raid1.c 44.1 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 * raid1.c : Multiple Devices driver for Linux
 *
 * Copyright (C) 1999, 2000, 2001 Ingo Molnar, Red Hat
 *
 * Copyright (C) 1996, 1997, 1998 Ingo Molnar, Miguel de Icaza, Gadi Oxman
 *
 * RAID-1 management functions.
 *
 * Better read-balancing code written by Mika Kuoppala <miku@iki.fi>, 2000
 *
 * Fixes to reconstruction by Jakob stergaard" <jakob@ostenfeld.dk>
 * Various fixes by Neil Brown <neilb@cse.unsw.edu.au>
 *
15 16 17 18 19 20 21 22 23
 * Changes by Peter T. Breuer <ptb@it.uc3m.es> 31/1/2003 to support
 * bitmapped intelligence in resync:
 *
 *      - bitmap marked during normal i/o
 *      - bitmap used to skip nondirty blocks during sync
 *
 * Additions to bitmap code, (C) 2003-2004 Paul Clements, SteelEye Technology:
 * - persistent bitmap code
 *
L
Linus Torvalds 已提交
24 25 26 27 28 29 30 31 32 33
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2, or (at your option)
 * any later version.
 *
 * You should have received a copy of the GNU General Public License
 * (for example /usr/src/linux/COPYING); if not, write to the Free
 * Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
 */

34
#include "dm-bio-list.h"
L
Linus Torvalds 已提交
35
#include <linux/raid/raid1.h>
36 37 38 39 40 41 42 43
#include <linux/raid/bitmap.h>

#define DEBUG 0
#if DEBUG
#define PRINTK(x...) printk(x)
#else
#define PRINTK(x...)
#endif
L
Linus Torvalds 已提交
44 45 46 47 48 49 50 51 52 53 54

/*
 * Number of guaranteed r1bios in case of extreme VM load:
 */
#define	NR_RAID1_BIOS 256

static mdk_personality_t raid1_personality;

static void unplug_slaves(mddev_t *mddev);


A
Al Viro 已提交
55
static void * r1bio_pool_alloc(gfp_t gfp_flags, void *data)
L
Linus Torvalds 已提交
56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
{
	struct pool_info *pi = data;
	r1bio_t *r1_bio;
	int size = offsetof(r1bio_t, bios[pi->raid_disks]);

	/* allocate a r1bio with room for raid_disks entries in the bios array */
	r1_bio = kmalloc(size, gfp_flags);
	if (r1_bio)
		memset(r1_bio, 0, size);
	else
		unplug_slaves(pi->mddev);

	return r1_bio;
}

static void r1bio_pool_free(void *r1_bio, void *data)
{
	kfree(r1_bio);
}

#define RESYNC_BLOCK_SIZE (64*1024)
//#define RESYNC_BLOCK_SIZE PAGE_SIZE
#define RESYNC_SECTORS (RESYNC_BLOCK_SIZE >> 9)
#define RESYNC_PAGES ((RESYNC_BLOCK_SIZE + PAGE_SIZE-1) / PAGE_SIZE)
#define RESYNC_WINDOW (2048*1024)

A
Al Viro 已提交
82
static void * r1buf_pool_alloc(gfp_t gfp_flags, void *data)
L
Linus Torvalds 已提交
83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224
{
	struct pool_info *pi = data;
	struct page *page;
	r1bio_t *r1_bio;
	struct bio *bio;
	int i, j;

	r1_bio = r1bio_pool_alloc(gfp_flags, pi);
	if (!r1_bio) {
		unplug_slaves(pi->mddev);
		return NULL;
	}

	/*
	 * Allocate bios : 1 for reading, n-1 for writing
	 */
	for (j = pi->raid_disks ; j-- ; ) {
		bio = bio_alloc(gfp_flags, RESYNC_PAGES);
		if (!bio)
			goto out_free_bio;
		r1_bio->bios[j] = bio;
	}
	/*
	 * Allocate RESYNC_PAGES data pages and attach them to
	 * the first bio;
	 */
	bio = r1_bio->bios[0];
	for (i = 0; i < RESYNC_PAGES; i++) {
		page = alloc_page(gfp_flags);
		if (unlikely(!page))
			goto out_free_pages;

		bio->bi_io_vec[i].bv_page = page;
	}

	r1_bio->master_bio = NULL;

	return r1_bio;

out_free_pages:
	for ( ; i > 0 ; i--)
		__free_page(bio->bi_io_vec[i-1].bv_page);
out_free_bio:
	while ( ++j < pi->raid_disks )
		bio_put(r1_bio->bios[j]);
	r1bio_pool_free(r1_bio, data);
	return NULL;
}

static void r1buf_pool_free(void *__r1_bio, void *data)
{
	struct pool_info *pi = data;
	int i;
	r1bio_t *r1bio = __r1_bio;
	struct bio *bio = r1bio->bios[0];

	for (i = 0; i < RESYNC_PAGES; i++) {
		__free_page(bio->bi_io_vec[i].bv_page);
		bio->bi_io_vec[i].bv_page = NULL;
	}
	for (i=0 ; i < pi->raid_disks; i++)
		bio_put(r1bio->bios[i]);

	r1bio_pool_free(r1bio, data);
}

static void put_all_bios(conf_t *conf, r1bio_t *r1_bio)
{
	int i;

	for (i = 0; i < conf->raid_disks; i++) {
		struct bio **bio = r1_bio->bios + i;
		if (*bio)
			bio_put(*bio);
		*bio = NULL;
	}
}

static inline void free_r1bio(r1bio_t *r1_bio)
{
	unsigned long flags;

	conf_t *conf = mddev_to_conf(r1_bio->mddev);

	/*
	 * Wake up any possible resync thread that waits for the device
	 * to go idle.
	 */
	spin_lock_irqsave(&conf->resync_lock, flags);
	if (!--conf->nr_pending) {
		wake_up(&conf->wait_idle);
		wake_up(&conf->wait_resume);
	}
	spin_unlock_irqrestore(&conf->resync_lock, flags);

	put_all_bios(conf, r1_bio);
	mempool_free(r1_bio, conf->r1bio_pool);
}

static inline void put_buf(r1bio_t *r1_bio)
{
	conf_t *conf = mddev_to_conf(r1_bio->mddev);
	unsigned long flags;

	mempool_free(r1_bio, conf->r1buf_pool);

	spin_lock_irqsave(&conf->resync_lock, flags);
	if (!conf->barrier)
		BUG();
	--conf->barrier;
	wake_up(&conf->wait_resume);
	wake_up(&conf->wait_idle);

	if (!--conf->nr_pending) {
		wake_up(&conf->wait_idle);
		wake_up(&conf->wait_resume);
	}
	spin_unlock_irqrestore(&conf->resync_lock, flags);
}

static void reschedule_retry(r1bio_t *r1_bio)
{
	unsigned long flags;
	mddev_t *mddev = r1_bio->mddev;
	conf_t *conf = mddev_to_conf(mddev);

	spin_lock_irqsave(&conf->device_lock, flags);
	list_add(&r1_bio->retry_list, &conf->retry_list);
	spin_unlock_irqrestore(&conf->device_lock, flags);

	md_wakeup_thread(mddev->thread);
}

/*
 * raid_end_bio_io() is called when we have finished servicing a mirrored
 * operation and are ready to return a success/failure code to the buffer
 * cache layer.
 */
static void raid_end_bio_io(r1bio_t *r1_bio)
{
	struct bio *bio = r1_bio->master_bio;

225 226 227 228 229 230 231 232 233 234 235
	/* if nobody has done the final endio yet, do it now */
	if (!test_and_set_bit(R1BIO_Returned, &r1_bio->state)) {
		PRINTK(KERN_DEBUG "raid1: sync end %s on sectors %llu-%llu\n",
			(bio_data_dir(bio) == WRITE) ? "write" : "read",
			(unsigned long long) bio->bi_sector,
			(unsigned long long) bio->bi_sector +
				(bio->bi_size >> 9) - 1);

		bio_endio(bio, bio->bi_size,
			test_bit(R1BIO_Uptodate, &r1_bio->state) ? 0 : -EIO);
	}
L
Linus Torvalds 已提交
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
	free_r1bio(r1_bio);
}

