提交
1255
	ret = -ENOMEM;
1256
	bio = bio_kmalloc(gfp_mask, nr_pages);
L
Linus Torvalds 已提交
1257 1258 1259 1260
	if (!bio)
		goto out_bmd;

	ret = 0;
1261 1262

	if (map_data) {
1263
		nr_pages = 1 << map_data->page_order;
1264 1265
		i = map_data->offset / PAGE_SIZE;
	}
L
Linus Torvalds 已提交
1266
	while (len) {
1267
		unsigned int bytes = PAGE_SIZE;
L
Linus Torvalds 已提交
1268

1269 1270
		bytes -= offset;

L
Linus Torvalds 已提交
1271 1272 1273
		if (bytes > len)
			bytes = len;

1274
		if (map_data) {
1275
			if (i == map_data->nr_entries * nr_pages) {
1276 1277 1278
				ret = -ENOMEM;
				break;
			}
1279 1280 1281 1282 1283 1284

			page = map_data->pages[i / nr_pages];
			page += (i % nr_pages);

			i++;
		} else {
1285
			page = alloc_page(q->bounce_gfp | gfp_mask);
1286 1287 1288 1289
			if (!page) {
				ret = -ENOMEM;
				break;
			}
L
Linus Torvalds 已提交
1290 1291
		}

1292
		if (bio_add_pc_page(q, bio, page, bytes, offset) < bytes)
L
Linus Torvalds 已提交
1293 1294 1295
			break;

		len -= bytes;
1296
		offset = 0;
L
Linus Torvalds 已提交
1297 1298 1299 1300 1301
	}

	if (ret)
		goto cleanup;

1302 1303 1304
	if (map_data)
		map_data->offset += bio->bi_iter.bi_size;

L
Linus Torvalds 已提交
1305 1306 1307
	/*
	 * success
	 */
1308
	if (((iter->type & WRITE) && (!map_data || !map_data->null_mapped)) ||
1309
	    (map_data && map_data->from_user)) {
1310
		ret = bio_copy_from_iter(bio, iter);
1311 1312
		if (ret)
			goto cleanup;
1313 1314
	} else {
		iov_iter_advance(iter, bio->bi_iter.bi_size);
L
Linus Torvalds 已提交
1315 1316
	}

1317
	bio->bi_private = bmd;
1318 1319
	if (map_data && map_data->null_mapped)
		bio_set_flag(bio, BIO_NULL_MAPPED);
L
Linus Torvalds 已提交
1320 1321
	return bio;
cleanup:
1322
	if (!map_data)
1323
		bio_free_pages(bio);
L
Linus Torvalds 已提交
1324 1325
	bio_put(bio);
out_bmd:
1326
	kfree(bmd);
L
Linus Torvalds 已提交
1327 1328 1329
	return ERR_PTR(ret);
}

1330 1331 1332 1333 1334 1335 1336 1337 1338 1339
/**
 *	bio_map_user_iov - map user iovec into bio
 *	@q:		the struct request_queue for the bio
 *	@iter:		iovec iterator
 *	@gfp_mask:	memory allocation flags
 *
 *	Map the user space address into a bio suitable for io to a block
 *	device. Returns an error pointer in case of error.
 */
struct bio *bio_map_user_iov(struct request_queue *q,
1340
			     struct iov_iter *iter,
1341
			     gfp_t gfp_mask)
L
Linus Torvalds 已提交
1342
{
1343
	int j;
L
Linus Torvalds 已提交
1344
	struct bio *bio;
1345
	int ret;
A
Al Viro 已提交
1346
	struct bio_vec *bvec;
L
Linus Torvalds 已提交
1347

1348
	if (!iov_iter_count(iter))
L
Linus Torvalds 已提交
1349 1350
		return ERR_PTR(-EINVAL);

1351
	bio = bio_kmalloc(gfp_mask, iov_iter_npages(iter, BIO_MAX_PAGES));
L
Linus Torvalds 已提交
1352 1353 1354
	if (!bio)
		return ERR_PTR(-ENOMEM);

1355
	while (iov_iter_count(iter)) {
1356
		struct page **pages;
1357 1358 1359
		ssize_t bytes;
		size_t offs, added = 0;
		int npages;
L
Linus Torvalds 已提交
1360

1361
		bytes = iov_iter_get_pages_alloc(iter, &pages, LONG_MAX, &offs);
1362 1363
		if (unlikely(bytes <= 0)) {
			ret = bytes ? bytes : -EFAULT;
1364
			goto out_unmap;
1365
		}
1366

