1 .. SPDX-License-Identifier: GPL-2.0
2
3 ==========================================
4 WHAT IS Flash-Friendly File System (F2FS)?
5 ==========================================
6
7 NAND flash memory-based storage devices, such
8 been equipped on a variety systems ranging fro
9 they are known to have different characteristi
10 disks, a file system, an upper layer to the st
11 changes from the sketch in the design level.
12
13 F2FS is a file system exploiting NAND flash me
14 is based on Log-structured File System (LFS).
15 addressing the fundamental issues in LFS, whic
16 tree and high cleaning overhead.
17
18 Since a NAND flash memory-based storage device
19 according to its internal geometry or flash me
20 F2FS and its tools support various parameters
21 layout, but also for selecting allocation and
22
23 The following git tree provides the file syste
24 a consistency checking tool (fsck.f2fs), and a
25
26 - git://git.kernel.org/pub/scm/linux/kernel/gi
27
28 For sending patches, please use the following
29
30 - linux-f2fs-devel@lists.sourceforge.net
31
32 For reporting bugs, please use the following f
33
34 - https://bugzilla.kernel.org/enter_bug.cgi?pr
35
36 Background and Design issues
37 ============================
38
39 Log-structured File System (LFS)
40 --------------------------------
41 "A log-structured file system writes all modif
42 a log-like structure, thereby speeding up bot
43 The log is the only structure on disk; it cont
44 files can be read back from the log efficientl
45 areas on disk for fast writing, we divide the
46 segment cleaner to compress the live informati
47 segments." from Rosenblum, M. and Ousterhout,
48 implementation of a log-structured file system
49 10, 1, 26–52.
50
51 Wandering Tree Problem
52 ----------------------
53 In LFS, when a file data is updated and writte
54 pointer block is updated due to the changed lo
55 block is also updated due to the direct pointe
56 the upper index structures such as inode, inod
57 also updated recursively. This problem is call
58 and in order to enhance the performance, it sh
59 propagation as much as possible.
60
61 [1] Bityutskiy, A. 2005. JFFS3 design issues.
62
63 Cleaning Overhead
64 -----------------
65 Since LFS is based on out-of-place writes, it
66 scattered across the whole storage. In order t
67 needs to reclaim these obsolete blocks seamles
68 as a cleaning process.
69
70 The process consists of three operations as fo
71
72 1. A victim segment is selected through refere
73 2. It loads parent index structures of all the
74 segment summary blocks.
75 3. It checks the cross-reference between the d
76 4. It moves valid data selectively.
77
78 This cleaning job may cause unexpected long de
79 is to hide the latencies to users. And also de
80 amount of valid data to be moved, and move the
81
82 Key Features
83 ============
84
85 Flash Awareness
86 ---------------
87 - Enlarge the random write area for better per
88 spatial locality
89 - Align FS data structures to the operational
90
91 Wandering Tree Problem
92 ----------------------
93 - Use a term, “node”, that represents inod
94 - Introduce Node Address Table (NAT) containin
95 blocks; this will cut off the update propaga
96
97 Cleaning Overhead
98 -----------------
99 - Support a background cleaning process
100 - Support greedy and cost-benefit algorithms f
101 - Support multi-head logs for static/dynamic h
102 - Introduce adaptive logging for efficient blo
103
104 Mount Options
105 =============
106
107
108 ======================== =====================
109 background_gc=%s Turn on/off cleaning
110 collection, triggered
111 idle. If background_g
112 collection and if bac
113 will be turned off. I
114 on synchronous garbag
115 Default value for thi
116 collection is on by d
117 gc_merge When background_gc is
118 let background GC thr
119 it can eliminate the
120 GC operation when GC
121 I/O and CPU resources
122 nogc_merge Disable GC merge feat
123 disable_roll_forward Disable the roll-forw
124 norecovery Disable the roll-forw
125 only (i.e., -o ro,dis
126 discard/nodiscard Enable/disable real-t
127 enabled, f2fs will is
128 segment is cleaned.
129 heap/no_heap Deprecated.
