raid5-cache.c 71.9 KB
Newer Older
S
Shaohua Li 已提交
1 2
/*
 * Copyright (C) 2015 Shaohua Li <shli@fb.com>
S
Song Liu 已提交
3
 * Copyright (C) 2016 Song Liu <songliubraving@fb.com>
S
Shaohua Li 已提交
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
 *
 * 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
20
#include <linux/crc32c.h>
S
Shaohua Li 已提交
21 22 23
#include <linux/random.h>
#include "md.h"
#include "raid5.h"
S
Song Liu 已提交
24
#include "bitmap.h"
S
Shaohua Li 已提交
25 26 27 28 29 30 31

/*
 * 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
32
/*
33 34 35 36
 * log->max_free_space is min(1/4 disk size, 10G reclaimable space).
 *
 * In write through mode, the reclaim runs every log->max_free_space.
 * This can prevent the recovery scans for too long
S
Shaohua Li 已提交
37 38 39 40
 */
#define RECLAIM_MAX_FREE_SPACE (10 * 1024 * 1024 * 2) /* sector */
#define RECLAIM_MAX_FREE_SPACE_SHIFT (2)

41 42 43 44 45 46 47
/* wake up reclaim thread periodically */
#define R5C_RECLAIM_WAKEUP_INTERVAL (30 * HZ)
/* start flush with these full stripes */
#define R5C_FULL_STRIPE_FLUSH_BATCH 256
/* reclaim stripes in groups */
#define R5C_RECLAIM_STRIPE_GROUP (NR_STRIPE_HASH_LOCKS * 2)

C
Christoph Hellwig 已提交
48 49 50 51 52 53
/*
 * 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

54 55 56 57 58 59 60 61 62 63
/*
 * r5c journal modes of the array: write-back or write-through.
 * write-through mode has identical behavior as existing log only
 * implementation.
 */
enum r5c_journal_mode {
	R5C_JOURNAL_MODE_WRITE_THROUGH = 0,
	R5C_JOURNAL_MODE_WRITE_BACK = 1,
};

S
Song Liu 已提交
64 65
static char *r5c_journal_mode_str[] = {"write-through",
				       "write-back"};
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96
/*
 * raid5 cache state machine
 *
 * With rhe RAID cache, each stripe works in two phases:
 *	- caching phase
 *	- writing-out phase
 *
 * These two phases are controlled by bit STRIPE_R5C_CACHING:
 *   if STRIPE_R5C_CACHING == 0, the stripe is in writing-out phase
 *   if STRIPE_R5C_CACHING == 1, the stripe is in caching phase
 *
 * When there is no journal, or the journal is in write-through mode,
 * the stripe is always in writing-out phase.
 *
 * For write-back journal, the stripe is sent to caching phase on write
 * (r5c_try_caching_write). r5c_make_stripe_write_out() kicks off
 * the write-out phase by clearing STRIPE_R5C_CACHING.
 *
 * Stripes in caching phase do not write the raid disks. Instead, all
 * writes are committed from the log device. Therefore, a stripe in
 * caching phase handles writes as:
 *	- write to log device
 *	- return IO
 *
 * Stripes in writing-out phase handle writes as:
 *	- calculate parity
 *	- write pending data and parity to journal
 *	- write data and parity to raid disks
 *	- return IO for pending writes
 */

S
Shaohua Li 已提交
97 98 99 100 101 102 103
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
104 105
	sector_t max_free_space;	/* reclaim run if free space is at
					 * this size */
S
Shaohua Li 已提交
106 107 108 109 110 111 112 113

	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 */

114 115 116
	sector_t next_checkpoint;
	u64 next_cp_seq;

S
Shaohua Li 已提交
117 118 119 120 121 122 123 124 125 126
	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 */
127 128
	struct list_head flushing_ios;	/* io_units which are waiting for log
					 * cache flush */
129
	struct list_head finished_ios;	/* io_units which settle down in log disk */
130
	struct bio flush_bio;
S
Shaohua Li 已提交
131

132 133
	struct list_head no_mem_stripes;   /* pending stripes, -ENOMEM */

S
Shaohua Li 已提交
134
	struct kmem_cache *io_kc;
135
	mempool_t *io_pool;
C
Christoph Hellwig 已提交
136
	struct bio_set *bs;
137
	mempool_t *meta_pool;
S
Shaohua Li 已提交
138

S
Shaohua Li 已提交
139 140 141 142 143 144 145 146
	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) */
147
	wait_queue_head_t iounit_wait;
S
Shaohua Li 已提交
148

S
Shaohua Li 已提交
149 150
	struct list_head no_space_stripes; /* pending stripes, log has no space */
	spinlock_t no_space_stripes_lock;
151 152

	bool need_cache_flush;
153 154 155

	/* for r5c_cache */
	enum r5c_journal_mode r5c_journal_mode;
156 157 158 159 160 161

	/* all stripes in r5cache, in the order of seq at sh->log_start */
	struct list_head stripe_in_journal_list;

	spinlock_t stripe_in_journal_lock;
	atomic_t stripe_in_journal_count;
S
Song Liu 已提交
162 163 164

	/* to submit async io_units, to fulfill ordering of flush */
	struct work_struct deferred_io_work;
S
Shaohua Li 已提交
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189
};

/*
 * 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
190
	bool need_split_bio;
S
Song Liu 已提交
191 192 193 194 195 196 197 198 199 200 201 202
	struct bio *split_bio;

	unsigned int has_flush:1;      /* include flush request */
	unsigned int has_fua:1;        /* include fua request */
	unsigned int has_null_flush:1; /* include empty flush request */
	/*
	 * io isn't sent yet, flush/fua request can only be submitted till it's
	 * the first IO in running_ios list
	 */
	unsigned int io_deferred:1;

	struct bio_list flush_barriers;   /* size == 0 flush bios */
S
Shaohua Li 已提交
203 204 205 206 207 208 209 210
};

/* 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 */
211
	IO_UNIT_STRIPE_END = 3,	/* stripes data finished writing to raid */
S
Shaohua Li 已提交
212 213
};

214 215 216 217 218 219
bool r5c_is_writeback(struct r5l_log *log)
{
	return (log != NULL &&
		log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_BACK);
}

S
Shaohua Li 已提交
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254
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;
}

S
Song Liu 已提交
255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291
static void
r5c_return_dev_pending_writes(struct r5conf *conf, struct r5dev *dev,
			      struct bio_list *return_bi)
{
	struct bio *wbi, *wbi2;

	wbi = dev->written;
	dev->written = NULL;
	while (wbi && wbi->bi_iter.bi_sector <
	       dev->sector + STRIPE_SECTORS) {
		wbi2 = r5_next_bio(wbi, dev->sector);
		if (!raid5_dec_bi_active_stripes(wbi)) {
			md_write_end(conf->mddev);
			bio_list_add(return_bi, wbi);
		}
		wbi = wbi2;
	}
}

void r5c_handle_cached_data_endio(struct r5conf *conf,
	  struct stripe_head *sh, int disks, struct bio_list *return_bi)
{
	int i;

	for (i = sh->disks; i--; ) {
		if (sh->dev[i].written) {
			set_bit(R5_UPTODATE, &sh->dev[i].flags);
			r5c_return_dev_pending_writes(conf, &sh->dev[i],
						      return_bi);
			bitmap_endwrite(conf->mddev->bitmap, sh->sector,
					STRIPE_SECTORS,
					!test_bit(STRIPE_DEGRADED, &sh->state),
					0);
		}
	}
}

292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389
/* Check whether we should flush some stripes to free up stripe cache */
void r5c_check_stripe_cache_usage(struct r5conf *conf)
{
	int total_cached;

	if (!r5c_is_writeback(conf->log))
		return;

	total_cached = atomic_read(&conf->r5c_cached_partial_stripes) +
		atomic_read(&conf->r5c_cached_full_stripes);

	/*
	 * The following condition is true for either of the following:
	 *   - stripe cache pressure high:
	 *          total_cached > 3/4 min_nr_stripes ||
	 *          empty_inactive_list_nr > 0
	 *   - stripe cache pressure moderate:
	 *          total_cached > 1/2 min_nr_stripes
	 */
	if (total_cached > conf->min_nr_stripes * 1 / 2 ||
	    atomic_read(&conf->empty_inactive_list_nr) > 0)
		r5l_wake_reclaim(conf->log, 0);
}

/*
 * flush cache when there are R5C_FULL_STRIPE_FLUSH_BATCH or more full
 * stripes in the cache
 */
void r5c_check_cached_full_stripe(struct r5conf *conf)
{
	if (!r5c_is_writeback(conf->log))
		return;

	/*
	 * wake up reclaim for R5C_FULL_STRIPE_FLUSH_BATCH cached stripes
	 * or a full stripe (chunk size / 4k stripes).
	 */
	if (atomic_read(&conf->r5c_cached_full_stripes) >=
	    min(R5C_FULL_STRIPE_FLUSH_BATCH,
		conf->chunk_sectors >> STRIPE_SHIFT))
		r5l_wake_reclaim(conf->log, 0);
}

/*
 * Total log space (in sectors) needed to flush all data in cache
 *
 * Currently, writing-out phase automatically includes all pending writes
 * to the same sector. So the reclaim of each stripe takes up to
 * (conf->raid_disks + 1) pages of log space.
 *
 * To totally avoid deadlock due to log space, the code reserves
 * (conf->raid_disks + 1) pages for each stripe in cache, which is not
 * necessary in most cases.
 *
 * To improve this, we will need writing-out phase to be able to NOT include
 * pending writes, which will reduce the requirement to
 * (conf->max_degraded + 1) pages per stripe in cache.
 */
static sector_t r5c_log_required_to_flush_cache(struct r5conf *conf)
{
	struct r5l_log *log = conf->log;

	if (!r5c_is_writeback(log))
		return 0;

	return BLOCK_SECTORS * (conf->raid_disks + 1) *
		atomic_read(&log->stripe_in_journal_count);
}