/*
 * Update disk head position estimator based on IRQ completion info.
 */
static inline void update_head_pos(int disk, r1bio_t *r1_bio)
{
	conf_t *conf = mddev_to_conf(r1_bio->mddev);

	conf->mirrors[disk].head_position =
		r1_bio->sector + (r1_bio->sectors);
}

static int raid1_end_read_request(struct bio *bio, unsigned int bytes_done, int error)
{
	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
	int mirror;
	conf_t *conf = mddev_to_conf(r1_bio->mddev);

	if (bio->bi_size)
		return 1;
	
	mirror = r1_bio->read_disk;
	/*
	 * this branch is our 'one mirror IO has finished' event handler:
	 */
	if (!uptodate)
		md_error(r1_bio->mddev, conf->mirrors[mirror].rdev);
	else
		/*
		 * Set R1BIO_Uptodate in our master bio, so that
		 * we will return a good error code for to the higher
		 * levels even if IO on some other mirrored buffer fails.
		 *
		 * The 'master' represents the composite IO operation to
		 * user-side. So if something waits for IO, then it will
		 * wait for the 'master' bio.
		 */
		set_bit(R1BIO_Uptodate, &r1_bio->state);

	update_head_pos(mirror, r1_bio);

	/*
	 * we have only one bio on the read side
	 */
	if (uptodate)
		raid_end_bio_io(r1_bio);
	else {
		/*
		 * oops, read error:
		 */
		char b[BDEVNAME_SIZE];
		if (printk_ratelimit())
			printk(KERN_ERR "raid1: %s: rescheduling sector %llu\n",
			       bdevname(conf->mirrors[mirror].rdev->bdev,b), (unsigned long long)r1_bio->sector);
		reschedule_retry(r1_bio);
	}

	rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev);
	return 0;
}

static int raid1_end_write_request(struct bio *bio, unsigned int bytes_done, int error)
{
	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
304
	int mirror, behind;
L
Linus Torvalds 已提交
305 306 307 308 309 310 311 312 313 314 315 316
	conf_t *conf = mddev_to_conf(r1_bio->mddev);

	if (bio->bi_size)
		return 1;

	for (mirror = 0; mirror < conf->raid_disks; mirror++)
		if (r1_bio->bios[mirror] == bio)
			break;

	/*
	 * this branch is our 'one mirror IO has finished' event handler:
	 */
317
	if (!uptodate) {
L
Linus Torvalds 已提交
318
		md_error(r1_bio->mddev, conf->mirrors[mirror].rdev);
319 320 321
		/* an I/O failed, we can't clear the bitmap */
		set_bit(R1BIO_Degraded, &r1_bio->state);
	} else
L
Linus Torvalds 已提交
322 323 324 325 326 327 328 329 330 331 332 333 334
		/*
		 * Set R1BIO_Uptodate in our master bio, so that
		 * we will return a good error code for to the higher
		 * levels even if IO on some other mirrored buffer fails.
		 *
		 * The 'master' represents the composite IO operation to
		 * user-side. So if something waits for IO, then it will
		 * wait for the 'master' bio.
		 */
		set_bit(R1BIO_Uptodate, &r1_bio->state);

	update_head_pos(mirror, r1_bio);

335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357
	behind = test_bit(R1BIO_BehindIO, &r1_bio->state);
	if (behind) {
		if (test_bit(WriteMostly, &conf->mirrors[mirror].rdev->flags))
			atomic_dec(&r1_bio->behind_remaining);

		/* In behind mode, we ACK the master bio once the I/O has safely
		 * reached all non-writemostly disks. Setting the Returned bit
		 * ensures that this gets done only once -- we don't ever want to
		 * return -EIO here, instead we'll wait */

		if (atomic_read(&r1_bio->behind_remaining) >= (atomic_read(&r1_bio->remaining)-1) &&
		    test_bit(R1BIO_Uptodate, &r1_bio->state)) {
			/* Maybe we can return now */
			if (!test_and_set_bit(R1BIO_Returned, &r1_bio->state)) {
				struct bio *mbio = r1_bio->master_bio;
				PRINTK(KERN_DEBUG "raid1: behind end write sectors %llu-%llu\n",
				       (unsigned long long) mbio->bi_sector,
				       (unsigned long long) mbio->bi_sector +
				       (mbio->bi_size >> 9) - 1);
				bio_endio(mbio, mbio->bi_size, 0);
			}
		}
	}
L
Linus Torvalds 已提交
358 359 360 361 362 363
	/*
	 *
	 * Let's see if all mirrored write operations have finished
	 * already.
	 */
	if (atomic_dec_and_test(&r1_bio->remaining)) {
364 365 366 367 368 369
		if (test_bit(R1BIO_BehindIO, &r1_bio->state)) {
			/* free extra copy of the data pages */
			int i = bio->bi_vcnt;
			while (i--)
				__free_page(bio->bi_io_vec[i].bv_page);
		}
370 371 372
		/* clear the bitmap if all writes complete successfully */
		bitmap_endwrite(r1_bio->mddev->bitmap, r1_bio->sector,
				r1_bio->sectors,
373 374
				!test_bit(R1BIO_Degraded, &r1_bio->state),
				behind);
L
Linus Torvalds 已提交
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401
		md_write_end(r1_bio->mddev);
		raid_end_bio_io(r1_bio);
	}

	rdev_dec_pending(conf->mirrors[mirror].rdev, conf->mddev);
	return 0;
}


/*
 * This routine returns the disk from which the requested read should
 * be done. There is a per-array 'next expected sequential IO' sector
 * number - if this matches on the next IO then we use the last disk.
 * There is also a per-disk 'last know head position' sector that is
 * maintained from IRQ contexts, both the normal and the resync IO
 * completion handlers update this position correctly. If there is no
 * perfect sequential match then we pick the disk whose head is closest.
 *
 * If there are 2 mirrors in the same 2 devices, performance degrades
 * because position is mirror, not device based.
 *
 * The rdev for the device selected will have nr_pending incremented.
 */
static int read_balance(conf_t *conf, r1bio_t *r1_bio)
{
	const unsigned long this_sector = r1_bio->sector;
	int new_disk = conf->last_used, disk = new_disk;
402
	int wonly_disk = -1;
L
Linus Torvalds 已提交
403 404
	const int sectors = r1_bio->sectors;
	sector_t new_distance, current_distance;
405
	mdk_rdev_t *rdev;
L
Linus Torvalds 已提交
406 407 408

	rcu_read_lock();
	/*
409
	 * Check if we can balance. We can balance on the whole
L
Linus Torvalds 已提交
410 411 412 413 414 415 416 417 418
	 * device if no resync is going on, or below the resync window.
	 * We take the first readable disk when above the resync window.
	 */
 retry:
	if (conf->mddev->recovery_cp < MaxSector &&
	    (this_sector + sectors >= conf->next_resync)) {
		/* Choose the first operation device, for consistancy */
		new_disk = 0;

419 420 421 422 423 424 425 426 427 428
		for (rdev = conf->mirrors[new_disk].rdev;
		     !rdev || !rdev->in_sync
			     || test_bit(WriteMostly, &rdev->flags);
		     rdev = conf->mirrors[++new_disk].rdev) {

			if (rdev && rdev->in_sync)
				wonly_disk = new_disk;

			if (new_disk == conf->raid_disks - 1) {
				new_disk = wonly_disk;
L
Linus Torvalds 已提交
429 430 431 432 433 434 435 436
				break;
			}
		}
		goto rb_out;
	}


	/* make sure the disk is operational */
437 438 439 440 441 442 443 444
	for (rdev = conf->mirrors[new_disk].rdev;
	     !rdev || !rdev->in_sync ||
		     test_bit(WriteMostly, &rdev->flags);
	     rdev = conf->mirrors[new_disk].rdev) {

		if (rdev && rdev->in_sync)
			wonly_disk = new_disk;