130 nouser_xattr Disable Extended User
131 by default if CONFIG_
132 noacl Disable POSIX Access
133 by default if CONFIG_
134 active_logs=%u Support configuring t
135 current design, f2fs
136 Default number is 6.
137 disable_ext_identify Disable the extension
138 is not aware of cold
139 inline_xattr Enable the inline xat
140 noinline_xattr Disable the inline xa
141 inline_xattr_size=%u Support configuring i
142 flexible inline xattr
143 inline_data Enable the inline dat
144 files can be written
145 inline_dentry Enable the inline dir
146 directory entries can
147 space of inode block
148 dentries is limited t
149 noinline_dentry Disable the inline de
150 flush_merge Merge concurrent cach
151 to eliminate redundan
152 device handles the ca
153 recommend to enable t
154 nobarrier This option can be us
155 its cached data shoul
156 If this option is set
157 but f2fs still guaran
158 data writes.
159 barrier If this option is set
160 issued.
161 fastboot This option is used w
162 time as much as possi
163 can be sacrificed.
164 extent_cache Enable an extent cach
165 as many as extent whi
166 address and physical
167 increasing the cache
168 noextent_cache Disable an extent cac
169 the above extent_cach
170 noinline_data Disable the inline da
171 enabled by default.
172 data_flush Enable data flushing
173 persist data of regul
174 reserve_root=%d Support configuring r
175 allocation from a pri
176 gid, unit: 4KB, the d
177 resuid=%d The user ID which may
178 resgid=%d The group ID which ma
179 fault_injection=%d Enable fault injectio
180 specified injection r
181 fault_type=%d Support configuring f
182 enabled with fault_in
183 is shown below, it su
184
185 =====================
186 Type_Name
187 =====================
188 FAULT_KMALLOC
189 FAULT_KVMALLOC
190 FAULT_PAGE_ALLOC
191 FAULT_PAGE_GET
192 FAULT_ALLOC_BIO
193 FAULT_ALLOC_NID
194 FAULT_ORPHAN
195 FAULT_BLOCK
196 FAULT_DIR_DEPTH
197 FAULT_EVICT_INODE
198 FAULT_TRUNCATE
199 FAULT_READ_IO
200 FAULT_CHECKPOINT
201 FAULT_DISCARD
202 FAULT_WRITE_IO
203 FAULT_SLAB_ALLOC
204 FAULT_DQUOT_INIT
205 FAULT_LOCK_OP
206 FAULT_BLKADDR_VALIDIT
207 FAULT_BLKADDR_CONSIST
208 FAULT_NO_SEGMENT
209 =====================
210 mode=%s Control block allocat
211 and "lfs". In "lfs" m
212 writes towards main a
213 "fragment:segment" an
214 These are developer o
215 fragmentation/after-G
216 modes to understand f
217 and eventually get so
218 In "fragment:segment"
219 position. With this,
220 In "fragment:block",
221 "max_fragment_chunk"
222 We added some randomn
223 it close to realistic
224 1..<max_fragment_chun
225 length of 1..<max_fra
226 allocated blocks will
227 Note that "fragment:b
228 option for more rando
229 Please, use these opt
230 recommend to re-forma
231 usrquota Enable plain user dis
232 grpquota Enable plain group di
233 prjquota Enable plain project
234 usrjquota=<file> Appoint specified fil
235 grpjquota=<file> information can be pr
236 prjjquota=<file> <quota file>: must be
237 jqfmt=<quota type> <quota type>: [vfsold
238 offusrjquota Turn off user journal
239 offgrpjquota Turn off group journa
240 offprjjquota Turn off project jour
241 quota Enable plain user dis
242 noquota Disable all plain dis
243 alloc_mode=%s Adjust block allocati
244 and "default".
