raid5-cache.c 33.7 KB
Newer Older
S
Shaohua Li 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
/*
 * Copyright (C) 2015 Shaohua Li <shli@fb.com>
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms and conditions of the GNU General Public License,
 * version 2, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
 * more details.
 *
 */
#include <linux/kernel.h>
#include <linux/wait.h>
#include <linux/blkdev.h>
#include <linux/slab.h>
#include <linux/raid/md_p.h>
S
Shaohua Li 已提交
19
#include <linux/crc32c.h>
S
Shaohua Li 已提交
20 21 22 23 24 25 26 27 28 29
#include <linux/random.h>
#include "md.h"
#include "raid5.h"

/*
 * metadata/data stored in disk with 4k size unit (a block) regardless
 * underneath hardware sector size. only works with PAGE_SIZE == 4096
 */
#define BLOCK_SECTORS (8)

S
Shaohua Li 已提交
30 31 32 33 34 35 36
/*
 * reclaim runs every 1/4 disk size or 10G reclaimable space. This can prevent
 * recovery scans a very long log
 */
#define RECLAIM_MAX_FREE_SPACE (10 * 1024 * 1024 * 2) /* sector */
#define RECLAIM_MAX_FREE_SPACE_SHIFT (2)

C
Christoph Hellwig 已提交
37 38 39 40 41 42
/*
 * We only need 2 bios per I/O unit to make progress, but ensure we
 * have a few more available to not get too tight.
 */
#define R5L_POOL_SIZE	4

S
Shaohua Li 已提交
43 44 45 46 47 48 49
struct r5l_log {
	struct md_rdev *rdev;

	u32 uuid_checksum;

	sector_t device_size;		/* log device size, round to
					 * BLOCK_SECTORS */
S
Shaohua Li 已提交
50 51
	sector_t max_free_space;	/* reclaim run if free space is at
					 * this size */
S
Shaohua Li 已提交
52 53 54 55 56 57 58 59

	sector_t last_checkpoint;	/* log tail. where recovery scan
					 * starts from */
	u64 last_cp_seq;		/* log tail sequence */

	sector_t log_start;		/* log head. where new data appends */
	u64 seq;			/* log head sequence */

60 61 62
	sector_t next_checkpoint;
	u64 next_cp_seq;

S
Shaohua Li 已提交
63 64 65 66 67 68 69 70 71 72
	struct mutex io_mutex;
	struct r5l_io_unit *current_io;	/* current io_unit accepting new data */

	spinlock_t io_list_lock;
	struct list_head running_ios;	/* io_units which are still running,
					 * and have not yet been completely
					 * written to the log */
	struct list_head io_end_ios;	/* io_units which have been completely
					 * written to the log but not yet written
					 * to the RAID */
73 74
	struct list_head flushing_ios;	/* io_units which are waiting for log
					 * cache flush */
75
	struct list_head finished_ios;	/* io_units which settle down in log disk */
76
	struct bio flush_bio;
S
Shaohua Li 已提交
77

78 79
	struct list_head no_mem_stripes;   /* pending stripes, -ENOMEM */

S
Shaohua Li 已提交
80
	struct kmem_cache *io_kc;
81
	mempool_t *io_pool;
C
Christoph Hellwig 已提交
82
	struct bio_set *bs;
83
	mempool_t *meta_pool;
S
Shaohua Li 已提交
84

S
Shaohua Li 已提交
85 86 87 88 89 90 91 92
	struct md_thread *reclaim_thread;
	unsigned long reclaim_target;	/* number of space that need to be
					 * reclaimed.  if it's 0, reclaim spaces
					 * used by io_units which are in
					 * IO_UNIT_STRIPE_END state (eg, reclaim
					 * dones't wait for specific io_unit
					 * switching to IO_UNIT_STRIPE_END
					 * state) */
93
	wait_queue_head_t iounit_wait;
S
Shaohua Li 已提交
94

S
Shaohua Li 已提交
95 96
	struct list_head no_space_stripes; /* pending stripes, log has no space */
	spinlock_t no_space_stripes_lock;
97 98

	bool need_cache_flush;
S
Shaohua Li 已提交
99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
};

/*
 * an IO range starts from a meta data block and end at the next meta data
 * block. The io unit's the meta data block tracks data/parity followed it. io
 * unit is written to log disk with normal write, as we always flush log disk
 * first and then start move data to raid disks, there is no requirement to
 * write io unit with FLUSH/FUA
 */
struct r5l_io_unit {
	struct r5l_log *log;

	struct page *meta_page;	/* store meta block */
	int meta_offset;	/* current offset in meta_page */

	struct bio *current_bio;/* current_bio accepting new data */

	atomic_t pending_stripe;/* how many stripes not flushed to raid */
	u64 seq;		/* seq number of the metablock */
	sector_t log_start;	/* where the io_unit starts */
	sector_t log_end;	/* where the io_unit ends */
	struct list_head log_sibling; /* log->running_ios */
	struct list_head stripe_list; /* stripes added to the io_unit */

	int state;
C
Christoph Hellwig 已提交
124
	bool need_split_bio;
S
Shaohua Li 已提交
125 126 127 128 129 130 131 132
};

/* r5l_io_unit state */
enum r5l_io_unit_state {
	IO_UNIT_RUNNING = 0,	/* accepting new IO */
	IO_UNIT_IO_START = 1,	/* io_unit bio start writing to log,
				 * don't accepting new bio */
	IO_UNIT_IO_END = 2,	/* io_unit bio finish writing to log */
133
	IO_UNIT_STRIPE_END = 3,	/* stripes data finished writing to raid */
S
Shaohua Li 已提交
134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
};

static sector_t r5l_ring_add(struct r5l_log *log, sector_t start, sector_t inc)
{
	start += inc;
	if (start >= log->device_size)
		start = start - log->device_size;
	return start;
}

static sector_t r5l_ring_distance(struct r5l_log *log, sector_t start,
				  sector_t end)
{
	if (end >= start)
		return end - start;
	else
		return end + log->device_size - start;
}

static bool r5l_has_free_space(struct r5l_log *log, sector_t size)
{
	sector_t used_size;

	used_size = r5l_ring_distance(log, log->last_checkpoint,
					log->log_start);

	return log->device_size > used_size + size;
}

static void __r5l_set_io_unit_state(struct r5l_io_unit *io,
				    enum r5l_io_unit_state state)
{
	if (WARN_ON(io->state >= state))
		return;
	io->state = state;
}

171 172 173 174 175 176 177 178 179 180 181
static void r5l_io_run_stripes(struct r5l_io_unit *io)
{
	struct stripe_head *sh, *next;

	list_for_each_entry_safe(sh, next, &io->stripe_list, log_list) {
		list_del_init(&sh->log_list);
		set_bit(STRIPE_HANDLE, &sh->state);
		raid5_release_stripe(sh);
	}
}