L
Linus Torvalds 已提交
445 446 447 448
		if (new_disk <= 0)
			new_disk = conf->raid_disks;
		new_disk--;
		if (new_disk == disk) {
449 450
			new_disk = wonly_disk;
			break;
L
Linus Torvalds 已提交
451 452
		}
	}
453 454 455 456

	if (new_disk < 0)
		goto rb_out;

L
Linus Torvalds 已提交
457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476
	disk = new_disk;
	/* now disk == new_disk == starting point for search */

	/*
	 * Don't change to another disk for sequential reads:
	 */
	if (conf->next_seq_sect == this_sector)
		goto rb_out;
	if (this_sector == conf->mirrors[new_disk].head_position)
		goto rb_out;

	current_distance = abs(this_sector - conf->mirrors[disk].head_position);

	/* Find the disk whose head is closest */

	do {
		if (disk <= 0)
			disk = conf->raid_disks;
		disk--;

477 478 479 480 481
		rdev = conf->mirrors[disk].rdev;

		if (!rdev ||
		    !rdev->in_sync ||
		    test_bit(WriteMostly, &rdev->flags))
L
Linus Torvalds 已提交
482 483 484 485 486 487 488 489 490 491 492 493 494
			continue;

		if (!atomic_read(&rdev->nr_pending)) {
			new_disk = disk;
			break;
		}
		new_distance = abs(this_sector - conf->mirrors[disk].head_position);
		if (new_distance < current_distance) {
			current_distance = new_distance;
			new_disk = disk;
		}
	} while (disk != conf->last_used);

495
 rb_out:
L
Linus Torvalds 已提交
496 497 498


	if (new_disk >= 0) {
499 500 501 502 503
		rdev = conf->mirrors[new_disk].rdev;
		if (!rdev)
			goto retry;
		atomic_inc(&rdev->nr_pending);
		if (!rdev->in_sync) {
L
Linus Torvalds 已提交
504 505 506
			/* cannot risk returning a device that failed
			 * before we inc'ed nr_pending
			 */
507
			atomic_dec(&rdev->nr_pending);
L
Linus Torvalds 已提交
508 509
			goto retry;
		}
510 511
		conf->next_seq_sect = this_sector + sectors;
		conf->last_used = new_disk;
L
Linus Torvalds 已提交
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543
	}
	rcu_read_unlock();

	return new_disk;
}

static void unplug_slaves(mddev_t *mddev)
{
	conf_t *conf = mddev_to_conf(mddev);
	int i;

	rcu_read_lock();
	for (i=0; i<mddev->raid_disks; i++) {
		mdk_rdev_t *rdev = conf->mirrors[i].rdev;
		if (rdev && !rdev->faulty && atomic_read(&rdev->nr_pending)) {
			request_queue_t *r_queue = bdev_get_queue(rdev->bdev);

			atomic_inc(&rdev->nr_pending);
			rcu_read_unlock();

			if (r_queue->unplug_fn)
				r_queue->unplug_fn(r_queue);

			rdev_dec_pending(rdev, mddev);
			rcu_read_lock();
		}
	}
	rcu_read_unlock();
}

static void raid1_unplug(request_queue_t *q)
{
544 545 546 547
	mddev_t *mddev = q->queuedata;

	unplug_slaves(mddev);
	md_wakeup_thread(mddev->thread);
L
Linus Torvalds 已提交
548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589
}

static int raid1_issue_flush(request_queue_t *q, struct gendisk *disk,
			     sector_t *error_sector)
{
	mddev_t *mddev = q->queuedata;
	conf_t *conf = mddev_to_conf(mddev);
	int i, ret = 0;

	rcu_read_lock();
	for (i=0; i<mddev->raid_disks && ret == 0; i++) {
		mdk_rdev_t *rdev = conf->mirrors[i].rdev;
		if (rdev && !rdev->faulty) {
			struct block_device *bdev = rdev->bdev;
			request_queue_t *r_queue = bdev_get_queue(bdev);

			if (!r_queue->issue_flush_fn)
				ret = -EOPNOTSUPP;
			else {
				atomic_inc(&rdev->nr_pending);
				rcu_read_unlock();
				ret = r_queue->issue_flush_fn(r_queue, bdev->bd_disk,
							      error_sector);
				rdev_dec_pending(rdev, mddev);
				rcu_read_lock();
			}
		}
	}
	rcu_read_unlock();
	return ret;
}

/*
 * Throttle resync depth, so that we can both get proper overlapping of
 * requests, but are still able to handle normal requests quickly.
 */
#define RESYNC_DEPTH 32

static void device_barrier(conf_t *conf, sector_t sect)
{
	spin_lock_irq(&conf->resync_lock);
	wait_event_lock_irq(conf->wait_idle, !waitqueue_active(&conf->wait_resume),
590
			    conf->resync_lock, raid1_unplug(conf->mddev->queue));
L
Linus Torvalds 已提交
591 592 593
	
	if (!conf->barrier++) {
		wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
594
				    conf->resync_lock, raid1_unplug(conf->mddev->queue));
L
Linus Torvalds 已提交
595 596 597 598
		if (conf->nr_pending)
			BUG();
	}
	wait_event_lock_irq(conf->wait_resume, conf->barrier < RESYNC_DEPTH,
599
			    conf->resync_lock, raid1_unplug(conf->mddev->queue));
L
Linus Torvalds 已提交
600 601 602 603
	conf->next_resync = sect;
	spin_unlock_irq(&conf->resync_lock);
}

604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636
/* duplicate the data pages for behind I/O */
static struct page **alloc_behind_pages(struct bio *bio)
{
	int i;
	struct bio_vec *bvec;
	struct page **pages = kmalloc(bio->bi_vcnt * sizeof(struct page *),
					GFP_NOIO);
	if (unlikely(!pages))
		goto do_sync_io;

	memset(pages, 0, bio->bi_vcnt * sizeof(struct page *));

	bio_for_each_segment(bvec, bio, i) {
		pages[i] = alloc_page(GFP_NOIO);
		if (unlikely(!pages[i]))
			goto do_sync_io;
		memcpy(kmap(pages[i]) + bvec->bv_offset,
			kmap(bvec->bv_page) + bvec->bv_offset, bvec->bv_len);
		kunmap(pages[i]);
		kunmap(bvec->bv_page);
	}

	return pages;

do_sync_io:
	if (pages)
		for (i = 0; i < bio->bi_vcnt && pages[i]; i++)
			__free_page(pages[i]);
	kfree(pages);
	PRINTK("%dB behind alloc failed, doing sync I/O\n", bio->bi_size);
	return NULL;
}

L
Linus Torvalds 已提交
637 638 639 640 641 642 643
static int make_request(request_queue_t *q, struct bio * bio)
{
	mddev_t *mddev = q->queuedata;
	conf_t *conf = mddev_to_conf(mddev);
	mirror_info_t *mirror;
	r1bio_t *r1_bio;
	struct bio *read_bio;
644
	int i, targets = 0, disks;
L
Linus Torvalds 已提交
645
	mdk_rdev_t *rdev;
646 647 648
	struct bitmap *bitmap = mddev->bitmap;
	unsigned long flags;
	struct bio_list bl;
649
	struct page **behind_pages = NULL;
650

651 652 653 654
	if (unlikely(bio_barrier(bio))) {
		bio_endio(bio, bio->bi_size, -EOPNOTSUPP);
		return 0;
	}
L
Linus Torvalds 已提交
655 656 657 658 659 660

	/*
	 * Register the new request and wait if the reconstruction
	 * thread has put up a bar for new requests.
	 * Continue immediately if no resync is active currently.
	 */
661 662
	md_write_start(mddev, bio); /* wait on superblock update early */

