raid5.c 134 KB
Newer Older
Linus Torvalds's avatar
Linus Torvalds committed
1 2 3 4
/*
 * raid5.c : Multiple Devices driver for Linux
 *	   Copyright (C) 1996, 1997 Ingo Molnar, Miguel de Icaza, Gadi Oxman
 *	   Copyright (C) 1999, 2000 Ingo Molnar
5
 *	   Copyright (C) 2002, 2003 H. Peter Anvin
Linus Torvalds's avatar
Linus Torvalds committed
6
 *
7 8 9
 * RAID-4/5/6 management functions.
 * Thanks to Penguin Computing for making the RAID-6 development possible
 * by donating a test server!
Linus Torvalds's avatar
Linus Torvalds committed
10 11 12 13 14 15 16 17 18 19 20
 *
 * 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.
 */

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
/*
 * BITMAP UNPLUGGING:
 *
 * The sequencing for updating the bitmap reliably is a little
 * subtle (and I got it wrong the first time) so it deserves some
 * explanation.
 *
 * We group bitmap updates into batches.  Each batch has a number.
 * We may write out several batches at once, but that isn't very important.
 * conf->bm_write is the number of the last batch successfully written.
 * conf->bm_flush is the number of the last batch that was closed to
 *    new additions.
 * When we discover that we will need to write to any block in a stripe
 * (in add_stripe_bio) we update the in-memory bitmap and record in sh->bm_seq
 * the number of the batch it will be in. This is bm_flush+1.
 * When we are ready to do a write, if that batch hasn't been written yet,
 *   we plug the array and queue the stripe for later.
 * When an unplug happens, we increment bm_flush, thus closing the current
 *   batch.
 * When we notice that bm_flush > bm_write, we write out all pending updates
 * to the bitmap, and advance bm_write to where bm_flush was.
 * This may occasionally write a bit out twice, but is sure never to
 * miss any bits.
 */
Linus Torvalds's avatar
Linus Torvalds committed
45 46 47 48 49

#include <linux/module.h>
#include <linux/slab.h>
#include <linux/highmem.h>
#include <linux/bitops.h>
50
#include <linux/kthread.h>
Linus Torvalds's avatar
Linus Torvalds committed
51
#include <asm/atomic.h>
52
#include "raid6.h"
Linus Torvalds's avatar
Linus Torvalds committed
53

54
#include <linux/raid/bitmap.h>
55
#include <linux/async_tx.h>
56

Linus Torvalds's avatar
Linus Torvalds committed
57 58 59 60 61 62 63 64 65
/*
 * Stripe cache
 */

#define NR_STRIPES		256
#define STRIPE_SIZE		PAGE_SIZE
#define STRIPE_SHIFT		(PAGE_SHIFT - 9)
#define STRIPE_SECTORS		(STRIPE_SIZE>>9)
#define	IO_THRESHOLD		1
66
#define NR_HASH			(PAGE_SIZE / sizeof(struct hlist_head))
Linus Torvalds's avatar
Linus Torvalds committed
67 68
#define HASH_MASK		(NR_HASH - 1)

69
#define stripe_hash(conf, sect)	(&((conf)->stripe_hashtbl[((sect) >> STRIPE_SHIFT) & HASH_MASK]))
Linus Torvalds's avatar
Linus Torvalds committed
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90

/* bio's attached to a stripe+device for I/O are linked together in bi_sector
 * order without overlap.  There may be several bio's per stripe+device, and
 * a bio could span several devices.
 * When walking this list for a particular stripe+device, we must never proceed
 * beyond a bio that extends past this device, as the next bio might no longer
 * be valid.
 * This macro is used to determine the 'next' bio in the list, given the sector
 * of the current stripe+device
 */
#define r5_next_bio(bio, sect) ( ( (bio)->bi_sector + ((bio)->bi_size>>9) < sect + STRIPE_SECTORS) ? (bio)->bi_next : NULL)
/*
 * The following can be used to debug the driver
 */
#define RAID5_PARANOIA	1
#if RAID5_PARANOIA && defined(CONFIG_SMP)
# define CHECK_DEVLOCK() assert_spin_locked(&conf->device_lock)
#else
# define CHECK_DEVLOCK()
#endif

91
#ifdef DEBUG
Linus Torvalds's avatar
Linus Torvalds committed
92 93 94 95
#define inline
#define __inline__
#endif

96 97 98 99 100 101 102 103 104 105
#if !RAID6_USE_EMPTY_ZERO_PAGE
/* In .bss so it's zeroed */
const char raid6_empty_zero_page[PAGE_SIZE] __attribute__((aligned(256)));
#endif

static inline int raid6_next_disk(int disk, int raid_disks)
{
	disk++;
	return (disk < raid_disks) ? disk : 0;
}
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122

static void return_io(struct bio *return_bi)
{
	struct bio *bi = return_bi;
	while (bi) {
		int bytes = bi->bi_size;

		return_bi = bi->bi_next;
		bi->bi_next = NULL;
		bi->bi_size = 0;
		bi->bi_end_io(bi, bytes,
			      test_bit(BIO_UPTODATE, &bi->bi_flags)
			        ? 0 : -EIO);
		bi = return_bi;
	}
}

Linus Torvalds's avatar
Linus Torvalds committed
123 124
static void print_raid5_conf (raid5_conf_t *conf);

125
static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh)
Linus Torvalds's avatar
Linus Torvalds committed
126 127
{
	if (atomic_dec_and_test(&sh->count)) {
128 129
		BUG_ON(!list_empty(&sh->lru));
		BUG_ON(atomic_read(&conf->active_stripes)==0);
Linus Torvalds's avatar
Linus Torvalds committed
130
		if (test_bit(STRIPE_HANDLE, &sh->state)) {
131
			if (test_bit(STRIPE_DELAYED, &sh->state)) {
Linus Torvalds's avatar
Linus Torvalds committed
132
				list_add_tail(&sh->lru, &conf->delayed_list);
133 134
				blk_plug_device(conf->mddev->queue);
			} else if (test_bit(STRIPE_BIT_DELAY, &sh->state) &&
135
				   sh->bm_seq - conf->seq_write > 0) {
136
				list_add_tail(&sh->lru, &conf->bitmap_list);
137 138
				blk_plug_device(conf->mddev->queue);
			} else {
139
				clear_bit(STRIPE_BIT_DELAY, &sh->state);
Linus Torvalds's avatar
Linus Torvalds committed
140
				list_add_tail(&sh->lru, &conf->handle_list);
141
			}
Linus Torvalds's avatar
Linus Torvalds committed
142 143
			md_wakeup_thread(conf->mddev->thread);
		} else {
144
			BUG_ON(sh->ops.pending);
Linus Torvalds's avatar
Linus Torvalds committed
145 146 147 148 149 150
			if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
				atomic_dec(&conf->preread_active_stripes);
				if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD)
					md_wakeup_thread(conf->mddev->thread);
			}
			atomic_dec(&conf->active_stripes);
151 152
			if (!test_bit(STRIPE_EXPANDING, &sh->state)) {
				list_add_tail(&sh->lru, &conf->inactive_list);
Linus Torvalds's avatar
Linus Torvalds committed
153
				wake_up(&conf->wait_for_stripe);
154 155
				if (conf->retry_read_aligned)
					md_wakeup_thread(conf->mddev->thread);
156
			}
Linus Torvalds's avatar
Linus Torvalds committed
157 158 159 160 161 162 163
		}
	}
}
static void release_stripe(struct stripe_head *sh)
{
	raid5_conf_t *conf = sh->raid_conf;
	unsigned long flags;
164

Linus Torvalds's avatar
Linus Torvalds committed
165 166 167 168 169
	spin_lock_irqsave(&conf->device_lock, flags);
	__release_stripe(conf, sh);
	spin_unlock_irqrestore(&conf->device_lock, flags);
}

170
static inline void remove_hash(struct stripe_head *sh)
Linus Torvalds's avatar
Linus Torvalds committed
171
{
172 173
	pr_debug("remove_hash(), stripe %llu\n",
		(unsigned long long)sh->sector);
Linus Torvalds's avatar
Linus Torvalds committed
174

175
	hlist_del_init(&sh->hash);
Linus Torvalds's avatar
Linus Torvalds committed
176 177
}

178
static inline void insert_hash(raid5_conf_t *conf, struct stripe_head *sh)
Linus Torvalds's avatar
Linus Torvalds committed
179
{
180
	struct hlist_head *hp = stripe_hash(conf, sh->sector);
Linus Torvalds's avatar
Linus Torvalds committed
181

182 183
	pr_debug("insert_hash(), stripe %llu\n",
		(unsigned long long)sh->sector);
Linus Torvalds's avatar
Linus Torvalds committed
184 185

	CHECK_DEVLOCK();
186
	hlist_add_head(&sh->hash, hp);
Linus Torvalds's avatar
Linus Torvalds committed
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
}


/* find an idle stripe, make sure it is unhashed, and return it. */
static struct stripe_head *get_free_stripe(raid5_conf_t *conf)
{
	struct stripe_head *sh = NULL;
	struct list_head *first;

	CHECK_DEVLOCK();
	if (list_empty(&conf->inactive_list))
		goto out;
	first = conf->inactive_list.next;
	sh = list_entry(first, struct stripe_head, lru);
	list_del_init(first);
	remove_hash(sh);
	atomic_inc(&conf->active_stripes);
out:
	return sh;
}

static void shrink_buffers(struct stripe_head *sh, int num)
{
	struct page *p;
	int i;

	for (i=0; i<num ; i++) {
		p = sh->dev[i].page;
		if (!p)
			continue;
		sh->dev[i].page = NULL;
218
		put_page(p);
Linus Torvalds's avatar
Linus Torvalds committed
219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
	}
}

static int grow_buffers(struct stripe_head *sh, int num)
{
	int i;

	for (i=0; i<num; i++) {
		struct page *page;

		if (!(page = alloc_page(GFP_KERNEL))) {
			return 1;
		}
		sh->dev[i].page = page;
	}
	return 0;
}

static void raid5_build_block (struct stripe_head *sh, int i);

239
static void init_stripe(struct stripe_head *sh, sector_t sector, int pd_idx, int disks)
Linus Torvalds's avatar
Linus Torvalds committed
240 241
{
	raid5_conf_t *conf = sh->raid_conf;
242
	int i;
Linus Torvalds's avatar
Linus Torvalds committed
243

244 245
	BUG_ON(atomic_read(&sh->count) != 0);
	BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));
246 247
	BUG_ON(sh->ops.pending || sh->ops.ack || sh->ops.complete);

Linus Torvalds's avatar
Linus Torvalds committed
248
	CHECK_DEVLOCK();
249
	pr_debug("init_stripe called, stripe %llu\n",
Linus Torvalds's avatar
Linus Torvalds committed
250 251 252
		(unsigned long long)sh->sector);

	remove_hash(sh);
253

Linus Torvalds's avatar
Linus Torvalds committed
254 255 256 257
	sh->sector = sector;
	sh->pd_idx = pd_idx;
	sh->state = 0;

258 259 260
	sh->disks = disks;

	for (i = sh->disks; i--; ) {
Linus Torvalds's avatar
Linus Torvalds committed
261 262
		struct r5dev *dev = &sh->dev[i];

263
		if (dev->toread || dev->read || dev->towrite || dev->written ||
Linus Torvalds's avatar
Linus Torvalds committed
264
		    test_bit(R5_LOCKED, &dev->flags)) {
265
			printk(KERN_ERR "sector=%llx i=%d %p %p %p %p %d\n",
Linus Torvalds's avatar
Linus Torvalds committed
266
			       (unsigned long long)sh->sector, i, dev->toread,
267
			       dev->read, dev->towrite, dev->written,
Linus Torvalds's avatar
Linus Torvalds committed
268 269 270 271 272 273 274 275 276
			       test_bit(R5_LOCKED, &dev->flags));
			BUG();
		}
		dev->flags = 0;
		raid5_build_block(sh, i);
	}
	insert_hash(conf, sh);
}

277
static struct stripe_head *__find_stripe(raid5_conf_t *conf, sector_t sector, int disks)
Linus Torvalds's avatar
Linus Torvalds committed
278 279
{
	struct stripe_head *sh;
280
	struct hlist_node *hn;
Linus Torvalds's avatar
Linus Torvalds committed
281 282

	CHECK_DEVLOCK();
283
	pr_debug("__find_stripe, sector %llu\n", (unsigned long long)sector);
284
	hlist_for_each_entry(sh, hn, stripe_hash(conf, sector), hash)
285
		if (sh->sector == sector && sh->disks == disks)
Linus Torvalds's avatar
Linus Torvalds committed
286
			return sh;
287
	pr_debug("__stripe %llu not in cache\n", (unsigned long long)sector);
Linus Torvalds's avatar
Linus Torvalds committed
288 289 290 291
	return NULL;
}

static void unplug_slaves(mddev_t *mddev);
292
static void raid5_unplug_device(struct request_queue *q);
Linus Torvalds's avatar
Linus Torvalds committed
293

294 295
static struct stripe_head *get_active_stripe(raid5_conf_t *conf, sector_t sector, int disks,
					     int pd_idx, int noblock)
Linus Torvalds's avatar
Linus Torvalds committed
296 297 298
{
	struct stripe_head *sh;

299
	pr_debug("get_stripe, sector %llu\n", (unsigned long long)sector);
Linus Torvalds's avatar
Linus Torvalds committed
300 301 302 303

	spin_lock_irq(&conf->device_lock);

	do {
304 305 306
		wait_event_lock_irq(conf->wait_for_stripe,
				    conf->quiesce == 0,
				    conf->device_lock, /* nothing */);
307
		sh = __find_stripe(conf, sector, disks);
Linus Torvalds's avatar
Linus Torvalds committed
308 309 310 311 312 313 314 315 316
		if (!sh) {
			if (!conf->inactive_blocked)
				sh = get_free_stripe(conf);
			if (noblock && sh == NULL)
				break;
			if (!sh) {
				conf->inactive_blocked = 1;
				wait_event_lock_irq(conf->wait_for_stripe,
						    !list_empty(&conf->inactive_list) &&
317 318
						    (atomic_read(&conf->active_stripes)
						     < (conf->max_nr_stripes *3/4)
Linus Torvalds's avatar
Linus Torvalds committed
319 320
						     || !conf->inactive_blocked),
						    conf->device_lock,
321
						    raid5_unplug_device(conf->mddev->queue)
Linus Torvalds's avatar
Linus Torvalds committed
322 323 324
					);
				conf->inactive_blocked = 0;
			} else
325
				init_stripe(sh, sector, pd_idx, disks);
Linus Torvalds's avatar
Linus Torvalds committed
326 327
		} else {
			if (atomic_read(&sh->count)) {
328
			  BUG_ON(!list_empty(&sh->lru));
Linus Torvalds's avatar
Linus Torvalds committed
329 330 331
			} else {
				if (!test_bit(STRIPE_HANDLE, &sh->state))
					atomic_inc(&conf->active_stripes);
332 333
				if (list_empty(&sh->lru) &&
				    !test_bit(STRIPE_EXPANDING, &sh->state))
334 335
					BUG();
				list_del_init(&sh->lru);
Linus Torvalds's avatar
Linus Torvalds committed
336 337 338 339 340 341 342 343 344 345 346
			}
		}
	} while (sh == NULL);

	if (sh)
		atomic_inc(&sh->count);

	spin_unlock_irq(&conf->device_lock);
	return sh;
}

347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379
/* test_and_ack_op() ensures that we only dequeue an operation once */
#define test_and_ack_op(op, pend) \
do {							\
	if (test_bit(op, &sh->ops.pending) &&		\
		!test_bit(op, &sh->ops.complete)) {	\
		if (test_and_set_bit(op, &sh->ops.ack)) \
			clear_bit(op, &pend);		\
		else					\
			ack++;				\
	} else						\
		clear_bit(op, &pend);			\
} while (0)

/* find new work to run, do not resubmit work that is already
 * in flight
 */
static unsigned long get_stripe_work(struct stripe_head *sh)
{
	unsigned long pending;
	int ack = 0;

	pending = sh->ops.pending;

	test_and_ack_op(STRIPE_OP_BIOFILL, pending);
	test_and_ack_op(STRIPE_OP_COMPUTE_BLK, pending);
	test_and_ack_op(STRIPE_OP_PREXOR, pending);
	test_and_ack_op(STRIPE_OP_BIODRAIN, pending);
	test_and_ack_op(STRIPE_OP_POSTXOR, pending);
	test_and_ack_op(STRIPE_OP_CHECK, pending);
	if (test_and_clear_bit(STRIPE_OP_IO, &sh->ops.pending))
		ack++;

	sh->ops.count -= ack;
380 381 382 383 384 385
	if (unlikely(sh->ops.count < 0)) {
		printk(KERN_ERR "pending: %#lx ops.pending: %#lx ops.ack: %#lx "
			"ops.complete: %#lx\n", pending, sh->ops.pending,
			sh->ops.ack, sh->ops.complete);
		BUG();
	}
386 387 388 389

	return pending;
}

390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500
static int
raid5_end_read_request(struct bio *bi, unsigned int bytes_done, int error);
static int
raid5_end_write_request (struct bio *bi, unsigned int bytes_done, int error);

static void ops_run_io(struct stripe_head *sh)
{
	raid5_conf_t *conf = sh->raid_conf;
	int i, disks = sh->disks;

	might_sleep();

	for (i = disks; i--; ) {
		int rw;
		struct bio *bi;
		mdk_rdev_t *rdev;
		if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
			rw = WRITE;
		else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
			rw = READ;
		else
			continue;

		bi = &sh->dev[i].req;

		bi->bi_rw = rw;
		if (rw == WRITE)
			bi->bi_end_io = raid5_end_write_request;
		else
			bi->bi_end_io = raid5_end_read_request;

		rcu_read_lock();
		rdev = rcu_dereference(conf->disks[i].rdev);
		if (rdev && test_bit(Faulty, &rdev->flags))
			rdev = NULL;
		if (rdev)
			atomic_inc(&rdev->nr_pending);
		rcu_read_unlock();

		if (rdev) {
			if (test_bit(STRIPE_SYNCING, &sh->state) ||
				test_bit(STRIPE_EXPAND_SOURCE, &sh->state) ||
				test_bit(STRIPE_EXPAND_READY, &sh->state))
				md_sync_acct(rdev->bdev, STRIPE_SECTORS);

			bi->bi_bdev = rdev->bdev;
			pr_debug("%s: for %llu schedule op %ld on disc %d\n",
				__FUNCTION__, (unsigned long long)sh->sector,
				bi->bi_rw, i);
			atomic_inc(&sh->count);
			bi->bi_sector = sh->sector + rdev->data_offset;
			bi->bi_flags = 1 << BIO_UPTODATE;
			bi->bi_vcnt = 1;
			bi->bi_max_vecs = 1;
			bi->bi_idx = 0;
			bi->bi_io_vec = &sh->dev[i].vec;
			bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
			bi->bi_io_vec[0].bv_offset = 0;
			bi->bi_size = STRIPE_SIZE;
			bi->bi_next = NULL;
			if (rw == WRITE &&
			    test_bit(R5_ReWrite, &sh->dev[i].flags))
				atomic_add(STRIPE_SECTORS,
					&rdev->corrected_errors);
			generic_make_request(bi);
		} else {
			if (rw == WRITE)
				set_bit(STRIPE_DEGRADED, &sh->state);
			pr_debug("skip op %ld on disc %d for sector %llu\n",
				bi->bi_rw, i, (unsigned long long)sh->sector);
			clear_bit(R5_LOCKED, &sh->dev[i].flags);
			set_bit(STRIPE_HANDLE, &sh->state);
		}
	}
}

static struct dma_async_tx_descriptor *
async_copy_data(int frombio, struct bio *bio, struct page *page,
	sector_t sector, struct dma_async_tx_descriptor *tx)
{
	struct bio_vec *bvl;
	struct page *bio_page;
	int i;
	int page_offset;

	if (bio->bi_sector >= sector)
		page_offset = (signed)(bio->bi_sector - sector) * 512;
	else
		page_offset = (signed)(sector - bio->bi_sector) * -512;
	bio_for_each_segment(bvl, bio, i) {
		int len = bio_iovec_idx(bio, i)->bv_len;
		int clen;
		int b_offset = 0;

		if (page_offset < 0) {
			b_offset = -page_offset;
			page_offset += b_offset;
			len -= b_offset;
		}

		if (len > 0 && page_offset + len > STRIPE_SIZE)
			clen = STRIPE_SIZE - page_offset;
		else
			clen = len;

		if (clen > 0) {
			b_offset += bio_iovec_idx(bio, i)->bv_offset;
			bio_page = bio_iovec_idx(bio, i)->bv_page;
			if (frombio)
				tx = async_memcpy(page, bio_page, page_offset,
					b_offset, clen,
501
					ASYNC_TX_DEP_ACK,
502 503 504 505
					tx, NULL, NULL);
			else
				tx = async_memcpy(bio_page, page, b_offset,
					page_offset, clen,
506
					ASYNC_TX_DEP_ACK,
507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
					tx, NULL, NULL);
		}
		if (clen < len) /* hit end of page */
			break;
		page_offset +=  len;
	}

	return tx;
}

static void ops_complete_biofill(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;
	struct bio *return_bi = NULL;
	raid5_conf_t *conf = sh->raid_conf;
522
	int i;
523 524 525 526 527 528 529 530 531

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	/* clear completed biofills */
	for (i = sh->disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];

		/* acknowledge completion of a biofill operation */
532 533 534 535 536
		/* and check if we need to reply to a read request,
		 * new R5_Wantfill requests are held off until
		 * !test_bit(STRIPE_OP_BIOFILL, &sh->ops.pending)
		 */
		if (test_and_clear_bit(R5_Wantfill, &dev->flags)) {
537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558
			struct bio *rbi, *rbi2;

			/* The access to dev->read is outside of the
			 * spin_lock_irq(&conf->device_lock), but is protected
			 * by the STRIPE_OP_BIOFILL pending bit
			 */
			BUG_ON(!dev->read);
			rbi = dev->read;
			dev->read = NULL;
			while (rbi && rbi->bi_sector <
				dev->sector + STRIPE_SECTORS) {
				rbi2 = r5_next_bio(rbi, dev->sector);
				spin_lock_irq(&conf->device_lock);
				if (--rbi->bi_phys_segments == 0) {
					rbi->bi_next = return_bi;
					return_bi = rbi;
				}
				spin_unlock_irq(&conf->device_lock);
				rbi = rbi2;
			}
		}
	}
559
	set_bit(STRIPE_OP_BIOFILL, &sh->ops.complete);
560 561 562

	return_io(return_bi);

563
	set_bit(STRIPE_HANDLE, &sh->state);
564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920
	release_stripe(sh);
}

static void ops_run_biofill(struct stripe_head *sh)
{
	struct dma_async_tx_descriptor *tx = NULL;
	raid5_conf_t *conf = sh->raid_conf;
	int i;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	for (i = sh->disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		if (test_bit(R5_Wantfill, &dev->flags)) {
			struct bio *rbi;
			spin_lock_irq(&conf->device_lock);
			dev->read = rbi = dev->toread;
			dev->toread = NULL;
			spin_unlock_irq(&conf->device_lock);
			while (rbi && rbi->bi_sector <
				dev->sector + STRIPE_SECTORS) {
				tx = async_copy_data(0, rbi, dev->page,
					dev->sector, tx);
				rbi = r5_next_bio(rbi, dev->sector);
			}
		}
	}

	atomic_inc(&sh->count);
	async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx,
		ops_complete_biofill, sh);
}

static void ops_complete_compute5(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;
	int target = sh->ops.target;
	struct r5dev *tgt = &sh->dev[target];

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	set_bit(R5_UPTODATE, &tgt->flags);
	BUG_ON(!test_bit(R5_Wantcompute, &tgt->flags));
	clear_bit(R5_Wantcompute, &tgt->flags);
	set_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.complete);
	set_bit(STRIPE_HANDLE, &sh->state);
	release_stripe(sh);
}

static struct dma_async_tx_descriptor *
ops_run_compute5(struct stripe_head *sh, unsigned long pending)
{
	/* kernel stack size limits the total number of disks */
	int disks = sh->disks;
	struct page *xor_srcs[disks];
	int target = sh->ops.target;
	struct r5dev *tgt = &sh->dev[target];
	struct page *xor_dest = tgt->page;
	int count = 0;
	struct dma_async_tx_descriptor *tx;
	int i;

	pr_debug("%s: stripe %llu block: %d\n",
		__FUNCTION__, (unsigned long long)sh->sector, target);
	BUG_ON(!test_bit(R5_Wantcompute, &tgt->flags));

	for (i = disks; i--; )
		if (i != target)
			xor_srcs[count++] = sh->dev[i].page;

	atomic_inc(&sh->count);

	if (unlikely(count == 1))
		tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE,
			0, NULL, ops_complete_compute5, sh);
	else
		tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
			ASYNC_TX_XOR_ZERO_DST, NULL,
			ops_complete_compute5, sh);

