blob: 8db50ddb0ec94cd6cf327c72ca9eddec20a97fbf [file] [log] [blame]
xjb04a4022021-11-25 15:01:52 +08001// SPDX-License-Identifier: GPL-2.0
2/*
3 * fs/f2fs/f2fs.h
4 *
5 * Copyright (c) 2012 Samsung Electronics Co., Ltd.
6 * http://www.samsung.com/
7 */
8#ifndef _LINUX_F2FS_H
9#define _LINUX_F2FS_H
10
11#include <linux/uio.h>
12#include <linux/types.h>
13#include <linux/page-flags.h>
14#include <linux/buffer_head.h>
15#include <linux/slab.h>
16#include <linux/crc32.h>
17#include <linux/magic.h>
18#include <linux/kobject.h>
19#include <linux/sched.h>
20#include <linux/cred.h>
21#include <linux/vmalloc.h>
22#include <linux/bio.h>
23#include <linux/blkdev.h>
24#include <linux/quotaops.h>
25#include <crypto/hash.h>
26
27#include <linux/fscrypt.h>
28#include <linux/fsverity.h>
29
30#ifdef CONFIG_F2FS_CHECK_FS
31#define f2fs_bug_on(sbi, condition) BUG_ON(condition)
32#else
33#define f2fs_bug_on(sbi, condition) \
34 do { \
35 if (unlikely(condition)) { \
36 WARN_ON(1); \
37 set_sbi_flag(sbi, SBI_NEED_FSCK); \
38 } \
39 } while (0)
40#endif
41
42enum {
43 FAULT_KMALLOC,
44 FAULT_KVMALLOC,
45 FAULT_PAGE_ALLOC,
46 FAULT_PAGE_GET,
47 FAULT_ALLOC_BIO,
48 FAULT_ALLOC_NID,
49 FAULT_ORPHAN,
50 FAULT_BLOCK,
51 FAULT_DIR_DEPTH,
52 FAULT_EVICT_INODE,
53 FAULT_TRUNCATE,
54 FAULT_READ_IO,
55 FAULT_CHECKPOINT,
56 FAULT_DISCARD,
57 FAULT_WRITE_IO,
58 FAULT_MAX,
59};
60
61#ifdef CONFIG_F2FS_FAULT_INJECTION
62#define F2FS_ALL_FAULT_TYPE ((1 << FAULT_MAX) - 1)
63
64struct f2fs_fault_info {
65 atomic_t inject_ops;
66 unsigned int inject_rate;
67 unsigned int inject_type;
68};
69
70extern const char *f2fs_fault_name[FAULT_MAX];
71#define IS_FAULT_SET(fi, type) ((fi)->inject_type & (1 << (type)))
72#endif
73
74/*
75 * For mount options
76 */
77#define F2FS_MOUNT_BG_GC 0x00000001
78#define F2FS_MOUNT_DISABLE_ROLL_FORWARD 0x00000002
79#define F2FS_MOUNT_DISCARD 0x00000004
80#define F2FS_MOUNT_NOHEAP 0x00000008
81#define F2FS_MOUNT_XATTR_USER 0x00000010
82#define F2FS_MOUNT_POSIX_ACL 0x00000020
83#define F2FS_MOUNT_DISABLE_EXT_IDENTIFY 0x00000040
84#define F2FS_MOUNT_INLINE_XATTR 0x00000080
85#define F2FS_MOUNT_INLINE_DATA 0x00000100
86#define F2FS_MOUNT_INLINE_DENTRY 0x00000200
87#define F2FS_MOUNT_FLUSH_MERGE 0x00000400
88#define F2FS_MOUNT_NOBARRIER 0x00000800
89#define F2FS_MOUNT_FASTBOOT 0x00001000
90#define F2FS_MOUNT_EXTENT_CACHE 0x00002000
91#define F2FS_MOUNT_FORCE_FG_GC 0x00004000
92#define F2FS_MOUNT_DATA_FLUSH 0x00008000
93#define F2FS_MOUNT_FAULT_INJECTION 0x00010000
94#define F2FS_MOUNT_ADAPTIVE 0x00020000
95#define F2FS_MOUNT_LFS 0x00040000
96#define F2FS_MOUNT_USRQUOTA 0x00080000
97#define F2FS_MOUNT_GRPQUOTA 0x00100000
98#define F2FS_MOUNT_PRJQUOTA 0x00200000
99#define F2FS_MOUNT_QUOTA 0x00400000
100#define F2FS_MOUNT_INLINE_XATTR_SIZE 0x00800000
101#define F2FS_MOUNT_RESERVE_ROOT 0x01000000
102#define F2FS_MOUNT_DISABLE_CHECKPOINT 0x02000000
103
104#define F2FS_OPTION(sbi) ((sbi)->mount_opt)
105#define clear_opt(sbi, option) (F2FS_OPTION(sbi).opt &= ~F2FS_MOUNT_##option)
106#define set_opt(sbi, option) (F2FS_OPTION(sbi).opt |= F2FS_MOUNT_##option)
107#define test_opt(sbi, option) (F2FS_OPTION(sbi).opt & F2FS_MOUNT_##option)
108
109#define ver_after(a, b) (typecheck(unsigned long long, a) && \
110 typecheck(unsigned long long, b) && \
111 ((long long)((a) - (b)) > 0))
112
113typedef u32 block_t; /*
114 * should not change u32, since it is the on-disk block
115 * address format, __le32.
116 */
117typedef u32 nid_t;
118
119struct f2fs_mount_info {
120 unsigned int opt;
121 int write_io_size_bits; /* Write IO size bits */
122 block_t root_reserved_blocks; /* root reserved blocks */
123 kuid_t s_resuid; /* reserved blocks for uid */
124 kgid_t s_resgid; /* reserved blocks for gid */
125 int active_logs; /* # of active logs */
126 int inline_xattr_size; /* inline xattr size */
127#ifdef CONFIG_F2FS_FAULT_INJECTION
128 struct f2fs_fault_info fault_info; /* For fault injection */
129#endif
130#ifdef CONFIG_QUOTA
131 /* Names of quota files with journalled quota */
132 char *s_qf_names[MAXQUOTAS];
133 int s_jquota_fmt; /* Format of quota to use */
134#endif
135 /* For which write hints are passed down to block layer */
136 int whint_mode;
137 int alloc_mode; /* segment allocation policy */
138 int fsync_mode; /* fsync policy */
139 bool test_dummy_encryption; /* test dummy encryption */
140#ifdef CONFIG_FS_ENCRYPTION
141 bool inlinecrypt; /* inline encryption enabled */
142#endif
143 block_t unusable_cap; /* Amount of space allowed to be
144 * unusable when disabling checkpoint
145 */
146};
147
148#define F2FS_FEATURE_ENCRYPT 0x0001
149#define F2FS_FEATURE_BLKZONED 0x0002
150#define F2FS_FEATURE_ATOMIC_WRITE 0x0004
151#define F2FS_FEATURE_EXTRA_ATTR 0x0008
152#define F2FS_FEATURE_PRJQUOTA 0x0010
153#define F2FS_FEATURE_INODE_CHKSUM 0x0020
154#define F2FS_FEATURE_FLEXIBLE_INLINE_XATTR 0x0040
155#define F2FS_FEATURE_QUOTA_INO 0x0080
156#define F2FS_FEATURE_INODE_CRTIME 0x0100
157#define F2FS_FEATURE_LOST_FOUND 0x0200
158#define F2FS_FEATURE_VERITY 0x0400
159#define F2FS_FEATURE_SB_CHKSUM 0x0800
160#define F2FS_FEATURE_CASEFOLD 0x1000
161
162#define __F2FS_HAS_FEATURE(raw_super, mask) \
163 ((raw_super->feature & cpu_to_le32(mask)) != 0)
164#define F2FS_HAS_FEATURE(sbi, mask) __F2FS_HAS_FEATURE(sbi->raw_super, mask)
165#define F2FS_SET_FEATURE(sbi, mask) \
166 (sbi->raw_super->feature |= cpu_to_le32(mask))
167#define F2FS_CLEAR_FEATURE(sbi, mask) \
168 (sbi->raw_super->feature &= ~cpu_to_le32(mask))
169
170/*
171 * Default values for user and/or group using reserved blocks
172 */
173#define F2FS_DEF_RESUID 0
174#define F2FS_DEF_RESGID 0
175
176/*
177 * For checkpoint manager
178 */
179enum {
180 NAT_BITMAP,
181 SIT_BITMAP
182};
183
184#define CP_UMOUNT 0x00000001
185#define CP_FASTBOOT 0x00000002
186#define CP_SYNC 0x00000004
187#define CP_RECOVERY 0x00000008
188#define CP_DISCARD 0x00000010
189#define CP_TRIMMED 0x00000020
190#define CP_PAUSE 0x00000040
191
192#define MAX_DISCARD_BLOCKS(sbi) BLKS_PER_SEC(sbi)
193#define DEF_MAX_DISCARD_REQUEST 8 /* issue 8 discards per round */
194#define DEF_MIN_DISCARD_ISSUE_TIME 50 /* 50 ms, if exists */
195#define DEF_MID_DISCARD_ISSUE_TIME 500 /* 500 ms, if device busy */
196#define DEF_MAX_DISCARD_ISSUE_TIME 60000 /* 60 s, if no candidates */
197#define DEF_DISCARD_URGENT_UTIL 80 /* do more discard over 80% */
198#define DEF_CP_INTERVAL 60 /* 60 secs */
199#define DEF_IDLE_INTERVAL 5 /* 5 secs */
200#define DEF_DISABLE_INTERVAL 5 /* 5 secs */
201#define DEF_DISABLE_QUICK_INTERVAL 1 /* 1 secs */
202#define DEF_UMOUNT_DISCARD_TIMEOUT 5 /* 5 secs */
203
204struct cp_control {
205 int reason;
206 __u64 trim_start;
207 __u64 trim_end;
208 __u64 trim_minlen;
209};
210
211/*
212 * indicate meta/data type
213 */
214enum {
215 META_CP,
216 META_NAT,
217 META_SIT,
218 META_SSA,
219 META_MAX,
220 META_POR,
221 DATA_GENERIC, /* check range only */
222 DATA_GENERIC_ENHANCE, /* strong check on range and segment bitmap */
223 DATA_GENERIC_ENHANCE_READ, /*
224 * strong check on range and segment
225 * bitmap but no warning due to race
226 * condition of read on truncated area
227 * by extent_cache
228 */
229 META_GENERIC,
230};
231
232/* for the list of ino */
233enum {
234 ORPHAN_INO, /* for orphan ino list */
235 APPEND_INO, /* for append ino list */
236 UPDATE_INO, /* for update ino list */
237 TRANS_DIR_INO, /* for trasactions dir ino list */
238 FLUSH_INO, /* for multiple device flushing */
239 MAX_INO_ENTRY, /* max. list */
240};
241
242struct ino_entry {
243 struct list_head list; /* list head */
244 nid_t ino; /* inode number */
245 unsigned int dirty_device; /* dirty device bitmap */
246};
247
248/* for the list of inodes to be GCed */
249struct inode_entry {
250 struct list_head list; /* list head */
251 struct inode *inode; /* vfs inode pointer */
252};
253
254struct fsync_node_entry {
255 struct list_head list; /* list head */
256 struct page *page; /* warm node page pointer */
257 unsigned int seq_id; /* sequence id */
258};
259
260/* for the bitmap indicate blocks to be discarded */
261struct discard_entry {
262 struct list_head list; /* list head */
263 block_t start_blkaddr; /* start blockaddr of current segment */
264 unsigned char discard_map[SIT_VBLOCK_MAP_SIZE]; /* segment discard bitmap */
265};
266
267/* default discard granularity of inner discard thread, unit: block count */
268#define DEFAULT_DISCARD_GRANULARITY 16
269
270/* max discard pend list number */
271#define MAX_PLIST_NUM 512
272#define plist_idx(blk_num) ((blk_num) >= MAX_PLIST_NUM ? \
273 (MAX_PLIST_NUM - 1) : ((blk_num) - 1))
274
275enum {
276 D_PREP, /* initial */
277 D_PARTIAL, /* partially submitted */
278 D_SUBMIT, /* all submitted */
279 D_DONE, /* finished */
280};
281
282struct discard_info {
283 block_t lstart; /* logical start address */
284 block_t len; /* length */
285 block_t start; /* actual start address in dev */
286};
287
288struct discard_cmd {
289 struct rb_node rb_node; /* rb node located in rb-tree */
290 union {
291 struct {
292 block_t lstart; /* logical start address */
293 block_t len; /* length */
294 block_t start; /* actual start address in dev */
295 };
296 struct discard_info di; /* discard info */
297
298 };
299 struct list_head list; /* command list */
300 struct completion wait; /* compleation */
301 struct block_device *bdev; /* bdev */
302 unsigned short ref; /* reference count */
303 unsigned char state; /* state */
304 unsigned char queued; /* queued discard */
305 int error; /* bio error */
306 spinlock_t lock; /* for state/bio_ref updating */
307 unsigned short bio_ref; /* bio reference count */
308};
309
310enum {
311 DPOLICY_BG,
312 DPOLICY_FORCE,
313 DPOLICY_FSTRIM,
314 DPOLICY_UMOUNT,
315 MAX_DPOLICY,
316};
317
318struct discard_policy {
319 int type; /* type of discard */
320 unsigned int min_interval; /* used for candidates exist */
321 unsigned int mid_interval; /* used for device busy */
322 unsigned int max_interval; /* used for candidates not exist */
323 unsigned int max_requests; /* # of discards issued per round */
324 unsigned int io_aware_gran; /* minimum granularity discard not be aware of I/O */
325 bool io_aware; /* issue discard in idle time */
326 bool sync; /* submit discard with REQ_SYNC flag */
327 bool ordered; /* issue discard by lba order */
328 unsigned int granularity; /* discard granularity */
329 int timeout; /* discard timeout for put_super */
330};
331
332struct discard_cmd_control {
333 struct task_struct *f2fs_issue_discard; /* discard thread */
334 struct list_head entry_list; /* 4KB discard entry list */
335 struct list_head pend_list[MAX_PLIST_NUM];/* store pending entries */
336 struct list_head wait_list; /* store on-flushing entries */
337 struct list_head fstrim_list; /* in-flight discard from fstrim */
338 wait_queue_head_t discard_wait_queue; /* waiting queue for wake-up */
339 unsigned int discard_wake; /* to wake up discard thread */
340 struct mutex cmd_lock;
341 unsigned int nr_discards; /* # of discards in the list */
342 unsigned int max_discards; /* max. discards to be issued */
343 unsigned int discard_granularity; /* discard granularity */
344 unsigned int undiscard_blks; /* # of undiscard blocks */
345 unsigned int next_pos; /* next discard position */
346 atomic_t issued_discard; /* # of issued discard */
347 atomic_t queued_discard; /* # of queued discard */
348 atomic_t discard_cmd_cnt; /* # of cached cmd count */
349 struct rb_root_cached root; /* root of discard rb-tree */
350 bool rbtree_check; /* config for consistence check */
351};
352
353/* for the list of fsync inodes, used only during recovery */
354struct fsync_inode_entry {
355 struct list_head list; /* list head */
356 struct inode *inode; /* vfs inode pointer */
357 block_t blkaddr; /* block address locating the last fsync */
358 block_t last_dentry; /* block address locating the last dentry */
359};
360
361#define nats_in_cursum(jnl) (le16_to_cpu((jnl)->n_nats))
362#define sits_in_cursum(jnl) (le16_to_cpu((jnl)->n_sits))
363
364#define nat_in_journal(jnl, i) ((jnl)->nat_j.entries[i].ne)
365#define nid_in_journal(jnl, i) ((jnl)->nat_j.entries[i].nid)
366#define sit_in_journal(jnl, i) ((jnl)->sit_j.entries[i].se)
367#define segno_in_journal(jnl, i) ((jnl)->sit_j.entries[i].segno)
368
369#define MAX_NAT_JENTRIES(jnl) (NAT_JOURNAL_ENTRIES - nats_in_cursum(jnl))
370#define MAX_SIT_JENTRIES(jnl) (SIT_JOURNAL_ENTRIES - sits_in_cursum(jnl))
371
372static inline int update_nats_in_cursum(struct f2fs_journal *journal, int i)
373{
374 int before = nats_in_cursum(journal);
375
376 journal->n_nats = cpu_to_le16(before + i);
377 return before;
378}
379
380static inline int update_sits_in_cursum(struct f2fs_journal *journal, int i)
381{
382 int before = sits_in_cursum(journal);
383
384 journal->n_sits = cpu_to_le16(before + i);
385 return before;
386}
387
388static inline bool __has_cursum_space(struct f2fs_journal *journal,
389 int size, int type)
390{
391 if (type == NAT_JOURNAL)
392 return size <= MAX_NAT_JENTRIES(journal);
393 return size <= MAX_SIT_JENTRIES(journal);
394}
395
396/*
397 * ioctl commands
398 */
399#define F2FS_IOC_GETFLAGS FS_IOC_GETFLAGS
400#define F2FS_IOC_SETFLAGS FS_IOC_SETFLAGS
401#define F2FS_IOC_GETVERSION FS_IOC_GETVERSION
402
403#define F2FS_IOCTL_MAGIC 0xf5
404#define F2FS_IOC_START_ATOMIC_WRITE _IO(F2FS_IOCTL_MAGIC, 1)
405#define F2FS_IOC_COMMIT_ATOMIC_WRITE _IO(F2FS_IOCTL_MAGIC, 2)
406#define F2FS_IOC_START_VOLATILE_WRITE _IO(F2FS_IOCTL_MAGIC, 3)
407#define F2FS_IOC_RELEASE_VOLATILE_WRITE _IO(F2FS_IOCTL_MAGIC, 4)
408#define F2FS_IOC_ABORT_VOLATILE_WRITE _IO(F2FS_IOCTL_MAGIC, 5)
409#define F2FS_IOC_GARBAGE_COLLECT _IOW(F2FS_IOCTL_MAGIC, 6, __u32)
410#define F2FS_IOC_WRITE_CHECKPOINT _IO(F2FS_IOCTL_MAGIC, 7)
411#define F2FS_IOC_DEFRAGMENT _IOWR(F2FS_IOCTL_MAGIC, 8, \
412 struct f2fs_defragment)
413#define F2FS_IOC_MOVE_RANGE _IOWR(F2FS_IOCTL_MAGIC, 9, \
414 struct f2fs_move_range)
415#define F2FS_IOC_FLUSH_DEVICE _IOW(F2FS_IOCTL_MAGIC, 10, \
416 struct f2fs_flush_device)
417#define F2FS_IOC_GARBAGE_COLLECT_RANGE _IOW(F2FS_IOCTL_MAGIC, 11, \
418 struct f2fs_gc_range)
419#define F2FS_IOC_GET_FEATURES _IOR(F2FS_IOCTL_MAGIC, 12, __u32)
420#define F2FS_IOC_SET_PIN_FILE _IOW(F2FS_IOCTL_MAGIC, 13, __u32)
421#define F2FS_IOC_GET_PIN_FILE _IOR(F2FS_IOCTL_MAGIC, 14, __u32)
422#define F2FS_IOC_PRECACHE_EXTENTS _IO(F2FS_IOCTL_MAGIC, 15)
423#define F2FS_IOC_RESIZE_FS _IOW(F2FS_IOCTL_MAGIC, 16, __u64)
424
425#define F2FS_IOC_GET_VOLUME_NAME FS_IOC_GETFSLABEL
426#define F2FS_IOC_SET_VOLUME_NAME FS_IOC_SETFSLABEL
427
428#define F2FS_IOC_SET_ENCRYPTION_POLICY FS_IOC_SET_ENCRYPTION_POLICY
429#define F2FS_IOC_GET_ENCRYPTION_POLICY FS_IOC_GET_ENCRYPTION_POLICY
430#define F2FS_IOC_GET_ENCRYPTION_PWSALT FS_IOC_GET_ENCRYPTION_PWSALT
431
432/*
433 * should be same as XFS_IOC_GOINGDOWN.