L
Linus Torvalds 已提交
663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684
	spin_lock_irq(&conf->resync_lock);
	wait_event_lock_irq(conf->wait_resume, !conf->barrier, conf->resync_lock, );
	conf->nr_pending++;
	spin_unlock_irq(&conf->resync_lock);

	if (bio_data_dir(bio)==WRITE) {
		disk_stat_inc(mddev->gendisk, writes);
		disk_stat_add(mddev->gendisk, write_sectors, bio_sectors(bio));
	} else {
		disk_stat_inc(mddev->gendisk, reads);
		disk_stat_add(mddev->gendisk, read_sectors, bio_sectors(bio));
	}

	/*
	 * make_request() can abort the operation when READA is being
	 * used and no empty request is available.
	 *
	 */
	r1_bio = mempool_alloc(conf->r1bio_pool, GFP_NOIO);

	r1_bio->master_bio = bio;
	r1_bio->sectors = bio->bi_size >> 9;
685
	r1_bio->state = 0;
L
Linus Torvalds 已提交
686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
	r1_bio->mddev = mddev;
	r1_bio->sector = bio->bi_sector;

	if (bio_data_dir(bio) == READ) {
		/*
		 * read balancing logic:
		 */
		int rdisk = read_balance(conf, r1_bio);

		if (rdisk < 0) {
			/* couldn't find anywhere to read from */
			raid_end_bio_io(r1_bio);
			return 0;
		}
		mirror = conf->mirrors + rdisk;

		r1_bio->read_disk = rdisk;

		read_bio = bio_clone(bio, GFP_NOIO);

		r1_bio->bios[rdisk] = read_bio;

		read_bio->bi_sector = r1_bio->sector + mirror->rdev->data_offset;
		read_bio->bi_bdev = mirror->rdev->bdev;
		read_bio->bi_end_io = raid1_end_read_request;
		read_bio->bi_rw = READ;
		read_bio->bi_private = r1_bio;

		generic_make_request(read_bio);
		return 0;
	}

	/*
	 * WRITE:
	 */
	/* first select target devices under spinlock and
	 * inc refcount on their rdev.  Record them by setting
	 * bios[x] to bio
	 */
	disks = conf->raid_disks;
726 727 728 729 730 731 732
#if 0
	{ static int first=1;
	if (first) printk("First Write sector %llu disks %d\n",
			  (unsigned long long)r1_bio->sector, disks);
	first = 0;
	}
#endif
L
Linus Torvalds 已提交
733 734 735 736 737 738 739 740 741 742
	rcu_read_lock();
	for (i = 0;  i < disks; i++) {
		if ((rdev=conf->mirrors[i].rdev) != NULL &&
		    !rdev->faulty) {
			atomic_inc(&rdev->nr_pending);
			if (rdev->faulty) {
				atomic_dec(&rdev->nr_pending);
				r1_bio->bios[i] = NULL;
			} else
				r1_bio->bios[i] = bio;
743
			targets++;
L
Linus Torvalds 已提交
744 745 746 747 748
		} else
			r1_bio->bios[i] = NULL;
	}
	rcu_read_unlock();

749 750
	BUG_ON(targets == 0); /* we never fail the last device */

751 752 753 754 755 756
	if (targets < conf->raid_disks) {
		/* array is degraded, we will not clear the bitmap
		 * on I/O completion (see raid1_end_write_request) */
		set_bit(R1BIO_Degraded, &r1_bio->state);
	}

757 758 759 760 761 762
	/* do behind I/O ? */
	if (bitmap &&
	    atomic_read(&bitmap->behind_writes) < bitmap->max_write_behind &&
	    (behind_pages = alloc_behind_pages(bio)) != NULL)
		set_bit(R1BIO_BehindIO, &r1_bio->state);

763
	atomic_set(&r1_bio->remaining, 0);
764
	atomic_set(&r1_bio->behind_remaining, 0);
765

766
	bio_list_init(&bl);
L
Linus Torvalds 已提交
767 768 769 770 771 772 773 774 775 776 777 778 779 780
	for (i = 0; i < disks; i++) {
		struct bio *mbio;
		if (!r1_bio->bios[i])
			continue;

		mbio = bio_clone(bio, GFP_NOIO);
		r1_bio->bios[i] = mbio;

		mbio->bi_sector	= r1_bio->sector + conf->mirrors[i].rdev->data_offset;
		mbio->bi_bdev = conf->mirrors[i].rdev->bdev;
		mbio->bi_end_io	= raid1_end_write_request;
		mbio->bi_rw = WRITE;
		mbio->bi_private = r1_bio;

781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797
		if (behind_pages) {
			struct bio_vec *bvec;
			int j;

			/* Yes, I really want the '__' version so that
			 * we clear any unused pointer in the io_vec, rather
			 * than leave them unchanged.  This is important
			 * because when we come to free the pages, we won't
			 * know the originial bi_idx, so we just free
			 * them all
			 */
			__bio_for_each_segment(bvec, mbio, j, 0)
				bvec->bv_page = behind_pages[j];
			if (test_bit(WriteMostly, &conf->mirrors[i].rdev->flags))
				atomic_inc(&r1_bio->behind_remaining);
		}

L
Linus Torvalds 已提交
798 799
		atomic_inc(&r1_bio->remaining);

800
		bio_list_add(&bl, mbio);
L
Linus Torvalds 已提交
801
	}
802
	kfree(behind_pages); /* the behind pages are attached to the bios now */
L
Linus Torvalds 已提交
803

804 805
	bitmap_startwrite(bitmap, bio->bi_sector, r1_bio->sectors,
				test_bit(R1BIO_BehindIO, &r1_bio->state));
806 807 808 809 810 811 812 813 814 815 816 817
	spin_lock_irqsave(&conf->device_lock, flags);
	bio_list_merge(&conf->pending_bio_list, &bl);
	bio_list_init(&bl);

	blk_plug_device(mddev->queue);
	spin_unlock_irqrestore(&conf->device_lock, flags);

#if 0
	while ((bio = bio_list_pop(&bl)) != NULL)
		generic_make_request(bio);
#endif

L
Linus Torvalds 已提交
818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896
	return 0;
}

static void status(struct seq_file *seq, mddev_t *mddev)
{
	conf_t *conf = mddev_to_conf(mddev);
	int i;

	seq_printf(seq, " [%d/%d] [", conf->raid_disks,
						conf->working_disks);
	for (i = 0; i < conf->raid_disks; i++)
		seq_printf(seq, "%s",
			      conf->mirrors[i].rdev &&
			      conf->mirrors[i].rdev->in_sync ? "U" : "_");
	seq_printf(seq, "]");
}


static void error(mddev_t *mddev, mdk_rdev_t *rdev)
{
	char b[BDEVNAME_SIZE];
	conf_t *conf = mddev_to_conf(mddev);

	/*
	 * If it is not operational, then we have already marked it as dead
	 * else if it is the last working disks, ignore the error, let the
	 * next level up know.
	 * else mark the drive as failed
	 */
	if (rdev->in_sync
	    && conf->working_disks == 1)
		/*
		 * Don't fail the drive, act as though we were just a
		 * normal single drive
		 */
		return;
	if (rdev->in_sync) {
		mddev->degraded++;
		conf->working_disks--;
		/*
		 * if recovery is running, make sure it aborts.
		 */
		set_bit(MD_RECOVERY_ERR, &mddev->recovery);
	}
	rdev->in_sync = 0;
	rdev->faulty = 1;
	mddev->sb_dirty = 1;
	printk(KERN_ALERT "raid1: Disk failure on %s, disabling device. \n"
		"	Operation continuing on %d devices\n",
		bdevname(rdev->bdev,b), conf->working_disks);
}

static void print_conf(conf_t *conf)
{
	int i;
	mirror_info_t *tmp;

	printk("RAID1 conf printout:\n");
	if (!conf) {
		printk("(!conf)\n");
		return;
	}
	printk(" --- wd:%d rd:%d\n", conf->working_disks,
		conf->raid_disks);

	for (i = 0; i < conf->raid_disks; i++) {
		char b[BDEVNAME_SIZE];
		tmp = conf->mirrors + i;
		if (tmp->rdev)
			printk(" disk %d, wo:%d, o:%d, dev:%s\n",
				i, !tmp->rdev->in_sync, !tmp->rdev->faulty,
				bdevname(tmp->rdev->bdev,b));
	}
}

static void close_sync(conf_t *conf)
{
	spin_lock_irq(&conf->resync_lock);
	wait_event_lock_irq(conf->wait_resume, !conf->barrier,
897
			    conf->resync_lock, 	raid1_unplug(conf->mddev->queue));
L
Linus Torvalds 已提交
898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936
	spin_unlock_irq(&conf->resync_lock);

	if (conf->barrier) BUG();
	if (waitqueue_active(&conf->wait_idle)) BUG();

	mempool_destroy(conf->r1buf_pool);
	conf->r1buf_pool = NULL;
}

static int raid1_spare_active(mddev_t *mddev)
{
	int i;
	conf_t *conf = mddev->private;
	mirror_info_t *tmp;