	/* ack now if postxor is not set to be run */
	if (tx && !test_bit(STRIPE_OP_POSTXOR, &pending))
		async_tx_ack(tx);

	return tx;
}

static void ops_complete_prexor(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	set_bit(STRIPE_OP_PREXOR, &sh->ops.complete);
}

static struct dma_async_tx_descriptor *
ops_run_prexor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
{
	/* kernel stack size limits the total number of disks */
	int disks = sh->disks;
	struct page *xor_srcs[disks];
	int count = 0, pd_idx = sh->pd_idx, i;

	/* existing parity data subtracted */
	struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		/* Only process blocks that are known to be uptodate */
		if (dev->towrite && test_bit(R5_Wantprexor, &dev->flags))
			xor_srcs[count++] = dev->page;
	}

	tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
		ASYNC_TX_DEP_ACK | ASYNC_TX_XOR_DROP_DST, tx,
		ops_complete_prexor, sh);

	return tx;
}

static struct dma_async_tx_descriptor *
ops_run_biodrain(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
{
	int disks = sh->disks;
	int pd_idx = sh->pd_idx, i;

	/* check if prexor is active which means only process blocks
	 * that are part of a read-modify-write (Wantprexor)
	 */
	int prexor = test_bit(STRIPE_OP_PREXOR, &sh->ops.pending);

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		struct bio *chosen;
		int towrite;

		towrite = 0;
		if (prexor) { /* rmw */
			if (dev->towrite &&
			    test_bit(R5_Wantprexor, &dev->flags))
				towrite = 1;
		} else { /* rcw */
			if (i != pd_idx && dev->towrite &&
				test_bit(R5_LOCKED, &dev->flags))
				towrite = 1;
		}

		if (towrite) {
			struct bio *wbi;

			spin_lock(&sh->lock);
			chosen = dev->towrite;
			dev->towrite = NULL;
			BUG_ON(dev->written);
			wbi = dev->written = chosen;
			spin_unlock(&sh->lock);

			while (wbi && wbi->bi_sector <
				dev->sector + STRIPE_SECTORS) {
				tx = async_copy_data(1, wbi, dev->page,
					dev->sector, tx);
				wbi = r5_next_bio(wbi, dev->sector);
			}
		}
	}

	return tx;
}

static void ops_complete_postxor(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	set_bit(STRIPE_OP_POSTXOR, &sh->ops.complete);
	set_bit(STRIPE_HANDLE, &sh->state);
	release_stripe(sh);
}

static void ops_complete_write(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;
	int disks = sh->disks, i, pd_idx = sh->pd_idx;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		if (dev->written || i == pd_idx)
			set_bit(R5_UPTODATE, &dev->flags);
	}

	set_bit(STRIPE_OP_BIODRAIN, &sh->ops.complete);
	set_bit(STRIPE_OP_POSTXOR, &sh->ops.complete);

	set_bit(STRIPE_HANDLE, &sh->state);
	release_stripe(sh);
}

static void
ops_run_postxor(struct stripe_head *sh, struct dma_async_tx_descriptor *tx)
{
	/* kernel stack size limits the total number of disks */
	int disks = sh->disks;
	struct page *xor_srcs[disks];

	int count = 0, pd_idx = sh->pd_idx, i;
	struct page *xor_dest;
	int prexor = test_bit(STRIPE_OP_PREXOR, &sh->ops.pending);
	unsigned long flags;
	dma_async_tx_callback callback;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	/* check if prexor is active which means only process blocks
	 * that are part of a read-modify-write (written)
	 */
	if (prexor) {
		xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (dev->written)
				xor_srcs[count++] = dev->page;
		}
	} else {
		xor_dest = sh->dev[pd_idx].page;
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (i != pd_idx)
				xor_srcs[count++] = dev->page;
		}
	}

	/* check whether this postxor is part of a write */
	callback = test_bit(STRIPE_OP_BIODRAIN, &sh->ops.pending) ?
		ops_complete_write : ops_complete_postxor;

	/* 1/ if we prexor'd then the dest is reused as a source
	 * 2/ if we did not prexor then we are redoing the parity
	 * set ASYNC_TX_XOR_DROP_DST and ASYNC_TX_XOR_ZERO_DST
	 * for the synchronous xor case
	 */
	flags = ASYNC_TX_DEP_ACK | ASYNC_TX_ACK |
		(prexor ? ASYNC_TX_XOR_DROP_DST : ASYNC_TX_XOR_ZERO_DST);

	atomic_inc(&sh->count);

	if (unlikely(count == 1)) {
		flags &= ~(ASYNC_TX_XOR_DROP_DST | ASYNC_TX_XOR_ZERO_DST);
		tx = async_memcpy(xor_dest, xor_srcs[0], 0, 0, STRIPE_SIZE,
			flags, tx, callback, sh);
	} else
		tx = async_xor(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
			flags, tx, callback, sh);
}

static void ops_complete_check(void *stripe_head_ref)
{
	struct stripe_head *sh = stripe_head_ref;
	int pd_idx = sh->pd_idx;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	if (test_and_clear_bit(STRIPE_OP_MOD_DMA_CHECK, &sh->ops.pending) &&
		sh->ops.zero_sum_result == 0)
		set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);

	set_bit(STRIPE_OP_CHECK, &sh->ops.complete);
	set_bit(STRIPE_HANDLE, &sh->state);
	release_stripe(sh);
}

static void ops_run_check(struct stripe_head *sh)
{
	/* kernel stack size limits the total number of disks */
	int disks = sh->disks;
	struct page *xor_srcs[disks];
	struct dma_async_tx_descriptor *tx;

	int count = 0, pd_idx = sh->pd_idx, i;
	struct page *xor_dest = xor_srcs[count++] = sh->dev[pd_idx].page;

	pr_debug("%s: stripe %llu\n", __FUNCTION__,
		(unsigned long long)sh->sector);

	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		if (i != pd_idx)
			xor_srcs[count++] = dev->page;
	}

	tx = async_xor_zero_sum(xor_dest, xor_srcs, 0, count, STRIPE_SIZE,
		&sh->ops.zero_sum_result, 0, NULL, NULL, NULL);

	if (tx)
		set_bit(STRIPE_OP_MOD_DMA_CHECK, &sh->ops.pending);
	else
		clear_bit(STRIPE_OP_MOD_DMA_CHECK, &sh->ops.pending);

	atomic_inc(&sh->count);
	tx = async_trigger_callback(ASYNC_TX_DEP_ACK | ASYNC_TX_ACK, tx,
		ops_complete_check, sh);
}

static void raid5_run_ops(struct stripe_head *sh, unsigned long pending)
{
	int overlap_clear = 0, i, disks = sh->disks;
	struct dma_async_tx_descriptor *tx = NULL;

	if (test_bit(STRIPE_OP_BIOFILL, &pending)) {
		ops_run_biofill(sh);
		overlap_clear++;
	}

	if (test_bit(STRIPE_OP_COMPUTE_BLK, &pending))
		tx = ops_run_compute5(sh, pending);

	if (test_bit(STRIPE_OP_PREXOR, &pending))
		tx = ops_run_prexor(sh, tx);

	if (test_bit(STRIPE_OP_BIODRAIN, &pending)) {
		tx = ops_run_biodrain(sh, tx);
		overlap_clear++;
	}

	if (test_bit(STRIPE_OP_POSTXOR, &pending))
		ops_run_postxor(sh, tx);

	if (test_bit(STRIPE_OP_CHECK, &pending))
		ops_run_check(sh);

	if (test_bit(STRIPE_OP_IO, &pending))
		ops_run_io(sh);

	if (overlap_clear)
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (test_and_clear_bit(R5_Overlap, &dev->flags))
				wake_up(&sh->raid_conf->wait_for_overlap);
		}
}

921
static int grow_one_stripe(raid5_conf_t *conf)
Linus Torvalds's avatar
Linus Torvalds committed
922 923
{
	struct stripe_head *sh;
924 925 926 927 928 929 930 931 932 933 934 935
	sh = kmem_cache_alloc(conf->slab_cache, GFP_KERNEL);
	if (!sh)
		return 0;
	memset(sh, 0, sizeof(*sh) + (conf->raid_disks-1)*sizeof(struct r5dev));
	sh->raid_conf = conf;
	spin_lock_init(&sh->lock);

	if (grow_buffers(sh, conf->raid_disks)) {
		shrink_buffers(sh, conf->raid_disks);
		kmem_cache_free(conf->slab_cache, sh);
		return 0;
	}
936
	sh->disks = conf->raid_disks;
937 938 939 940 941 942 943 944 945 946
	/* we just created an active stripe so... */
	atomic_set(&sh->count, 1);
	atomic_inc(&conf->active_stripes);
	INIT_LIST_HEAD(&sh->lru);
	release_stripe(sh);
	return 1;
}

static int grow_stripes(raid5_conf_t *conf, int num)
{
947
	struct kmem_cache *sc;
Linus Torvalds's avatar
Linus Torvalds committed
948 949
	int devs = conf->raid_disks;

950 951
	sprintf(conf->cache_name[0], "raid5-%s", mdname(conf->mddev));
	sprintf(conf->cache_name[1], "raid5-%s-alt", mdname(conf->mddev));
952 953
	conf->active_name = 0;
	sc = kmem_cache_create(conf->cache_name[conf->active_name],
Linus Torvalds's avatar
Linus Torvalds committed
954
			       sizeof(struct stripe_head)+(devs-1)*sizeof(struct r5dev),
955
			       0, 0, NULL);
Linus Torvalds's avatar
Linus Torvalds committed
956 957 958
	if (!sc)
		return 1;
	conf->slab_cache = sc;
959
	conf->pool_size = devs;
960
	while (num--)
961
		if (!grow_one_stripe(conf))
Linus Torvalds's avatar
Linus Torvalds committed
962 963 964
			return 1;
	return 0;
}
965 966

#ifdef CONFIG_MD_RAID5_RESHAPE
967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995
static int resize_stripes(raid5_conf_t *conf, int newsize)
{
	/* Make all the stripes able to hold 'newsize' devices.
	 * New slots in each stripe get 'page' set to a new page.
	 *
	 * This happens in stages:
	 * 1/ create a new kmem_cache and allocate the required number of
	 *    stripe_heads.
	 * 2/ gather all the old stripe_heads and tranfer the pages across
	 *    to the new stripe_heads.  This will have the side effect of
	 *    freezing the array as once all stripe_heads have been collected,
	 *    no IO will be possible.  Old stripe heads are freed once their
	 *    pages have been transferred over, and the old kmem_cache is
	 *    freed when all stripes are done.
	 * 3/ reallocate conf->disks to be suitable bigger.  If this fails,
	 *    we simple return a failre status - no need to clean anything up.
	 * 4/ allocate new pages for the new slots in the new stripe_heads.
	 *    If this fails, we don't bother trying the shrink the
	 *    stripe_heads down again, we just leave them as they are.
	 *    As each stripe_head is processed the new one is released into
	 *    active service.
	 *
	 * Once step2 is started, we cannot afford to wait for a write,
	 * so we use GFP_NOIO allocations.
	 */
	struct stripe_head *osh, *nsh;
	LIST_HEAD(newstripes);
	struct disk_info *ndisks;
	int err = 0;
996
	struct kmem_cache *sc;
997 998 999 1000 1001
	int i;

	if (newsize <= conf->pool_size)
		return 0; /* never bother to shrink */

1002 1003
	md_allow_write(conf->mddev);

1004 1005 1006
	/* Step 1 */
	sc = kmem_cache_create(conf->cache_name[1-conf->active_name],
			       sizeof(struct stripe_head)+(newsize-1)*sizeof(struct r5dev),
1007
			       0, 0, NULL);
1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041
	if (!sc)
		return -ENOMEM;

	for (i = conf->max_nr_stripes; i; i--) {
		nsh = kmem_cache_alloc(sc, GFP_KERNEL);
		if (!nsh)
			break;

		memset(nsh, 0, sizeof(*nsh) + (newsize-1)*sizeof(struct r5dev));

		nsh->raid_conf = conf;
		spin_lock_init(&nsh->lock);

		list_add(&nsh->lru, &newstripes);
	}
	if (i) {
		/* didn't get enough, give up */
		while (!list_empty(&newstripes)) {
			nsh = list_entry(newstripes.next, struct stripe_head, lru);
			list_del(&nsh->lru);
			kmem_cache_free(sc, nsh);
		}
		kmem_cache_destroy(sc);
		return -ENOMEM;
	}
	/* Step 2 - Must use GFP_NOIO now.
	 * OK, we have enough stripes, start collecting inactive
	 * stripes and copying them over
	 */
	list_for_each_entry(nsh, &newstripes, lru) {
		spin_lock_irq(&conf->device_lock);
		wait_event_lock_irq(conf->wait_for_stripe,
				    !list_empty(&conf->inactive_list),
				    conf->device_lock,
1042
				    unplug_slaves(conf->mddev)
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 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
			);
		osh = get_free_stripe(conf);
		spin_unlock_irq(&conf->device_lock);
		atomic_set(&nsh->count, 1);
		for(i=0; i<conf->pool_size; i++)
			nsh->dev[i].page = osh->dev[i].page;
		for( ; i<newsize; i++)
			nsh->dev[i].page = NULL;
		kmem_cache_free(conf->slab_cache, osh);
	}
	kmem_cache_destroy(conf->slab_cache);

	/* Step 3.
	 * At this point, we are holding all the stripes so the array
	 * is completely stalled, so now is a good time to resize
	 * conf->disks.
	 */
	ndisks = kzalloc(newsize * sizeof(struct disk_info), GFP_NOIO);
	if (ndisks) {
		for (i=0; i<conf->raid_disks; i++)
			ndisks[i] = conf->disks[i];
		kfree(conf->disks);
		conf->disks = ndisks;
	} else
		err = -ENOMEM;

	/* Step 4, return new stripes to service */
	while(!list_empty(&newstripes)) {
		nsh = list_entry(newstripes.next, struct stripe_head, lru);
		list_del_init(&nsh->lru);
		for (i=conf->raid_disks; i < newsize; i++)
			if (nsh->dev[i].page == NULL) {
				struct page *p = alloc_page(GFP_NOIO);
				nsh->dev[i].page = p;
				if (!p)
					err = -ENOMEM;
			}
		release_stripe(nsh);
	}
	/* critical section pass, GFP_NOIO no longer needed */

	conf->slab_cache = sc;
	conf->active_name = 1-conf->active_name;
	conf->pool_size = newsize;
	return err;
}
1089
#endif
Linus Torvalds's avatar
Linus Torvalds committed
1090

1091
static int drop_one_stripe(raid5_conf_t *conf)
Linus Torvalds's avatar
Linus Torvalds committed
1092 1093 1094
{
	struct stripe_head *sh;

1095 1096 1097 1098 1099
	spin_lock_irq(&conf->device_lock);
	sh = get_free_stripe(conf);
	spin_unlock_irq(&conf->device_lock);
	if (!sh)
		return 0;
1100
	BUG_ON(atomic_read(&sh->count));
1101
	shrink_buffers(sh, conf->pool_size);
1102 1103 1104 1105 1106 1107 1108 1109 1110 1111
	kmem_cache_free(conf->slab_cache, sh);
	atomic_dec(&conf->active_stripes);
	return 1;
}

static void shrink_stripes(raid5_conf_t *conf)
{
	while (drop_one_stripe(conf))
		;

1112 1113
	if (conf->slab_cache)
		kmem_cache_destroy(conf->slab_cache);
Linus Torvalds's avatar
Linus Torvalds committed
1114 1115 1116
	conf->slab_cache = NULL;
}

1117
static int raid5_end_read_request(struct bio * bi, unsigned int bytes_done,
Linus Torvalds's avatar
Linus Torvalds committed
1118 1119 1120 1121
				   int error)
{
 	struct stripe_head *sh = bi->bi_private;
	raid5_conf_t *conf = sh->raid_conf;
1122
	int disks = sh->disks, i;
Linus Torvalds's avatar
Linus Torvalds committed
1123
	int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
1124 1125
	char b[BDEVNAME_SIZE];
	mdk_rdev_t *rdev;
Linus Torvalds's avatar
Linus Torvalds committed
1126 1127 1128 1129 1130 1131 1132 1133

	if (bi->bi_size)
		return 1;

	for (i=0 ; i<disks; i++)
		if (bi == &sh->dev[i].req)
			break;

1134 1135
	pr_debug("end_read_request %llu/%d, count: %d, uptodate %d.\n",
		(unsigned long long)sh->sector, i, atomic_read(&sh->count),
Linus Torvalds's avatar
Linus Torvalds committed
1136 1137 1138 1139 1140 1141 1142 1143
		uptodate);
	if (i == disks) {
		BUG();
		return 0;
	}

	if (uptodate) {
		set_bit(R5_UPTODATE, &sh->dev[i].flags);
1144
		if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
1145 1146 1147 1148 1149
			rdev = conf->disks[i].rdev;
			printk(KERN_INFO "raid5:%s: read error corrected (%lu sectors at %llu on %s)\n",
			       mdname(conf->mddev), STRIPE_SECTORS,
			       (unsigned long long)sh->sector + rdev->data_offset,
			       bdevname(rdev->bdev, b));
1150 1151 1152
			clear_bit(R5_ReadError, &sh->dev[i].flags);
			clear_bit(R5_ReWrite, &sh->dev[i].flags);
		}
1153 1154
		if (atomic_read(&conf->disks[i].rdev->read_errors))
			atomic_set(&conf->disks[i].rdev->read_errors, 0);
Linus Torvalds's avatar
Linus Torvalds committed
1155
	} else {
1156
		const char *bdn = bdevname(conf->disks[i].rdev->bdev, b);
1157
		int retry = 0;
1158 1159
		rdev = conf->disks[i].rdev;

Linus Torvalds's avatar
Linus Torvalds committed
1160
		clear_bit(R5_UPTODATE, &sh->dev[i].flags);
1161
		atomic_inc(&rdev->read_errors);
1162
		if (conf->mddev->degraded)
1163 1164 1165 1166
			printk(KERN_WARNING "raid5:%s: read error not correctable (sector %llu on %s).\n",
			       mdname(conf->mddev),
			       (unsigned long long)sh->sector + rdev->data_offset,
			       bdn);
1167
		else if (test_bit(R5_ReWrite, &sh->dev[i].flags))
1168
			/* Oh, no!!! */
1169 1170 1171 1172 1173
			printk(KERN_WARNING "raid5:%s: read error NOT corrected!! (sector %llu on %s).\n",
			       mdname(conf->mddev),
			       (unsigned long long)sh->sector + rdev->data_offset,
			       bdn);
		else if (atomic_read(&rdev->read_errors)
1174
			 > conf->max_nr_stripes)
1175
			printk(KERN_WARNING
1176 1177
			       "raid5:%s: Too many read errors, failing device %s.\n",
			       mdname(conf->mddev), bdn);
1178 1179 1180 1181 1182
		else
			retry = 1;
		if (retry)
			set_bit(R5_ReadError, &sh->dev[i].flags);
		else {
1183 1184
			clear_bit(R5_ReadError, &sh->dev[i].flags);
			clear_bit(R5_ReWrite, &sh->dev[i].flags);
1185
			md_error(conf->mddev, rdev);
1186
		}
Linus Torvalds's avatar
Linus Torvalds committed
1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199
	}
	rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
	clear_bit(R5_LOCKED, &sh->dev[i].flags);
	set_bit(STRIPE_HANDLE, &sh->state);
	release_stripe(sh);
	return 0;
}

static int raid5_end_write_request (struct bio *bi, unsigned int bytes_done,
				    int error)
{
 	struct stripe_head *sh = bi->bi_private;
	raid5_conf_t *conf = sh->raid_conf;
1200
	int disks = sh->disks, i;
Linus Torvalds's avatar
Linus Torvalds committed
1201 1202 1203 1204 1205 1206 1207 1208 1209
	int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);

	if (bi->bi_size)
		return 1;

	for (i=0 ; i<disks; i++)
		if (bi == &sh->dev[i].req)
			break;

1210
	pr_debug("end_write_request %llu/%d, count %d, uptodate: %d.\n",
Linus Torvalds's avatar
Linus Torvalds committed
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
		(unsigned long long)sh->sector, i, atomic_read(&sh->count),
		uptodate);
	if (i == disks) {
		BUG();
		return 0;
	}

	if (!uptodate)
		md_error(conf->mddev, conf->disks[i].rdev);

	rdev_dec_pending(conf->disks[i].rdev, conf->mddev);
	
	clear_bit(R5_LOCKED, &sh->dev[i].flags);
	set_bit(STRIPE_HANDLE, &sh->state);
1225
	release_stripe(sh);
Linus Torvalds's avatar
Linus Torvalds committed
1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247
	return 0;
}


static sector_t compute_blocknr(struct stripe_head *sh, int i);
	
static void raid5_build_block (struct stripe_head *sh, int i)
{
	struct r5dev *dev = &sh->dev[i];

	bio_init(&dev->req);
	dev->req.bi_io_vec = &dev->vec;
	dev->req.bi_vcnt++;
	dev->req.bi_max_vecs++;
	dev->vec.bv_page = dev->page;
	dev->vec.bv_len = STRIPE_SIZE;
	dev->vec.bv_offset = 0;

	dev->req.bi_sector = sh->sector;
	dev->req.bi_private = sh;

	dev->flags = 0;
1248
	dev->sector = compute_blocknr(sh, i);
Linus Torvalds's avatar
Linus Torvalds committed
1249 1250 1251 1252 1253 1254
}

static void error(mddev_t *mddev, mdk_rdev_t *rdev)
{
	char b[BDEVNAME_SIZE];
	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
1255
	pr_debug("raid5: error called\n");
Linus Torvalds's avatar
Linus Torvalds committed
1256

1257
	if (!test_bit(Faulty, &rdev->flags)) {
1258
		set_bit(MD_CHANGE_DEVS, &mddev->flags);
1259 1260 1261
		if (test_and_clear_bit(In_sync, &rdev->flags)) {
			unsigned long flags;
			spin_lock_irqsave(&conf->device_lock, flags);
Linus Torvalds's avatar
Linus Torvalds committed
1262
			mddev->degraded++;
1263
			spin_unlock_irqrestore(&conf->device_lock, flags);
Linus Torvalds's avatar
Linus Torvalds committed
1264 1265 1266 1267 1268
			/*
			 * if recovery was running, make sure it aborts.
			 */
			set_bit(MD_RECOVERY_ERR, &mddev->recovery);
		}
1269
		set_bit(Faulty, &rdev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
1270 1271 1272
		printk (KERN_ALERT
			"raid5: Disk failure on %s, disabling device."
			" Operation continuing on %d devices\n",
1273
			bdevname(rdev->bdev,b), conf->raid_disks - mddev->degraded);
Linus Torvalds's avatar
Linus Torvalds committed
1274
	}
1275
}
Linus Torvalds's avatar
Linus Torvalds committed
1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312

/*
 * Input: a 'big' sector number,
 * Output: index of the data and parity disk, and the sector # in them.
 */
static sector_t raid5_compute_sector(sector_t r_sector, unsigned int raid_disks,
			unsigned int data_disks, unsigned int * dd_idx,
			unsigned int * pd_idx, raid5_conf_t *conf)
{
	long stripe;
	unsigned long chunk_number;
	unsigned int chunk_offset;
	sector_t new_sector;
	int sectors_per_chunk = conf->chunk_size >> 9;

	/* First compute the information on this sector */

	/*
	 * Compute the chunk number and the sector offset inside the chunk
	 */
	chunk_offset = sector_div(r_sector, sectors_per_chunk);
	chunk_number = r_sector;
	BUG_ON(r_sector != chunk_number);

	/*
	 * Compute the stripe number
	 */
	stripe = chunk_number / data_disks;

	/*
	 * Compute the data disk and parity disk indexes inside the stripe
	 */
	*dd_idx = chunk_number % data_disks;