182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197
static void r5l_log_run_stripes(struct r5l_log *log)
{
	struct r5l_io_unit *io, *next;

	assert_spin_locked(&log->io_list_lock);

	list_for_each_entry_safe(io, next, &log->running_ios, log_sibling) {
		/* don't change list order */
		if (io->state < IO_UNIT_IO_END)
			break;

		list_move_tail(&io->log_sibling, &log->finished_ios);
		r5l_io_run_stripes(io);
	}
}

198 199 200 201 202 203 204 205 206 207 208 209 210 211
static void r5l_move_to_end_ios(struct r5l_log *log)
{
	struct r5l_io_unit *io, *next;

	assert_spin_locked(&log->io_list_lock);

	list_for_each_entry_safe(io, next, &log->running_ios, log_sibling) {
		/* don't change list order */
		if (io->state < IO_UNIT_IO_END)
			break;
		list_move_tail(&io->log_sibling, &log->io_end_ios);
	}
}

S
Shaohua Li 已提交
212 213 214 215
static void r5l_log_endio(struct bio *bio)
{
	struct r5l_io_unit *io = bio->bi_private;
	struct r5l_log *log = io->log;
216
	unsigned long flags;
S
Shaohua Li 已提交
217

S
Shaohua Li 已提交
218 219 220
	if (bio->bi_error)
		md_error(log->rdev->mddev, log->rdev);

S
Shaohua Li 已提交
221
	bio_put(bio);
222
	mempool_free(io->meta_page, log->meta_pool);
S
Shaohua Li 已提交
223

224 225
	spin_lock_irqsave(&log->io_list_lock, flags);
	__r5l_set_io_unit_state(io, IO_UNIT_IO_END);
226
	if (log->need_cache_flush)
227
		r5l_move_to_end_ios(log);
228 229
	else
		r5l_log_run_stripes(log);
230 231
	spin_unlock_irqrestore(&log->io_list_lock, flags);

232 233
	if (log->need_cache_flush)
		md_wakeup_thread(log->rdev->mddev->thread);
S
Shaohua Li 已提交
234 235 236 237 238 239
}

static void r5l_submit_current_io(struct r5l_log *log)
{
	struct r5l_io_unit *io = log->current_io;
	struct r5l_meta_block *block;
240
	unsigned long flags;
S
Shaohua Li 已提交
241 242 243 244 245 246 247
	u32 crc;

	if (!io)
		return;

	block = page_address(io->meta_page);
	block->meta_size = cpu_to_le32(io->meta_offset);
S
Shaohua Li 已提交
248
	crc = crc32c_le(log->uuid_checksum, block, PAGE_SIZE);
S
Shaohua Li 已提交
249 250 251
	block->checksum = cpu_to_le32(crc);

	log->current_io = NULL;
252 253 254
	spin_lock_irqsave(&log->io_list_lock, flags);
	__r5l_set_io_unit_state(io, IO_UNIT_IO_START);
	spin_unlock_irqrestore(&log->io_list_lock, flags);
S
Shaohua Li 已提交
255

256
	submit_bio(io->current_bio);
S
Shaohua Li 已提交
257 258
}

C
Christoph Hellwig 已提交
259
static struct bio *r5l_bio_alloc(struct r5l_log *log)
260
{
C
Christoph Hellwig 已提交
261
	struct bio *bio = bio_alloc_bioset(GFP_NOIO, BIO_MAX_PAGES, log->bs);
262

M
Mike Christie 已提交
263
	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
264
	bio->bi_bdev = log->rdev->bdev;
265
	bio->bi_iter.bi_sector = log->rdev->data_offset + log->log_start;
266 267 268 269

	return bio;
}

270 271 272 273 274 275 276 277 278 279 280 281
static void r5_reserve_log_entry(struct r5l_log *log, struct r5l_io_unit *io)
{
	log->log_start = r5l_ring_add(log, log->log_start, BLOCK_SECTORS);

	/*
	 * If we filled up the log device start from the beginning again,
	 * which will require a new bio.
	 *
	 * Note: for this to work properly the log size needs to me a multiple
	 * of BLOCK_SECTORS.
	 */
	if (log->log_start == 0)
C
Christoph Hellwig 已提交
282
		io->need_split_bio = true;
283 284 285 286

	io->log_end = log->log_start;
}

S
Shaohua Li 已提交
287 288 289 290 291
static struct r5l_io_unit *r5l_new_meta(struct r5l_log *log)
{
	struct r5l_io_unit *io;
	struct r5l_meta_block *block;

292 293 294 295 296
	io = mempool_alloc(log->io_pool, GFP_ATOMIC);
	if (!io)
		return NULL;
	memset(io, 0, sizeof(*io));

297 298 299 300
	io->log = log;
	INIT_LIST_HEAD(&io->log_sibling);
	INIT_LIST_HEAD(&io->stripe_list);
	io->state = IO_UNIT_RUNNING;
S
Shaohua Li 已提交
301

302
	io->meta_page = mempool_alloc(log->meta_pool, GFP_NOIO);
S
Shaohua Li 已提交
303
	block = page_address(io->meta_page);
304
	clear_page(block);
S
Shaohua Li 已提交
305 306 307 308 309 310 311
	block->magic = cpu_to_le32(R5LOG_MAGIC);
	block->version = R5LOG_VERSION;
	block->seq = cpu_to_le64(log->seq);
	block->position = cpu_to_le64(log->log_start);

	io->log_start = log->log_start;
	io->meta_offset = sizeof(struct r5l_meta_block);
312
	io->seq = log->seq++;
S
Shaohua Li 已提交
313

C
Christoph Hellwig 已提交
314 315 316
	io->current_bio = r5l_bio_alloc(log);
	io->current_bio->bi_end_io = r5l_log_endio;
	io->current_bio->bi_private = io;
317
	bio_add_page(io->current_bio, io->meta_page, PAGE_SIZE, 0);
S
Shaohua Li 已提交
318

319
	r5_reserve_log_entry(log, io);
S
Shaohua Li 已提交
320 321 322 323 324 325 326 327 328 329

	spin_lock_irq(&log->io_list_lock);
	list_add_tail(&io->log_sibling, &log->running_ios);
	spin_unlock_irq(&log->io_list_lock);

	return io;
}

static int r5l_get_meta(struct r5l_log *log, unsigned int payload_size)
{
330 331
	if (log->current_io &&
	    log->current_io->meta_offset + payload_size > PAGE_SIZE)
S
Shaohua Li 已提交
332 333
		r5l_submit_current_io(log);

334
	if (!log->current_io) {
335
		log->current_io = r5l_new_meta(log);
336 337 338 339
		if (!log->current_io)
			return -ENOMEM;
	}