	/*
	 * Find all failed disks within the RAID1 configuration 
	 * and mark them readable
	 */
	for (i = 0; i < conf->raid_disks; i++) {
		tmp = conf->mirrors + i;
		if (tmp->rdev 
		    && !tmp->rdev->faulty
		    && !tmp->rdev->in_sync) {
			conf->working_disks++;
			mddev->degraded--;
			tmp->rdev->in_sync = 1;
		}
	}

	print_conf(conf);
	return 0;
}


static int raid1_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
{
	conf_t *conf = mddev->private;
	int found = 0;
937
	int mirror = 0;
L
Linus Torvalds 已提交
938 939
	mirror_info_t *p;

940 941 942
	if (rdev->saved_raid_disk >= 0 &&
	    conf->mirrors[rdev->saved_raid_disk].rdev == NULL)
		mirror = rdev->saved_raid_disk;
L
Linus Torvalds 已提交
943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958
	for (mirror=0; mirror < mddev->raid_disks; mirror++)
		if ( !(p=conf->mirrors+mirror)->rdev) {

			blk_queue_stack_limits(mddev->queue,
					       rdev->bdev->bd_disk->queue);
			/* as we don't honour merge_bvec_fn, we must never risk
			 * violating it, so limit ->max_sector to one PAGE, as
			 * a one page request is never in violation.
			 */
			if (rdev->bdev->bd_disk->queue->merge_bvec_fn &&
			    mddev->queue->max_sectors > (PAGE_SIZE>>9))
				blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);

			p->head_position = 0;
			rdev->raid_disk = mirror;
			found = 1;
959 960
			if (rdev->saved_raid_disk != mirror)
				conf->fullsync = 1;
L
Linus Torvalds 已提交
961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984
			p->rdev = rdev;
			break;
		}

	print_conf(conf);
	return found;
}

static int raid1_remove_disk(mddev_t *mddev, int number)
{
	conf_t *conf = mddev->private;
	int err = 0;
	mdk_rdev_t *rdev;
	mirror_info_t *p = conf->mirrors+ number;

	print_conf(conf);
	rdev = p->rdev;
	if (rdev) {
		if (rdev->in_sync ||
		    atomic_read(&rdev->nr_pending)) {
			err = -EBUSY;
			goto abort;
		}
		p->rdev = NULL;
985
		synchronize_rcu();
L
Linus Torvalds 已提交
986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015
		if (atomic_read(&rdev->nr_pending)) {
			/* lost the race, try later */
			err = -EBUSY;
			p->rdev = rdev;
		}
	}
abort:

	print_conf(conf);
	return err;
}


static int end_sync_read(struct bio *bio, unsigned int bytes_done, int error)
{
	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
	conf_t *conf = mddev_to_conf(r1_bio->mddev);

	if (bio->bi_size)
		return 1;

	if (r1_bio->bios[r1_bio->read_disk] != bio)
		BUG();
	update_head_pos(r1_bio->read_disk, r1_bio);
	/*
	 * we have read a block, now it needs to be re-written,
	 * or re-read if the read failed.
	 * We don't do much here, just schedule handling by raid1d
	 */
1016
	if (!uptodate) {
L
Linus Torvalds 已提交
1017 1018
		md_error(r1_bio->mddev,
			 conf->mirrors[r1_bio->read_disk].rdev);
1019
	} else
L
Linus Torvalds 已提交
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042
		set_bit(R1BIO_Uptodate, &r1_bio->state);
	rdev_dec_pending(conf->mirrors[r1_bio->read_disk].rdev, conf->mddev);
	reschedule_retry(r1_bio);
	return 0;
}

static int end_sync_write(struct bio *bio, unsigned int bytes_done, int error)
{
	int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags);
	r1bio_t * r1_bio = (r1bio_t *)(bio->bi_private);
	mddev_t *mddev = r1_bio->mddev;
	conf_t *conf = mddev_to_conf(mddev);
	int i;
	int mirror=0;

	if (bio->bi_size)
		return 1;

	for (i = 0; i < conf->raid_disks; i++)
		if (r1_bio->bios[i] == bio) {
			mirror = i;
			break;
		}
1043
	if (!uptodate)
L
Linus Torvalds 已提交
1044
		md_error(mddev, conf->mirrors[mirror].rdev);
1045

L
Linus Torvalds 已提交
1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064
	update_head_pos(mirror, r1_bio);

	if (atomic_dec_and_test(&r1_bio->remaining)) {
		md_done_sync(mddev, r1_bio->sectors, uptodate);
		put_buf(r1_bio);
	}
	rdev_dec_pending(conf->mirrors[mirror].rdev, mddev);
	return 0;
}

static void sync_request_write(mddev_t *mddev, r1bio_t *r1_bio)
{
	conf_t *conf = mddev_to_conf(mddev);
	int i;
	int disks = conf->raid_disks;
	struct bio *bio, *wbio;

	bio = r1_bio->bios[r1_bio->read_disk];

1065 1066 1067
/*
	if (r1_bio->sector == 0) printk("First sync write startss\n");
*/
L
Linus Torvalds 已提交
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094
	/*
	 * schedule writes
	 */
	if (!test_bit(R1BIO_Uptodate, &r1_bio->state)) {
		/*
		 * There is no point trying a read-for-reconstruct as
		 * reconstruct is about to be aborted
		 */
		char b[BDEVNAME_SIZE];
		printk(KERN_ALERT "raid1: %s: unrecoverable I/O read error"
			" for block %llu\n",
			bdevname(bio->bi_bdev,b), 
			(unsigned long long)r1_bio->sector);
		md_done_sync(mddev, r1_bio->sectors, 0);
		put_buf(r1_bio);
		return;
	}

	atomic_set(&r1_bio->remaining, 1);
	for (i = 0; i < disks ; i++) {
		wbio = r1_bio->bios[i];
		if (wbio->bi_end_io != end_sync_write)
			continue;

		atomic_inc(&conf->mirrors[i].rdev->nr_pending);
		atomic_inc(&r1_bio->remaining);
		md_sync_acct(conf->mirrors[i].rdev->bdev, wbio->bi_size >> 9);
1095

L
Linus Torvalds 已提交
1096 1097 1098 1099
		generic_make_request(wbio);
	}

	if (atomic_dec_and_test(&r1_bio->remaining)) {
1100
		/* if we're here, all write(s) have completed, so clean up */
L
Linus Torvalds 已提交
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128
		md_done_sync(mddev, r1_bio->sectors, 1);
		put_buf(r1_bio);
	}
}