	/*
	 * Select the parity disk based on the user selected algorithm.
	 */
1313 1314
	switch(conf->level) {
	case 4:
Linus Torvalds's avatar
Linus Torvalds committed
1315
		*pd_idx = data_disks;
1316 1317 1318
		break;
	case 5:
		switch (conf->algorithm) {
Linus Torvalds's avatar
Linus Torvalds committed
1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337
		case ALGORITHM_LEFT_ASYMMETRIC:
			*pd_idx = data_disks - stripe % raid_disks;
			if (*dd_idx >= *pd_idx)
				(*dd_idx)++;
			break;
		case ALGORITHM_RIGHT_ASYMMETRIC:
			*pd_idx = stripe % raid_disks;
			if (*dd_idx >= *pd_idx)
				(*dd_idx)++;
			break;
		case ALGORITHM_LEFT_SYMMETRIC:
			*pd_idx = data_disks - stripe % raid_disks;
			*dd_idx = (*pd_idx + 1 + *dd_idx) % raid_disks;
			break;
		case ALGORITHM_RIGHT_SYMMETRIC:
			*pd_idx = stripe % raid_disks;
			*dd_idx = (*pd_idx + 1 + *dd_idx) % raid_disks;
			break;
		default:
1338
			printk(KERN_ERR "raid5: unsupported algorithm %d\n",
Linus Torvalds's avatar
Linus Torvalds committed
1339
				conf->algorithm);
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
		}
		break;
	case 6:

		/**** FIX THIS ****/
		switch (conf->algorithm) {
		case ALGORITHM_LEFT_ASYMMETRIC:
			*pd_idx = raid_disks - 1 - (stripe % raid_disks);
			if (*pd_idx == raid_disks-1)
				(*dd_idx)++; 	/* Q D D D P */
			else if (*dd_idx >= *pd_idx)
				(*dd_idx) += 2; /* D D P Q D */
			break;
		case ALGORITHM_RIGHT_ASYMMETRIC:
			*pd_idx = stripe % raid_disks;
			if (*pd_idx == raid_disks-1)
				(*dd_idx)++; 	/* Q D D D P */
			else if (*dd_idx >= *pd_idx)
				(*dd_idx) += 2; /* D D P Q D */
			break;
		case ALGORITHM_LEFT_SYMMETRIC:
			*pd_idx = raid_disks - 1 - (stripe % raid_disks);
			*dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
			break;
		case ALGORITHM_RIGHT_SYMMETRIC:
			*pd_idx = stripe % raid_disks;
			*dd_idx = (*pd_idx + 2 + *dd_idx) % raid_disks;
			break;
		default:
			printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
				conf->algorithm);
		}
		break;
Linus Torvalds's avatar
Linus Torvalds committed
1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
	}

	/*
	 * Finally, compute the new sector number
	 */
	new_sector = (sector_t)stripe * sectors_per_chunk + chunk_offset;
	return new_sector;
}


static sector_t compute_blocknr(struct stripe_head *sh, int i)
{
	raid5_conf_t *conf = sh->raid_conf;
1386 1387
	int raid_disks = sh->disks;
	int data_disks = raid_disks - conf->max_degraded;
Linus Torvalds's avatar
Linus Torvalds committed
1388 1389 1390 1391 1392 1393 1394
	sector_t new_sector = sh->sector, check;
	int sectors_per_chunk = conf->chunk_size >> 9;
	sector_t stripe;
	int chunk_offset;
	int chunk_number, dummy1, dummy2, dd_idx = i;
	sector_t r_sector;

1395

Linus Torvalds's avatar
Linus Torvalds committed
1396 1397 1398 1399
	chunk_offset = sector_div(new_sector, sectors_per_chunk);
	stripe = new_sector;
	BUG_ON(new_sector != stripe);

1400 1401 1402 1403 1404 1405
	if (i == sh->pd_idx)
		return 0;
	switch(conf->level) {
	case 4: break;
	case 5:
		switch (conf->algorithm) {
Linus Torvalds's avatar
Linus Torvalds committed
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417
		case ALGORITHM_LEFT_ASYMMETRIC:
		case ALGORITHM_RIGHT_ASYMMETRIC:
			if (i > sh->pd_idx)
				i--;
			break;
		case ALGORITHM_LEFT_SYMMETRIC:
		case ALGORITHM_RIGHT_SYMMETRIC:
			if (i < sh->pd_idx)
				i += raid_disks;
			i -= (sh->pd_idx + 1);
			break;
		default:
1418
			printk(KERN_ERR "raid5: unsupported algorithm %d\n",
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
			       conf->algorithm);
		}
		break;
	case 6:
		if (i == raid6_next_disk(sh->pd_idx, raid_disks))
			return 0; /* It is the Q disk */
		switch (conf->algorithm) {
		case ALGORITHM_LEFT_ASYMMETRIC:
		case ALGORITHM_RIGHT_ASYMMETRIC:
		  	if (sh->pd_idx == raid_disks-1)
				i--; 	/* Q D D D P */
			else if (i > sh->pd_idx)
				i -= 2; /* D D P Q D */
			break;
		case ALGORITHM_LEFT_SYMMETRIC:
		case ALGORITHM_RIGHT_SYMMETRIC:
			if (sh->pd_idx == raid_disks-1)
				i--; /* Q D D D P */
			else {
				/* D D P Q D */
				if (i < sh->pd_idx)
					i += raid_disks;
				i -= (sh->pd_idx + 2);
			}
			break;
		default:
			printk (KERN_CRIT "raid6: unsupported algorithm %d\n",
Linus Torvalds's avatar
Linus Torvalds committed
1446
				conf->algorithm);
1447 1448
		}
		break;
Linus Torvalds's avatar
Linus Torvalds committed
1449 1450 1451 1452 1453 1454 1455
	}

	chunk_number = stripe * data_disks + i;
	r_sector = (sector_t)chunk_number * sectors_per_chunk + chunk_offset;

	check = raid5_compute_sector (r_sector, raid_disks, data_disks, &dummy1, &dummy2, conf);
	if (check != sh->sector || dummy1 != dd_idx || dummy2 != sh->pd_idx) {
1456
		printk(KERN_ERR "compute_blocknr: map not correct\n");
Linus Torvalds's avatar
Linus Torvalds committed
1457 1458 1459 1460 1461 1462 1463 1464
		return 0;
	}
	return r_sector;
}



/*
1465 1466 1467 1468 1469
 * Copy data between a page in the stripe cache, and one or more bion
 * The page could align with the middle of the bio, or there could be
 * several bion, each with several bio_vecs, which cover part of the page
 * Multiple bion are linked together on bi_next.  There may be extras
 * at the end of this list.  We ignore them.
Linus Torvalds's avatar
Linus Torvalds committed
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
 */
static void copy_data(int frombio, struct bio *bio,
		     struct page *page,
		     sector_t sector)
{
	char *pa = page_address(page);
	struct bio_vec *bvl;
	int i;
	int page_offset;

	if (bio->bi_sector >= sector)
		page_offset = (signed)(bio->bi_sector - sector) * 512;
	else
		page_offset = (signed)(sector - bio->bi_sector) * -512;
	bio_for_each_segment(bvl, bio, i) {
		int len = bio_iovec_idx(bio,i)->bv_len;
		int clen;
		int b_offset = 0;

		if (page_offset < 0) {
			b_offset = -page_offset;
			page_offset += b_offset;
			len -= b_offset;
		}

		if (len > 0 && page_offset + len > STRIPE_SIZE)
			clen = STRIPE_SIZE - page_offset;
		else clen = len;
1498

Linus Torvalds's avatar
Linus Torvalds committed
1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512
		if (clen > 0) {
			char *ba = __bio_kmap_atomic(bio, i, KM_USER0);
			if (frombio)
				memcpy(pa+page_offset, ba+b_offset, clen);
			else
				memcpy(ba+b_offset, pa+page_offset, clen);
			__bio_kunmap_atomic(ba, KM_USER0);
		}
		if (clen < len) /* hit end of page */
			break;
		page_offset +=  len;
	}
}

1513 1514 1515 1516 1517
#define check_xor()	do {						  \
				if (count == MAX_XOR_BLOCKS) {		  \
				xor_blocks(count, STRIPE_SIZE, dest, ptr);\
				count = 0;				  \
			   }						  \
Linus Torvalds's avatar
Linus Torvalds committed
1518 1519
			} while(0)

1520 1521 1522
static void compute_parity6(struct stripe_head *sh, int method)
{
	raid6_conf_t *conf = sh->raid_conf;
1523
	int i, pd_idx = sh->pd_idx, qd_idx, d0_idx, disks = sh->disks, count;
1524 1525 1526 1527 1528 1529 1530
	struct bio *chosen;
	/**** FIX THIS: This could be very bad if disks is close to 256 ****/
	void *ptrs[disks];

	qd_idx = raid6_next_disk(pd_idx, disks);
	d0_idx = raid6_next_disk(qd_idx, disks);

1531
	pr_debug("compute_parity, stripe %llu, method %d\n",
1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
		(unsigned long long)sh->sector, method);

	switch(method) {
	case READ_MODIFY_WRITE:
		BUG();		/* READ_MODIFY_WRITE N/A for RAID-6 */
	case RECONSTRUCT_WRITE:
		for (i= disks; i-- ;)
			if ( i != pd_idx && i != qd_idx && sh->dev[i].towrite ) {
				chosen = sh->dev[i].towrite;
				sh->dev[i].towrite = NULL;

				if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
					wake_up(&conf->wait_for_overlap);

1546
				BUG_ON(sh->dev[i].written);
1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603
				sh->dev[i].written = chosen;
			}
		break;
	case CHECK_PARITY:
		BUG();		/* Not implemented yet */
	}

	for (i = disks; i--;)
		if (sh->dev[i].written) {
			sector_t sector = sh->dev[i].sector;
			struct bio *wbi = sh->dev[i].written;
			while (wbi && wbi->bi_sector < sector + STRIPE_SECTORS) {
				copy_data(1, wbi, sh->dev[i].page, sector);
				wbi = r5_next_bio(wbi, sector);
			}

			set_bit(R5_LOCKED, &sh->dev[i].flags);
			set_bit(R5_UPTODATE, &sh->dev[i].flags);
		}

//	switch(method) {
//	case RECONSTRUCT_WRITE:
//	case CHECK_PARITY:
//	case UPDATE_PARITY:
		/* Note that unlike RAID-5, the ordering of the disks matters greatly. */
		/* FIX: Is this ordering of drives even remotely optimal? */
		count = 0;
		i = d0_idx;
		do {
			ptrs[count++] = page_address(sh->dev[i].page);
			if (count <= disks-2 && !test_bit(R5_UPTODATE, &sh->dev[i].flags))
				printk("block %d/%d not uptodate on parity calc\n", i,count);
			i = raid6_next_disk(i, disks);
		} while ( i != d0_idx );
//		break;
//	}

	raid6_call.gen_syndrome(disks, STRIPE_SIZE, ptrs);

	switch(method) {
	case RECONSTRUCT_WRITE:
		set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
		set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
		set_bit(R5_LOCKED,   &sh->dev[pd_idx].flags);
		set_bit(R5_LOCKED,   &sh->dev[qd_idx].flags);
		break;
	case UPDATE_PARITY:
		set_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
		set_bit(R5_UPTODATE, &sh->dev[qd_idx].flags);
		break;
	}
}


/* Compute one missing block */
static void compute_block_1(struct stripe_head *sh, int dd_idx, int nozero)
{
1604
	int i, count, disks = sh->disks;
1605
	void *ptr[MAX_XOR_BLOCKS], *dest, *p;
1606 1607 1608
	int pd_idx = sh->pd_idx;
	int qd_idx = raid6_next_disk(pd_idx, disks);

1609
	pr_debug("compute_block_1, stripe %llu, idx %d\n",
1610 1611 1612 1613 1614 1615
		(unsigned long long)sh->sector, dd_idx);

	if ( dd_idx == qd_idx ) {
		/* We're actually computing the Q drive */
		compute_parity6(sh, UPDATE_PARITY);
	} else {
1616 1617 1618
		dest = page_address(sh->dev[dd_idx].page);
		if (!nozero) memset(dest, 0, STRIPE_SIZE);
		count = 0;
1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631
		for (i = disks ; i--; ) {
			if (i == dd_idx || i == qd_idx)
				continue;
			p = page_address(sh->dev[i].page);
			if (test_bit(R5_UPTODATE, &sh->dev[i].flags))
				ptr[count++] = p;
			else
				printk("compute_block() %d, stripe %llu, %d"
				       " not present\n", dd_idx,
				       (unsigned long long)sh->sector, i);

			check_xor();
		}
1632 1633
		if (count)
			xor_blocks(count, STRIPE_SIZE, dest, ptr);
1634 1635 1636 1637 1638 1639 1640 1641
		if (!nozero) set_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
		else clear_bit(R5_UPTODATE, &sh->dev[dd_idx].flags);
	}
}

/* Compute two missing blocks */
static void compute_block_2(struct stripe_head *sh, int dd_idx1, int dd_idx2)
{
1642
	int i, count, disks = sh->disks;
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655
	int pd_idx = sh->pd_idx;
	int qd_idx = raid6_next_disk(pd_idx, disks);
	int d0_idx = raid6_next_disk(qd_idx, disks);
	int faila, failb;

	/* faila and failb are disk numbers relative to d0_idx */
	/* pd_idx become disks-2 and qd_idx become disks-1 */
	faila = (dd_idx1 < d0_idx) ? dd_idx1+(disks-d0_idx) : dd_idx1-d0_idx;
	failb = (dd_idx2 < d0_idx) ? dd_idx2+(disks-d0_idx) : dd_idx2-d0_idx;

	BUG_ON(faila == failb);
	if ( failb < faila ) { int tmp = faila; faila = failb; failb = tmp; }

1656
	pr_debug("compute_block_2, stripe %llu, idx %d,%d (%d,%d)\n",
1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701
	       (unsigned long long)sh->sector, dd_idx1, dd_idx2, faila, failb);

	if ( failb == disks-1 ) {
		/* Q disk is one of the missing disks */
		if ( faila == disks-2 ) {
			/* Missing P+Q, just recompute */
			compute_parity6(sh, UPDATE_PARITY);
			return;
		} else {
			/* We're missing D+Q; recompute D from P */
			compute_block_1(sh, (dd_idx1 == qd_idx) ? dd_idx2 : dd_idx1, 0);
			compute_parity6(sh, UPDATE_PARITY); /* Is this necessary? */
			return;
		}
	}

	/* We're missing D+P or D+D; build pointer table */
	{
		/**** FIX THIS: This could be very bad if disks is close to 256 ****/
		void *ptrs[disks];

		count = 0;
		i = d0_idx;
		do {
			ptrs[count++] = page_address(sh->dev[i].page);
			i = raid6_next_disk(i, disks);
			if (i != dd_idx1 && i != dd_idx2 &&
			    !test_bit(R5_UPTODATE, &sh->dev[i].flags))
				printk("compute_2 with missing block %d/%d\n", count, i);
		} while ( i != d0_idx );

		if ( failb == disks-2 ) {
			/* We're missing D+P. */
			raid6_datap_recov(disks, STRIPE_SIZE, faila, ptrs);
		} else {
			/* We're missing D+D. */
			raid6_2data_recov(disks, STRIPE_SIZE, faila, failb, ptrs);
		}

		/* Both the above update both missing blocks */
		set_bit(R5_UPTODATE, &sh->dev[dd_idx1].flags);
		set_bit(R5_UPTODATE, &sh->dev[dd_idx2].flags);
	}
}

1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716
static int
handle_write_operations5(struct stripe_head *sh, int rcw, int expand)
{
	int i, pd_idx = sh->pd_idx, disks = sh->disks;
	int locked = 0;

	if (rcw) {
		/* if we are not expanding this is a proper write request, and
		 * there will be bios with new data to be drained into the
		 * stripe cache
		 */
		if (!expand) {
			set_bit(STRIPE_OP_BIODRAIN, &sh->ops.pending);
			sh->ops.count++;
		}
1717

1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774
		set_bit(STRIPE_OP_POSTXOR, &sh->ops.pending);
		sh->ops.count++;

		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];

			if (dev->towrite) {
				set_bit(R5_LOCKED, &dev->flags);
				if (!expand)
					clear_bit(R5_UPTODATE, &dev->flags);
				locked++;
			}
		}
	} else {
		BUG_ON(!(test_bit(R5_UPTODATE, &sh->dev[pd_idx].flags) ||
			test_bit(R5_Wantcompute, &sh->dev[pd_idx].flags)));

		set_bit(STRIPE_OP_PREXOR, &sh->ops.pending);
		set_bit(STRIPE_OP_BIODRAIN, &sh->ops.pending);
		set_bit(STRIPE_OP_POSTXOR, &sh->ops.pending);

		sh->ops.count += 3;

		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (i == pd_idx)
				continue;

			/* For a read-modify write there may be blocks that are
			 * locked for reading while others are ready to be
			 * written so we distinguish these blocks by the
			 * R5_Wantprexor bit
			 */
			if (dev->towrite &&
			    (test_bit(R5_UPTODATE, &dev->flags) ||
			    test_bit(R5_Wantcompute, &dev->flags))) {
				set_bit(R5_Wantprexor, &dev->flags);
				set_bit(R5_LOCKED, &dev->flags);
				clear_bit(R5_UPTODATE, &dev->flags);
				locked++;
			}
		}
	}

	/* keep the parity disk locked while asynchronous operations
	 * are in flight
	 */
	set_bit(R5_LOCKED, &sh->dev[pd_idx].flags);
	clear_bit(R5_UPTODATE, &sh->dev[pd_idx].flags);
	locked++;

	pr_debug("%s: stripe %llu locked: %d pending: %lx\n",
		__FUNCTION__, (unsigned long long)sh->sector,
		locked, sh->ops.pending);

	return locked;
}
1775

Linus Torvalds's avatar
Linus Torvalds committed
1776 1777
/*
 * Each stripe/dev can have one or more bion attached.
1778
 * toread/towrite point to the first in a chain.
Linus Torvalds's avatar
Linus Torvalds committed
1779 1780 1781 1782 1783 1784
 * The bi_next chain must be in order.
 */
static int add_stripe_bio(struct stripe_head *sh, struct bio *bi, int dd_idx, int forwrite)
{
	struct bio **bip;
	raid5_conf_t *conf = sh->raid_conf;
1785
	int firstwrite=0;
Linus Torvalds's avatar
Linus Torvalds committed
1786

1787
	pr_debug("adding bh b#%llu to stripe s#%llu\n",
Linus Torvalds's avatar
Linus Torvalds committed
1788 1789 1790 1791 1792 1793
		(unsigned long long)bi->bi_sector,
		(unsigned long long)sh->sector);


	spin_lock(&sh->lock);
	spin_lock_irq(&conf->device_lock);
1794
	if (forwrite) {
Linus Torvalds's avatar
Linus Torvalds committed
1795
		bip = &sh->dev[dd_idx].towrite;
1796 1797 1798
		if (*bip == NULL && sh->dev[dd_idx].written == NULL)
			firstwrite = 1;
	} else
Linus Torvalds's avatar
Linus Torvalds committed
1799 1800 1801 1802 1803 1804 1805 1806 1807
		bip = &sh->dev[dd_idx].toread;
	while (*bip && (*bip)->bi_sector < bi->bi_sector) {
		if ((*bip)->bi_sector + ((*bip)->bi_size >> 9) > bi->bi_sector)
			goto overlap;
		bip = & (*bip)->bi_next;
	}
	if (*bip && (*bip)->bi_sector < bi->bi_sector + ((bi->bi_size)>>9))
		goto overlap;

1808
	BUG_ON(*bip && bi->bi_next && (*bip) != bi->bi_next);
Linus Torvalds's avatar
Linus Torvalds committed
1809 1810 1811 1812 1813 1814 1815
	if (*bip)
		bi->bi_next = *bip;
	*bip = bi;
	bi->bi_phys_segments ++;
	spin_unlock_irq(&conf->device_lock);
	spin_unlock(&sh->lock);

1816
	pr_debug("added bi b#%llu to stripe s#%llu, disk %d.\n",
Linus Torvalds's avatar
Linus Torvalds committed
1817 1818 1819
		(unsigned long long)bi->bi_sector,
		(unsigned long long)sh->sector, dd_idx);

1820 1821 1822
	if (conf->mddev->bitmap && firstwrite) {
		bitmap_startwrite(conf->mddev->bitmap, sh->sector,
				  STRIPE_SECTORS, 0);
1823
		sh->bm_seq = conf->seq_flush+1;
1824 1825 1826
		set_bit(STRIPE_BIT_DELAY, &sh->state);
	}

Linus Torvalds's avatar
Linus Torvalds committed
1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848
	if (forwrite) {
		/* check if page is covered */
		sector_t sector = sh->dev[dd_idx].sector;
		for (bi=sh->dev[dd_idx].towrite;
		     sector < sh->dev[dd_idx].sector + STRIPE_SECTORS &&
			     bi && bi->bi_sector <= sector;
		     bi = r5_next_bio(bi, sh->dev[dd_idx].sector)) {
			if (bi->bi_sector + (bi->bi_size>>9) >= sector)
				sector = bi->bi_sector + (bi->bi_size>>9);
		}
		if (sector >= sh->dev[dd_idx].sector + STRIPE_SECTORS)
			set_bit(R5_OVERWRITE, &sh->dev[dd_idx].flags);
	}
	return 1;

 overlap:
	set_bit(R5_Overlap, &sh->dev[dd_idx].flags);
	spin_unlock_irq(&conf->device_lock);
	spin_unlock(&sh->lock);
	return 0;
}

1849 1850
static void end_reshape(raid5_conf_t *conf);

1851 1852 1853 1854 1855 1856 1857
static int page_is_zero(struct page *p)
{
	char *a = page_address(p);
	return ((*(u32*)a) == 0 &&
		memcmp(a, a+4, STRIPE_SIZE-4)==0);
}

1858 1859 1860 1861
static int stripe_to_pdidx(sector_t stripe, raid5_conf_t *conf, int disks)
{
	int sectors_per_chunk = conf->chunk_size >> 9;
	int pd_idx, dd_idx;
1862 1863
	int chunk_offset = sector_div(stripe, sectors_per_chunk);

1864 1865 1866 1867
	raid5_compute_sector(stripe * (disks - conf->max_degraded)
			     *sectors_per_chunk + chunk_offset,
			     disks, disks - conf->max_degraded,
			     &dd_idx, &pd_idx, conf);
1868 1869 1870
	return pd_idx;
}

1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928
static void
handle_requests_to_failed_array(raid5_conf_t *conf, struct stripe_head *sh,
				struct stripe_head_state *s, int disks,
				struct bio **return_bi)
{
	int i;
	for (i = disks; i--; ) {
		struct bio *bi;
		int bitmap_end = 0;

		if (test_bit(R5_ReadError, &sh->dev[i].flags)) {
			mdk_rdev_t *rdev;
			rcu_read_lock();
			rdev = rcu_dereference(conf->disks[i].rdev);
			if (rdev && test_bit(In_sync, &rdev->flags))
				/* multiple read failures in one stripe */
				md_error(conf->mddev, rdev);
			rcu_read_unlock();
		}
		spin_lock_irq(&conf->device_lock);
		/* fail all writes first */
		bi = sh->dev[i].towrite;
		sh->dev[i].towrite = NULL;
		if (bi) {
			s->to_write--;
			bitmap_end = 1;
		}

		if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
			wake_up(&conf->wait_for_overlap);

		while (bi && bi->bi_sector <
			sh->dev[i].sector + STRIPE_SECTORS) {
			struct bio *nextbi = r5_next_bio(bi, sh->dev[i].sector);
			clear_bit(BIO_UPTODATE, &bi->bi_flags);
			if (--bi->bi_phys_segments == 0) {
				md_write_end(conf->mddev);
				bi->bi_next = *return_bi;
				*return_bi = bi;
			}
			bi = nextbi;
		}
		/* and fail all 'written' */
		bi = sh->dev[i].written;
		sh->dev[i].written = NULL;
		if (bi) bitmap_end = 1;
		while (bi && bi->bi_sector <
		       sh->dev[i].sector + STRIPE_SECTORS) {
			struct bio *bi2 = r5_next_bio(bi, sh->dev[i].sector);
			clear_bit(BIO_UPTODATE, &bi->bi_flags);
			if (--bi->bi_phys_segments == 0) {
				md_write_end(conf->mddev);
				bi->bi_next = *return_bi;
				*return_bi = bi;
			}
			bi = bi2;
		}