1367
		npages = DIV_ROUND_UP(offs + bytes, PAGE_SIZE);
1368

1369 1370 1371 1372 1373 1374 1375 1376
		if (unlikely(offs & queue_dma_alignment(q))) {
			ret = -EINVAL;
			j = 0;
		} else {
			for (j = 0; j < npages; j++) {
				struct page *page = pages[j];
				unsigned int n = PAGE_SIZE - offs;
				unsigned short prev_bi_vcnt = bio->bi_vcnt;
1377

1378 1379
				if (n > bytes)
					n = bytes;
1380

1381 1382
				if (!bio_add_pc_page(q, bio, page, n, offs))
					break;
L
Linus Torvalds 已提交
1383

1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394
				/*
				 * check if vector was merged with previous
				 * drop page reference if needed
				 */
				if (bio->bi_vcnt == prev_bi_vcnt)
					put_page(page);

				added += n;
				bytes -= n;
				offs = 0;
			}
1395
			iov_iter_advance(iter, added);
1396
		}
L
Linus Torvalds 已提交
1397
		/*
1398
		 * release the pages we didn't map into the bio, if any
L
Linus Torvalds 已提交
1399
		 */
1400
		while (j < npages)
1401
			put_page(pages[j++]);
1402
		kvfree(pages);
1403 1404 1405
		/* couldn't stuff something into bio? */
		if (bytes)
			break;
L
Linus Torvalds 已提交
1406 1407
	}

1408
	bio_set_flag(bio, BIO_USER_MAPPED);
1409 1410

	/*
1411
	 * subtle -- if bio_map_user_iov() ended up bouncing a bio,
1412 1413 1414 1415 1416
	 * it would normally disappear when its bi_end_io is run.
	 * however, we need it for the unmap, so grab an extra
	 * reference to it
	 */
	bio_get(bio);
L
Linus Torvalds 已提交
1417
	return bio;
1418 1419

 out_unmap:
A
Al Viro 已提交
1420 1421
	bio_for_each_segment_all(bvec, bio, j) {
		put_page(bvec->bv_page);
1422
	}
L
Linus Torvalds 已提交
1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434
	bio_put(bio);
	return ERR_PTR(ret);
}

static void __bio_unmap_user(struct bio *bio)
{
	struct bio_vec *bvec;
	int i;

	/*
	 * make sure we dirty pages we wrote to
	 */
1435
	bio_for_each_segment_all(bvec, bio, i) {
L
Linus Torvalds 已提交
1436 1437 1438
		if (bio_data_dir(bio) == READ)
			set_page_dirty_lock(bvec->bv_page);

1439
		put_page(bvec->bv_page);
L
Linus Torvalds 已提交
1440 1441 1442 1443 1444 1445 1446 1447 1448
	}

	bio_put(bio);
}

/**
 *	bio_unmap_user	-	unmap a bio
 *	@bio:		the bio being unmapped
 *
1449 1450
 *	Unmap a bio previously mapped by bio_map_user_iov(). Must be called from
 *	process context.
L
Linus Torvalds 已提交
1451 1452 1453 1454 1455 1456 1457 1458 1459
 *
 *	bio_unmap_user() may sleep.
 */
void bio_unmap_user(struct bio *bio)
{
	__bio_unmap_user(bio);
	bio_put(bio);
}

1460
static void bio_map_kern_endio(struct bio *bio)
1461 1462 1463 1464
{
	bio_put(bio);
}

1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476
/**
 *	bio_map_kern	-	map kernel address into bio
 *	@q: the struct request_queue for the bio
 *	@data: pointer to buffer to map
 *	@len: length in bytes
 *	@gfp_mask: allocation flags for bio allocation
 *
 *	Map the kernel address into a bio suitable for io to a block
 *	device. Returns an error pointer in case of error.
 */
struct bio *bio_map_kern(struct request_queue *q, void *data, unsigned int len,
			 gfp_t gfp_mask)
M
Mike Christie 已提交
1477 1478 1479 1480 1481 1482 1483 1484
{
	unsigned long kaddr = (unsigned long)data;
	unsigned long end = (kaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
	unsigned long start = kaddr >> PAGE_SHIFT;
	const int nr_pages = end - start;
	int offset, i;
	struct bio *bio;