S
Shaohua Li 已提交
340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368
	return 0;
}

static void r5l_append_payload_meta(struct r5l_log *log, u16 type,
				    sector_t location,
				    u32 checksum1, u32 checksum2,
				    bool checksum2_valid)
{
	struct r5l_io_unit *io = log->current_io;
	struct r5l_payload_data_parity *payload;

	payload = page_address(io->meta_page) + io->meta_offset;
	payload->header.type = cpu_to_le16(type);
	payload->header.flags = cpu_to_le16(0);
	payload->size = cpu_to_le32((1 + !!checksum2_valid) <<
				    (PAGE_SHIFT - 9));
	payload->location = cpu_to_le64(location);
	payload->checksum[0] = cpu_to_le32(checksum1);
	if (checksum2_valid)
		payload->checksum[1] = cpu_to_le32(checksum2);

	io->meta_offset += sizeof(struct r5l_payload_data_parity) +
		sizeof(__le32) * (1 + !!checksum2_valid);
}

static void r5l_append_payload_page(struct r5l_log *log, struct page *page)
{
	struct r5l_io_unit *io = log->current_io;

C
Christoph Hellwig 已提交
369 370
	if (io->need_split_bio) {
		struct bio *prev = io->current_bio;
371

C
Christoph Hellwig 已提交
372 373 374
		io->current_bio = r5l_bio_alloc(log);
		bio_chain(io->current_bio, prev);

375
		submit_bio(prev);
S
Shaohua Li 已提交
376 377
	}

C
Christoph Hellwig 已提交
378 379 380
	if (!bio_add_page(io->current_bio, page, PAGE_SIZE, 0))
		BUG();

381
	r5_reserve_log_entry(log, io);
S
Shaohua Li 已提交
382 383
}

384
static int r5l_log_stripe(struct r5l_log *log, struct stripe_head *sh,
S
Shaohua Li 已提交
385 386 387 388
			   int data_pages, int parity_pages)
{
	int i;
	int meta_size;
389
	int ret;
S
Shaohua Li 已提交
390 391 392 393 394 395 396 397
	struct r5l_io_unit *io;

	meta_size =
		((sizeof(struct r5l_payload_data_parity) + sizeof(__le32))
		 * data_pages) +
		sizeof(struct r5l_payload_data_parity) +
		sizeof(__le32) * parity_pages;

398 399 400 401
	ret = r5l_get_meta(log, meta_size);
	if (ret)
		return ret;

S
Shaohua Li 已提交
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
	io = log->current_io;

	for (i = 0; i < sh->disks; i++) {
		if (!test_bit(R5_Wantwrite, &sh->dev[i].flags))
			continue;
		if (i == sh->pd_idx || i == sh->qd_idx)
			continue;
		r5l_append_payload_meta(log, R5LOG_PAYLOAD_DATA,
					raid5_compute_blocknr(sh, i, 0),
					sh->dev[i].log_checksum, 0, false);
		r5l_append_payload_page(log, sh->dev[i].page);
	}

	if (sh->qd_idx >= 0) {
		r5l_append_payload_meta(log, R5LOG_PAYLOAD_PARITY,
					sh->sector, sh->dev[sh->pd_idx].log_checksum,
					sh->dev[sh->qd_idx].log_checksum, true);
		r5l_append_payload_page(log, sh->dev[sh->pd_idx].page);
		r5l_append_payload_page(log, sh->dev[sh->qd_idx].page);
	} else {
		r5l_append_payload_meta(log, R5LOG_PAYLOAD_PARITY,
					sh->sector, sh->dev[sh->pd_idx].log_checksum,
					0, false);
		r5l_append_payload_page(log, sh->dev[sh->pd_idx].page);
	}

	list_add_tail(&sh->log_list, &io->stripe_list);
	atomic_inc(&io->pending_stripe);
	sh->log_io = io;
431 432

	return 0;
S
Shaohua Li 已提交
433 434
}

435
static void r5l_wake_reclaim(struct r5l_log *log, sector_t space);
S
Shaohua Li 已提交
436 437 438 439 440 441 442 443 444 445 446
/*
 * running in raid5d, where reclaim could wait for raid5d too (when it flushes
 * data from log to raid disks), so we shouldn't wait for reclaim here
 */
int r5l_write_stripe(struct r5l_log *log, struct stripe_head *sh)
{
	int write_disks = 0;
	int data_pages, parity_pages;
	int meta_size;
	int reserve;
	int i;
447
	int ret = 0;
S
Shaohua Li 已提交
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468

	if (!log)
		return -EAGAIN;
	/* Don't support stripe batch */
	if (sh->log_io || !test_bit(R5_Wantwrite, &sh->dev[sh->pd_idx].flags) ||
	    test_bit(STRIPE_SYNCING, &sh->state)) {
		/* the stripe is written to log, we start writing it to raid */
		clear_bit(STRIPE_LOG_TRAPPED, &sh->state);
		return -EAGAIN;
	}

	for (i = 0; i < sh->disks; i++) {
		void *addr;

		if (!test_bit(R5_Wantwrite, &sh->dev[i].flags))
			continue;
		write_disks++;
		/* checksum is already calculated in last run */
		if (test_bit(STRIPE_LOG_TRAPPED, &sh->state))
			continue;
		addr = kmap_atomic(sh->dev[i].page);
S
Shaohua Li 已提交
469 470
		sh->dev[i].log_checksum = crc32c_le(log->uuid_checksum,
						    addr, PAGE_SIZE);
S
Shaohua Li 已提交
471 472 473 474 475 476 477 478 479 480 481 482 483 484 485
		kunmap_atomic(addr);
	}
	parity_pages = 1 + !!(sh->qd_idx >= 0);
	data_pages = write_disks - parity_pages;

	meta_size =
		((sizeof(struct r5l_payload_data_parity) + sizeof(__le32))
		 * data_pages) +
		sizeof(struct r5l_payload_data_parity) +
		sizeof(__le32) * parity_pages;
	/* Doesn't work with very big raid array */
	if (meta_size + sizeof(struct r5l_meta_block) > PAGE_SIZE)
		return -EINVAL;

	set_bit(STRIPE_LOG_TRAPPED, &sh->state);
486 487 488 489 490
	/*
	 * The stripe must enter state machine again to finish the write, so
	 * don't delay.
	 */
	clear_bit(STRIPE_DELAYED, &sh->state);
S
Shaohua Li 已提交
491 492 493 494 495
	atomic_inc(&sh->count);

	mutex_lock(&log->io_mutex);
	/* meta + data */
	reserve = (1 + write_disks) << (PAGE_SHIFT - 9);
496
	if (!r5l_has_free_space(log, reserve)) {
S
Shaohua Li 已提交
497 498 499 500 501
		spin_lock(&log->no_space_stripes_lock);
		list_add_tail(&sh->log_list, &log->no_space_stripes);
		spin_unlock(&log->no_space_stripes_lock);

		r5l_wake_reclaim(log, reserve);
502 503 504 505 506 507 508
	} else {
		ret = r5l_log_stripe(log, sh, data_pages, parity_pages);
		if (ret) {
			spin_lock_irq(&log->io_list_lock);
			list_add_tail(&sh->log_list, &log->no_mem_stripes);
			spin_unlock_irq(&log->io_list_lock);
		}
S
Shaohua Li 已提交
509 510
	}