1929 1930 1931 1932 1933 1934
		/* fail any reads if this device is non-operational and
		 * the data has not reached the cache yet.
		 */
		if (!test_bit(R5_Wantfill, &sh->dev[i].flags) &&
		    (!test_bit(R5_Insync, &sh->dev[i].flags) ||
		      test_bit(R5_ReadError, &sh->dev[i].flags))) {
1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959
			bi = sh->dev[i].toread;
			sh->dev[i].toread = NULL;
			if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
				wake_up(&conf->wait_for_overlap);
			if (bi) s->to_read--;
			while (bi && bi->bi_sector <
			       sh->dev[i].sector + STRIPE_SECTORS) {
				struct bio *nextbi =
					r5_next_bio(bi, sh->dev[i].sector);
				clear_bit(BIO_UPTODATE, &bi->bi_flags);
				if (--bi->bi_phys_segments == 0) {
					bi->bi_next = *return_bi;
					*return_bi = bi;
				}
				bi = nextbi;
			}
		}
		spin_unlock_irq(&conf->device_lock);
		if (bitmap_end)
			bitmap_endwrite(conf->mddev->bitmap, sh->sector,
					STRIPE_SECTORS, 0, 0);
	}

}

1960 1961 1962 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 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
/* __handle_issuing_new_read_requests5 - returns 0 if there are no more disks
 * to process
 */
static int __handle_issuing_new_read_requests5(struct stripe_head *sh,
			struct stripe_head_state *s, int disk_idx, int disks)
{
	struct r5dev *dev = &sh->dev[disk_idx];
	struct r5dev *failed_dev = &sh->dev[s->failed_num];

	/* don't schedule compute operations or reads on the parity block while
	 * a check is in flight
	 */
	if ((disk_idx == sh->pd_idx) &&
	     test_bit(STRIPE_OP_CHECK, &sh->ops.pending))
		return ~0;

	/* is the data in this block needed, and can we get it? */
	if (!test_bit(R5_LOCKED, &dev->flags) &&
	    !test_bit(R5_UPTODATE, &dev->flags) && (dev->toread ||
	    (dev->towrite && !test_bit(R5_OVERWRITE, &dev->flags)) ||
	     s->syncing || s->expanding || (s->failed &&
	     (failed_dev->toread || (failed_dev->towrite &&
	     !test_bit(R5_OVERWRITE, &failed_dev->flags)
	     ))))) {
		/* 1/ We would like to get this block, possibly by computing it,
		 * but we might not be able to.
		 *
		 * 2/ Since parity check operations potentially make the parity
		 * block !uptodate it will need to be refreshed before any
		 * compute operations on data disks are scheduled.
		 *
		 * 3/ We hold off parity block re-reads until check operations
		 * have quiesced.
		 */
		if ((s->uptodate == disks - 1) &&
		    !test_bit(STRIPE_OP_CHECK, &sh->ops.pending)) {
			set_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending);
			set_bit(R5_Wantcompute, &dev->flags);
			sh->ops.target = disk_idx;
			s->req_compute = 1;
			sh->ops.count++;
			/* Careful: from this point on 'uptodate' is in the eye
			 * of raid5_run_ops which services 'compute' operations
			 * before writes. R5_Wantcompute flags a block that will
			 * be R5_UPTODATE by the time it is needed for a
			 * subsequent operation.
			 */
			s->uptodate++;
			return 0; /* uptodate + compute == disks */
		} else if ((s->uptodate < disks - 1) &&
			test_bit(R5_Insync, &dev->flags)) {
			/* Note: we hold off compute operations while checks are
			 * in flight, but we still prefer 'compute' over 'read'
			 * hence we only read if (uptodate < * disks-1)
			 */
			set_bit(R5_LOCKED, &dev->flags);
			set_bit(R5_Wantread, &dev->flags);
			if (!test_and_set_bit(STRIPE_OP_IO, &sh->ops.pending))
				sh->ops.count++;
			s->locked++;
			pr_debug("Reading block %d (sync=%d)\n", disk_idx,
				s->syncing);
		}
	}

	return ~0;
}

2028 2029 2030 2031
static void handle_issuing_new_read_requests5(struct stripe_head *sh,
			struct stripe_head_state *s, int disks)
{
	int i;
2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054

	/* Clear completed compute operations.  Parity recovery
	 * (STRIPE_OP_MOD_REPAIR_PD) implies a write-back which is handled
	 * later on in this routine
	 */
	if (test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.complete) &&
		!test_bit(STRIPE_OP_MOD_REPAIR_PD, &sh->ops.pending)) {
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.complete);
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.ack);
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending);
	}

	/* look for blocks to read/compute, skip this if a compute
	 * is already in flight, or if the stripe contents are in the
	 * midst of changing due to a write
	 */
	if (!test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending) &&
		!test_bit(STRIPE_OP_PREXOR, &sh->ops.pending) &&
		!test_bit(STRIPE_OP_POSTXOR, &sh->ops.pending)) {
		for (i = disks; i--; )
			if (__handle_issuing_new_read_requests5(
				sh, s, i, disks) == 0)
				break;
2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080
	}
	set_bit(STRIPE_HANDLE, &sh->state);
}

static void handle_issuing_new_read_requests6(struct stripe_head *sh,
			struct stripe_head_state *s, struct r6_state *r6s,
			int disks)
{
	int i;
	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		if (!test_bit(R5_LOCKED, &dev->flags) &&
		    !test_bit(R5_UPTODATE, &dev->flags) &&
		    (dev->toread || (dev->towrite &&
		     !test_bit(R5_OVERWRITE, &dev->flags)) ||
		     s->syncing || s->expanding ||
		     (s->failed >= 1 &&
		      (sh->dev[r6s->failed_num[0]].toread ||
		       s->to_write)) ||
		     (s->failed >= 2 &&
		      (sh->dev[r6s->failed_num[1]].toread ||
		       s->to_write)))) {
			/* we would like to get this block, possibly
			 * by computing it, but we might not be able to
			 */
			if (s->uptodate == disks-1) {
2081
				pr_debug("Computing stripe %llu block %d\n",
2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097
				       (unsigned long long)sh->sector, i);
				compute_block_1(sh, i, 0);
				s->uptodate++;
			} else if ( s->uptodate == disks-2 && s->failed >= 2 ) {
				/* Computing 2-failure is *very* expensive; only
				 * do it if failed >= 2
				 */
				int other;
				for (other = disks; other--; ) {
					if (other == i)
						continue;
					if (!test_bit(R5_UPTODATE,
					      &sh->dev[other].flags))
						break;
				}
				BUG_ON(other < 0);
2098
				pr_debug("Computing stripe %llu blocks %d,%d\n",
2099 2100 2101 2102 2103 2104 2105 2106
				       (unsigned long long)sh->sector,
				       i, other);
				compute_block_2(sh, i, other);
				s->uptodate += 2;
			} else if (test_bit(R5_Insync, &dev->flags)) {
				set_bit(R5_LOCKED, &dev->flags);
				set_bit(R5_Wantread, &dev->flags);
				s->locked++;
2107
				pr_debug("Reading block %d (sync=%d)\n",
2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134
					i, s->syncing);
			}
		}
	}
	set_bit(STRIPE_HANDLE, &sh->state);
}


/* handle_completed_write_requests
 * any written block on an uptodate or failed drive can be returned.
 * Note that if we 'wrote' to a failed drive, it will be UPTODATE, but
 * never LOCKED, so we don't need to test 'failed' directly.
 */
static void handle_completed_write_requests(raid5_conf_t *conf,
	struct stripe_head *sh, int disks, struct bio **return_bi)
{
	int i;
	struct r5dev *dev;

	for (i = disks; i--; )
		if (sh->dev[i].written) {
			dev = &sh->dev[i];
			if (!test_bit(R5_LOCKED, &dev->flags) &&
				test_bit(R5_UPTODATE, &dev->flags)) {
				/* We can return any write requests */
				struct bio *wbi, *wbi2;
				int bitmap_end = 0;
2135
				pr_debug("Return write for disc %d\n", i);
2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170
				spin_lock_irq(&conf->device_lock);
				wbi = dev->written;
				dev->written = NULL;
				while (wbi && wbi->bi_sector <
					dev->sector + STRIPE_SECTORS) {
					wbi2 = r5_next_bio(wbi, dev->sector);
					if (--wbi->bi_phys_segments == 0) {
						md_write_end(conf->mddev);
						wbi->bi_next = *return_bi;
						*return_bi = wbi;
					}
					wbi = wbi2;
				}
				if (dev->towrite == NULL)
					bitmap_end = 1;
				spin_unlock_irq(&conf->device_lock);
				if (bitmap_end)
					bitmap_endwrite(conf->mddev->bitmap,
							sh->sector,
							STRIPE_SECTORS,
					 !test_bit(STRIPE_DEGRADED, &sh->state),
							0);
			}
		}
}

static void handle_issuing_new_write_requests5(raid5_conf_t *conf,
		struct stripe_head *sh,	struct stripe_head_state *s, int disks)
{
	int rmw = 0, rcw = 0, i;
	for (i = disks; i--; ) {
		/* would I have to read this buffer for read_modify_write */
		struct r5dev *dev = &sh->dev[i];
		if ((dev->towrite || i == sh->pd_idx) &&
		    !test_bit(R5_LOCKED, &dev->flags) &&
2171 2172
		    !(test_bit(R5_UPTODATE, &dev->flags) ||
		      test_bit(R5_Wantcompute, &dev->flags))) {
2173 2174 2175 2176 2177 2178 2179 2180
			if (test_bit(R5_Insync, &dev->flags))
				rmw++;
			else
				rmw += 2*disks;  /* cannot read it */
		}
		/* Would I have to read this buffer for reconstruct_write */
		if (!test_bit(R5_OVERWRITE, &dev->flags) && i != sh->pd_idx &&
		    !test_bit(R5_LOCKED, &dev->flags) &&
2181 2182 2183
		    !(test_bit(R5_UPTODATE, &dev->flags) ||
		    test_bit(R5_Wantcompute, &dev->flags))) {
			if (test_bit(R5_Insync, &dev->flags)) rcw++;
2184 2185 2186 2187
			else
				rcw += 2*disks;
		}
	}
2188
	pr_debug("for sector %llu, rmw=%d rcw=%d\n",
2189 2190 2191 2192 2193 2194 2195 2196
		(unsigned long long)sh->sector, rmw, rcw);
	set_bit(STRIPE_HANDLE, &sh->state);
	if (rmw < rcw && rmw > 0)
		/* prefer read-modify-write, but need to get some data */
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if ((dev->towrite || i == sh->pd_idx) &&
			    !test_bit(R5_LOCKED, &dev->flags) &&
2197 2198
			    !(test_bit(R5_UPTODATE, &dev->flags) ||
			    test_bit(R5_Wantcompute, &dev->flags)) &&
2199 2200 2201
			    test_bit(R5_Insync, &dev->flags)) {
				if (
				  test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
2202
					pr_debug("Read_old block "
2203 2204 2205
						"%d for r-m-w\n", i);
					set_bit(R5_LOCKED, &dev->flags);
					set_bit(R5_Wantread, &dev->flags);
2206 2207 2208
					if (!test_and_set_bit(
						STRIPE_OP_IO, &sh->ops.pending))
						sh->ops.count++;
2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222
					s->locked++;
				} else {
					set_bit(STRIPE_DELAYED, &sh->state);
					set_bit(STRIPE_HANDLE, &sh->state);
				}
			}
		}
	if (rcw <= rmw && rcw > 0)
		/* want reconstruct write, but need to get some data */
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (!test_bit(R5_OVERWRITE, &dev->flags) &&
			    i != sh->pd_idx &&
			    !test_bit(R5_LOCKED, &dev->flags) &&
2223 2224
			    !(test_bit(R5_UPTODATE, &dev->flags) ||
			    test_bit(R5_Wantcompute, &dev->flags)) &&
2225 2226 2227
			    test_bit(R5_Insync, &dev->flags)) {
				if (
				  test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
2228
					pr_debug("Read_old block "
2229 2230 2231
						"%d for Reconstruct\n", i);
					set_bit(R5_LOCKED, &dev->flags);
					set_bit(R5_Wantread, &dev->flags);
2232 2233 2234
					if (!test_and_set_bit(
						STRIPE_OP_IO, &sh->ops.pending))
						sh->ops.count++;
2235 2236 2237 2238 2239 2240 2241 2242 2243 2244
					s->locked++;
				} else {
					set_bit(STRIPE_DELAYED, &sh->state);
					set_bit(STRIPE_HANDLE, &sh->state);
				}
			}
		}
	/* now if nothing is locked, and if we have enough data,
	 * we can start a write request
	 */
2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255
	/* since handle_stripe can be called at any time we need to handle the
	 * case where a compute block operation has been submitted and then a
	 * subsequent call wants to start a write request.  raid5_run_ops only
	 * handles the case where compute block and postxor are requested
	 * simultaneously.  If this is not the case then new writes need to be
	 * held off until the compute completes.
	 */
	if ((s->req_compute ||
	    !test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending)) &&
		(s->locked == 0 && (rcw == 0 || rmw == 0) &&
		!test_bit(STRIPE_BIT_DELAY, &sh->state)))
2256
		s->locked += handle_write_operations5(sh, rcw == 0, 0);
2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274
}

static void handle_issuing_new_write_requests6(raid5_conf_t *conf,
		struct stripe_head *sh,	struct stripe_head_state *s,
		struct r6_state *r6s, int disks)
{
	int rcw = 0, must_compute = 0, pd_idx = sh->pd_idx, i;
	int qd_idx = r6s->qd_idx;
	for (i = disks; i--; ) {
		struct r5dev *dev = &sh->dev[i];
		/* Would I have to read this buffer for reconstruct_write */
		if (!test_bit(R5_OVERWRITE, &dev->flags)
		    && i != pd_idx && i != qd_idx
		    && (!test_bit(R5_LOCKED, &dev->flags)
			    ) &&
		    !test_bit(R5_UPTODATE, &dev->flags)) {
			if (test_bit(R5_Insync, &dev->flags)) rcw++;
			else {
2275
				pr_debug("raid6: must_compute: "
2276 2277 2278 2279 2280
					"disk %d flags=%#lx\n", i, dev->flags);
				must_compute++;
			}
		}
	}
2281
	pr_debug("for sector %llu, rcw=%d, must_compute=%d\n",
2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295
	       (unsigned long long)sh->sector, rcw, must_compute);
	set_bit(STRIPE_HANDLE, &sh->state);

	if (rcw > 0)
		/* want reconstruct write, but need to get some data */
		for (i = disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			if (!test_bit(R5_OVERWRITE, &dev->flags)
			    && !(s->failed == 0 && (i == pd_idx || i == qd_idx))
			    && !test_bit(R5_LOCKED, &dev->flags) &&
			    !test_bit(R5_UPTODATE, &dev->flags) &&
			    test_bit(R5_Insync, &dev->flags)) {
				if (
				  test_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
2296
					pr_debug("Read_old stripe %llu "
2297 2298 2299 2300 2301 2302
						"block %d for Reconstruct\n",
					     (unsigned long long)sh->sector, i);
					set_bit(R5_LOCKED, &dev->flags);
					set_bit(R5_Wantread, &dev->flags);
					s->locked++;
				} else {
2303
					pr_debug("Request delayed stripe %llu "
2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332
						"block %d for Reconstruct\n",
					     (unsigned long long)sh->sector, i);
					set_bit(STRIPE_DELAYED, &sh->state);
					set_bit(STRIPE_HANDLE, &sh->state);
				}
			}
		}
	/* now if nothing is locked, and if we have enough data, we can start a
	 * write request
	 */
	if (s->locked == 0 && rcw == 0 &&
	    !test_bit(STRIPE_BIT_DELAY, &sh->state)) {
		if (must_compute > 0) {
			/* We have failed blocks and need to compute them */
			switch (s->failed) {
			case 0:
				BUG();
			case 1:
				compute_block_1(sh, r6s->failed_num[0], 0);
				break;
			case 2:
				compute_block_2(sh, r6s->failed_num[0],
						r6s->failed_num[1]);
				break;
			default: /* This request should have been failed? */
				BUG();
			}
		}

2333
		pr_debug("Computing parity for stripe %llu\n",
2334 2335 2336 2337 2338
			(unsigned long long)sh->sector);
		compute_parity6(sh, RECONSTRUCT_WRITE);
		/* now every locked buffer is ready to be written */
		for (i = disks; i--; )
			if (test_bit(R5_LOCKED, &sh->dev[i].flags)) {
2339
				pr_debug("Writing stripe %llu block %d\n",
2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359
				       (unsigned long long)sh->sector, i);
				s->locked++;
				set_bit(R5_Wantwrite, &sh->dev[i].flags);
			}
		/* after a RECONSTRUCT_WRITE, the stripe MUST be in-sync */
		set_bit(STRIPE_INSYNC, &sh->state);

		if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
			atomic_dec(&conf->preread_active_stripes);
			if (atomic_read(&conf->preread_active_stripes) <
			    IO_THRESHOLD)
				md_wakeup_thread(conf->mddev->thread);
		}
	}
}

static void handle_parity_checks5(raid5_conf_t *conf, struct stripe_head *sh,
				struct stripe_head_state *s, int disks)
{
	set_bit(STRIPE_HANDLE, &sh->state);
2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381
	/* Take one of the following actions:
	 * 1/ start a check parity operation if (uptodate == disks)
	 * 2/ finish a check parity operation and act on the result
	 * 3/ skip to the writeback section if we previously
	 *    initiated a recovery operation
	 */
	if (s->failed == 0 &&
	    !test_bit(STRIPE_OP_MOD_REPAIR_PD, &sh->ops.pending)) {
		if (!test_and_set_bit(STRIPE_OP_CHECK, &sh->ops.pending)) {
			BUG_ON(s->uptodate != disks);
			clear_bit(R5_UPTODATE, &sh->dev[sh->pd_idx].flags);
			sh->ops.count++;
			s->uptodate--;
		} else if (
		       test_and_clear_bit(STRIPE_OP_CHECK, &sh->ops.complete)) {
			clear_bit(STRIPE_OP_CHECK, &sh->ops.ack);
			clear_bit(STRIPE_OP_CHECK, &sh->ops.pending);

			if (sh->ops.zero_sum_result == 0)
				/* parity is correct (on disc,
				 * not in buffer any more)
				 */
2382 2383
				set_bit(STRIPE_INSYNC, &sh->state);
			else {
2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400
				conf->mddev->resync_mismatches +=
					STRIPE_SECTORS;
				if (test_bit(
				     MD_RECOVERY_CHECK, &conf->mddev->recovery))
					/* don't try to repair!! */
					set_bit(STRIPE_INSYNC, &sh->state);
				else {
					set_bit(STRIPE_OP_COMPUTE_BLK,
						&sh->ops.pending);
					set_bit(STRIPE_OP_MOD_REPAIR_PD,
						&sh->ops.pending);
					set_bit(R5_Wantcompute,
						&sh->dev[sh->pd_idx].flags);
					sh->ops.target = sh->pd_idx;
					sh->ops.count++;
					s->uptodate++;
				}
2401 2402 2403
			}
		}
	}
2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420

	/* check if we can clear a parity disk reconstruct */
	if (test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.complete) &&
		test_bit(STRIPE_OP_MOD_REPAIR_PD, &sh->ops.pending)) {

		clear_bit(STRIPE_OP_MOD_REPAIR_PD, &sh->ops.pending);
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.complete);
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.ack);
		clear_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending);
	}

	/* Wait for check parity and compute block operations to complete
	 * before write-back
	 */
	if (!test_bit(STRIPE_INSYNC, &sh->state) &&
		!test_bit(STRIPE_OP_CHECK, &sh->ops.pending) &&
		!test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending)) {
2421 2422 2423 2424 2425 2426 2427 2428 2429 2430
		struct r5dev *dev;
		/* either failed parity check, or recovery is happening */
		if (s->failed == 0)
			s->failed_num = sh->pd_idx;
		dev = &sh->dev[s->failed_num];
		BUG_ON(!test_bit(R5_UPTODATE, &dev->flags));
		BUG_ON(s->uptodate != disks);

		set_bit(R5_LOCKED, &dev->flags);
		set_bit(R5_Wantwrite, &dev->flags);
2431 2432 2433
		if (!test_and_set_bit(STRIPE_OP_IO, &sh->ops.pending))
			sh->ops.count++;

2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541
		clear_bit(STRIPE_DEGRADED, &sh->state);
		s->locked++;
		set_bit(STRIPE_INSYNC, &sh->state);
	}
}


static void handle_parity_checks6(raid5_conf_t *conf, struct stripe_head *sh,
				struct stripe_head_state *s,
				struct r6_state *r6s, struct page *tmp_page,
				int disks)
{
	int update_p = 0, update_q = 0;
	struct r5dev *dev;
	int pd_idx = sh->pd_idx;
	int qd_idx = r6s->qd_idx;

	set_bit(STRIPE_HANDLE, &sh->state);

	BUG_ON(s->failed > 2);
	BUG_ON(s->uptodate < disks);
	/* Want to check and possibly repair P and Q.
	 * However there could be one 'failed' device, in which
	 * case we can only check one of them, possibly using the
	 * other to generate missing data
	 */

	/* If !tmp_page, we cannot do the calculations,
	 * but as we have set STRIPE_HANDLE, we will soon be called
	 * by stripe_handle with a tmp_page - just wait until then.
	 */
	if (tmp_page) {
		if (s->failed == r6s->q_failed) {
			/* The only possible failed device holds 'Q', so it
			 * makes sense to check P (If anything else were failed,
			 * we would have used P to recreate it).
			 */
			compute_block_1(sh, pd_idx, 1);
			if (!page_is_zero(sh->dev[pd_idx].page)) {
				compute_block_1(sh, pd_idx, 0);
				update_p = 1;
			}
		}
		if (!r6s->q_failed && s->failed < 2) {
			/* q is not failed, and we didn't use it to generate
			 * anything, so it makes sense to check it
			 */
			memcpy(page_address(tmp_page),
			       page_address(sh->dev[qd_idx].page),
			       STRIPE_SIZE);
			compute_parity6(sh, UPDATE_PARITY);
			if (memcmp(page_address(tmp_page),
				   page_address(sh->dev[qd_idx].page),
				   STRIPE_SIZE) != 0) {
				clear_bit(STRIPE_INSYNC, &sh->state);
				update_q = 1;
			}
		}
		if (update_p || update_q) {
			conf->mddev->resync_mismatches += STRIPE_SECTORS;
			if (test_bit(MD_RECOVERY_CHECK, &conf->mddev->recovery))
				/* don't try to repair!! */
				update_p = update_q = 0;
		}

		/* now write out any block on a failed drive,
		 * or P or Q if they need it
		 */

		if (s->failed == 2) {
			dev = &sh->dev[r6s->failed_num[1]];
			s->locked++;
			set_bit(R5_LOCKED, &dev->flags);
			set_bit(R5_Wantwrite, &dev->flags);
		}
		if (s->failed >= 1) {
			dev = &sh->dev[r6s->failed_num[0]];
			s->locked++;
			set_bit(R5_LOCKED, &dev->flags);
			set_bit(R5_Wantwrite, &dev->flags);
		}

		if (update_p) {
			dev = &sh->dev[pd_idx];
			s->locked++;
			set_bit(R5_LOCKED, &dev->flags);
			set_bit(R5_Wantwrite, &dev->flags);
		}
		if (update_q) {
			dev = &sh->dev[qd_idx];
			s->locked++;
			set_bit(R5_LOCKED, &dev->flags);
			set_bit(R5_Wantwrite, &dev->flags);
		}
		clear_bit(STRIPE_DEGRADED, &sh->state);

		set_bit(STRIPE_INSYNC, &sh->state);
	}
}

static void handle_stripe_expansion(raid5_conf_t *conf, struct stripe_head *sh,
				struct r6_state *r6s)
{
	int i;

	/* We have read all the blocks in this stripe and now we need to
	 * copy some of them into a target stripe for expand.
	 */
2542
	struct dma_async_tx_descriptor *tx = NULL;
2543 2544
	clear_bit(STRIPE_EXPAND_SOURCE, &sh->state);
	for (i = 0; i < sh->disks; i++)
2545
		if (i != sh->pd_idx && (!r6s || i != r6s->qd_idx)) {
2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567
			int dd_idx, pd_idx, j;
			struct stripe_head *sh2;

			sector_t bn = compute_blocknr(sh, i);
			sector_t s = raid5_compute_sector(bn, conf->raid_disks,
						conf->raid_disks -
						conf->max_degraded, &dd_idx,
						&pd_idx, conf);
			sh2 = get_active_stripe(conf, s, conf->raid_disks,
						pd_idx, 1);
			if (sh2 == NULL)
				/* so far only the early blocks of this stripe
				 * have been requested.  When later blocks
				 * get requested, we will try again
				 */
				continue;
			if (!test_bit(STRIPE_EXPANDING, &sh2->state) ||
			   test_bit(R5_Expanded, &sh2->dev[dd_idx].flags)) {
				/* must have already done this block */
				release_stripe(sh2);
				continue;
			}
2568 2569 2570 2571 2572 2573