1485
	bio = bio_kmalloc(gfp_mask, nr_pages);
M
Mike Christie 已提交
1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
	if (!bio)
		return ERR_PTR(-ENOMEM);

	offset = offset_in_page(kaddr);
	for (i = 0; i < nr_pages; i++) {
		unsigned int bytes = PAGE_SIZE - offset;

		if (len <= 0)
			break;

		if (bytes > len)
			bytes = len;

1499
		if (bio_add_pc_page(q, bio, virt_to_page(data), bytes,
1500 1501 1502 1503 1504
				    offset) < bytes) {
			/* we don't support partial mappings */
			bio_put(bio);
			return ERR_PTR(-EINVAL);
		}
M
Mike Christie 已提交
1505 1506 1507 1508 1509 1510

		data += bytes;
		len -= bytes;
		offset = 0;
	}

1511
	bio->bi_end_io = bio_map_kern_endio;
M
Mike Christie 已提交
1512 1513
	return bio;
}
1514
EXPORT_SYMBOL(bio_map_kern);
M
Mike Christie 已提交
1515

1516
static void bio_copy_kern_endio(struct bio *bio)
1517
{
1518 1519 1520 1521
	bio_free_pages(bio);
	bio_put(bio);
}

1522
static void bio_copy_kern_endio_read(struct bio *bio)
1523
{
C
Christoph Hellwig 已提交
1524
	char *p = bio->bi_private;
1525
	struct bio_vec *bvec;
1526 1527
	int i;

1528
	bio_for_each_segment_all(bvec, bio, i) {
1529
		memcpy(p, page_address(bvec->bv_page), bvec->bv_len);
1530
		p += bvec->bv_len;
1531 1532
	}

1533
	bio_copy_kern_endio(bio);
1534 1535 1536 1537 1538 1539 1540 1541
}

/**
 *	bio_copy_kern	-	copy kernel address into bio
 *	@q: the struct request_queue for the bio
 *	@data: pointer to buffer to copy
 *	@len: length in bytes
 *	@gfp_mask: allocation flags for bio and page allocation
1542
 *	@reading: data direction is READ
1543 1544 1545 1546 1547 1548 1549
 *
 *	copy the kernel address into a bio suitable for io to a block
 *	device. Returns an error pointer in case of error.
 */
struct bio *bio_copy_kern(struct request_queue *q, void *data, unsigned int len,
			  gfp_t gfp_mask, int reading)
{
C
Christoph Hellwig 已提交
1550 1551 1552 1553 1554
	unsigned long kaddr = (unsigned long)data;
	unsigned long end = (kaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
	unsigned long start = kaddr >> PAGE_SHIFT;
	struct bio *bio;
	void *p = data;
1555
	int nr_pages = 0;
1556

C
Christoph Hellwig 已提交
1557 1558 1559 1560 1561
	/*
	 * Overflow, abort
	 */
	if (end < start)
		return ERR_PTR(-EINVAL);
1562

C
Christoph Hellwig 已提交
1563 1564 1565 1566
	nr_pages = end - start;
	bio = bio_kmalloc(gfp_mask, nr_pages);
	if (!bio)
		return ERR_PTR(-ENOMEM);
1567

C
Christoph Hellwig 已提交
1568 1569 1570
	while (len) {
		struct page *page;
		unsigned int bytes = PAGE_SIZE;
1571

C
Christoph Hellwig 已提交
1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586
		if (bytes > len)
			bytes = len;

		page = alloc_page(q->bounce_gfp | gfp_mask);
		if (!page)
			goto cleanup;

		if (!reading)
			memcpy(page_address(page), p, bytes);

		if (bio_add_pc_page(q, bio, page, bytes, 0) < bytes)
			break;

		len -= bytes;
		p += bytes;
1587 1588
	}

1589 1590 1591 1592 1593 1594
	if (reading) {
		bio->bi_end_io = bio_copy_kern_endio_read;
		bio->bi_private = data;
	} else {
		bio->bi_end_io = bio_copy_kern_endio;
	}
1595

1596
	return bio;
C
Christoph Hellwig 已提交
1597 1598

cleanup:
1599
	bio_free_pages(bio);
C
Christoph Hellwig 已提交
1600 1601
	bio_put(bio);
	return ERR_PTR(-ENOMEM);
1602 1603
}