511
	mutex_unlock(&log->io_mutex);
S
Shaohua Li 已提交
512 513 514 515 516 517 518 519 520 521 522 523
	return 0;
}

void r5l_write_stripe_run(struct r5l_log *log)
{
	if (!log)
		return;
	mutex_lock(&log->io_mutex);
	r5l_submit_current_io(log);
	mutex_unlock(&log->io_mutex);
}

524 525 526 527 528 529 530 531 532 533 534 535 536 537
int r5l_handle_flush_request(struct r5l_log *log, struct bio *bio)
{
	if (!log)
		return -ENODEV;
	/*
	 * we flush log disk cache first, then write stripe data to raid disks.
	 * So if bio is finished, the log disk cache is flushed already. The
	 * recovery guarantees we can recovery the bio from log disk, so we
	 * don't need to flush again
	 */
	if (bio->bi_iter.bi_size == 0) {
		bio_endio(bio);
		return 0;
	}
J
Jens Axboe 已提交
538
	bio->bi_opf &= ~REQ_PREFLUSH;
539 540 541
	return -EAGAIN;
}

S
Shaohua Li 已提交
542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557
/* This will run after log space is reclaimed */
static void r5l_run_no_space_stripes(struct r5l_log *log)
{
	struct stripe_head *sh;

	spin_lock(&log->no_space_stripes_lock);
	while (!list_empty(&log->no_space_stripes)) {
		sh = list_first_entry(&log->no_space_stripes,
				      struct stripe_head, log_list);
		list_del_init(&sh->log_list);
		set_bit(STRIPE_HANDLE, &sh->state);
		raid5_release_stripe(sh);
	}
	spin_unlock(&log->no_space_stripes_lock);
}

558 559 560 561 562 563
static sector_t r5l_reclaimable_space(struct r5l_log *log)
{
	return r5l_ring_distance(log, log->last_checkpoint,
				 log->next_checkpoint);
}

564 565 566 567 568 569 570 571 572 573 574 575 576 577 578
static void r5l_run_no_mem_stripe(struct r5l_log *log)
{
	struct stripe_head *sh;

	assert_spin_locked(&log->io_list_lock);

	if (!list_empty(&log->no_mem_stripes)) {
		sh = list_first_entry(&log->no_mem_stripes,
				      struct stripe_head, log_list);
		list_del_init(&sh->log_list);
		set_bit(STRIPE_HANDLE, &sh->state);
		raid5_release_stripe(sh);
	}
}

579
static bool r5l_complete_finished_ios(struct r5l_log *log)
580 581 582 583 584 585
{
	struct r5l_io_unit *io, *next;
	bool found = false;

	assert_spin_locked(&log->io_list_lock);

586
	list_for_each_entry_safe(io, next, &log->finished_ios, log_sibling) {
587 588 589 590 591 592 593 594
		/* don't change list order */
		if (io->state < IO_UNIT_STRIPE_END)
			break;

		log->next_checkpoint = io->log_start;
		log->next_cp_seq = io->seq;

		list_del(&io->log_sibling);
595 596
		mempool_free(io, log->io_pool);
		r5l_run_no_mem_stripe(log);
597 598 599 600 601 602 603

		found = true;
	}

	return found;
}

604 605 606 607 608 609 610
static void __r5l_stripe_write_finished(struct r5l_io_unit *io)
{
	struct r5l_log *log = io->log;
	unsigned long flags;

	spin_lock_irqsave(&log->io_list_lock, flags);
	__r5l_set_io_unit_state(io, IO_UNIT_STRIPE_END);
611

612
	if (!r5l_complete_finished_ios(log)) {
613 614 615
		spin_unlock_irqrestore(&log->io_list_lock, flags);
		return;
	}
616

617
	if (r5l_reclaimable_space(log) > log->max_free_space)
618 619 620 621 622 623
		r5l_wake_reclaim(log, 0);

	spin_unlock_irqrestore(&log->io_list_lock, flags);
	wake_up(&log->iounit_wait);
}

S
Shaohua Li 已提交
624 625 626 627 628 629 630
void r5l_stripe_write_finished(struct stripe_head *sh)
{
	struct r5l_io_unit *io;

	io = sh->log_io;
	sh->log_io = NULL;

631 632
	if (io && atomic_dec_and_test(&io->pending_stripe))
		__r5l_stripe_write_finished(io);
S
Shaohua Li 已提交
633 634
}

635 636 637 638 639 640 641
static void r5l_log_flush_endio(struct bio *bio)
{
	struct r5l_log *log = container_of(bio, struct r5l_log,
		flush_bio);
	unsigned long flags;
	struct r5l_io_unit *io;

S
Shaohua Li 已提交
642 643 644
	if (bio->bi_error)
		md_error(log->rdev->mddev, log->rdev);

645
	spin_lock_irqsave(&log->io_list_lock, flags);
646 647
	list_for_each_entry(io, &log->flushing_ios, log_sibling)
		r5l_io_run_stripes(io);
648
	list_splice_tail_init(&log->flushing_ios, &log->finished_ios);
649 650 651
	spin_unlock_irqrestore(&log->io_list_lock, flags);
}