434 * Flags for going down operation used by FS_IOC_GOINGDOWN
435 */
436#define F2FS_IOC_SHUTDOWN _IOR('X', 125, __u32) /* Shutdown */
437#define F2FS_GOING_DOWN_FULLSYNC 0x0 /* going down with full sync */
438#define F2FS_GOING_DOWN_METASYNC 0x1 /* going down with metadata */
439#define F2FS_GOING_DOWN_NOSYNC 0x2 /* going down */
440#define F2FS_GOING_DOWN_METAFLUSH 0x3 /* going down with meta flush */
441#define F2FS_GOING_DOWN_NEED_FSCK 0x4 /* going down to trigger fsck */
442
443#if defined(__KERNEL__) && defined(CONFIG_COMPAT)
444/*
445 * ioctl commands in 32 bit emulation
446 */
447#define F2FS_IOC32_GETFLAGS FS_IOC32_GETFLAGS
448#define F2FS_IOC32_SETFLAGS FS_IOC32_SETFLAGS
449#define F2FS_IOC32_GETVERSION FS_IOC32_GETVERSION
450#endif
451
452#define F2FS_IOC_FSGETXATTR FS_IOC_FSGETXATTR
453#define F2FS_IOC_FSSETXATTR FS_IOC_FSSETXATTR
454
455struct f2fs_gc_range {
456 u32 sync;
457 u64 start;
458 u64 len;
459};
460
461struct f2fs_defragment {
462 u64 start;
463 u64 len;
464};
465
466struct f2fs_move_range {
467 u32 dst_fd; /* destination fd */
468 u64 pos_in; /* start position in src_fd */
469 u64 pos_out; /* start position in dst_fd */
470 u64 len; /* size to move */
471};
472
473struct f2fs_flush_device {
474 u32 dev_num; /* device number to flush */
475 u32 segments; /* # of segments to flush */
476};
477
478/* for inline stuff */
479#define DEF_INLINE_RESERVED_SIZE 1
480static inline int get_extra_isize(struct inode *inode);
481static inline int get_inline_xattr_addrs(struct inode *inode);
482#define MAX_INLINE_DATA(inode) (sizeof(__le32) * \
483 (CUR_ADDRS_PER_INODE(inode) - \
484 get_inline_xattr_addrs(inode) - \
485 DEF_INLINE_RESERVED_SIZE))
486
487/* for inline dir */
488#define NR_INLINE_DENTRY(inode) (MAX_INLINE_DATA(inode) * BITS_PER_BYTE / \
489 ((SIZE_OF_DIR_ENTRY + F2FS_SLOT_LEN) * \
490 BITS_PER_BYTE + 1))
491#define INLINE_DENTRY_BITMAP_SIZE(inode) \
492 DIV_ROUND_UP(NR_INLINE_DENTRY(inode), BITS_PER_BYTE)
493#define INLINE_RESERVED_SIZE(inode) (MAX_INLINE_DATA(inode) - \
494 ((SIZE_OF_DIR_ENTRY + F2FS_SLOT_LEN) * \
495 NR_INLINE_DENTRY(inode) + \
496 INLINE_DENTRY_BITMAP_SIZE(inode)))
497
498/*
499 * For INODE and NODE manager
500 */
501/* for directory operations */
502struct f2fs_dentry_ptr {
503 struct inode *inode;
504 void *bitmap;
505 struct f2fs_dir_entry *dentry;
506 __u8 (*filename)[F2FS_SLOT_LEN];
507 int max;
508 int nr_bitmap;
509};
510
511static inline void make_dentry_ptr_block(struct inode *inode,
512 struct f2fs_dentry_ptr *d, struct f2fs_dentry_block *t)
513{
514 d->inode = inode;
515 d->max = NR_DENTRY_IN_BLOCK;
516 d->nr_bitmap = SIZE_OF_DENTRY_BITMAP;
517 d->bitmap = t->dentry_bitmap;
518 d->dentry = t->dentry;
519 d->filename = t->filename;
520}
521
522static inline void make_dentry_ptr_inline(struct inode *inode,
523 struct f2fs_dentry_ptr *d, void *t)
524{
525 int entry_cnt = NR_INLINE_DENTRY(inode);
526 int bitmap_size = INLINE_DENTRY_BITMAP_SIZE(inode);
527 int reserved_size = INLINE_RESERVED_SIZE(inode);
528
529 d->inode = inode;
530 d->max = entry_cnt;
531 d->nr_bitmap = bitmap_size;
532 d->bitmap = t;
533 d->dentry = t + bitmap_size + reserved_size;
534 d->filename = t + bitmap_size + reserved_size +
535 SIZE_OF_DIR_ENTRY * entry_cnt;
536}
537
538/*
539 * XATTR_NODE_OFFSET stores xattrs to one node block per file keeping -1
540 * as its node offset to distinguish from index node blocks.
541 * But some bits are used to mark the node block.
542 */
543#define XATTR_NODE_OFFSET ((((unsigned int)-1) << OFFSET_BIT_SHIFT) \
544 >> OFFSET_BIT_SHIFT)
545enum {
546 ALLOC_NODE, /* allocate a new node page if needed */
547 LOOKUP_NODE, /* look up a node without readahead */
548 LOOKUP_NODE_RA, /*
549 * look up a node with readahead called
550 * by get_data_block.
551 */
552};
553
554#define DEFAULT_RETRY_IO_COUNT 8 /* maximum retry read IO count */
555
556/* maximum retry quota flush count */
557#define DEFAULT_RETRY_QUOTA_FLUSH_COUNT 8
558
559#define F2FS_LINK_MAX 0xffffffff /* maximum link count per file */
560
561#define MAX_DIR_RA_PAGES 4 /* maximum ra pages of dir */
562
563/* for in-memory extent cache entry */
564#define F2FS_MIN_EXTENT_LEN 64 /* minimum extent length */
565
566/* number of extent info in extent cache we try to shrink */
567#define EXTENT_CACHE_SHRINK_NUMBER 128
568
569struct rb_entry {
570 struct rb_node rb_node; /* rb node located in rb-tree */
571 unsigned int ofs; /* start offset of the entry */
572 unsigned int len; /* length of the entry */
573};
574
575struct extent_info {
576 unsigned int fofs; /* start offset in a file */
577 unsigned int len; /* length of the extent */
578 u32 blk; /* start block address of the extent */
579};
580
581struct extent_node {
582 struct rb_node rb_node; /* rb node located in rb-tree */
583 struct extent_info ei; /* extent info */
584 struct list_head list; /* node in global extent list of sbi */
585 struct extent_tree *et; /* extent tree pointer */
586};
587
588struct extent_tree {
589 nid_t ino; /* inode number */
590 struct rb_root_cached root; /* root of extent info rb-tree */
591 struct extent_node *cached_en; /* recently accessed extent node */
592 struct extent_info largest; /* largested extent info */
593 struct list_head list; /* to be used by sbi->zombie_list */
594 rwlock_t lock; /* protect extent info rb-tree */
595 atomic_t node_cnt; /* # of extent node in rb-tree*/
596 bool largest_updated; /* largest extent updated */
597};
598
599/*
600 * This structure is taken from ext4_map_blocks.
601 *
602 * Note that, however, f2fs uses NEW and MAPPED flags for f2fs_map_blocks().
603 */
604#define F2FS_MAP_NEW (1 << BH_New)
605#define F2FS_MAP_MAPPED (1 << BH_Mapped)
606#define F2FS_MAP_UNWRITTEN (1 << BH_Unwritten)
607#define F2FS_MAP_FLAGS (F2FS_MAP_NEW | F2FS_MAP_MAPPED |\
608 F2FS_MAP_UNWRITTEN)
609
610struct f2fs_map_blocks {
611 block_t m_pblk;
612 block_t m_lblk;
613 unsigned int m_len;
614 unsigned int m_flags;
615 pgoff_t *m_next_pgofs; /* point next possible non-hole pgofs */
616 pgoff_t *m_next_extent; /* point to next possible extent */
617 int m_seg_type;
618 bool m_may_create; /* indicate it is from write path */
619};
620
621/* for flag in get_data_block */
622enum {
623 F2FS_GET_BLOCK_DEFAULT,
624 F2FS_GET_BLOCK_FIEMAP,
625 F2FS_GET_BLOCK_BMAP,
626 F2FS_GET_BLOCK_DIO,
627 F2FS_GET_BLOCK_PRE_DIO,
628 F2FS_GET_BLOCK_PRE_AIO,
629 F2FS_GET_BLOCK_PRECACHE,
630};
631
632/*
633 * i_advise uses FADVISE_XXX_BIT. We can add additional hints later.
634 */
635#define FADVISE_COLD_BIT 0x01
636#define FADVISE_LOST_PINO_BIT 0x02
637#define FADVISE_ENCRYPT_BIT 0x04
638#define FADVISE_ENC_NAME_BIT 0x08
639#define FADVISE_KEEP_SIZE_BIT 0x10
640#define FADVISE_HOT_BIT 0x20
641#define FADVISE_VERITY_BIT 0x40
642
643#define FADVISE_MODIFIABLE_BITS (FADVISE_COLD_BIT | FADVISE_HOT_BIT)
644
645#define file_is_cold(inode) is_file(inode, FADVISE_COLD_BIT)
646#define file_wrong_pino(inode) is_file(inode, FADVISE_LOST_PINO_BIT)
647#define file_set_cold(inode) set_file(inode, FADVISE_COLD_BIT)
648#define file_lost_pino(inode) set_file(inode, FADVISE_LOST_PINO_BIT)
649#define file_clear_cold(inode) clear_file(inode, FADVISE_COLD_BIT)
650#define file_got_pino(inode) clear_file(inode, FADVISE_LOST_PINO_BIT)
651#define file_is_encrypt(inode) is_file(inode, FADVISE_ENCRYPT_BIT)
652#define file_set_encrypt(inode) set_file(inode, FADVISE_ENCRYPT_BIT)
653#define file_clear_encrypt(inode) clear_file(inode, FADVISE_ENCRYPT_BIT)
654#define file_enc_name(inode) is_file(inode, FADVISE_ENC_NAME_BIT)
655#define file_set_enc_name(inode) set_file(inode, FADVISE_ENC_NAME_BIT)
656#define file_keep_isize(inode) is_file(inode, FADVISE_KEEP_SIZE_BIT)
657#define file_set_keep_isize(inode) set_file(inode, FADVISE_KEEP_SIZE_BIT)
658#define file_is_hot(inode) is_file(inode, FADVISE_HOT_BIT)
659#define file_set_hot(inode) set_file(inode, FADVISE_HOT_BIT)
660#define file_clear_hot(inode) clear_file(inode, FADVISE_HOT_BIT)
661#define file_is_verity(inode) is_file(inode, FADVISE_VERITY_BIT)
662#define file_set_verity(inode) set_file(inode, FADVISE_VERITY_BIT)
663
664#define DEF_DIR_LEVEL 0
665
666enum {
667 GC_FAILURE_PIN,
668 GC_FAILURE_ATOMIC,
669 MAX_GC_FAILURE
670};
671
672struct f2fs_inode_info {
673 struct inode vfs_inode; /* serve a vfs inode */
674 unsigned long i_flags; /* keep an inode flags for ioctl */
675 unsigned char i_advise; /* use to give file attribute hints */
676 unsigned char i_dir_level; /* use for dentry level for large dir */
677 unsigned int i_current_depth; /* only for directory depth */
678 /* for gc failure statistic */
679 unsigned int i_gc_failures[MAX_GC_FAILURE];
680 unsigned int i_pino; /* parent inode number */
681 umode_t i_acl_mode; /* keep file acl mode temporarily */
682
683 /* Use below internally in f2fs*/
684 unsigned long flags; /* use to pass per-file flags */
685 struct rw_semaphore i_sem; /* protect fi info */
686 atomic_t dirty_pages; /* # of dirty pages */
687 f2fs_hash_t chash; /* hash value of given file name */
688 unsigned int clevel; /* maximum level of given file name */
689 struct task_struct *task; /* lookup and create consistency */
690 struct task_struct *cp_task; /* separate cp/wb IO stats*/
691 nid_t i_xattr_nid; /* node id that contains xattrs */
692 loff_t last_disk_size; /* lastly written file size */
693
694#ifdef CONFIG_QUOTA
695 struct dquot *i_dquot[MAXQUOTAS];
696
697 /* quota space reservation, managed internally by quota code */
698 qsize_t i_reserved_quota;
699#endif
700 struct list_head dirty_list; /* dirty list for dirs and files */
701 struct list_head gdirty_list; /* linked in global dirty list */
702 struct list_head inmem_ilist; /* list for inmem inodes */
703 struct list_head inmem_pages; /* inmemory pages managed by f2fs */
704 struct task_struct *inmem_task; /* store inmemory task */
705 struct mutex inmem_lock; /* lock for inmemory pages */
706 struct extent_tree *extent_tree; /* cached extent_tree entry */
707
708 /* avoid racing between foreground op and gc */
709 struct rw_semaphore i_gc_rwsem[2];
710 struct rw_semaphore i_mmap_sem;
711 struct rw_semaphore i_xattr_sem; /* avoid racing between reading and changing EAs */
712
713 int i_extra_isize; /* size of extra space located in i_addr */
714 kprojid_t i_projid; /* id for project quota */
715 int i_inline_xattr_size; /* inline xattr size */
716 struct timespec64 i_crtime; /* inode creation time */
717 struct timespec64 i_disk_time[4];/* inode disk times */
718};
719
720static inline void get_extent_info(struct extent_info *ext,
721 struct f2fs_extent *i_ext)
722{
723 ext->fofs = le32_to_cpu(i_ext->fofs);
724 ext->blk = le32_to_cpu(i_ext->blk);
725 ext->len = le32_to_cpu(i_ext->len);
726}
727
728static inline void set_raw_extent(struct extent_info *ext,
729 struct f2fs_extent *i_ext)
730{
731 i_ext->fofs = cpu_to_le32(ext->fofs);
732 i_ext->blk = cpu_to_le32(ext->blk);
733 i_ext->len = cpu_to_le32(ext->len);
734}
735
736static inline void set_extent_info(struct extent_info *ei, unsigned int fofs,
737 u32 blk, unsigned int len)
738{
739 ei->fofs = fofs;
740 ei->blk = blk;
741 ei->len = len;
742}
743
744static inline bool __is_discard_mergeable(struct discard_info *back,
745 struct discard_info *front, unsigned int max_len)
746{
747 return (back->lstart + back->len == front->lstart) &&
748 (back->len + front->len <= max_len);
749}
750
751static inline bool __is_discard_back_mergeable(struct discard_info *cur,
752 struct discard_info *back, unsigned int max_len)
753{
754 return __is_discard_mergeable(back, cur, max_len);
755}
756
757static inline bool __is_discard_front_mergeable(struct discard_info *cur,
758 struct discard_info *front, unsigned int max_len)
759{
760 return __is_discard_mergeable(cur, front, max_len);
761}
762
763static inline bool __is_extent_mergeable(struct extent_info *back,
764 struct extent_info *front)
765{
766 return (back->fofs + back->len == front->fofs &&
767 back->blk + back->len == front->blk);
768}
769
770static inline bool __is_back_mergeable(struct extent_info *cur,
771 struct extent_info *back)
772{
773 return __is_extent_mergeable(back, cur);
774}
775
776static inline bool __is_front_mergeable(struct extent_info *cur,
777 struct extent_info *front)
778{
779 return __is_extent_mergeable(cur, front);
780}
781
782extern void f2fs_mark_inode_dirty_sync(struct inode *inode, bool sync);
783static inline void __try_update_largest_extent(struct extent_tree *et,
784 struct extent_node *en)
785{
786 if (en->ei.len > et->largest.len) {
787 et->largest = en->ei;
788 et->largest_updated = true;
789 }
790}
791
792/*
793 * For free nid management
794 */
795enum nid_state {
796 FREE_NID, /* newly added to free nid list */
797 PREALLOC_NID, /* it is preallocated */
798 MAX_NID_STATE,
799};
800
801struct f2fs_nm_info {
802 block_t nat_blkaddr; /* base disk address of NAT */
803 nid_t max_nid; /* maximum possible node ids */
804 nid_t available_nids; /* # of available node ids */
805 nid_t next_scan_nid; /* the next nid to be scanned */
806 unsigned int ram_thresh; /* control the memory footprint */
807 unsigned int ra_nid_pages; /* # of nid pages to be readaheaded */
808 unsigned int dirty_nats_ratio; /* control dirty nats ratio threshold */
809
810 /* NAT cache management */
811 struct radix_tree_root nat_root;/* root of the nat entry cache */
812 struct radix_tree_root nat_set_root;/* root of the nat set cache */
813 struct rw_semaphore nat_tree_lock; /* protect nat_tree_lock */
814 struct list_head nat_entries; /* cached nat entry list (clean) */
815 spinlock_t nat_list_lock; /* protect clean nat entry list */
816 unsigned int nat_cnt; /* the # of cached nat entries */
817 unsigned int dirty_nat_cnt; /* total num of nat entries in set */
818 unsigned int nat_blocks; /* # of nat blocks */
819
820 /* free node ids management */
821 struct radix_tree_root free_nid_root;/* root of the free_nid cache */
822 struct list_head free_nid_list; /* list for free nids excluding preallocated nids */
823 unsigned int nid_cnt[MAX_NID_STATE]; /* the number of free node id */
824 spinlock_t nid_list_lock; /* protect nid lists ops */
825 struct mutex build_lock; /* lock for build free nids */
826 unsigned char **free_nid_bitmap;
827 unsigned char *nat_block_bitmap;
828 unsigned short *free_nid_count; /* free nid count of NAT block */
829
830 /* for checkpoint */
831 char *nat_bitmap; /* NAT bitmap pointer */
832
833 unsigned int nat_bits_blocks; /* # of nat bits blocks */
834 unsigned char *nat_bits; /* NAT bits blocks */
835 unsigned char *full_nat_bits; /* full NAT pages */
836 unsigned char *empty_nat_bits; /* empty NAT pages */
837#ifdef CONFIG_F2FS_CHECK_FS
838 char *nat_bitmap_mir; /* NAT bitmap mirror */
839#endif
840 int bitmap_size; /* bitmap size */
841};
842
843/*
844 * this structure is used as one of function parameters.
845 * all the information are dedicated to a given direct node block determined
846 * by the data offset in a file.
847 */
848struct dnode_of_data {
849 struct inode *inode; /* vfs inode pointer */
850 struct page *inode_page; /* its inode page, NULL is possible */
851 struct page *node_page; /* cached direct node page */
852 nid_t nid; /* node id of the direct node block */
853 unsigned int ofs_in_node; /* data offset in the node page */
854 bool inode_page_locked; /* inode page is locked or not */
855 bool node_changed; /* is node block changed */
856 char cur_level; /* level of hole node page */
857 char max_level; /* level of current page located */
858 block_t data_blkaddr; /* block address of the node block */
859};
860
861static inline void set_new_dnode(struct dnode_of_data *dn, struct inode *inode,
862 struct page *ipage, struct page *npage, nid_t nid)
863{
864 memset(dn, 0, sizeof(*dn));
865 dn->inode = inode;
866 dn->inode_page = ipage;
867 dn->node_page = npage;
868 dn->nid = nid;
869}
870
871/*
872 * For SIT manager
873 *
874 * By default, there are 6 active log areas across the whole main area.
875 * When considering hot and cold data separation to reduce cleaning overhead,
876 * we split 3 for data logs and 3 for node logs as hot, warm, and cold types,
877 * respectively.
878 * In the current design, you should not change the numbers intentionally.
879 * Instead, as a mount option such as active_logs=x, you can use 2, 4, and 6
880 * logs individually according to the underlying devices. (default: 6)
881 * Just in case, on-disk layout covers maximum 16 logs that consist of 8 for
882 * data and 8 for node logs.
883 */
884#define NR_CURSEG_DATA_TYPE (3)
885#define NR_CURSEG_NODE_TYPE (3)
886#define NR_CURSEG_TYPE (NR_CURSEG_DATA_TYPE + NR_CURSEG_NODE_TYPE)
887
888enum {
889 CURSEG_HOT_DATA = 0, /* directory entry blocks */
890 CURSEG_WARM_DATA, /* data blocks */
891 CURSEG_COLD_DATA, /* multimedia or GCed data blocks */
892 CURSEG_HOT_NODE, /* direct node blocks of directory files */
893 CURSEG_WARM_NODE, /* direct node blocks of normal files */
894 CURSEG_COLD_NODE, /* indirect node blocks */
895 NO_CHECK_TYPE,
896 CURSEG_COLD_DATA_PINNED,/* cold data for pinned file */
897};
898
899struct flush_cmd {
900 struct completion wait;
901 struct llist_node llnode;
902 nid_t ino;
903 int ret;
904};
905
906struct flush_cmd_control {
907 struct task_struct *f2fs_issue_flush; /* flush thread */
908 wait_queue_head_t flush_wait_queue; /* waiting queue for wake-up */
909 atomic_t issued_flush; /* # of issued flushes */
910 atomic_t queued_flush; /* # of queued flushes */
911 struct llist_head issue_list; /* list for command issue */
912 struct llist_node *dispatch_list; /* list for command dispatch */
913};
914
915struct f2fs_sm_info {
916 struct sit_info *sit_info; /* whole segment information */
917 struct free_segmap_info *free_info; /* free segment information */
918 struct dirty_seglist_info *dirty_info; /* dirty segment information */
919 struct curseg_info *curseg_array; /* active segment information */
920
921 struct rw_semaphore curseg_lock; /* for preventing curseg change */
922
923 block_t seg0_blkaddr; /* block address of 0'th segment */
924 block_t main_blkaddr; /* start block address of main area */
925 block_t ssa_blkaddr; /* start block address of SSA area */
926
927 unsigned int segment_count; /* total # of segments */
928 unsigned int main_segments; /* # of segments in main area */
929 unsigned int reserved_segments; /* # of reserved segments */
930 unsigned int ovp_segments; /* # of overprovision segments */
931
932 /* a threshold to reclaim prefree segments */
933 unsigned int rec_prefree_segments;
934
935 /* for batched trimming */
936 unsigned int trim_sections; /* # of sections to trim */
937
938 struct list_head sit_entry_set; /* sit entry set list */
939
940 unsigned int ipu_policy; /* in-place-update policy */
941 unsigned int min_ipu_util; /* in-place-update threshold */
942 unsigned int min_fsync_blocks; /* threshold for fsync */
943 unsigned int min_seq_blocks; /* threshold for sequential blocks */
944 unsigned int min_hot_blocks; /* threshold for hot block allocation */
945 unsigned int min_ssr_sections; /* threshold to trigger SSR allocation */
946
947 /* for flush command control */
948 struct flush_cmd_control *fcc_info;
949
950 /* for discard command control */
951 struct discard_cmd_control *dcc_info;
952};
953
954/*
955 * For superblock
956 */
957/*
958 * COUNT_TYPE for monitoring
959 *
960 * f2fs monitors the number of several block types such as on-writeback,
961 * dirty dentry blocks, dirty node blocks, and dirty meta blocks.
