segment.h 27.4 KB
Newer Older
1
/* SPDX-License-Identifier: GPL-2.0 */
2
/*
3 4 5 6 7
 * fs/f2fs/segment.h
 *
 * Copyright (c) 2012 Samsung Electronics Co., Ltd.
 *             http://www.samsung.com/
 */
8
#include <linux/blkdev.h>
9
#include <linux/backing-dev.h>
10

11 12
/* constant macro */
#define NULL_SEGNO			((unsigned int)(~0))
13
#define NULL_SECNO			((unsigned int)(~0))
14

15
#define DEF_RECLAIM_PREFREE_SEGMENTS	5	/* 5% over total segments */
16
#define DEF_MAX_RECLAIM_PREFREE_SEGMENTS	4096	/* 8GB in maximum */
17

Jaegeuk Kim's avatar
Jaegeuk Kim committed
18
#define F2FS_MIN_SEGMENTS	9 /* SB + 2 (CP + SIT + NAT) + SSA + MAIN */
19
#define F2FS_MIN_META_SEGMENTS	8 /* SB + 2 (CP + SIT + NAT) + SSA */
Jaegeuk Kim's avatar
Jaegeuk Kim committed
20

Namjae Jeon's avatar
Namjae Jeon committed
21
/* L: Logical segment # in volume, R: Relative segment # in main area */
22 23
#define GET_L2R_SEGNO(free_i, segno)	((segno) - (free_i)->start_segno)
#define GET_R2L_SEGNO(free_i, segno)	((segno) + (free_i)->start_segno)
24

25
#define IS_DATASEG(t)	((t) <= CURSEG_COLD_DATA)
Chao Yu's avatar
Chao Yu committed
26
#define IS_NODESEG(t)	((t) >= CURSEG_HOT_NODE && (t) <= CURSEG_COLD_NODE)
27

28 29 30 31 32 33
static inline void sanity_check_seg_type(struct f2fs_sb_info *sbi,
						unsigned short seg_type)
{
	f2fs_bug_on(sbi, seg_type >= NR_PERSISTENT_LOG);
}

Jaegeuk Kim's avatar
Jaegeuk Kim committed
34 35 36 37
#define IS_HOT(t)	((t) == CURSEG_HOT_NODE || (t) == CURSEG_HOT_DATA)
#define IS_WARM(t)	((t) == CURSEG_WARM_NODE || (t) == CURSEG_WARM_DATA)
#define IS_COLD(t)	((t) == CURSEG_COLD_NODE || (t) == CURSEG_COLD_DATA)

38
#define IS_CURSEG(sbi, seg)						\
39 40 41 42 43
	(((seg) == CURSEG_I(sbi, CURSEG_HOT_DATA)->segno) ||	\
	 ((seg) == CURSEG_I(sbi, CURSEG_WARM_DATA)->segno) ||	\
	 ((seg) == CURSEG_I(sbi, CURSEG_COLD_DATA)->segno) ||	\
	 ((seg) == CURSEG_I(sbi, CURSEG_HOT_NODE)->segno) ||	\
	 ((seg) == CURSEG_I(sbi, CURSEG_WARM_NODE)->segno) ||	\
Chao Yu's avatar
Chao Yu committed
44
	 ((seg) == CURSEG_I(sbi, CURSEG_COLD_NODE)->segno) ||	\
45 46
	 ((seg) == CURSEG_I(sbi, CURSEG_COLD_DATA_PINNED)->segno) ||	\
	 ((seg) == CURSEG_I(sbi, CURSEG_ALL_DATA_ATGC)->segno))
47 48

#define IS_CURSEC(sbi, secno)						\
49 50 51 52 53 54 55 56 57 58 59
	(((secno) == CURSEG_I(sbi, CURSEG_HOT_DATA)->segno /		\
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_WARM_DATA)->segno /		\
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_COLD_DATA)->segno /		\
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_HOT_NODE)->segno /		\
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_WARM_NODE)->segno /		\
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_COLD_NODE)->segno /		\
Chao Yu's avatar
Chao Yu committed
60 61
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_COLD_DATA_PINNED)->segno /	\
62 63
	  (sbi)->segs_per_sec) ||	\
	 ((secno) == CURSEG_I(sbi, CURSEG_ALL_DATA_ATGC)->segno /	\
Chao Yu's avatar
Chao Yu committed
64
	  (sbi)->segs_per_sec))
65

66 67 68 69 70 71
#define MAIN_BLKADDR(sbi)						\
	(SM_I(sbi) ? SM_I(sbi)->main_blkaddr : 				\
		le32_to_cpu(F2FS_RAW_SUPER(sbi)->main_blkaddr))
#define SEG0_BLKADDR(sbi)						\
	(SM_I(sbi) ? SM_I(sbi)->seg0_blkaddr : 				\
		le32_to_cpu(F2FS_RAW_SUPER(sbi)->segment0_blkaddr))
72 73

#define MAIN_SEGS(sbi)	(SM_I(sbi)->main_segments)
74
#define MAIN_SECS(sbi)	((sbi)->total_sections)
75

76 77 78
#define TOTAL_SEGS(sbi)							\
	(SM_I(sbi) ? SM_I(sbi)->segment_count : 				\
		le32_to_cpu(F2FS_RAW_SUPER(sbi)->segment_count))
79
#define TOTAL_BLKS(sbi)	(TOTAL_SEGS(sbi) << (sbi)->log_blocks_per_seg)
80 81