S
Shaohua Li 已提交
652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667
/*
 * Starting dispatch IO to raid.
 * io_unit(meta) consists of a log. There is one situation we want to avoid. A
 * broken meta in the middle of a log causes recovery can't find meta at the
 * head of log. If operations require meta at the head persistent in log, we
 * must make sure meta before it persistent in log too. A case is:
 *
 * stripe data/parity is in log, we start write stripe to raid disks. stripe
 * data/parity must be persistent in log before we do the write to raid disks.
 *
 * The solution is we restrictly maintain io_unit list order. In this case, we
 * only write stripes of an io_unit to raid disks till the io_unit is the first
 * one whose data/parity is in log.
 */
void r5l_flush_stripe_to_raid(struct r5l_log *log)
{
668
	bool do_flush;
669 670

	if (!log || !log->need_cache_flush)
S
Shaohua Li 已提交
671 672 673
		return;

	spin_lock_irq(&log->io_list_lock);
674 675 676 677
	/* flush bio is running */
	if (!list_empty(&log->flushing_ios)) {
		spin_unlock_irq(&log->io_list_lock);
		return;
S
Shaohua Li 已提交
678
	}
679 680
	list_splice_tail_init(&log->io_end_ios, &log->flushing_ios);
	do_flush = !list_empty(&log->flushing_ios);
S
Shaohua Li 已提交
681
	spin_unlock_irq(&log->io_list_lock);
682 683 684 685 686 687

	if (!do_flush)
		return;
	bio_reset(&log->flush_bio);
	log->flush_bio.bi_bdev = log->rdev->bdev;
	log->flush_bio.bi_end_io = r5l_log_flush_endio;
M
Mike Christie 已提交
688
	bio_set_op_attrs(&log->flush_bio, REQ_OP_WRITE, WRITE_FLUSH);
689
	submit_bio(&log->flush_bio);
S
Shaohua Li 已提交
690 691 692
}

static void r5l_write_super(struct r5l_log *log, sector_t cp);
S
Shaohua Li 已提交
693 694 695 696 697 698 699 700 701 702 703 704 705
static void r5l_write_super_and_discard_space(struct r5l_log *log,
	sector_t end)
{
	struct block_device *bdev = log->rdev->bdev;
	struct mddev *mddev;

	r5l_write_super(log, end);

	if (!blk_queue_discard(bdev_get_queue(bdev)))
		return;

	mddev = log->rdev->mddev;
	/*
706 707 708 709 710 711 712 713 714
	 * Discard could zero data, so before discard we must make sure
	 * superblock is updated to new log tail. Updating superblock (either
	 * directly call md_update_sb() or depend on md thread) must hold
	 * reconfig mutex. On the other hand, raid5_quiesce is called with
	 * reconfig_mutex hold. The first step of raid5_quiesce() is waitting
	 * for all IO finish, hence waitting for reclaim thread, while reclaim
	 * thread is calling this function and waitting for reconfig mutex. So
	 * there is a deadlock. We workaround this issue with a trylock.
	 * FIXME: we could miss discard if we can't take reconfig mutex
S
Shaohua Li 已提交
715
	 */
716 717 718 719 720 721
	set_mask_bits(&mddev->flags, 0,
		BIT(MD_CHANGE_DEVS) | BIT(MD_CHANGE_PENDING));
	if (!mddev_trylock(mddev))
		return;
	md_update_sb(mddev, 1);
	mddev_unlock(mddev);
S
Shaohua Li 已提交
722

S
Shaohua Li 已提交
723
	/* discard IO error really doesn't matter, ignore it */
S
Shaohua Li 已提交
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738
	if (log->last_checkpoint < end) {
		blkdev_issue_discard(bdev,
				log->last_checkpoint + log->rdev->data_offset,
				end - log->last_checkpoint, GFP_NOIO, 0);
	} else {
		blkdev_issue_discard(bdev,
				log->last_checkpoint + log->rdev->data_offset,
				log->device_size - log->last_checkpoint,
				GFP_NOIO, 0);
		blkdev_issue_discard(bdev, log->rdev->data_offset, end,
				GFP_NOIO, 0);
	}
}


S
Shaohua Li 已提交
739 740 741
static void r5l_do_reclaim(struct r5l_log *log)
{
	sector_t reclaim_target = xchg(&log->reclaim_target, 0);
742 743 744
	sector_t reclaimable;
	sector_t next_checkpoint;
	u64 next_cp_seq;
S
Shaohua Li 已提交
745 746 747 748 749 750 751 752

	spin_lock_irq(&log->io_list_lock);
	/*
	 * move proper io_unit to reclaim list. We should not change the order.
	 * reclaimable/unreclaimable io_unit can be mixed in the list, we
	 * shouldn't reuse space of an unreclaimable io_unit
	 */
	while (1) {
753 754
		reclaimable = r5l_reclaimable_space(log);
		if (reclaimable >= reclaim_target ||
S
Shaohua Li 已提交
755 756
		    (list_empty(&log->running_ios) &&
		     list_empty(&log->io_end_ios) &&
757
		     list_empty(&log->flushing_ios) &&
758
		     list_empty(&log->finished_ios)))
S
Shaohua Li 已提交
759 760
			break;

761 762 763 764
		md_wakeup_thread(log->rdev->mddev->thread);
		wait_event_lock_irq(log->iounit_wait,
				    r5l_reclaimable_space(log) > reclaimable,
				    log->io_list_lock);
S
Shaohua Li 已提交
765
	}
766 767 768

	next_checkpoint = log->next_checkpoint;
	next_cp_seq = log->next_cp_seq;
S
Shaohua Li 已提交
769 770
	spin_unlock_irq(&log->io_list_lock);

771 772
	BUG_ON(reclaimable < 0);
	if (reclaimable == 0)
S
Shaohua Li 已提交
773 774 775 776 777 778 779
		return;

	/*
	 * write_super will flush cache of each raid disk. We must write super
	 * here, because the log area might be reused soon and we don't want to
	 * confuse recovery
	 */
S
Shaohua Li 已提交
780
	r5l_write_super_and_discard_space(log, next_checkpoint);
S
Shaohua Li 已提交
781 782

	mutex_lock(&log->io_mutex);
783 784
	log->last_checkpoint = next_checkpoint;
	log->last_cp_seq = next_cp_seq;
S
Shaohua Li 已提交
785 786
	mutex_unlock(&log->io_mutex);

787
	r5l_run_no_space_stripes(log);
S
Shaohua Li 已提交
788 789 790 791 792 793 794 795 796 797 798 799 800
}

static void r5l_reclaim_thread(struct md_thread *thread)
{
	struct mddev *mddev = thread->mddev;
	struct r5conf *conf = mddev->private;
	struct r5l_log *log = conf->log;

	if (!log)
		return;
	r5l_do_reclaim(log);
}

S
Shaohua Li 已提交
801 802
static void r5l_wake_reclaim(struct r5l_log *log, sector_t space)
{
S
Shaohua Li 已提交
803 804 805 806 807 808 809 810 811
	unsigned long target;
	unsigned long new = (unsigned long)space; /* overflow in theory */

	do {
		target = log->reclaim_target;
		if (new < target)
			return;
	} while (cmpxchg(&log->reclaim_target, target, new) != target);
	md_wakeup_thread(log->reclaim_thread);
S
Shaohua Li 已提交
812 813
}

