bio-integrity.c 17.4 KB
Newer Older
1 2 3
/*
 * bio-integrity.c - bio data integrity extensions
 *
4
 * Copyright (C) 2007, 2008, 2009 Oracle Corporation
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
 * Written by: Martin K. Petersen <martin.petersen@oracle.com>
 *
 * 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 License
 * along with this program; see the file COPYING.  If not, write to
 * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139,
 * USA.
 *
 */

#include <linux/blkdev.h>
#include <linux/mempool.h>
25
#include <linux/export.h>
26 27
#include <linux/bio.h>
#include <linux/workqueue.h>
28
#include <linux/slab.h>
29

30
#define BIP_INLINE_VECS	4
31

32
static struct kmem_cache *bip_slab;
33 34 35
static struct workqueue_struct *kintegrityd_wq;

/**
36
 * bio_integrity_alloc - Allocate integrity payload and attach it to bio
37 38 39 40 41 42 43 44
 * @bio:	bio to attach integrity metadata to
 * @gfp_mask:	Memory allocation mask
 * @nr_vecs:	Number of integrity metadata scatter-gather elements
 *
 * Description: This function prepares a bio for attaching integrity
 * metadata.  nr_vecs specifies the maximum number of pages containing
 * integrity metadata that can be attached.
 */
45 46 47
struct bio_integrity_payload *bio_integrity_alloc(struct bio *bio,
						  gfp_t gfp_mask,
						  unsigned int nr_vecs)
48 49
{
	struct bio_integrity_payload *bip;
50
	struct bio_set *bs = bio->bi_pool;
51 52 53 54 55 56 57 58
	unsigned long idx = BIO_POOL_NONE;
	unsigned inline_vecs;

	if (!bs) {
		bip = kmalloc(sizeof(struct bio_integrity_payload) +
			      sizeof(struct bio_vec) * nr_vecs, gfp_mask);
		inline_vecs = nr_vecs;
	} else {
59
		bip = mempool_alloc(bs->bio_integrity_pool, gfp_mask);
60
		inline_vecs = BIP_INLINE_VECS;
61 62
	}

63 64 65
	if (unlikely(!bip))
		return NULL;

66 67
	memset(bip, 0, sizeof(*bip));

68 69 70 71 72
	if (nr_vecs > inline_vecs) {
		bip->bip_vec = bvec_alloc(gfp_mask, nr_vecs, &idx,
					  bs->bvec_integrity_pool);
		if (!bip->bip_vec)
			goto err;
73
		bip->bip_max_vcnt = bvec_nr_vecs(idx);
74 75
	} else {
		bip->bip_vec = bip->bip_inline_vecs;
76
		bip->bip_max_vcnt = inline_vecs;
77 78
	}

79
	bip->bip_slab = idx;
80 81 82 83
	bip->bip_bio = bio;
	bio->bi_integrity = bip;

	return bip;
84 85 86
err:
	mempool_free(bip, bs->bio_integrity_pool);
	return NULL;
87 88 89 90 91 92 93 94 95 96
}
EXPORT_SYMBOL(bio_integrity_alloc);

/**
 * bio_integrity_free - Free bio integrity payload
 * @bio:	bio containing bip to be freed
 *
 * Description: Used to free the integrity portion of a bio. Usually
 * called from bio_free().
 */
97
void bio_integrity_free(struct bio *bio)
98 99
{
	struct bio_integrity_payload *bip = bio->bi_integrity;
100 101
	struct bio_set *bs = bio->bi_pool;

102
	if (bip->bip_owns_buf)
103 104
		kfree(bip->bip_buf);

105 106 107 108 109
	if (bs) {
		if (bip->bip_slab != BIO_POOL_NONE)
			bvec_free(bs->bvec_integrity_pool, bip->bip_vec,
				  bip->bip_slab);