#define MAX_BLKADDR(sbi)	(SEG0_BLKADDR(sbi) + TOTAL_BLKS(sbi))
82 83
#define SEGMENT_SIZE(sbi)	(1ULL << ((sbi)->log_blocksize +	\
					(sbi)->log_blocks_per_seg))
84 85

#define START_BLOCK(sbi, segno)	(SEG0_BLKADDR(sbi) +			\
86
	 (GET_R2L_SEGNO(FREE_I(sbi), segno) << (sbi)->log_blocks_per_seg))
87

88
#define NEXT_FREE_BLKADDR(sbi, curseg)					\
89
	(START_BLOCK(sbi, (curseg)->segno) + (curseg)->next_blkoff)
90

91
#define GET_SEGOFF_FROM_SEG0(sbi, blk_addr)	((blk_addr) - SEG0_BLKADDR(sbi))
92
#define GET_SEGNO_FROM_SEG0(sbi, blk_addr)				\
93
	(GET_SEGOFF_FROM_SEG0(sbi, blk_addr) >> (sbi)->log_blocks_per_seg)
Jaegeuk Kim's avatar
Jaegeuk Kim committed
94
#define GET_BLKOFF_FROM_SEG0(sbi, blk_addr)				\
95
	(GET_SEGOFF_FROM_SEG0(sbi, blk_addr) & ((sbi)->blocks_per_seg - 1))
Jaegeuk Kim's avatar
Jaegeuk Kim committed
96

97
#define GET_SEGNO(sbi, blk_addr)					\
98
	((!__is_valid_data_blkaddr(blk_addr)) ?			\
99 100
	NULL_SEGNO : GET_L2R_SEGNO(FREE_I(sbi),			\
		GET_SEGNO_FROM_SEG0(sbi, blk_addr)))
101 102 103
#define BLKS_PER_SEC(sbi)					\
	((sbi)->segs_per_sec * (sbi)->blocks_per_seg)
#define GET_SEC_FROM_SEG(sbi, segno)				\
104
	((segno) / (sbi)->segs_per_sec)
105
#define GET_SEG_FROM_SEC(sbi, secno)				\
106
	((secno) * (sbi)->segs_per_sec)
107 108 109 110
#define GET_ZONE_FROM_SEC(sbi, secno)				\
	((secno) / (sbi)->secs_per_zone)
#define GET_ZONE_FROM_SEG(sbi, segno)				\
	GET_ZONE_FROM_SEC(sbi, GET_SEC_FROM_SEG(sbi, segno))
111 112

#define GET_SUM_BLOCK(sbi, segno)				\
113
	((sbi)->sm_info->ssa_blkaddr + (segno))
114 115

#define GET_SUM_TYPE(footer) ((footer)->entry_type)
116
#define SET_SUM_TYPE(footer, type) ((footer)->entry_type = (type))
117 118

#define SIT_ENTRY_OFFSET(sit_i, segno)					\
119
	((segno) % (sit_i)->sents_per_block)
120
#define SIT_BLOCK_OFFSET(segno)					\
121
	((segno) / SIT_ENTRY_PER_BLOCK)
122 123
#define	START_SEGNO(segno)		\
	(SIT_BLOCK_OFFSET(segno) * SIT_ENTRY_PER_BLOCK)
124
#define SIT_BLK_CNT(sbi)			\
125
	DIV_ROUND_UP(MAIN_SEGS(sbi), SIT_ENTRY_PER_BLOCK)
126 127 128
#define f2fs_bitmap_size(nr)			\
	(BITS_TO_LONGS(nr) * sizeof(unsigned long))

Chao Yu's avatar
Chao Yu committed
129 130 131
#define SECTOR_FROM_BLOCK(blk_addr)					\
	(((sector_t)blk_addr) << F2FS_LOG_SECTORS_PER_BLOCK)
#define SECTOR_TO_BLOCK(sectors)					\
132
	((sectors) >> F2FS_LOG_SECTORS_PER_BLOCK)
133

134 135 136 137 138 139 140 141 142 143 144 145 146 147
/*
 * indicate a block allocation direction: RIGHT and LEFT.
 * RIGHT means allocating new sections towards the end of volume.
 * LEFT means the opposite direction.
 */
enum {
	ALLOC_RIGHT = 0,
	ALLOC_LEFT
};

/*
 * In the victim_sel_policy->alloc_mode, there are two block allocation modes.
 * LFS writes data sequentially with cleaning operations.
 * SSR (Slack Space Recycle) reuses obsolete space without cleaning operations.
148 149
 * AT_SSR (Age Threshold based Slack Space Recycle) merges fragments into
 * fragmented segment which has similar aging degree.
150 151 152
 */
enum {
	LFS = 0,
153 154
	SSR,
	AT_SSR,
155 156 157 158 159 160
};

/*
 * In the victim_sel_policy->gc_mode, there are two gc, aka cleaning, modes.
 * GC_CB is based on cost-benefit algorithm.
 * GC_GREEDY is based on greedy algorithm.
161
 * GC_AT is based on age-threshold algorithm.
162 163 164
 */
enum {
	GC_CB = 0,
165
	GC_GREEDY,
166
	GC_AT,
167 168 169
	ALLOC_NEXT,
	FLUSH_DEVICE,
	MAX_GC_POLICY,
170 171 172 173 174
};