			/* place all the copies on one channel */
			tx = async_memcpy(sh2->dev[dd_idx].page,
				sh->dev[i].page, 0, 0, STRIPE_SIZE,
				ASYNC_TX_DEP_ACK, tx, NULL, NULL);

2574 2575 2576 2577
			set_bit(R5_Expanded, &sh2->dev[dd_idx].flags);
			set_bit(R5_UPTODATE, &sh2->dev[dd_idx].flags);
			for (j = 0; j < conf->raid_disks; j++)
				if (j != sh2->pd_idx &&
2578 2579
				    (!r6s || j != raid6_next_disk(sh2->pd_idx,
								 sh2->disks)) &&
2580 2581 2582 2583 2584 2585 2586
				    !test_bit(R5_Expanded, &sh2->dev[j].flags))
					break;
			if (j == conf->raid_disks) {
				set_bit(STRIPE_EXPAND_READY, &sh2->state);
				set_bit(STRIPE_HANDLE, &sh2->state);
			}
			release_stripe(sh2);
2587

2588
		}
2589 2590 2591 2592 2593
	/* done submitting copies, wait for them to complete */
	if (tx) {
		async_tx_ack(tx);
		dma_wait_for_async_tx(tx);
	}
2594
}
Linus Torvalds's avatar
Linus Torvalds committed
2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611

/*
 * handle_stripe - do things to a stripe.
 *
 * We lock the stripe and then examine the state of various bits
 * to see what needs to be done.
 * Possible results:
 *    return some read request which now have data
 *    return some write requests which are safely on disc
 *    schedule a read on some buffers
 *    schedule a write of some buffers
 *    return confirmation of parity correctness
 *
 * buffers are taken off read_list or write_list, and bh_cache buffers
 * get BH_Lock set before the stripe lock is released.
 *
 */
2612

2613
static void handle_stripe5(struct stripe_head *sh)
Linus Torvalds's avatar
Linus Torvalds committed
2614 2615
{
	raid5_conf_t *conf = sh->raid_conf;
2616 2617 2618
	int disks = sh->disks, i;
	struct bio *return_bi = NULL;
	struct stripe_head_state s;
Linus Torvalds's avatar
Linus Torvalds committed
2619
	struct r5dev *dev;
2620
	unsigned long pending = 0;
Linus Torvalds's avatar
Linus Torvalds committed
2621

2622
	memset(&s, 0, sizeof(s));
2623 2624 2625 2626
	pr_debug("handling stripe %llu, state=%#lx cnt=%d, pd_idx=%d "
		"ops=%lx:%lx:%lx\n", (unsigned long long)sh->sector, sh->state,
		atomic_read(&sh->count), sh->pd_idx,
		sh->ops.pending, sh->ops.ack, sh->ops.complete);
Linus Torvalds's avatar
Linus Torvalds committed
2627 2628 2629 2630 2631

	spin_lock(&sh->lock);
	clear_bit(STRIPE_HANDLE, &sh->state);
	clear_bit(STRIPE_DELAYED, &sh->state);

2632 2633 2634
	s.syncing = test_bit(STRIPE_SYNCING, &sh->state);
	s.expanding = test_bit(STRIPE_EXPAND_SOURCE, &sh->state);
	s.expanded = test_bit(STRIPE_EXPAND_READY, &sh->state);
Linus Torvalds's avatar
Linus Torvalds committed
2635 2636
	/* Now to look around and see what can be done */

2637 2638 2639 2640 2641 2642 2643
	/* clean-up completed biofill operations */
	if (test_bit(STRIPE_OP_BIOFILL, &sh->ops.complete)) {
		clear_bit(STRIPE_OP_BIOFILL, &sh->ops.pending);
		clear_bit(STRIPE_OP_BIOFILL, &sh->ops.ack);
		clear_bit(STRIPE_OP_BIOFILL, &sh->ops.complete);
	}

2644
	rcu_read_lock();
Linus Torvalds's avatar
Linus Torvalds committed
2645 2646
	for (i=disks; i--; ) {
		mdk_rdev_t *rdev;
2647
		struct r5dev *dev = &sh->dev[i];
Linus Torvalds's avatar
Linus Torvalds committed
2648 2649
		clear_bit(R5_Insync, &dev->flags);

2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661
		pr_debug("check %d: state 0x%lx toread %p read %p write %p "
			"written %p\n",	i, dev->flags, dev->toread, dev->read,
			dev->towrite, dev->written);

		/* maybe we can request a biofill operation
		 *
		 * new wantfill requests are only permitted while
		 * STRIPE_OP_BIOFILL is clear
		 */
		if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread &&
			!test_bit(STRIPE_OP_BIOFILL, &sh->ops.pending))
			set_bit(R5_Wantfill, &dev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
2662 2663

		/* now count some things */
2664 2665
		if (test_bit(R5_LOCKED, &dev->flags)) s.locked++;
		if (test_bit(R5_UPTODATE, &dev->flags)) s.uptodate++;
2666
		if (test_bit(R5_Wantcompute, &dev->flags)) s.compute++;
Linus Torvalds's avatar
Linus Torvalds committed
2667

2668 2669 2670
		if (test_bit(R5_Wantfill, &dev->flags))
			s.to_fill++;
		else if (dev->toread)
2671
			s.to_read++;
Linus Torvalds's avatar
Linus Torvalds committed
2672
		if (dev->towrite) {
2673
			s.to_write++;
Linus Torvalds's avatar
Linus Torvalds committed
2674
			if (!test_bit(R5_OVERWRITE, &dev->flags))
2675
				s.non_overwrite++;
Linus Torvalds's avatar
Linus Torvalds committed
2676
		}
2677 2678
		if (dev->written)
			s.written++;
2679
		rdev = rcu_dereference(conf->disks[i].rdev);
2680
		if (!rdev || !test_bit(In_sync, &rdev->flags)) {
2681
			/* The ReadError flag will just be confusing now */
2682 2683 2684
			clear_bit(R5_ReadError, &dev->flags);
			clear_bit(R5_ReWrite, &dev->flags);
		}
2685
		if (!rdev || !test_bit(In_sync, &rdev->flags)
2686
		    || test_bit(R5_ReadError, &dev->flags)) {
2687 2688
			s.failed++;
			s.failed_num = i;
Linus Torvalds's avatar
Linus Torvalds committed
2689 2690 2691
		} else
			set_bit(R5_Insync, &dev->flags);
	}
2692
	rcu_read_unlock();
2693 2694 2695 2696

	if (s.to_fill && !test_and_set_bit(STRIPE_OP_BIOFILL, &sh->ops.pending))
		sh->ops.count++;

2697
	pr_debug("locked=%d uptodate=%d to_read=%d"
Linus Torvalds's avatar
Linus Torvalds committed
2698
		" to_write=%d failed=%d failed_num=%d\n",
2699 2700
		s.locked, s.uptodate, s.to_read, s.to_write,
		s.failed, s.failed_num);
Linus Torvalds's avatar
Linus Torvalds committed
2701 2702 2703
	/* check if the array has lost two devices and, if so, some requests might
	 * need to be failed
	 */
2704 2705 2706 2707
	if (s.failed > 1 && s.to_read+s.to_write+s.written)
		handle_requests_to_failed_array(conf, sh, &s, disks,
						&return_bi);
	if (s.failed > 1 && s.syncing) {
Linus Torvalds's avatar
Linus Torvalds committed
2708 2709
		md_done_sync(conf->mddev, STRIPE_SECTORS,0);
		clear_bit(STRIPE_SYNCING, &sh->state);
2710
		s.syncing = 0;
Linus Torvalds's avatar
Linus Torvalds committed
2711 2712 2713 2714 2715 2716
	}

	/* might be able to return some write requests if the parity block
	 * is safe, or on a failed drive
	 */
	dev = &sh->dev[sh->pd_idx];
2717 2718 2719 2720 2721 2722
	if ( s.written &&
	     ((test_bit(R5_Insync, &dev->flags) &&
	       !test_bit(R5_LOCKED, &dev->flags) &&
	       test_bit(R5_UPTODATE, &dev->flags)) ||
	       (s.failed == 1 && s.failed_num == sh->pd_idx)))
		handle_completed_write_requests(conf, sh, disks, &return_bi);
Linus Torvalds's avatar
Linus Torvalds committed
2723 2724 2725 2726 2727

	/* Now we might consider reading some blocks, either to check/generate
	 * parity, or to satisfy requests
	 * or to load a block that is being partially written.
	 */
2728
	if (s.to_read || s.non_overwrite ||
2729 2730
	    (s.syncing && (s.uptodate + s.compute < disks)) || s.expanding ||
	    test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending))
2731
		handle_issuing_new_read_requests5(sh, &s, disks);
Linus Torvalds's avatar
Linus Torvalds committed
2732

2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796
	/* Now we check to see if any write operations have recently
	 * completed
	 */

	/* leave prexor set until postxor is done, allows us to distinguish
	 * a rmw from a rcw during biodrain
	 */
	if (test_bit(STRIPE_OP_PREXOR, &sh->ops.complete) &&
		test_bit(STRIPE_OP_POSTXOR, &sh->ops.complete)) {

		clear_bit(STRIPE_OP_PREXOR, &sh->ops.complete);
		clear_bit(STRIPE_OP_PREXOR, &sh->ops.ack);
		clear_bit(STRIPE_OP_PREXOR, &sh->ops.pending);

		for (i = disks; i--; )
			clear_bit(R5_Wantprexor, &sh->dev[i].flags);
	}

	/* if only POSTXOR is set then this is an 'expand' postxor */
	if (test_bit(STRIPE_OP_BIODRAIN, &sh->ops.complete) &&
		test_bit(STRIPE_OP_POSTXOR, &sh->ops.complete)) {

		clear_bit(STRIPE_OP_BIODRAIN, &sh->ops.complete);
		clear_bit(STRIPE_OP_BIODRAIN, &sh->ops.ack);
		clear_bit(STRIPE_OP_BIODRAIN, &sh->ops.pending);

		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.complete);
		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.ack);
		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.pending);

		/* All the 'written' buffers and the parity block are ready to
		 * be written back to disk
		 */
		BUG_ON(!test_bit(R5_UPTODATE, &sh->dev[sh->pd_idx].flags));
		for (i = disks; i--; ) {
			dev = &sh->dev[i];
			if (test_bit(R5_LOCKED, &dev->flags) &&
				(i == sh->pd_idx || dev->written)) {
				pr_debug("Writing block %d\n", i);
				set_bit(R5_Wantwrite, &dev->flags);
				if (!test_and_set_bit(
				    STRIPE_OP_IO, &sh->ops.pending))
					sh->ops.count++;
				if (!test_bit(R5_Insync, &dev->flags) ||
				    (i == sh->pd_idx && s.failed == 0))
					set_bit(STRIPE_INSYNC, &sh->state);
			}
		}
		if (test_and_clear_bit(STRIPE_PREREAD_ACTIVE, &sh->state)) {
			atomic_dec(&conf->preread_active_stripes);
			if (atomic_read(&conf->preread_active_stripes) <
				IO_THRESHOLD)
				md_wakeup_thread(conf->mddev->thread);
		}
	}

	/* Now to consider new write requests and what else, if anything
	 * should be read.  We do not handle new writes when:
	 * 1/ A 'write' operation (copy+xor) is already in flight.
	 * 2/ A 'check' operation is in flight, as it may clobber the parity
	 *    block.
	 */
	if (s.to_write && !test_bit(STRIPE_OP_POSTXOR, &sh->ops.pending) &&
			  !test_bit(STRIPE_OP_CHECK, &sh->ops.pending))
2797
		handle_issuing_new_write_requests5(conf, sh, &s, disks);
Linus Torvalds's avatar
Linus Torvalds committed
2798 2799

	/* maybe we need to check and possibly fix the parity for this stripe
2800 2801 2802
	 * Any reads will already have been scheduled, so we just see if enough
	 * data is available.  The parity check is held off while parity
	 * dependent operations are in flight.
Linus Torvalds's avatar
Linus Torvalds committed
2803
	 */
2804 2805 2806 2807 2808
	if ((s.syncing && s.locked == 0 &&
	     !test_bit(STRIPE_OP_COMPUTE_BLK, &sh->ops.pending) &&
	     !test_bit(STRIPE_INSYNC, &sh->state)) ||
	      test_bit(STRIPE_OP_CHECK, &sh->ops.pending) ||
	      test_bit(STRIPE_OP_MOD_REPAIR_PD, &sh->ops.pending))
2809
		handle_parity_checks5(conf, sh, &s, disks);
2810

2811
	if (s.syncing && s.locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
Linus Torvalds's avatar
Linus Torvalds committed
2812 2813 2814
		md_done_sync(conf->mddev, STRIPE_SECTORS,1);
		clear_bit(STRIPE_SYNCING, &sh->state);
	}
2815 2816 2817 2818

	/* If the failed drive is just a ReadError, then we might need to progress
	 * the repair/check process
	 */
2819 2820 2821 2822
	if (s.failed == 1 && !conf->mddev->ro &&
	    test_bit(R5_ReadError, &sh->dev[s.failed_num].flags)
	    && !test_bit(R5_LOCKED, &sh->dev[s.failed_num].flags)
	    && test_bit(R5_UPTODATE, &sh->dev[s.failed_num].flags)
2823
		) {
2824
		dev = &sh->dev[s.failed_num];
2825 2826
		if (!test_bit(R5_ReWrite, &dev->flags)) {
			set_bit(R5_Wantwrite, &dev->flags);
2827 2828
			if (!test_and_set_bit(STRIPE_OP_IO, &sh->ops.pending))
				sh->ops.count++;
2829 2830
			set_bit(R5_ReWrite, &dev->flags);
			set_bit(R5_LOCKED, &dev->flags);
2831
			s.locked++;
2832 2833 2834
		} else {
			/* let's read it back */
			set_bit(R5_Wantread, &dev->flags);
2835 2836
			if (!test_and_set_bit(STRIPE_OP_IO, &sh->ops.pending))
				sh->ops.count++;
2837
			set_bit(R5_LOCKED, &dev->flags);
2838
			s.locked++;
2839 2840 2841
		}
	}

2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853
	/* Finish postxor operations initiated by the expansion
	 * process
	 */
	if (test_bit(STRIPE_OP_POSTXOR, &sh->ops.complete) &&
		!test_bit(STRIPE_OP_BIODRAIN, &sh->ops.pending)) {

		clear_bit(STRIPE_EXPANDING, &sh->state);

		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.pending);
		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.ack);
		clear_bit(STRIPE_OP_POSTXOR, &sh->ops.complete);

2854
		for (i = conf->raid_disks; i--; ) {
2855
			set_bit(R5_Wantwrite, &sh->dev[i].flags);
2856 2857
			if (!test_and_set_bit(STRIPE_OP_IO, &sh->ops.pending))
				sh->ops.count++;
2858
		}
2859 2860 2861 2862 2863 2864 2865 2866
	}

	if (s.expanded && test_bit(STRIPE_EXPANDING, &sh->state) &&
		!test_bit(STRIPE_OP_POSTXOR, &sh->ops.pending)) {
		/* Need to write out all blocks after computing parity */
		sh->disks = conf->raid_disks;
		sh->pd_idx = stripe_to_pdidx(sh->sector, conf,
			conf->raid_disks);
2867
		s.locked += handle_write_operations5(sh, 1, 1);
2868 2869
	} else if (s.expanded &&
		!test_bit(STRIPE_OP_POSTXOR, &sh->ops.pending)) {
2870
		clear_bit(STRIPE_EXPAND_READY, &sh->state);
2871
		atomic_dec(&conf->reshape_stripes);
2872 2873 2874 2875
		wake_up(&conf->wait_for_overlap);
		md_done_sync(conf->mddev, STRIPE_SECTORS, 1);
	}

2876 2877
	if (s.expanding && s.locked == 0)
		handle_stripe_expansion(conf, sh, NULL);
2878

2879 2880 2881
	if (sh->ops.count)
		pending = get_stripe_work(sh);

Linus Torvalds's avatar
Linus Torvalds committed
2882 2883
	spin_unlock(&sh->lock);

2884 2885 2886
	if (pending)
		raid5_run_ops(sh, pending);

2887
	return_io(return_bi);
Linus Torvalds's avatar
Linus Torvalds committed
2888 2889 2890

}

2891
static void handle_stripe6(struct stripe_head *sh, struct page *tmp_page)
Linus Torvalds's avatar
Linus Torvalds committed
2892
{
2893
	raid6_conf_t *conf = sh->raid_conf;
2894
	int disks = sh->disks;
2895 2896 2897 2898
	struct bio *return_bi = NULL;
	int i, pd_idx = sh->pd_idx;
	struct stripe_head_state s;
	struct r6_state r6s;
2899
	struct r5dev *dev, *pdev, *qdev;
Linus Torvalds's avatar
Linus Torvalds committed
2900

2901
	r6s.qd_idx = raid6_next_disk(pd_idx, disks);
2902
	pr_debug("handling stripe %llu, state=%#lx cnt=%d, "
2903 2904 2905 2906
		"pd_idx=%d, qd_idx=%d\n",
	       (unsigned long long)sh->sector, sh->state,
	       atomic_read(&sh->count), pd_idx, r6s.qd_idx);
	memset(&s, 0, sizeof(s));
2907

2908 2909 2910 2911
	spin_lock(&sh->lock);
	clear_bit(STRIPE_HANDLE, &sh->state);
	clear_bit(STRIPE_DELAYED, &sh->state);

2912 2913 2914
	s.syncing = test_bit(STRIPE_SYNCING, &sh->state);
	s.expanding = test_bit(STRIPE_EXPAND_SOURCE, &sh->state);
	s.expanded = test_bit(STRIPE_EXPAND_READY, &sh->state);
2915
	/* Now to look around and see what can be done */
Linus Torvalds's avatar
Linus Torvalds committed
2916 2917

	rcu_read_lock();
2918 2919 2920 2921
	for (i=disks; i--; ) {
		mdk_rdev_t *rdev;
		dev = &sh->dev[i];
		clear_bit(R5_Insync, &dev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
2922

2923
		pr_debug("check %d: state 0x%lx read %p write %p written %p\n",
2924 2925 2926 2927
			i, dev->flags, dev->toread, dev->towrite, dev->written);
		/* maybe we can reply to a read */
		if (test_bit(R5_UPTODATE, &dev->flags) && dev->toread) {
			struct bio *rbi, *rbi2;
2928
			pr_debug("Return read for disc %d\n", i);
2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946
			spin_lock_irq(&conf->device_lock);
			rbi = dev->toread;
			dev->toread = NULL;
			if (test_and_clear_bit(R5_Overlap, &dev->flags))
				wake_up(&conf->wait_for_overlap);
			spin_unlock_irq(&conf->device_lock);
			while (rbi && rbi->bi_sector < dev->sector + STRIPE_SECTORS) {
				copy_data(0, rbi, dev->page, dev->sector);
				rbi2 = r5_next_bio(rbi, dev->sector);
				spin_lock_irq(&conf->device_lock);
				if (--rbi->bi_phys_segments == 0) {
					rbi->bi_next = return_bi;
					return_bi = rbi;
				}
				spin_unlock_irq(&conf->device_lock);
				rbi = rbi2;
			}
		}
Linus Torvalds's avatar
Linus Torvalds committed
2947

2948
		/* now count some things */
2949 2950
		if (test_bit(R5_LOCKED, &dev->flags)) s.locked++;
		if (test_bit(R5_UPTODATE, &dev->flags)) s.uptodate++;
Linus Torvalds's avatar
Linus Torvalds committed
2951

2952

2953 2954
		if (dev->toread)
			s.to_read++;
2955
		if (dev->towrite) {
2956
			s.to_write++;
2957
			if (!test_bit(R5_OVERWRITE, &dev->flags))
2958
				s.non_overwrite++;
2959
		}
2960 2961
		if (dev->written)
			s.written++;
2962 2963 2964 2965 2966
		rdev = rcu_dereference(conf->disks[i].rdev);
		if (!rdev || !test_bit(In_sync, &rdev->flags)) {
			/* The ReadError flag will just be confusing now */
			clear_bit(R5_ReadError, &dev->flags);
			clear_bit(R5_ReWrite, &dev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
2967
		}
2968 2969
		if (!rdev || !test_bit(In_sync, &rdev->flags)
		    || test_bit(R5_ReadError, &dev->flags)) {
2970 2971 2972
			if (s.failed < 2)
				r6s.failed_num[s.failed] = i;
			s.failed++;
2973 2974
		} else
			set_bit(R5_Insync, &dev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
2975 2976
	}
	rcu_read_unlock();
2977
	pr_debug("locked=%d uptodate=%d to_read=%d"
2978
	       " to_write=%d failed=%d failed_num=%d,%d\n",
2979 2980 2981 2982
	       s.locked, s.uptodate, s.to_read, s.to_write, s.failed,
	       r6s.failed_num[0], r6s.failed_num[1]);
	/* check if the array has lost >2 devices and, if so, some requests
	 * might need to be failed
2983
	 */
2984 2985 2986 2987
	if (s.failed > 2 && s.to_read+s.to_write+s.written)
		handle_requests_to_failed_array(conf, sh, &s, disks,
						&return_bi);
	if (s.failed > 2 && s.syncing) {
2988 2989
		md_done_sync(conf->mddev, STRIPE_SECTORS,0);
		clear_bit(STRIPE_SYNCING, &sh->state);
2990
		s.syncing = 0;
2991 2992 2993 2994 2995 2996 2997
	}

	/*
	 * might be able to return some write requests if the parity blocks
	 * are safe, or on a failed drive
	 */
	pdev = &sh->dev[pd_idx];
2998 2999 3000 3001 3002 3003 3004 3005
	r6s.p_failed = (s.failed >= 1 && r6s.failed_num[0] == pd_idx)
		|| (s.failed >= 2 && r6s.failed_num[1] == pd_idx);
	qdev = &sh->dev[r6s.qd_idx];
	r6s.q_failed = (s.failed >= 1 && r6s.failed_num[0] == r6s.qd_idx)
		|| (s.failed >= 2 && r6s.failed_num[1] == r6s.qd_idx);

	if ( s.written &&
	     ( r6s.p_failed || ((test_bit(R5_Insync, &pdev->flags)
3006
			     && !test_bit(R5_LOCKED, &pdev->flags)
3007 3008
			     && test_bit(R5_UPTODATE, &pdev->flags)))) &&
	     ( r6s.q_failed || ((test_bit(R5_Insync, &qdev->flags)
3009
			     && !test_bit(R5_LOCKED, &qdev->flags)
3010 3011
			     && test_bit(R5_UPTODATE, &qdev->flags)))))
		handle_completed_write_requests(conf, sh, disks, &return_bi);
3012 3013 3014 3015 3016

	/* Now we might consider reading some blocks, either to check/generate
	 * parity, or to satisfy requests
	 * or to load a block that is being partially written.
	 */
3017 3018 3019
	if (s.to_read || s.non_overwrite || (s.to_write && s.failed) ||
	    (s.syncing && (s.uptodate < disks)) || s.expanding)
		handle_issuing_new_read_requests6(sh, &s, &r6s, disks);
3020 3021

	/* now to consider writing and what else, if anything should be read */
3022 3023
	if (s.to_write)
		handle_issuing_new_write_requests6(conf, sh, &s, &r6s, disks);
3024 3025

	/* maybe we need to check and possibly fix the parity for this stripe
3026 3027
	 * Any reads will already have been scheduled, so we just see if enough
	 * data is available
3028
	 */
3029 3030
	if (s.syncing && s.locked == 0 && !test_bit(STRIPE_INSYNC, &sh->state))
		handle_parity_checks6(conf, sh, &s, &r6s, tmp_page, disks);
3031

3032
	if (s.syncing && s.locked == 0 && test_bit(STRIPE_INSYNC, &sh->state)) {
3033 3034 3035 3036 3037 3038 3039
		md_done_sync(conf->mddev, STRIPE_SECTORS,1);
		clear_bit(STRIPE_SYNCING, &sh->state);
	}