110
		mempool_free(bip, bs->bio_integrity_pool);
111 112 113
	} else {
		kfree(bip);
	}
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133

	bio->bi_integrity = NULL;
}
EXPORT_SYMBOL(bio_integrity_free);

/**
 * bio_integrity_add_page - Attach integrity metadata
 * @bio:	bio to update
 * @page:	page containing integrity metadata
 * @len:	number of bytes of integrity metadata in page
 * @offset:	start offset within page
 *
 * Description: Attach a page containing integrity metadata to bio.
 */
int bio_integrity_add_page(struct bio *bio, struct page *page,
			   unsigned int len, unsigned int offset)
{
	struct bio_integrity_payload *bip = bio->bi_integrity;
	struct bio_vec *iv;

134
	if (bip->bip_vcnt >= bip->bip_max_vcnt) {
135 136 137 138
		printk(KERN_ERR "%s: bip_vec full\n", __func__);
		return 0;
	}

139
	iv = bip->bip_vec + bip->bip_vcnt;
140 141 142 143 144 145 146 147 148 149

	iv->bv_page = page;
	iv->bv_len = len;
	iv->bv_offset = offset;
	bip->bip_vcnt++;

	return len;
}
EXPORT_SYMBOL(bio_integrity_add_page);

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167
static int bdev_integrity_enabled(struct block_device *bdev, int rw)
{
	struct blk_integrity *bi = bdev_get_integrity(bdev);

	if (bi == NULL)
		return 0;

	if (rw == READ && bi->verify_fn != NULL &&
	    (bi->flags & INTEGRITY_FLAG_READ))
		return 1;

	if (rw == WRITE && bi->generate_fn != NULL &&
	    (bi->flags & INTEGRITY_FLAG_WRITE))
		return 1;

	return 0;
}

168 169 170 171 172 173 174 175 176 177 178
/**
 * bio_integrity_enabled - Check whether integrity can be passed
 * @bio:	bio to check
 *
 * Description: Determines whether bio_integrity_prep() can be called
 * on this bio or not.	bio data direction and target device must be
 * set prior to calling.  The functions honors the write_generate and
 * read_verify flags in sysfs.
 */
int bio_integrity_enabled(struct bio *bio)
{
179 180 181
	if (!bio_is_rw(bio))
		return 0;

182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199
	/* Already protected? */
	if (bio_integrity(bio))
		return 0;

	return bdev_integrity_enabled(bio->bi_bdev, bio_data_dir(bio));
}
EXPORT_SYMBOL(bio_integrity_enabled);

/**
 * bio_integrity_hw_sectors - Convert 512b sectors to hardware ditto
 * @bi:		blk_integrity profile for device
 * @sectors:	Number of 512 sectors to convert
 *
 * Description: The block layer calculates everything in 512 byte
 * sectors but integrity metadata is done in terms of the hardware
 * sector size of the storage device.  Convert the block layer sectors
 * to physical sectors.
 */
200 201
static inline unsigned int bio_integrity_hw_sectors(struct blk_integrity *bi,
						    unsigned int sectors)
202 203 204 205 206 207 208 209
{
	/* At this point there are only 512b or 4096b DIF/EPP devices */
	if (bi->sector_size == 4096)
		return sectors >>= 3;

	return sectors;
}

210 211 212 213 214 215
static inline unsigned int bio_integrity_bytes(struct blk_integrity *bi,
					       unsigned int sectors)
{
	return bio_integrity_hw_sectors(bi, sectors) * bi->tuple_size;
}

216 217 218 219 220 221 222 223 224 225 226 227
/**
 * bio_integrity_tag_size - Retrieve integrity tag space
 * @bio:	bio to inspect
 *
 * Description: Returns the maximum number of tag bytes that can be
 * attached to this bio. Filesystems can use this to determine how
 * much metadata to attach to an I/O.
 */
unsigned int bio_integrity_tag_size(struct bio *bio)
{
	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);