245 fsync_mode=%s Control the policy of
246 "strict", and "nobarr
247 default, fsync will f
248 light operation to im
249 In "strict" mode, fsy
250 with xfs, ext4 and bt
251 pass, but the perform
252 based on "posix", but
253 non-atomic files like
254 test_dummy_encryption
255 test_dummy_encryption=%s
256 Enable dummy encrypti
257 context. The fake fsc
258 The argument may be e
259 select the correspond
260 checkpoint=%s[:%u[%]] Set to "disable" to t
261 to reenable checkpoin
262 disabled, any unmount
263 the filesystem conten
264 filesystem was mounte
265 While mounting with c
266 run garbage collectio
267 be used. If this take
268 EAGAIN. You may optio
269 of the disk you would
270 avoid additional garb
271 number of blocks, or
272 with checkpoint=disab
273 hide up to all remain
274 would be unusable can
275 This space is reclaim
276 checkpoint_merge When checkpoint is en
277 daemon and make it to
278 much as possible to e
279 we can eliminate the
280 operation when the ch
281 a cgroup having low i
282 do better, we set the
283 to "3", to give one h
284 This is the same way
285 journaling thread of
286 nocheckpoint_merge Disable checkpoint me
287 compress_algorithm=%s Control compress algo
288 "lz4", "zstd" and "lz
289 compress_algorithm=%s:%d Control compress algo
290 "lz4" and "zstd" supp
291 algorithm level
292 lz4 3 - 16
293 zstd 1 - 22
294 compress_log_size=%u Support configuring c
295 be 4KB * (1 << %u). T
296 compress_extension=%s Support adding specif
297 compression on those
298 with '.ext' has high
299 on compression extens
300 these file by default
301 For other files, we c
302 Note that, there is o
303 can be set to enable
304 nocompress_extension=%s Support adding specif
305 compression on those
306 If you know exactly w
307 The same extension na
308 extension at the same
309 If the compress exten
310 nocompress extension
311 Don't allow use '*' t
312 After add nocompress_
313 dir_flag < comp_exten
314 See more in compressi
315
316 compress_chksum Support verifying chk
317 compress_mode=%s Control file compress
318 modes. In "fs" mode (
319 on the compression en
320 the automaic compress
321 choosing the target f
322 compression/decompres
323 ioctls.
324 compress_cache Support to use addres
325 cache compressed bloc
326 random read.
327 inlinecrypt When possible, encryp
328 files using the blk-c
329 filesystem-layer encr
330 inline encryption har
331 unaffected. For more
332 Documentation/block/i
333 atgc Enable age-threshold
334 effectiveness and eff
335 discard_unit=%s Control discard unit,
336 and "section", issued
337 aligned to the unit,
338 so that small discard
339 For blkzoned device,
340 default, it is helpfu
341 reduce memory cost by
342 discard.
343 memory=%s Control memory mode.
344 "low" mode is introdu
345 Because of the nature
346 will try to save memo
347 "normal" mode is the
348 age_extent_cache Enable an age extent
349 data block update fre
350 order to provide bett
351 allocation.
352 errors=%s Specify f2fs behavior
353 "panic", "continue" a
354 panic immediately, co
355 the partition in read
356 mode.
357 =====================
358 mode
359 =====================
360 access ops
361 syscall errors
362 mount option
363 pending dir write
364 pending non-dir write
365 pending node write
366 pending meta write
367 =====================
368 ======================== =====================
369
370 Debugfs Entries
371 ===============
372
373 /sys/kernel/debug/f2fs/ contains information a
374 f2fs. Each file shows the whole f2fs informati
375
376 /sys/kernel/debug/f2fs/status includes:
377
378 - major file system information managed by f2
379 - average SIT information about whole segment
380 - current memory footprint consumed by f2fs.
381
382 Sysfs Entries
383 =============
384
385 Information about mounted f2fs file systems ca
386 /sys/fs/f2fs. Each mounted filesystem will ha
387 /sys/fs/f2fs based on its device name (i.e., /
388 The files in each per-device directory are sho
389
390 Files in /sys/fs/f2fs/<devname>
391 (see also Documentation/ABI/testing/sysfs-fs-f
392
393 Usage
394 =====
395
396 1. Download userland tools and compile them.