962 */
963#define WB_DATA_TYPE(p) (__is_cp_guaranteed(p) ? F2FS_WB_CP_DATA : F2FS_WB_DATA)
964enum count_type {
965 F2FS_DIRTY_DENTS,
966 F2FS_DIRTY_DATA,
967 F2FS_DIRTY_QDATA,
968 F2FS_DIRTY_NODES,
969 F2FS_DIRTY_META,
970 F2FS_INMEM_PAGES,
971 F2FS_DIRTY_IMETA,
972 F2FS_WB_CP_DATA,
973 F2FS_WB_DATA,
974 F2FS_RD_DATA,
975 F2FS_RD_NODE,
976 F2FS_RD_META,
977 F2FS_DIO_WRITE,
978 F2FS_DIO_READ,
979 NR_COUNT_TYPE,
980};
981
982/*
983 * The below are the page types of bios used in submit_bio().
984 * The available types are:
985 * DATA User data pages. It operates as async mode.
986 * NODE Node pages. It operates as async mode.
987 * META FS metadata pages such as SIT, NAT, CP.
988 * NR_PAGE_TYPE The number of page types.
989 * META_FLUSH Make sure the previous pages are written
990 * with waiting the bio's completion
991 * ... Only can be used with META.
992 */
993#define PAGE_TYPE_OF_BIO(type) ((type) > META ? META : (type))
994enum page_type {
995 DATA,
996 NODE,
997 META,
998 NR_PAGE_TYPE,
999 META_FLUSH,
1000 INMEM, /* the below types are used by tracepoints only. */
1001 INMEM_DROP,
1002 INMEM_INVALIDATE,
1003 INMEM_REVOKE,
1004 IPU,
1005 OPU,
1006};
1007
1008enum temp_type {
1009 HOT = 0, /* must be zero for meta bio */
1010 WARM,
1011 COLD,
1012 NR_TEMP_TYPE,
1013};
1014
1015enum need_lock_type {
1016 LOCK_REQ = 0,
1017 LOCK_DONE,
1018 LOCK_RETRY,
1019};
1020
1021enum cp_reason_type {
1022 CP_NO_NEEDED,
1023 CP_NON_REGULAR,
1024 CP_HARDLINK,
1025 CP_SB_NEED_CP,
1026 CP_WRONG_PINO,
1027 CP_NO_SPC_ROLL,
1028 CP_NODE_NEED_CP,
1029 CP_FASTBOOT_MODE,
1030 CP_SPEC_LOG_NUM,
1031 CP_RECOVER_DIR,
1032};
1033
1034enum iostat_type {
1035 APP_DIRECT_IO, /* app direct IOs */
1036 APP_BUFFERED_IO, /* app buffered IOs */
1037 APP_WRITE_IO, /* app write IOs */
1038 APP_MAPPED_IO, /* app mapped IOs */
1039 FS_DATA_IO, /* data IOs from kworker/fsync/reclaimer */
1040 FS_NODE_IO, /* node IOs from kworker/fsync/reclaimer */
1041 FS_META_IO, /* meta IOs from kworker/reclaimer */
1042 FS_GC_DATA_IO, /* data IOs from forground gc */
1043 FS_GC_NODE_IO, /* node IOs from forground gc */
1044 FS_CP_DATA_IO, /* data IOs from checkpoint */
1045 FS_CP_NODE_IO, /* node IOs from checkpoint */
1046 FS_CP_META_IO, /* meta IOs from checkpoint */
1047 FS_DISCARD, /* discard */
1048 NR_IO_TYPE,
1049};
1050
1051struct f2fs_io_info {
1052 struct f2fs_sb_info *sbi; /* f2fs_sb_info pointer */
1053 nid_t ino; /* inode number */
1054 enum page_type type; /* contains DATA/NODE/META/META_FLUSH */
1055 enum temp_type temp; /* contains HOT/WARM/COLD */
1056 int op; /* contains REQ_OP_ */
1057 int op_flags; /* req_flag_bits */
1058 block_t new_blkaddr; /* new block address to be written */
1059 block_t old_blkaddr; /* old block address before Cow */
1060 struct page *page; /* page to be written */
1061 struct page *encrypted_page; /* encrypted page */
1062 struct list_head list; /* serialize IOs */
1063 bool submitted; /* indicate IO submission */
1064 int need_lock; /* indicate we need to lock cp_rwsem */
1065 bool in_list; /* indicate fio is in io_list */
1066 bool is_por; /* indicate IO is from recovery or not */
1067 bool retry; /* need to reallocate block address */
1068 enum iostat_type io_type; /* io type */
1069 struct writeback_control *io_wbc; /* writeback control */
1070 struct bio **bio; /* bio for ipu */
1071 sector_t *last_block; /* last block number in bio */
1072 unsigned char version; /* version of the node */
1073};
1074
1075struct bio_entry {
1076 struct bio *bio;
1077 struct list_head list;
1078};
1079
1080#define is_read_io(rw) ((rw) == READ)
1081struct f2fs_bio_info {
1082 struct f2fs_sb_info *sbi; /* f2fs superblock */
1083 struct bio *bio; /* bios to merge */
1084 sector_t last_block_in_bio; /* last block number */
1085 struct f2fs_io_info fio; /* store buffered io info. */
1086 struct rw_semaphore io_rwsem; /* blocking op for bio */
1087 spinlock_t io_lock; /* serialize DATA/NODE IOs */
1088 struct list_head io_list; /* track fios */
1089 struct list_head bio_list; /* bio entry list head */
1090 struct rw_semaphore bio_list_lock; /* lock to protect bio entry list */
1091};
1092
1093#define FDEV(i) (sbi->devs[i])
1094#define RDEV(i) (raw_super->devs[i])
1095struct f2fs_dev_info {
1096 struct block_device *bdev;
1097 char path[MAX_PATH_LEN];
1098 unsigned int total_segments;
1099 block_t start_blk;
1100 block_t end_blk;
1101#ifdef CONFIG_BLK_DEV_ZONED
1102 unsigned int nr_blkz; /* Total number of zones */
1103 unsigned long *blkz_seq; /* Bitmap indicating sequential zones */
1104#endif
1105};
1106
1107enum inode_type {
1108 DIR_INODE, /* for dirty dir inode */
1109 FILE_INODE, /* for dirty regular/symlink inode */
1110 DIRTY_META, /* for all dirtied inode metadata */
1111 ATOMIC_FILE, /* for all atomic files */
1112 NR_INODE_TYPE,
1113};
1114
1115/* for inner inode cache management */
1116struct inode_management {
1117 struct radix_tree_root ino_root; /* ino entry array */
1118 spinlock_t ino_lock; /* for ino entry lock */
1119 struct list_head ino_list; /* inode list head */
1120 unsigned long ino_num; /* number of entries */
1121};
1122
1123/* For s_flag in struct f2fs_sb_info */
1124enum {
1125 SBI_IS_DIRTY, /* dirty flag for checkpoint */
1126 SBI_IS_CLOSE, /* specify unmounting */
1127 SBI_NEED_FSCK, /* need fsck.f2fs to fix */
1128 SBI_POR_DOING, /* recovery is doing or not */
1129 SBI_NEED_SB_WRITE, /* need to recover superblock */
1130 SBI_NEED_CP, /* need to checkpoint */
1131 SBI_IS_SHUTDOWN, /* shutdown by ioctl */
1132 SBI_IS_RECOVERED, /* recovered orphan/data */
1133 SBI_CP_DISABLED, /* CP was disabled last mount */
1134 SBI_CP_DISABLED_QUICK, /* CP was disabled quickly */
1135 SBI_QUOTA_NEED_FLUSH, /* need to flush quota info in CP */
1136 SBI_QUOTA_SKIP_FLUSH, /* skip flushing quota in current CP */
1137 SBI_QUOTA_NEED_REPAIR, /* quota file may be corrupted */
1138 SBI_IS_RESIZEFS, /* resizefs is in process */
1139};
1140
1141enum {
1142 CP_TIME,
1143 REQ_TIME,
1144 DISCARD_TIME,
1145 GC_TIME,
1146 DISABLE_TIME,
1147 UMOUNT_DISCARD_TIMEOUT,
1148 MAX_TIME,
1149};
1150
1151enum {
1152 GC_NORMAL,
1153 GC_IDLE_CB,
1154 GC_IDLE_GREEDY,
1155 GC_URGENT,
1156};
1157
1158enum {
1159 WHINT_MODE_OFF, /* not pass down write hints */
1160 WHINT_MODE_USER, /* try to pass down hints given by users */
1161 WHINT_MODE_FS, /* pass down hints with F2FS policy */
1162};
1163
1164enum {
1165 ALLOC_MODE_DEFAULT, /* stay default */
1166 ALLOC_MODE_REUSE, /* reuse segments as much as possible */
1167};
1168
1169enum fsync_mode {
1170 FSYNC_MODE_POSIX, /* fsync follows posix semantics */
1171 FSYNC_MODE_STRICT, /* fsync behaves in line with ext4 */
1172 FSYNC_MODE_NOBARRIER, /* fsync behaves nobarrier based on posix */
1173};
1174
1175#ifdef CONFIG_FS_ENCRYPTION
1176#define DUMMY_ENCRYPTION_ENABLED(sbi) \
1177 (unlikely(F2FS_OPTION(sbi).test_dummy_encryption))
1178#else
1179#define DUMMY_ENCRYPTION_ENABLED(sbi) (0)
1180#endif
1181
1182struct f2fs_sb_info {
1183 struct super_block *sb; /* pointer to VFS super block */
1184 struct proc_dir_entry *s_proc; /* proc entry */
1185 struct f2fs_super_block *raw_super; /* raw super block pointer */
1186 struct rw_semaphore sb_lock; /* lock for raw super block */
1187 int valid_super_block; /* valid super block no */
1188 unsigned long s_flag; /* flags for sbi */
1189 struct mutex writepages; /* mutex for writepages() */
1190#ifdef CONFIG_UNICODE
1191 struct unicode_map *s_encoding;
1192 __u16 s_encoding_flags;
1193#endif
1194
1195#ifdef CONFIG_BLK_DEV_ZONED
1196 unsigned int blocks_per_blkz; /* F2FS blocks per zone */
1197 unsigned int log_blocks_per_blkz; /* log2 F2FS blocks per zone */
1198#endif
1199
1200 /* for node-related operations */
1201 struct f2fs_nm_info *nm_info; /* node manager */
1202 struct inode *node_inode; /* cache node blocks */
1203
1204 /* for segment-related operations */
1205 struct f2fs_sm_info *sm_info; /* segment manager */
1206
1207 /* for bio operations */
1208 struct f2fs_bio_info *write_io[NR_PAGE_TYPE]; /* for write bios */
1209 /* keep migration IO order for LFS mode */
1210 struct rw_semaphore io_order_lock;
1211 mempool_t *write_io_dummy; /* Dummy pages */
1212
1213 /* for checkpoint */
1214 struct f2fs_checkpoint *ckpt; /* raw checkpoint pointer */
1215 int cur_cp_pack; /* remain current cp pack */
1216 spinlock_t cp_lock; /* for flag in ckpt */
1217 struct inode *meta_inode; /* cache meta blocks */
1218 struct mutex cp_mutex; /* checkpoint procedure lock */
1219 struct rw_semaphore cp_rwsem; /* blocking FS operations */
1220 struct rw_semaphore node_write; /* locking node writes */
1221 struct rw_semaphore node_change; /* locking node change */
1222 wait_queue_head_t cp_wait;
1223 unsigned long last_time[MAX_TIME]; /* to store time in jiffies */
1224 long interval_time[MAX_TIME]; /* to store thresholds */
1225
1226 struct inode_management im[MAX_INO_ENTRY]; /* manage inode cache */
1227
1228 spinlock_t fsync_node_lock; /* for node entry lock */
1229 struct list_head fsync_node_list; /* node list head */
1230 unsigned int fsync_seg_id; /* sequence id */
1231 unsigned int fsync_node_num; /* number of node entries */
1232
1233 /* for orphan inode, use 0'th array */
1234 unsigned int max_orphans; /* max orphan inodes */
1235
1236 /* for inode management */
1237 struct list_head inode_list[NR_INODE_TYPE]; /* dirty inode list */
1238 spinlock_t inode_lock[NR_INODE_TYPE]; /* for dirty inode list lock */
1239 struct mutex flush_lock; /* for flush exclusion */
1240
1241 /* for extent tree cache */
1242 struct radix_tree_root extent_tree_root;/* cache extent cache entries */
1243 struct mutex extent_tree_lock; /* locking extent radix tree */
1244 struct list_head extent_list; /* lru list for shrinker */
1245 spinlock_t extent_lock; /* locking extent lru list */
1246 atomic_t total_ext_tree; /* extent tree count */
1247 struct list_head zombie_list; /* extent zombie tree list */
1248 atomic_t total_zombie_tree; /* extent zombie tree count */
1249 atomic_t total_ext_node; /* extent info count */
1250
1251 /* basic filesystem units */
1252 unsigned int log_sectors_per_block; /* log2 sectors per block */
1253 unsigned int log_blocksize; /* log2 block size */
1254 unsigned int blocksize; /* block size */
1255 unsigned int root_ino_num; /* root inode number*/
1256 unsigned int node_ino_num; /* node inode number*/
1257 unsigned int meta_ino_num; /* meta inode number*/
1258 unsigned int log_blocks_per_seg; /* log2 blocks per segment */
1259 unsigned int blocks_per_seg; /* blocks per segment */
1260 unsigned int segs_per_sec; /* segments per section */
1261 unsigned int secs_per_zone; /* sections per zone */
1262 unsigned int total_sections; /* total section count */
1263 struct mutex resize_mutex; /* for resize exclusion */
1264 unsigned int total_node_count; /* total node block count */
1265 unsigned int total_valid_node_count; /* valid node block count */
1266 loff_t max_file_blocks; /* max block index of file */
1267 int dir_level; /* directory level */
1268 int readdir_ra; /* readahead inode in readdir */
1269
1270 block_t user_block_count; /* # of user blocks */
1271 block_t total_valid_block_count; /* # of valid blocks */
1272 block_t discard_blks; /* discard command candidats */
1273 block_t last_valid_block_count; /* for recovery */
1274 block_t reserved_blocks; /* configurable reserved blocks */
1275 block_t current_reserved_blocks; /* current reserved blocks */
1276
1277 /* Additional tracking for no checkpoint mode */
1278 block_t unusable_block_count; /* # of blocks saved by last cp */
1279
1280 unsigned int nquota_files; /* # of quota sysfile */
1281 struct rw_semaphore quota_sem; /* blocking cp for flags */
1282
1283 /* # of pages, see count_type */
1284 atomic_t nr_pages[NR_COUNT_TYPE];
1285 /* # of allocated blocks */
1286 struct percpu_counter alloc_valid_block_count;
1287
1288 /* writeback control */
1289 atomic_t wb_sync_req[META]; /* count # of WB_SYNC threads */
1290
1291 /* valid inode count */
1292 struct percpu_counter total_valid_inode_count;
1293
1294 struct f2fs_mount_info mount_opt; /* mount options */
1295
1296 /* for cleaning operations */
1297 struct mutex gc_mutex; /* mutex for GC */
1298 struct f2fs_gc_kthread *gc_thread; /* GC thread */
1299 unsigned int cur_victim_sec; /* current victim section num */
1300 unsigned int gc_mode; /* current GC state */
1301 unsigned int next_victim_seg[2]; /* next segment in victim section */
1302 /* for skip statistic */
1303 unsigned int atomic_files; /* # of opened atomic file */
1304 unsigned long long skipped_atomic_files[2]; /* FG_GC and BG_GC */
1305 unsigned long long skipped_gc_rwsem; /* FG_GC only */
1306
1307 /* threshold for gc trials on pinned files */
1308 u64 gc_pin_file_threshold;
1309 struct rw_semaphore pin_sem;
1310
1311 /* maximum # of trials to find a victim segment for SSR and GC */
1312 unsigned int max_victim_search;
1313 /* migration granularity of garbage collection, unit: segment */
1314 unsigned int migration_granularity;
1315
1316 /*
1317 * for stat information.
1318 * one is for the LFS mode, and the other is for the SSR mode.
1319 */
1320#ifdef CONFIG_F2FS_STAT_FS
1321 struct f2fs_stat_info *stat_info; /* FS status information */
1322 atomic_t meta_count[META_MAX]; /* # of meta blocks */
1323 unsigned int segment_count[2]; /* # of allocated segments */
1324 unsigned int block_count[2]; /* # of allocated blocks */
1325 atomic_t inplace_count; /* # of inplace update */
1326 atomic64_t total_hit_ext; /* # of lookup extent cache */
1327 atomic64_t read_hit_rbtree; /* # of hit rbtree extent node */
1328 atomic64_t read_hit_largest; /* # of hit largest extent node */
1329 atomic64_t read_hit_cached; /* # of hit cached extent node */
1330 atomic_t inline_xattr; /* # of inline_xattr inodes */
1331 atomic_t inline_inode; /* # of inline_data inodes */
1332 atomic_t inline_dir; /* # of inline_dentry inodes */
1333 atomic_t aw_cnt; /* # of atomic writes */
1334 atomic_t vw_cnt; /* # of volatile writes */
1335 atomic_t max_aw_cnt; /* max # of atomic writes */
1336 atomic_t max_vw_cnt; /* max # of volatile writes */
1337 int bg_gc; /* background gc calls */
1338 unsigned int io_skip_bggc; /* skip background gc for in-flight IO */
1339 unsigned int other_skip_bggc; /* skip background gc for other reasons */
1340 unsigned int ndirty_inode[NR_INODE_TYPE]; /* # of dirty inodes */
1341#endif
1342 spinlock_t stat_lock; /* lock for stat operations */
1343
1344 /* For app/fs IO statistics */
1345 spinlock_t iostat_lock;
1346 unsigned long long write_iostat[NR_IO_TYPE];
1347 bool iostat_enable;
1348
1349 /* For sysfs suppport */
1350 struct kobject s_kobj;
1351 struct completion s_kobj_unregister;
1352
1353 /* For shrinker support */
1354 struct list_head s_list;
1355 int s_ndevs; /* number of devices */
1356 struct f2fs_dev_info *devs; /* for device list */
1357 unsigned int dirty_device; /* for checkpoint data flush */
1358 spinlock_t dev_lock; /* protect dirty_device */
1359 struct mutex umount_mutex;
1360 unsigned int shrinker_run_no;
1361
1362 /* For write statistics */
1363 u64 sectors_written_start;
1364 u64 kbytes_written;
1365
1366 /* Reference to checksum algorithm driver via cryptoapi */
1367 struct crypto_shash *s_chksum_driver;
1368
1369 /* Precomputed FS UUID checksum for seeding other checksums */
1370 __u32 s_chksum_seed;
1371};
1372
1373struct f2fs_private_dio {
1374 struct inode *inode;
1375 void *orig_private;
1376 bio_end_io_t *orig_end_io;
1377 bool write;
1378};
1379
1380#ifdef CONFIG_F2FS_FAULT_INJECTION
1381#define f2fs_show_injection_info(sbi, type) \
1382 printk_ratelimited("%sF2FS-fs (%s) : inject %s in %s of %pS\n", \
1383 KERN_INFO, sbi->sb->s_id, \
1384 f2fs_fault_name[type], \
1385 __func__, __builtin_return_address(0))
1386static inline bool time_to_inject(struct f2fs_sb_info *sbi, int type)
1387{
1388 struct f2fs_fault_info *ffi = &F2FS_OPTION(sbi).fault_info;
1389
1390 if (!ffi->inject_rate)
1391 return false;
1392
1393 if (!IS_FAULT_SET(ffi, type))
1394 return false;
1395
1396 atomic_inc(&ffi->inject_ops);
1397 if (atomic_read(&ffi->inject_ops) >= ffi->inject_rate) {
1398 atomic_set(&ffi->inject_ops, 0);
1399 return true;
1400 }
1401 return false;
1402}
1403#else
1404#define f2fs_show_injection_info(sbi, type) do { } while (0)
1405static inline bool time_to_inject(struct f2fs_sb_info *sbi, int type)
1406{
1407 return false;
1408}
1409#endif
1410
1411/*
1412 * Test if the mounted volume is a multi-device volume.
1413 * - For a single regular disk volume, sbi->s_ndevs is 0.
1414 * - For a single zoned disk volume, sbi->s_ndevs is 1.
1415 * - For a multi-device volume, sbi->s_ndevs is always 2 or more.
1416 */
1417static inline bool f2fs_is_multi_device(struct f2fs_sb_info *sbi)
1418{
1419 return sbi->s_ndevs > 1;
1420}
1421
1422/* For write statistics. Suppose sector size is 512 bytes,
1423 * and the return value is in kbytes. s is of struct f2fs_sb_info.