814 815
void r5l_quiesce(struct r5l_log *log, int state)
{
S
Shaohua Li 已提交
816
	struct mddev *mddev;
817 818 819
	if (!log || state == 2)
		return;
	if (state == 0) {
820 821 822 823 824 825 826
		/*
		 * This is a special case for hotadd. In suspend, the array has
		 * no journal. In resume, journal is initialized as well as the
		 * reclaim thread.
		 */
		if (log->reclaim_thread)
			return;
827 828 829
		log->reclaim_thread = md_register_thread(r5l_reclaim_thread,
					log->rdev->mddev, "reclaim");
	} else if (state == 1) {
S
Shaohua Li 已提交
830 831 832
		/* make sure r5l_write_super_and_discard_space exits */
		mddev = log->rdev->mddev;
		wake_up(&mddev->sb_wait);
833 834 835 836 837 838
		r5l_wake_reclaim(log, -1L);
		md_unregister_thread(&log->reclaim_thread);
		r5l_do_reclaim(log);
	}
}

S
Shaohua Li 已提交
839 840
bool r5l_log_disk_error(struct r5conf *conf)
{
841 842
	struct r5l_log *log;
	bool ret;
843
	/* don't allow write if journal disk is missing */
844 845 846 847 848 849 850 851 852
	rcu_read_lock();
	log = rcu_dereference(conf->log);

	if (!log)
		ret = test_bit(MD_HAS_JOURNAL, &conf->mddev->flags);
	else
		ret = test_bit(Faulty, &log->rdev->flags);
	rcu_read_unlock();
	return ret;
S
Shaohua Li 已提交
853 854
}

S
Shaohua Li 已提交
855 856 857 858 859 860 861 862 863 864 865 866 867 868
struct r5l_recovery_ctx {
	struct page *meta_page;		/* current meta */
	sector_t meta_total_blocks;	/* total size of current meta and data */
	sector_t pos;			/* recovery position */
	u64 seq;			/* recovery position seq */
};

static int r5l_read_meta_block(struct r5l_log *log,
			       struct r5l_recovery_ctx *ctx)
{
	struct page *page = ctx->meta_page;
	struct r5l_meta_block *mb;
	u32 crc, stored_crc;

M
Mike Christie 已提交
869 870
	if (!sync_page_io(log->rdev, ctx->pos, PAGE_SIZE, page, REQ_OP_READ, 0,
			  false))
S
Shaohua Li 已提交
871 872 873 874 875 876 877 878 879 880 881 882
		return -EIO;

	mb = page_address(page);
	stored_crc = le32_to_cpu(mb->checksum);
	mb->checksum = 0;

	if (le32_to_cpu(mb->magic) != R5LOG_MAGIC ||
	    le64_to_cpu(mb->seq) != ctx->seq ||
	    mb->version != R5LOG_VERSION ||
	    le64_to_cpu(mb->position) != ctx->pos)
		return -EINVAL;

S
Shaohua Li 已提交
883
	crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
884 885 886 887 888 889 890 891 892 893 894 895 896 897
	if (stored_crc != crc)
		return -EINVAL;

	if (le32_to_cpu(mb->meta_size) > PAGE_SIZE)
		return -EINVAL;

	ctx->meta_total_blocks = BLOCK_SECTORS;

	return 0;
}

static int r5l_recovery_flush_one_stripe(struct r5l_log *log,
					 struct r5l_recovery_ctx *ctx,
					 sector_t stripe_sect,
898
					 int *offset)
S
Shaohua Li 已提交
899 900 901 902 903 904 905 906
{
	struct r5conf *conf = log->rdev->mddev->private;
	struct stripe_head *sh;
	struct r5l_payload_data_parity *payload;
	int disk_index;

	sh = raid5_get_active_stripe(conf, stripe_sect, 0, 0, 0);
	while (1) {
907 908
		sector_t log_offset = r5l_ring_add(log, ctx->pos,
				ctx->meta_total_blocks);
S
Shaohua Li 已提交
909 910 911 912 913 914 915
		payload = page_address(ctx->meta_page) + *offset;

		if (le16_to_cpu(payload->header.type) == R5LOG_PAYLOAD_DATA) {
			raid5_compute_sector(conf,
					     le64_to_cpu(payload->location), 0,
					     &disk_index, sh);

916
			sync_page_io(log->rdev, log_offset, PAGE_SIZE,
M
Mike Christie 已提交
917 918
				     sh->dev[disk_index].page, REQ_OP_READ, 0,
				     false);
S
Shaohua Li 已提交
919 920 921 922 923
			sh->dev[disk_index].log_checksum =
				le32_to_cpu(payload->checksum[0]);
			set_bit(R5_Wantwrite, &sh->dev[disk_index].flags);
		} else {
			disk_index = sh->pd_idx;
924
			sync_page_io(log->rdev, log_offset, PAGE_SIZE,
M
Mike Christie 已提交
925 926
				     sh->dev[disk_index].page, REQ_OP_READ, 0,
				     false);
S
Shaohua Li 已提交
927 928 929 930 931 932 933
			sh->dev[disk_index].log_checksum =
				le32_to_cpu(payload->checksum[0]);
			set_bit(R5_Wantwrite, &sh->dev[disk_index].flags);

			if (sh->qd_idx >= 0) {
				disk_index = sh->qd_idx;
				sync_page_io(log->rdev,
934
					     r5l_ring_add(log, log_offset, BLOCK_SECTORS),
S
Shaohua Li 已提交
935
					     PAGE_SIZE, sh->dev[disk_index].page,
M
Mike Christie 已提交
936
					     REQ_OP_READ, 0, false);
S
Shaohua Li 已提交
937 938 939 940 941 942 943
				sh->dev[disk_index].log_checksum =
					le32_to_cpu(payload->checksum[1]);
				set_bit(R5_Wantwrite,
					&sh->dev[disk_index].flags);
			}
		}

944
		ctx->meta_total_blocks += le32_to_cpu(payload->size);
S
Shaohua Li 已提交
945 946 947 948 949 950 951 952 953 954 955 956 957 958
		*offset += sizeof(struct r5l_payload_data_parity) +
			sizeof(__le32) *
			(le32_to_cpu(payload->size) >> (PAGE_SHIFT - 9));
		if (le16_to_cpu(payload->header.type) == R5LOG_PAYLOAD_PARITY)
			break;
	}

	for (disk_index = 0; disk_index < sh->disks; disk_index++) {
		void *addr;
		u32 checksum;

		if (!test_bit(R5_Wantwrite, &sh->dev[disk_index].flags))
			continue;
		addr = kmap_atomic(sh->dev[disk_index].page);
S
Shaohua Li 已提交
959
		checksum = crc32c_le(log->uuid_checksum, addr, PAGE_SIZE);
S
Shaohua Li 已提交
960 961 962 963 964 965 966 967 968 969 970 971 972
		kunmap_atomic(addr);
		if (checksum != sh->dev[disk_index].log_checksum)
			goto error;
	}

	for (disk_index = 0; disk_index < sh->disks; disk_index++) {
		struct md_rdev *rdev, *rrdev;

		if (!test_and_clear_bit(R5_Wantwrite,
					&sh->dev[disk_index].flags))
			continue;