397
398 2. Skip, if f2fs was compiled statically insid
399 Otherwise, insert the f2fs.ko module::
400
401 # insmod f2fs.ko
402
403 3. Create a directory to use when mounting::
404
405 # mkdir /mnt/f2fs
406
407 4. Format the block device, and then mount as
408
409 # mkfs.f2fs -l label /dev/block_device
410 # mount -t f2fs /dev/block_device /mnt
411
412 mkfs.f2fs
413 ---------
414 The mkfs.f2fs is for the use of formatting a p
415 which builds a basic on-disk layout.
416
417 The quick options consist of:
418
419 =============== ===========================
420 ``-l [label]`` Give a volume label, up to
421 ``-a [0 or 1]`` Split start location of eac
422
423 1 is set by default, which
424 ``-o [int]`` Set overprovision ratio in
425
426 5 is set by default.
427 ``-s [int]`` Set the number of segments
428
429 1 is set by default.
430 ``-z [int]`` Set the number of sections
431
432 1 is set by default.
433 ``-e [str]`` Set basic extension list. e
434 ``-t [0 or 1]`` Disable discard command or
435
436 1 is set by default, which
437 =============== ===========================
438
439 Note: please refer to the manpage of mkfs.f2fs
440
441 fsck.f2fs
442 ---------
443 The fsck.f2fs is a tool to check the consisten
444 partition, which examines whether the filesyst
445 are cross-referenced correctly or not.
446 Note that, initial version of the tool does no
447
448 The quick options consist of::
449
450 -d debug level [default:0]
451
452 Note: please refer to the manpage of fsck.f2fs
453
454 dump.f2fs
455 ---------
456 The dump.f2fs shows the information of specifi
457 file. Each file is dump_ssa and dump_sit.
458
459 The dump.f2fs is used to debug on-disk data st
460 It shows on-disk inode information recognized
461 able to dump all the SSA and SIT entries into
462 ./dump_sit respectively.
463
464 The options consist of::
465
466 -d debug level [default:0]
467 -i inode no (hex)
468 -s [SIT dump segno from #1~#2 (decimal), for
469 -a [SSA dump segno from #1~#2 (decimal), for
470
471 Examples::
472
473 # dump.f2fs -i [ino] /dev/sdx
474 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump)
475 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump)
476
477 Note: please refer to the manpage of dump.f2fs
478
479 sload.f2fs
480 ----------
481 The sload.f2fs gives a way to insert files and
482 image. This tool is useful when building f2fs
483
484 Note: please refer to the manpage of sload.f2f
485
486 resize.f2fs
487 -----------
488 The resize.f2fs lets a user resize the f2fs-fo
489 all the files and directories stored in the im
490
491 Note: please refer to the manpage of resize.f2
492
493 defrag.f2fs
494 -----------
495 The defrag.f2fs can be used to defragment scat
496 filesystem metadata across the disk. This can
497 more free consecutive space.
498
499 Note: please refer to the manpage of defrag.f2
500
501 f2fs_io
502 -------
503 The f2fs_io is a simple tool to issue various
504 f2fs-specific ones, which is very useful for Q
505
506 Note: please refer to the manpage of f2fs_io(8
507
508 Design
509 ======
510
511 On-disk Layout
512 --------------
513
514 F2FS divides the whole volume into a number of
515 to 2MB in size. A section is composed of conse
516 consists of a set of sections. By default, sec
517 segment size identically, but users can easily
518
519 F2FS splits the entire volume into six areas,
520 consist of multiple segments as described belo
521
522 al
523 |-> align with the segment si
524 _________________________________________
525 | | | Segment |
526 | Superblock | Checkpoint | Info. |
527 | (SB) | (CP) | Table (SIT) |
528 |____________|_____2______|______N______|_
529
530
531
532 ._________
533 |_Segment_
534 .
535 ._________
536 |_section_
537 .
538 .________.
539 |__zone__|
540
541 - Superblock (SB)
542 It is located at the beginning of the parti
543 to avoid file system crash. It contains bas
544 default parameters of f2fs.
545
546 - Checkpoint (CP)
547 It contains file system information, bitmap
548 inode lists, and summary entries of current
549
550 - Segment Information Table (SIT)
551 It contains segment information such as val
552 validity of all the blocks.