L
Linus Torvalds 已提交
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622
/*
 * bio_set_pages_dirty() and bio_check_pages_dirty() are support functions
 * for performing direct-IO in BIOs.
 *
 * The problem is that we cannot run set_page_dirty() from interrupt context
 * because the required locks are not interrupt-safe.  So what we can do is to
 * mark the pages dirty _before_ performing IO.  And in interrupt context,
 * check that the pages are still dirty.   If so, fine.  If not, redirty them
 * in process context.
 *
 * We special-case compound pages here: normally this means reads into hugetlb
 * pages.  The logic in here doesn't really work right for compound pages
 * because the VM does not uniformly chase down the head page in all cases.
 * But dirtiness of compound pages is pretty meaningless anyway: the VM doesn't
 * handle them at all.  So we skip compound pages here at an early stage.
 *
 * Note that this code is very hard to test under normal circumstances because
 * direct-io pins the pages with get_user_pages().  This makes
 * is_page_cache_freeable return false, and the VM will not clean the pages.
1623
 * But other code (eg, flusher threads) could clean the pages if they are mapped
L
Linus Torvalds 已提交
1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634
 * pagecache.
 *
 * Simply disabling the call to bio_set_pages_dirty() is a good way to test the
 * deferred bio dirtying paths.
 */

/*
 * bio_set_pages_dirty() will mark all the bio's pages as dirty.
 */
void bio_set_pages_dirty(struct bio *bio)
{
1635
	struct bio_vec *bvec;
L
Linus Torvalds 已提交
1636 1637
	int i;

1638 1639
	bio_for_each_segment_all(bvec, bio, i) {
		struct page *page = bvec->bv_page;
L
Linus Torvalds 已提交
1640 1641 1642 1643 1644

		if (page && !PageCompound(page))
			set_page_dirty_lock(page);
	}
}
1645
EXPORT_SYMBOL_GPL(bio_set_pages_dirty);
L
Linus Torvalds 已提交
1646

1647
static void bio_release_pages(struct bio *bio)
L
Linus Torvalds 已提交
1648
{
1649
	struct bio_vec *bvec;
L
Linus Torvalds 已提交
1650 1651
	int i;

1652 1653
	bio_for_each_segment_all(bvec, bio, i) {
		struct page *page = bvec->bv_page;
L
Linus Torvalds 已提交
1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666

		if (page)
			put_page(page);
	}
}

/*
 * bio_check_pages_dirty() will check that all the BIO's pages are still dirty.
 * If they are, then fine.  If, however, some pages are clean then they must
 * have been written out during the direct-IO read.  So we take another ref on
 * the BIO and the offending pages and re-dirty the pages in process context.
 *
 * It is expected that bio_check_pages_dirty() will wholly own the BIO from
1667 1668
 * here on.  It will run one put_page() against each page and will run one
 * bio_put() against the BIO.
L
Linus Torvalds 已提交
1669 1670
 */

1671
static void bio_dirty_fn(struct work_struct *work);
L
Linus Torvalds 已提交
1672

1673
static DECLARE_WORK(bio_dirty_work, bio_dirty_fn);
L
Linus Torvalds 已提交
1674 1675 1676 1677 1678 1679
static DEFINE_SPINLOCK(bio_dirty_lock);
static struct bio *bio_dirty_list;

/*
 * This runs in process context
 */
1680
static void bio_dirty_fn(struct work_struct *work)
L
Linus Torvalds 已提交
1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701
{
	unsigned long flags;
	struct bio *bio;

	spin_lock_irqsave(&bio_dirty_lock, flags);
	bio = bio_dirty_list;
	bio_dirty_list = NULL;
	spin_unlock_irqrestore(&bio_dirty_lock, flags);

	while (bio) {
		struct bio *next = bio->bi_private;

		bio_set_pages_dirty(bio);
		bio_release_pages(bio);
		bio_put(bio);
		bio = next;
	}
}

void bio_check_pages_dirty(struct bio *bio)
{
1702
	struct bio_vec *bvec;
L
Linus Torvalds 已提交
1703 1704 1705
	int nr_clean_pages = 0;
	int i;

1706 1707
	bio_for_each_segment_all(bvec, bio, i) {
		struct page *page = bvec->bv_page;
L
Linus Torvalds 已提交
1708 1709

		if (PageDirty(page) || PageCompound(page)) {
1710
			put_page(page);
1711
			bvec->bv_page = NULL;
L
Linus Torvalds 已提交
1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728
		} else {
			nr_clean_pages++;
		}
	}

	if (nr_clean_pages) {
		unsigned long flags;

		spin_lock_irqsave(&bio_dirty_lock, flags);
		bio->bi_private = bio_dirty_list;
		bio_dirty_list = bio;
		spin_unlock_irqrestore(&bio_dirty_lock, flags);
		schedule_work(&bio_dirty_work);
	} else {
		bio_put(bio);
	}
}
1729
EXPORT_SYMBOL_GPL(bio_check_pages_dirty);
L
Linus Torvalds 已提交
1730