/*
 * BG_GC means the background cleaning job.
 * FG_GC means the on-demand cleaning job.
175
 * FORCE_FG_GC means on-demand cleaning job in background.
176 177 178
 */
enum {
	BG_GC = 0,
179 180
	FG_GC,
	FORCE_FG_GC,
181 182 183 184 185 186
};

/* for a function parameter to select a victim segment */
struct victim_sel_policy {
	int alloc_mode;			/* LFS or SSR */
	int gc_mode;			/* GC_CB or GC_GREEDY */
187 188 189 190 191
	unsigned long *dirty_bitmap;	/* dirty segment/section bitmap */
	unsigned int max_search;	/*
					 * maximum # of segments/sections
					 * to search
					 */
192 193 194
	unsigned int offset;		/* last scanned bitmap offset */
	unsigned int ofs_unit;		/* bitmap search unit */
	unsigned int min_cost;		/* minimum cost */
195
	unsigned long long oldest_age;	/* oldest age of segments having the same min cost */
196
	unsigned int min_segno;		/* segment # having min. cost */
197 198
	unsigned long long age;		/* mtime of GCed section*/
	unsigned long long age_threshold;/* age threshold */
199 200 201
};

struct seg_entry {
202 203 204 205
	unsigned int type:6;		/* segment type like CURSEG_XXX_TYPE */
	unsigned int valid_blocks:10;	/* # of valid blocks */
	unsigned int ckpt_valid_blocks:10;	/* # of valid blocks last cp */
	unsigned int padding:6;		/* padding */
206
	unsigned char *cur_valid_map;	/* validity bitmap of blocks */
207 208 209
#ifdef CONFIG_F2FS_CHECK_FS
	unsigned char *cur_valid_map_mir;	/* mirror of current valid bitmap */
#endif
210
	/*
211
	 * # of valid blocks and the validity bitmap stored in the last
212 213
	 * checkpoint pack. This information is used by the SSR mode.
	 */
214
	unsigned char *ckpt_valid_map;	/* validity bitmap of blocks last cp */
215
	unsigned char *discard_map;
216 217 218 219 220 221 222 223 224 225 226
	unsigned long long mtime;	/* modification time of the segment */
};

struct sec_entry {
	unsigned int valid_blocks;	/* # of valid blocks in a section */
};

struct segment_allocation {
	void (*allocate_segment)(struct f2fs_sb_info *, int, bool);
};

227
#define MAX_SKIP_GC_COUNT			16
228

Jaegeuk Kim's avatar
Jaegeuk Kim committed
229 230 231
struct inmem_pages {
	struct list_head list;
	struct page *page;
232
	block_t old_addr;		/* for revoking when fail to commit */
Jaegeuk Kim's avatar
Jaegeuk Kim committed
233 234
};

235 236 237 238 239 240
struct sit_info {
	const struct segment_allocation *s_ops;

	block_t sit_base_addr;		/* start block address of SIT area */
	block_t sit_blocks;		/* # of blocks used by SIT area */
	block_t written_valid_blocks;	/* # of valid blocks in main area */
241
	char *bitmap;			/* all bitmaps pointer */
242
	char *sit_bitmap;		/* SIT bitmap pointer */
243 244
#ifdef CONFIG_F2FS_CHECK_FS
	char *sit_bitmap_mir;		/* SIT bitmap mirror */
245 246 247

	/* bitmap of segments to be ignored by GC in case of errors */
	unsigned long *invalid_segmap;
248
#endif
249 250
	unsigned int bitmap_size;	/* SIT bitmap size */

251
	unsigned long *tmp_map;			/* bitmap for temporal use */
252 253 254
	unsigned long *dirty_sentries_bitmap;	/* bitmap for dirty sentries */
	unsigned int dirty_sentries;		/* # of dirty sentries */
	unsigned int sents_per_block;		/* # of SIT entries per block */
255
	struct rw_semaphore sentry_lock;	/* to protect SIT cache */
256 257 258 259 260 261 262 263
	struct seg_entry *sentries;		/* SIT segment-level cache */
	struct sec_entry *sec_entries;		/* SIT section-level cache */

	/* for cost-benefit algorithm in cleaning procedure */
	unsigned long long elapsed_time;	/* elapsed time after mount */
	unsigned long long mounted_time;	/* mount time */
	unsigned long long min_mtime;		/* min. modification time */
	unsigned long long max_mtime;		/* max. modification time */
264 265
	unsigned long long dirty_min_mtime;	/* rerange candidates in GC_AT */
	unsigned long long dirty_max_mtime;	/* rerange candidates in GC_AT */
266 267

	unsigned int last_victim[MAX_GC_POLICY]; /* last victim segment # */
268 269 270 271 272 273
};

struct free_segmap_info {
	unsigned int start_segno;	/* start segment number logically */
	unsigned int free_segments;	/* # of free segments */
	unsigned int free_sections;	/* # of free sections */
274
	spinlock_t segmap_lock;		/* free segmap lock */
275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294
	unsigned long *free_segmap;	/* free segment bitmap */
	unsigned long *free_secmap;	/* free section bitmap */
};