		/* in case device is broken */
S
Shaohua Li 已提交
973
		rcu_read_lock();
S
Shaohua Li 已提交
974
		rdev = rcu_dereference(conf->disks[disk_index].rdev);
S
Shaohua Li 已提交
975 976 977
		if (rdev) {
			atomic_inc(&rdev->nr_pending);
			rcu_read_unlock();
S
Shaohua Li 已提交
978
			sync_page_io(rdev, stripe_sect, PAGE_SIZE,
M
Mike Christie 已提交
979 980
				     sh->dev[disk_index].page, REQ_OP_WRITE, 0,
				     false);
S
Shaohua Li 已提交
981 982 983
			rdev_dec_pending(rdev, rdev->mddev);
			rcu_read_lock();
		}
S
Shaohua Li 已提交
984
		rrdev = rcu_dereference(conf->disks[disk_index].replacement);
S
Shaohua Li 已提交
985 986 987
		if (rrdev) {
			atomic_inc(&rrdev->nr_pending);
			rcu_read_unlock();
S
Shaohua Li 已提交
988
			sync_page_io(rrdev, stripe_sect, PAGE_SIZE,
M
Mike Christie 已提交
989 990
				     sh->dev[disk_index].page, REQ_OP_WRITE, 0,
				     false);
S
Shaohua Li 已提交
991 992 993 994
			rdev_dec_pending(rrdev, rrdev->mddev);
			rcu_read_lock();
		}
		rcu_read_unlock();
S
Shaohua Li 已提交
995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
	}
	raid5_release_stripe(sh);
	return 0;

error:
	for (disk_index = 0; disk_index < sh->disks; disk_index++)
		sh->dev[disk_index].flags = 0;
	raid5_release_stripe(sh);
	return -EINVAL;
}

static int r5l_recovery_flush_one_meta(struct r5l_log *log,
				       struct r5l_recovery_ctx *ctx)
{
	struct r5conf *conf = log->rdev->mddev->private;
	struct r5l_payload_data_parity *payload;
	struct r5l_meta_block *mb;
	int offset;
	sector_t stripe_sector;

	mb = page_address(ctx->meta_page);
	offset = sizeof(struct r5l_meta_block);

	while (offset < le32_to_cpu(mb->meta_size)) {
		int dd;

		payload = (void *)mb + offset;
		stripe_sector = raid5_compute_sector(conf,
						     le64_to_cpu(payload->location), 0, &dd, NULL);
		if (r5l_recovery_flush_one_stripe(log, ctx, stripe_sector,
1025
						  &offset))
S
Shaohua Li 已提交
1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060
			return -EINVAL;
	}
	return 0;
}

/* copy data/parity from log to raid disks */
static void r5l_recovery_flush_log(struct r5l_log *log,
				   struct r5l_recovery_ctx *ctx)
{
	while (1) {
		if (r5l_read_meta_block(log, ctx))
			return;
		if (r5l_recovery_flush_one_meta(log, ctx))
			return;
		ctx->seq++;
		ctx->pos = r5l_ring_add(log, ctx->pos, ctx->meta_total_blocks);
	}
}

static int r5l_log_write_empty_meta_block(struct r5l_log *log, sector_t pos,
					  u64 seq)
{
	struct page *page;
	struct r5l_meta_block *mb;
	u32 crc;

	page = alloc_page(GFP_KERNEL | __GFP_ZERO);
	if (!page)
		return -ENOMEM;
	mb = page_address(page);
	mb->magic = cpu_to_le32(R5LOG_MAGIC);
	mb->version = R5LOG_VERSION;
	mb->meta_size = cpu_to_le32(sizeof(struct r5l_meta_block));
	mb->seq = cpu_to_le64(seq);
	mb->position = cpu_to_le64(pos);
S
Shaohua Li 已提交
1061
	crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
1062 1063
	mb->checksum = cpu_to_le32(crc);

M
Mike Christie 已提交
1064 1065
	if (!sync_page_io(log->rdev, pos, PAGE_SIZE, page, REQ_OP_WRITE,
			  WRITE_FUA, false)) {
S
Shaohua Li 已提交
1066 1067 1068 1069 1070 1071 1072
		__free_page(page);
		return -EIO;
	}
	__free_page(page);
	return 0;
}

S
Shaohua Li 已提交
1073 1074
static int r5l_recovery_log(struct r5l_log *log)
{
S
Shaohua Li 已提交
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098
	struct r5l_recovery_ctx ctx;

	ctx.pos = log->last_checkpoint;
	ctx.seq = log->last_cp_seq;
	ctx.meta_page = alloc_page(GFP_KERNEL);
	if (!ctx.meta_page)
		return -ENOMEM;

	r5l_recovery_flush_log(log, &ctx);
	__free_page(ctx.meta_page);

	/*
	 * we did a recovery. Now ctx.pos points to an invalid meta block. New
	 * log will start here. but we can't let superblock point to last valid
	 * meta block. The log might looks like:
	 * | meta 1| meta 2| meta 3|
	 * meta 1 is valid, meta 2 is invalid. meta 3 could be valid. If
	 * superblock points to meta 1, we write a new valid meta 2n.  if crash
	 * happens again, new recovery will start from meta 1. Since meta 2n is
	 * valid now, recovery will think meta 3 is valid, which is wrong.
	 * The solution is we create a new meta in meta2 with its seq == meta
	 * 1's seq + 10 and let superblock points to meta2. The same recovery will
	 * not think meta 3 is a valid meta, because its seq doesn't match
	 */
1099
	if (ctx.seq > log->last_cp_seq) {
S
Shaohua Li 已提交
1100 1101 1102 1103 1104 1105 1106 1107
		int ret;

		ret = r5l_log_write_empty_meta_block(log, ctx.pos, ctx.seq + 10);
		if (ret)
			return ret;
		log->seq = ctx.seq + 11;
		log->log_start = r5l_ring_add(log, ctx.pos, BLOCK_SECTORS);
		r5l_write_super(log, ctx.pos);
1108 1109
		log->last_checkpoint = ctx.pos;
		log->next_checkpoint = ctx.pos;
S
Shaohua Li 已提交
1110 1111 1112 1113
	} else {
		log->log_start = ctx.pos;
		log->seq = ctx.seq;
	}
S
Shaohua Li 已提交
1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141
	return 0;
}

static void r5l_write_super(struct r5l_log *log, sector_t cp)
{
	struct mddev *mddev = log->rdev->mddev;

	log->rdev->journal_tail = cp;
	set_bit(MD_CHANGE_DEVS, &mddev->flags);
}

static int r5l_load_log(struct r5l_log *log)
{
	struct md_rdev *rdev = log->rdev;
	struct page *page;
	struct r5l_meta_block *mb;
	sector_t cp = log->rdev->journal_tail;
	u32 stored_crc, expected_crc;
	bool create_super = false;
	int ret;