1424 */
1425#define BD_PART_WRITTEN(s) \
1426(((u64)part_stat_read((s)->sb->s_bdev->bd_part, sectors[STAT_WRITE]) - \
1427 (s)->sectors_written_start) >> 1)
1428
1429static inline void f2fs_update_time(struct f2fs_sb_info *sbi, int type)
1430{
1431 unsigned long now = jiffies;
1432
1433 sbi->last_time[type] = now;
1434
1435 /* DISCARD_TIME and GC_TIME are based on REQ_TIME */
1436 if (type == REQ_TIME) {
1437 sbi->last_time[DISCARD_TIME] = now;
1438 sbi->last_time[GC_TIME] = now;
1439 }
1440}
1441
1442static inline bool f2fs_time_over(struct f2fs_sb_info *sbi, int type)
1443{
1444 unsigned long interval = sbi->interval_time[type] * HZ;
1445
1446 return time_after(jiffies, sbi->last_time[type] + interval);
1447}
1448
1449static inline unsigned int f2fs_time_to_wait(struct f2fs_sb_info *sbi,
1450 int type)
1451{
1452 unsigned long interval = sbi->interval_time[type] * HZ;
1453 unsigned int wait_ms = 0;
1454 long delta;
1455
1456 delta = (sbi->last_time[type] + interval) - jiffies;
1457 if (delta > 0)
1458 wait_ms = jiffies_to_msecs(delta);
1459
1460 return wait_ms;
1461}
1462
1463/*
1464 * Inline functions
1465 */
1466static inline u32 __f2fs_crc32(struct f2fs_sb_info *sbi, u32 crc,
1467 const void *address, unsigned int length)
1468{
1469 struct {
1470 struct shash_desc shash;
1471 char ctx[4];
1472 } desc;
1473 int err;
1474
1475 BUG_ON(crypto_shash_descsize(sbi->s_chksum_driver) != sizeof(desc.ctx));
1476
1477 desc.shash.tfm = sbi->s_chksum_driver;
1478 desc.shash.flags = 0;
1479 *(u32 *)desc.ctx = crc;
1480
1481 err = crypto_shash_update(&desc.shash, address, length);
1482 BUG_ON(err);
1483
1484 return *(u32 *)desc.ctx;
1485}
1486
1487static inline u32 f2fs_crc32(struct f2fs_sb_info *sbi, const void *address,
1488 unsigned int length)
1489{
1490 return __f2fs_crc32(sbi, F2FS_SUPER_MAGIC, address, length);
1491}
1492
1493static inline bool f2fs_crc_valid(struct f2fs_sb_info *sbi, __u32 blk_crc,
1494 void *buf, size_t buf_size)
1495{
1496 return f2fs_crc32(sbi, buf, buf_size) == blk_crc;
1497}
1498
1499static inline u32 f2fs_chksum(struct f2fs_sb_info *sbi, u32 crc,
1500 const void *address, unsigned int length)
1501{
1502 return __f2fs_crc32(sbi, crc, address, length);
1503}
1504
1505static inline struct f2fs_inode_info *F2FS_I(struct inode *inode)
1506{
1507 return container_of(inode, struct f2fs_inode_info, vfs_inode);
1508}
1509
1510static inline struct f2fs_sb_info *F2FS_SB(struct super_block *sb)
1511{
1512 return sb->s_fs_info;
1513}
1514
1515static inline struct f2fs_sb_info *F2FS_I_SB(struct inode *inode)
1516{
1517 return F2FS_SB(inode->i_sb);
1518}
1519
1520static inline struct f2fs_sb_info *F2FS_M_SB(struct address_space *mapping)
1521{
1522 return F2FS_I_SB(mapping->host);
1523}
1524
1525static inline struct f2fs_sb_info *F2FS_P_SB(struct page *page)
1526{
1527 return F2FS_M_SB(page_file_mapping(page));
1528}
1529
1530static inline struct f2fs_super_block *F2FS_RAW_SUPER(struct f2fs_sb_info *sbi)
1531{
1532 return (struct f2fs_super_block *)(sbi->raw_super);
1533}
1534
1535static inline struct f2fs_checkpoint *F2FS_CKPT(struct f2fs_sb_info *sbi)
1536{
1537 return (struct f2fs_checkpoint *)(sbi->ckpt);
1538}
1539
1540static inline struct f2fs_node *F2FS_NODE(struct page *page)
1541{
1542 return (struct f2fs_node *)page_address(page);
1543}
1544
1545static inline struct f2fs_inode *F2FS_INODE(struct page *page)
1546{
1547 return &((struct f2fs_node *)page_address(page))->i;
1548}
1549
1550static inline struct f2fs_nm_info *NM_I(struct f2fs_sb_info *sbi)
1551{
1552 return (struct f2fs_nm_info *)(sbi->nm_info);
1553}
1554
1555static inline struct f2fs_sm_info *SM_I(struct f2fs_sb_info *sbi)
1556{
1557 return (struct f2fs_sm_info *)(sbi->sm_info);
1558}
1559
1560static inline struct sit_info *SIT_I(struct f2fs_sb_info *sbi)
1561{
1562 return (struct sit_info *)(SM_I(sbi)->sit_info);
1563}
1564
1565static inline struct free_segmap_info *FREE_I(struct f2fs_sb_info *sbi)
1566{
1567 return (struct free_segmap_info *)(SM_I(sbi)->free_info);
1568}
1569
1570static inline struct dirty_seglist_info *DIRTY_I(struct f2fs_sb_info *sbi)
1571{
1572 return (struct dirty_seglist_info *)(SM_I(sbi)->dirty_info);
1573}
1574
1575static inline struct address_space *META_MAPPING(struct f2fs_sb_info *sbi)
1576{
1577 return sbi->meta_inode->i_mapping;
1578}
1579
1580static inline struct address_space *NODE_MAPPING(struct f2fs_sb_info *sbi)
1581{
1582 return sbi->node_inode->i_mapping;
1583}
1584
1585static inline bool is_sbi_flag_set(struct f2fs_sb_info *sbi, unsigned int type)
1586{
1587 return test_bit(type, &sbi->s_flag);
1588}
1589
1590static inline void set_sbi_flag(struct f2fs_sb_info *sbi, unsigned int type)
1591{
1592 set_bit(type, &sbi->s_flag);
1593}
1594
1595static inline void clear_sbi_flag(struct f2fs_sb_info *sbi, unsigned int type)
1596{
1597 clear_bit(type, &sbi->s_flag);
1598}
1599
1600static inline unsigned long long cur_cp_version(struct f2fs_checkpoint *cp)
1601{
1602 return le64_to_cpu(cp->checkpoint_ver);
1603}
1604
1605static inline unsigned long f2fs_qf_ino(struct super_block *sb, int type)
1606{
1607 if (type < F2FS_MAX_QUOTAS)
1608 return le32_to_cpu(F2FS_SB(sb)->raw_super->qf_ino[type]);
1609 return 0;
1610}
1611
1612static inline __u64 cur_cp_crc(struct f2fs_checkpoint *cp)
1613{
1614 size_t crc_offset = le32_to_cpu(cp->checksum_offset);
1615 return le32_to_cpu(*((__le32 *)((unsigned char *)cp + crc_offset)));
1616}
1617
1618static inline bool __is_set_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
1619{
1620 unsigned int ckpt_flags = le32_to_cpu(cp->ckpt_flags);
1621
1622 return ckpt_flags & f;
1623}
1624
1625static inline bool is_set_ckpt_flags(struct f2fs_sb_info *sbi, unsigned int f)
1626{
1627 return __is_set_ckpt_flags(F2FS_CKPT(sbi), f);
1628}
1629
1630static inline void __set_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
1631{
1632 unsigned int ckpt_flags;
1633
1634 ckpt_flags = le32_to_cpu(cp->ckpt_flags);
1635 ckpt_flags |= f;
1636 cp->ckpt_flags = cpu_to_le32(ckpt_flags);
1637}
1638
1639static inline void set_ckpt_flags(struct f2fs_sb_info *sbi, unsigned int f)
1640{
1641 unsigned long flags;
1642
1643 spin_lock_irqsave(&sbi->cp_lock, flags);
1644 __set_ckpt_flags(F2FS_CKPT(sbi), f);
1645 spin_unlock_irqrestore(&sbi->cp_lock, flags);
1646}
1647
1648static inline void __clear_ckpt_flags(struct f2fs_checkpoint *cp, unsigned int f)
1649{
1650 unsigned int ckpt_flags;
1651
1652 ckpt_flags = le32_to_cpu(cp->ckpt_flags);
1653 ckpt_flags &= (~f);
1654 cp->ckpt_flags = cpu_to_le32(ckpt_flags);
1655}
1656
1657static inline void clear_ckpt_flags(struct f2fs_sb_info *sbi, unsigned int f)
1658{
1659 unsigned long flags;
1660
1661 spin_lock_irqsave(&sbi->cp_lock, flags);
1662 __clear_ckpt_flags(F2FS_CKPT(sbi), f);
1663 spin_unlock_irqrestore(&sbi->cp_lock, flags);
1664}
1665
1666static inline void disable_nat_bits(struct f2fs_sb_info *sbi, bool lock)
1667{
1668 unsigned long flags;
1669 unsigned char *nat_bits;
1670
1671 /*
1672 * In order to re-enable nat_bits we need to call fsck.f2fs by
1673 * set_sbi_flag(sbi, SBI_NEED_FSCK). But it may give huge cost,
1674 * so let's rely on regular fsck or unclean shutdown.
1675 */
1676
1677 if (lock)
1678 spin_lock_irqsave(&sbi->cp_lock, flags);
1679 __clear_ckpt_flags(F2FS_CKPT(sbi), CP_NAT_BITS_FLAG);
1680 nat_bits = NM_I(sbi)->nat_bits;
1681 NM_I(sbi)->nat_bits = NULL;
1682 if (lock)
1683 spin_unlock_irqrestore(&sbi->cp_lock, flags);
1684
1685 kvfree(nat_bits);
1686}
1687
1688static inline bool enabled_nat_bits(struct f2fs_sb_info *sbi,
1689 struct cp_control *cpc)
1690{
1691 bool set = is_set_ckpt_flags(sbi, CP_NAT_BITS_FLAG);
1692
1693 return (cpc) ? (cpc->reason & CP_UMOUNT) && set : set;
1694}
1695
1696static inline void f2fs_lock_op(struct f2fs_sb_info *sbi)
1697{
1698 down_read(&sbi->cp_rwsem);
1699}
1700
1701static inline int f2fs_trylock_op(struct f2fs_sb_info *sbi)
1702{
1703 return down_read_trylock(&sbi->cp_rwsem);
1704}
1705
1706static inline void f2fs_unlock_op(struct f2fs_sb_info *sbi)
1707{
1708 up_read(&sbi->cp_rwsem);
1709}
1710
1711static inline void f2fs_lock_all(struct f2fs_sb_info *sbi)
1712{
1713 down_write(&sbi->cp_rwsem);
1714}
1715
1716static inline void f2fs_unlock_all(struct f2fs_sb_info *sbi)
1717{
1718 up_write(&sbi->cp_rwsem);
1719}
1720
1721static inline int __get_cp_reason(struct f2fs_sb_info *sbi)
1722{
1723 int reason = CP_SYNC;
1724
1725 if (test_opt(sbi, FASTBOOT))
1726 reason = CP_FASTBOOT;
1727 if (is_sbi_flag_set(sbi, SBI_IS_CLOSE))
1728 reason = CP_UMOUNT;
1729 return reason;
1730}
1731
1732static inline bool __remain_node_summaries(int reason)
1733{
1734 return (reason & (CP_UMOUNT | CP_FASTBOOT));
1735}
1736
1737static inline bool __exist_node_summaries(struct f2fs_sb_info *sbi)
1738{
1739 return (is_set_ckpt_flags(sbi, CP_UMOUNT_FLAG) ||
1740 is_set_ckpt_flags(sbi, CP_FASTBOOT_FLAG));
1741}
1742
1743/*
1744 * Check whether the inode has blocks or not
1745 */
1746static inline int F2FS_HAS_BLOCKS(struct inode *inode)
1747{
1748 block_t xattr_block = F2FS_I(inode)->i_xattr_nid ? 1 : 0;
1749
1750 return (inode->i_blocks >> F2FS_LOG_SECTORS_PER_BLOCK) > xattr_block;
1751}
1752
1753static inline bool f2fs_has_xattr_block(unsigned int ofs)
1754{
1755 return ofs == XATTR_NODE_OFFSET;
1756}
1757
1758static inline bool __allow_reserved_blocks(struct f2fs_sb_info *sbi,
1759 struct inode *inode, bool cap)
1760{
1761 if (!inode)
1762 return true;
1763 if (!test_opt(sbi, RESERVE_ROOT))
1764 return false;
1765 if (IS_NOQUOTA(inode))
1766 return true;
1767 if (uid_eq(F2FS_OPTION(sbi).s_resuid, current_fsuid()))
1768 return true;
1769 if (!gid_eq(F2FS_OPTION(sbi).s_resgid, GLOBAL_ROOT_GID) &&
1770 in_group_p(F2FS_OPTION(sbi).s_resgid))
1771 return true;
1772 if (cap && capable(CAP_SYS_RESOURCE))
1773 return true;
1774 return false;
1775}
1776
1777static inline void f2fs_i_blocks_write(struct inode *, block_t, bool, bool);
1778static inline int inc_valid_block_count(struct f2fs_sb_info *sbi,
1779 struct inode *inode, blkcnt_t *count)
1780{
1781 blkcnt_t diff = 0, release = 0;
1782 block_t avail_user_block_count;
1783 int ret;
1784
1785 ret = dquot_reserve_block(inode, *count);
1786 if (ret)
1787 return ret;
1788
1789 if (time_to_inject(sbi, FAULT_BLOCK)) {
1790 f2fs_show_injection_info(sbi, FAULT_BLOCK);
1791 release = *count;
1792 goto release_quota;
1793 }
1794
1795 /*
1796 * let's increase this in prior to actual block count change in order
1797 * for f2fs_sync_file to avoid data races when deciding checkpoint.
1798 */
1799 percpu_counter_add(&sbi->alloc_valid_block_count, (*count));
1800
1801 spin_lock(&sbi->stat_lock);
1802 sbi->total_valid_block_count += (block_t)(*count);
1803 avail_user_block_count = sbi->user_block_count -
1804 sbi->current_reserved_blocks;
1805
1806 if (!__allow_reserved_blocks(sbi, inode, true))
1807 avail_user_block_count -= F2FS_OPTION(sbi).root_reserved_blocks;
1808 if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED))) {
1809 if (avail_user_block_count > sbi->unusable_block_count)
1810 avail_user_block_count -= sbi->unusable_block_count;
1811 else
1812 avail_user_block_count = 0;
1813 }
1814 if (unlikely(sbi->total_valid_block_count > avail_user_block_count)) {
1815 diff = sbi->total_valid_block_count - avail_user_block_count;
1816 if (diff > *count)
1817 diff = *count;
1818 *count -= diff;
1819 release = diff;
1820 sbi->total_valid_block_count -= diff;
1821 if (!*count) {
1822 spin_unlock(&sbi->stat_lock);
1823 goto enospc;
1824 }
1825 }
1826 spin_unlock(&sbi->stat_lock);
1827
1828 if (unlikely(release)) {
1829 percpu_counter_sub(&sbi->alloc_valid_block_count, release);
1830 dquot_release_reservation_block(inode, release);
1831 }
1832 f2fs_i_blocks_write(inode, *count, true, true);
1833 return 0;
1834
1835enospc:
1836 percpu_counter_sub(&sbi->alloc_valid_block_count, release);
1837release_quota:
1838 dquot_release_reservation_block(inode, release);
1839 return -ENOSPC;
1840}
1841
1842__printf(2, 3)
1843void f2fs_printk(struct f2fs_sb_info *sbi, const char *fmt, ...);
1844
1845#define f2fs_err(sbi, fmt, ...) \
1846 f2fs_printk(sbi, KERN_ERR fmt, ##__VA_ARGS__)
1847#define f2fs_warn(sbi, fmt, ...) \
1848 f2fs_printk(sbi, KERN_WARNING fmt, ##__VA_ARGS__)
1849#define f2fs_notice(sbi, fmt, ...) \
1850 f2fs_printk(sbi, KERN_NOTICE fmt, ##__VA_ARGS__)
1851#define f2fs_info(sbi, fmt, ...) \
1852 f2fs_printk(sbi, KERN_INFO fmt, ##__VA_ARGS__)
1853#define f2fs_debug(sbi, fmt, ...) \
1854 f2fs_printk(sbi, KERN_DEBUG fmt, ##__VA_ARGS__)
1855
1856static inline void dec_valid_block_count(struct f2fs_sb_info *sbi,
1857 struct inode *inode,
1858 block_t count)
1859{
1860 blkcnt_t sectors = count << F2FS_LOG_SECTORS_PER_BLOCK;
1861
1862 spin_lock(&sbi->stat_lock);
1863 f2fs_bug_on(sbi, sbi->total_valid_block_count < (block_t) count);
1864 sbi->total_valid_block_count -= (block_t)count;
1865 if (sbi->reserved_blocks &&
1866 sbi->current_reserved_blocks < sbi->reserved_blocks)
1867 sbi->current_reserved_blocks = min(sbi->reserved_blocks,
1868 sbi->current_reserved_blocks + count);
1869 spin_unlock(&sbi->stat_lock);
1870 if (unlikely(inode->i_blocks < sectors)) {
1871 f2fs_warn(sbi, "Inconsistent i_blocks, ino:%lu, iblocks:%llu, sectors:%llu",
1872 inode->i_ino,
1873 (unsigned long long)inode->i_blocks,
1874 (unsigned long long)sectors);
1875 set_sbi_flag(sbi, SBI_NEED_FSCK);
1876 return;
1877 }
1878 f2fs_i_blocks_write(inode, count, false, true);
1879}
1880
1881static inline void inc_page_count(struct f2fs_sb_info *sbi, int count_type)
1882{
1883 atomic_inc(&sbi->nr_pages[count_type]);
1884
1885 if (count_type == F2FS_DIRTY_DENTS ||
1886 count_type == F2FS_DIRTY_NODES ||
1887 count_type == F2FS_DIRTY_META ||
1888 count_type == F2FS_DIRTY_QDATA ||
1889 count_type == F2FS_DIRTY_IMETA)
1890 set_sbi_flag(sbi, SBI_IS_DIRTY);
1891}
1892
1893static inline void inode_inc_dirty_pages(struct inode *inode)
1894{
1895 atomic_inc(&F2FS_I(inode)->dirty_pages);
1896 inc_page_count(F2FS_I_SB(inode), S_ISDIR(inode->i_mode) ?
1897 F2FS_DIRTY_DENTS : F2FS_DIRTY_DATA);
1898 if (IS_NOQUOTA(inode))
1899 inc_page_count(F2FS_I_SB(inode), F2FS_DIRTY_QDATA);
1900}
1901
1902static inline void dec_page_count(struct f2fs_sb_info *sbi, int count_type)
1903{
1904 atomic_dec(&sbi->nr_pages[count_type]);
1905}
1906
1907static inline void inode_dec_dirty_pages(struct inode *inode)
1908{
1909 if (!S_ISDIR(inode->i_mode) && !S_ISREG(inode->i_mode) &&
1910 !S_ISLNK(inode->i_mode))
1911 return;
1912
1913 atomic_dec(&F2FS_I(inode)->dirty_pages);
1914 dec_page_count(F2FS_I_SB(inode), S_ISDIR(inode->i_mode) ?
1915 F2FS_DIRTY_DENTS : F2FS_DIRTY_DATA);
1916 if (IS_NOQUOTA(inode))
1917 dec_page_count(F2FS_I_SB(inode), F2FS_DIRTY_QDATA);
1918}
1919
1920static inline s64 get_pages(struct f2fs_sb_info *sbi, int count_type)
1921{
1922 return atomic_read(&sbi->nr_pages[count_type]);
1923}
1924
1925static inline int get_dirty_pages(struct inode *inode)
1926{
1927 return atomic_read(&F2FS_I(inode)->dirty_pages);
1928}
1929
1930static inline int get_blocktype_secs(struct f2fs_sb_info *sbi, int block_type)
1931{
1932 unsigned int pages_per_sec = sbi->segs_per_sec * sbi->blocks_per_seg;
1933 unsigned int segs = (get_pages(sbi, block_type) + pages_per_sec - 1) >>
1934 sbi->log_blocks_per_seg;
1935
1936 return segs / sbi->segs_per_sec;
1937}
1938
1939static inline block_t valid_user_blocks(struct f2fs_sb_info *sbi)
1940{
1941 return sbi->total_valid_block_count;
1942}
1943
1944static inline block_t discard_blocks(struct f2fs_sb_info *sbi)
1945{
1946 return sbi->discard_blks;
1947}
1948
1949static inline unsigned long __bitmap_size(struct f2fs_sb_info *sbi, int flag)
1950{
1951 struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
1952
1953 /* return NAT or SIT bitmap */
1954 if (flag == NAT_BITMAP)
1955 return le32_to_cpu(ckpt->nat_ver_bitmap_bytesize);
1956 else if (flag == SIT_BITMAP)
1957 return le32_to_cpu(ckpt->sit_ver_bitmap_bytesize);
1958
1959 return 0;
1960}
1961
1962static inline block_t __cp_payload(struct f2fs_sb_info *sbi)
1963{
1964 return le32_to_cpu(F2FS_RAW_SUPER(sbi)->cp_payload);
1965}
1966
1967static inline void *__bitmap_ptr(struct f2fs_sb_info *sbi, int flag)
1968{
1969 struct f2fs_checkpoint *ckpt = F2FS_CKPT(sbi);
1970 int offset;
1971
1972 if (is_set_ckpt_flags(sbi, CP_LARGE_NAT_BITMAP_FLAG)) {
1973 offset = (flag == SIT_BITMAP) ?