	/* If the failed drives are just a ReadError, then we might need
	 * to progress the repair/check process
	 */
3040 3041 3042
	if (s.failed <= 2 && !conf->mddev->ro)
		for (i = 0; i < s.failed; i++) {
			dev = &sh->dev[r6s.failed_num[i]];
3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057
			if (test_bit(R5_ReadError, &dev->flags)
			    && !test_bit(R5_LOCKED, &dev->flags)
			    && test_bit(R5_UPTODATE, &dev->flags)
				) {
				if (!test_bit(R5_ReWrite, &dev->flags)) {
					set_bit(R5_Wantwrite, &dev->flags);
					set_bit(R5_ReWrite, &dev->flags);
					set_bit(R5_LOCKED, &dev->flags);
				} else {
					/* let's read it back */
					set_bit(R5_Wantread, &dev->flags);
					set_bit(R5_LOCKED, &dev->flags);
				}
			}
		}
3058

3059
	if (s.expanded && test_bit(STRIPE_EXPANDING, &sh->state)) {
3060 3061 3062 3063 3064 3065 3066
		/* Need to write out all blocks after computing P&Q */
		sh->disks = conf->raid_disks;
		sh->pd_idx = stripe_to_pdidx(sh->sector, conf,
					     conf->raid_disks);
		compute_parity6(sh, RECONSTRUCT_WRITE);
		for (i = conf->raid_disks ; i-- ;  ) {
			set_bit(R5_LOCKED, &sh->dev[i].flags);
3067
			s.locked++;
3068 3069 3070
			set_bit(R5_Wantwrite, &sh->dev[i].flags);
		}
		clear_bit(STRIPE_EXPANDING, &sh->state);
3071
	} else if (s.expanded) {
3072 3073 3074 3075 3076 3077
		clear_bit(STRIPE_EXPAND_READY, &sh->state);
		atomic_dec(&conf->reshape_stripes);
		wake_up(&conf->wait_for_overlap);
		md_done_sync(conf->mddev, STRIPE_SECTORS, 1);
	}

3078 3079
	if (s.expanding && s.locked == 0)
		handle_stripe_expansion(conf, sh, &r6s);
3080

3081 3082
	spin_unlock(&sh->lock);

3083
	return_io(return_bi);
3084 3085 3086 3087 3088 3089

	for (i=disks; i-- ;) {
		int rw;
		struct bio *bi;
		mdk_rdev_t *rdev;
		if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags))
3090
			rw = WRITE;
3091
		else if (test_and_clear_bit(R5_Wantread, &sh->dev[i].flags))
3092
			rw = READ;
3093 3094 3095 3096 3097 3098
		else
			continue;

		bi = &sh->dev[i].req;

		bi->bi_rw = rw;
3099
		if (rw == WRITE)
3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112
			bi->bi_end_io = raid5_end_write_request;
		else
			bi->bi_end_io = raid5_end_read_request;

		rcu_read_lock();
		rdev = rcu_dereference(conf->disks[i].rdev);
		if (rdev && test_bit(Faulty, &rdev->flags))
			rdev = NULL;
		if (rdev)
			atomic_inc(&rdev->nr_pending);
		rcu_read_unlock();

		if (rdev) {
3113
			if (s.syncing || s.expanding || s.expanded)
3114 3115 3116
				md_sync_acct(rdev->bdev, STRIPE_SECTORS);

			bi->bi_bdev = rdev->bdev;
3117
			pr_debug("for %llu schedule op %ld on disc %d\n",
3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134
				(unsigned long long)sh->sector, bi->bi_rw, i);
			atomic_inc(&sh->count);
			bi->bi_sector = sh->sector + rdev->data_offset;
			bi->bi_flags = 1 << BIO_UPTODATE;
			bi->bi_vcnt = 1;
			bi->bi_max_vecs = 1;
			bi->bi_idx = 0;
			bi->bi_io_vec = &sh->dev[i].vec;
			bi->bi_io_vec[0].bv_len = STRIPE_SIZE;
			bi->bi_io_vec[0].bv_offset = 0;
			bi->bi_size = STRIPE_SIZE;
			bi->bi_next = NULL;
			if (rw == WRITE &&
			    test_bit(R5_ReWrite, &sh->dev[i].flags))
				atomic_add(STRIPE_SECTORS, &rdev->corrected_errors);
			generic_make_request(bi);
		} else {
3135
			if (rw == WRITE)
3136
				set_bit(STRIPE_DEGRADED, &sh->state);
3137
			pr_debug("skip op %ld on disc %d for sector %llu\n",
3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193
				bi->bi_rw, i, (unsigned long long)sh->sector);
			clear_bit(R5_LOCKED, &sh->dev[i].flags);
			set_bit(STRIPE_HANDLE, &sh->state);
		}
	}
}

static void handle_stripe(struct stripe_head *sh, struct page *tmp_page)
{
	if (sh->raid_conf->level == 6)
		handle_stripe6(sh, tmp_page);
	else
		handle_stripe5(sh);
}



static void raid5_activate_delayed(raid5_conf_t *conf)
{
	if (atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD) {
		while (!list_empty(&conf->delayed_list)) {
			struct list_head *l = conf->delayed_list.next;
			struct stripe_head *sh;
			sh = list_entry(l, struct stripe_head, lru);
			list_del_init(l);
			clear_bit(STRIPE_DELAYED, &sh->state);
			if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
				atomic_inc(&conf->preread_active_stripes);
			list_add_tail(&sh->lru, &conf->handle_list);
		}
	}
}

static void activate_bit_delay(raid5_conf_t *conf)
{
	/* device_lock is held */
	struct list_head head;
	list_add(&head, &conf->bitmap_list);
	list_del_init(&conf->bitmap_list);
	while (!list_empty(&head)) {
		struct stripe_head *sh = list_entry(head.next, struct stripe_head, lru);
		list_del_init(&sh->lru);
		atomic_inc(&sh->count);
		__release_stripe(conf, sh);
	}
}

static void unplug_slaves(mddev_t *mddev)
{
	raid5_conf_t *conf = mddev_to_conf(mddev);
	int i;

	rcu_read_lock();
	for (i=0; i<mddev->raid_disks; i++) {
		mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
		if (rdev && !test_bit(Faulty, &rdev->flags) && atomic_read(&rdev->nr_pending)) {
3194
			struct request_queue *r_queue = bdev_get_queue(rdev->bdev);
3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208

			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();
}

3209
static void raid5_unplug_device(struct request_queue *q)
3210 3211 3212 3213 3214 3215 3216 3217 3218 3219
{
	mddev_t *mddev = q->queuedata;
	raid5_conf_t *conf = mddev_to_conf(mddev);
	unsigned long flags;

	spin_lock_irqsave(&conf->device_lock, flags);

	if (blk_remove_plug(q)) {
		conf->seq_flush++;
		raid5_activate_delayed(conf);
3220
	}
Linus Torvalds's avatar
Linus Torvalds committed
3221 3222 3223 3224 3225 3226 3227
	md_wakeup_thread(mddev->thread);

	spin_unlock_irqrestore(&conf->device_lock, flags);

	unplug_slaves(mddev);
}

3228
static int raid5_issue_flush(struct request_queue *q, struct gendisk *disk,
Linus Torvalds's avatar
Linus Torvalds committed
3229 3230 3231 3232 3233 3234 3235 3236
			     sector_t *error_sector)
{
	mddev_t *mddev = q->queuedata;
	raid5_conf_t *conf = mddev_to_conf(mddev);
	int i, ret = 0;

	rcu_read_lock();
	for (i=0; i<mddev->raid_disks && ret == 0; i++) {
3237
		mdk_rdev_t *rdev = rcu_dereference(conf->disks[i].rdev);
3238
		if (rdev && !test_bit(Faulty, &rdev->flags)) {
Linus Torvalds's avatar
Linus Torvalds committed
3239
			struct block_device *bdev = rdev->bdev;
3240
			struct request_queue *r_queue = bdev_get_queue(bdev);
Linus Torvalds's avatar
Linus Torvalds committed
3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257

			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;
}

3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275
static int raid5_congested(void *data, int bits)
{
	mddev_t *mddev = data;
	raid5_conf_t *conf = mddev_to_conf(mddev);

	/* No difference between reads and writes.  Just check
	 * how busy the stripe_cache is
	 */
	if (conf->inactive_blocked)
		return 1;
	if (conf->quiesce)
		return 1;
	if (list_empty_careful(&conf->inactive_list))
		return 1;

	return 0;
}

3276 3277 3278
/* We want read requests to align with chunks where possible,
 * but write requests don't need to.
 */
3279
static int raid5_mergeable_bvec(struct request_queue *q, struct bio *bio, struct bio_vec *biovec)
3280 3281 3282 3283 3284 3285 3286
{
	mddev_t *mddev = q->queuedata;
	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
	int max;
	unsigned int chunk_sectors = mddev->chunk_size >> 9;
	unsigned int bio_sectors = bio->bi_size >> 9;

3287
	if (bio_data_dir(bio) == WRITE)
3288 3289 3290 3291 3292 3293 3294 3295 3296 3297
		return biovec->bv_len; /* always allow writes to be mergeable */

	max =  (chunk_sectors - ((sector & (chunk_sectors - 1)) + bio_sectors)) << 9;
	if (max < 0) max = 0;
	if (max <= biovec->bv_len && bio_sectors == 0)
		return biovec->bv_len;
	else
		return max;
}

3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308

static int in_chunk_boundary(mddev_t *mddev, struct bio *bio)
{
	sector_t sector = bio->bi_sector + get_start_sect(bio->bi_bdev);
	unsigned int chunk_sectors = mddev->chunk_size >> 9;
	unsigned int bio_sectors = bio->bi_size >> 9;

	return  chunk_sectors >=
		((sector & (chunk_sectors - 1)) + bio_sectors);
}

3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337
/*
 *  add bio to the retry LIFO  ( in O(1) ... we are in interrupt )
 *  later sampled by raid5d.
 */
static void add_bio_to_retry(struct bio *bi,raid5_conf_t *conf)
{
	unsigned long flags;

	spin_lock_irqsave(&conf->device_lock, flags);

	bi->bi_next = conf->retry_read_aligned_list;
	conf->retry_read_aligned_list = bi;

	spin_unlock_irqrestore(&conf->device_lock, flags);
	md_wakeup_thread(conf->mddev->thread);
}


static struct bio *remove_bio_from_retry(raid5_conf_t *conf)
{
	struct bio *bi;

	bi = conf->retry_read_aligned;
	if (bi) {
		conf->retry_read_aligned = NULL;
		return bi;
	}
	bi = conf->retry_read_aligned_list;
	if(bi) {
3338
		conf->retry_read_aligned_list = bi->bi_next;
3339 3340 3341 3342 3343 3344 3345 3346 3347
		bi->bi_next = NULL;
		bi->bi_phys_segments = 1; /* biased count of active stripes */
		bi->bi_hw_segments = 0; /* count of processed stripes */
	}

	return bi;
}


3348 3349 3350 3351 3352 3353
/*
 *  The "raid5_align_endio" should check if the read succeeded and if it
 *  did, call bio_endio on the original bio (having bio_put the new bio
 *  first).
 *  If the read failed..
 */
3354
static int raid5_align_endio(struct bio *bi, unsigned int bytes, int error)
3355 3356
{
	struct bio* raid_bi  = bi->bi_private;
3357 3358 3359 3360 3361
	mddev_t *mddev;
	raid5_conf_t *conf;
	int uptodate = test_bit(BIO_UPTODATE, &bi->bi_flags);
	mdk_rdev_t *rdev;

3362 3363 3364
	if (bi->bi_size)
		return 1;
	bio_put(bi);
3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380

	mddev = raid_bi->bi_bdev->bd_disk->queue->queuedata;
	conf = mddev_to_conf(mddev);
	rdev = (void*)raid_bi->bi_next;
	raid_bi->bi_next = NULL;

	rdev_dec_pending(rdev, conf->mddev);

	if (!error && uptodate) {
		bio_endio(raid_bi, bytes, 0);
		if (atomic_dec_and_test(&conf->active_aligned_reads))
			wake_up(&conf->wait_for_stripe);
		return 0;
	}


3381
	pr_debug("raid5_align_endio : io error...handing IO for a retry\n");
3382 3383

	add_bio_to_retry(raid_bi, conf);
3384 3385 3386
	return 0;
}

3387 3388
static int bio_fits_rdev(struct bio *bi)
{
3389
	struct request_queue *q = bdev_get_queue(bi->bi_bdev);
3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407

	if ((bi->bi_size>>9) > q->max_sectors)
		return 0;
	blk_recount_segments(q, bi);
	if (bi->bi_phys_segments > q->max_phys_segments ||
	    bi->bi_hw_segments > q->max_hw_segments)
		return 0;

	if (q->merge_bvec_fn)
		/* it's too hard to apply the merge_bvec_fn at this stage,
		 * just just give up
		 */
		return 0;

	return 1;
}


3408
static int chunk_aligned_read(struct request_queue *q, struct bio * raid_bio)
3409 3410 3411 3412
{
	mddev_t *mddev = q->queuedata;
	raid5_conf_t *conf = mddev_to_conf(mddev);
	const unsigned int raid_disks = conf->raid_disks;
3413
	const unsigned int data_disks = raid_disks - conf->max_degraded;
3414 3415 3416 3417 3418
	unsigned int dd_idx, pd_idx;
	struct bio* align_bi;
	mdk_rdev_t *rdev;

	if (!in_chunk_boundary(mddev, raid_bio)) {
3419
		pr_debug("chunk_aligned_read : non aligned\n");
3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448
		return 0;
	}
	/*
 	 * use bio_clone to make a copy of the bio
	 */
	align_bi = bio_clone(raid_bio, GFP_NOIO);
	if (!align_bi)
		return 0;
	/*
	 *   set bi_end_io to a new function, and set bi_private to the
	 *     original bio.
	 */
	align_bi->bi_end_io  = raid5_align_endio;
	align_bi->bi_private = raid_bio;
	/*
	 *	compute position
	 */
	align_bi->bi_sector =  raid5_compute_sector(raid_bio->bi_sector,
					raid_disks,
					data_disks,
					&dd_idx,
					&pd_idx,
					conf);

	rcu_read_lock();
	rdev = rcu_dereference(conf->disks[dd_idx].rdev);
	if (rdev && test_bit(In_sync, &rdev->flags)) {
		atomic_inc(&rdev->nr_pending);
		rcu_read_unlock();
3449 3450 3451 3452 3453
		raid_bio->bi_next = (void*)rdev;
		align_bi->bi_bdev =  rdev->bdev;
		align_bi->bi_flags &= ~(1 << BIO_SEG_VALID);
		align_bi->bi_sector += rdev->data_offset;

3454 3455 3456 3457 3458 3459 3460
		if (!bio_fits_rdev(align_bi)) {
			/* too big in some way */
			bio_put(align_bi);
			rdev_dec_pending(rdev, mddev);
			return 0;
		}

3461 3462 3463 3464 3465 3466 3467
		spin_lock_irq(&conf->device_lock);
		wait_event_lock_irq(conf->wait_for_stripe,
				    conf->quiesce == 0,
				    conf->device_lock, /* nothing */);
		atomic_inc(&conf->active_aligned_reads);
		spin_unlock_irq(&conf->device_lock);

3468 3469 3470 3471
		generic_make_request(align_bi);
		return 1;
	} else {
		rcu_read_unlock();
3472
		bio_put(align_bi);
3473 3474 3475 3476 3477
		return 0;
	}
}


3478
static int make_request(struct request_queue *q, struct bio * bi)
Linus Torvalds's avatar
Linus Torvalds committed
3479 3480 3481 3482 3483 3484 3485
{
	mddev_t *mddev = q->queuedata;
	raid5_conf_t *conf = mddev_to_conf(mddev);
	unsigned int dd_idx, pd_idx;
	sector_t new_sector;
	sector_t logical_sector, last_sector;
	struct stripe_head *sh;
3486
	const int rw = bio_data_dir(bi);
3487
	int remaining;
Linus Torvalds's avatar
Linus Torvalds committed
3488

3489 3490 3491 3492 3493
	if (unlikely(bio_barrier(bi))) {
		bio_endio(bi, bi->bi_size, -EOPNOTSUPP);
		return 0;
	}

3494
	md_write_start(mddev, bi);
3495

3496 3497
	disk_stat_inc(mddev->gendisk, ios[rw]);
	disk_stat_add(mddev->gendisk, sectors[rw], bio_sectors(bi));
Linus Torvalds's avatar
Linus Torvalds committed
3498

3499
	if (rw == READ &&
3500 3501 3502 3503
	     mddev->reshape_position == MaxSector &&
	     chunk_aligned_read(q,bi))
            	return 0;

Linus Torvalds's avatar
Linus Torvalds committed
3504 3505 3506 3507
	logical_sector = bi->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
	last_sector = bi->bi_sector + (bi->bi_size>>9);
	bi->bi_next = NULL;
	bi->bi_phys_segments = 1;	/* over-loaded to count active stripes */
3508

Linus Torvalds's avatar
Linus Torvalds committed
3509 3510
	for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) {
		DEFINE_WAIT(w);
3511
		int disks, data_disks;
3512

3513
	retry:
3514
		prepare_to_wait(&conf->wait_for_overlap, &w, TASK_UNINTERRUPTIBLE);
3515 3516 3517
		if (likely(conf->expand_progress == MaxSector))
			disks = conf->raid_disks;
		else {
3518 3519 3520 3521 3522 3523 3524 3525
			/* spinlock is needed as expand_progress may be
			 * 64bit on a 32bit platform, and so it might be
			 * possible to see a half-updated value
			 * Ofcourse expand_progress could change after
			 * the lock is dropped, so once we get a reference
			 * to the stripe that we think it is, we will have
			 * to check again.
			 */
3526 3527 3528 3529
			spin_lock_irq(&conf->device_lock);
			disks = conf->raid_disks;
			if (logical_sector >= conf->expand_progress)
				disks = conf->previous_raid_disks;
3530 3531 3532 3533 3534 3535 3536
			else {
				if (logical_sector >= conf->expand_lo) {
					spin_unlock_irq(&conf->device_lock);
					schedule();
					goto retry;
				}
			}
3537 3538
			spin_unlock_irq(&conf->device_lock);
		}
3539 3540 3541
		data_disks = disks - conf->max_degraded;

 		new_sector = raid5_compute_sector(logical_sector, disks, data_disks,
3542
						  &dd_idx, &pd_idx, conf);
3543
		pr_debug("raid5: make_request, sector %llu logical %llu\n",
Linus Torvalds's avatar
Linus Torvalds committed
3544 3545 3546
			(unsigned long long)new_sector, 
			(unsigned long long)logical_sector);

3547
		sh = get_active_stripe(conf, new_sector, disks, pd_idx, (bi->bi_rw&RWA_MASK));
Linus Torvalds's avatar
Linus Torvalds committed
3548
		if (sh) {
3549 3550
			if (unlikely(conf->expand_progress != MaxSector)) {
				/* expansion might have moved on while waiting for a
3551 3552 3553 3554 3555 3556
				 * stripe, so we must do the range check again.
				 * Expansion could still move past after this
				 * test, but as we are holding a reference to
				 * 'sh', we know that if that happens,
				 *  STRIPE_EXPANDING will get set and the expansion
				 * won't proceed until we finish with the stripe.
3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569
				 */
				int must_retry = 0;
				spin_lock_irq(&conf->device_lock);
				if (logical_sector <  conf->expand_progress &&
				    disks == conf->previous_raid_disks)
					/* mismatch, need to try again */
					must_retry = 1;
				spin_unlock_irq(&conf->device_lock);
				if (must_retry) {
					release_stripe(sh);
					goto retry;
				}
			}
3570 3571 3572 3573 3574 3575 3576 3577 3578
			/* FIXME what if we get a false positive because these
			 * are being updated.
			 */
			if (logical_sector >= mddev->suspend_lo &&
			    logical_sector < mddev->suspend_hi) {
				release_stripe(sh);
				schedule();
				goto retry;
			}
3579 3580 3581 3582 3583

			if (test_bit(STRIPE_EXPANDING, &sh->state) ||
			    !add_stripe_bio(sh, bi, dd_idx, (bi->bi_rw&RW_MASK))) {
				/* Stripe is busy expanding or
				 * add failed due to overlap.  Flush everything
Linus Torvalds's avatar
Linus Torvalds committed
3584 3585 3586 3587 3588 3589 3590 3591
				 * and wait a while
				 */
				raid5_unplug_device(mddev->queue);
				release_stripe(sh);
				schedule();
				goto retry;
			}
			finish_wait(&conf->wait_for_overlap, &w);
3592
			handle_stripe(sh, NULL);
Linus Torvalds's avatar
Linus Torvalds committed
3593 3594 3595 3596 3597 3598 3599 3600 3601 3602
			release_stripe(sh);
		} else {
			/* cannot get stripe for read-ahead, just give-up */
			clear_bit(BIO_UPTODATE, &bi->bi_flags);
			finish_wait(&conf->wait_for_overlap, &w);
			break;
		}
			
	}
	spin_lock_irq(&conf->device_lock);
3603 3604 3605
	remaining = --bi->bi_phys_segments;
	spin_unlock_irq(&conf->device_lock);
	if (remaining == 0) {
Linus Torvalds's avatar
Linus Torvalds committed
3606 3607
		int bytes = bi->bi_size;

3608
		if ( rw == WRITE )
Linus Torvalds's avatar
Linus Torvalds committed
3609 3610
			md_write_end(mddev);
		bi->bi_size = 0;
3611 3612 3613
		bi->bi_end_io(bi, bytes,
			      test_bit(BIO_UPTODATE, &bi->bi_flags)
			        ? 0 : -EIO);
Linus Torvalds's avatar
Linus Torvalds committed
3614 3615 3616 3617
	}
	return 0;
}

3618
static sector_t reshape_request(mddev_t *mddev, sector_t sector_nr, int *skipped)
Linus Torvalds's avatar
Linus Torvalds committed
3619
{
3620 3621 3622 3623 3624 3625 3626 3627 3628
	/* reshaping is quite different to recovery/resync so it is
	 * handled quite separately ... here.
	 *
	 * On each call to sync_request, we gather one chunk worth of
	 * destination stripes and flag them as expanding.
	 * Then we find all the source stripes and request reads.
	 * As the reads complete, handle_stripe will copy the data
	 * into the destination stripe and release that stripe.
	 */
Linus Torvalds's avatar
Linus Torvalds committed
3629 3630
	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
	struct stripe_head *sh;
3631 3632
	int pd_idx;
	sector_t first_sector, last_sector;
3633 3634 3635
	int raid_disks = conf->previous_raid_disks;
	int data_disks = raid_disks - conf->max_degraded;
	int new_data_disks = conf->raid_disks - conf->max_degraded;
3636 3637 3638 3639 3640 3641 3642 3643
	int i;
	int dd_idx;
	sector_t writepos, safepos, gap;

	if (sector_nr == 0 &&
	    conf->expand_progress != 0) {
		/* restarting in the middle, skip the initial sectors */
		sector_nr = conf->expand_progress;
3644
		sector_div(sector_nr, new_data_disks);
3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657
		*skipped = 1;
		return sector_nr;
	}