	/* Make sure it's valid */
	if (cp >= rdev->sectors || round_down(cp, BLOCK_SECTORS) != cp)
		cp = 0;
	page = alloc_page(GFP_KERNEL);
	if (!page)
		return -ENOMEM;

M
Mike Christie 已提交
1142
	if (!sync_page_io(rdev, cp, PAGE_SIZE, page, REQ_OP_READ, 0, false)) {
S
Shaohua Li 已提交
1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
		ret = -EIO;
		goto ioerr;
	}
	mb = page_address(page);

	if (le32_to_cpu(mb->magic) != R5LOG_MAGIC ||
	    mb->version != R5LOG_VERSION) {
		create_super = true;
		goto create;
	}
	stored_crc = le32_to_cpu(mb->checksum);
	mb->checksum = 0;
S
Shaohua Li 已提交
1155
	expected_crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167
	if (stored_crc != expected_crc) {
		create_super = true;
		goto create;
	}
	if (le64_to_cpu(mb->position) != cp) {
		create_super = true;
		goto create;
	}
create:
	if (create_super) {
		log->last_cp_seq = prandom_u32();
		cp = 0;
1168
		r5l_log_write_empty_meta_block(log, cp, log->last_cp_seq);
S
Shaohua Li 已提交
1169 1170 1171 1172 1173 1174 1175 1176 1177 1178
		/*
		 * Make sure super points to correct address. Log might have
		 * data very soon. If super hasn't correct log tail address,
		 * recovery can't find the log
		 */
		r5l_write_super(log, cp);
	} else
		log->last_cp_seq = le64_to_cpu(mb->seq);

	log->device_size = round_down(rdev->sectors, BLOCK_SECTORS);
S
Shaohua Li 已提交
1179 1180 1181
	log->max_free_space = log->device_size >> RECLAIM_MAX_FREE_SPACE_SHIFT;
	if (log->max_free_space > RECLAIM_MAX_FREE_SPACE)
		log->max_free_space = RECLAIM_MAX_FREE_SPACE;
S
Shaohua Li 已提交
1182
	log->last_checkpoint = cp;
1183
	log->next_checkpoint = cp;
S
Shaohua Li 已提交
1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194

	__free_page(page);

	return r5l_recovery_log(log);
ioerr:
	__free_page(page);
	return ret;
}

int r5l_init_log(struct r5conf *conf, struct md_rdev *rdev)
{
J
Jens Axboe 已提交
1195
	struct request_queue *q = bdev_get_queue(rdev->bdev);
S
Shaohua Li 已提交
1196 1197 1198 1199 1200 1201 1202 1203 1204
	struct r5l_log *log;

	if (PAGE_SIZE != 4096)
		return -EINVAL;
	log = kzalloc(sizeof(*log), GFP_KERNEL);
	if (!log)
		return -ENOMEM;
	log->rdev = rdev;

J
Jens Axboe 已提交
1205
	log->need_cache_flush = test_bit(QUEUE_FLAG_WC, &q->queue_flags) != 0;
1206

S
Shaohua Li 已提交
1207 1208
	log->uuid_checksum = crc32c_le(~0, rdev->mddev->uuid,
				       sizeof(rdev->mddev->uuid));
S
Shaohua Li 已提交
1209 1210 1211 1212 1213

	mutex_init(&log->io_mutex);

	spin_lock_init(&log->io_list_lock);
	INIT_LIST_HEAD(&log->running_ios);
S
Shaohua Li 已提交
1214
	INIT_LIST_HEAD(&log->io_end_ios);
1215
	INIT_LIST_HEAD(&log->flushing_ios);
1216
	INIT_LIST_HEAD(&log->finished_ios);
1217
	bio_init(&log->flush_bio);
S
Shaohua Li 已提交
1218 1219 1220 1221 1222

	log->io_kc = KMEM_CACHE(r5l_io_unit, 0);
	if (!log->io_kc)
		goto io_kc;

1223 1224 1225 1226
	log->io_pool = mempool_create_slab_pool(R5L_POOL_SIZE, log->io_kc);
	if (!log->io_pool)
		goto io_pool;

C
Christoph Hellwig 已提交
1227 1228 1229 1230
	log->bs = bioset_create(R5L_POOL_SIZE, 0);
	if (!log->bs)
		goto io_bs;

1231 1232 1233 1234
	log->meta_pool = mempool_create_page_pool(R5L_POOL_SIZE, 0);
	if (!log->meta_pool)
		goto out_mempool;

S
Shaohua Li 已提交
1235 1236 1237 1238
	log->reclaim_thread = md_register_thread(r5l_reclaim_thread,
						 log->rdev->mddev, "reclaim");
	if (!log->reclaim_thread)
		goto reclaim_thread;
1239
	init_waitqueue_head(&log->iounit_wait);
S
Shaohua Li 已提交
1240

1241 1242
	INIT_LIST_HEAD(&log->no_mem_stripes);

S
Shaohua Li 已提交
1243 1244 1245 1246 1247 1248
	INIT_LIST_HEAD(&log->no_space_stripes);
	spin_lock_init(&log->no_space_stripes_lock);

	if (r5l_load_log(log))
		goto error;

1249
	rcu_assign_pointer(conf->log, log);
1250
	set_bit(MD_HAS_JOURNAL, &conf->mddev->flags);
S
Shaohua Li 已提交
1251
	return 0;
1252

S
Shaohua Li 已提交
1253
error:
S
Shaohua Li 已提交
1254 1255
	md_unregister_thread(&log->reclaim_thread);
reclaim_thread:
1256 1257
	mempool_destroy(log->meta_pool);
out_mempool:
C
Christoph Hellwig 已提交
1258 1259
	bioset_free(log->bs);
io_bs:
1260 1261
	mempool_destroy(log->io_pool);
io_pool:
S
Shaohua Li 已提交
1262 1263 1264 1265 1266 1267 1268 1269
	kmem_cache_destroy(log->io_kc);
io_kc:
	kfree(log);
	return -EINVAL;
}

void r5l_exit_log(struct r5l_log *log)
{
S
Shaohua Li 已提交
1270
	md_unregister_thread(&log->reclaim_thread);
1271
	mempool_destroy(log->meta_pool);
C
Christoph Hellwig 已提交
1272
	bioset_free(log->bs);
1273
	mempool_destroy(log->io_pool);
S
Shaohua Li 已提交
1274 1275 1276
	kmem_cache_destroy(log->io_kc);
	kfree(log);
}