1974 le32_to_cpu(ckpt->nat_ver_bitmap_bytesize) : 0;
1975 /*
1976 * if large_nat_bitmap feature is enabled, leave checksum
1977 * protection for all nat/sit bitmaps.
1978 */
1979 return &ckpt->sit_nat_version_bitmap + offset + sizeof(__le32);
1980 }
1981
1982 if (__cp_payload(sbi) > 0) {
1983 if (flag == NAT_BITMAP)
1984 return &ckpt->sit_nat_version_bitmap;
1985 else
1986 return (unsigned char *)ckpt + F2FS_BLKSIZE;
1987 } else {
1988 offset = (flag == NAT_BITMAP) ?
1989 le32_to_cpu(ckpt->sit_ver_bitmap_bytesize) : 0;
1990 return &ckpt->sit_nat_version_bitmap + offset;
1991 }
1992}
1993
1994static inline block_t __start_cp_addr(struct f2fs_sb_info *sbi)
1995{
1996 block_t start_addr = le32_to_cpu(F2FS_RAW_SUPER(sbi)->cp_blkaddr);
1997
1998 if (sbi->cur_cp_pack == 2)
1999 start_addr += sbi->blocks_per_seg;
2000 return start_addr;
2001}
2002
2003static inline block_t __start_cp_next_addr(struct f2fs_sb_info *sbi)
2004{
2005 block_t start_addr = le32_to_cpu(F2FS_RAW_SUPER(sbi)->cp_blkaddr);
2006
2007 if (sbi->cur_cp_pack == 1)
2008 start_addr += sbi->blocks_per_seg;
2009 return start_addr;
2010}
2011
2012static inline void __set_cp_next_pack(struct f2fs_sb_info *sbi)
2013{
2014 sbi->cur_cp_pack = (sbi->cur_cp_pack == 1) ? 2 : 1;
2015}
2016
2017static inline block_t __start_sum_addr(struct f2fs_sb_info *sbi)
2018{
2019 return le32_to_cpu(F2FS_CKPT(sbi)->cp_pack_start_sum);
2020}
2021
2022static inline int inc_valid_node_count(struct f2fs_sb_info *sbi,
2023 struct inode *inode, bool is_inode)
2024{
2025 block_t valid_block_count;
2026 unsigned int valid_node_count, user_block_count;
2027 int err;
2028
2029 if (is_inode) {
2030 if (inode) {
2031 err = dquot_alloc_inode(inode);
2032 if (err)
2033 return err;
2034 }
2035 } else {
2036 err = dquot_reserve_block(inode, 1);
2037 if (err)
2038 return err;
2039 }
2040
2041 if (time_to_inject(sbi, FAULT_BLOCK)) {
2042 f2fs_show_injection_info(sbi, FAULT_BLOCK);
2043 goto enospc;
2044 }
2045
2046 spin_lock(&sbi->stat_lock);
2047
2048 valid_block_count = sbi->total_valid_block_count +
2049 sbi->current_reserved_blocks + 1;
2050
2051 if (!__allow_reserved_blocks(sbi, inode, false))
2052 valid_block_count += F2FS_OPTION(sbi).root_reserved_blocks;
2053 user_block_count = sbi->user_block_count;
2054 if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED)))
2055 user_block_count -= sbi->unusable_block_count;
2056
2057 if (unlikely(valid_block_count > user_block_count)) {
2058 spin_unlock(&sbi->stat_lock);
2059 goto enospc;
2060 }
2061
2062 valid_node_count = sbi->total_valid_node_count + 1;
2063 if (unlikely(valid_node_count > sbi->total_node_count)) {
2064 spin_unlock(&sbi->stat_lock);
2065 goto enospc;
2066 }
2067
2068 sbi->total_valid_node_count++;
2069 sbi->total_valid_block_count++;
2070 spin_unlock(&sbi->stat_lock);
2071
2072 if (inode) {
2073 if (is_inode)
2074 f2fs_mark_inode_dirty_sync(inode, true);
2075 else
2076 f2fs_i_blocks_write(inode, 1, true, true);
2077 }
2078
2079 percpu_counter_inc(&sbi->alloc_valid_block_count);
2080 return 0;
2081
2082enospc:
2083 if (is_inode) {
2084 if (inode)
2085 dquot_free_inode(inode);
2086 } else {
2087 dquot_release_reservation_block(inode, 1);
2088 }
2089 return -ENOSPC;
2090}
2091
2092static inline void dec_valid_node_count(struct f2fs_sb_info *sbi,
2093 struct inode *inode, bool is_inode)
2094{
2095 spin_lock(&sbi->stat_lock);
2096
2097 f2fs_bug_on(sbi, !sbi->total_valid_block_count);
2098 f2fs_bug_on(sbi, !sbi->total_valid_node_count);
2099
2100 sbi->total_valid_node_count--;
2101 sbi->total_valid_block_count--;
2102 if (sbi->reserved_blocks &&
2103 sbi->current_reserved_blocks < sbi->reserved_blocks)
2104 sbi->current_reserved_blocks++;
2105
2106 spin_unlock(&sbi->stat_lock);
2107
2108 if (is_inode) {
2109 dquot_free_inode(inode);
2110 } else {
2111 if (unlikely(inode->i_blocks == 0)) {
2112 f2fs_warn(sbi, "Inconsistent i_blocks, ino:%lu, iblocks:%llu",
2113 inode->i_ino,
2114 (unsigned long long)inode->i_blocks);
2115 set_sbi_flag(sbi, SBI_NEED_FSCK);
2116 return;
2117 }
2118 f2fs_i_blocks_write(inode, 1, false, true);
2119 }
2120}
2121
2122static inline unsigned int valid_node_count(struct f2fs_sb_info *sbi)
2123{
2124 return sbi->total_valid_node_count;
2125}
2126
2127static inline void inc_valid_inode_count(struct f2fs_sb_info *sbi)
2128{
2129 percpu_counter_inc(&sbi->total_valid_inode_count);
2130}
2131
2132static inline void dec_valid_inode_count(struct f2fs_sb_info *sbi)
2133{
2134 percpu_counter_dec(&sbi->total_valid_inode_count);
2135}
2136
2137static inline s64 valid_inode_count(struct f2fs_sb_info *sbi)
2138{
2139 return percpu_counter_sum_positive(&sbi->total_valid_inode_count);
2140}
2141
2142static inline struct page *f2fs_grab_cache_page(struct address_space *mapping,
2143 pgoff_t index, bool for_write)
2144{
2145 struct page *page;
2146
2147 if (IS_ENABLED(CONFIG_F2FS_FAULT_INJECTION)) {
2148 if (!for_write)
2149 page = find_get_page_flags(mapping, index,
2150 FGP_LOCK | FGP_ACCESSED);
2151 else
2152 page = find_lock_page(mapping, index);
2153 if (page)
2154 return page;
2155
2156 if (time_to_inject(F2FS_M_SB(mapping), FAULT_PAGE_ALLOC)) {
2157 f2fs_show_injection_info(F2FS_M_SB(mapping),
2158 FAULT_PAGE_ALLOC);
2159 return NULL;
2160 }
2161 }
2162
2163 if (!for_write)
2164 return grab_cache_page(mapping, index);
2165 return grab_cache_page_write_begin(mapping, index, AOP_FLAG_NOFS);
2166}
2167
2168static inline struct page *f2fs_pagecache_get_page(
2169 struct address_space *mapping, pgoff_t index,
2170 int fgp_flags, gfp_t gfp_mask)
2171{
2172 if (time_to_inject(F2FS_M_SB(mapping), FAULT_PAGE_GET)) {
2173 f2fs_show_injection_info(F2FS_M_SB(mapping), FAULT_PAGE_GET);
2174 return NULL;
2175 }
2176
2177 return pagecache_get_page(mapping, index, fgp_flags, gfp_mask);
2178}
2179
2180static inline void f2fs_copy_page(struct page *src, struct page *dst)
2181{
2182 char *src_kaddr = kmap(src);
2183 char *dst_kaddr = kmap(dst);
2184
2185 memcpy(dst_kaddr, src_kaddr, PAGE_SIZE);
2186 kunmap(dst);
2187 kunmap(src);
2188}
2189
2190static inline void f2fs_put_page(struct page *page, int unlock)
2191{
2192 if (!page)
2193 return;
2194
2195 if (unlock) {
2196 f2fs_bug_on(F2FS_P_SB(page), !PageLocked(page));
2197 unlock_page(page);
2198 }
2199 put_page(page);
2200}
2201
2202static inline void f2fs_put_dnode(struct dnode_of_data *dn)
2203{
2204 if (dn->node_page)
2205 f2fs_put_page(dn->node_page, 1);
2206 if (dn->inode_page && dn->node_page != dn->inode_page)
2207 f2fs_put_page(dn->inode_page, 0);
2208 dn->node_page = NULL;
2209 dn->inode_page = NULL;
2210}
2211
2212static inline struct kmem_cache *f2fs_kmem_cache_create(const char *name,
2213 size_t size)
2214{
2215 return kmem_cache_create(name, size, 0, SLAB_RECLAIM_ACCOUNT, NULL);
2216}
2217
2218static inline void *f2fs_kmem_cache_alloc(struct kmem_cache *cachep,
2219 gfp_t flags)
2220{
2221 void *entry;
2222
2223 entry = kmem_cache_alloc(cachep, flags);
2224 if (!entry)
2225 entry = kmem_cache_alloc(cachep, flags | __GFP_NOFAIL);
2226 return entry;
2227}
2228
2229static inline struct bio *f2fs_bio_alloc(struct f2fs_sb_info *sbi,
2230 int npages, bool no_fail)
2231{
2232 struct bio *bio;
2233
2234 if (no_fail) {
2235 /* No failure on bio allocation */
2236 bio = bio_alloc(GFP_NOIO, npages);
2237 if (!bio)
2238 bio = bio_alloc(GFP_NOIO | __GFP_NOFAIL, npages);
2239 return bio;
2240 }
2241 if (time_to_inject(sbi, FAULT_ALLOC_BIO)) {
2242 f2fs_show_injection_info(sbi, FAULT_ALLOC_BIO);
2243 return NULL;
2244 }
2245
2246 return bio_alloc(GFP_KERNEL, npages);
2247}
2248
2249static inline bool is_idle(struct f2fs_sb_info *sbi, int type)
2250{
2251 if (sbi->gc_mode == GC_URGENT)
2252 return true;
2253
2254 if (get_pages(sbi, F2FS_RD_DATA) || get_pages(sbi, F2FS_RD_NODE) ||
2255 get_pages(sbi, F2FS_RD_META) || get_pages(sbi, F2FS_WB_DATA) ||
2256 get_pages(sbi, F2FS_WB_CP_DATA) ||
2257 get_pages(sbi, F2FS_DIO_READ) ||
2258 get_pages(sbi, F2FS_DIO_WRITE))
2259 return false;
2260
2261 if (type != DISCARD_TIME && SM_I(sbi) && SM_I(sbi)->dcc_info &&
2262 atomic_read(&SM_I(sbi)->dcc_info->queued_discard))
2263 return false;
2264
2265 if (SM_I(sbi) && SM_I(sbi)->fcc_info &&
2266 atomic_read(&SM_I(sbi)->fcc_info->queued_flush))
2267 return false;
2268
2269 return f2fs_time_over(sbi, type);
2270}
2271
2272static inline void f2fs_radix_tree_insert(struct radix_tree_root *root,
2273 unsigned long index, void *item)
2274{
2275 while (radix_tree_insert(root, index, item))
2276 cond_resched();
2277}
2278
2279#define RAW_IS_INODE(p) ((p)->footer.nid == (p)->footer.ino)
2280
2281static inline bool IS_INODE(struct page *page)
2282{
2283 struct f2fs_node *p = F2FS_NODE(page);
2284
2285 return RAW_IS_INODE(p);
2286}
2287
2288static inline int offset_in_addr(struct f2fs_inode *i)
2289{
2290 return (i->i_inline & F2FS_EXTRA_ATTR) ?
2291 (le16_to_cpu(i->i_extra_isize) / sizeof(__le32)) : 0;
2292}
2293
2294static inline __le32 *blkaddr_in_node(struct f2fs_node *node)
2295{
2296 return RAW_IS_INODE(node) ? node->i.i_addr : node->dn.addr;
2297}
2298
2299static inline int f2fs_has_extra_attr(struct inode *inode);
2300static inline block_t datablock_addr(struct inode *inode,
2301 struct page *node_page, unsigned int offset)
2302{
2303 struct f2fs_node *raw_node;
2304 __le32 *addr_array;
2305 int base = 0;
2306 bool is_inode = IS_INODE(node_page);
2307
2308 raw_node = F2FS_NODE(node_page);
2309
2310 /* from GC path only */
2311 if (is_inode) {
2312 if (!inode)
2313 base = offset_in_addr(&raw_node->i);
2314 else if (f2fs_has_extra_attr(inode))
2315 base = get_extra_isize(inode);
2316 }
2317
2318 addr_array = blkaddr_in_node(raw_node);
2319 return le32_to_cpu(addr_array[base + offset]);
2320}
2321
2322static inline int f2fs_test_bit(unsigned int nr, char *addr)
2323{
2324 int mask;
2325
2326 addr += (nr >> 3);
2327 mask = 1 << (7 - (nr & 0x07));
2328 return mask & *addr;
2329}
2330
2331static inline void f2fs_set_bit(unsigned int nr, char *addr)
2332{
2333 int mask;
2334
2335 addr += (nr >> 3);
2336 mask = 1 << (7 - (nr & 0x07));
2337 *addr |= mask;
2338}
2339
2340static inline void f2fs_clear_bit(unsigned int nr, char *addr)
2341{
2342 int mask;
2343
2344 addr += (nr >> 3);
2345 mask = 1 << (7 - (nr & 0x07));
2346 *addr &= ~mask;
2347}
2348
2349static inline int f2fs_test_and_set_bit(unsigned int nr, char *addr)
2350{
2351 int mask;
2352 int ret;
2353
2354 addr += (nr >> 3);
2355 mask = 1 << (7 - (nr & 0x07));
2356 ret = mask & *addr;
2357 *addr |= mask;
2358 return ret;
2359}
2360
2361static inline int f2fs_test_and_clear_bit(unsigned int nr, char *addr)
2362{
2363 int mask;
2364 int ret;
2365
2366 addr += (nr >> 3);
2367 mask = 1 << (7 - (nr & 0x07));
2368 ret = mask & *addr;
2369 *addr &= ~mask;
2370 return ret;
2371}
2372
2373static inline void f2fs_change_bit(unsigned int nr, char *addr)
2374{
2375 int mask;
2376
2377 addr += (nr >> 3);
2378 mask = 1 << (7 - (nr & 0x07));
2379 *addr ^= mask;
2380}
2381
2382/*
2383 * On-disk inode flags (f2fs_inode::i_flags)
2384 */
2385#define F2FS_SYNC_FL 0x00000008 /* Synchronous updates */
2386#define F2FS_IMMUTABLE_FL 0x00000010 /* Immutable file */
2387#define F2FS_APPEND_FL 0x00000020 /* writes to file may only append */
2388#define F2FS_NODUMP_FL 0x00000040 /* do not dump file */
2389#define F2FS_NOATIME_FL 0x00000080 /* do not update atime */
2390#define F2FS_INDEX_FL 0x00001000 /* hash-indexed directory */
2391#define F2FS_DIRSYNC_FL 0x00010000 /* dirsync behaviour (directories only) */
2392#define F2FS_PROJINHERIT_FL 0x20000000 /* Create with parents projid */
2393#define F2FS_CASEFOLD_FL 0x40000000 /* Casefolded file */
2394
2395/* Flags that should be inherited by new inodes from their parent. */
2396#define F2FS_FL_INHERITED (F2FS_SYNC_FL | F2FS_NODUMP_FL | F2FS_NOATIME_FL | \
2397 F2FS_DIRSYNC_FL | F2FS_PROJINHERIT_FL | \
2398 F2FS_CASEFOLD_FL)
2399
2400/* Flags that are appropriate for regular files (all but dir-specific ones). */
2401#define F2FS_REG_FLMASK (~(F2FS_DIRSYNC_FL | F2FS_PROJINHERIT_FL | \
2402 F2FS_CASEFOLD_FL))
2403
2404/* Flags that are appropriate for non-directories/regular files. */
2405#define F2FS_OTHER_FLMASK (F2FS_NODUMP_FL | F2FS_NOATIME_FL)
2406
2407static inline __u32 f2fs_mask_flags(umode_t mode, __u32 flags)
2408{
2409 if (S_ISDIR(mode))
2410 return flags;
2411 else if (S_ISREG(mode))
2412 return flags & F2FS_REG_FLMASK;
2413 else
2414 return flags & F2FS_OTHER_FLMASK;
2415}
2416
2417/* used for f2fs_inode_info->flags */
2418enum {
2419 FI_NEW_INODE, /* indicate newly allocated inode */
2420 FI_DIRTY_INODE, /* indicate inode is dirty or not */
2421 FI_AUTO_RECOVER, /* indicate inode is recoverable */
2422 FI_DIRTY_DIR, /* indicate directory has dirty pages */
2423 FI_INC_LINK, /* need to increment i_nlink */
2424 FI_ACL_MODE, /* indicate acl mode */
2425 FI_NO_ALLOC, /* should not allocate any blocks */
2426 FI_FREE_NID, /* free allocated nide */
2427 FI_NO_EXTENT, /* not to use the extent cache */
2428 FI_INLINE_XATTR, /* used for inline xattr */
2429 FI_INLINE_DATA, /* used for inline data*/
2430 FI_INLINE_DENTRY, /* used for inline dentry */
2431 FI_APPEND_WRITE, /* inode has appended data */
2432 FI_UPDATE_WRITE, /* inode has in-place-update data */
2433 FI_NEED_IPU, /* used for ipu per file */
2434 FI_ATOMIC_FILE, /* indicate atomic file */
2435 FI_ATOMIC_COMMIT, /* indicate the state of atomical committing */
2436 FI_VOLATILE_FILE, /* indicate volatile file */
2437 FI_FIRST_BLOCK_WRITTEN, /* indicate #0 data block was written */
2438 FI_DROP_CACHE, /* drop dirty page cache */
2439 FI_DATA_EXIST, /* indicate data exists */
2440 FI_INLINE_DOTS, /* indicate inline dot dentries */
2441 FI_DO_DEFRAG, /* indicate defragment is running */
2442 FI_DIRTY_FILE, /* indicate regular/symlink has dirty pages */
2443 FI_NO_PREALLOC, /* indicate skipped preallocated blocks */
2444 FI_HOT_DATA, /* indicate file is hot */
2445 FI_EXTRA_ATTR, /* indicate file has extra attribute */
2446 FI_PROJ_INHERIT, /* indicate file inherits projectid */
2447 FI_PIN_FILE, /* indicate file should not be gced */
2448 FI_ATOMIC_REVOKE_REQUEST, /* request to drop atomic data */
2449 FI_VERITY_IN_PROGRESS, /* building fs-verity Merkle tree */
2450};
2451
2452static inline void __mark_inode_dirty_flag(struct inode *inode,
2453 int flag, bool set)
2454{
2455 switch (flag) {
2456 case FI_INLINE_XATTR:
2457 case FI_INLINE_DATA:
2458 case FI_INLINE_DENTRY:
2459 case FI_NEW_INODE:
2460 if (set)
2461 return;
2462 /* fall through */
2463 case FI_DATA_EXIST:
2464 case FI_INLINE_DOTS:
2465 case FI_PIN_FILE:
2466 f2fs_mark_inode_dirty_sync(inode, true);
2467 }
2468}
2469
2470static inline void set_inode_flag(struct inode *inode, int flag)
2471{
2472 if (!test_bit(flag, &F2FS_I(inode)->flags))
2473 set_bit(flag, &F2FS_I(inode)->flags);
2474 __mark_inode_dirty_flag(inode, flag, true);
2475}
2476
2477static inline int is_inode_flag_set(struct inode *inode, int flag)
2478{
2479 return test_bit(flag, &F2FS_I(inode)->flags);
2480}
2481
2482static inline void clear_inode_flag(struct inode *inode, int flag)
2483{
2484 if (test_bit(flag, &F2FS_I(inode)->flags))
2485 clear_bit(flag, &F2FS_I(inode)->flags);
2486 __mark_inode_dirty_flag(inode, flag, false);
2487}
2488
2489static inline bool f2fs_verity_in_progress(struct inode *inode)
2490{
2491 return IS_ENABLED(CONFIG_FS_VERITY) &&
2492 is_inode_flag_set(inode, FI_VERITY_IN_PROGRESS);
2493}
2494
2495static inline void set_acl_inode(struct inode *inode, umode_t mode)
2496{
2497 F2FS_I(inode)->i_acl_mode = mode;
2498 set_inode_flag(inode, FI_ACL_MODE);
2499 f2fs_mark_inode_dirty_sync(inode, false);
2500}
2501
2502static inline void f2fs_i_links_write(struct inode *inode, bool inc)
2503{
2504 if (inc)
2505 inc_nlink(inode);
2506 else
2507 drop_nlink(inode);
2508 f2fs_mark_inode_dirty_sync(inode, true);
2509}
2510
2511static inline void f2fs_i_blocks_write(struct inode *inode,
2512 block_t diff, bool add, bool claim)
2513{
2514 bool clean = !is_inode_flag_set(inode, FI_DIRTY_INODE);
2515 bool recover = is_inode_flag_set(inode, FI_AUTO_RECOVER);
2516
2517 /* add = 1, claim = 1 should be dquot_reserve_block in pair */
2518 if (add) {
2519 if (claim)
2520 dquot_claim_block(inode, diff);
2521 else
2522 dquot_alloc_block_nofail(inode, diff);
2523 } else {
2524 dquot_free_block(inode, diff);
2525 }
2526
2527 f2fs_mark_inode_dirty_sync(inode, true);