1731 1732
void generic_start_io_acct(struct request_queue *q, int rw,
			   unsigned long sectors, struct hd_struct *part)
1733 1734 1735
{
	int cpu = part_stat_lock();

1736
	part_round_stats(q, cpu, part);
1737 1738
	part_stat_inc(cpu, part, ios[rw]);
	part_stat_add(cpu, part, sectors[rw], sectors);
1739
	part_inc_in_flight(q, part, rw);
1740 1741 1742 1743 1744

	part_stat_unlock();
}
EXPORT_SYMBOL(generic_start_io_acct);

1745 1746
void generic_end_io_acct(struct request_queue *q, int rw,
			 struct hd_struct *part, unsigned long start_time)
1747 1748 1749 1750 1751
{
	unsigned long duration = jiffies - start_time;
	int cpu = part_stat_lock();

	part_stat_add(cpu, part, ticks[rw], duration);
1752 1753
	part_round_stats(q, cpu, part);
	part_dec_in_flight(q, part, rw);
1754 1755 1756 1757 1758

	part_stat_unlock();
}
EXPORT_SYMBOL(generic_end_io_acct);

1759 1760 1761
#if ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE
void bio_flush_dcache_pages(struct bio *bi)
{
1762 1763
	struct bio_vec bvec;
	struct bvec_iter iter;
1764

1765 1766
	bio_for_each_segment(bvec, bi, iter)
		flush_dcache_page(bvec.bv_page);
1767 1768 1769 1770
}
EXPORT_SYMBOL(bio_flush_dcache_pages);
#endif

1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781
static inline bool bio_remaining_done(struct bio *bio)
{
	/*
	 * If we're not chaining, then ->__bi_remaining is always 1 and
	 * we always end io on the first invocation.
	 */
	if (!bio_flagged(bio, BIO_CHAIN))
		return true;

	BUG_ON(atomic_read(&bio->__bi_remaining) <= 0);

1782
	if (atomic_dec_and_test(&bio->__bi_remaining)) {
1783
		bio_clear_flag(bio, BIO_CHAIN);
1784
		return true;
1785
	}
1786 1787 1788 1789

	return false;
}

L
Linus Torvalds 已提交
1790 1791 1792 1793 1794
/**
 * bio_endio - end I/O on a bio
 * @bio:	bio
 *
 * Description:
1795 1796 1797
 *   bio_endio() will end I/O on the whole bio. bio_endio() is the preferred
 *   way to end I/O on a bio. No one should call bi_end_io() directly on a
 *   bio unless they own it and thus know that it has an end_io function.
N
NeilBrown 已提交
1798 1799 1800 1801 1802
 *
 *   bio_endio() can be called several times on a bio that has been chained
 *   using bio_chain().  The ->bi_end_io() function will only be called the
 *   last time.  At this point the BLK_TA_COMPLETE tracing event will be
 *   generated if BIO_TRACE_COMPLETION is set.
L
Linus Torvalds 已提交
1803
 **/
1804
void bio_endio(struct bio *bio)
L
Linus Torvalds 已提交
1805
{
C
Christoph Hellwig 已提交
1806
again:
1807
	if (!bio_remaining_done(bio))
C
Christoph Hellwig 已提交
1808
		return;
1809 1810
	if (!bio_integrity_endio(bio))
		return;
L
Linus Torvalds 已提交
1811

J
Josef Bacik 已提交
1812 1813 1814
	if (bio->bi_disk)
		rq_qos_done_bio(bio->bi_disk->queue, bio);

C
Christoph Hellwig 已提交
1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825
	/*
	 * Need to have a real endio function for chained bios, otherwise
	 * various corner cases will break (like stacking block devices that
	 * save/restore bi_end_io) - however, we want to avoid unbounded
	 * recursion and blowing the stack. Tail call optimization would
	 * handle this, but compiling with frame pointers also disables
	 * gcc's sibling call optimization.
	 */
	if (bio->bi_end_io == bio_chain_endio) {
		bio = __bio_chain_endio(bio);
		goto again;
K
Kent Overstreet 已提交
1826
	}
C
Christoph Hellwig 已提交
1827