/*
 * This is a kernel thread which:
 *
 *	1.	Retries failed read operations on working mirrors.
 *	2.	Updates the raid superblock when problems encounter.
 *	3.	Performs writes following reads for array syncronising.
 */

static void raid1d(mddev_t *mddev)
{
	r1bio_t *r1_bio;
	struct bio *bio;
	unsigned long flags;
	conf_t *conf = mddev_to_conf(mddev);
	struct list_head *head = &conf->retry_list;
	int unplug=0;
	mdk_rdev_t *rdev;

	md_check_recovery(mddev);
	
	for (;;) {
		char b[BDEVNAME_SIZE];
		spin_lock_irqsave(&conf->device_lock, flags);
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148

		if (conf->pending_bio_list.head) {
			bio = bio_list_get(&conf->pending_bio_list);
			blk_remove_plug(mddev->queue);
			spin_unlock_irqrestore(&conf->device_lock, flags);
			/* flush any pending bitmap writes to disk before proceeding w/ I/O */
			if (bitmap_unplug(mddev->bitmap) != 0)
				printk("%s: bitmap file write failed!\n", mdname(mddev));

			while (bio) { /* submit pending writes */
				struct bio *next = bio->bi_next;
				bio->bi_next = NULL;
				generic_make_request(bio);
				bio = next;
			}
			unplug = 1;

			continue;
		}

L
Linus Torvalds 已提交
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 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221
		if (list_empty(head))
			break;
		r1_bio = list_entry(head->prev, r1bio_t, retry_list);
		list_del(head->prev);
		spin_unlock_irqrestore(&conf->device_lock, flags);

		mddev = r1_bio->mddev;
		conf = mddev_to_conf(mddev);
		if (test_bit(R1BIO_IsSync, &r1_bio->state)) {
			sync_request_write(mddev, r1_bio);
			unplug = 1;
		} else {
			int disk;
			bio = r1_bio->bios[r1_bio->read_disk];
			if ((disk=read_balance(conf, r1_bio)) == -1) {
				printk(KERN_ALERT "raid1: %s: unrecoverable I/O"
				       " read error for block %llu\n",
				       bdevname(bio->bi_bdev,b),
				       (unsigned long long)r1_bio->sector);
				raid_end_bio_io(r1_bio);
			} else {
				r1_bio->bios[r1_bio->read_disk] = NULL;
				r1_bio->read_disk = disk;
				bio_put(bio);
				bio = bio_clone(r1_bio->master_bio, GFP_NOIO);
				r1_bio->bios[r1_bio->read_disk] = bio;
				rdev = conf->mirrors[disk].rdev;
				if (printk_ratelimit())
					printk(KERN_ERR "raid1: %s: redirecting sector %llu to"
					       " another mirror\n",
					       bdevname(rdev->bdev,b),
					       (unsigned long long)r1_bio->sector);
				bio->bi_sector = r1_bio->sector + rdev->data_offset;
				bio->bi_bdev = rdev->bdev;
				bio->bi_end_io = raid1_end_read_request;
				bio->bi_rw = READ;
				bio->bi_private = r1_bio;
				unplug = 1;
				generic_make_request(bio);
			}
		}
	}
	spin_unlock_irqrestore(&conf->device_lock, flags);
	if (unplug)
		unplug_slaves(mddev);
}


static int init_resync(conf_t *conf)
{
	int buffs;

	buffs = RESYNC_WINDOW / RESYNC_BLOCK_SIZE;
	if (conf->r1buf_pool)
		BUG();
	conf->r1buf_pool = mempool_create(buffs, r1buf_pool_alloc, r1buf_pool_free,
					  conf->poolinfo);
	if (!conf->r1buf_pool)
		return -ENOMEM;
	conf->next_resync = 0;
	return 0;
}

/*
 * perform a "sync" on one "block"
 *
 * We need to make sure that no normal I/O request - particularly write
 * requests - conflict with active sync requests.
 *
 * This is achieved by tracking pending requests and a 'barrier' concept
 * that can be installed to exclude normal IO requests.
 */

1222
static sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
L
Linus Torvalds 已提交
1223 1224 1225 1226 1227 1228 1229 1230
{
	conf_t *conf = mddev_to_conf(mddev);
	mirror_info_t *mirror;
	r1bio_t *r1_bio;
	struct bio *bio;
	sector_t max_sector, nr_sectors;
	int disk;
	int i;
1231
	int wonly;
L
Linus Torvalds 已提交
1232
	int write_targets = 0;
1233
	int sync_blocks;
1234
	int still_degraded = 0;
L
Linus Torvalds 已提交
1235 1236

	if (!conf->r1buf_pool)
1237 1238 1239 1240
	{
/*
		printk("sync start - bitmap %p\n", mddev->bitmap);
*/
L
Linus Torvalds 已提交
1241
		if (init_resync(conf))
1242
			return 0;
1243
	}
L
Linus Torvalds 已提交
1244 1245 1246

	max_sector = mddev->size << 1;
	if (sector_nr >= max_sector) {
1247 1248 1249 1250 1251
		/* If we aborted, we need to abort the
		 * sync on the 'current' bitmap chunk (there will
		 * only be one in raid1 resync.
		 * We can find the current addess in mddev->curr_resync
		 */
1252 1253
		if (mddev->curr_resync < max_sector) /* aborted */
			bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
1254
						&sync_blocks, 1);
1255
		else /* completed sync */
1256
			conf->fullsync = 0;
1257 1258

		bitmap_close_sync(mddev->bitmap);
L
Linus Torvalds 已提交
1259 1260 1261 1262
		close_sync(conf);
		return 0;
	}

1263 1264 1265 1266
	/* before building a request, check if we can skip these blocks..
	 * This call the bitmap_start_sync doesn't actually record anything
	 */
	if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) &&
1267
	    !conf->fullsync) {
1268 1269 1270 1271
		/* We can skip this block, and probably several more */
		*skipped = 1;
		return sync_blocks;
	}
L
Linus Torvalds 已提交
1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286
	/*
	 * If there is non-resync activity waiting for us then
	 * put in a delay to throttle resync.
	 */
	if (!go_faster && waitqueue_active(&conf->wait_resume))
		msleep_interruptible(1000);
	device_barrier(conf, sector_nr + RESYNC_SECTORS);

	/*
	 * If reconstructing, and >1 working disc,
	 * could dedicate one to rebuild and others to
	 * service read requests ..
	 */
	disk = conf->last_used;
	/* make sure disk is operational */
1287
	wonly = disk;
L
Linus Torvalds 已提交
1288
	while (conf->mirrors[disk].rdev == NULL ||
1289 1290 1291 1292 1293 1294
	       !conf->mirrors[disk].rdev->in_sync ||
	       test_bit(WriteMostly, &conf->mirrors[disk].rdev->flags)
		) {
		if (conf->mirrors[disk].rdev  &&
		    conf->mirrors[disk].rdev->in_sync)
			wonly = disk;
L
Linus Torvalds 已提交
1295 1296 1297
		if (disk <= 0)
			disk = conf->raid_disks;
		disk--;
1298 1299
		if (disk == conf->last_used) {
			disk = wonly;
L
Linus Torvalds 已提交
1300
			break;
1301
		}
L
Linus Torvalds 已提交
1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316
	}
	conf->last_used = disk;
	atomic_inc(&conf->mirrors[disk].rdev->nr_pending);


	mirror = conf->mirrors + disk;

	r1_bio = mempool_alloc(conf->r1buf_pool, GFP_NOIO);

	spin_lock_irq(&conf->resync_lock);
	conf->nr_pending++;
	spin_unlock_irq(&conf->resync_lock);

	r1_bio->mddev = mddev;
	r1_bio->sector = sector_nr;
1317
	r1_bio->state = 0;
L
Linus Torvalds 已提交
1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338
	set_bit(R1BIO_IsSync, &r1_bio->state);
	r1_bio->read_disk = disk;

	for (i=0; i < conf->raid_disks; i++) {
		bio = r1_bio->bios[i];

		/* take from bio_init */
		bio->bi_next = NULL;
		bio->bi_flags |= 1 << BIO_UPTODATE;
		bio->bi_rw = 0;
		bio->bi_vcnt = 0;
		bio->bi_idx = 0;
		bio->bi_phys_segments = 0;
		bio->bi_hw_segments = 0;
		bio->bi_size = 0;
		bio->bi_end_io = NULL;
		bio->bi_private = NULL;

		if (i == disk) {
			bio->bi_rw = READ;
			bio->bi_end_io = end_sync_read;
1339 1340 1341 1342 1343 1344
		} else if (conf->mirrors[i].rdev == NULL ||
			   conf->mirrors[i].rdev->faulty) {
			still_degraded = 1;
			continue;
		} else if (!conf->mirrors[i].rdev->in_sync ||
			   sector_nr + RESYNC_SECTORS > mddev->recovery_cp) {
L
Linus Torvalds 已提交
1345 1346 1347 1348
			bio->bi_rw = WRITE;
			bio->bi_end_io = end_sync_write;
			write_targets ++;
		} else
1349
			/* no need to read or write here */
L
Linus Torvalds 已提交
1350 1351 1352 1353 1354
			continue;
		bio->bi_sector = sector_nr + conf->mirrors[i].rdev->data_offset;
		bio->bi_bdev = conf->mirrors[i].rdev->bdev;
		bio->bi_private = r1_bio;
	}
1355