/*
 * evaluate log space usage and update R5C_LOG_TIGHT and R5C_LOG_CRITICAL
 *
 * R5C_LOG_TIGHT is set when free space on the log device is less than 3x of
 * reclaim_required_space. R5C_LOG_CRITICAL is set when free space on the log
 * device is less than 2x of reclaim_required_space.
 */
static inline void r5c_update_log_state(struct r5l_log *log)
{
	struct r5conf *conf = log->rdev->mddev->private;
	sector_t free_space;
	sector_t reclaim_space;

	if (!r5c_is_writeback(log))
		return;

	free_space = r5l_ring_distance(log, log->log_start,
				       log->last_checkpoint);
	reclaim_space = r5c_log_required_to_flush_cache(conf);
	if (free_space < 2 * reclaim_space)
		set_bit(R5C_LOG_CRITICAL, &conf->cache_state);
	else
		clear_bit(R5C_LOG_CRITICAL, &conf->cache_state);
	if (free_space < 3 * reclaim_space)
		set_bit(R5C_LOG_TIGHT, &conf->cache_state);
	else
		clear_bit(R5C_LOG_TIGHT, &conf->cache_state);
}

390 391 392 393
/*
 * Put the stripe into writing-out phase by clearing STRIPE_R5C_CACHING.
 * This function should only be called in write-back mode.
 */
394
void r5c_make_stripe_write_out(struct stripe_head *sh)
395 396 397 398 399 400 401 402
{
	struct r5conf *conf = sh->raid_conf;
	struct r5l_log *log = conf->log;

	BUG_ON(!r5c_is_writeback(log));

	WARN_ON(!test_bit(STRIPE_R5C_CACHING, &sh->state));
	clear_bit(STRIPE_R5C_CACHING, &sh->state);
S
Song Liu 已提交
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

	if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
		atomic_inc(&conf->preread_active_stripes);

	if (test_and_clear_bit(STRIPE_R5C_PARTIAL_STRIPE, &sh->state)) {
		BUG_ON(atomic_read(&conf->r5c_cached_partial_stripes) == 0);
		atomic_dec(&conf->r5c_cached_partial_stripes);
	}

	if (test_and_clear_bit(STRIPE_R5C_FULL_STRIPE, &sh->state)) {
		BUG_ON(atomic_read(&conf->r5c_cached_full_stripes) == 0);
		atomic_dec(&conf->r5c_cached_full_stripes);
	}
}

static void r5c_handle_data_cached(struct stripe_head *sh)
{
	int i;

	for (i = sh->disks; i--; )
		if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags)) {
			set_bit(R5_InJournal, &sh->dev[i].flags);
			clear_bit(R5_LOCKED, &sh->dev[i].flags);
		}
	clear_bit(STRIPE_LOG_TRAPPED, &sh->state);
}

/*
 * this journal write must contain full parity,
 * it may also contain some data pages
 */
static void r5c_handle_parity_cached(struct stripe_head *sh)
{
	int i;

	for (i = sh->disks; i--; )
		if (test_bit(R5_InJournal, &sh->dev[i].flags))
			set_bit(R5_Wantwrite, &sh->dev[i].flags);
441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459
}

/*
 * Setting proper flags after writing (or flushing) data and/or parity to the
 * log device. This is called from r5l_log_endio() or r5l_log_flush_endio().
 */
static void r5c_finish_cache_stripe(struct stripe_head *sh)
{
	struct r5l_log *log = sh->raid_conf->log;

	if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) {
		BUG_ON(test_bit(STRIPE_R5C_CACHING, &sh->state));
		/*
		 * Set R5_InJournal for parity dev[pd_idx]. This means
		 * all data AND parity in the journal. For RAID 6, it is
		 * NOT necessary to set the flag for dev[qd_idx], as the
		 * two parities are written out together.
		 */
		set_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags);
S
Song Liu 已提交
460 461 462 463 464 465
	} else if (test_bit(STRIPE_R5C_CACHING, &sh->state)) {
		r5c_handle_data_cached(sh);
	} else {
		r5c_handle_parity_cached(sh);
		set_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags);
	}
466 467
}

468 469 470 471 472 473
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);
474 475 476

		r5c_finish_cache_stripe(sh);

477 478 479 480 481
		set_bit(STRIPE_HANDLE, &sh->state);
		raid5_release_stripe(sh);
	}
}

482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497
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);
	}
}

498 499 500 501 502 503 504 505 506 507 508 509 510 511
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
Song Liu 已提交
512
static void __r5l_stripe_write_finished(struct r5l_io_unit *io);
S
Shaohua Li 已提交
513 514 515
static void r5l_log_endio(struct bio *bio)
{
	struct r5l_io_unit *io = bio->bi_private;
S
Song Liu 已提交
516
	struct r5l_io_unit *io_deferred;
S
Shaohua Li 已提交
517
	struct r5l_log *log = io->log;
518
	unsigned long flags;
S
Shaohua Li 已提交
519

S
Shaohua Li 已提交
520 521 522
	if (bio->bi_error)
		md_error(log->rdev->mddev, log->rdev);

S
Shaohua Li 已提交
523
	bio_put(bio);
524
	mempool_free(io->meta_page, log->meta_pool);
S
Shaohua Li 已提交
525

526 527
	spin_lock_irqsave(&log->io_list_lock, flags);
	__r5l_set_io_unit_state(io, IO_UNIT_IO_END);
528
	if (log->need_cache_flush)
529
		r5l_move_to_end_ios(log);
530 531
	else
		r5l_log_run_stripes(log);
S
Song Liu 已提交
532 533 534 535 536 537 538 539 540 541 542
	if (!list_empty(&log->running_ios)) {
		/*
		 * FLUSH/FUA io_unit is deferred because of ordering, now we
		 * can dispatch it
		 */
		io_deferred = list_first_entry(&log->running_ios,
					       struct r5l_io_unit, log_sibling);
		if (io_deferred->io_deferred)
			schedule_work(&log->deferred_io_work);
	}

543 544
	spin_unlock_irqrestore(&log->io_list_lock, flags);

545 546
	if (log->need_cache_flush)
		md_wakeup_thread(log->rdev->mddev->thread);
S
Song Liu 已提交
547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604

	if (io->has_null_flush) {
		struct bio *bi;

		WARN_ON(bio_list_empty(&io->flush_barriers));
		while ((bi = bio_list_pop(&io->flush_barriers)) != NULL) {
			bio_endio(bi);
			atomic_dec(&io->pending_stripe);
		}
		if (atomic_read(&io->pending_stripe) == 0)
			__r5l_stripe_write_finished(io);
	}
}

static void r5l_do_submit_io(struct r5l_log *log, struct r5l_io_unit *io)
{
	unsigned long flags;

	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);

	if (io->has_flush)
		bio_set_op_attrs(io->current_bio, REQ_OP_WRITE, WRITE_FLUSH);
	if (io->has_fua)
		bio_set_op_attrs(io->current_bio, REQ_OP_WRITE, WRITE_FUA);
	submit_bio(io->current_bio);

	if (!io->split_bio)
		return;

	if (io->has_flush)
		bio_set_op_attrs(io->split_bio, REQ_OP_WRITE, WRITE_FLUSH);
	if (io->has_fua)
		bio_set_op_attrs(io->split_bio, REQ_OP_WRITE, WRITE_FUA);
	submit_bio(io->split_bio);
}

/* deferred io_unit will be dispatched here */
static void r5l_submit_io_async(struct work_struct *work)
{
	struct r5l_log *log = container_of(work, struct r5l_log,
					   deferred_io_work);
	struct r5l_io_unit *io = NULL;
	unsigned long flags;

	spin_lock_irqsave(&log->io_list_lock, flags);
	if (!list_empty(&log->running_ios)) {
		io = list_first_entry(&log->running_ios, struct r5l_io_unit,
				      log_sibling);
		if (!io->io_deferred)
			io = NULL;
		else
			io->io_deferred = 0;
	}
	spin_unlock_irqrestore(&log->io_list_lock, flags);
	if (io)
		r5l_do_submit_io(log, io);
S
Shaohua Li 已提交
605 606 607 608 609
}

static void r5l_submit_current_io(struct r5l_log *log)
{
	struct r5l_io_unit *io = log->current_io;
S
Song Liu 已提交
610
	struct bio *bio;
S
Shaohua Li 已提交
611
	struct r5l_meta_block *block;
612
	unsigned long flags;
S
Shaohua Li 已提交
613
	u32 crc;
S
Song Liu 已提交
614
	bool do_submit = true;
S
Shaohua Li 已提交
615 616 617 618 619 620

	if (!io)
		return;

	block = page_address(io->meta_page);
	block->meta_size = cpu_to_le32(io->meta_offset);
S
Shaohua Li 已提交
621
	crc = crc32c_le(log->uuid_checksum, block, PAGE_SIZE);
S
Shaohua Li 已提交
622
	block->checksum = cpu_to_le32(crc);
S
Song Liu 已提交
623
	bio = io->current_bio;
S
Shaohua Li 已提交
624 625

	log->current_io = NULL;
626
	spin_lock_irqsave(&log->io_list_lock, flags);
S
Song Liu 已提交
627 628 629 630 631 632 633
	if (io->has_flush || io->has_fua) {
		if (io != list_first_entry(&log->running_ios,
					   struct r5l_io_unit, log_sibling)) {
			io->io_deferred = 1;
			do_submit = false;
		}
	}
634
	spin_unlock_irqrestore(&log->io_list_lock, flags);
S
Song Liu 已提交
635 636
	if (do_submit)
		r5l_do_submit_io(log, io);
S
Shaohua Li 已提交
637 638
}

C
Christoph Hellwig 已提交
639
static struct bio *r5l_bio_alloc(struct r5l_log *log)
640
{
C
Christoph Hellwig 已提交
641
	struct bio *bio = bio_alloc_bioset(GFP_NOIO, BIO_MAX_PAGES, log->bs);
642

M
Mike Christie 已提交
643
	bio_set_op_attrs(bio, REQ_OP_WRITE, 0);
644
	bio->bi_bdev = log->rdev->bdev;
645
	bio->bi_iter.bi_sector = log->rdev->data_offset + log->log_start;
646 647 648 649

	return bio;
}

650 651 652 653
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);

654
	r5c_update_log_state(log);
655 656 657 658 659 660 661 662
	/*
	 * 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
663
		io->need_split_bio = true;
664 665 666 667

	io->log_end = log->log_start;
}

S
Shaohua Li 已提交
668 669 670 671 672
static struct r5l_io_unit *r5l_new_meta(struct r5l_log *log)
{
	struct r5l_io_unit *io;
	struct r5l_meta_block *block;

673 674 675 676 677
	io = mempool_alloc(log->io_pool, GFP_ATOMIC);
	if (!io)
		return NULL;
	memset(io, 0, sizeof(*io));