1828 1829
	if (bio->bi_disk && bio_flagged(bio, BIO_TRACE_COMPLETION)) {
		trace_block_bio_complete(bio->bi_disk->queue, bio,
1830
					 blk_status_to_errno(bio->bi_status));
N
NeilBrown 已提交
1831 1832 1833
		bio_clear_flag(bio, BIO_TRACE_COMPLETION);
	}

1834
	blk_throtl_bio_endio(bio);
S
Shaohua Li 已提交
1835 1836
	/* release cgroup info */
	bio_uninit(bio);
C
Christoph Hellwig 已提交
1837 1838
	if (bio->bi_end_io)
		bio->bi_end_io(bio);
L
Linus Torvalds 已提交
1839
}
1840
EXPORT_SYMBOL(bio_endio);
L
Linus Torvalds 已提交
1841

K
Kent Overstreet 已提交
1842 1843 1844 1845 1846 1847 1848 1849 1850 1851
/**
 * bio_split - split a bio
 * @bio:	bio to split
 * @sectors:	number of sectors to split from the front of @bio
 * @gfp:	gfp mask
 * @bs:		bio set to allocate from
 *
 * Allocates and returns a new bio which represents @sectors from the start of
 * @bio, and updates @bio to represent the remaining sectors.
 *
1852 1853 1854
 * Unless this is a discard request the newly allocated bio will point
 * to @bio's bi_io_vec; it is the caller's responsibility to ensure that
 * @bio is not freed before the split.
K
Kent Overstreet 已提交
1855 1856 1857 1858
 */
struct bio *bio_split(struct bio *bio, int sectors,
		      gfp_t gfp, struct bio_set *bs)
{
1859
	struct bio *split;
K
Kent Overstreet 已提交
1860 1861 1862 1863

	BUG_ON(sectors <= 0);
	BUG_ON(sectors >= bio_sectors(bio));

1864
	split = bio_clone_fast(bio, gfp, bs);
K
Kent Overstreet 已提交
1865 1866 1867 1868 1869 1870
	if (!split)
		return NULL;

	split->bi_iter.bi_size = sectors << 9;

	if (bio_integrity(split))
1871
		bio_integrity_trim(split);
K
Kent Overstreet 已提交
1872 1873 1874

	bio_advance(bio, split->bi_iter.bi_size);

N
NeilBrown 已提交
1875
	if (bio_flagged(bio, BIO_TRACE_COMPLETION))
1876
		bio_set_flag(split, BIO_TRACE_COMPLETION);
N
NeilBrown 已提交
1877

K
Kent Overstreet 已提交
1878 1879 1880 1881
	return split;
}
EXPORT_SYMBOL(bio_split);

1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894
/**
 * bio_trim - trim a bio
 * @bio:	bio to trim
 * @offset:	number of sectors to trim from the front of @bio
 * @size:	size we want to trim @bio to, in sectors
 */
void bio_trim(struct bio *bio, int offset, int size)
{
	/* 'bio' is a cloned bio which we need to trim to match
	 * the given offset and size.
	 */

	size <<= 9;
1895
	if (offset == 0 && size == bio->bi_iter.bi_size)
1896 1897
		return;

1898
	bio_clear_flag(bio, BIO_SEG_VALID);
1899 1900 1901

	bio_advance(bio, offset << 9);

1902
	bio->bi_iter.bi_size = size;
1903 1904

	if (bio_integrity(bio))
1905
		bio_integrity_trim(bio);
1906

1907 1908 1909
}
EXPORT_SYMBOL_GPL(bio_trim);

L
Linus Torvalds 已提交
1910 1911 1912 1913
/*
 * create memory pools for biovec's in a bio_set.
 * use the global biovec slabs created for general use.
 */
1914
int biovec_init_pool(mempool_t *pool, int pool_entries)
L
Linus Torvalds 已提交
1915
{
1916
	struct biovec_slab *bp = bvec_slabs + BVEC_POOL_MAX;
L
Linus Torvalds 已提交
1917

1918
	return mempool_init_slab_pool(pool, pool_entries, bp->slab);
L
Linus Torvalds 已提交
1919 1920
}

1921 1922 1923 1924 1925 1926 1927
/*
 * bioset_exit - exit a bioset initialized with bioset_init()
 *
 * May be called on a zeroed but uninitialized bioset (i.e. allocated with
 * kzalloc()).
 */
void bioset_exit(struct bio_set *bs)
L
Linus Torvalds 已提交
1928
{
1929 1930
	if (bs->rescue_workqueue)
		destroy_workqueue(bs->rescue_workqueue);
1931
	bs->rescue_workqueue = NULL;
1932