	/* we update the metadata when there is more than 3Meg
	 * in the block range (that is rather arbitrary, should
	 * probably be time based) or when the data about to be
	 * copied would over-write the source of the data at
	 * the front of the range.
	 * i.e. one new_stripe forward from expand_progress new_maps
	 * to after where expand_lo old_maps to
	 */
	writepos = conf->expand_progress +
3658 3659
		conf->chunk_size/512*(new_data_disks);
	sector_div(writepos, new_data_disks);
3660
	safepos = conf->expand_lo;
3661
	sector_div(safepos, data_disks);
3662 3663 3664
	gap = conf->expand_progress - conf->expand_lo;

	if (writepos >= safepos ||
3665
	    gap > (new_data_disks)*3000*2 /*3Meg*/) {
3666 3667 3668 3669
		/* Cannot proceed until we've updated the superblock... */
		wait_event(conf->wait_for_overlap,
			   atomic_read(&conf->reshape_stripes)==0);
		mddev->reshape_position = conf->expand_progress;
3670
		set_bit(MD_CHANGE_DEVS, &mddev->flags);
3671
		md_wakeup_thread(mddev->thread);
3672
		wait_event(mddev->sb_wait, mddev->flags == 0 ||
3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694
			   kthread_should_stop());
		spin_lock_irq(&conf->device_lock);
		conf->expand_lo = mddev->reshape_position;
		spin_unlock_irq(&conf->device_lock);
		wake_up(&conf->wait_for_overlap);
	}

	for (i=0; i < conf->chunk_size/512; i+= STRIPE_SECTORS) {
		int j;
		int skipped = 0;
		pd_idx = stripe_to_pdidx(sector_nr+i, conf, conf->raid_disks);
		sh = get_active_stripe(conf, sector_nr+i,
				       conf->raid_disks, pd_idx, 0);
		set_bit(STRIPE_EXPANDING, &sh->state);
		atomic_inc(&conf->reshape_stripes);
		/* If any of this stripe is beyond the end of the old
		 * array, then we need to zero those blocks
		 */
		for (j=sh->disks; j--;) {
			sector_t s;
			if (j == sh->pd_idx)
				continue;
3695 3696 3697
			if (conf->level == 6 &&
			    j == raid6_next_disk(sh->pd_idx, sh->disks))
				continue;
3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713
			s = compute_blocknr(sh, j);
			if (s < (mddev->array_size<<1)) {
				skipped = 1;
				continue;
			}
			memset(page_address(sh->dev[j].page), 0, STRIPE_SIZE);
			set_bit(R5_Expanded, &sh->dev[j].flags);
			set_bit(R5_UPTODATE, &sh->dev[j].flags);
		}
		if (!skipped) {
			set_bit(STRIPE_EXPAND_READY, &sh->state);
			set_bit(STRIPE_HANDLE, &sh->state);
		}
		release_stripe(sh);
	}
	spin_lock_irq(&conf->device_lock);
3714
	conf->expand_progress = (sector_nr + i) * new_data_disks;
3715 3716 3717 3718 3719 3720 3721
	spin_unlock_irq(&conf->device_lock);
	/* Ok, those stripe are ready. We can start scheduling
	 * reads on the source stripes.
	 * The source stripes are determined by mapping the first and last
	 * block on the destination stripes.
	 */
	first_sector =
3722
		raid5_compute_sector(sector_nr*(new_data_disks),
3723 3724 3725 3726
				     raid_disks, data_disks,
				     &dd_idx, &pd_idx, conf);
	last_sector =
		raid5_compute_sector((sector_nr+conf->chunk_size/512)
3727
				     *(new_data_disks) -1,
3728 3729 3730 3731 3732
				     raid_disks, data_disks,
				     &dd_idx, &pd_idx, conf);
	if (last_sector >= (mddev->size<<1))
		last_sector = (mddev->size<<1)-1;
	while (first_sector <= last_sector) {
3733 3734
		pd_idx = stripe_to_pdidx(first_sector, conf,
					 conf->previous_raid_disks);
3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750
		sh = get_active_stripe(conf, first_sector,
				       conf->previous_raid_disks, pd_idx, 0);
		set_bit(STRIPE_EXPAND_SOURCE, &sh->state);
		set_bit(STRIPE_HANDLE, &sh->state);
		release_stripe(sh);
		first_sector += STRIPE_SECTORS;
	}
	return conf->chunk_size>>9;
}

/* FIXME go_faster isn't used */
static inline sector_t sync_request(mddev_t *mddev, sector_t sector_nr, int *skipped, int go_faster)
{
	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
	struct stripe_head *sh;
	int pd_idx;
Linus Torvalds's avatar
Linus Torvalds committed
3751
	int raid_disks = conf->raid_disks;
3752 3753
	sector_t max_sector = mddev->size << 1;
	int sync_blocks;
3754 3755
	int still_degraded = 0;
	int i;
Linus Torvalds's avatar
Linus Torvalds committed
3756

3757
	if (sector_nr >= max_sector) {
Linus Torvalds's avatar
Linus Torvalds committed
3758 3759
		/* just being told to finish up .. nothing much to do */
		unplug_slaves(mddev);
3760 3761 3762 3763
		if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)) {
			end_reshape(conf);
			return 0;
		}
3764 3765 3766 3767

		if (mddev->curr_resync < max_sector) /* aborted */
			bitmap_end_sync(mddev->bitmap, mddev->curr_resync,
					&sync_blocks, 1);
3768
		else /* completed sync */
3769 3770 3771
			conf->fullsync = 0;
		bitmap_close_sync(mddev->bitmap);

Linus Torvalds's avatar
Linus Torvalds committed
3772 3773
		return 0;
	}
3774

3775 3776
	if (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
		return reshape_request(mddev, sector_nr, skipped);
3777

3778
	/* if there is too many failed drives and we are trying
Linus Torvalds's avatar
Linus Torvalds committed
3779 3780 3781
	 * to resync, then assert that we are finished, because there is
	 * nothing we can do.
	 */
3782
	if (mddev->degraded >= conf->max_degraded &&
3783
	    test_bit(MD_RECOVERY_SYNC, &mddev->recovery)) {
3784 3785
		sector_t rv = (mddev->size << 1) - sector_nr;
		*skipped = 1;
Linus Torvalds's avatar
Linus Torvalds committed
3786 3787
		return rv;
	}
3788
	if (!bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, 1) &&
3789
	    !test_bit(MD_RECOVERY_REQUESTED, &mddev->recovery) &&
3790 3791 3792 3793 3794 3795
	    !conf->fullsync && sync_blocks >= STRIPE_SECTORS) {
		/* we can skip this block, and probably more */
		sync_blocks /= STRIPE_SECTORS;
		*skipped = 1;
		return sync_blocks * STRIPE_SECTORS; /* keep things rounded to whole stripes */
	}
Linus Torvalds's avatar
Linus Torvalds committed
3796

3797
	pd_idx = stripe_to_pdidx(sector_nr, conf, raid_disks);
3798
	sh = get_active_stripe(conf, sector_nr, raid_disks, pd_idx, 1);
Linus Torvalds's avatar
Linus Torvalds committed
3799
	if (sh == NULL) {
3800
		sh = get_active_stripe(conf, sector_nr, raid_disks, pd_idx, 0);
Linus Torvalds's avatar
Linus Torvalds committed
3801
		/* make sure we don't swamp the stripe cache if someone else
3802
		 * is trying to get access
Linus Torvalds's avatar
Linus Torvalds committed
3803
		 */
3804
		schedule_timeout_uninterruptible(1);
Linus Torvalds's avatar
Linus Torvalds committed
3805
	}
3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816
	/* Need to check if array will still be degraded after recovery/resync
	 * We don't need to check the 'failed' flag as when that gets set,
	 * recovery aborts.
	 */
	for (i=0; i<mddev->raid_disks; i++)
		if (conf->disks[i].rdev == NULL)
			still_degraded = 1;

	bitmap_start_sync(mddev->bitmap, sector_nr, &sync_blocks, still_degraded);

	spin_lock(&sh->lock);
Linus Torvalds's avatar
Linus Torvalds committed
3817 3818 3819 3820
	set_bit(STRIPE_SYNCING, &sh->state);
	clear_bit(STRIPE_INSYNC, &sh->state);
	spin_unlock(&sh->lock);

3821
	handle_stripe(sh, NULL);
Linus Torvalds's avatar
Linus Torvalds committed
3822 3823 3824 3825 3826
	release_stripe(sh);

	return STRIPE_SECTORS;
}

3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855
static int  retry_aligned_read(raid5_conf_t *conf, struct bio *raid_bio)
{
	/* We may not be able to submit a whole bio at once as there
	 * may not be enough stripe_heads available.
	 * We cannot pre-allocate enough stripe_heads as we may need
	 * more than exist in the cache (if we allow ever large chunks).
	 * So we do one stripe head at a time and record in
	 * ->bi_hw_segments how many have been done.
	 *
	 * We *know* that this entire raid_bio is in one chunk, so
	 * it will be only one 'dd_idx' and only need one call to raid5_compute_sector.
	 */
	struct stripe_head *sh;
	int dd_idx, pd_idx;
	sector_t sector, logical_sector, last_sector;
	int scnt = 0;
	int remaining;
	int handled = 0;

	logical_sector = raid_bio->bi_sector & ~((sector_t)STRIPE_SECTORS-1);
	sector = raid5_compute_sector(	logical_sector,
					conf->raid_disks,
					conf->raid_disks - conf->max_degraded,
					&dd_idx,
					&pd_idx,
					conf);
	last_sector = raid_bio->bi_sector + (raid_bio->bi_size>>9);

	for (; logical_sector < last_sector;
3856 3857 3858
	     logical_sector += STRIPE_SECTORS,
		     sector += STRIPE_SECTORS,
		     scnt++) {
3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873

		if (scnt < raid_bio->bi_hw_segments)
			/* already done this stripe */
			continue;

		sh = get_active_stripe(conf, sector, conf->raid_disks, pd_idx, 1);

		if (!sh) {
			/* failed to get a stripe - must wait */
			raid_bio->bi_hw_segments = scnt;
			conf->retry_read_aligned = raid_bio;
			return handled;
		}

		set_bit(R5_ReadError, &sh->dev[dd_idx].flags);
3874 3875 3876 3877 3878 3879 3880
		if (!add_stripe_bio(sh, raid_bio, dd_idx, 0)) {
			release_stripe(sh);
			raid_bio->bi_hw_segments = scnt;
			conf->retry_read_aligned = raid_bio;
			return handled;
		}

3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891
		handle_stripe(sh, NULL);
		release_stripe(sh);
		handled++;
	}
	spin_lock_irq(&conf->device_lock);
	remaining = --raid_bio->bi_phys_segments;
	spin_unlock_irq(&conf->device_lock);
	if (remaining == 0) {
		int bytes = raid_bio->bi_size;

		raid_bio->bi_size = 0;
3892 3893 3894
		raid_bio->bi_end_io(raid_bio, bytes,
			      test_bit(BIO_UPTODATE, &raid_bio->bi_flags)
			        ? 0 : -EIO);
3895 3896 3897 3898 3899 3900 3901 3902
	}
	if (atomic_dec_and_test(&conf->active_aligned_reads))
		wake_up(&conf->wait_for_stripe);
	return handled;
}



Linus Torvalds's avatar
Linus Torvalds committed
3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915
/*
 * This is our raid5 kernel thread.
 *
 * We scan the hash table for stripes which can be handled now.
 * During the scan, completed stripes are saved for us by the interrupt
 * handler, so that they will not have to wait for our next wakeup.
 */
static void raid5d (mddev_t *mddev)
{
	struct stripe_head *sh;
	raid5_conf_t *conf = mddev_to_conf(mddev);
	int handled;

3916
	pr_debug("+++ raid5d active\n");
Linus Torvalds's avatar
Linus Torvalds committed
3917 3918 3919 3920 3921 3922 3923

	md_check_recovery(mddev);

	handled = 0;
	spin_lock_irq(&conf->device_lock);
	while (1) {
		struct list_head *first;
3924
		struct bio *bio;
Linus Torvalds's avatar
Linus Torvalds committed
3925

3926
		if (conf->seq_flush != conf->seq_write) {
3927
			int seq = conf->seq_flush;
3928
			spin_unlock_irq(&conf->device_lock);
3929
			bitmap_unplug(mddev->bitmap);
3930
			spin_lock_irq(&conf->device_lock);
3931 3932 3933 3934
			conf->seq_write = seq;
			activate_bit_delay(conf);
		}

Linus Torvalds's avatar
Linus Torvalds committed
3935 3936 3937 3938 3939 3940
		if (list_empty(&conf->handle_list) &&
		    atomic_read(&conf->preread_active_stripes) < IO_THRESHOLD &&
		    !blk_queue_plugged(mddev->queue) &&
		    !list_empty(&conf->delayed_list))
			raid5_activate_delayed(conf);

3941 3942 3943 3944 3945 3946 3947 3948 3949 3950
		while ((bio = remove_bio_from_retry(conf))) {
			int ok;
			spin_unlock_irq(&conf->device_lock);
			ok = retry_aligned_read(conf, bio);
			spin_lock_irq(&conf->device_lock);
			if (!ok)
				break;
			handled++;
		}

3951 3952
		if (list_empty(&conf->handle_list)) {
			async_tx_issue_pending_all();
Linus Torvalds's avatar
Linus Torvalds committed
3953
			break;
3954
		}
Linus Torvalds's avatar
Linus Torvalds committed
3955 3956 3957 3958 3959 3960

		first = conf->handle_list.next;
		sh = list_entry(first, struct stripe_head, lru);

		list_del_init(first);
		atomic_inc(&sh->count);
3961
		BUG_ON(atomic_read(&sh->count)!= 1);
Linus Torvalds's avatar
Linus Torvalds committed
3962 3963 3964
		spin_unlock_irq(&conf->device_lock);
		
		handled++;
3965
		handle_stripe(sh, conf->spare_page);
Linus Torvalds's avatar
Linus Torvalds committed
3966 3967 3968 3969
		release_stripe(sh);

		spin_lock_irq(&conf->device_lock);
	}
3970
	pr_debug("%d stripes handled\n", handled);
Linus Torvalds's avatar
Linus Torvalds committed
3971 3972 3973 3974 3975

	spin_unlock_irq(&conf->device_lock);

	unplug_slaves(mddev);

3976
	pr_debug("--- raid5d inactive\n");
Linus Torvalds's avatar
Linus Torvalds committed
3977 3978
}

3979
static ssize_t
3980
raid5_show_stripe_cache_size(mddev_t *mddev, char *page)
3981
{
3982
	raid5_conf_t *conf = mddev_to_conf(mddev);
3983 3984 3985 3986
	if (conf)
		return sprintf(page, "%d\n", conf->max_nr_stripes);
	else
		return 0;
3987 3988 3989
}

static ssize_t
3990
raid5_store_stripe_cache_size(mddev_t *mddev, const char *page, size_t len)
3991
{
3992
	raid5_conf_t *conf = mddev_to_conf(mddev);
3993 3994 3995 3996
	char *end;
	int new;
	if (len >= PAGE_SIZE)
		return -EINVAL;
3997 3998
	if (!conf)
		return -ENODEV;
3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010

	new = simple_strtoul(page, &end, 10);
	if (!*page || (*end && *end != '\n') )
		return -EINVAL;
	if (new <= 16 || new > 32768)
		return -EINVAL;
	while (new < conf->max_nr_stripes) {
		if (drop_one_stripe(conf))
			conf->max_nr_stripes--;
		else
			break;
	}
4011
	md_allow_write(mddev);
4012 4013 4014 4015 4016 4017 4018
	while (new > conf->max_nr_stripes) {
		if (grow_one_stripe(conf))
			conf->max_nr_stripes++;
		else break;
	}
	return len;
}
4019

4020 4021 4022 4023
static struct md_sysfs_entry
raid5_stripecache_size = __ATTR(stripe_cache_size, S_IRUGO | S_IWUSR,
				raid5_show_stripe_cache_size,
				raid5_store_stripe_cache_size);
4024 4025

static ssize_t
4026
stripe_cache_active_show(mddev_t *mddev, char *page)
4027
{
4028
	raid5_conf_t *conf = mddev_to_conf(mddev);
4029 4030 4031 4032
	if (conf)
		return sprintf(page, "%d\n", atomic_read(&conf->active_stripes));
	else
		return 0;
4033 4034
}

4035 4036
static struct md_sysfs_entry
raid5_stripecache_active = __ATTR_RO(stripe_cache_active);
4037

4038
static struct attribute *raid5_attrs[] =  {
4039 4040 4041 4042
	&raid5_stripecache_size.attr,
	&raid5_stripecache_active.attr,
	NULL,
};
4043 4044 4045
static struct attribute_group raid5_attrs_group = {
	.name = NULL,
	.attrs = raid5_attrs,
4046 4047
};

4048
static int run(mddev_t *mddev)
Linus Torvalds's avatar
Linus Torvalds committed
4049 4050 4051 4052 4053 4054
{
	raid5_conf_t *conf;
	int raid_disk, memory;
	mdk_rdev_t *rdev;
	struct disk_info *disk;
	struct list_head *tmp;
4055
	int working_disks = 0;
Linus Torvalds's avatar
Linus Torvalds committed
4056

4057 4058
	if (mddev->level != 5 && mddev->level != 4 && mddev->level != 6) {
		printk(KERN_ERR "raid5: %s: raid level not set to 4/5/6 (%d)\n",
4059
		       mdname(mddev), mddev->level);
Linus Torvalds's avatar
Linus Torvalds committed
4060 4061 4062
		return -EIO;
	}

4063 4064 4065 4066 4067 4068 4069 4070
	if (mddev->reshape_position != MaxSector) {
		/* Check that we can continue the reshape.
		 * Currently only disks can change, it must
		 * increase, and we must be past the point where
		 * a stripe over-writes itself
		 */
		sector_t here_new, here_old;
		int old_disks;
4071
		int max_degraded = (mddev->level == 5 ? 1 : 2);
4072 4073 4074 4075

		if (mddev->new_level != mddev->level ||
		    mddev->new_layout != mddev->layout ||
		    mddev->new_chunk != mddev->chunk_size) {
4076 4077
			printk(KERN_ERR "raid5: %s: unsupported reshape "
			       "required - aborting.\n",
4078 4079 4080 4081
			       mdname(mddev));
			return -EINVAL;
		}
		if (mddev->delta_disks <= 0) {
4082 4083
			printk(KERN_ERR "raid5: %s: unsupported reshape "
			       "(reduce disks) required - aborting.\n",
4084 4085 4086 4087 4088
			       mdname(mddev));
			return -EINVAL;
		}
		old_disks = mddev->raid_disks - mddev->delta_disks;
		/* reshape_position must be on a new-stripe boundary, and one
4089 4090
		 * further up in new geometry must map after here in old
		 * geometry.
4091 4092
		 */
		here_new = mddev->reshape_position;
4093 4094 4095 4096
		if (sector_div(here_new, (mddev->chunk_size>>9)*
			       (mddev->raid_disks - max_degraded))) {
			printk(KERN_ERR "raid5: reshape_position not "
			       "on a stripe boundary\n");
4097 4098 4099 4100
			return -EINVAL;
		}
		/* here_new is the stripe we will write to */
		here_old = mddev->reshape_position;
4101 4102 4103 4104
		sector_div(here_old, (mddev->chunk_size>>9)*
			   (old_disks-max_degraded));
		/* here_old is the first stripe that we might need to read
		 * from */
4105 4106
		if (here_new >= here_old) {
			/* Reading from the same stripe as writing to - bad */
4107 4108
			printk(KERN_ERR "raid5: reshape_position too early for "
			       "auto-recovery - aborting.\n");
4109 4110 4111 4112 4113 4114 4115
			return -EINVAL;
		}
		printk(KERN_INFO "raid5: reshape will continue\n");
		/* OK, we should be able to continue; */
	}


4116
	mddev->private = kzalloc(sizeof (raid5_conf_t), GFP_KERNEL);
Linus Torvalds's avatar
Linus Torvalds committed
4117 4118
	if ((conf = mddev->private) == NULL)
		goto abort;
4119 4120 4121 4122 4123 4124 4125 4126
	if (mddev->reshape_position == MaxSector) {
		conf->previous_raid_disks = conf->raid_disks = mddev->raid_disks;
	} else {
		conf->raid_disks = mddev->raid_disks;
		conf->previous_raid_disks = mddev->raid_disks - mddev->delta_disks;
	}

	conf->disks = kzalloc(conf->raid_disks * sizeof(struct disk_info),
4127 4128 4129
			      GFP_KERNEL);
	if (!conf->disks)
		goto abort;
4130

Linus Torvalds's avatar
Linus Torvalds committed
4131 4132
	conf->mddev = mddev;

4133
	if ((conf->stripe_hashtbl = kzalloc(PAGE_SIZE, GFP_KERNEL)) == NULL)
Linus Torvalds's avatar
Linus Torvalds committed
4134 4135
		goto abort;

4136 4137 4138 4139 4140
	if (mddev->level == 6) {
		conf->spare_page = alloc_page(GFP_KERNEL);
		if (!conf->spare_page)
			goto abort;
	}
Linus Torvalds's avatar
Linus Torvalds committed
4141 4142 4143 4144 4145
	spin_lock_init(&conf->device_lock);
	init_waitqueue_head(&conf->wait_for_stripe);
	init_waitqueue_head(&conf->wait_for_overlap);
	INIT_LIST_HEAD(&conf->handle_list);
	INIT_LIST_HEAD(&conf->delayed_list);
4146
	INIT_LIST_HEAD(&conf->bitmap_list);
Linus Torvalds's avatar
Linus Torvalds committed
4147 4148 4149
	INIT_LIST_HEAD(&conf->inactive_list);
	atomic_set(&conf->active_stripes, 0);
	atomic_set(&conf->preread_active_stripes, 0);
4150
	atomic_set(&conf->active_aligned_reads, 0);
Linus Torvalds's avatar
Linus Torvalds committed
4151

4152
	pr_debug("raid5: run(%s) called.\n", mdname(mddev));
Linus Torvalds's avatar
Linus Torvalds committed
4153 4154 4155

	ITERATE_RDEV(mddev,rdev,tmp) {
		raid_disk = rdev->raid_disk;
4156
		if (raid_disk >= conf->raid_disks
Linus Torvalds's avatar
Linus Torvalds committed
4157 4158 4159 4160 4161 4162
		    || raid_disk < 0)
			continue;
		disk = conf->disks + raid_disk;

		disk->rdev = rdev;

4163
		if (test_bit(In_sync, &rdev->flags)) {
Linus Torvalds's avatar
Linus Torvalds committed
4164 4165 4166 4167
			char b[BDEVNAME_SIZE];
			printk(KERN_INFO "raid5: device %s operational as raid"
				" disk %d\n", bdevname(rdev->bdev,b),
				raid_disk);
4168
			working_disks++;
Linus Torvalds's avatar
Linus Torvalds committed
4169 4170 4171 4172
		}
	}

	/*
4173
	 * 0 for a fully functional array, 1 or 2 for a degraded array.
Linus Torvalds's avatar
Linus Torvalds committed
4174
	 */
4175
	mddev->degraded = conf->raid_disks - working_disks;
Linus Torvalds's avatar
Linus Torvalds committed
4176 4177 4178
	conf->mddev = mddev;
	conf->chunk_size = mddev->chunk_size;
	conf->level = mddev->level;
4179 4180 4181 4182
	if (conf->level == 6)
		conf->max_degraded = 2;
	else
		conf->max_degraded = 1;
Linus Torvalds's avatar
Linus Torvalds committed
4183 4184
	conf->algorithm = mddev->layout;
	conf->max_nr_stripes = NR_STRIPES;
4185
	conf->expand_progress = mddev->reshape_position;
Linus Torvalds's avatar
Linus Torvalds committed
4186 4187 4188

	/* device size must be a multiple of chunk size */
	mddev->size &= ~(mddev->chunk_size/1024 -1);
4189
	mddev->resync_max_sectors = mddev->size << 1;
Linus Torvalds's avatar
Linus Torvalds committed
4190

4191 4192 4193 4194 4195
	if (conf->level == 6 && conf->raid_disks < 4) {
		printk(KERN_ERR "raid6: not enough configured devices for %s (%d, minimum 4)\n",
		       mdname(mddev), conf->raid_disks);
		goto abort;
	}
Linus Torvalds's avatar
Linus Torvalds committed
4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206
	if (!conf->chunk_size || conf->chunk_size % 4) {
		printk(KERN_ERR "raid5: invalid chunk size %d for %s\n",
			conf->chunk_size, mdname(mddev));
		goto abort;
	}
	if (conf->algorithm > ALGORITHM_RIGHT_SYMMETRIC) {
		printk(KERN_ERR 
			"raid5: unsupported parity algorithm %d for %s\n",
			conf->algorithm, mdname(mddev));
		goto abort;
	}
4207
	if (mddev->degraded > conf->max_degraded) {
Linus Torvalds's avatar
Linus Torvalds committed
4208 4209
		printk(KERN_ERR "raid5: not enough operational devices for %s"
			" (%d/%d failed)\n",
4210
			mdname(mddev), mddev->degraded, conf->raid_disks);
Linus Torvalds's avatar
Linus Torvalds committed
4211 4212 4213
		goto abort;
	}

4214
	if (mddev->degraded > 0 &&
Linus Torvalds's avatar
Linus Torvalds committed
4215
	    mddev->recovery_cp != MaxSector) {
4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226
		if (mddev->ok_start_degraded)
			printk(KERN_WARNING
			       "raid5: starting dirty degraded array: %s"
			       "- data corruption possible.\n",
			       mdname(mddev));
		else {
			printk(KERN_ERR
			       "raid5: cannot start dirty degraded array for %s\n",
			       mdname(mddev));
			goto abort;
		}
Linus Torvalds's avatar
Linus Torvalds committed
4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237
	}