228
	BUG_ON(bio->bi_iter.bi_size == 0);
229

230
	return bi->tag_size * (bio->bi_iter.bi_size / bi->sector_size);
231 232 233
}
EXPORT_SYMBOL(bio_integrity_tag_size);

234 235
static int bio_integrity_tag(struct bio *bio, void *tag_buf, unsigned int len,
			     int set)
236 237 238 239 240 241 242 243 244 245
{
	struct bio_integrity_payload *bip = bio->bi_integrity;
	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);
	unsigned int nr_sectors;

	BUG_ON(bip->bip_buf == NULL);

	if (bi->tag_size == 0)
		return -1;

246 247
	nr_sectors = bio_integrity_hw_sectors(bi,
					DIV_ROUND_UP(len, bi->tag_size));
248

249 250 251
	if (nr_sectors * bi->tuple_size > bip->bip_iter.bi_size) {
		printk(KERN_ERR "%s: tag too big for bio: %u > %u\n", __func__,
		       nr_sectors * bi->tuple_size, bip->bip_iter.bi_size);
252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300
		return -1;
	}

	if (set)
		bi->set_tag_fn(bip->bip_buf, tag_buf, nr_sectors);
	else
		bi->get_tag_fn(bip->bip_buf, tag_buf, nr_sectors);

	return 0;
}

/**
 * bio_integrity_set_tag - Attach a tag buffer to a bio
 * @bio:	bio to attach buffer to
 * @tag_buf:	Pointer to a buffer containing tag data
 * @len:	Length of the included buffer
 *
 * Description: Use this function to tag a bio by leveraging the extra
 * space provided by devices formatted with integrity protection.  The
 * size of the integrity buffer must be <= to the size reported by
 * bio_integrity_tag_size().
 */
int bio_integrity_set_tag(struct bio *bio, void *tag_buf, unsigned int len)
{
	BUG_ON(bio_data_dir(bio) != WRITE);

	return bio_integrity_tag(bio, tag_buf, len, 1);
}
EXPORT_SYMBOL(bio_integrity_set_tag);

/**
 * bio_integrity_get_tag - Retrieve a tag buffer from a bio
 * @bio:	bio to retrieve buffer from
 * @tag_buf:	Pointer to a buffer for the tag data
 * @len:	Length of the target buffer
 *
 * Description: Use this function to retrieve the tag buffer from a
 * completed I/O. The size of the integrity buffer must be <= to the
 * size reported by bio_integrity_tag_size().
 */
int bio_integrity_get_tag(struct bio *bio, void *tag_buf, unsigned int len)
{
	BUG_ON(bio_data_dir(bio) != READ);

	return bio_integrity_tag(bio, tag_buf, len, 0);
}
EXPORT_SYMBOL(bio_integrity_get_tag);

/**
301 302 303
 * bio_integrity_generate_verify - Generate/verify integrity metadata for a bio
 * @bio:	bio to generate/verify integrity metadata for
 * @operate:	operate number, 1 for generate, 0 for verify
304
 */
305
static int bio_integrity_generate_verify(struct bio *bio, int operate)
306 307 308
{
	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);
	struct blk_integrity_exchg bix;
309
	struct bio_vec *bv;
310
	sector_t sector;
311
	unsigned int sectors, ret = 0, i;
312 313
	void *prot_buf = bio->bi_integrity->bip_buf;

314 315 316 317 318
	if (operate)
		sector = bio->bi_iter.bi_sector;
	else
		sector = bio->bi_integrity->bip_iter.bi_sector;

319 320 321
	bix.disk_name = bio->bi_bdev->bd_disk->disk_name;
	bix.sector_size = bi->sector_size;

322 323 324 325
	bio_for_each_segment_all(bv, bio, i) {
		void *kaddr = kmap_atomic(bv->bv_page);
		bix.data_buf = kaddr + bv->bv_offset;
		bix.data_size = bv->bv_len;
326 327 328
		bix.prot_buf = prot_buf;
		bix.sector = sector;