L
Linus Torvalds 已提交
1356 1357 1358 1359
	if (write_targets == 0) {
		/* There is nowhere to write, so all non-sync
		 * drives must be failed - so we are finished
		 */
1360 1361
		sector_t rv = max_sector - sector_nr;
		*skipped = 1;
L
Linus Torvalds 已提交
1362 1363 1364 1365 1366 1367
		put_buf(r1_bio);
		rdev_dec_pending(conf->mirrors[disk].rdev, mddev);
		return rv;
	}

	nr_sectors = 0;
1368
	sync_blocks = 0;
L
Linus Torvalds 已提交
1369 1370 1371 1372 1373 1374 1375
	do {
		struct page *page;
		int len = PAGE_SIZE;
		if (sector_nr + (len>>9) > max_sector)
			len = (max_sector - sector_nr) << 9;
		if (len == 0)
			break;
1376 1377
		if (sync_blocks == 0) {
			if (!bitmap_start_sync(mddev->bitmap, sector_nr,
1378
					&sync_blocks, still_degraded) &&
1379 1380 1381 1382 1383 1384
					!conf->fullsync)
				break;
			if (sync_blocks < (PAGE_SIZE>>9))
				BUG();
			if (len > (sync_blocks<<9))
				len = sync_blocks<<9;
1385
		}
1386

L
Linus Torvalds 已提交
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396
		for (i=0 ; i < conf->raid_disks; i++) {
			bio = r1_bio->bios[i];
			if (bio->bi_end_io) {
				page = r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page;
				if (bio_add_page(bio, page, len, 0) == 0) {
					/* stop here */
					r1_bio->bios[0]->bi_io_vec[bio->bi_vcnt].bv_page = page;
					while (i > 0) {
						i--;
						bio = r1_bio->bios[i];
1397 1398
						if (bio->bi_end_io==NULL)
							continue;
L
Linus Torvalds 已提交
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409
						/* remove last page from this bio */
						bio->bi_vcnt--;
						bio->bi_size -= len;
						bio->bi_flags &= ~(1<< BIO_SEG_VALID);
					}
					goto bio_full;
				}
			}
		}
		nr_sectors += len>>9;
		sector_nr += len>>9;
1410
		sync_blocks -= (len>>9);
L
Linus Torvalds 已提交
1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498
	} while (r1_bio->bios[disk]->bi_vcnt < RESYNC_PAGES);
 bio_full:
	bio = r1_bio->bios[disk];
	r1_bio->sectors = nr_sectors;

	md_sync_acct(mirror->rdev->bdev, nr_sectors);

	generic_make_request(bio);

	return nr_sectors;
}

static int run(mddev_t *mddev)
{
	conf_t *conf;
	int i, j, disk_idx;
	mirror_info_t *disk;
	mdk_rdev_t *rdev;
	struct list_head *tmp;

	if (mddev->level != 1) {
		printk("raid1: %s: raid level not set to mirroring (%d)\n",
		       mdname(mddev), mddev->level);
		goto out;
	}
	/*
	 * copy the already verified devices into our private RAID1
	 * bookkeeping area. [whatever we allocate in run(),
	 * should be freed in stop()]
	 */
	conf = kmalloc(sizeof(conf_t), GFP_KERNEL);
	mddev->private = conf;
	if (!conf)
		goto out_no_mem;

	memset(conf, 0, sizeof(*conf));
	conf->mirrors = kmalloc(sizeof(struct mirror_info)*mddev->raid_disks, 
				 GFP_KERNEL);
	if (!conf->mirrors)
		goto out_no_mem;

	memset(conf->mirrors, 0, sizeof(struct mirror_info)*mddev->raid_disks);

	conf->poolinfo = kmalloc(sizeof(*conf->poolinfo), GFP_KERNEL);
	if (!conf->poolinfo)
		goto out_no_mem;
	conf->poolinfo->mddev = mddev;
	conf->poolinfo->raid_disks = mddev->raid_disks;
	conf->r1bio_pool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc,
					  r1bio_pool_free,
					  conf->poolinfo);
	if (!conf->r1bio_pool)
		goto out_no_mem;

	ITERATE_RDEV(mddev, rdev, tmp) {
		disk_idx = rdev->raid_disk;
		if (disk_idx >= mddev->raid_disks
		    || disk_idx < 0)
			continue;
		disk = conf->mirrors + disk_idx;

		disk->rdev = rdev;

		blk_queue_stack_limits(mddev->queue,
				       rdev->bdev->bd_disk->queue);
		/* as we don't honour merge_bvec_fn, we must never risk
		 * violating it, so limit ->max_sector to one PAGE, as
		 * a one page request is never in violation.
		 */
		if (rdev->bdev->bd_disk->queue->merge_bvec_fn &&
		    mddev->queue->max_sectors > (PAGE_SIZE>>9))
			blk_queue_max_sectors(mddev->queue, PAGE_SIZE>>9);

		disk->head_position = 0;
		if (!rdev->faulty && rdev->in_sync)
			conf->working_disks++;
	}
	conf->raid_disks = mddev->raid_disks;
	conf->mddev = mddev;
	spin_lock_init(&conf->device_lock);
	INIT_LIST_HEAD(&conf->retry_list);
	if (conf->working_disks == 1)
		mddev->recovery_cp = MaxSector;

	spin_lock_init(&conf->resync_lock);
	init_waitqueue_head(&conf->wait_idle);
	init_waitqueue_head(&conf->wait_resume);

1499 1500 1501
	bio_list_init(&conf->pending_bio_list);
	bio_list_init(&conf->flushing_bio_list);

L
Linus Torvalds 已提交
1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529
	if (!conf->working_disks) {
		printk(KERN_ERR "raid1: no operational mirrors for %s\n",
			mdname(mddev));
		goto out_free_conf;
	}

	mddev->degraded = 0;
	for (i = 0; i < conf->raid_disks; i++) {

		disk = conf->mirrors + i;

		if (!disk->rdev) {
			disk->head_position = 0;
			mddev->degraded++;
		}
	}

	/*
	 * find the first working one and use it as a starting point
	 * to read balancing.
	 */
	for (j = 0; j < conf->raid_disks &&
		     (!conf->mirrors[j].rdev ||
		      !conf->mirrors[j].rdev->in_sync) ; j++)
		/* nothing */;
	conf->last_used = j;


1530 1531 1532 1533 1534 1535
	mddev->thread = md_register_thread(raid1d, mddev, "%s_raid1");
	if (!mddev->thread) {
		printk(KERN_ERR
		       "raid1: couldn't allocate thread for %s\n",
		       mdname(mddev));
		goto out_free_conf;
L
Linus Torvalds 已提交
1536
	}
1537 1538
	if (mddev->bitmap) mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;

L
Linus Torvalds 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547
	printk(KERN_INFO 
		"raid1: raid set %s active with %d out of %d mirrors\n",
		mdname(mddev), mddev->raid_disks - mddev->degraded, 
		mddev->raid_disks);
	/*
	 * Ok, everything is just fine now
	 */
	mddev->array_size = mddev->size;

1548 1549 1550
	mddev->queue->unplug_fn = raid1_unplug;
	mddev->queue->issue_flush_fn = raid1_issue_flush;