/* Notice: The order of dirty type is same with CURSEG_XXX in f2fs.h */
enum dirty_type {
	DIRTY_HOT_DATA,		/* dirty segments assigned as hot data logs */
	DIRTY_WARM_DATA,	/* dirty segments assigned as warm data logs */
	DIRTY_COLD_DATA,	/* dirty segments assigned as cold data logs */
	DIRTY_HOT_NODE,		/* dirty segments assigned as hot node logs */
	DIRTY_WARM_NODE,	/* dirty segments assigned as warm node logs */
	DIRTY_COLD_NODE,	/* dirty segments assigned as cold node logs */
	DIRTY,			/* to count # of dirty segments */
	PRE,			/* to count # of entirely obsolete segments */
	NR_DIRTY_TYPE
};

struct dirty_seglist_info {
	const struct victim_selection *v_ops;	/* victim selction operation */
	unsigned long *dirty_segmap[NR_DIRTY_TYPE];
295
	unsigned long *dirty_secmap;
296 297
	struct mutex seglist_lock;		/* lock for segment bitmaps */
	int nr_dirty[NR_DIRTY_TYPE];		/* # of dirty segments */
298
	unsigned long *victim_secmap;		/* background GC victims */
299 300 301 302 303
};

/* victim selection function for cleaning and SSR */
struct victim_selection {
	int (*get_victim)(struct f2fs_sb_info *, unsigned int *,
304
					int, int, char, unsigned long long);
305 306 307 308 309 310
};

/* for active log information */
struct curseg_info {
	struct mutex curseg_mutex;		/* lock for consistency */
	struct f2fs_summary_block *sum_blk;	/* cached summary block */
311 312
	struct rw_semaphore journal_rwsem;	/* protect journal area */
	struct f2fs_journal *journal;		/* cached journal info */
313
	unsigned char alloc_type;		/* current allocation type */
Chao Yu's avatar
Chao Yu committed
314
	unsigned short seg_type;		/* segment type like CURSEG_XXX_TYPE */
315 316 317 318
	unsigned int segno;			/* current segment number */
	unsigned short next_blkoff;		/* next block offset to write */
	unsigned int zone;			/* current zone number */
	unsigned int next_segno;		/* preallocated segment */
Chao Yu's avatar
Chao Yu committed
319
	bool inited;				/* indicate inmem log is inited */
320 321
};

322 323 324 325 326 327
struct sit_entry_set {
	struct list_head set_list;	/* link with all sit sets */
	unsigned int start_segno;	/* start segno of sits in set */
	unsigned int entry_cnt;		/* the # of sit entries in set */
};

328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346
/*
 * inline functions
 */
static inline struct curseg_info *CURSEG_I(struct f2fs_sb_info *sbi, int type)
{
	return (struct curseg_info *)(SM_I(sbi)->curseg_array + type);
}

static inline struct seg_entry *get_seg_entry(struct f2fs_sb_info *sbi,
						unsigned int segno)
{
	struct sit_info *sit_i = SIT_I(sbi);
	return &sit_i->sentries[segno];
}

static inline struct sec_entry *get_sec_entry(struct f2fs_sb_info *sbi,
						unsigned int segno)
{
	struct sit_info *sit_i = SIT_I(sbi);
347
	return &sit_i->sec_entries[GET_SEC_FROM_SEG(sbi, segno)];
348 349 350
}

static inline unsigned int get_valid_blocks(struct f2fs_sb_info *sbi,
351
				unsigned int segno, bool use_section)
352 353 354 355 356
{
	/*
	 * In order to get # of valid blocks in a section instantly from many
	 * segments, f2fs manages two counting structures separately.
	 */
357
	if (use_section && __is_large_section(sbi))
358 359 360 361 362
		return get_sec_entry(sbi, segno)->valid_blocks;
	else
		return get_seg_entry(sbi, segno)->valid_blocks;
}

363 364 365 366 367 368
static inline unsigned int get_ckpt_valid_blocks(struct f2fs_sb_info *sbi,
				unsigned int segno)
{
	return get_seg_entry(sbi, segno)->ckpt_valid_blocks;
}

369 370 371 372 373 374 375
static inline void seg_info_from_raw_sit(struct seg_entry *se,
					struct f2fs_sit_entry *rs)
{
	se->valid_blocks = GET_SIT_VBLOCKS(rs);
	se->ckpt_valid_blocks = GET_SIT_VBLOCKS(rs);
	memcpy(se->cur_valid_map, rs->valid_map, SIT_VBLOCK_MAP_SIZE);
	memcpy(se->ckpt_valid_map, rs->valid_map, SIT_VBLOCK_MAP_SIZE);
376 377 378
#ifdef CONFIG_F2FS_CHECK_FS
	memcpy(se->cur_valid_map_mir, rs->valid_map, SIT_VBLOCK_MAP_SIZE);
#endif
379 380 381 382
	se->type = GET_SIT_TYPE(rs);
	se->mtime = le64_to_cpu(rs->mtime);
}