	{
		mddev->thread = md_register_thread(raid5d, mddev, "%s_raid5");
		if (!mddev->thread) {
			printk(KERN_ERR 
				"raid5: couldn't allocate thread for %s\n",
				mdname(mddev));
			goto abort;
		}
	}
4238
	memory = conf->max_nr_stripes * (sizeof(struct stripe_head) +
Linus Torvalds's avatar
Linus Torvalds committed
4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262
		 conf->raid_disks * ((sizeof(struct bio) + PAGE_SIZE))) / 1024;
	if (grow_stripes(conf, conf->max_nr_stripes)) {
		printk(KERN_ERR 
			"raid5: couldn't allocate %dkB for buffers\n", memory);
		shrink_stripes(conf);
		md_unregister_thread(mddev->thread);
		goto abort;
	} else
		printk(KERN_INFO "raid5: allocated %dkB for %s\n",
			memory, mdname(mddev));

	if (mddev->degraded == 0)
		printk("raid5: raid level %d set %s active with %d out of %d"
			" devices, algorithm %d\n", conf->level, mdname(mddev), 
			mddev->raid_disks-mddev->degraded, mddev->raid_disks,
			conf->algorithm);
	else
		printk(KERN_ALERT "raid5: raid level %d set %s active with %d"
			" out of %d devices, algorithm %d\n", conf->level,
			mdname(mddev), mddev->raid_disks - mddev->degraded,
			mddev->raid_disks, conf->algorithm);

	print_raid5_conf(conf);

4263 4264
	if (conf->expand_progress != MaxSector) {
		printk("...ok start reshape thread\n");
4265
		conf->expand_lo = conf->expand_progress;
4266 4267 4268 4269 4270 4271 4272 4273 4274
		atomic_set(&conf->reshape_stripes, 0);
		clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
		clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
		set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
		set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
		mddev->sync_thread = md_register_thread(md_do_sync, mddev,
							"%s_reshape");
	}

Linus Torvalds's avatar
Linus Torvalds committed
4275
	/* read-ahead size must cover two whole stripes, which is
4276
	 * 2 * (datadisks) * chunksize where 'n' is the number of raid devices
Linus Torvalds's avatar
Linus Torvalds committed
4277 4278
	 */
	{
4279 4280
		int data_disks = conf->previous_raid_disks - conf->max_degraded;
		int stripe = data_disks *
4281
			(mddev->chunk_size / PAGE_SIZE);
Linus Torvalds's avatar
Linus Torvalds committed
4282 4283 4284 4285 4286
		if (mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
			mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
	}

	/* Ok, everything is just fine now */
4287 4288 4289 4290
	if (sysfs_create_group(&mddev->kobj, &raid5_attrs_group))
		printk(KERN_WARNING
		       "raid5: failed to create sysfs attributes for %s\n",
		       mdname(mddev));
4291 4292 4293

	mddev->queue->unplug_fn = raid5_unplug_device;
	mddev->queue->issue_flush_fn = raid5_issue_flush;
4294
	mddev->queue->backing_dev_info.congested_data = mddev;
4295
	mddev->queue->backing_dev_info.congested_fn = raid5_congested;
4296

4297 4298
	mddev->array_size =  mddev->size * (conf->previous_raid_disks -
					    conf->max_degraded);
4299

4300 4301
	blk_queue_merge_bvec(mddev->queue, raid5_mergeable_bvec);

Linus Torvalds's avatar
Linus Torvalds committed
4302 4303 4304 4305
	return 0;
abort:
	if (conf) {
		print_raid5_conf(conf);
4306
		safe_put_page(conf->spare_page);
4307
		kfree(conf->disks);
4308
		kfree(conf->stripe_hashtbl);
Linus Torvalds's avatar
Linus Torvalds committed
4309 4310 4311 4312 4313 4314 4315 4316 4317
		kfree(conf);
	}
	mddev->private = NULL;
	printk(KERN_ALERT "raid5: failed to run raid set %s\n", mdname(mddev));
	return -EIO;
}



4318
static int stop(mddev_t *mddev)
Linus Torvalds's avatar
Linus Torvalds committed
4319 4320 4321 4322 4323 4324
{
	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;

	md_unregister_thread(mddev->thread);
	mddev->thread = NULL;
	shrink_stripes(conf);
4325
	kfree(conf->stripe_hashtbl);
4326
	mddev->queue->backing_dev_info.congested_fn = NULL;
Linus Torvalds's avatar
Linus Torvalds committed
4327
	blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/
4328
	sysfs_remove_group(&mddev->kobj, &raid5_attrs_group);
4329
	kfree(conf->disks);
4330
	kfree(conf);
Linus Torvalds's avatar
Linus Torvalds committed
4331 4332 4333 4334
	mddev->private = NULL;
	return 0;
}

4335
#ifdef DEBUG
4336
static void print_sh (struct seq_file *seq, struct stripe_head *sh)
Linus Torvalds's avatar
Linus Torvalds committed
4337 4338 4339
{
	int i;

4340 4341 4342 4343 4344
	seq_printf(seq, "sh %llu, pd_idx %d, state %ld.\n",
		   (unsigned long long)sh->sector, sh->pd_idx, sh->state);
	seq_printf(seq, "sh %llu,  count %d.\n",
		   (unsigned long long)sh->sector, atomic_read(&sh->count));
	seq_printf(seq, "sh %llu, ", (unsigned long long)sh->sector);
4345
	for (i = 0; i < sh->disks; i++) {
4346 4347
		seq_printf(seq, "(cache%d: %p %ld) ",
			   i, sh->dev[i].page, sh->dev[i].flags);
Linus Torvalds's avatar
Linus Torvalds committed
4348
	}
4349
	seq_printf(seq, "\n");
Linus Torvalds's avatar
Linus Torvalds committed
4350 4351
}

4352
static void printall (struct seq_file *seq, raid5_conf_t *conf)
Linus Torvalds's avatar
Linus Torvalds committed
4353 4354
{
	struct stripe_head *sh;
4355
	struct hlist_node *hn;
Linus Torvalds's avatar
Linus Torvalds committed
4356 4357 4358 4359
	int i;

	spin_lock_irq(&conf->device_lock);
	for (i = 0; i < NR_HASH; i++) {
4360
		hlist_for_each_entry(sh, hn, &conf->stripe_hashtbl[i], hash) {
Linus Torvalds's avatar
Linus Torvalds committed
4361 4362
			if (sh->raid_conf != conf)
				continue;
4363
			print_sh(seq, sh);
Linus Torvalds's avatar
Linus Torvalds committed
4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375
		}
	}
	spin_unlock_irq(&conf->device_lock);
}
#endif

static void status (struct seq_file *seq, mddev_t *mddev)
{
	raid5_conf_t *conf = (raid5_conf_t *) mddev->private;
	int i;

	seq_printf (seq, " level %d, %dk chunk, algorithm %d", mddev->level, mddev->chunk_size >> 10, mddev->layout);
4376
	seq_printf (seq, " [%d/%d] [", conf->raid_disks, conf->raid_disks - mddev->degraded);
Linus Torvalds's avatar
Linus Torvalds committed
4377 4378 4379
	for (i = 0; i < conf->raid_disks; i++)
		seq_printf (seq, "%s",
			       conf->disks[i].rdev &&
4380
			       test_bit(In_sync, &conf->disks[i].rdev->flags) ? "U" : "_");
Linus Torvalds's avatar
Linus Torvalds committed
4381
	seq_printf (seq, "]");
4382
#ifdef DEBUG
4383 4384
	seq_printf (seq, "\n");
	printall(seq, conf);
Linus Torvalds's avatar
Linus Torvalds committed
4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397
#endif
}

static void print_raid5_conf (raid5_conf_t *conf)
{
	int i;
	struct disk_info *tmp;

	printk("RAID5 conf printout:\n");
	if (!conf) {
		printk("(conf==NULL)\n");
		return;
	}
4398 4399
	printk(" --- rd:%d wd:%d\n", conf->raid_disks,
		 conf->raid_disks - conf->mddev->degraded);
Linus Torvalds's avatar
Linus Torvalds committed
4400 4401 4402 4403 4404 4405

	for (i = 0; i < conf->raid_disks; i++) {
		char b[BDEVNAME_SIZE];
		tmp = conf->disks + i;
		if (tmp->rdev)
		printk(" disk %d, o:%d, dev:%s\n",
4406
			i, !test_bit(Faulty, &tmp->rdev->flags),
Linus Torvalds's avatar
Linus Torvalds committed
4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419
			bdevname(tmp->rdev->bdev,b));
	}
}

static int raid5_spare_active(mddev_t *mddev)
{
	int i;
	raid5_conf_t *conf = mddev->private;
	struct disk_info *tmp;

	for (i = 0; i < conf->raid_disks; i++) {
		tmp = conf->disks + i;
		if (tmp->rdev
4420
		    && !test_bit(Faulty, &tmp->rdev->flags)
4421 4422 4423
		    && !test_and_set_bit(In_sync, &tmp->rdev->flags)) {
			unsigned long flags;
			spin_lock_irqsave(&conf->device_lock, flags);
Linus Torvalds's avatar
Linus Torvalds committed
4424
			mddev->degraded--;
4425
			spin_unlock_irqrestore(&conf->device_lock, flags);
Linus Torvalds's avatar
Linus Torvalds committed
4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441
		}
	}
	print_raid5_conf(conf);
	return 0;
}

static int raid5_remove_disk(mddev_t *mddev, int number)
{
	raid5_conf_t *conf = mddev->private;
	int err = 0;
	mdk_rdev_t *rdev;
	struct disk_info *p = conf->disks + number;

	print_raid5_conf(conf);
	rdev = p->rdev;
	if (rdev) {
4442
		if (test_bit(In_sync, &rdev->flags) ||
Linus Torvalds's avatar
Linus Torvalds committed
4443 4444 4445 4446 4447
		    atomic_read(&rdev->nr_pending)) {
			err = -EBUSY;
			goto abort;
		}
		p->rdev = NULL;
4448
		synchronize_rcu();
Linus Torvalds's avatar
Linus Torvalds committed
4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467
		if (atomic_read(&rdev->nr_pending)) {
			/* lost the race, try later */
			err = -EBUSY;
			p->rdev = rdev;
		}
	}
abort:

	print_raid5_conf(conf);
	return err;
}

static int raid5_add_disk(mddev_t *mddev, mdk_rdev_t *rdev)
{
	raid5_conf_t *conf = mddev->private;
	int found = 0;
	int disk;
	struct disk_info *p;

4468
	if (mddev->degraded > conf->max_degraded)
Linus Torvalds's avatar
Linus Torvalds committed
4469 4470 4471 4472
		/* no point adding a device */
		return 0;

	/*
4473 4474
	 * find the disk ... but prefer rdev->saved_raid_disk
	 * if possible.
Linus Torvalds's avatar
Linus Torvalds committed
4475
	 */
4476 4477 4478 4479 4480 4481
	if (rdev->saved_raid_disk >= 0 &&
	    conf->disks[rdev->saved_raid_disk].rdev == NULL)
		disk = rdev->saved_raid_disk;
	else
		disk = 0;
	for ( ; disk < conf->raid_disks; disk++)
Linus Torvalds's avatar
Linus Torvalds committed
4482
		if ((p=conf->disks + disk)->rdev == NULL) {
4483
			clear_bit(In_sync, &rdev->flags);
Linus Torvalds's avatar
Linus Torvalds committed
4484 4485
			rdev->raid_disk = disk;
			found = 1;
4486 4487
			if (rdev->saved_raid_disk != disk)
				conf->fullsync = 1;
4488
			rcu_assign_pointer(p->rdev, rdev);
Linus Torvalds's avatar
Linus Torvalds committed
4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503
			break;
		}
	print_raid5_conf(conf);
	return found;
}

static int raid5_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.
	 */
4504 4505
	raid5_conf_t *conf = mddev_to_conf(mddev);

Linus Torvalds's avatar
Linus Torvalds committed
4506
	sectors &= ~((sector_t)mddev->chunk_size/512 - 1);
4507
	mddev->array_size = (sectors * (mddev->raid_disks-conf->max_degraded))>>1;
Linus Torvalds's avatar
Linus Torvalds committed
4508
	set_capacity(mddev->gendisk, mddev->array_size << 1);
4509
	mddev->changed = 1;
Linus Torvalds's avatar
Linus Torvalds committed
4510 4511 4512 4513 4514
	if (sectors/2  > mddev->size && mddev->recovery_cp == MaxSector) {
		mddev->recovery_cp = mddev->size << 1;
		set_bit(MD_RECOVERY_NEEDED, &mddev->recovery);
	}
	mddev->size = sectors /2;
4515
	mddev->resync_max_sectors = sectors;
Linus Torvalds's avatar
Linus Torvalds committed
4516 4517 4518
	return 0;
}

4519
#ifdef CONFIG_MD_RAID5_RESHAPE
4520
static int raid5_check_reshape(mddev_t *mddev)
4521 4522 4523 4524
{
	raid5_conf_t *conf = mddev_to_conf(mddev);
	int err;

4525 4526 4527 4528
	if (mddev->delta_disks < 0 ||
	    mddev->new_level != mddev->level)
		return -EINVAL; /* Cannot shrink array or change level yet */
	if (mddev->delta_disks == 0)
4529 4530 4531 4532 4533 4534 4535 4536 4537 4538
		return 0; /* nothing to do */

	/* Can only proceed if there are plenty of stripe_heads.
	 * We need a minimum of one full stripe,, and for sensible progress
	 * it is best to have about 4 times that.
	 * If we require 4 times, then the default 256 4K stripe_heads will
	 * allow for chunk sizes up to 256K, which is probably OK.
	 * If the chunk size is greater, user-space should request more
	 * stripe_heads first.
	 */
4539 4540
	if ((mddev->chunk_size / STRIPE_SIZE) * 4 > conf->max_nr_stripes ||
	    (mddev->new_chunk / STRIPE_SIZE) * 4 > conf->max_nr_stripes) {
4541 4542 4543 4544 4545
		printk(KERN_WARNING "raid5: reshape: not enough stripes.  Needed %lu\n",
		       (mddev->chunk_size / STRIPE_SIZE)*4);
		return -ENOSPC;
	}

4546 4547 4548 4549
	err = resize_stripes(conf, conf->raid_disks + mddev->delta_disks);
	if (err)
		return err;

4550 4551
	if (mddev->degraded > conf->max_degraded)
		return -EINVAL;
4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562
	/* looks like we might be able to manage this */
	return 0;
}

static int raid5_start_reshape(mddev_t *mddev)
{
	raid5_conf_t *conf = mddev_to_conf(mddev);
	mdk_rdev_t *rdev;
	struct list_head *rtmp;
	int spares = 0;
	int added_devices = 0;
4563
	unsigned long flags;
4564

4565
	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery))
4566 4567
		return -EBUSY;

4568 4569 4570 4571
	ITERATE_RDEV(mddev, rdev, rtmp)
		if (rdev->raid_disk < 0 &&
		    !test_bit(Faulty, &rdev->flags))
			spares++;
4572

4573
	if (spares - mddev->degraded < mddev->delta_disks - conf->max_degraded)
4574 4575 4576 4577 4578
		/* Not enough devices even to make a degraded array
		 * of that size
		 */
		return -EINVAL;

4579
	atomic_set(&conf->reshape_stripes, 0);
4580 4581
	spin_lock_irq(&conf->device_lock);
	conf->previous_raid_disks = conf->raid_disks;
4582
	conf->raid_disks += mddev->delta_disks;
4583
	conf->expand_progress = 0;
4584
	conf->expand_lo = 0;
4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596
	spin_unlock_irq(&conf->device_lock);

	/* Add some new drives, as many as will fit.
	 * We know there are enough to make the newly sized array work.
	 */
	ITERATE_RDEV(mddev, rdev, rtmp)
		if (rdev->raid_disk < 0 &&
		    !test_bit(Faulty, &rdev->flags)) {
			if (raid5_add_disk(mddev, rdev)) {
				char nm[20];
				set_bit(In_sync, &rdev->flags);
				added_devices++;
4597
				rdev->recovery_offset = 0;
4598
				sprintf(nm, "rd%d", rdev->raid_disk);
4599 4600 4601 4602 4603 4604
				if (sysfs_create_link(&mddev->kobj,
						      &rdev->kobj, nm))
					printk(KERN_WARNING
					       "raid5: failed to create "
					       " link %s for %s\n",
					       nm, mdname(mddev));
4605 4606 4607 4608
			} else
				break;
		}

4609
	spin_lock_irqsave(&conf->device_lock, flags);
4610
	mddev->degraded = (conf->raid_disks - conf->previous_raid_disks) - added_devices;
4611
	spin_unlock_irqrestore(&conf->device_lock, flags);
4612
	mddev->raid_disks = conf->raid_disks;
4613
	mddev->reshape_position = 0;
4614
	set_bit(MD_CHANGE_DEVS, &mddev->flags);
4615

4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639
	clear_bit(MD_RECOVERY_SYNC, &mddev->recovery);
	clear_bit(MD_RECOVERY_CHECK, &mddev->recovery);
	set_bit(MD_RECOVERY_RESHAPE, &mddev->recovery);
	set_bit(MD_RECOVERY_RUNNING, &mddev->recovery);
	mddev->sync_thread = md_register_thread(md_do_sync, mddev,
						"%s_reshape");
	if (!mddev->sync_thread) {
		mddev->recovery = 0;
		spin_lock_irq(&conf->device_lock);
		mddev->raid_disks = conf->raid_disks = conf->previous_raid_disks;
		conf->expand_progress = MaxSector;
		spin_unlock_irq(&conf->device_lock);
		return -EAGAIN;
	}
	md_wakeup_thread(mddev->sync_thread);
	md_new_event(mddev);
	return 0;
}
#endif

static void end_reshape(raid5_conf_t *conf)
{
	struct block_device *bdev;

4640
	if (!test_bit(MD_RECOVERY_INTR, &conf->mddev->recovery)) {
4641 4642
		conf->mddev->array_size = conf->mddev->size *
			(conf->raid_disks - conf->max_degraded);
4643
		set_capacity(conf->mddev->gendisk, conf->mddev->array_size << 1);
4644
		conf->mddev->changed = 1;
4645 4646 4647 4648

		bdev = bdget_disk(conf->mddev->gendisk, 0);
		if (bdev) {
			mutex_lock(&bdev->bd_inode->i_mutex);
4649
			i_size_write(bdev->bd_inode, (loff_t)conf->mddev->array_size << 10);
4650 4651 4652 4653 4654 4655 4656
			mutex_unlock(&bdev->bd_inode->i_mutex);
			bdput(bdev);
		}
		spin_lock_irq(&conf->device_lock);
		conf->expand_progress = MaxSector;
		spin_unlock_irq(&conf->device_lock);
		conf->mddev->reshape_position = MaxSector;
4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667

		/* read-ahead size must cover two whole stripes, which is
		 * 2 * (datadisks) * chunksize where 'n' is the number of raid devices
		 */
		{
			int data_disks = conf->previous_raid_disks - conf->max_degraded;
			int stripe = data_disks *
				(conf->mddev->chunk_size / PAGE_SIZE);
			if (conf->mddev->queue->backing_dev_info.ra_pages < 2 * stripe)
				conf->mddev->queue->backing_dev_info.ra_pages = 2 * stripe;
		}
4668 4669 4670
	}
}

4671 4672 4673 4674 4675
static void raid5_quiesce(mddev_t *mddev, int state)
{
	raid5_conf_t *conf = mddev_to_conf(mddev);

	switch(state) {
4676 4677 4678 4679
	case 2: /* resume for a suspend */
		wake_up(&conf->wait_for_overlap);
		break;

4680 4681 4682 4683
	case 1: /* stop all writes */
		spin_lock_irq(&conf->device_lock);
		conf->quiesce = 1;
		wait_event_lock_irq(conf->wait_for_stripe,
4684 4685
				    atomic_read(&conf->active_stripes) == 0 &&
				    atomic_read(&conf->active_aligned_reads) == 0,
4686 4687 4688 4689 4690 4691 4692 4693
				    conf->device_lock, /* nothing */);
		spin_unlock_irq(&conf->device_lock);
		break;

	case 0: /* re-enable writes */
		spin_lock_irq(&conf->device_lock);
		conf->quiesce = 0;
		wake_up(&conf->wait_for_stripe);
4694
		wake_up(&conf->wait_for_overlap);
4695 4696 4697 4698
		spin_unlock_irq(&conf->device_lock);
		break;
	}
}
4699

4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714
static struct mdk_personality raid6_personality =
{
	.name		= "raid6",
	.level		= 6,
	.owner		= THIS_MODULE,
	.make_request	= make_request,
	.run		= run,
	.stop		= stop,
	.status		= status,
	.error_handler	= error,
	.hot_add_disk	= raid5_add_disk,
	.hot_remove_disk= raid5_remove_disk,
	.spare_active	= raid5_spare_active,
	.sync_request	= sync_request,
	.resize		= raid5_resize,
4715 4716 4717 4718
#ifdef CONFIG_MD_RAID5_RESHAPE
	.check_reshape	= raid5_check_reshape,
	.start_reshape  = raid5_start_reshape,
#endif
4719 4720
	.quiesce	= raid5_quiesce,
};
4721
static struct mdk_personality raid5_personality =
Linus Torvalds's avatar
Linus Torvalds committed
4722 4723
{
	.name		= "raid5",
4724
	.level		= 5,
Linus Torvalds's avatar
Linus Torvalds committed
4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735
	.owner		= THIS_MODULE,
	.make_request	= make_request,
	.run		= run,
	.stop		= stop,
	.status		= status,
	.error_handler	= error,
	.hot_add_disk	= raid5_add_disk,
	.hot_remove_disk= raid5_remove_disk,
	.spare_active	= raid5_spare_active,
	.sync_request	= sync_request,
	.resize		= raid5_resize,
4736
#ifdef CONFIG_MD_RAID5_RESHAPE
4737 4738
	.check_reshape	= raid5_check_reshape,
	.start_reshape  = raid5_start_reshape,
4739
#endif
4740
	.quiesce	= raid5_quiesce,
Linus Torvalds's avatar
Linus Torvalds committed
4741 4742
};

4743
static struct mdk_personality raid4_personality =
Linus Torvalds's avatar
Linus Torvalds committed
4744
{
4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757
	.name		= "raid4",
	.level		= 4,
	.owner		= THIS_MODULE,
	.make_request	= make_request,
	.run		= run,
	.stop		= stop,
	.status		= status,
	.error_handler	= error,
	.hot_add_disk	= raid5_add_disk,
	.hot_remove_disk= raid5_remove_disk,
	.spare_active	= raid5_spare_active,
	.sync_request	= sync_request,
	.resize		= raid5_resize,
4758 4759 4760 4761
#ifdef CONFIG_MD_RAID5_RESHAPE
	.check_reshape	= raid5_check_reshape,
	.start_reshape  = raid5_start_reshape,
#endif
4762 4763 4764 4765 4766
	.quiesce	= raid5_quiesce,
};

static int __init raid5_init(void)
{
4767 4768 4769 4770 4771 4772
	int e;

	e = raid6_select_algo();
	if ( e )
		return e;
	register_md_personality(&raid6_personality);
4773 4774 4775
	register_md_personality(&raid5_personality);
	register_md_personality(&raid4_personality);
	return 0;
Linus Torvalds's avatar
Linus Torvalds committed
4776 4777
}

4778
static void raid5_exit(void)
Linus Torvalds's avatar
Linus Torvalds committed
4779
{
4780
	unregister_md_personality(&raid6_personality);
4781 4782
	unregister_md_personality(&raid5_personality);
	unregister_md_personality(&raid4_personality);
Linus Torvalds's avatar
Linus Torvalds committed
4783 4784 4785 4786 4787 4788
}

module_init(raid5_init);
module_exit(raid5_exit);
MODULE_LICENSE("GPL");
MODULE_ALIAS("md-personality-4"); /* RAID5 */
4789 4790
MODULE_ALIAS("md-raid5");
MODULE_ALIAS("md-raid4");
4791 4792
MODULE_ALIAS("md-level-5");
MODULE_ALIAS("md-level-4");
4793 4794 4795 4796 4797 4798 4799
MODULE_ALIAS("md-personality-8"); /* RAID6 */
MODULE_ALIAS("md-raid6");
MODULE_ALIAS("md-level-6");

/* This used to be two separate modules, they were: */
MODULE_ALIAS("raid5");
MODULE_ALIAS("raid6");