L
Linus Torvalds 已提交
1551 1552 1553 1554 1555 1556 1557 1558 1559 1560
	return 0;

out_no_mem:
	printk(KERN_ERR "raid1: couldn't allocate memory for %s\n",
	       mdname(mddev));

out_free_conf:
	if (conf) {
		if (conf->r1bio_pool)
			mempool_destroy(conf->r1bio_pool);
1561 1562
		kfree(conf->mirrors);
		kfree(conf->poolinfo);
L
Linus Torvalds 已提交
1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
		kfree(conf);
		mddev->private = NULL;
	}
out:
	return -EIO;
}

static int stop(mddev_t *mddev)
{
	conf_t *conf = mddev_to_conf(mddev);
1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
	struct bitmap *bitmap = mddev->bitmap;
	int behind_wait = 0;

	/* wait for behind writes to complete */
	while (bitmap && atomic_read(&bitmap->behind_writes) > 0) {
		behind_wait++;
		printk(KERN_INFO "raid1: behind writes in progress on device %s, waiting to stop (%d)\n", mdname(mddev), behind_wait);
		set_current_state(TASK_UNINTERRUPTIBLE);
		schedule_timeout(HZ); /* wait a second */
		/* need to kick something here to make sure I/O goes? */
	}
L
Linus Torvalds 已提交
1584 1585 1586 1587 1588 1589

	md_unregister_thread(mddev->thread);
	mddev->thread = NULL;
	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
	if (conf->r1bio_pool)
		mempool_destroy(conf->r1bio_pool);
1590 1591
	kfree(conf->mirrors);
	kfree(conf->poolinfo);
L
Linus Torvalds 已提交
1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613
	kfree(conf);
	mddev->private = NULL;
	return 0;
}

static int raid1_resize(mddev_t *mddev, sector_t sectors)
{
	/* no resync is happening, and there is enough space
	 * on all devices, so we can resize.
	 * We need to make sure resync covers any new space.
	 * If the array is shrinking we should possibly wait until
	 * any io in the removed space completes, but it hardly seems
	 * worth it.
	 */
	mddev->array_size = sectors>>1;
	set_capacity(mddev->gendisk, mddev->array_size << 1);
	mddev->changed = 1;
	if (mddev->array_size > mddev->size && mddev->recovery_cp == MaxSector) {
		mddev->recovery_cp = mddev->size << 1;
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	}
	mddev->size = mddev->array_size;
1614
	mddev->resync_max_sectors = sectors;
L
Linus Torvalds 已提交
1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626
	return 0;
}

static int raid1_reshape(mddev_t *mddev, int raid_disks)
{
	/* We need to:
	 * 1/ resize the r1bio_pool
	 * 2/ resize conf->mirrors
	 *
	 * We allocate a new r1bio_pool if we can.
	 * Then raise a device barrier and wait until all IO stops.
	 * Then resize conf->mirrors and swap in the new r1bio pool.
1627 1628 1629
	 *
	 * At the same time, we "pack" the devices so that all the missing
	 * devices have the higher raid_disk numbers.
L
Linus Torvalds 已提交
1630 1631 1632 1633 1634
	 */
	mempool_t *newpool, *oldpool;
	struct pool_info *newpoolinfo;
	mirror_info_t *newmirrors;
	conf_t *conf = mddev_to_conf(mddev);
1635
	int cnt;
L
Linus Torvalds 已提交
1636

1637
	int d, d2;
L
Linus Torvalds 已提交
1638

1639 1640 1641 1642 1643 1644
	if (raid_disks < conf->raid_disks) {
		cnt=0;
		for (d= 0; d < conf->raid_disks; d++)
			if (conf->mirrors[d].rdev)
				cnt++;
		if (cnt > raid_disks)
L
Linus Torvalds 已提交
1645
			return -EBUSY;
1646
	}
L
Linus Torvalds 已提交
1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670

	newpoolinfo = kmalloc(sizeof(*newpoolinfo), GFP_KERNEL);
	if (!newpoolinfo)
		return -ENOMEM;
	newpoolinfo->mddev = mddev;
	newpoolinfo->raid_disks = raid_disks;

	newpool = mempool_create(NR_RAID1_BIOS, r1bio_pool_alloc,
				 r1bio_pool_free, newpoolinfo);
	if (!newpool) {
		kfree(newpoolinfo);
		return -ENOMEM;
	}
	newmirrors = kmalloc(sizeof(struct mirror_info) * raid_disks, GFP_KERNEL);
	if (!newmirrors) {
		kfree(newpoolinfo);
		mempool_destroy(newpool);
		return -ENOMEM;
	}
	memset(newmirrors, 0, sizeof(struct mirror_info)*raid_disks);

	spin_lock_irq(&conf->resync_lock);
	conf->barrier++;
	wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
1671
			    conf->resync_lock, raid1_unplug(mddev->queue));
L
Linus Torvalds 已提交
1672 1673 1674 1675 1676
	spin_unlock_irq(&conf->resync_lock);

	/* ok, everything is stopped */
	oldpool = conf->r1bio_pool;
	conf->r1bio_pool = newpool;
1677 1678 1679 1680 1681 1682

	for (d=d2=0; d < conf->raid_disks; d++)
		if (conf->mirrors[d].rdev) {
			conf->mirrors[d].rdev->raid_disk = d2;
			newmirrors[d2++].rdev = conf->mirrors[d].rdev;
		}
L
Linus Torvalds 已提交
1683 1684 1685 1686 1687 1688 1689 1690
	kfree(conf->mirrors);
	conf->mirrors = newmirrors;
	kfree(conf->poolinfo);
	conf->poolinfo = newpoolinfo;

	mddev->degraded += (raid_disks - conf->raid_disks);
	conf->raid_disks = mddev->raid_disks = raid_disks;

1691
	conf->last_used = 0; /* just make sure it is in-range */
L
Linus Torvalds 已提交
1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705
	spin_lock_irq(&conf->resync_lock);
	conf->barrier--;
	spin_unlock_irq(&conf->resync_lock);
	wake_up(&conf->wait_resume);
	wake_up(&conf->wait_idle);


	set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	md_wakeup_thread(mddev->thread);

	mempool_destroy(oldpool);
	return 0;
}

1706
static void raid1_quiesce(mddev_t *mddev, int state)
1707 1708 1709 1710
{
	conf_t *conf = mddev_to_conf(mddev);

	switch(state) {
1711
	case 1:
1712 1713 1714 1715 1716 1717
		spin_lock_irq(&conf->resync_lock);
		conf->barrier++;
		wait_event_lock_irq(conf->wait_idle, !conf->nr_pending,
				    conf->resync_lock, raid1_unplug(mddev->queue));
		spin_unlock_irq(&conf->resync_lock);
		break;
1718
	case 0:
1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734
		spin_lock_irq(&conf->resync_lock);
		conf->barrier--;
		spin_unlock_irq(&conf->resync_lock);
		wake_up(&conf->wait_resume);
		wake_up(&conf->wait_idle);
		break;
	}
	if (mddev->thread) {
		if (mddev->bitmap)
			mddev->thread->timeout = mddev->bitmap->daemon_sleep * HZ;
		else
			mddev->thread->timeout = MAX_SCHEDULE_TIMEOUT;
		md_wakeup_thread(mddev->thread);
	}
}

L
Linus Torvalds 已提交
1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750

static mdk_personality_t raid1_personality =
{
	.name		= "raid1",
	.owner		= THIS_MODULE,
	.make_request	= make_request,
	.run		= run,
	.stop		= stop,
	.status		= status,
	.error_handler	= error,
	.hot_add_disk	= raid1_add_disk,
	.hot_remove_disk= raid1_remove_disk,
	.spare_active	= raid1_spare_active,
	.sync_request	= sync_request,
	.resize		= raid1_resize,
	.reshape	= raid1_reshape,
1751
	.quiesce	= raid1_quiesce,
L
Linus Torvalds 已提交
1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767
};

static int __init raid_init(void)
{
	return register_md_personality(RAID1, &raid1_personality);
}

static void raid_exit(void)
{
	unregister_md_personality(RAID1);
}

module_init(raid_init);
module_exit(raid_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-3"); /* RAID1 */