2528 if (clean || recover)
2529 set_inode_flag(inode, FI_AUTO_RECOVER);
2530}
2531
2532static inline void f2fs_i_size_write(struct inode *inode, loff_t i_size)
2533{
2534 bool clean = !is_inode_flag_set(inode, FI_DIRTY_INODE);
2535 bool recover = is_inode_flag_set(inode, FI_AUTO_RECOVER);
2536
2537 if (i_size_read(inode) == i_size)
2538 return;
2539
2540 i_size_write(inode, i_size);
2541 f2fs_mark_inode_dirty_sync(inode, true);
2542 if (clean || recover)
2543 set_inode_flag(inode, FI_AUTO_RECOVER);
2544}
2545
2546static inline void f2fs_i_depth_write(struct inode *inode, unsigned int depth)
2547{
2548 F2FS_I(inode)->i_current_depth = depth;
2549 f2fs_mark_inode_dirty_sync(inode, true);
2550}
2551
2552static inline void f2fs_i_gc_failures_write(struct inode *inode,
2553 unsigned int count)
2554{
2555 F2FS_I(inode)->i_gc_failures[GC_FAILURE_PIN] = count;
2556 f2fs_mark_inode_dirty_sync(inode, true);
2557}
2558
2559static inline void f2fs_i_xnid_write(struct inode *inode, nid_t xnid)
2560{
2561 F2FS_I(inode)->i_xattr_nid = xnid;
2562 f2fs_mark_inode_dirty_sync(inode, true);
2563}
2564
2565static inline void f2fs_i_pino_write(struct inode *inode, nid_t pino)
2566{
2567 F2FS_I(inode)->i_pino = pino;
2568 f2fs_mark_inode_dirty_sync(inode, true);
2569}
2570
2571static inline void get_inline_info(struct inode *inode, struct f2fs_inode *ri)
2572{
2573 struct f2fs_inode_info *fi = F2FS_I(inode);
2574
2575 if (ri->i_inline & F2FS_INLINE_XATTR)
2576 set_bit(FI_INLINE_XATTR, &fi->flags);
2577 if (ri->i_inline & F2FS_INLINE_DATA)
2578 set_bit(FI_INLINE_DATA, &fi->flags);
2579 if (ri->i_inline & F2FS_INLINE_DENTRY)
2580 set_bit(FI_INLINE_DENTRY, &fi->flags);
2581 if (ri->i_inline & F2FS_DATA_EXIST)
2582 set_bit(FI_DATA_EXIST, &fi->flags);
2583 if (ri->i_inline & F2FS_INLINE_DOTS)
2584 set_bit(FI_INLINE_DOTS, &fi->flags);
2585 if (ri->i_inline & F2FS_EXTRA_ATTR)
2586 set_bit(FI_EXTRA_ATTR, &fi->flags);
2587 if (ri->i_inline & F2FS_PIN_FILE)
2588 set_bit(FI_PIN_FILE, &fi->flags);
2589}
2590
2591static inline void set_raw_inline(struct inode *inode, struct f2fs_inode *ri)
2592{
2593 ri->i_inline = 0;
2594
2595 if (is_inode_flag_set(inode, FI_INLINE_XATTR))
2596 ri->i_inline |= F2FS_INLINE_XATTR;
2597 if (is_inode_flag_set(inode, FI_INLINE_DATA))
2598 ri->i_inline |= F2FS_INLINE_DATA;
2599 if (is_inode_flag_set(inode, FI_INLINE_DENTRY))
2600 ri->i_inline |= F2FS_INLINE_DENTRY;
2601 if (is_inode_flag_set(inode, FI_DATA_EXIST))
2602 ri->i_inline |= F2FS_DATA_EXIST;
2603 if (is_inode_flag_set(inode, FI_INLINE_DOTS))
2604 ri->i_inline |= F2FS_INLINE_DOTS;
2605 if (is_inode_flag_set(inode, FI_EXTRA_ATTR))
2606 ri->i_inline |= F2FS_EXTRA_ATTR;
2607 if (is_inode_flag_set(inode, FI_PIN_FILE))
2608 ri->i_inline |= F2FS_PIN_FILE;
2609}
2610
2611static inline int f2fs_has_extra_attr(struct inode *inode)
2612{
2613 return is_inode_flag_set(inode, FI_EXTRA_ATTR);
2614}
2615
2616static inline int f2fs_has_inline_xattr(struct inode *inode)
2617{
2618 return is_inode_flag_set(inode, FI_INLINE_XATTR);
2619}
2620
2621static inline unsigned int addrs_per_inode(struct inode *inode)
2622{
2623 unsigned int addrs = CUR_ADDRS_PER_INODE(inode) -
2624 get_inline_xattr_addrs(inode);
2625 return ALIGN_DOWN(addrs, 1);
2626}
2627
2628static inline unsigned int addrs_per_block(struct inode *inode)
2629{
2630 return ALIGN_DOWN(DEF_ADDRS_PER_BLOCK, 1);
2631}
2632
2633static inline void *inline_xattr_addr(struct inode *inode, struct page *page)
2634{
2635 struct f2fs_inode *ri = F2FS_INODE(page);
2636
2637 return (void *)&(ri->i_addr[DEF_ADDRS_PER_INODE -
2638 get_inline_xattr_addrs(inode)]);
2639}
2640
2641static inline int inline_xattr_size(struct inode *inode)
2642{
2643 if (f2fs_has_inline_xattr(inode))
2644 return get_inline_xattr_addrs(inode) * sizeof(__le32);
2645 return 0;
2646}
2647
2648static inline int f2fs_has_inline_data(struct inode *inode)
2649{
2650 return is_inode_flag_set(inode, FI_INLINE_DATA);
2651}
2652
2653static inline int f2fs_exist_data(struct inode *inode)
2654{
2655 return is_inode_flag_set(inode, FI_DATA_EXIST);
2656}
2657
2658static inline int f2fs_has_inline_dots(struct inode *inode)
2659{
2660 return is_inode_flag_set(inode, FI_INLINE_DOTS);
2661}
2662
2663static inline bool f2fs_is_pinned_file(struct inode *inode)
2664{
2665 return is_inode_flag_set(inode, FI_PIN_FILE);
2666}
2667
2668static inline bool f2fs_is_atomic_file(struct inode *inode)
2669{
2670 return is_inode_flag_set(inode, FI_ATOMIC_FILE);
2671}
2672
2673static inline bool f2fs_is_commit_atomic_write(struct inode *inode)
2674{
2675 return is_inode_flag_set(inode, FI_ATOMIC_COMMIT);
2676}
2677
2678static inline bool f2fs_is_volatile_file(struct inode *inode)
2679{
2680 return is_inode_flag_set(inode, FI_VOLATILE_FILE);
2681}
2682
2683static inline bool f2fs_is_first_block_written(struct inode *inode)
2684{
2685 return is_inode_flag_set(inode, FI_FIRST_BLOCK_WRITTEN);
2686}
2687
2688static inline bool f2fs_is_drop_cache(struct inode *inode)
2689{
2690 return is_inode_flag_set(inode, FI_DROP_CACHE);
2691}
2692
2693static inline void *inline_data_addr(struct inode *inode, struct page *page)
2694{
2695 struct f2fs_inode *ri = F2FS_INODE(page);
2696 int extra_size = get_extra_isize(inode);
2697
2698 return (void *)&(ri->i_addr[extra_size + DEF_INLINE_RESERVED_SIZE]);
2699}
2700
2701static inline int f2fs_has_inline_dentry(struct inode *inode)
2702{
2703 return is_inode_flag_set(inode, FI_INLINE_DENTRY);
2704}
2705
2706static inline int is_file(struct inode *inode, int type)
2707{
2708 return F2FS_I(inode)->i_advise & type;
2709}
2710
2711static inline void set_file(struct inode *inode, int type)
2712{
2713 F2FS_I(inode)->i_advise |= type;
2714 f2fs_mark_inode_dirty_sync(inode, true);
2715}
2716
2717static inline void clear_file(struct inode *inode, int type)
2718{
2719 F2FS_I(inode)->i_advise &= ~type;
2720 f2fs_mark_inode_dirty_sync(inode, true);
2721}
2722
2723static inline bool f2fs_is_time_consistent(struct inode *inode)
2724{
2725 if (!timespec64_equal(F2FS_I(inode)->i_disk_time, &inode->i_atime))
2726 return false;
2727 if (!timespec64_equal(F2FS_I(inode)->i_disk_time + 1, &inode->i_ctime))
2728 return false;
2729 if (!timespec64_equal(F2FS_I(inode)->i_disk_time + 2, &inode->i_mtime))
2730 return false;
2731 if (!timespec64_equal(F2FS_I(inode)->i_disk_time + 3,
2732 &F2FS_I(inode)->i_crtime))
2733 return false;
2734 return true;
2735}
2736
2737static inline bool f2fs_skip_inode_update(struct inode *inode, int dsync)
2738{
2739 bool ret;
2740
2741 if (dsync) {
2742 struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
2743
2744 spin_lock(&sbi->inode_lock[DIRTY_META]);
2745 ret = list_empty(&F2FS_I(inode)->gdirty_list);
2746 spin_unlock(&sbi->inode_lock[DIRTY_META]);
2747 return ret;
2748 }
2749 if (!is_inode_flag_set(inode, FI_AUTO_RECOVER) ||
2750 file_keep_isize(inode) ||
2751 i_size_read(inode) & ~PAGE_MASK)
2752 return false;
2753
2754 if (!f2fs_is_time_consistent(inode))
2755 return false;
2756
2757 down_read(&F2FS_I(inode)->i_sem);
2758 ret = F2FS_I(inode)->last_disk_size == i_size_read(inode);
2759 up_read(&F2FS_I(inode)->i_sem);
2760
2761 return ret;
2762}
2763
2764static inline bool f2fs_readonly(struct super_block *sb)
2765{
2766 return sb_rdonly(sb);
2767}
2768
2769static inline bool f2fs_cp_error(struct f2fs_sb_info *sbi)
2770{
2771 return is_set_ckpt_flags(sbi, CP_ERROR_FLAG);
2772}
2773
2774static inline bool is_dot_dotdot(const struct qstr *str)
2775{
2776 if (str->len == 1 && str->name[0] == '.')
2777 return true;
2778
2779 if (str->len == 2 && str->name[0] == '.' && str->name[1] == '.')
2780 return true;
2781
2782 return false;
2783}
2784
2785static inline bool f2fs_may_extent_tree(struct inode *inode)
2786{
2787 struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
2788
2789 if (!test_opt(sbi, EXTENT_CACHE) ||
2790 is_inode_flag_set(inode, FI_NO_EXTENT))
2791 return false;
2792
2793 /*
2794 * for recovered files during mount do not create extents
2795 * if shrinker is not registered.
2796 */
2797 if (list_empty(&sbi->s_list))
2798 return false;
2799
2800 return S_ISREG(inode->i_mode);
2801}
2802
2803static inline void *f2fs_kmalloc(struct f2fs_sb_info *sbi,
2804 size_t size, gfp_t flags)
2805{
2806 void *ret;
2807
2808 if (time_to_inject(sbi, FAULT_KMALLOC)) {
2809 f2fs_show_injection_info(sbi, FAULT_KMALLOC);
2810 return NULL;
2811 }
2812
2813 ret = kmalloc(size, flags);
2814 if (ret)
2815 return ret;
2816
2817 return kvmalloc(size, flags);
2818}
2819
2820static inline void *f2fs_kzalloc(struct f2fs_sb_info *sbi,
2821 size_t size, gfp_t flags)
2822{
2823 return f2fs_kmalloc(sbi, size, flags | __GFP_ZERO);
2824}
2825
2826static inline void *f2fs_kvmalloc(struct f2fs_sb_info *sbi,
2827 size_t size, gfp_t flags)
2828{
2829 if (time_to_inject(sbi, FAULT_KVMALLOC)) {
2830 f2fs_show_injection_info(sbi, FAULT_KVMALLOC);
2831 return NULL;
2832 }
2833
2834 return kvmalloc(size, flags);
2835}
2836
2837static inline void *f2fs_kvzalloc(struct f2fs_sb_info *sbi,
2838 size_t size, gfp_t flags)
2839{
2840 return f2fs_kvmalloc(sbi, size, flags | __GFP_ZERO);
2841}
2842
2843static inline int get_extra_isize(struct inode *inode)
2844{
2845 return F2FS_I(inode)->i_extra_isize / sizeof(__le32);
2846}
2847
2848static inline int get_inline_xattr_addrs(struct inode *inode)
2849{
2850 return F2FS_I(inode)->i_inline_xattr_size;
2851}
2852
2853#define f2fs_get_inode_mode(i) \
2854 ((is_inode_flag_set(i, FI_ACL_MODE)) ? \
2855 (F2FS_I(i)->i_acl_mode) : ((i)->i_mode))
2856
2857#define F2FS_TOTAL_EXTRA_ATTR_SIZE \
2858 (offsetof(struct f2fs_inode, i_extra_end) - \
2859 offsetof(struct f2fs_inode, i_extra_isize)) \
2860
2861#define F2FS_OLD_ATTRIBUTE_SIZE (offsetof(struct f2fs_inode, i_addr))
2862#define F2FS_FITS_IN_INODE(f2fs_inode, extra_isize, field) \
2863 ((offsetof(typeof(*(f2fs_inode)), field) + \
2864 sizeof((f2fs_inode)->field)) \
2865 <= (F2FS_OLD_ATTRIBUTE_SIZE + (extra_isize))) \
2866
2867static inline void f2fs_reset_iostat(struct f2fs_sb_info *sbi)
2868{
2869 int i;
2870
2871 spin_lock(&sbi->iostat_lock);
2872 for (i = 0; i < NR_IO_TYPE; i++)
2873 sbi->write_iostat[i] = 0;
2874 spin_unlock(&sbi->iostat_lock);
2875}
2876
2877static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi,
2878 enum iostat_type type, unsigned long long io_bytes)
2879{
2880 if (!sbi->iostat_enable)
2881 return;
2882 spin_lock(&sbi->iostat_lock);
2883 sbi->write_iostat[type] += io_bytes;
2884
2885 if (type == APP_WRITE_IO || type == APP_DIRECT_IO)
2886 sbi->write_iostat[APP_BUFFERED_IO] =
2887 sbi->write_iostat[APP_WRITE_IO] -
2888 sbi->write_iostat[APP_DIRECT_IO];
2889 spin_unlock(&sbi->iostat_lock);
2890}
2891
2892#define __is_large_section(sbi) ((sbi)->segs_per_sec > 1)
2893
2894#define __is_meta_io(fio) (PAGE_TYPE_OF_BIO((fio)->type) == META)
2895
2896bool f2fs_is_valid_blkaddr(struct f2fs_sb_info *sbi,
2897 block_t blkaddr, int type);
2898static inline void verify_blkaddr(struct f2fs_sb_info *sbi,
2899 block_t blkaddr, int type)
2900{
2901 if (!f2fs_is_valid_blkaddr(sbi, blkaddr, type)) {
2902 f2fs_err(sbi, "invalid blkaddr: %u, type: %d, run fsck to fix.",
2903 blkaddr, type);
2904 f2fs_bug_on(sbi, 1);
2905 }
2906}
2907
2908static inline bool __is_valid_data_blkaddr(block_t blkaddr)
2909{
2910 if (blkaddr == NEW_ADDR || blkaddr == NULL_ADDR)
2911 return false;
2912 return true;
2913}
2914
2915static inline void f2fs_set_page_private(struct page *page,
2916 unsigned long data)
2917{
2918 if (PagePrivate(page))
2919 return;
2920
2921 get_page(page);
2922 SetPagePrivate(page);
2923 set_page_private(page, data);
2924}
2925
2926static inline void f2fs_clear_page_private(struct page *page)
2927{
2928 if (!PagePrivate(page))
2929 return;
2930
2931 set_page_private(page, 0);
2932 ClearPagePrivate(page);
2933 f2fs_put_page(page, 0);
2934}
2935
2936/*
2937 * file.c
2938 */
2939int f2fs_sync_file(struct file *file, loff_t start, loff_t end, int datasync);
2940void f2fs_truncate_data_blocks(struct dnode_of_data *dn);
2941int f2fs_truncate_blocks(struct inode *inode, u64 from, bool lock);
2942int f2fs_truncate(struct inode *inode);
2943int f2fs_getattr(const struct path *path, struct kstat *stat,
2944 u32 request_mask, unsigned int flags);
2945int f2fs_setattr(struct dentry *dentry, struct iattr *attr);
2946int f2fs_truncate_hole(struct inode *inode, pgoff_t pg_start, pgoff_t pg_end);
2947void f2fs_truncate_data_blocks_range(struct dnode_of_data *dn, int count);
2948int f2fs_precache_extents(struct inode *inode);
2949long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg);
2950long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
2951int f2fs_transfer_project_quota(struct inode *inode, kprojid_t kprojid);
2952int f2fs_pin_file_control(struct inode *inode, bool inc);
2953
2954/*
2955 * inode.c
2956 */
2957void f2fs_set_inode_flags(struct inode *inode);
2958bool f2fs_inode_chksum_verify(struct f2fs_sb_info *sbi, struct page *page);
2959void f2fs_inode_chksum_set(struct f2fs_sb_info *sbi, struct page *page);
2960struct inode *f2fs_iget(struct super_block *sb, unsigned long ino);
2961struct inode *f2fs_iget_retry(struct super_block *sb, unsigned long ino);
2962int f2fs_try_to_free_nats(struct f2fs_sb_info *sbi, int nr_shrink);
2963void f2fs_update_inode(struct inode *inode, struct page *node_page);
2964void f2fs_update_inode_page(struct inode *inode);
2965int f2fs_write_inode(struct inode *inode, struct writeback_control *wbc);
2966void f2fs_evict_inode(struct inode *inode);
2967void f2fs_handle_failed_inode(struct inode *inode);
2968
2969/*
2970 * namei.c
2971 */
2972int f2fs_update_extension_list(struct f2fs_sb_info *sbi, const char *name,
2973 bool hot, bool set);
2974struct dentry *f2fs_get_parent(struct dentry *child);
2975
2976extern int f2fs_ci_compare(const struct inode *parent,
2977 const struct qstr *name,
2978 const struct qstr *entry,
2979 bool quick);
2980
2981/*
2982 * dir.c
2983 */
2984unsigned char f2fs_get_de_type(struct f2fs_dir_entry *de);
2985struct f2fs_dir_entry *f2fs_find_target_dentry(struct fscrypt_name *fname,
2986 f2fs_hash_t namehash, int *max_slots,
2987 struct f2fs_dentry_ptr *d);
2988int f2fs_fill_dentries(struct dir_context *ctx, struct f2fs_dentry_ptr *d,
2989 unsigned int start_pos, struct fscrypt_str *fstr);
2990void f2fs_do_make_empty_dir(struct inode *inode, struct inode *parent,
2991 struct f2fs_dentry_ptr *d);
2992struct page *f2fs_init_inode_metadata(struct inode *inode, struct inode *dir,
2993 const struct qstr *new_name,
2994 const struct qstr *orig_name, struct page *dpage);
2995void f2fs_update_parent_metadata(struct inode *dir, struct inode *inode,
2996 unsigned int current_depth);
2997int f2fs_room_for_filename(const void *bitmap, int slots, int max_slots);
2998void f2fs_drop_nlink(struct inode *dir, struct inode *inode);
2999struct f2fs_dir_entry *__f2fs_find_entry(struct inode *dir,
3000 struct fscrypt_name *fname, struct page **res_page);
3001struct f2fs_dir_entry *f2fs_find_entry(struct inode *dir,
3002 const struct qstr *child, struct page **res_page);
3003struct f2fs_dir_entry *f2fs_parent_dir(struct inode *dir, struct page **p);
3004ino_t f2fs_inode_by_name(struct inode *dir, const struct qstr *qstr,
3005 struct page **page);
3006void f2fs_set_link(struct inode *dir, struct f2fs_dir_entry *de,
3007 struct page *page, struct inode *inode);
3008void f2fs_update_dentry(nid_t ino, umode_t mode, struct f2fs_dentry_ptr *d,
3009 const struct qstr *name, f2fs_hash_t name_hash,
3010 unsigned int bit_pos);
3011int f2fs_add_regular_entry(struct inode *dir, const struct qstr *new_name,
3012 const struct qstr *orig_name,
3013 struct inode *inode, nid_t ino, umode_t mode);
3014int f2fs_add_dentry(struct inode *dir, struct fscrypt_name *fname,
3015 struct inode *inode, nid_t ino, umode_t mode);
3016int f2fs_do_add_link(struct inode *dir, const struct qstr *name,
3017 struct inode *inode, nid_t ino, umode_t mode);
3018void f2fs_delete_entry(struct f2fs_dir_entry *dentry, struct page *page,
3019 struct inode *dir, struct inode *inode);
3020int f2fs_do_tmpfile(struct inode *inode, struct inode *dir);
3021bool f2fs_empty_dir(struct inode *dir);
3022
3023static inline int f2fs_add_link(struct dentry *dentry, struct inode *inode)
3024{
3025 return f2fs_do_add_link(d_inode(dentry->d_parent), &dentry->d_name,
3026 inode, inode->i_ino, inode->i_mode);
3027}
3028
3029/*