678 679 680
	io->log = log;
	INIT_LIST_HEAD(&io->log_sibling);
	INIT_LIST_HEAD(&io->stripe_list);
S
Song Liu 已提交
681
	bio_list_init(&io->flush_barriers);
682
	io->state = IO_UNIT_RUNNING;
S
Shaohua Li 已提交
683

684
	io->meta_page = mempool_alloc(log->meta_pool, GFP_NOIO);
S
Shaohua Li 已提交
685
	block = page_address(io->meta_page);
686
	clear_page(block);
S
Shaohua Li 已提交
687 688 689 690 691 692 693
	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);
694
	io->seq = log->seq++;
S
Shaohua Li 已提交
695

C
Christoph Hellwig 已提交
696 697 698
	io->current_bio = r5l_bio_alloc(log);
	io->current_bio->bi_end_io = r5l_log_endio;
	io->current_bio->bi_private = io;
699
	bio_add_page(io->current_bio, io->meta_page, PAGE_SIZE, 0);
S
Shaohua Li 已提交
700

701
	r5_reserve_log_entry(log, io);
S
Shaohua Li 已提交
702 703 704 705 706 707 708 709 710 711

	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)
{
712 713
	if (log->current_io &&
	    log->current_io->meta_offset + payload_size > PAGE_SIZE)
S
Shaohua Li 已提交
714 715
		r5l_submit_current_io(log);

716
	if (!log->current_io) {
717
		log->current_io = r5l_new_meta(log);
718 719 720 721
		if (!log->current_io)
			return -ENOMEM;
	}

S
Shaohua Li 已提交
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
	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
751
	if (io->need_split_bio) {
S
Song Liu 已提交
752 753
		BUG_ON(io->split_bio);
		io->split_bio = io->current_bio;
C
Christoph Hellwig 已提交
754
		io->current_bio = r5l_bio_alloc(log);
S
Song Liu 已提交
755 756
		bio_chain(io->current_bio, io->split_bio);
		io->need_split_bio = false;
S
Shaohua Li 已提交
757 758
	}

C
Christoph Hellwig 已提交
759 760 761
	if (!bio_add_page(io->current_bio, page, PAGE_SIZE, 0))
		BUG();

762
	r5_reserve_log_entry(log, io);
S
Shaohua Li 已提交
763 764
}

765
static int r5l_log_stripe(struct r5l_log *log, struct stripe_head *sh,
S
Shaohua Li 已提交
766 767 768 769
			   int data_pages, int parity_pages)
{
	int i;
	int meta_size;
770
	int ret;
S
Shaohua Li 已提交
771 772 773 774 775 776 777 778
	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;

779 780 781 782
	ret = r5l_get_meta(log, meta_size);
	if (ret)
		return ret;

S
Shaohua Li 已提交
783 784
	io = log->current_io;

S
Song Liu 已提交
785 786 787
	if (test_and_clear_bit(STRIPE_R5C_PREFLUSH, &sh->state))
		io->has_flush = 1;

S
Shaohua Li 已提交
788
	for (i = 0; i < sh->disks; i++) {
S
Song Liu 已提交
789 790
		if (!test_bit(R5_Wantwrite, &sh->dev[i].flags) ||
		    test_bit(R5_InJournal, &sh->dev[i].flags))
S
Shaohua Li 已提交
791 792 793
			continue;
		if (i == sh->pd_idx || i == sh->qd_idx)
			continue;
S
Song Liu 已提交
794 795 796 797 798 799 800 801 802
		if (test_bit(R5_WantFUA, &sh->dev[i].flags) &&
		    log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_BACK) {
			io->has_fua = 1;
			/*
			 * we need to flush journal to make sure recovery can
			 * reach the data with fua flag
			 */
			io->has_flush = 1;
		}
S
Shaohua Li 已提交
803 804 805 806 807 808
		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);
	}

809
	if (parity_pages == 2) {
S
Shaohua Li 已提交
810 811 812 813 814
		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);
815
	} else if (parity_pages == 1) {
S
Shaohua Li 已提交
816 817 818 819
		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);
820 821
	} else  /* Just writing data, not parity, in caching phase */
		BUG_ON(parity_pages != 0);
S
Shaohua Li 已提交
822 823 824 825

	list_add_tail(&sh->log_list, &io->stripe_list);
	atomic_inc(&io->pending_stripe);
	sh->log_io = io;
826

827 828 829 830 831 832 833 834 835 836 837 838
	if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH)
		return 0;

	if (sh->log_start == MaxSector) {
		BUG_ON(!list_empty(&sh->r5c));
		sh->log_start = io->log_start;
		spin_lock_irq(&log->stripe_in_journal_lock);
		list_add_tail(&sh->r5c,
			      &log->stripe_in_journal_list);
		spin_unlock_irq(&log->stripe_in_journal_lock);
		atomic_inc(&log->stripe_in_journal_count);
	}
839
	return 0;
S
Shaohua Li 已提交
840 841
}

842 843 844 845 846 847 848 849 850
/* add stripe to no_space_stripes, and then wake up reclaim */
static inline void r5l_add_no_space_stripe(struct r5l_log *log,
					   struct stripe_head *sh)
{
	spin_lock(&log->no_space_stripes_lock);
	list_add_tail(&sh->log_list, &log->no_space_stripes);
	spin_unlock(&log->no_space_stripes_lock);
}

S
Shaohua Li 已提交
851 852 853 854 855 856
/*
 * 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)
{
857
	struct r5conf *conf = sh->raid_conf;
S
Shaohua Li 已提交
858 859 860 861
	int write_disks = 0;
	int data_pages, parity_pages;
	int reserve;
	int i;
862
	int ret = 0;
863
	bool wake_reclaim = false;
S
Shaohua Li 已提交
864 865 866 867 868 869 870 871 872 873 874

	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;
	}

875 876
	WARN_ON(test_bit(STRIPE_R5C_CACHING, &sh->state));

S
Shaohua Li 已提交
877 878 879
	for (i = 0; i < sh->disks; i++) {
		void *addr;

S
Song Liu 已提交
880 881
		if (!test_bit(R5_Wantwrite, &sh->dev[i].flags) ||
		    test_bit(R5_InJournal, &sh->dev[i].flags))
S
Shaohua Li 已提交
882
			continue;
S
Song Liu 已提交
883

S
Shaohua Li 已提交
884 885 886 887 888
		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
889 890
		sh->dev[i].log_checksum = crc32c_le(log->uuid_checksum,
						    addr, PAGE_SIZE);
S
Shaohua Li 已提交
891 892 893 894 895 896
		kunmap_atomic(addr);
	}
	parity_pages = 1 + !!(sh->qd_idx >= 0);
	data_pages = write_disks - parity_pages;

	set_bit(STRIPE_LOG_TRAPPED, &sh->state);
897 898 899 900 901
	/*
	 * 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
902 903 904 905 906 907
	atomic_inc(&sh->count);

	mutex_lock(&log->io_mutex);
	/* meta + data */
	reserve = (1 + write_disks) << (PAGE_SHIFT - 9);

908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943
	if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) {
		if (!r5l_has_free_space(log, reserve)) {
			r5l_add_no_space_stripe(log, sh);
			wake_reclaim = true;
		} 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);
			}
		}
	} else {  /* R5C_JOURNAL_MODE_WRITE_BACK */
		/*
		 * log space critical, do not process stripes that are
		 * not in cache yet (sh->log_start == MaxSector).
		 */
		if (test_bit(R5C_LOG_CRITICAL, &conf->cache_state) &&
		    sh->log_start == MaxSector) {
			r5l_add_no_space_stripe(log, sh);
			wake_reclaim = true;
			reserve = 0;
		} else if (!r5l_has_free_space(log, reserve)) {
			if (sh->log_start == log->last_checkpoint)
				BUG();
			else
				r5l_add_no_space_stripe(log, sh);
		} 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);
			}
944
		}
S
Shaohua Li 已提交
945 946
	}

947
	mutex_unlock(&log->io_mutex);
948 949
	if (wake_reclaim)
		r5l_wake_reclaim(log, reserve);
S
Shaohua Li 已提交
950 951 952 953 954 955 956 957 958 959 960 961
	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);
}

962 963 964 965
int r5l_handle_flush_request(struct r5l_log *log, struct bio *bio)
{
	if (!log)
		return -ENODEV;
S
Song Liu 已提交
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992

	if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) {
		/*
		 * in write through (journal only)
		 * 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;
		}
		bio->bi_opf &= ~REQ_PREFLUSH;
	} else {
		/* write back (with cache) */
		if (bio->bi_iter.bi_size == 0) {
			mutex_lock(&log->io_mutex);
			r5l_get_meta(log, 0);
			bio_list_add(&log->current_io->flush_barriers, bio);
			log->current_io->has_flush = 1;
			log->current_io->has_null_flush = 1;
			atomic_inc(&log->current_io->pending_stripe);
			r5l_submit_current_io(log);
			mutex_unlock(&log->io_mutex);
			return 0;
		}
993 994 995 996
	}
	return -EAGAIN;
}

S
Shaohua Li 已提交
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
/* 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);
}

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
/*
 * calculate new last_checkpoint
 * for write through mode, returns log->next_checkpoint
 * for write back, returns log_start of first sh in stripe_in_journal_list
 */
static sector_t r5c_calculate_new_cp(struct r5conf *conf)
{
	struct stripe_head *sh;
	struct r5l_log *log = conf->log;
	sector_t new_cp;
	unsigned long flags;

	if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH)
		return log->next_checkpoint;

	spin_lock_irqsave(&log->stripe_in_journal_lock, flags);
	if (list_empty(&conf->log->stripe_in_journal_list)) {
		/* all stripes flushed */
		spin_unlock(&log->stripe_in_journal_lock);
		return log->next_checkpoint;
	}
	sh = list_first_entry(&conf->log->stripe_in_journal_list,
			      struct stripe_head, r5c);
	new_cp = sh->log_start;
	spin_unlock_irqrestore(&log->stripe_in_journal_lock, flags);
	return new_cp;
}

1041 1042
static sector_t r5l_reclaimable_space(struct r5l_log *log)
{
1043 1044
	struct r5conf *conf = log->rdev->mddev->private;

1045
	return r5l_ring_distance(log, log->last_checkpoint,
1046
				 r5c_calculate_new_cp(conf));
1047 1048
}

1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063
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);
	}
}