329
		if (operate)
330
			bi->generate_fn(&bix);
331
		else {
332 333 334 335 336 337
			ret = bi->verify_fn(&bix);
			if (ret) {
				kunmap_atomic(kaddr);
				return ret;
			}
		}
338

339
		sectors = bv->bv_len / bi->sector_size;
340 341 342
		sector += sectors;
		prot_buf += sectors * bi->tuple_size;

343
		kunmap_atomic(kaddr);
344
	}
345 346 347 348 349 350 351 352 353 354 355 356 357 358 359
	return ret;
}

/**
 * bio_integrity_generate - Generate integrity metadata for a bio
 * @bio:	bio to generate integrity metadata for
 *
 * Description: Generates integrity metadata for a bio by calling the
 * block device's generation callback function.  The bio must have a
 * bip attached with enough room to accommodate the generated
 * integrity metadata.
 */
static void bio_integrity_generate(struct bio *bio)
{
	bio_integrity_generate_verify(bio, 1);
360 361
}

362 363 364 365 366 367 368 369
static inline unsigned short blk_integrity_tuple_size(struct blk_integrity *bi)
{
	if (bi)
		return bi->tuple_size;

	return 0;
}

370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400
/**
 * bio_integrity_prep - Prepare bio for integrity I/O
 * @bio:	bio to prepare
 *
 * Description: Allocates a buffer for integrity metadata, maps the
 * pages and attaches them to a bio.  The bio must have data
 * direction, target device and start sector set priot to calling.  In
 * the WRITE case, integrity metadata will be generated using the
 * block device's integrity function.  In the READ case, the buffer
 * will be prepared for DMA and a suitable end_io handler set up.
 */
int bio_integrity_prep(struct bio *bio)
{
	struct bio_integrity_payload *bip;
	struct blk_integrity *bi;
	struct request_queue *q;
	void *buf;
	unsigned long start, end;
	unsigned int len, nr_pages;
	unsigned int bytes, offset, i;
	unsigned int sectors;

	bi = bdev_get_integrity(bio->bi_bdev);
	q = bdev_get_queue(bio->bi_bdev);
	BUG_ON(bi == NULL);
	BUG_ON(bio_integrity(bio));

	sectors = bio_integrity_hw_sectors(bi, bio_sectors(bio));

	/* Allocate kernel buffer for protection data */
	len = sectors * blk_integrity_tuple_size(bi);
401
	buf = kmalloc(len, GFP_NOIO | q->bounce_gfp);
402 403
	if (unlikely(buf == NULL)) {
		printk(KERN_ERR "could not allocate integrity buffer\n");
404
		return -ENOMEM;
405 406 407 408 409 410 411 412 413 414 415 416 417 418
	}

	end = (((unsigned long) buf) + len + PAGE_SIZE - 1) >> PAGE_SHIFT;
	start = ((unsigned long) buf) >> PAGE_SHIFT;
	nr_pages = end - start;

	/* Allocate bio integrity payload and integrity vectors */
	bip = bio_integrity_alloc(bio, GFP_NOIO, nr_pages);
	if (unlikely(bip == NULL)) {
		printk(KERN_ERR "could not allocate data integrity bioset\n");
		kfree(buf);
		return -EIO;
	}

419
	bip->bip_owns_buf = 1;
420
	bip->bip_buf = buf;
421 422
	bip->bip_iter.bi_size = len;
	bip->bip_iter.bi_sector = bio->bi_iter.bi_sector;
423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473

	/* Map it */
	offset = offset_in_page(buf);
	for (i = 0 ; i < nr_pages ; i++) {
		int ret;
		bytes = PAGE_SIZE - offset;

		if (len <= 0)
			break;

		if (bytes > len)
			bytes = len;

		ret = bio_integrity_add_page(bio, virt_to_page(buf),
					     bytes, offset);

		if (ret == 0)
			return 0;

		if (ret < bytes)
			break;