3030 * super.c
3031 */
3032int f2fs_inode_dirtied(struct inode *inode, bool sync);
3033void f2fs_inode_synced(struct inode *inode);
3034int f2fs_enable_quota_files(struct f2fs_sb_info *sbi, bool rdonly);
3035int f2fs_quota_sync(struct super_block *sb, int type);
3036void f2fs_quota_off_umount(struct super_block *sb);
3037int f2fs_commit_super(struct f2fs_sb_info *sbi, bool recover);
3038int f2fs_sync_fs(struct super_block *sb, int sync);
3039int f2fs_sanity_check_ckpt(struct f2fs_sb_info *sbi);
3040
3041/*
3042 * hash.c
3043 */
3044f2fs_hash_t f2fs_dentry_hash(const struct inode *dir,
3045 const struct qstr *name_info, struct fscrypt_name *fname);
3046
3047/*
3048 * node.c
3049 */
3050struct dnode_of_data;
3051struct node_info;
3052
3053int f2fs_check_nid_range(struct f2fs_sb_info *sbi, nid_t nid);
3054bool f2fs_available_free_memory(struct f2fs_sb_info *sbi, int type);
3055bool f2fs_in_warm_node_list(struct f2fs_sb_info *sbi, struct page *page);
3056void f2fs_init_fsync_node_info(struct f2fs_sb_info *sbi);
3057void f2fs_del_fsync_node_entry(struct f2fs_sb_info *sbi, struct page *page);
3058void f2fs_reset_fsync_node_info(struct f2fs_sb_info *sbi);
3059int f2fs_need_dentry_mark(struct f2fs_sb_info *sbi, nid_t nid);
3060bool f2fs_is_checkpointed_node(struct f2fs_sb_info *sbi, nid_t nid);
3061bool f2fs_need_inode_block_update(struct f2fs_sb_info *sbi, nid_t ino);
3062int f2fs_get_node_info(struct f2fs_sb_info *sbi, nid_t nid,
3063 struct node_info *ni);
3064pgoff_t f2fs_get_next_page_offset(struct dnode_of_data *dn, pgoff_t pgofs);
3065int f2fs_get_dnode_of_data(struct dnode_of_data *dn, pgoff_t index, int mode);
3066int f2fs_truncate_inode_blocks(struct inode *inode, pgoff_t from);
3067int f2fs_truncate_xattr_node(struct inode *inode);
3068int f2fs_wait_on_node_pages_writeback(struct f2fs_sb_info *sbi,
3069 unsigned int seq_id);
3070int f2fs_remove_inode_page(struct inode *inode);
3071struct page *f2fs_new_inode_page(struct inode *inode);
3072struct page *f2fs_new_node_page(struct dnode_of_data *dn, unsigned int ofs);
3073void f2fs_ra_node_page(struct f2fs_sb_info *sbi, nid_t nid);
3074struct page *f2fs_get_node_page(struct f2fs_sb_info *sbi, pgoff_t nid);
3075struct page *f2fs_get_node_page_ra(struct page *parent, int start);
3076int f2fs_move_node_page(struct page *node_page, int gc_type);
3077int f2fs_fsync_node_pages(struct f2fs_sb_info *sbi, struct inode *inode,
3078 struct writeback_control *wbc, bool atomic,
3079 unsigned int *seq_id);
3080int f2fs_sync_node_pages(struct f2fs_sb_info *sbi,
3081 struct writeback_control *wbc,
3082 bool do_balance, enum iostat_type io_type);
3083int f2fs_build_free_nids(struct f2fs_sb_info *sbi, bool sync, bool mount);
3084bool f2fs_alloc_nid(struct f2fs_sb_info *sbi, nid_t *nid);
3085void f2fs_alloc_nid_done(struct f2fs_sb_info *sbi, nid_t nid);
3086void f2fs_alloc_nid_failed(struct f2fs_sb_info *sbi, nid_t nid);
3087int f2fs_try_to_free_nids(struct f2fs_sb_info *sbi, int nr_shrink);
3088void f2fs_recover_inline_xattr(struct inode *inode, struct page *page);
3089int f2fs_recover_xattr_data(struct inode *inode, struct page *page);
3090int f2fs_recover_inode_page(struct f2fs_sb_info *sbi, struct page *page);
3091int f2fs_restore_node_summary(struct f2fs_sb_info *sbi,
3092 unsigned int segno, struct f2fs_summary_block *sum);
3093int f2fs_flush_nat_entries(struct f2fs_sb_info *sbi, struct cp_control *cpc);
3094int f2fs_build_node_manager(struct f2fs_sb_info *sbi);
3095void f2fs_destroy_node_manager(struct f2fs_sb_info *sbi);
3096int __init f2fs_create_node_manager_caches(void);
3097void f2fs_destroy_node_manager_caches(void);
3098
3099/*
3100 * segment.c
3101 */
3102bool f2fs_need_SSR(struct f2fs_sb_info *sbi);
3103void f2fs_register_inmem_page(struct inode *inode, struct page *page);
3104void f2fs_drop_inmem_pages_all(struct f2fs_sb_info *sbi, bool gc_failure);
3105void f2fs_drop_inmem_pages(struct inode *inode);
3106void f2fs_drop_inmem_page(struct inode *inode, struct page *page);
3107int f2fs_commit_inmem_pages(struct inode *inode);
3108void f2fs_balance_fs(struct f2fs_sb_info *sbi, bool need);
3109void f2fs_balance_fs_bg(struct f2fs_sb_info *sbi);
3110int f2fs_issue_flush(struct f2fs_sb_info *sbi, nid_t ino);
3111int f2fs_create_flush_cmd_control(struct f2fs_sb_info *sbi);
3112int f2fs_flush_device_cache(struct f2fs_sb_info *sbi);
3113void f2fs_destroy_flush_cmd_control(struct f2fs_sb_info *sbi, bool free);
3114void f2fs_invalidate_blocks(struct f2fs_sb_info *sbi, block_t addr);
3115bool f2fs_is_checkpointed_data(struct f2fs_sb_info *sbi, block_t blkaddr);
3116void f2fs_drop_discard_cmd(struct f2fs_sb_info *sbi);
3117void f2fs_stop_discard_thread(struct f2fs_sb_info *sbi);
3118bool f2fs_issue_discard_timeout(struct f2fs_sb_info *sbi);
3119void f2fs_clear_prefree_segments(struct f2fs_sb_info *sbi,
3120 struct cp_control *cpc);
3121void f2fs_dirty_to_prefree(struct f2fs_sb_info *sbi);
3122block_t f2fs_get_unusable_blocks(struct f2fs_sb_info *sbi);
3123int f2fs_disable_cp_again(struct f2fs_sb_info *sbi, block_t unusable);
3124void f2fs_release_discard_addrs(struct f2fs_sb_info *sbi);
3125int f2fs_npages_for_summary_flush(struct f2fs_sb_info *sbi, bool for_ra);
3126void allocate_segment_for_resize(struct f2fs_sb_info *sbi, int type,
3127 unsigned int start, unsigned int end);
3128void f2fs_allocate_new_segments(struct f2fs_sb_info *sbi, int type);
3129int f2fs_trim_fs(struct f2fs_sb_info *sbi, struct fstrim_range *range);
3130bool f2fs_exist_trim_candidates(struct f2fs_sb_info *sbi,
3131 struct cp_control *cpc);
3132struct page *f2fs_get_sum_page(struct f2fs_sb_info *sbi, unsigned int segno);
3133void f2fs_update_meta_page(struct f2fs_sb_info *sbi, void *src,
3134 block_t blk_addr);
3135void f2fs_do_write_meta_page(struct f2fs_sb_info *sbi, struct page *page,
3136 enum iostat_type io_type);
3137void f2fs_do_write_node_page(unsigned int nid, struct f2fs_io_info *fio);
3138void f2fs_outplace_write_data(struct dnode_of_data *dn,
3139 struct f2fs_io_info *fio);
3140int f2fs_inplace_write_data(struct f2fs_io_info *fio);
3141void f2fs_do_replace_block(struct f2fs_sb_info *sbi, struct f2fs_summary *sum,
3142 block_t old_blkaddr, block_t new_blkaddr,
3143 bool recover_curseg, bool recover_newaddr);
3144void f2fs_replace_block(struct f2fs_sb_info *sbi, struct dnode_of_data *dn,
3145 block_t old_addr, block_t new_addr,
3146 unsigned char version, bool recover_curseg,
3147 bool recover_newaddr);
3148void f2fs_allocate_data_block(struct f2fs_sb_info *sbi, struct page *page,
3149 block_t old_blkaddr, block_t *new_blkaddr,
3150 struct f2fs_summary *sum, int type,
3151 struct f2fs_io_info *fio, bool add_list);
3152void f2fs_wait_on_page_writeback(struct page *page,
3153 enum page_type type, bool ordered, bool locked);
3154void f2fs_wait_on_block_writeback(struct inode *inode, block_t blkaddr);
3155void f2fs_wait_on_block_writeback_range(struct inode *inode, block_t blkaddr,
3156 block_t len);
3157void f2fs_write_data_summaries(struct f2fs_sb_info *sbi, block_t start_blk);
3158void f2fs_write_node_summaries(struct f2fs_sb_info *sbi, block_t start_blk);
3159int f2fs_lookup_journal_in_cursum(struct f2fs_journal *journal, int type,
3160 unsigned int val, int alloc);
3161void f2fs_flush_sit_entries(struct f2fs_sb_info *sbi, struct cp_control *cpc);
3162int f2fs_build_segment_manager(struct f2fs_sb_info *sbi);
3163void f2fs_destroy_segment_manager(struct f2fs_sb_info *sbi);
3164int __init f2fs_create_segment_manager_caches(void);
3165void f2fs_destroy_segment_manager_caches(void);
3166int f2fs_rw_hint_to_seg_type(enum rw_hint hint);
3167enum rw_hint f2fs_io_type_to_rw_hint(struct f2fs_sb_info *sbi,
3168 enum page_type type, enum temp_type temp);
3169
3170/*
3171 * checkpoint.c
3172 */
3173void f2fs_stop_checkpoint(struct f2fs_sb_info *sbi, bool end_io);
3174struct page *f2fs_grab_meta_page(struct f2fs_sb_info *sbi, pgoff_t index);
3175struct page *f2fs_get_meta_page(struct f2fs_sb_info *sbi, pgoff_t index);
3176struct page *f2fs_get_meta_page_nofail(struct f2fs_sb_info *sbi, pgoff_t index);
3177struct page *f2fs_get_tmp_page(struct f2fs_sb_info *sbi, pgoff_t index);
3178bool f2fs_is_valid_blkaddr(struct f2fs_sb_info *sbi,
3179 block_t blkaddr, int type);
3180int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_t start, int nrpages,
3181 int type, bool sync);
3182void f2fs_ra_meta_pages_cond(struct f2fs_sb_info *sbi, pgoff_t index);
3183long f2fs_sync_meta_pages(struct f2fs_sb_info *sbi, enum page_type type,
3184 long nr_to_write, enum iostat_type io_type);
3185void f2fs_add_ino_entry(struct f2fs_sb_info *sbi, nid_t ino, int type);
3186void f2fs_remove_ino_entry(struct f2fs_sb_info *sbi, nid_t ino, int type);
3187void f2fs_release_ino_entry(struct f2fs_sb_info *sbi, bool all);
3188bool f2fs_exist_written_data(struct f2fs_sb_info *sbi, nid_t ino, int mode);
3189void f2fs_set_dirty_device(struct f2fs_sb_info *sbi, nid_t ino,
3190 unsigned int devidx, int type);
3191bool f2fs_is_dirty_device(struct f2fs_sb_info *sbi, nid_t ino,
3192 unsigned int devidx, int type);
3193int f2fs_sync_inode_meta(struct f2fs_sb_info *sbi);
3194int f2fs_acquire_orphan_inode(struct f2fs_sb_info *sbi);
3195void f2fs_release_orphan_inode(struct f2fs_sb_info *sbi);
3196void f2fs_add_orphan_inode(struct inode *inode);
3197void f2fs_remove_orphan_inode(struct f2fs_sb_info *sbi, nid_t ino);
3198int f2fs_recover_orphan_inodes(struct f2fs_sb_info *sbi);
3199int f2fs_get_valid_checkpoint(struct f2fs_sb_info *sbi);
3200void f2fs_update_dirty_page(struct inode *inode, struct page *page);
3201void f2fs_remove_dirty_inode(struct inode *inode);
3202int f2fs_sync_dirty_inodes(struct f2fs_sb_info *sbi, enum inode_type type);
3203void f2fs_wait_on_all_pages_writeback(struct f2fs_sb_info *sbi);
3204int f2fs_write_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc);
3205void f2fs_init_ino_entry_info(struct f2fs_sb_info *sbi);
3206int __init f2fs_create_checkpoint_caches(void);
3207void f2fs_destroy_checkpoint_caches(void);
3208
3209/*
3210 * data.c
3211 */
3212int f2fs_init_post_read_processing(void);
3213void f2fs_destroy_post_read_processing(void);
3214int f2fs_init_bio_entry_cache(void);
3215void f2fs_destroy_bio_entry_cache(void);
3216void f2fs_submit_merged_write(struct f2fs_sb_info *sbi, enum page_type type);
3217void f2fs_submit_merged_write_cond(struct f2fs_sb_info *sbi,
3218 struct inode *inode, struct page *page,
3219 nid_t ino, enum page_type type);
3220void f2fs_submit_merged_ipu_write(struct f2fs_sb_info *sbi,
3221 struct bio **bio, struct page *page);
3222void f2fs_flush_merged_writes(struct f2fs_sb_info *sbi);
3223int f2fs_submit_page_bio(struct f2fs_io_info *fio);
3224int f2fs_merge_page_bio(struct f2fs_io_info *fio);
3225void f2fs_submit_page_write(struct f2fs_io_info *fio);
3226struct block_device *f2fs_target_device(struct f2fs_sb_info *sbi,
3227 block_t blk_addr, struct bio *bio);
3228int f2fs_target_device_index(struct f2fs_sb_info *sbi, block_t blkaddr);
3229void f2fs_set_data_blkaddr(struct dnode_of_data *dn);
3230void f2fs_update_data_blkaddr(struct dnode_of_data *dn, block_t blkaddr);
3231int f2fs_reserve_new_blocks(struct dnode_of_data *dn, blkcnt_t count);
3232int f2fs_reserve_new_block(struct dnode_of_data *dn);
3233int f2fs_get_block(struct dnode_of_data *dn, pgoff_t index);
3234int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *from);
3235int f2fs_reserve_block(struct dnode_of_data *dn, pgoff_t index);
3236struct page *f2fs_get_read_data_page(struct inode *inode, pgoff_t index,
3237 int op_flags, bool for_write);
3238struct page *f2fs_find_data_page(struct inode *inode, pgoff_t index);
3239struct page *f2fs_get_lock_data_page(struct inode *inode, pgoff_t index,
3240 bool for_write);
3241struct page *f2fs_get_new_data_page(struct inode *inode,
3242 struct page *ipage, pgoff_t index, bool new_i_size);
3243int f2fs_do_write_data_page(struct f2fs_io_info *fio);
3244void __do_map_lock(struct f2fs_sb_info *sbi, int flag, bool lock);
3245int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map,
3246 int create, int flag);
3247int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
3248 u64 start, u64 len);
3249bool f2fs_should_update_inplace(struct inode *inode, struct f2fs_io_info *fio);
3250bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio);
3251void f2fs_invalidate_page(struct page *page, unsigned int offset,
3252 unsigned int length);
3253int f2fs_release_page(struct page *page, gfp_t wait);
3254#ifdef CONFIG_MIGRATION
3255int f2fs_migrate_page(struct address_space *mapping, struct page *newpage,
3256 struct page *page, enum migrate_mode mode);
3257#endif
3258bool f2fs_overwrite_io(struct inode *inode, loff_t pos, size_t len);
3259void f2fs_clear_radix_tree_dirty_tag(struct page *page);
3260
3261/*
3262 * gc.c
3263 */
3264int f2fs_start_gc_thread(struct f2fs_sb_info *sbi);
3265void f2fs_stop_gc_thread(struct f2fs_sb_info *sbi);
3266block_t f2fs_start_bidx_of_node(unsigned int node_ofs, struct inode *inode);
3267int f2fs_gc(struct f2fs_sb_info *sbi, bool sync, bool background,
3268 unsigned int segno);
3269void f2fs_build_gc_manager(struct f2fs_sb_info *sbi);
3270int f2fs_resize_fs(struct f2fs_sb_info *sbi, __u64 block_count);
3271
3272/*
3273 * recovery.c
3274 */
3275int f2fs_recover_fsync_data(struct f2fs_sb_info *sbi, bool check_only);
3276bool f2fs_space_for_roll_forward(struct f2fs_sb_info *sbi);
3277
3278/*
3279 * debug.c
3280 */
3281#ifdef CONFIG_F2FS_STAT_FS
3282struct f2fs_stat_info {
3283 struct list_head stat_list;
3284 struct f2fs_sb_info *sbi;
3285 int all_area_segs, sit_area_segs, nat_area_segs, ssa_area_segs;
3286 int main_area_segs, main_area_sections, main_area_zones;
3287 unsigned long long hit_largest, hit_cached, hit_rbtree;
3288 unsigned long long hit_total, total_ext;
3289 int ext_tree, zombie_tree, ext_node;
3290 int ndirty_node, ndirty_dent, ndirty_meta, ndirty_imeta;
3291 int ndirty_data, ndirty_qdata;
3292 int inmem_pages;
3293 unsigned int ndirty_dirs, ndirty_files, nquota_files, ndirty_all;
3294 int nats, dirty_nats, sits, dirty_sits;
3295 int free_nids, avail_nids, alloc_nids;
3296 int total_count, utilization;
3297 int bg_gc, nr_wb_cp_data, nr_wb_data;
3298 int nr_rd_data, nr_rd_node, nr_rd_meta;
3299 int nr_dio_read, nr_dio_write;
3300 unsigned int io_skip_bggc, other_skip_bggc;
3301 int nr_flushing, nr_flushed, flush_list_empty;
3302 int nr_discarding, nr_discarded;
3303 int nr_discard_cmd;
3304 unsigned int undiscard_blks;
3305 int inline_xattr, inline_inode, inline_dir, append, update, orphans;
3306 int aw_cnt, max_aw_cnt, vw_cnt, max_vw_cnt;
3307 unsigned int valid_count, valid_node_count, valid_inode_count, discard_blks;
3308 unsigned int bimodal, avg_vblocks;
3309 int util_free, util_valid, util_invalid;
3310 int rsvd_segs, overp_segs;
3311 int dirty_count, node_pages, meta_pages;
3312 int prefree_count, call_count, cp_count, bg_cp_count;
3313 int tot_segs, node_segs, data_segs, free_segs, free_secs;
3314 int bg_node_segs, bg_data_segs;
3315 int tot_blks, data_blks, node_blks;
3316 int bg_data_blks, bg_node_blks;
3317 unsigned long long skipped_atomic_files[2];
3318 int curseg[NR_CURSEG_TYPE];
3319 int cursec[NR_CURSEG_TYPE];
3320 int curzone[NR_CURSEG_TYPE];
3321
3322 unsigned int meta_count[META_MAX];
3323 unsigned int segment_count[2];
3324 unsigned int block_count[2];
3325 unsigned int inplace_count;
3326 unsigned long long base_mem, cache_mem, page_mem;
3327};
3328
3329static inline struct f2fs_stat_info *F2FS_STAT(struct f2fs_sb_info *sbi)
3330{
3331 return (struct f2fs_stat_info *)sbi->stat_info;
3332}
3333
3334#define stat_inc_cp_count(si) ((si)->cp_count++)
3335#define stat_inc_bg_cp_count(si) ((si)->bg_cp_count++)
3336#define stat_inc_call_count(si) ((si)->call_count++)
3337#define stat_inc_bggc_count(sbi) ((sbi)->bg_gc++)
3338#define stat_io_skip_bggc_count(sbi) ((sbi)->io_skip_bggc++)
3339#define stat_other_skip_bggc_count(sbi) ((sbi)->other_skip_bggc++)
3340#define stat_inc_dirty_inode(sbi, type) ((sbi)->ndirty_inode[type]++)
3341#define stat_dec_dirty_inode(sbi, type) ((sbi)->ndirty_inode[type]--)
3342#define stat_inc_total_hit(sbi) (atomic64_inc(&(sbi)->total_hit_ext))
3343#define stat_inc_rbtree_node_hit(sbi) (atomic64_inc(&(sbi)->read_hit_rbtree))
3344#define stat_inc_largest_node_hit(sbi) (atomic64_inc(&(sbi)->read_hit_largest))