1064
static bool r5l_complete_finished_ios(struct r5l_log *log)
1065 1066 1067 1068 1069 1070
{
	struct r5l_io_unit *io, *next;
	bool found = false;

	assert_spin_locked(&log->io_list_lock);

1071
	list_for_each_entry_safe(io, next, &log->finished_ios, log_sibling) {
1072 1073 1074 1075 1076 1077 1078 1079
		/* 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);
1080 1081
		mempool_free(io, log->io_pool);
		r5l_run_no_mem_stripe(log);
1082 1083 1084 1085 1086 1087 1088

		found = true;
	}

	return found;
}

1089 1090 1091
static void __r5l_stripe_write_finished(struct r5l_io_unit *io)
{
	struct r5l_log *log = io->log;
1092
	struct r5conf *conf = log->rdev->mddev->private;
1093 1094 1095 1096
	unsigned long flags;

	spin_lock_irqsave(&log->io_list_lock, flags);
	__r5l_set_io_unit_state(io, IO_UNIT_STRIPE_END);
1097

1098
	if (!r5l_complete_finished_ios(log)) {
1099 1100 1101
		spin_unlock_irqrestore(&log->io_list_lock, flags);
		return;
	}
1102

1103 1104
	if (r5l_reclaimable_space(log) > log->max_free_space ||
	    test_bit(R5C_LOG_TIGHT, &conf->cache_state))
1105 1106 1107 1108 1109 1110
		r5l_wake_reclaim(log, 0);

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

S
Shaohua Li 已提交
1111 1112 1113 1114 1115 1116 1117
void r5l_stripe_write_finished(struct stripe_head *sh)
{
	struct r5l_io_unit *io;

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

1118 1119
	if (io && atomic_dec_and_test(&io->pending_stripe))
		__r5l_stripe_write_finished(io);
S
Shaohua Li 已提交
1120 1121
}

1122 1123 1124 1125 1126 1127 1128
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
1129 1130 1131
	if (bio->bi_error)
		md_error(log->rdev->mddev, log->rdev);

1132
	spin_lock_irqsave(&log->io_list_lock, flags);
1133 1134
	list_for_each_entry(io, &log->flushing_ios, log_sibling)
		r5l_io_run_stripes(io);
1135
	list_splice_tail_init(&log->flushing_ios, &log->finished_ios);
1136 1137 1138
	spin_unlock_irqrestore(&log->io_list_lock, flags);
}

S
Shaohua Li 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154
/*
 * 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)
{
1155
	bool do_flush;
1156 1157

	if (!log || !log->need_cache_flush)
S
Shaohua Li 已提交
1158 1159 1160
		return;

	spin_lock_irq(&log->io_list_lock);
1161 1162 1163 1164
	/* flush bio is running */
	if (!list_empty(&log->flushing_ios)) {
		spin_unlock_irq(&log->io_list_lock);
		return;
S
Shaohua Li 已提交
1165
	}
1166 1167
	list_splice_tail_init(&log->io_end_ios, &log->flushing_ios);
	do_flush = !list_empty(&log->flushing_ios);
S
Shaohua Li 已提交
1168
	spin_unlock_irq(&log->io_list_lock);
1169 1170 1171 1172 1173 1174

	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
1175
	bio_set_op_attrs(&log->flush_bio, REQ_OP_WRITE, WRITE_FLUSH);
1176
	submit_bio(&log->flush_bio);
S
Shaohua Li 已提交
1177 1178 1179
}

static void r5l_write_super(struct r5l_log *log, sector_t cp);
S
Shaohua Li 已提交
1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192
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;
	/*
1193 1194 1195 1196 1197 1198 1199 1200 1201
	 * 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
1202
	 */
1203 1204 1205 1206 1207 1208
	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
1209

S
Shaohua Li 已提交
1210
	/* discard IO error really doesn't matter, ignore it */
S
Shaohua Li 已提交
1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
	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);
	}
}

1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 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 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353
/*
 * r5c_flush_stripe moves stripe from cached list to handle_list. When called,
 * the stripe must be on r5c_cached_full_stripes or r5c_cached_partial_stripes.
 *
 * must hold conf->device_lock
 */
static void r5c_flush_stripe(struct r5conf *conf, struct stripe_head *sh)
{
	BUG_ON(list_empty(&sh->lru));
	BUG_ON(!test_bit(STRIPE_R5C_CACHING, &sh->state));
	BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));

	/*
	 * The stripe is not ON_RELEASE_LIST, so it is safe to call
	 * raid5_release_stripe() while holding conf->device_lock
	 */
	BUG_ON(test_bit(STRIPE_ON_RELEASE_LIST, &sh->state));
	assert_spin_locked(&conf->device_lock);

	list_del_init(&sh->lru);
	atomic_inc(&sh->count);

	set_bit(STRIPE_HANDLE, &sh->state);
	atomic_inc(&conf->active_stripes);
	r5c_make_stripe_write_out(sh);

	if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
		atomic_inc(&conf->preread_active_stripes);
	raid5_release_stripe(sh);
}

/*
 * if num == 0, flush all full stripes
 * if num > 0, flush all full stripes. If less than num full stripes are
 *             flushed, flush some partial stripes until totally num stripes are
 *             flushed or there is no more cached stripes.
 */
void r5c_flush_cache(struct r5conf *conf, int num)
{
	int count;
	struct stripe_head *sh, *next;

	assert_spin_locked(&conf->device_lock);
	if (!conf->log)
		return;

	count = 0;
	list_for_each_entry_safe(sh, next, &conf->r5c_full_stripe_list, lru) {
		r5c_flush_stripe(conf, sh);
		count++;
	}

	if (count >= num)
		return;
	list_for_each_entry_safe(sh, next,
				 &conf->r5c_partial_stripe_list, lru) {
		r5c_flush_stripe(conf, sh);
		if (++count >= num)
			break;
	}
}

static void r5c_do_reclaim(struct r5conf *conf)
{
	struct r5l_log *log = conf->log;
	struct stripe_head *sh;
	int count = 0;
	unsigned long flags;
	int total_cached;
	int stripes_to_flush;

	if (!r5c_is_writeback(log))
		return;

	total_cached = atomic_read(&conf->r5c_cached_partial_stripes) +
		atomic_read(&conf->r5c_cached_full_stripes);

	if (total_cached > conf->min_nr_stripes * 3 / 4 ||
	    atomic_read(&conf->empty_inactive_list_nr) > 0)
		/*
		 * if stripe cache pressure high, flush all full stripes and
		 * some partial stripes
		 */
		stripes_to_flush = R5C_RECLAIM_STRIPE_GROUP;
	else if (total_cached > conf->min_nr_stripes * 1 / 2 ||
		 atomic_read(&conf->r5c_cached_full_stripes) >
		 R5C_FULL_STRIPE_FLUSH_BATCH)
		/*
		 * if stripe cache pressure moderate, or if there is many full
		 * stripes,flush all full stripes
		 */
		stripes_to_flush = 0;
	else
		/* no need to flush */
		stripes_to_flush = -1;

	if (stripes_to_flush >= 0) {
		spin_lock_irqsave(&conf->device_lock, flags);
		r5c_flush_cache(conf, stripes_to_flush);
		spin_unlock_irqrestore(&conf->device_lock, flags);
	}

	/* if log space is tight, flush stripes on stripe_in_journal_list */
	if (test_bit(R5C_LOG_TIGHT, &conf->cache_state)) {
		spin_lock_irqsave(&log->stripe_in_journal_lock, flags);
		spin_lock(&conf->device_lock);
		list_for_each_entry(sh, &log->stripe_in_journal_list, r5c) {
			/*
			 * stripes on stripe_in_journal_list could be in any
			 * state of the stripe_cache state machine. In this
			 * case, we only want to flush stripe on
			 * r5c_cached_full/partial_stripes. The following
			 * condition makes sure the stripe is on one of the
			 * two lists.
			 */
			if (!list_empty(&sh->lru) &&
			    !test_bit(STRIPE_HANDLE, &sh->state) &&
			    atomic_read(&sh->count) == 0) {
				r5c_flush_stripe(conf, sh);
			}
			if (count++ >= R5C_RECLAIM_STRIPE_GROUP)
				break;
		}
		spin_unlock(&conf->device_lock);
		spin_unlock_irqrestore(&log->stripe_in_journal_lock, flags);
	}
	md_wakeup_thread(conf->mddev->thread);
}

S
Shaohua Li 已提交
1354 1355
static void r5l_do_reclaim(struct r5l_log *log)
{
1356
	struct r5conf *conf = log->rdev->mddev->private;
S
Shaohua Li 已提交
1357
	sector_t reclaim_target = xchg(&log->reclaim_target, 0);
1358 1359
	sector_t reclaimable;
	sector_t next_checkpoint;
1360
	bool write_super;
S
Shaohua Li 已提交
1361 1362

	spin_lock_irq(&log->io_list_lock);
1363 1364
	write_super = r5l_reclaimable_space(log) > log->max_free_space ||
		reclaim_target != 0 || !list_empty(&log->no_space_stripes);
S
Shaohua Li 已提交
1365 1366 1367 1368 1369 1370
	/*
	 * 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) {
1371 1372
		reclaimable = r5l_reclaimable_space(log);
		if (reclaimable >= reclaim_target ||
S
Shaohua Li 已提交
1373 1374
		    (list_empty(&log->running_ios) &&
		     list_empty(&log->io_end_ios) &&
1375
		     list_empty(&log->flushing_ios) &&
1376
		     list_empty(&log->finished_ios)))
S
Shaohua Li 已提交
1377 1378
			break;

1379 1380 1381 1382
		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
1383
	}
1384

1385
	next_checkpoint = r5c_calculate_new_cp(conf);
S
Shaohua Li 已提交
1386 1387
	spin_unlock_irq(&log->io_list_lock);

1388
	BUG_ON(reclaimable < 0);
1389 1390

	if (reclaimable == 0 || !write_super)
S
Shaohua Li 已提交
1391 1392 1393 1394 1395 1396 1397
		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
1398
	r5l_write_super_and_discard_space(log, next_checkpoint);
S
Shaohua Li 已提交
1399 1400

	mutex_lock(&log->io_mutex);
1401
	log->last_checkpoint = next_checkpoint;
1402
	r5c_update_log_state(log);
S
Shaohua Li 已提交
1403 1404
	mutex_unlock(&log->io_mutex);

1405
	r5l_run_no_space_stripes(log);
S
Shaohua Li 已提交
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415
}