553
554 - Node Address Table (NAT)
555 It is composed of a block address table for
556 Main area.
557
558 - Segment Summary Area (SSA)
559 It contains summary entries which contains
560 data and node blocks stored in Main area.
561
562 - Main Area
563 It contains file and directory data includi
564
565 In order to avoid misalignment between file sy
566 aligns the start block address of CP with the
567 start block address of Main area with the zone
568 in SSA area.
569
570 Reference the following survey for additional
571 https://wiki.linaro.org/WorkingGroups/Kernel/P
572
573 File System Metadata Structure
574 ------------------------------
575
576 F2FS adopts the checkpointing scheme to mainta
577 mount time, F2FS first tries to find the last
578 CP area. In order to reduce the scanning time,
579 One of them always indicates the last valid da
580 mechanism. In addition to CP, NAT and SIT also
581
582 For file system consistency, each CP points to
583 valid, as shown as below::
584
585 +--------+----------+---------+
586 | CP | SIT | NAT |
587 +--------+----------+---------+
588 . . . .
589 . . . .
590 . . .
591 +-------+-------+--------+--------+--------+
592 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 |
593 +-------+-------+--------+--------+--------+
594 | ^
595 | |
596 `----------------------------------------
597
598 Index Structure
599 ---------------
600
601 The key data structure to manage the data loca
602 traditional file structures, F2FS has three ty
603 indirect node. F2FS assigns 4KB to an inode bl
604 indices, two direct node pointers, two indirec
605 indirect node pointer as described below. One
606 data blocks, and one indirect node block conta
607 one inode block (i.e., a file) covers::
608
609 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 10
610
611 Inode block (4KB)
612 |- data (923)
613 |- direct node (2)
614 | `- data (1018)
615 |- indirect node (2)
616 | `- direct node (1018)
617 | `- data (1018)
618 `- double indirect node (1)
619 `- indirect node (101
620 `- direc
621
622
623 Note that all the node blocks are mapped by NA
624 each node is translated by the NAT table. In t
625 tree problem, F2FS is able to cut off the prop
626 leaf data writes.
627
628 Directory Structure
629 -------------------
630
631 A directory entry occupies 11 bytes, which con
632
633 - hash hash value of the file name
634 - ino inode number
635 - len the length of file name
636 - type file type such as directory, s
637
638 A dentry block consists of 214 dentry slots an
639 used to represent whether each dentry is valid
640 4KB with the following composition.
641
642 ::
643
644 Dentry Block(4 K) = bitmap (27 bytes) + rese
645 dentries(11 * 214 bytes)
646
647 [Bucket]
648 +--------------------------------
649 |dentry block 1 | dentry block 2
650 +--------------------------------
651 . .
652 . .
653 . [Dentry Block Structure: 4KB]
654 +--------+----------+----------+------------
655 | bitmap | reserved | dentries | file names
656 +--------+----------+----------+------------
657 [Dentry Block: 4KB] . .
658 . .
659 . .
660 +------+------+-----+------+
661 | hash | ino | len | type |
662 +------+------+-----+------+
663 [Dentry Structure: 11 bytes]
664
665 F2FS implements multi-level hash tables for di
666 a hash table with dedicated number of hash buc
667 "A(2B)" means a bucket includes 2 data blocks.
668
669 ::
670
671 ----------------------
672 A : bucket
673 B : block
674 N : MAX_DIR_HASH_DEPTH
675 ----------------------
676
677 level #0 | A(2B)
678 |
679 level #1 | A(2B) - A(2B)
680 |
681 level #2 | A(2B) - A(2B) - A(2B) - A(2B)
682 . | . . . .
683 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B)
684 . | . . . .