383
static inline void __seg_info_to_raw_sit(struct seg_entry *se,
384 385 386 387 388 389
					struct f2fs_sit_entry *rs)
{
	unsigned short raw_vblocks = (se->type << SIT_VBLOCKS_SHIFT) |
					se->valid_blocks;
	rs->vblocks = cpu_to_le16(raw_vblocks);
	memcpy(rs->valid_map, se->cur_valid_map, SIT_VBLOCK_MAP_SIZE);
390 391 392 393 394 395 396 397 398 399 400 401 402 403
	rs->mtime = cpu_to_le64(se->mtime);
}

static inline void seg_info_to_sit_page(struct f2fs_sb_info *sbi,
				struct page *page, unsigned int start)
{
	struct f2fs_sit_block *raw_sit;
	struct seg_entry *se;
	struct f2fs_sit_entry *rs;
	unsigned int end = min(start + SIT_ENTRY_PER_BLOCK,
					(unsigned long)MAIN_SEGS(sbi));
	int i;

	raw_sit = (struct f2fs_sit_block *)page_address(page);
404
	memset(raw_sit, 0, PAGE_SIZE);
405 406 407 408 409 410 411 412 413 414 415 416
	for (i = 0; i < end - start; i++) {
		rs = &raw_sit->entries[i];
		se = get_seg_entry(sbi, start + i);
		__seg_info_to_raw_sit(se, rs);
	}
}

static inline void seg_info_to_raw_sit(struct seg_entry *se,
					struct f2fs_sit_entry *rs)
{
	__seg_info_to_raw_sit(se, rs);

417 418 419 420 421 422 423 424
	memcpy(se->ckpt_valid_map, rs->valid_map, SIT_VBLOCK_MAP_SIZE);
	se->ckpt_valid_blocks = se->valid_blocks;
}

static inline unsigned int find_next_inuse(struct free_segmap_info *free_i,
		unsigned int max, unsigned int segno)
{
	unsigned int ret;
425
	spin_lock(&free_i->segmap_lock);
426
	ret = find_next_bit(free_i->free_segmap, max, segno);
427
	spin_unlock(&free_i->segmap_lock);
428 429 430 431 432 433
	return ret;
}

static inline void __set_free(struct f2fs_sb_info *sbi, unsigned int segno)
{
	struct free_segmap_info *free_i = FREE_I(sbi);
434 435
	unsigned int secno = GET_SEC_FROM_SEG(sbi, segno);
	unsigned int start_segno = GET_SEG_FROM_SEC(sbi, secno);
436
	unsigned int next;
437
	unsigned int usable_segs = f2fs_usable_segs_in_sec(sbi, segno);
438

439
	spin_lock(&free_i->segmap_lock);
440 441 442
	clear_bit(segno, free_i->free_segmap);
	free_i->free_segments++;

443 444
	next = find_next_bit(free_i->free_segmap,
			start_segno + sbi->segs_per_sec, start_segno);
445
	if (next >= start_segno + usable_segs) {
446 447 448
		clear_bit(secno, free_i->free_secmap);
		free_i->free_sections++;
	}
449
	spin_unlock(&free_i->segmap_lock);
450 451 452 453 454 455
}

static inline void __set_inuse(struct f2fs_sb_info *sbi,
		unsigned int segno)
{
	struct free_segmap_info *free_i = FREE_I(sbi);
456 457
	unsigned int secno = GET_SEC_FROM_SEG(sbi, segno);

458 459 460 461 462 463 464
	set_bit(segno, free_i->free_segmap);
	free_i->free_segments--;
	if (!test_and_set_bit(secno, free_i->free_secmap))
		free_i->free_sections--;
}

static inline void __set_test_and_free(struct f2fs_sb_info *sbi,
Chao Yu's avatar
Chao Yu committed
465
		unsigned int segno, bool inmem)
466 467
{
	struct free_segmap_info *free_i = FREE_I(sbi);
468 469
	unsigned int secno = GET_SEC_FROM_SEG(sbi, segno);
	unsigned int start_segno = GET_SEG_FROM_SEC(sbi, secno);
470
	unsigned int next;
471
	unsigned int usable_segs = f2fs_usable_segs_in_sec(sbi, segno);
472

473
	spin_lock(&free_i->segmap_lock);
474 475 476
	if (test_and_clear_bit(segno, free_i->free_segmap)) {
		free_i->free_segments++;

Chao Yu's avatar
Chao Yu committed
477
		if (!inmem && IS_CURSEC(sbi, secno))
478
			goto skip_free;
479 480
		next = find_next_bit(free_i->free_segmap,
				start_segno + sbi->segs_per_sec, start_segno);
481
		if (next >= start_segno + usable_segs) {
482 483 484 485
			if (test_and_clear_bit(secno, free_i->free_secmap))
				free_i->free_sections++;
		}
	}
486
skip_free:
487
	spin_unlock(&free_i->segmap_lock);
488 489 490 491 492 493
}

static inline void __set_test_and_inuse(struct f2fs_sb_info *sbi,
		unsigned int segno)
{
	struct free_segmap_info *free_i = FREE_I(sbi);
494 495
	unsigned int secno = GET_SEC_FROM_SEG(sbi, segno);

496
	spin_lock(&free_i->segmap_lock);
497 498 499 500 501
	if (!test_and_set_bit(segno, free_i->free_segmap)) {
		free_i->free_segments--;
		if (!test_and_set_bit(secno, free_i->free_secmap))
			free_i->free_sections--;
	}
502
	spin_unlock(&free_i->segmap_lock);
503 504 505 506 507 508
}

static inline void get_sit_bitmap(struct f2fs_sb_info *sbi,
		void *dst_addr)
{
	struct sit_info *sit_i = SIT_I(sbi);
509 510 511 512 513 514

#ifdef CONFIG_F2FS_CHECK_FS
	if (memcmp(sit_i->sit_bitmap, sit_i->sit_bitmap_mir,
						sit_i->bitmap_size))
		f2fs_bug_on(sbi, 1);
#endif
515 516 517 518 519
	memcpy(dst_addr, sit_i->sit_bitmap, sit_i->bitmap_size);
}

static inline block_t written_block_count(struct f2fs_sb_info *sbi)
{
520
	return SIT_I(sbi)->written_valid_blocks;
521 522 523 524
}

static inline unsigned int free_segments(struct f2fs_sb_info *sbi)
{
525
	return FREE_I(sbi)->free_segments;
526 527
}