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;
1416
	r5c_do_reclaim(conf);
S
Shaohua Li 已提交
1417 1418 1419
	r5l_do_reclaim(log);
}

1420
void r5l_wake_reclaim(struct r5l_log *log, sector_t space)
S
Shaohua Li 已提交
1421
{
S
Shaohua Li 已提交
1422 1423 1424
	unsigned long target;
	unsigned long new = (unsigned long)space; /* overflow in theory */

1425 1426
	if (!log)
		return;
S
Shaohua Li 已提交
1427 1428 1429 1430 1431 1432
	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
1433 1434
}

1435 1436
void r5l_quiesce(struct r5l_log *log, int state)
{
S
Shaohua Li 已提交
1437
	struct mddev *mddev;
1438 1439 1440
	if (!log || state == 2)
		return;
	if (state == 0) {
1441 1442 1443 1444 1445 1446 1447
		/*
		 * 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;
1448 1449
		log->reclaim_thread = md_register_thread(r5l_reclaim_thread,
					log->rdev->mddev, "reclaim");
1450
		log->reclaim_thread->timeout = R5C_RECLAIM_WAKEUP_INTERVAL;
1451
	} else if (state == 1) {
S
Shaohua Li 已提交
1452 1453 1454
		/* make sure r5l_write_super_and_discard_space exits */
		mddev = log->rdev->mddev;
		wake_up(&mddev->sb_wait);
1455
		r5l_wake_reclaim(log, MaxSector);
1456 1457 1458 1459 1460
		md_unregister_thread(&log->reclaim_thread);
		r5l_do_reclaim(log);
	}
}

S
Shaohua Li 已提交
1461 1462
bool r5l_log_disk_error(struct r5conf *conf)
{
1463 1464
	struct r5l_log *log;
	bool ret;
1465
	/* don't allow write if journal disk is missing */
1466 1467 1468 1469 1470 1471 1472 1473 1474
	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
1475 1476
}

S
Shaohua Li 已提交
1477 1478 1479 1480 1481
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 */
S
Song Liu 已提交
1482 1483 1484
	int data_parity_stripes;	/* number of data_parity stripes */
	int data_only_stripes;		/* number of data_only stripes */
	struct list_head cached_list;
S
Shaohua Li 已提交
1485 1486
};

1487 1488
static int r5l_recovery_read_meta_block(struct r5l_log *log,
					struct r5l_recovery_ctx *ctx)
S
Shaohua Li 已提交
1489 1490 1491 1492 1493
{
	struct page *page = ctx->meta_page;
	struct r5l_meta_block *mb;
	u32 crc, stored_crc;

M
Mike Christie 已提交
1494 1495
	if (!sync_page_io(log->rdev, ctx->pos, PAGE_SIZE, page, REQ_OP_READ, 0,
			  false))
S
Shaohua Li 已提交
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507
		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
1508
	crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519
	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;
}

1520 1521 1522 1523
static void
r5l_recovery_create_empty_meta_block(struct r5l_log *log,
				     struct page *page,
				     sector_t pos, u64 seq)
S
Shaohua Li 已提交
1524 1525 1526 1527 1528
{
	struct r5l_meta_block *mb;
	u32 crc;

	mb = page_address(page);
1529
	clear_page(mb);
S
Shaohua Li 已提交
1530 1531 1532 1533 1534
	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
1535
	crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
1536
	mb->checksum = cpu_to_le32(crc);
1537
}
S
Shaohua Li 已提交
1538

1539 1540 1541 1542 1543 1544 1545 1546 1547
static int r5l_log_write_empty_meta_block(struct r5l_log *log, sector_t pos,
					  u64 seq)
{
	struct page *page;

	page = alloc_page(GFP_KERNEL);
	if (!page)
		return -ENOMEM;
	r5l_recovery_create_empty_meta_block(log, page, pos, seq);
M
Mike Christie 已提交
1548 1549
	if (!sync_page_io(log->rdev, pos, PAGE_SIZE, page, REQ_OP_WRITE,
			  WRITE_FUA, false)) {
S
Shaohua Li 已提交
1550 1551 1552 1553 1554 1555 1556
		__free_page(page);
		return -EIO;
	}
	__free_page(page);
	return 0;
}

S
Song Liu 已提交
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 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 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 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 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 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 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 1929 1930 1931 1932 1933 1934 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 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 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 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 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 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 2135 2136 2137 2138 2139 2140
/*
 * r5l_recovery_load_data and r5l_recovery_load_parity uses flag R5_Wantwrite
 * to mark valid (potentially not flushed) data in the journal.
 *
 * We already verified checksum in r5l_recovery_verify_data_checksum_for_mb,
 * so there should not be any mismatch here.
 */
static void r5l_recovery_load_data(struct r5l_log *log,
				   struct stripe_head *sh,
				   struct r5l_recovery_ctx *ctx,
				   struct r5l_payload_data_parity *payload,
				   sector_t log_offset)
{
	struct mddev *mddev = log->rdev->mddev;
	struct r5conf *conf = mddev->private;
	int dd_idx;

	raid5_compute_sector(conf,
			     le64_to_cpu(payload->location), 0,
			     &dd_idx, sh);
	sync_page_io(log->rdev, log_offset, PAGE_SIZE,
		     sh->dev[dd_idx].page, REQ_OP_READ, 0, false);
	sh->dev[dd_idx].log_checksum =
		le32_to_cpu(payload->checksum[0]);
	ctx->meta_total_blocks += BLOCK_SECTORS;

	set_bit(R5_Wantwrite, &sh->dev[dd_idx].flags);
	set_bit(STRIPE_R5C_CACHING, &sh->state);
}

static void r5l_recovery_load_parity(struct r5l_log *log,
				     struct stripe_head *sh,
				     struct r5l_recovery_ctx *ctx,
				     struct r5l_payload_data_parity *payload,
				     sector_t log_offset)
{
	struct mddev *mddev = log->rdev->mddev;
	struct r5conf *conf = mddev->private;

	ctx->meta_total_blocks += BLOCK_SECTORS * conf->max_degraded;
	sync_page_io(log->rdev, log_offset, PAGE_SIZE,
		     sh->dev[sh->pd_idx].page, REQ_OP_READ, 0, false);
	sh->dev[sh->pd_idx].log_checksum =
		le32_to_cpu(payload->checksum[0]);
	set_bit(R5_Wantwrite, &sh->dev[sh->pd_idx].flags);

	if (sh->qd_idx >= 0) {
		sync_page_io(log->rdev,
			     r5l_ring_add(log, log_offset, BLOCK_SECTORS),
			     PAGE_SIZE, sh->dev[sh->qd_idx].page,
			     REQ_OP_READ, 0, false);
		sh->dev[sh->qd_idx].log_checksum =
			le32_to_cpu(payload->checksum[1]);
		set_bit(R5_Wantwrite, &sh->dev[sh->qd_idx].flags);
	}
	clear_bit(STRIPE_R5C_CACHING, &sh->state);
}

static void r5l_recovery_reset_stripe(struct stripe_head *sh)
{
	int i;

	sh->state = 0;
	sh->log_start = MaxSector;
	for (i = sh->disks; i--; )
		sh->dev[i].flags = 0;
}

static void
r5l_recovery_replay_one_stripe(struct r5conf *conf,
			       struct stripe_head *sh,
			       struct r5l_recovery_ctx *ctx)
{
	struct md_rdev *rdev, *rrdev;
	int disk_index;
	int data_count = 0;

	for (disk_index = 0; disk_index < sh->disks; disk_index++) {
		if (!test_bit(R5_Wantwrite, &sh->dev[disk_index].flags))
			continue;
		if (disk_index == sh->qd_idx || disk_index == sh->pd_idx)
			continue;
		data_count++;
	}

	/*
	 * stripes that only have parity must have been flushed
	 * before the crash that we are now recovering from, so
	 * there is nothing more to recovery.
	 */
	if (data_count == 0)
		goto out;

	for (disk_index = 0; disk_index < sh->disks; disk_index++) {
		if (!test_bit(R5_Wantwrite, &sh->dev[disk_index].flags))
			continue;

		/* in case device is broken */
		rcu_read_lock();
		rdev = rcu_dereference(conf->disks[disk_index].rdev);
		if (rdev) {
			atomic_inc(&rdev->nr_pending);
			rcu_read_unlock();
			sync_page_io(rdev, sh->sector, PAGE_SIZE,
				     sh->dev[disk_index].page, REQ_OP_WRITE, 0,
				     false);
			rdev_dec_pending(rdev, rdev->mddev);
			rcu_read_lock();
		}
		rrdev = rcu_dereference(conf->disks[disk_index].replacement);
		if (rrdev) {
			atomic_inc(&rrdev->nr_pending);
			rcu_read_unlock();
			sync_page_io(rrdev, sh->sector, PAGE_SIZE,
				     sh->dev[disk_index].page, REQ_OP_WRITE, 0,
				     false);
			rdev_dec_pending(rrdev, rrdev->mddev);
			rcu_read_lock();
		}
		rcu_read_unlock();
	}
	ctx->data_parity_stripes++;
out:
	r5l_recovery_reset_stripe(sh);
}

static struct stripe_head *
r5c_recovery_alloc_stripe(struct r5conf *conf,
			  struct list_head *recovery_list,
			  sector_t stripe_sect,
			  sector_t log_start)
{
	struct stripe_head *sh;

	sh = raid5_get_active_stripe(conf, stripe_sect, 0, 1, 0);
	if (!sh)
		return NULL;  /* no more stripe available */

	r5l_recovery_reset_stripe(sh);
	sh->log_start = log_start;

	return sh;
}

static struct stripe_head *
r5c_recovery_lookup_stripe(struct list_head *list, sector_t sect)
{
	struct stripe_head *sh;

	list_for_each_entry(sh, list, lru)
		if (sh->sector == sect)
			return sh;
	return NULL;
}

static void
r5c_recovery_drop_stripes(struct list_head *cached_stripe_list,
			  struct r5l_recovery_ctx *ctx)
{
	struct stripe_head *sh, *next;

	list_for_each_entry_safe(sh, next, cached_stripe_list, lru) {
		r5l_recovery_reset_stripe(sh);
		list_del_init(&sh->lru);
		raid5_release_stripe(sh);
	}
}

static void
r5c_recovery_replay_stripes(struct list_head *cached_stripe_list,
			    struct r5l_recovery_ctx *ctx)
{
	struct stripe_head *sh, *next;

	list_for_each_entry_safe(sh, next, cached_stripe_list, lru)
		if (!test_bit(STRIPE_R5C_CACHING, &sh->state)) {
			r5l_recovery_replay_one_stripe(sh->raid_conf, sh, ctx);
			list_del_init(&sh->lru);
			raid5_release_stripe(sh);
		}
}