685 level #N | A(4B) - A(4B) - A(4B) - A(4B)
686
687 The number of blocks and buckets are determine
688
689 ,- 2, if n < MAX_D
690 # of blocks in level #n = |
691 `- 4, Otherwise
692
693 ,- 2^(n + dir_lev
694 | if n + d
695 # of buckets in level #n = |
696 `- 2^((MAX_DIR_HA
697 Otherwis
698
699 When F2FS finds a file name in a directory, at
700 name is calculated. Then, F2FS scans the hash
701 dentry consisting of the file name and its ino
702 scans the next hash table in level #1. In this
703 each levels incrementally from 1 to N. In each
704 one bucket determined by the following equatio
705 complexity::
706
707 bucket number to scan in level #n = (hash va
708
709 In the case of file creation, F2FS finds empty
710 file name. F2FS searches the empty slots in th
711 1 to N in the same way as the lookup operation
712
713 The following figure shows an example of two c
714
715 --------------> Dir <--------------
716 | |
717 child child
718
719 child - child [hole] -
720
721 child - child - child [hole] -
722
723 Case 1: Case 2:
724 Number of children = 6, Number of
725 File size = 7 File size
726
727 Default Block Allocation
728 ------------------------
729
730 At runtime, F2FS manages six active logs insid
731 and Hot/Warm/Cold data.
732
733 - Hot node contains direct node blocks of
734 - Warm node contains direct node blocks ex
735 - Cold node contains indirect node blocks
736 - Hot data contains dentry blocks
737 - Warm data contains data blocks except ho
738 - Cold data contains multimedia data or mi
739
740 LFS has two schemes for free space management:
741 tion. The copy-and-compaction scheme which is
742 for devices showing very good sequential write
743 are served all the time for writing new data.
744 overhead under high utilization. Contrarily, t
745 from random writes, but no cleaning process is
746 scheme where the copy-and-compaction scheme is
747 policy is dynamically changed to the threaded
748 system status.
749
750 In order to align F2FS with underlying flash-b
751 segment in a unit of section. F2FS expects tha
752 same as the unit size of garbage collection in
753 to the mapping granularity in FTL, F2FS alloca
754 logs from different zones as much as possible,
755 the active logs into one allocation unit accor
756
757 Cleaning process
758 ----------------
759
760 F2FS does cleaning both on demand and in the b
761 triggered when there are not enough free segme
762 cleaner is operated by a kernel thread, and tr
763 system is idle.
764
765 F2FS supports two victim selection policies: g
766 In the greedy algorithm, F2FS selects a victim
767 of valid blocks. In the cost-benefit algorithm
768 according to the segment age and the number of
769 log block thrashing problem in the greedy algo
770 algorithm for on-demand cleaner, while backgro
771 algorithm.
772
773 In order to identify whether the data in the v
774 F2FS manages a bitmap. Each bit represents the
775 bitmap is composed of a bit stream covering wh
776
777 Write-hint Policy
778 -----------------
779
780 F2FS sets the whint all the time with the belo
781
782 ===================== ========================
783 User F2FS
784 ===================== ========================
785 N/A META
786 N/A HOT_NODE
787 N/A WARM_NODE
788 N/A COLD_NODE
789 ioctl(COLD) COLD_DATA
790 extension list "
791
792 -- buffered io
793 N/A COLD_DATA
794 N/A HOT_DATA
795 N/A WARM_DATA
796
797 -- direct io
798 WRITE_LIFE_EXTREME COLD_DATA
799 WRITE_LIFE_SHORT HOT_DATA
800 WRITE_LIFE_NOT_SET WARM_DATA
801 WRITE_LIFE_NONE "
802 WRITE_LIFE_MEDIUM "
803 WRITE_LIFE_LONG "
804 ===================== ========================
805
806 Fallocate(2) Policy
807 -------------------
808
809 The default policy follows the below POSIX rul
810
811 Allocating disk space
812 The default operation (i.e., mode is zero)
813 the disk space within the range specified
814 file size (as reported by stat(2)) will be
815 greater than the file size. Any subregion
816 by offset and len that did not contain dat
817 initialized to zero. This default behavio
818 behavior of the posix_fallocate(3) library
819 as a method of optimally implementing that
820
821 However, once F2FS receives ioctl(fd, F2FS_IOC
822 fallocate(fd, DEFAULT_MODE), it allocates on-d
823 zero or random data, which is useful to the be
824
825 1. create(fd)
826 2. ioctl(fd, F2FS_IOC_SET_PIN_FILE)
827 3. fallocate(fd, 0, 0, size)
828 4. address = fibmap(fd, offset)
829 5. open(blkdev)
830 6. write(blkdev, address)
831
832 Compression implementation
833 --------------------------
834
835 - New term named cluster is defined as basic u
836 be divided into multiple clusters logically.