528
static inline unsigned int reserved_segments(struct f2fs_sb_info *sbi)
529 530 531 532 533 534
{
	return SM_I(sbi)->reserved_segments;
}

static inline unsigned int free_sections(struct f2fs_sb_info *sbi)
{
535
	return FREE_I(sbi)->free_sections;
536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559
}

static inline unsigned int prefree_segments(struct f2fs_sb_info *sbi)
{
	return DIRTY_I(sbi)->nr_dirty[PRE];
}

static inline unsigned int dirty_segments(struct f2fs_sb_info *sbi)
{
	return DIRTY_I(sbi)->nr_dirty[DIRTY_HOT_DATA] +
		DIRTY_I(sbi)->nr_dirty[DIRTY_WARM_DATA] +
		DIRTY_I(sbi)->nr_dirty[DIRTY_COLD_DATA] +
		DIRTY_I(sbi)->nr_dirty[DIRTY_HOT_NODE] +
		DIRTY_I(sbi)->nr_dirty[DIRTY_WARM_NODE] +
		DIRTY_I(sbi)->nr_dirty[DIRTY_COLD_NODE];
}

static inline int overprovision_segments(struct f2fs_sb_info *sbi)
{
	return SM_I(sbi)->ovp_segments;
}

static inline int reserved_sections(struct f2fs_sb_info *sbi)
{
560
	return GET_SEC_FROM_SEG(sbi, reserved_segments(sbi));
561 562
}

563 564 565 566 567 568 569 570 571 572 573
static inline bool has_curseg_enough_space(struct f2fs_sb_info *sbi)
{
	unsigned int node_blocks = get_pages(sbi, F2FS_DIRTY_NODES) +
					get_pages(sbi, F2FS_DIRTY_DENTS);
	unsigned int dent_blocks = get_pages(sbi, F2FS_DIRTY_DENTS);
	unsigned int segno, left_blocks;
	int i;

	/* check current node segment */
	for (i = CURSEG_HOT_NODE; i <= CURSEG_COLD_NODE; i++) {
		segno = CURSEG_I(sbi, i)->segno;
574 575
		left_blocks = f2fs_usable_blks_in_seg(sbi, segno) -
				get_seg_entry(sbi, segno)->ckpt_valid_blocks;
576 577 578 579 580 581 582

		if (node_blocks > left_blocks)
			return false;
	}

	/* check current data segment */
	segno = CURSEG_I(sbi, CURSEG_HOT_DATA)->segno;
583
	left_blocks = f2fs_usable_blks_in_seg(sbi, segno) -
584 585 586 587 588 589
			get_seg_entry(sbi, segno)->ckpt_valid_blocks;
	if (dent_blocks > left_blocks)
		return false;
	return true;
}

590 591
static inline bool has_not_enough_free_secs(struct f2fs_sb_info *sbi,
					int freed, int needed)
592
{
593 594
	int node_secs = get_blocktype_secs(sbi, F2FS_DIRTY_NODES);
	int dent_secs = get_blocktype_secs(sbi, F2FS_DIRTY_DENTS);
595
	int imeta_secs = get_blocktype_secs(sbi, F2FS_DIRTY_IMETA);
596

597
	if (unlikely(is_sbi_flag_set(sbi, SBI_POR_DOING)))
598 599
		return false;

600 601 602
	if (free_sections(sbi) + freed == reserved_sections(sbi) + needed &&
			has_curseg_enough_space(sbi))
		return false;
603
	return (free_sections(sbi) + freed) <=
604 605
		(node_secs + 2 * dent_secs + imeta_secs +
		reserved_sections(sbi) + needed);
606 607
}

608
static inline bool f2fs_is_checkpoint_ready(struct f2fs_sb_info *sbi)
609 610
{
	if (likely(!is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
611
		return true;
612
	if (likely(!has_not_enough_free_secs(sbi, 0, 0)))
613 614
		return true;
	return false;
615 616
}

617 618
static inline bool excess_prefree_segs(struct f2fs_sb_info *sbi)
{
619
	return prefree_segments(sbi) > SM_I(sbi)->rec_prefree_segments;
620 621
}

622 623
static inline int utilization(struct f2fs_sb_info *sbi)
{
624 625
	return div_u64((u64)valid_user_blocks(sbi) * 100,
					sbi->user_block_count);
626 627 628 629
}

/*
 * Sometimes f2fs may be better to drop out-of-place update policy.
630 631 632 633 634 635 636
 * And, users can control the policy through sysfs entries.
 * There are five policies with triggering conditions as follows.
 * F2FS_IPU_FORCE - all the time,
 * F2FS_IPU_SSR - if SSR mode is activated,
 * F2FS_IPU_UTIL - if FS utilization is over threashold,
 * F2FS_IPU_SSR_UTIL - if SSR mode is activated and FS utilization is over
 *                     threashold,
637 638
 * F2FS_IPU_FSYNC - activated in fsync path only for high performance flash
 *                     storages. IPU will be triggered only if the # of dirty
639 640
 *                     pages over min_fsync_blocks. (=default option)
 * F2FS_IPU_ASYNC - do IPU given by asynchronous write requests.
641
 * F2FS_IPU_NOCACHE - disable IPU bio cache.
642
 * F2FS_IPUT_DISABLE - disable IPU. (=default option in LFS mode)
643
 */
644
#define DEF_MIN_IPU_UTIL	70
645
#define DEF_MIN_FSYNC_BLOCKS	8
646
#define DEF_MIN_HOT_BLOCKS	16
647