/* if matches return 0; otherwise return -EINVAL */
static int
r5l_recovery_verify_data_checksum(struct r5l_log *log, struct page *page,
				  sector_t log_offset, __le32 log_checksum)
{
	void *addr;
	u32 checksum;

	sync_page_io(log->rdev, log_offset, PAGE_SIZE,
		     page, REQ_OP_READ, 0, false);
	addr = kmap_atomic(page);
	checksum = crc32c_le(log->uuid_checksum, addr, PAGE_SIZE);
	kunmap_atomic(addr);
	return (le32_to_cpu(log_checksum) == checksum) ? 0 : -EINVAL;
}

/*
 * before loading data to stripe cache, we need verify checksum for all data,
 * if there is mismatch for any data page, we drop all data in the mata block
 */
static int
r5l_recovery_verify_data_checksum_for_mb(struct r5l_log *log,
					 struct r5l_recovery_ctx *ctx)
{
	struct mddev *mddev = log->rdev->mddev;
	struct r5conf *conf = mddev->private;
	struct r5l_meta_block *mb = page_address(ctx->meta_page);
	sector_t mb_offset = sizeof(struct r5l_meta_block);
	sector_t log_offset = r5l_ring_add(log, ctx->pos, BLOCK_SECTORS);
	struct page *page;
	struct r5l_payload_data_parity *payload;

	page = alloc_page(GFP_KERNEL);
	if (!page)
		return -ENOMEM;

	while (mb_offset < le32_to_cpu(mb->meta_size)) {
		payload = (void *)mb + mb_offset;

		if (payload->header.type == R5LOG_PAYLOAD_DATA) {
			if (r5l_recovery_verify_data_checksum(
				    log, page, log_offset,
				    payload->checksum[0]) < 0)
				goto mismatch;
		} else if (payload->header.type == R5LOG_PAYLOAD_PARITY) {
			if (r5l_recovery_verify_data_checksum(
				    log, page, log_offset,
				    payload->checksum[0]) < 0)
				goto mismatch;
			if (conf->max_degraded == 2 && /* q for RAID 6 */
			    r5l_recovery_verify_data_checksum(
				    log, page,
				    r5l_ring_add(log, log_offset,
						 BLOCK_SECTORS),
				    payload->checksum[1]) < 0)
				goto mismatch;
		} else /* not R5LOG_PAYLOAD_DATA or R5LOG_PAYLOAD_PARITY */
			goto mismatch;

		log_offset = r5l_ring_add(log, log_offset,
					  le32_to_cpu(payload->size));

		mb_offset += sizeof(struct r5l_payload_data_parity) +
			sizeof(__le32) *
			(le32_to_cpu(payload->size) >> (PAGE_SHIFT - 9));
	}

	put_page(page);
	return 0;

mismatch:
	put_page(page);
	return -EINVAL;
}

/*
 * Analyze all data/parity pages in one meta block
 * Returns:
 * 0 for success
 * -EINVAL for unknown playload type
 * -EAGAIN for checksum mismatch of data page
 * -ENOMEM for run out of memory (alloc_page failed or run out of stripes)
 */
static int
r5c_recovery_analyze_meta_block(struct r5l_log *log,
				struct r5l_recovery_ctx *ctx,
				struct list_head *cached_stripe_list)
{
	struct mddev *mddev = log->rdev->mddev;
	struct r5conf *conf = mddev->private;
	struct r5l_meta_block *mb;
	struct r5l_payload_data_parity *payload;
	int mb_offset;
	sector_t log_offset;
	sector_t stripe_sect;
	struct stripe_head *sh;
	int ret;

	/*
	 * for mismatch in data blocks, we will drop all data in this mb, but
	 * we will still read next mb for other data with FLUSH flag, as
	 * io_unit could finish out of order.
	 */
	ret = r5l_recovery_verify_data_checksum_for_mb(log, ctx);
	if (ret == -EINVAL)
		return -EAGAIN;
	else if (ret)
		return ret;   /* -ENOMEM duo to alloc_page() failed */

	mb = page_address(ctx->meta_page);
	mb_offset = sizeof(struct r5l_meta_block);
	log_offset = r5l_ring_add(log, ctx->pos, BLOCK_SECTORS);

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

		payload = (void *)mb + mb_offset;
		stripe_sect = (payload->header.type == R5LOG_PAYLOAD_DATA) ?
			raid5_compute_sector(
				conf, le64_to_cpu(payload->location), 0, &dd,
				NULL)
			: le64_to_cpu(payload->location);

		sh = r5c_recovery_lookup_stripe(cached_stripe_list,
						stripe_sect);

		if (!sh) {
			sh = r5c_recovery_alloc_stripe(conf, cached_stripe_list,
						       stripe_sect, ctx->pos);
			/*
			 * cannot get stripe from raid5_get_active_stripe
			 * try replay some stripes
			 */
			if (!sh) {
				r5c_recovery_replay_stripes(
					cached_stripe_list, ctx);
				sh = r5c_recovery_alloc_stripe(
					conf, cached_stripe_list,
					stripe_sect, ctx->pos);
			}
			if (!sh) {
				pr_debug("md/raid:%s: Increasing stripe cache size to %d to recovery data on journal.\n",
					mdname(mddev),
					conf->min_nr_stripes * 2);
				raid5_set_cache_size(mddev,
						     conf->min_nr_stripes * 2);
				sh = r5c_recovery_alloc_stripe(
					conf, cached_stripe_list, stripe_sect,
					ctx->pos);
			}
			if (!sh) {
				pr_err("md/raid:%s: Cannot get enough stripes due to memory pressure. Recovery failed.\n",
				       mdname(mddev));
				return -ENOMEM;
			}
			list_add_tail(&sh->lru, cached_stripe_list);
		}

		if (payload->header.type == R5LOG_PAYLOAD_DATA) {
			if (!test_bit(STRIPE_R5C_CACHING, &sh->state)) {
				r5l_recovery_replay_one_stripe(conf, sh, ctx);
				r5l_recovery_reset_stripe(sh);
				sh->log_start = ctx->pos;
				list_move_tail(&sh->lru, cached_stripe_list);
			}
			r5l_recovery_load_data(log, sh, ctx, payload,
					       log_offset);
		} else if (payload->header.type == R5LOG_PAYLOAD_PARITY)
			r5l_recovery_load_parity(log, sh, ctx, payload,
						 log_offset);
		else
			return -EINVAL;

		log_offset = r5l_ring_add(log, log_offset,
					  le32_to_cpu(payload->size));

		mb_offset += sizeof(struct r5l_payload_data_parity) +
			sizeof(__le32) *
			(le32_to_cpu(payload->size) >> (PAGE_SHIFT - 9));
	}

	return 0;
}

/*
 * Load the stripe into cache. The stripe will be written out later by
 * the stripe cache state machine.
 */
static void r5c_recovery_load_one_stripe(struct r5l_log *log,
					 struct stripe_head *sh)
{
	struct r5conf *conf = sh->raid_conf;
	struct r5dev *dev;
	int i;

	for (i = sh->disks; i--; ) {
		dev = sh->dev + i;
		if (test_and_clear_bit(R5_Wantwrite, &dev->flags)) {
			set_bit(R5_InJournal, &dev->flags);
			set_bit(R5_UPTODATE, &dev->flags);
		}
	}
	set_bit(STRIPE_R5C_PARTIAL_STRIPE, &sh->state);
	atomic_inc(&conf->r5c_cached_partial_stripes);
	list_add_tail(&sh->r5c, &log->stripe_in_journal_list);
}

/*
 * Scan through the log for all to-be-flushed data
 *
 * For stripes with data and parity, namely Data-Parity stripe
 * (STRIPE_R5C_CACHING == 0), we simply replay all the writes.
 *
 * For stripes with only data, namely Data-Only stripe
 * (STRIPE_R5C_CACHING == 1), we load them to stripe cache state machine.
 *
 * For a stripe, if we see data after parity, we should discard all previous
 * data and parity for this stripe, as these data are already flushed to
 * the array.
 *
 * At the end of the scan, we return the new journal_tail, which points to
 * first data-only stripe on the journal device, or next invalid meta block.
 */
static int r5c_recovery_flush_log(struct r5l_log *log,
				  struct r5l_recovery_ctx *ctx)
{
	struct stripe_head *sh, *next;
	int ret = 0;

	/* scan through the log */
	while (1) {
		if (r5l_recovery_read_meta_block(log, ctx))
			break;

		ret = r5c_recovery_analyze_meta_block(log, ctx,
						      &ctx->cached_list);
		/*
		 * -EAGAIN means mismatch in data block, in this case, we still
		 * try scan the next metablock
		 */
		if (ret && ret != -EAGAIN)
			break;   /* ret == -EINVAL or -ENOMEM */
		ctx->seq++;
		ctx->pos = r5l_ring_add(log, ctx->pos, ctx->meta_total_blocks);
	}

	if (ret == -ENOMEM) {
		r5c_recovery_drop_stripes(&ctx->cached_list, ctx);
		return ret;
	}

	/* replay data-parity stripes */
	r5c_recovery_replay_stripes(&ctx->cached_list, ctx);