1933 1934
	mempool_exit(&bs->bio_pool);
	mempool_exit(&bs->bvec_pool);
1935

1936
	bioset_integrity_free(bs);
1937 1938 1939 1940 1941
	if (bs->bio_slab)
		bio_put_slab(bs);
	bs->bio_slab = NULL;
}
EXPORT_SYMBOL(bioset_exit);
L
Linus Torvalds 已提交
1942

1943 1944
/**
 * bioset_init - Initialize a bio_set
K
Kent Overstreet 已提交
1945
 * @bs:		pool to initialize
1946 1947 1948 1949 1950
 * @pool_size:	Number of bio and bio_vecs to cache in the mempool
 * @front_pad:	Number of bytes to allocate in front of the returned bio
 * @flags:	Flags to modify behavior, currently %BIOSET_NEED_BVECS
 *              and %BIOSET_NEED_RESCUER
 *
K
Kent Overstreet 已提交
1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962
 * Description:
 *    Set up a bio_set to be used with @bio_alloc_bioset. Allows the caller
 *    to ask for a number of bytes to be allocated in front of the bio.
 *    Front pad allocation is useful for embedding the bio inside
 *    another structure, to avoid allocating extra data to go with the bio.
 *    Note that the bio must be embedded at the END of that structure always,
 *    or things will break badly.
 *    If %BIOSET_NEED_BVECS is set in @flags, a separate pool will be allocated
 *    for allocating iovecs.  This pool is not needed e.g. for bio_clone_fast().
 *    If %BIOSET_NEED_RESCUER is set, a workqueue is created which can be used to
 *    dispatch queued requests when the mempool runs out of space.
 *
1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
 */
int bioset_init(struct bio_set *bs,
		unsigned int pool_size,
		unsigned int front_pad,
		int flags)
{
	unsigned int back_pad = BIO_INLINE_VECS * sizeof(struct bio_vec);

	bs->front_pad = front_pad;

	spin_lock_init(&bs->rescue_lock);
	bio_list_init(&bs->rescue_list);
	INIT_WORK(&bs->rescue_work, bio_alloc_rescue);

	bs->bio_slab = bio_find_or_create_slab(front_pad + back_pad);
	if (!bs->bio_slab)
		return -ENOMEM;

	if (mempool_init_slab_pool(&bs->bio_pool, pool_size, bs->bio_slab))
		goto bad;

	if ((flags & BIOSET_NEED_BVECS) &&
	    biovec_init_pool(&bs->bvec_pool, pool_size))
		goto bad;

	if (!(flags & BIOSET_NEED_RESCUER))
		return 0;

	bs->rescue_workqueue = alloc_workqueue("bioset", WQ_MEM_RECLAIM, 0);
	if (!bs->rescue_workqueue)
		goto bad;

	return 0;
bad:
	bioset_exit(bs);
	return -ENOMEM;
}
EXPORT_SYMBOL(bioset_init);

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
/*
 * Initialize and setup a new bio_set, based on the settings from
 * another bio_set.
 */
int bioset_init_from_src(struct bio_set *bs, struct bio_set *src)
{
	int flags;

	flags = 0;
	if (src->bvec_pool.min_nr)
		flags |= BIOSET_NEED_BVECS;
	if (src->rescue_workqueue)
		flags |= BIOSET_NEED_RESCUER;

	return bioset_init(bs, src->bio_pool.min_nr, src->front_pad, flags);
}
EXPORT_SYMBOL(bioset_init_from_src);

2020
#ifdef CONFIG_BLK_CGROUP
2021

2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045
#ifdef CONFIG_MEMCG
/**
 * bio_associate_blkcg_from_page - associate a bio with the page's blkcg
 * @bio: target bio
 * @page: the page to lookup the blkcg from
 *
 * Associate @bio with the blkcg from @page's owning memcg.  This works like
 * every other associate function wrt references.
 */
int bio_associate_blkcg_from_page(struct bio *bio, struct page *page)
{
	struct cgroup_subsys_state *blkcg_css;

	if (unlikely(bio->bi_css))
		return -EBUSY;
	if (!page->mem_cgroup)
		return 0;
	blkcg_css = cgroup_get_e_css(page->mem_cgroup->css.cgroup,
				     &io_cgrp_subsys);
	bio->bi_css = blkcg_css;
	return 0;
}
#endif /* CONFIG_MEMCG */