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

	/* Install custom I/O completion handler if read verify is enabled */
	if (bio_data_dir(bio) == READ) {
		bip->bip_end_io = bio->bi_end_io;
		bio->bi_end_io = bio_integrity_endio;
	}

	/* Auto-generate integrity metadata if this is a write */
	if (bio_data_dir(bio) == WRITE)
		bio_integrity_generate(bio);

	return 0;
}
EXPORT_SYMBOL(bio_integrity_prep);

/**
 * bio_integrity_verify - Verify integrity metadata for a bio
 * @bio:	bio to verify
 *
 * Description: This function is called to verify the integrity of a
 * bio.	 The data in the bio io_vec is compared to the integrity
 * metadata returned by the HBA.
 */
static int bio_integrity_verify(struct bio *bio)
{
474
	return bio_integrity_generate_verify(bio, 0);
475 476 477 478 479 480 481 482 483 484 485 486
}

/**
 * bio_integrity_verify_fn - Integrity I/O completion worker
 * @work:	Work struct stored in bio to be verified
 *
 * Description: This workqueue function is called to complete a READ
 * request.  The function verifies the transferred integrity metadata
 * and then calls the original bio end_io function.
 */
static void bio_integrity_verify_fn(struct work_struct *work)
{
487
	struct bio_integrity_payload *bip =
488 489
		container_of(work, struct bio_integrity_payload, bip_work);
	struct bio *bio = bip->bip_bio;
490
	int error;
491

492
	error = bio_integrity_verify(bio);
493 494 495

	/* Restore original bio completion handler */
	bio->bi_end_io = bip->bip_end_io;
K
Kent Overstreet 已提交
496
	bio_endio_nodec(bio, error);
497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516
}

/**
 * bio_integrity_endio - Integrity I/O completion function
 * @bio:	Protected bio
 * @error:	Pointer to errno
 *
 * Description: Completion for integrity I/O
 *
 * Normally I/O completion is done in interrupt context.  However,
 * verifying I/O integrity is a time-consuming task which must be run
 * in process context.	This function postpones completion
 * accordingly.
 */
void bio_integrity_endio(struct bio *bio, int error)
{
	struct bio_integrity_payload *bip = bio->bi_integrity;

	BUG_ON(bip->bip_bio != bio);

517 518 519 520 521 522
	/* In case of an I/O error there is no point in verifying the
	 * integrity metadata.  Restore original bio end_io handler
	 * and run it.
	 */
	if (error) {
		bio->bi_end_io = bip->bip_end_io;
523
		bio_endio_nodec(bio, error);
524 525 526 527

		return;
	}

528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545
	INIT_WORK(&bip->bip_work, bio_integrity_verify_fn);
	queue_work(kintegrityd_wq, &bip->bip_work);
}
EXPORT_SYMBOL(bio_integrity_endio);

/**
 * bio_integrity_advance - Advance integrity vector
 * @bio:	bio whose integrity vector to update
 * @bytes_done:	number of data bytes that have been completed
 *
 * Description: This function calculates how many integrity bytes the
 * number of completed data bytes correspond to and advances the
 * integrity vector accordingly.
 */
void bio_integrity_advance(struct bio *bio, unsigned int bytes_done)
{
	struct bio_integrity_payload *bip = bio->bi_integrity;
	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);
546
	unsigned bytes = bio_integrity_bytes(bi, bytes_done >> 9);
547

548
	bvec_iter_advance(bip->bip_vec, &bip->bip_iter, bytes);
549 550 551 552 553 554 555 556 557 558 559 560 561 562
}
EXPORT_SYMBOL(bio_integrity_advance);