	/* load data-only stripes to stripe cache */
	list_for_each_entry_safe(sh, next, &ctx->cached_list, lru) {
		WARN_ON(!test_bit(STRIPE_R5C_CACHING, &sh->state));
		r5c_recovery_load_one_stripe(log, sh);
		list_del_init(&sh->lru);
		raid5_release_stripe(sh);
		ctx->data_only_stripes++;
	}

	return 0;
}

/*
 * 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
 */

/*
 * Before recovery, the log looks like the following
 *
 *   ---------------------------------------------
 *   |           valid log        | invalid log  |
 *   ---------------------------------------------
 *   ^
 *   |- log->last_checkpoint
 *   |- log->last_cp_seq
 *
 * Now we scan through the log until we see invalid entry
 *
 *   ---------------------------------------------
 *   |           valid log        | invalid log  |
 *   ---------------------------------------------
 *   ^                            ^
 *   |- log->last_checkpoint      |- ctx->pos
 *   |- log->last_cp_seq          |- ctx->seq
 *
 * From this point, we need to increase seq number by 10 to avoid
 * confusing next recovery.
 *
 *   ---------------------------------------------
 *   |           valid log        | invalid log  |
 *   ---------------------------------------------
 *   ^                              ^
 *   |- log->last_checkpoint        |- ctx->pos+1
 *   |- log->last_cp_seq            |- ctx->seq+11
 *
 * However, it is not safe to start the state machine yet, because data only
 * parities are not yet secured in RAID. To save these data only parities, we
 * rewrite them from seq+11.
 *
 *   -----------------------------------------------------------------
 *   |           valid log        | data only stripes | invalid log  |
 *   -----------------------------------------------------------------
 *   ^                                                ^
 *   |- log->last_checkpoint                          |- ctx->pos+n
 *   |- log->last_cp_seq                              |- ctx->seq+10+n
 *
 * If failure happens again during this process, the recovery can safe start
 * again from log->last_checkpoint.
 *
 * Once data only stripes are rewritten to journal, we move log_tail
 *
 *   -----------------------------------------------------------------
 *   |     old log        |    data only stripes    | invalid log  |
 *   -----------------------------------------------------------------
 *                        ^                         ^
 *                        |- log->last_checkpoint   |- ctx->pos+n
 *                        |- log->last_cp_seq       |- ctx->seq+10+n
 *
 * Then we can safely start the state machine. If failure happens from this
 * point on, the recovery will start from new log->last_checkpoint.
 */
static int
r5c_recovery_rewrite_data_only_stripes(struct r5l_log *log,
				       struct r5l_recovery_ctx *ctx)
{
	struct stripe_head *sh;
	struct mddev *mddev = log->rdev->mddev;
	struct page *page;

	page = alloc_page(GFP_KERNEL);
	if (!page) {
		pr_err("md/raid:%s: cannot allocate memory to rewrite data only stripes\n",
		       mdname(mddev));
		return -ENOMEM;
	}

	ctx->seq += 10;
	list_for_each_entry(sh, &ctx->cached_list, lru) {
		struct r5l_meta_block *mb;
		int i;
		int offset;
		sector_t write_pos;

		WARN_ON(!test_bit(STRIPE_R5C_CACHING, &sh->state));
		r5l_recovery_create_empty_meta_block(log, page,
						     ctx->pos, ctx->seq);
		mb = page_address(page);
		offset = le32_to_cpu(mb->meta_size);
		write_pos = ctx->pos + BLOCK_SECTORS;

		for (i = sh->disks; i--; ) {
			struct r5dev *dev = &sh->dev[i];
			struct r5l_payload_data_parity *payload;
			void *addr;

			if (test_bit(R5_InJournal, &dev->flags)) {
				payload = (void *)mb + offset;
				payload->header.type = cpu_to_le16(
					R5LOG_PAYLOAD_DATA);
				payload->size = BLOCK_SECTORS;
				payload->location = cpu_to_le64(
					raid5_compute_blocknr(sh, i, 0));
				addr = kmap_atomic(dev->page);
				payload->checksum[0] = cpu_to_le32(
					crc32c_le(log->uuid_checksum, addr,
						  PAGE_SIZE));
				kunmap_atomic(addr);
				sync_page_io(log->rdev, write_pos, PAGE_SIZE,
					     dev->page, REQ_OP_WRITE, 0, false);
				write_pos = r5l_ring_add(log, write_pos,
							 BLOCK_SECTORS);
				offset += sizeof(__le32) +
					sizeof(struct r5l_payload_data_parity);

			}
		}
		mb->meta_size = cpu_to_le32(offset);
		mb->checksum = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
		sync_page_io(log->rdev, ctx->pos, PAGE_SIZE, page,
			     REQ_OP_WRITE, WRITE_FUA, false);
		sh->log_start = ctx->pos;
		ctx->pos = write_pos;
		ctx->seq += 1;
	}
	__free_page(page);
	return 0;
}

S
Shaohua Li 已提交
2141 2142
static int r5l_recovery_log(struct r5l_log *log)
{
S
Song Liu 已提交
2143
	struct mddev *mddev = log->rdev->mddev;
S
Shaohua Li 已提交
2144
	struct r5l_recovery_ctx ctx;
S
Song Liu 已提交
2145
	int ret;
S
Shaohua Li 已提交
2146 2147 2148 2149

	ctx.pos = log->last_checkpoint;
	ctx.seq = log->last_cp_seq;
	ctx.meta_page = alloc_page(GFP_KERNEL);
S
Song Liu 已提交
2150 2151 2152 2153
	ctx.data_only_stripes = 0;
	ctx.data_parity_stripes = 0;
	INIT_LIST_HEAD(&ctx.cached_list);

S
Shaohua Li 已提交
2154 2155 2156
	if (!ctx.meta_page)
		return -ENOMEM;

S
Song Liu 已提交
2157
	ret = r5c_recovery_flush_log(log, &ctx);
S
Shaohua Li 已提交
2158 2159
	__free_page(ctx.meta_page);

S
Song Liu 已提交
2160 2161
	if (ret)
		return ret;
S
Song Liu 已提交
2162

S
Song Liu 已提交
2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176
	if ((ctx.data_only_stripes == 0) && (ctx.data_parity_stripes == 0))
		pr_debug("md/raid:%s: starting from clean shutdown\n",
			 mdname(mddev));
	else {
		pr_debug("md/raid:%s: recoverying %d data-only stripes and %d data-parity stripes\n",
			 mdname(mddev), ctx.data_only_stripes,
			 ctx.data_parity_stripes);

		if (ctx.data_only_stripes > 0)
			if (r5c_recovery_rewrite_data_only_stripes(log, &ctx)) {
				pr_err("md/raid:%s: failed to rewrite stripes to journal\n",
				       mdname(mddev));
				return -EIO;
			}
S
Song Liu 已提交
2177 2178
	}

S
Song Liu 已提交
2179 2180 2181 2182 2183
	log->log_start = ctx.pos;
	log->next_checkpoint = ctx.pos;
	log->seq = ctx.seq;
	r5l_log_write_empty_meta_block(log, ctx.pos, ctx.seq);
	r5l_write_super(log, ctx.pos);
S
Shaohua Li 已提交
2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194
	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);
}

S
Song Liu 已提交
2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257
static ssize_t r5c_journal_mode_show(struct mddev *mddev, char *page)
{
	struct r5conf *conf = mddev->private;
	int ret;

	if (!conf->log)
		return 0;

	switch (conf->log->r5c_journal_mode) {
	case R5C_JOURNAL_MODE_WRITE_THROUGH:
		ret = snprintf(
			page, PAGE_SIZE, "[%s] %s\n",
			r5c_journal_mode_str[R5C_JOURNAL_MODE_WRITE_THROUGH],
			r5c_journal_mode_str[R5C_JOURNAL_MODE_WRITE_BACK]);
		break;
	case R5C_JOURNAL_MODE_WRITE_BACK:
		ret = snprintf(
			page, PAGE_SIZE, "%s [%s]\n",
			r5c_journal_mode_str[R5C_JOURNAL_MODE_WRITE_THROUGH],
			r5c_journal_mode_str[R5C_JOURNAL_MODE_WRITE_BACK]);
		break;
	default:
		ret = 0;
	}
	return ret;
}

static ssize_t r5c_journal_mode_store(struct mddev *mddev,
				      const char *page, size_t length)
{
	struct r5conf *conf = mddev->private;
	struct r5l_log *log = conf->log;
	int val = -1, i;
	int len = length;

	if (!log)
		return -ENODEV;

	if (len && page[len - 1] == '\n')
		len -= 1;
	for (i = 0; i < ARRAY_SIZE(r5c_journal_mode_str); i++)
		if (strlen(r5c_journal_mode_str[i]) == len &&
		    strncmp(page, r5c_journal_mode_str[i], len) == 0) {
			val = i;
			break;
		}
	if (val < R5C_JOURNAL_MODE_WRITE_THROUGH ||
	    val > R5C_JOURNAL_MODE_WRITE_BACK)
		return -EINVAL;

	mddev_suspend(mddev);
	conf->log->r5c_journal_mode = val;
	mddev_resume(mddev);

	pr_debug("md/raid:%s: setting r5c cache mode to %d: %s\n",
		 mdname(mddev), val, r5c_journal_mode_str[val]);
	return length;
}

struct md_sysfs_entry
r5c_journal_mode = __ATTR(journal_mode, 0644,
			  r5c_journal_mode_show, r5c_journal_mode_store);

2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271
/*
 * Try handle write operation in caching phase. This function should only
 * be called in write-back mode.
 *
 * If all outstanding writes can be handled in caching phase, returns 0
 * If writes requires write-out phase, call r5c_make_stripe_write_out()
 * and returns -EAGAIN
 */
int r5c_try_caching_write(struct r5conf *conf,
			  struct stripe_head *sh,
			  struct stripe_head_state *s,
			  int disks)
{
	struct r5l_log *log = conf->log;
S
Song Liu 已提交
2272 2273 2274
	int i;
	struct r5dev *dev;
	int to_cache = 0;
2275 2276 2277

	BUG_ON(!r5c_is_writeback(log));

S
Song Liu 已提交
2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 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 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345
	if (!test_bit(STRIPE_R5C_CACHING, &sh->state)) {
		/*
		 * There are two different scenarios here:
		 *  1. The stripe has some data cached, and it is sent to
		 *     write-out phase for reclaim
		 *  2. The stripe is clean, and this is the first write
		 *
		 * For 1, return -EAGAIN, so we continue with
		 * handle_stripe_dirtying().
		 *
		 * For 2, set STRIPE_R5C_CACHING and continue with caching
		 * write.
		 */