837 (n >= 0) logical pages, compression size is
838 cluster can be compressed or not.
839
840 - In cluster metadata layout, one special bloc
841 a cluster is a compressed one or normal one;
842 metadata maps cluster to [1, 4 << n - 1] phy
843 stores data including compress header and co
844
845 - In order to eliminate write amplification du
846 support compression on write-once file, data
847 all logical blocks in cluster contain valid
848 cluster data is lower than specified thresho
849
850 - To enable compression on regular inode, ther
851
852 * chattr +c file
853 * chattr +c dir; touch dir/file
854 * mount w/ -o compress_extension=ext; touch
855 * mount w/ -o compress_extension=*; touch an
856
857 - To disable compression on regular inode, the
858
859 * chattr -c file
860 * mount w/ -o nocompress_extension=ext; touc
861
862 - Priority in between FS_COMPR_FL, FS_NOCOMP_F
863
864 * compress_extension=so; nocompress_extensio
865 dir/foo.so; touch dir/bar.zip; touch dir/b
866 should be compresse, bar.zip should be non
867 can enable compress on bar.zip.
868 * compress_extension=so; nocompress_extensio
869 dir/foo.so; touch dir/bar.zip; touch dir/b
870 compresse, bar.zip and baz.txt should be n
871 chattr+c dir/bar.zip; chattr+c dir/baz.txt
872 and baz.txt.
873
874 - At this point, compression feature doesn't e
875 directly in order to guarantee potential dat
876 Instead, the main goal is to reduce data wri
877 possible, resulting in extending disk life t
878 congestion. Alternatively, we've added ioctl
879 interface to reclaim compressed space and sh
880 special flag to the inode. Once the compress
881 will block writing data to the file until ei
882 reserved via ioctl(F2FS_IOC_RESERVE_COMPRESS
883 truncated to zero.
884
885 Compress metadata layout::
886
887 [Dnode Structu
888 +-----------------------------
889 | cluster 1 | cluster 2 | ....
890 +-----------------------------
891 . .
892 . .
893 . Compressed Cluster .
894 +----------+---------+---------+---------+
895 |compr flag| block 1 | block 2 | block 3 |
896 +----------+---------+---------+---------+
897 . .
898 .
899 .
900 +-------------+-------------+---------
901 | data length | data chksum | reserved
902 +-------------+-------------+---------
903
904 Compression mode
905 --------------------------
906
907 f2fs supports "fs" and "user" compression mode
908 With this option, f2fs provides a choice to se
909 compression enabled files (refer to "Compressi
910 enable compression on a regular inode).
911
912 1) compress_mode=fs
913 This is the default option. f2fs does automati
914 compression enabled files.
915
916 2) compress_mode=user
917 This disables the automatic compression and gi
918 target file and the timing. The user can do ma
919 compression enabled files using F2FS_IOC_DECOM
920 ioctls like the below.
921
922 To decompress a file,
923
924 fd = open(filename, O_WRONLY, 0);
925 ret = ioctl(fd, F2FS_IOC_DECOMPRESS_FILE);
926
927 To compress a file,
928
929 fd = open(filename, O_WRONLY, 0);
930 ret = ioctl(fd, F2FS_IOC_COMPRESS_FILE);
931
932 NVMe Zoned Namespace devices
933 ----------------------------
934
935 - ZNS defines a per-zone capacity which can be
936 zone-size. Zone-capacity is the number of us
937 F2FS checks if zone-capacity is less than zo
938 segment which starts after the zone-capacity
939 the free segment bitmap at initial mount tim
940 as permanently used so they are not allocate
941 consequently are not needed to be garbage co
942 zone-capacity is not aligned to default segm
943 can start before the zone-capacity and span
944 Such spanning segments are also considered a
945 past the zone-capacity are considered unusab
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.