648 649
#define SMALL_VOLUME_SEGMENTS	(16 * 512)	/* 16GB */

650 651 652 653 654
enum {
	F2FS_IPU_FORCE,
	F2FS_IPU_SSR,
	F2FS_IPU_UTIL,
	F2FS_IPU_SSR_UTIL,
655
	F2FS_IPU_FSYNC,
656
	F2FS_IPU_ASYNC,
657
	F2FS_IPU_NOCACHE,
658 659
};

660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681
static inline unsigned int curseg_segno(struct f2fs_sb_info *sbi,
		int type)
{
	struct curseg_info *curseg = CURSEG_I(sbi, type);
	return curseg->segno;
}

static inline unsigned char curseg_alloc_type(struct f2fs_sb_info *sbi,
		int type)
{
	struct curseg_info *curseg = CURSEG_I(sbi, type);
	return curseg->alloc_type;
}

static inline unsigned short curseg_blkoff(struct f2fs_sb_info *sbi, int type)
{
	struct curseg_info *curseg = CURSEG_I(sbi, type);
	return curseg->next_blkoff;
}

static inline void check_seg_range(struct f2fs_sb_info *sbi, unsigned int segno)
{
682
	f2fs_bug_on(sbi, segno > TOTAL_SEGS(sbi) - 1);
683 684
}

685
static inline void verify_fio_blkaddr(struct f2fs_io_info *fio)
686
{
687 688
	struct f2fs_sb_info *sbi = fio->sbi;

689 690 691 692 693
	if (__is_valid_data_blkaddr(fio->old_blkaddr))
		verify_blkaddr(sbi, fio->old_blkaddr, __is_meta_io(fio) ?
					META_GENERIC : DATA_GENERIC);
	verify_blkaddr(sbi, fio->new_blkaddr, __is_meta_io(fio) ?
					META_GENERIC : DATA_GENERIC_ENHANCE);
694 695 696
}

/*
arter97's avatar
arter97 committed
697
 * Summary block is always treated as an invalid block
698
 */
699
static inline int check_block_count(struct f2fs_sb_info *sbi,
700 701
		int segno, struct f2fs_sit_entry *raw_sit)
{
702
	bool is_valid  = test_bit_le(0, raw_sit->valid_map) ? true : false;
703
	int valid_blocks = 0;
704
	int cur_pos = 0, next_pos;
705
	unsigned int usable_blks_per_seg = f2fs_usable_blks_in_seg(sbi, segno);
706 707

	/* check bitmap with valid block count */
708 709 710
	do {
		if (is_valid) {
			next_pos = find_next_zero_bit_le(&raw_sit->valid_map,
711
					usable_blks_per_seg,
712 713 714 715
					cur_pos);
			valid_blocks += next_pos - cur_pos;
		} else
			next_pos = find_next_bit_le(&raw_sit->valid_map,
716
					usable_blks_per_seg,
717 718 719
					cur_pos);
		cur_pos = next_pos;
		is_valid = !is_valid;
720
	} while (cur_pos < usable_blks_per_seg);
721 722

	if (unlikely(GET_SIT_VBLOCKS(raw_sit) != valid_blocks)) {
723 724
		f2fs_err(sbi, "Mismatch valid blocks %d vs. %d",
			 GET_SIT_VBLOCKS(raw_sit), valid_blocks);
725
		set_sbi_flag(sbi, SBI_NEED_FSCK);
726
		return -EFSCORRUPTED;
727
	}
728

729 730 731 732 733
	if (usable_blks_per_seg < sbi->blocks_per_seg)
		f2fs_bug_on(sbi, find_next_bit_le(&raw_sit->valid_map,
				sbi->blocks_per_seg,
				usable_blks_per_seg) != sbi->blocks_per_seg);

Jaegeuk Kim's avatar
Jaegeuk Kim committed
734
	/* check segment usage, and check boundary of a given segment number */
735
	if (unlikely(GET_SIT_VBLOCKS(raw_sit) > usable_blks_per_seg
736
					|| segno > TOTAL_SEGS(sbi) - 1)) {
737 738
		f2fs_err(sbi, "Wrong valid blocks %d or segno %u",
			 GET_SIT_VBLOCKS(raw_sit), segno);
739
		set_sbi_flag(sbi, SBI_NEED_FSCK);
740
		return -EFSCORRUPTED;
741 742
	}
	return 0;
743
}
744 745 746 747 748

static inline pgoff_t current_sit_addr(struct f2fs_sb_info *sbi,
						unsigned int start)
{
	struct sit_info *sit_i = SIT_I(sbi);
749
	unsigned int offset = SIT_BLOCK_OFFSET(start);
750 751 752 753
	block_t blk_addr = sit_i->sit_base_addr + offset;

	check_seg_range(sbi, start);

754 755 756 757 758 759
#ifdef CONFIG_F2FS_CHECK_FS
	if (f2fs_test_bit(offset, sit_i->sit_bitmap) !=
			f2fs_test_bit(offset, sit_i->sit_bitmap_mir))
		f2fs_bug_on(sbi, 1);
#endif