		/* case 1: anything injournal or anything in written */
		if (s->injournal > 0 || s->written > 0)
			return -EAGAIN;
		/* case 2 */
		set_bit(STRIPE_R5C_CACHING, &sh->state);
	}

	for (i = disks; i--; ) {
		dev = &sh->dev[i];
		/* if non-overwrite, use writing-out phase */
		if (dev->towrite && !test_bit(R5_OVERWRITE, &dev->flags) &&
		    !test_bit(R5_InJournal, &dev->flags)) {
			r5c_make_stripe_write_out(sh);
			return -EAGAIN;
		}
	}

	for (i = disks; i--; ) {
		dev = &sh->dev[i];
		if (dev->towrite) {
			set_bit(R5_Wantwrite, &dev->flags);
			set_bit(R5_Wantdrain, &dev->flags);
			set_bit(R5_LOCKED, &dev->flags);
			to_cache++;
		}
	}

	if (to_cache) {
		set_bit(STRIPE_OP_BIODRAIN, &s->ops_request);
		/*
		 * set STRIPE_LOG_TRAPPED, which triggers r5c_cache_data()
		 * in ops_run_io(). STRIPE_LOG_TRAPPED will be cleared in
		 * r5c_handle_data_cached()
		 */
		set_bit(STRIPE_LOG_TRAPPED, &sh->state);
	}

	return 0;
}

/*
 * free extra pages (orig_page) we allocated for prexor
 */
void r5c_release_extra_page(struct stripe_head *sh)
{
	int i;

	for (i = sh->disks; i--; )
		if (sh->dev[i].page != sh->dev[i].orig_page) {
			struct page *p = sh->dev[i].orig_page;

			sh->dev[i].orig_page = sh->dev[i].page;
			put_page(p);
		}
2346 2347 2348 2349 2350 2351 2352 2353 2354 2355
}

/*
 * clean up the stripe (clear R5_InJournal for dev[pd_idx] etc.) after the
 * stripe is committed to RAID disks.
 */
void r5c_finish_stripe_write_out(struct r5conf *conf,
				 struct stripe_head *sh,
				 struct stripe_head_state *s)
{
S
Song Liu 已提交
2356 2357 2358
	int i;
	int do_wakeup = 0;

2359 2360 2361 2362 2363 2364 2365 2366 2367
	if (!conf->log ||
	    !test_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags))
		return;

	WARN_ON(test_bit(STRIPE_R5C_CACHING, &sh->state));
	clear_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags);

	if (conf->log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH)
		return;
S
Song Liu 已提交
2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386

	for (i = sh->disks; i--; ) {
		clear_bit(R5_InJournal, &sh->dev[i].flags);
		if (test_and_clear_bit(R5_Overlap, &sh->dev[i].flags))
			do_wakeup = 1;
	}

	/*
	 * analyse_stripe() runs before r5c_finish_stripe_write_out(),
	 * We updated R5_InJournal, so we also update s->injournal.
	 */
	s->injournal = 0;

	if (test_and_clear_bit(STRIPE_FULL_WRITE, &sh->state))
		if (atomic_dec_and_test(&conf->pending_full_writes))
			md_wakeup_thread(conf->mddev->thread);

	if (do_wakeup)
		wake_up(&conf->wait_for_overlap);
2387 2388 2389 2390 2391 2392 2393 2394 2395

	if (conf->log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH)
		return;

	spin_lock_irq(&conf->log->stripe_in_journal_lock);
	list_del_init(&sh->r5c);
	spin_unlock_irq(&conf->log->stripe_in_journal_lock);
	sh->log_start = MaxSector;
	atomic_dec(&conf->log->stripe_in_journal_count);
S
Song Liu 已提交
2396 2397 2398 2399 2400 2401
}

int
r5c_cache_data(struct r5l_log *log, struct stripe_head *sh,
	       struct stripe_head_state *s)
{
2402
	struct r5conf *conf = sh->raid_conf;
S
Song Liu 已提交
2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433
	int pages = 0;
	int reserve;
	int i;
	int ret = 0;

	BUG_ON(!log);

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

		if (!test_bit(R5_Wantwrite, &sh->dev[i].flags))
			continue;
		addr = kmap_atomic(sh->dev[i].page);
		sh->dev[i].log_checksum = crc32c_le(log->uuid_checksum,
						    addr, PAGE_SIZE);
		kunmap_atomic(addr);
		pages++;
	}
	WARN_ON(pages == 0);

	/*
	 * The stripe must enter state machine again to call endio, so
	 * don't delay.
	 */
	clear_bit(STRIPE_DELAYED, &sh->state);
	atomic_inc(&sh->count);

	mutex_lock(&log->io_mutex);
	/* meta + data */
	reserve = (1 + pages) << (PAGE_SHIFT - 9);

2434 2435 2436 2437 2438 2439 2440 2441
	if (test_bit(R5C_LOG_CRITICAL, &conf->cache_state) &&
	    sh->log_start == MaxSector)
		r5l_add_no_space_stripe(log, sh);
	else if (!r5l_has_free_space(log, reserve)) {
		if (sh->log_start == log->last_checkpoint)
			BUG();
		else
			r5l_add_no_space_stripe(log, sh);
S
Song Liu 已提交
2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452
	} else {
		ret = r5l_log_stripe(log, sh, pages, 0);
		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);
		}
	}

	mutex_unlock(&log->io_mutex);
	return 0;
2453 2454
}

S
Shaohua Li 已提交
2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471
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
2472
	if (!sync_page_io(rdev, cp, PAGE_SIZE, page, REQ_OP_READ, 0, false)) {
S
Shaohua Li 已提交
2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484
		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
2485
	expected_crc = crc32c_le(log->uuid_checksum, mb, PAGE_SIZE);
S
Shaohua Li 已提交
2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497
	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;
2498
		r5l_log_write_empty_meta_block(log, cp, log->last_cp_seq);
S
Shaohua Li 已提交
2499 2500 2501 2502 2503 2504 2505 2506 2507 2508
		/*
		 * 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
2509 2510 2511
	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
2512
	log->last_checkpoint = cp;
2513
	log->next_checkpoint = cp;
2514 2515 2516
	mutex_lock(&log->io_mutex);
	r5c_update_log_state(log);
	mutex_unlock(&log->io_mutex);
S
Shaohua Li 已提交
2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527

	__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
2528
	struct request_queue *q = bdev_get_queue(rdev->bdev);
S
Shaohua Li 已提交
2529 2530 2531 2532
	struct r5l_log *log;

	if (PAGE_SIZE != 4096)
		return -EINVAL;
2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548

	/*
	 * The PAGE_SIZE must be big enough to hold 1 r5l_meta_block and
	 * raid_disks r5l_payload_data_parity.
	 *
	 * Write journal and cache does not work for very big array
	 * (raid_disks > 203)
	 */
	if (sizeof(struct r5l_meta_block) +
	    ((sizeof(struct r5l_payload_data_parity) + sizeof(__le32)) *
	     conf->raid_disks) > PAGE_SIZE) {
		pr_err("md/raid:%s: write journal/cache doesn't work for array with %d disks\n",
		       mdname(conf->mddev), conf->raid_disks);
		return -EINVAL;
	}

S
Shaohua Li 已提交
2549 2550 2551 2552 2553
	log = kzalloc(sizeof(*log), GFP_KERNEL);
	if (!log)
		return -ENOMEM;
	log->rdev = rdev;

J
Jens Axboe 已提交
2554
	log->need_cache_flush = test_bit(QUEUE_FLAG_WC, &q->queue_flags) != 0;
2555

S
Shaohua Li 已提交
2556 2557
	log->uuid_checksum = crc32c_le(~0, rdev->mddev->uuid,
				       sizeof(rdev->mddev->uuid));
S
Shaohua Li 已提交
2558 2559 2560 2561 2562

	mutex_init(&log->io_mutex);

	spin_lock_init(&log->io_list_lock);
	INIT_LIST_HEAD(&log->running_ios);
S
Shaohua Li 已提交
2563
	INIT_LIST_HEAD(&log->io_end_ios);
2564
	INIT_LIST_HEAD(&log->flushing_ios);
2565
	INIT_LIST_HEAD(&log->finished_ios);
2566
	bio_init(&log->flush_bio);
S
Shaohua Li 已提交
2567 2568 2569 2570 2571

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

2572 2573 2574 2575
	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
2576 2577 2578 2579
	log->bs = bioset_create(R5L_POOL_SIZE, 0);
	if (!log->bs)
		goto io_bs;

2580 2581 2582 2583
	log->meta_pool = mempool_create_page_pool(R5L_POOL_SIZE, 0);
	if (!log->meta_pool)
		goto out_mempool;

S
Shaohua Li 已提交
2584 2585 2586 2587
	log->reclaim_thread = md_register_thread(r5l_reclaim_thread,
						 log->rdev->mddev, "reclaim");
	if (!log->reclaim_thread)
		goto reclaim_thread;
2588 2589
	log->reclaim_thread->timeout = R5C_RECLAIM_WAKEUP_INTERVAL;

2590
	init_waitqueue_head(&log->iounit_wait);
S
Shaohua Li 已提交
2591

2592 2593
	INIT_LIST_HEAD(&log->no_mem_stripes);

S
Shaohua Li 已提交
2594 2595 2596
	INIT_LIST_HEAD(&log->no_space_stripes);
	spin_lock_init(&log->no_space_stripes_lock);

S
Song Liu 已提交
2597 2598
	INIT_WORK(&log->deferred_io_work, r5l_submit_io_async);

2599
	log->r5c_journal_mode = R5C_JOURNAL_MODE_WRITE_THROUGH;
2600 2601 2602
	INIT_LIST_HEAD(&log->stripe_in_journal_list);
	spin_lock_init(&log->stripe_in_journal_lock);
	atomic_set(&log->stripe_in_journal_count, 0);
2603

S
Shaohua Li 已提交
2604 2605 2606
	if (r5l_load_log(log))
		goto error;

2607
	rcu_assign_pointer(conf->log, log);
2608
	set_bit(MD_HAS_JOURNAL, &conf->mddev->flags);
S
Shaohua Li 已提交
2609
	return 0;
2610

S
Shaohua Li 已提交
2611
error:
S
Shaohua Li 已提交
2612 2613
	md_unregister_thread(&log->reclaim_thread);
reclaim_thread:
2614 2615
	mempool_destroy(log->meta_pool);
out_mempool:
C
Christoph Hellwig 已提交
2616 2617
	bioset_free(log->bs);
io_bs:
2618 2619
	mempool_destroy(log->io_pool);
io_pool:
S
Shaohua Li 已提交
2620 2621 2622 2623 2624 2625 2626 2627
	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
2628
	md_unregister_thread(&log->reclaim_thread);
2629
	mempool_destroy(log->meta_pool);
C
Christoph Hellwig 已提交
2630
	bioset_free(log->bs);
2631
	mempool_destroy(log->io_pool);
S
Shaohua Li 已提交
2632 2633 2634
	kmem_cache_destroy(log->io_kc);
	kfree(log);
}