2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065
/**
 * bio_associate_blkcg - associate a bio with the specified blkcg
 * @bio: target bio
 * @blkcg_css: css of the blkcg to associate
 *
 * Associate @bio with the blkcg specified by @blkcg_css.  Block layer will
 * treat @bio as if it were issued by a task which belongs to the blkcg.
 *
 * This function takes an extra reference of @blkcg_css which will be put
 * when @bio is released.  The caller must own @bio and is responsible for
 * synchronizing calls to this function.
 */
int bio_associate_blkcg(struct bio *bio, struct cgroup_subsys_state *blkcg_css)
{
	if (unlikely(bio->bi_css))
		return -EBUSY;
	css_get(blkcg_css);
	bio->bi_css = blkcg_css;
	return 0;
}
2066
EXPORT_SYMBOL_GPL(bio_associate_blkcg);
2067

2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085
/**
 * bio_associate_blkg - associate a bio with the specified blkg
 * @bio: target bio
 * @blkg: the blkg to associate
 *
 * Associate @bio with the blkg specified by @blkg.  This is the queue specific
 * blkcg information associated with the @bio, a reference will be taken on the
 * @blkg and will be freed when the bio is freed.
 */
int bio_associate_blkg(struct bio *bio, struct blkcg_gq *blkg)
{
	if (unlikely(bio->bi_blkg))
		return -EBUSY;
	blkg_get(blkg);
	bio->bi_blkg = blkg;
	return 0;
}

2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099
/**
 * bio_disassociate_task - undo bio_associate_current()
 * @bio: target bio
 */
void bio_disassociate_task(struct bio *bio)
{
	if (bio->bi_ioc) {
		put_io_context(bio->bi_ioc);
		bio->bi_ioc = NULL;
	}
	if (bio->bi_css) {
		css_put(bio->bi_css);
		bio->bi_css = NULL;
	}
2100 2101 2102 2103
	if (bio->bi_blkg) {
		blkg_put(bio->bi_blkg);
		bio->bi_blkg = NULL;
	}
2104 2105
}

2106 2107 2108 2109 2110 2111 2112 2113 2114 2115
/**
 * bio_clone_blkcg_association - clone blkcg association from src to dst bio
 * @dst: destination bio
 * @src: source bio
 */
void bio_clone_blkcg_association(struct bio *dst, struct bio *src)
{
	if (src->bi_css)
		WARN_ON(bio_associate_blkcg(dst, src->bi_css));
}
2116
EXPORT_SYMBOL_GPL(bio_clone_blkcg_association);
2117 2118
#endif /* CONFIG_BLK_CGROUP */

L
Linus Torvalds 已提交
2119 2120 2121 2122
static void __init biovec_init_slabs(void)
{
	int i;

2123
	for (i = 0; i < BVEC_POOL_NR; i++) {
L
Linus Torvalds 已提交
2124 2125 2126
		int size;
		struct biovec_slab *bvs = bvec_slabs + i;

2127 2128 2129 2130 2131
		if (bvs->nr_vecs <= BIO_INLINE_VECS) {
			bvs->slab = NULL;
			continue;
		}

L
Linus Torvalds 已提交
2132 2133
		size = bvs->nr_vecs * sizeof(struct bio_vec);
		bvs->slab = kmem_cache_create(bvs->name, size, 0,
2134
                                SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
L
Linus Torvalds 已提交
2135 2136 2137 2138 2139
	}
}

static int __init init_bio(void)
{
2140 2141
	bio_slab_max = 2;
	bio_slab_nr = 0;
K
Kees Cook 已提交
2142 2143
	bio_slabs = kcalloc(bio_slab_max, sizeof(struct bio_slab),
			    GFP_KERNEL);
2144 2145
	if (!bio_slabs)
		panic("bio: can't allocate bios\n");
L
Linus Torvalds 已提交
2146

2147
	bio_integrity_init();
L
Linus Torvalds 已提交
2148 2149
	biovec_init_slabs();

2150
	if (bioset_init(&fs_bio_set, BIO_POOL_SIZE, 0, BIOSET_NEED_BVECS))
L
Linus Torvalds 已提交
2151 2152
		panic("bio: can't allocate bios\n");

2153
	if (bioset_integrity_create(&fs_bio_set, BIO_POOL_SIZE))
2154 2155
		panic("bio: can't create integrity pool\n");

L
Linus Torvalds 已提交
2156 2157 2158
	return 0;
}
subsys_initcall(init_bio);