760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781
	/* calculate sit block address */
	if (f2fs_test_bit(offset, sit_i->sit_bitmap))
		blk_addr += sit_i->sit_blocks;

	return blk_addr;
}

static inline pgoff_t next_sit_addr(struct f2fs_sb_info *sbi,
						pgoff_t block_addr)
{
	struct sit_info *sit_i = SIT_I(sbi);
	block_addr -= sit_i->sit_base_addr;
	if (block_addr < sit_i->sit_blocks)
		block_addr += sit_i->sit_blocks;
	else
		block_addr -= sit_i->sit_blocks;

	return block_addr + sit_i->sit_base_addr;
}

static inline void set_to_next_sit(struct sit_info *sit_i, unsigned int start)
{
782
	unsigned int block_off = SIT_BLOCK_OFFSET(start);
783

784
	f2fs_change_bit(block_off, sit_i->sit_bitmap);
785 786 787
#ifdef CONFIG_F2FS_CHECK_FS
	f2fs_change_bit(block_off, sit_i->sit_bitmap_mir);
#endif
788 789
}

790 791
static inline unsigned long long get_mtime(struct f2fs_sb_info *sbi,
						bool base_time)
792 793
{
	struct sit_info *sit_i = SIT_I(sbi);
Jaegeuk Kim's avatar
Jaegeuk Kim committed
794
	time64_t diff, now = ktime_get_boottime_seconds();
795

796 797 798 799 800 801 802 803 804 805 806
	if (now >= sit_i->mounted_time)
		return sit_i->elapsed_time + now - sit_i->mounted_time;

	/* system time is set to the past */
	if (!base_time) {
		diff = sit_i->mounted_time - now;
		if (sit_i->elapsed_time >= diff)
			return sit_i->elapsed_time - diff;
		return 0;
	}
	return sit_i->elapsed_time;
807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828
}

static inline void set_summary(struct f2fs_summary *sum, nid_t nid,
			unsigned int ofs_in_node, unsigned char version)
{
	sum->nid = cpu_to_le32(nid);
	sum->ofs_in_node = cpu_to_le16(ofs_in_node);
	sum->version = version;
}

static inline block_t start_sum_block(struct f2fs_sb_info *sbi)
{
	return __start_cp_addr(sbi) +
		le32_to_cpu(F2FS_CKPT(sbi)->cp_pack_start_sum);
}

static inline block_t sum_blk_addr(struct f2fs_sb_info *sbi, int base, int type)
{
	return __start_cp_addr(sbi) +
		le32_to_cpu(F2FS_CKPT(sbi)->cp_pack_total_block_count)
				- (base + 1) + type;
}
829 830 831 832 833 834 835

static inline bool sec_usage_check(struct f2fs_sb_info *sbi, unsigned int secno)
{
	if (IS_CURSEC(sbi, secno) || (sbi->cur_victim_sec == secno))
		return true;
	return false;
}
836

837 838 839 840
/*
 * It is very important to gather dirty pages and write at once, so that we can
 * submit a big bio without interfering other data writes.
 * By default, 512 pages for directory data,
841 842
 * 512 pages (2MB) * 8 for nodes, and
 * 256 pages * 8 for meta are set.
843 844 845
 */
static inline int nr_pages_to_skip(struct f2fs_sb_info *sbi, int type)
{
846
	if (sbi->sb->s_bdi->wb.dirty_exceeded)
847 848
		return 0;

849 850 851
	if (type == DATA)
		return sbi->blocks_per_seg;
	else if (type == NODE)
852
		return 8 * sbi->blocks_per_seg;
853
	else if (type == META)
854
		return 8 * BIO_MAX_PAGES;
855 856 857
	else
		return 0;
}
858 859 860 861 862 863 864 865 866 867 868 869 870

/*
 * When writing pages, it'd better align nr_to_write for segment size.
 */
static inline long nr_pages_to_write(struct f2fs_sb_info *sbi, int type,
					struct writeback_control *wbc)
{
	long nr_to_write, desired;

	if (wbc->sync_mode != WB_SYNC_NONE)
		return 0;

	nr_to_write = wbc->nr_to_write;
871
	desired = BIO_MAX_PAGES;
872
	if (type == NODE)
873
		desired <<= 1;
874 875 876 877

	wbc->nr_to_write = desired;
	return desired - nr_to_write;
}
878 879 880 881 882 883 884 885 886 887 888

static inline void wake_up_discard_thread(struct f2fs_sb_info *sbi, bool force)
{
	struct discard_cmd_control *dcc = SM_I(sbi)->dcc_info;
	bool wakeup = false;
	int i;

	if (force)
		goto wake_up;

	mutex_lock(&dcc->cmd_lock);
Chao Yu's avatar
Chao Yu committed
889 890 891
	for (i = MAX_PLIST_NUM - 1; i >= 0; i--) {
		if (i + 1 < dcc->discard_granularity)
			break;
892 893 894 895 896 897
		if (!list_empty(&dcc->pend_list[i])) {
			wakeup = true;
			break;
		}
	}
	mutex_unlock(&dcc->cmd_lock);
898
	if (!wakeup || !is_idle(sbi, DISCARD_TIME))
899 900 901 902 903
		return;
wake_up:
	dcc->discard_wake = 1;
	wake_up_interruptible_all(&dcc->discard_wait_queue);
}