/**
 * bio_integrity_trim - Trim integrity vector
 * @bio:	bio whose integrity vector to update
 * @offset:	offset to first data sector
 * @sectors:	number of data sectors
 *
 * Description: Used to trim the integrity vector in a cloned bio.
 * The ivec will be advanced corresponding to 'offset' data sectors
 * and the length will be truncated corresponding to 'len' data
 * sectors.
 */
563 564
void bio_integrity_trim(struct bio *bio, unsigned int offset,
			unsigned int sectors)
565 566 567 568
{
	struct bio_integrity_payload *bip = bio->bi_integrity;
	struct blk_integrity *bi = bdev_get_integrity(bio->bi_bdev);

569 570
	bio_integrity_advance(bio, offset << 9);
	bip->bip_iter.bi_size = bio_integrity_bytes(bi, sectors);
571 572 573 574 575 576 577
}
EXPORT_SYMBOL(bio_integrity_trim);

/**
 * bio_integrity_clone - Callback for cloning bios with integrity metadata
 * @bio:	New bio
 * @bio_src:	Original bio
578
 * @gfp_mask:	Memory allocation mask
579 580 581
 *
 * Description:	Called to allocate a bip when cloning a bio
 */
582
int bio_integrity_clone(struct bio *bio, struct bio *bio_src,
583
			gfp_t gfp_mask)
584 585 586 587 588 589
{
	struct bio_integrity_payload *bip_src = bio_src->bi_integrity;
	struct bio_integrity_payload *bip;

	BUG_ON(bip_src == NULL);

590
	bip = bio_integrity_alloc(bio, gfp_mask, bip_src->bip_vcnt);
591 592 593 594 595 596 597 598

	if (bip == NULL)
		return -EIO;

	memcpy(bip->bip_vec, bip_src->bip_vec,
	       bip_src->bip_vcnt * sizeof(struct bio_vec));

	bip->bip_vcnt = bip_src->bip_vcnt;
599
	bip->bip_iter = bip_src->bip_iter;
600 601 602 603 604

	return 0;
}
EXPORT_SYMBOL(bio_integrity_clone);

605
int bioset_integrity_create(struct bio_set *bs, int pool_size)
606
{
607 608 609
	if (bs->bio_integrity_pool)
		return 0;

610
	bs->bio_integrity_pool = mempool_create_slab_pool(pool_size, bip_slab);
611
	if (!bs->bio_integrity_pool)
612
		return -1;
613

614
	bs->bvec_integrity_pool = biovec_create_pool(pool_size);
615 616
	if (!bs->bvec_integrity_pool) {
		mempool_destroy(bs->bio_integrity_pool);
617
		return -1;
618
	}
619 620 621 622 623 624 625 626 627

	return 0;
}
EXPORT_SYMBOL(bioset_integrity_create);

void bioset_integrity_free(struct bio_set *bs)
{
	if (bs->bio_integrity_pool)
		mempool_destroy(bs->bio_integrity_pool);
628 629

	if (bs->bvec_integrity_pool)
630
		mempool_destroy(bs->bvec_integrity_pool);
631 632 633 634 635
}
EXPORT_SYMBOL(bioset_integrity_free);

void __init bio_integrity_init(void)
{
636 637 638 639 640 641
	/*
	 * kintegrityd won't block much but may burn a lot of CPU cycles.
	 * Make it highpri CPU intensive wq with max concurrency of 1.
	 */
	kintegrityd_wq = alloc_workqueue("kintegrityd", WQ_MEM_RECLAIM |
					 WQ_HIGHPRI | WQ_CPU_INTENSIVE, 1);
642 643
	if (!kintegrityd_wq)
		panic("Failed to create kintegrityd\n");
644

645 646 647 648 649 650
	bip_slab = kmem_cache_create("bio_integrity_payload",
				     sizeof(struct bio_integrity_payload) +
				     sizeof(struct bio_vec) * BIP_INLINE_VECS,
				     0, SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
	if (!bip_slab)
		panic("Failed to create slab\n");
651
}