3345#define stat_inc_cached_node_hit(sbi) (atomic64_inc(&(sbi)->read_hit_cached))
3346#define stat_inc_inline_xattr(inode) \
3347 do { \
3348 if (f2fs_has_inline_xattr(inode)) \
3349 (atomic_inc(&F2FS_I_SB(inode)->inline_xattr)); \
3350 } while (0)
3351#define stat_dec_inline_xattr(inode) \
3352 do { \
3353 if (f2fs_has_inline_xattr(inode)) \
3354 (atomic_dec(&F2FS_I_SB(inode)->inline_xattr)); \
3355 } while (0)
3356#define stat_inc_inline_inode(inode) \
3357 do { \
3358 if (f2fs_has_inline_data(inode)) \
3359 (atomic_inc(&F2FS_I_SB(inode)->inline_inode)); \
3360 } while (0)
3361#define stat_dec_inline_inode(inode) \
3362 do { \
3363 if (f2fs_has_inline_data(inode)) \
3364 (atomic_dec(&F2FS_I_SB(inode)->inline_inode)); \
3365 } while (0)
3366#define stat_inc_inline_dir(inode) \
3367 do { \
3368 if (f2fs_has_inline_dentry(inode)) \
3369 (atomic_inc(&F2FS_I_SB(inode)->inline_dir)); \
3370 } while (0)
3371#define stat_dec_inline_dir(inode) \
3372 do { \
3373 if (f2fs_has_inline_dentry(inode)) \
3374 (atomic_dec(&F2FS_I_SB(inode)->inline_dir)); \
3375 } while (0)
3376#define stat_inc_meta_count(sbi, blkaddr) \
3377 do { \
3378 if (blkaddr < SIT_I(sbi)->sit_base_addr) \
3379 atomic_inc(&(sbi)->meta_count[META_CP]); \
3380 else if (blkaddr < NM_I(sbi)->nat_blkaddr) \
3381 atomic_inc(&(sbi)->meta_count[META_SIT]); \
3382 else if (blkaddr < SM_I(sbi)->ssa_blkaddr) \
3383 atomic_inc(&(sbi)->meta_count[META_NAT]); \
3384 else if (blkaddr < SM_I(sbi)->main_blkaddr) \
3385 atomic_inc(&(sbi)->meta_count[META_SSA]); \
3386 } while (0)
3387#define stat_inc_seg_type(sbi, curseg) \
3388 ((sbi)->segment_count[(curseg)->alloc_type]++)
3389#define stat_inc_block_count(sbi, curseg) \
3390 ((sbi)->block_count[(curseg)->alloc_type]++)
3391#define stat_inc_inplace_blocks(sbi) \
3392 (atomic_inc(&(sbi)->inplace_count))
3393#define stat_inc_atomic_write(inode) \
3394 (atomic_inc(&F2FS_I_SB(inode)->aw_cnt))
3395#define stat_dec_atomic_write(inode) \
3396 (atomic_dec(&F2FS_I_SB(inode)->aw_cnt))
3397#define stat_update_max_atomic_write(inode) \
3398 do { \
3399 int cur = atomic_read(&F2FS_I_SB(inode)->aw_cnt); \
3400 int max = atomic_read(&F2FS_I_SB(inode)->max_aw_cnt); \
3401 if (cur > max) \
3402 atomic_set(&F2FS_I_SB(inode)->max_aw_cnt, cur); \
3403 } while (0)
3404#define stat_inc_volatile_write(inode) \
3405 (atomic_inc(&F2FS_I_SB(inode)->vw_cnt))
3406#define stat_dec_volatile_write(inode) \
3407 (atomic_dec(&F2FS_I_SB(inode)->vw_cnt))
3408#define stat_update_max_volatile_write(inode) \
3409 do { \
3410 int cur = atomic_read(&F2FS_I_SB(inode)->vw_cnt); \
3411 int max = atomic_read(&F2FS_I_SB(inode)->max_vw_cnt); \
3412 if (cur > max) \
3413 atomic_set(&F2FS_I_SB(inode)->max_vw_cnt, cur); \
3414 } while (0)
3415#define stat_inc_seg_count(sbi, type, gc_type) \
3416 do { \
3417 struct f2fs_stat_info *si = F2FS_STAT(sbi); \
3418 si->tot_segs++; \
3419 if ((type) == SUM_TYPE_DATA) { \
3420 si->data_segs++; \
3421 si->bg_data_segs += (gc_type == BG_GC) ? 1 : 0; \
3422 } else { \
3423 si->node_segs++; \
3424 si->bg_node_segs += (gc_type == BG_GC) ? 1 : 0; \
3425 } \
3426 } while (0)
3427
3428#define stat_inc_tot_blk_count(si, blks) \
3429 ((si)->tot_blks += (blks))
3430
3431#define stat_inc_data_blk_count(sbi, blks, gc_type) \
3432 do { \
3433 struct f2fs_stat_info *si = F2FS_STAT(sbi); \
3434 stat_inc_tot_blk_count(si, blks); \
3435 si->data_blks += (blks); \
3436 si->bg_data_blks += ((gc_type) == BG_GC) ? (blks) : 0; \
3437 } while (0)
3438
3439#define stat_inc_node_blk_count(sbi, blks, gc_type) \
3440 do { \
3441 struct f2fs_stat_info *si = F2FS_STAT(sbi); \
3442 stat_inc_tot_blk_count(si, blks); \
3443 si->node_blks += (blks); \
3444 si->bg_node_blks += ((gc_type) == BG_GC) ? (blks) : 0; \
3445 } while (0)
3446
3447int f2fs_build_stats(struct f2fs_sb_info *sbi);
3448void f2fs_destroy_stats(struct f2fs_sb_info *sbi);
3449void __init f2fs_create_root_stats(void);
3450void f2fs_destroy_root_stats(void);
3451#else
3452#define stat_inc_cp_count(si) do { } while (0)
3453#define stat_inc_bg_cp_count(si) do { } while (0)
3454#define stat_inc_call_count(si) do { } while (0)
3455#define stat_inc_bggc_count(si) do { } while (0)
3456#define stat_io_skip_bggc_count(sbi) do { } while (0)
3457#define stat_other_skip_bggc_count(sbi) do { } while (0)
3458#define stat_inc_dirty_inode(sbi, type) do { } while (0)
3459#define stat_dec_dirty_inode(sbi, type) do { } while (0)
3460#define stat_inc_total_hit(sb) do { } while (0)
3461#define stat_inc_rbtree_node_hit(sb) do { } while (0)
3462#define stat_inc_largest_node_hit(sbi) do { } while (0)
3463#define stat_inc_cached_node_hit(sbi) do { } while (0)
3464#define stat_inc_inline_xattr(inode) do { } while (0)
3465#define stat_dec_inline_xattr(inode) do { } while (0)
3466#define stat_inc_inline_inode(inode) do { } while (0)
3467#define stat_dec_inline_inode(inode) do { } while (0)
3468#define stat_inc_inline_dir(inode) do { } while (0)
3469#define stat_dec_inline_dir(inode) do { } while (0)
3470#define stat_inc_atomic_write(inode) do { } while (0)
3471#define stat_dec_atomic_write(inode) do { } while (0)
3472#define stat_update_max_atomic_write(inode) do { } while (0)
3473#define stat_inc_volatile_write(inode) do { } while (0)
3474#define stat_dec_volatile_write(inode) do { } while (0)
3475#define stat_update_max_volatile_write(inode) do { } while (0)
3476#define stat_inc_meta_count(sbi, blkaddr) do { } while (0)
3477#define stat_inc_seg_type(sbi, curseg) do { } while (0)
3478#define stat_inc_block_count(sbi, curseg) do { } while (0)
3479#define stat_inc_inplace_blocks(sbi) do { } while (0)
3480#define stat_inc_seg_count(sbi, type, gc_type) do { } while (0)
3481#define stat_inc_tot_blk_count(si, blks) do { } while (0)
3482#define stat_inc_data_blk_count(sbi, blks, gc_type) do { } while (0)
3483#define stat_inc_node_blk_count(sbi, blks, gc_type) do { } while (0)
3484
3485static inline int f2fs_build_stats(struct f2fs_sb_info *sbi) { return 0; }
3486static inline void f2fs_destroy_stats(struct f2fs_sb_info *sbi) { }
3487static inline void __init f2fs_create_root_stats(void) { }
3488static inline void f2fs_destroy_root_stats(void) { }
3489#endif
3490
3491extern const struct file_operations f2fs_dir_operations;
3492#ifdef CONFIG_UNICODE
3493extern const struct dentry_operations f2fs_dentry_ops;
3494#endif
3495extern const struct file_operations f2fs_file_operations;
3496extern const struct inode_operations f2fs_file_inode_operations;
3497extern const struct address_space_operations f2fs_dblock_aops;
3498extern const struct address_space_operations f2fs_node_aops;
3499extern const struct address_space_operations f2fs_meta_aops;
3500extern const struct inode_operations f2fs_dir_inode_operations;
3501extern const struct inode_operations f2fs_symlink_inode_operations;
3502extern const struct inode_operations f2fs_encrypted_symlink_inode_operations;
3503extern const struct inode_operations f2fs_special_inode_operations;
3504extern struct kmem_cache *f2fs_inode_entry_slab;
3505
3506/*
3507 * inline.c
3508 */
3509bool f2fs_may_inline_data(struct inode *inode);
3510bool f2fs_may_inline_dentry(struct inode *inode);
3511void f2fs_do_read_inline_data(struct page *page, struct page *ipage);
3512void f2fs_truncate_inline_inode(struct inode *inode,
3513 struct page *ipage, u64 from);
3514int f2fs_read_inline_data(struct inode *inode, struct page *page);
3515int f2fs_convert_inline_page(struct dnode_of_data *dn, struct page *page);
3516int f2fs_convert_inline_inode(struct inode *inode);
3517int f2fs_write_inline_data(struct inode *inode, struct page *page);
3518bool f2fs_recover_inline_data(struct inode *inode, struct page *npage);
3519struct f2fs_dir_entry *f2fs_find_in_inline_dir(struct inode *dir,
3520 struct fscrypt_name *fname, struct page **res_page);
3521int f2fs_make_empty_inline_dir(struct inode *inode, struct inode *parent,
3522 struct page *ipage);
3523int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name,
3524 const struct qstr *orig_name,
3525 struct inode *inode, nid_t ino, umode_t mode);
3526void f2fs_delete_inline_entry(struct f2fs_dir_entry *dentry,
3527 struct page *page, struct inode *dir,
3528 struct inode *inode);
3529bool f2fs_empty_inline_dir(struct inode *dir);
3530int f2fs_read_inline_dir(struct file *file, struct dir_context *ctx,
3531 struct fscrypt_str *fstr);
3532int f2fs_inline_data_fiemap(struct inode *inode,
3533 struct fiemap_extent_info *fieinfo,
3534 __u64 start, __u64 len);
3535
3536/*
3537 * shrinker.c
3538 */
3539unsigned long f2fs_shrink_count(struct shrinker *shrink,
3540 struct shrink_control *sc);
3541unsigned long f2fs_shrink_scan(struct shrinker *shrink,
3542 struct shrink_control *sc);
3543void f2fs_join_shrinker(struct f2fs_sb_info *sbi);
3544void f2fs_leave_shrinker(struct f2fs_sb_info *sbi);
3545
3546/*
3547 * extent_cache.c
3548 */
3549struct rb_entry *f2fs_lookup_rb_tree(struct rb_root_cached *root,
3550 struct rb_entry *cached_re, unsigned int ofs);
3551struct rb_node **f2fs_lookup_rb_tree_for_insert(struct f2fs_sb_info *sbi,
3552 struct rb_root_cached *root,
3553 struct rb_node **parent,
3554 unsigned int ofs, bool *leftmost);
3555struct rb_entry *f2fs_lookup_rb_tree_ret(struct rb_root_cached *root,
3556 struct rb_entry *cached_re, unsigned int ofs,
3557 struct rb_entry **prev_entry, struct rb_entry **next_entry,
3558 struct rb_node ***insert_p, struct rb_node **insert_parent,
3559 bool force, bool *leftmost);
3560bool f2fs_check_rb_tree_consistence(struct f2fs_sb_info *sbi,
3561 struct rb_root_cached *root);
3562unsigned int f2fs_shrink_extent_tree(struct f2fs_sb_info *sbi, int nr_shrink);
3563bool f2fs_init_extent_tree(struct inode *inode, struct f2fs_extent *i_ext);
3564void f2fs_drop_extent_tree(struct inode *inode);
3565unsigned int f2fs_destroy_extent_node(struct inode *inode);
3566void f2fs_destroy_extent_tree(struct inode *inode);
3567bool f2fs_lookup_extent_cache(struct inode *inode, pgoff_t pgofs,
3568 struct extent_info *ei);
3569void f2fs_update_extent_cache(struct dnode_of_data *dn);
3570void f2fs_update_extent_cache_range(struct dnode_of_data *dn,
3571 pgoff_t fofs, block_t blkaddr, unsigned int len);
3572void f2fs_init_extent_cache_info(struct f2fs_sb_info *sbi);
3573int __init f2fs_create_extent_cache(void);
3574void f2fs_destroy_extent_cache(void);
3575
3576/*
3577 * sysfs.c
3578 */
3579int __init f2fs_init_sysfs(void);
3580void f2fs_exit_sysfs(void);
3581int f2fs_register_sysfs(struct f2fs_sb_info *sbi);
3582void f2fs_unregister_sysfs(struct f2fs_sb_info *sbi);
3583
3584/* verity.c */
3585extern const struct fsverity_operations f2fs_verityops;
3586
3587/*
3588 * crypto support
3589 */
3590static inline bool f2fs_encrypted_file(struct inode *inode)
3591{
3592 return IS_ENCRYPTED(inode) && S_ISREG(inode->i_mode);
3593}
3594
3595static inline void f2fs_set_encrypted_inode(struct inode *inode)
3596{
3597#ifdef CONFIG_FS_ENCRYPTION
3598 file_set_encrypt(inode);
3599 f2fs_set_inode_flags(inode);
3600#endif
3601}
3602
3603/*
3604 * Returns true if the reads of the inode's data need to undergo some
3605 * postprocessing step, like decryption or authenticity verification.
3606 */
3607static inline bool f2fs_post_read_required(struct inode *inode)
3608{
3609 return f2fs_encrypted_file(inode) || fsverity_active(inode);
3610}
3611
3612#define F2FS_FEATURE_FUNCS(name, flagname) \
3613static inline int f2fs_sb_has_##name(struct f2fs_sb_info *sbi) \
3614{ \
3615 return F2FS_HAS_FEATURE(sbi, F2FS_FEATURE_##flagname); \
3616}
3617
3618F2FS_FEATURE_FUNCS(encrypt, ENCRYPT);
3619F2FS_FEATURE_FUNCS(blkzoned, BLKZONED);
3620F2FS_FEATURE_FUNCS(extra_attr, EXTRA_ATTR);
3621F2FS_FEATURE_FUNCS(project_quota, PRJQUOTA);
3622F2FS_FEATURE_FUNCS(inode_chksum, INODE_CHKSUM);
3623F2FS_FEATURE_FUNCS(flexible_inline_xattr, FLEXIBLE_INLINE_XATTR);
3624F2FS_FEATURE_FUNCS(quota_ino, QUOTA_INO);
3625F2FS_FEATURE_FUNCS(inode_crtime, INODE_CRTIME);
3626F2FS_FEATURE_FUNCS(lost_found, LOST_FOUND);
3627F2FS_FEATURE_FUNCS(verity, VERITY);
3628F2FS_FEATURE_FUNCS(sb_chksum, SB_CHKSUM);
3629F2FS_FEATURE_FUNCS(casefold, CASEFOLD);
3630
3631#ifdef CONFIG_BLK_DEV_ZONED
3632static inline bool f2fs_blkz_is_seq(struct f2fs_sb_info *sbi, int devi,
3633 block_t blkaddr)
3634{
3635 unsigned int zno = blkaddr >> sbi->log_blocks_per_blkz;
3636
3637 return test_bit(zno, FDEV(devi).blkz_seq);
3638}
3639#endif
3640
3641static inline bool f2fs_hw_should_discard(struct f2fs_sb_info *sbi)
3642{
3643 return f2fs_sb_has_blkzoned(sbi);
3644}
3645
3646static inline bool f2fs_bdev_support_discard(struct block_device *bdev)
3647{
3648 return blk_queue_discard(bdev_get_queue(bdev)) ||
3649 bdev_is_zoned(bdev);
3650}
3651
3652static inline bool f2fs_hw_support_discard(struct f2fs_sb_info *sbi)
3653{
3654 int i;
3655
3656 if (!f2fs_is_multi_device(sbi))
3657 return f2fs_bdev_support_discard(sbi->sb->s_bdev);
3658
3659 for (i = 0; i < sbi->s_ndevs; i++)
3660 if (f2fs_bdev_support_discard(FDEV(i).bdev))
3661 return true;
3662 return false;
3663}
3664
3665static inline bool f2fs_realtime_discard_enable(struct f2fs_sb_info *sbi)
3666{
3667 return (test_opt(sbi, DISCARD) && f2fs_hw_support_discard(sbi)) ||
3668 f2fs_hw_should_discard(sbi);
3669}
3670
3671static inline bool f2fs_hw_is_readonly(struct f2fs_sb_info *sbi)
3672{
3673 int i;
3674
3675 if (!f2fs_is_multi_device(sbi))
3676 return bdev_read_only(sbi->sb->s_bdev);
3677
3678 for (i = 0; i < sbi->s_ndevs; i++)
3679 if (bdev_read_only(FDEV(i).bdev))
3680 return true;
3681 return false;
3682}
3683
3684
3685static inline void set_opt_mode(struct f2fs_sb_info *sbi, unsigned int mt)
3686{
3687 clear_opt(sbi, ADAPTIVE);
3688 clear_opt(sbi, LFS);
3689
3690 switch (mt) {
3691 case F2FS_MOUNT_ADAPTIVE:
3692 set_opt(sbi, ADAPTIVE);
3693 break;
3694 case F2FS_MOUNT_LFS:
3695 set_opt(sbi, LFS);
3696 break;
3697 }
3698}
3699
3700static inline bool f2fs_may_encrypt(struct inode *inode)
3701{
3702#ifdef CONFIG_FS_ENCRYPTION
3703 umode_t mode = inode->i_mode;
3704
3705 return (S_ISREG(mode) || S_ISDIR(mode) || S_ISLNK(mode));
3706#else
3707 return false;
3708#endif
3709}
3710
3711static inline int block_unaligned_IO(struct inode *inode,
3712 struct kiocb *iocb, struct iov_iter *iter)
3713{
3714 unsigned int i_blkbits = READ_ONCE(inode->i_blkbits);
3715 unsigned int blocksize_mask = (1 << i_blkbits) - 1;
3716 loff_t offset = iocb->ki_pos;
3717 unsigned long align = offset | iov_iter_alignment(iter);
3718
3719 return align & blocksize_mask;
3720}
3721
3722static inline int allow_outplace_dio(struct inode *inode,
3723 struct kiocb *iocb, struct iov_iter *iter)
3724{
3725 struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
3726 int rw = iov_iter_rw(iter);
3727
3728 return (test_opt(sbi, LFS) && (rw == WRITE) &&
3729 !block_unaligned_IO(inode, iocb, iter));
3730}
3731
3732static inline bool f2fs_force_buffered_io(struct inode *inode,
3733 struct kiocb *iocb, struct iov_iter *iter)
3734{
3735 struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
3736 int rw = iov_iter_rw(iter);
3737
3738 if (f2fs_post_read_required(inode))
3739 return true;
3740 if (f2fs_is_multi_device(sbi))
3741 return true;
3742 /*
3743 * for blkzoned device, fallback direct IO to buffered IO, so
3744 * all IOs can be serialized by log-structured write.
3745 */
3746 if (f2fs_sb_has_blkzoned(sbi))
3747 return true;
3748 if (test_opt(sbi, LFS) && (rw == WRITE)) {
3749 if (block_unaligned_IO(inode, iocb, iter))
3750 return true;
3751 if (F2FS_IO_ALIGNED(sbi))
3752 return true;
3753 }
3754 if (is_sbi_flag_set(F2FS_I_SB(inode), SBI_CP_DISABLED) &&
3755 !IS_SWAPFILE(inode))
3756 return true;
3757
3758 return false;
3759}
3760
3761#ifdef CONFIG_F2FS_FAULT_INJECTION
3762extern void f2fs_build_fault_attr(struct f2fs_sb_info *sbi, unsigned int rate,
3763 unsigned int type);
3764#else
3765#define f2fs_build_fault_attr(sbi, rate, type) do { } while (0)
3766#endif
3767
3768static inline bool is_journalled_quota(struct f2fs_sb_info *sbi)
3769{
3770#ifdef CONFIG_QUOTA
3771 if (f2fs_sb_has_quota_ino(sbi))
3772 return true;
3773 if (F2FS_OPTION(sbi).s_qf_names[USRQUOTA] ||
3774 F2FS_OPTION(sbi).s_qf_names[GRPQUOTA] ||
3775 F2FS_OPTION(sbi).s_qf_names[PRJQUOTA])
3776 return true;
3777#endif
3778 return false;
3779}
3780
3781#define EFSBADCRC EBADMSG /* Bad CRC detected */
3782#define EFSCORRUPTED EUCLEAN /* Filesystem is corrupted */
3783
3784#endif /* _LINUX_F2FS_H */