net/mlx5: Unmap only the relevant IO memory mapping
[deliverable/linux.git] / Documentation / filesystems / f2fs.txt
1 ================================================================================
2 WHAT IS Flash-Friendly File System (F2FS)?
3 ================================================================================
4
5 NAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have
6 been equipped on a variety systems ranging from mobile to server systems. Since
7 they are known to have different characteristics from the conventional rotating
8 disks, a file system, an upper layer to the storage device, should adapt to the
9 changes from the sketch in the design level.
10
11 F2FS is a file system exploiting NAND flash memory-based storage devices, which
12 is based on Log-structured File System (LFS). The design has been focused on
13 addressing the fundamental issues in LFS, which are snowball effect of wandering
14 tree and high cleaning overhead.
15
16 Since a NAND flash memory-based storage device shows different characteristic
17 according to its internal geometry or flash memory management scheme, namely FTL,
18 F2FS and its tools support various parameters not only for configuring on-disk
19 layout, but also for selecting allocation and cleaning algorithms.
20
21 The following git tree provides the file system formatting tool (mkfs.f2fs),
22 a consistency checking tool (fsck.f2fs), and a debugging tool (dump.f2fs).
23 >> git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git
24
25 For reporting bugs and sending patches, please use the following mailing list:
26 >> linux-f2fs-devel@lists.sourceforge.net
27
28 ================================================================================
29 BACKGROUND AND DESIGN ISSUES
30 ================================================================================
31
32 Log-structured File System (LFS)
33 --------------------------------
34 "A log-structured file system writes all modifications to disk sequentially in
35 a log-like structure, thereby speeding up both file writing and crash recovery.
36 The log is the only structure on disk; it contains indexing information so that
37 files can be read back from the log efficiently. In order to maintain large free
38 areas on disk for fast writing, we divide the log into segments and use a
39 segment cleaner to compress the live information from heavily fragmented
40 segments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and
41 implementation of a log-structured file system", ACM Trans. Computer Systems
42 10, 1, 26–52.
43
44 Wandering Tree Problem
45 ----------------------
46 In LFS, when a file data is updated and written to the end of log, its direct
47 pointer block is updated due to the changed location. Then the indirect pointer
48 block is also updated due to the direct pointer block update. In this manner,
49 the upper index structures such as inode, inode map, and checkpoint block are
50 also updated recursively. This problem is called as wandering tree problem [1],
51 and in order to enhance the performance, it should eliminate or relax the update
52 propagation as much as possible.
53
54 [1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/
55
56 Cleaning Overhead
57 -----------------
58 Since LFS is based on out-of-place writes, it produces so many obsolete blocks
59 scattered across the whole storage. In order to serve new empty log space, it
60 needs to reclaim these obsolete blocks seamlessly to users. This job is called
61 as a cleaning process.
62
63 The process consists of three operations as follows.
64 1. A victim segment is selected through referencing segment usage table.
65 2. It loads parent index structures of all the data in the victim identified by
66 segment summary blocks.
67 3. It checks the cross-reference between the data and its parent index structure.
68 4. It moves valid data selectively.
69
70 This cleaning job may cause unexpected long delays, so the most important goal
71 is to hide the latencies to users. And also definitely, it should reduce the
72 amount of valid data to be moved, and move them quickly as well.
73
74 ================================================================================
75 KEY FEATURES
76 ================================================================================
77
78 Flash Awareness
79 ---------------
80 - Enlarge the random write area for better performance, but provide the high
81 spatial locality
82 - Align FS data structures to the operational units in FTL as best efforts
83
84 Wandering Tree Problem
85 ----------------------
86 - Use a term, “node”, that represents inodes as well as various pointer blocks
87 - Introduce Node Address Table (NAT) containing the locations of all the “node”
88 blocks; this will cut off the update propagation.
89
90 Cleaning Overhead
91 -----------------
92 - Support a background cleaning process
93 - Support greedy and cost-benefit algorithms for victim selection policies
94 - Support multi-head logs for static/dynamic hot and cold data separation
95 - Introduce adaptive logging for efficient block allocation
96
97 ================================================================================
98 MOUNT OPTIONS
99 ================================================================================
100
101 background_gc=%s Turn on/off cleaning operations, namely garbage
102 collection, triggered in background when I/O subsystem is
103 idle. If background_gc=on, it will turn on the garbage
104 collection and if background_gc=off, garbage collection
105 will be turned off. If background_gc=sync, it will turn
106 on synchronous garbage collection running in background.
107 Default value for this option is on. So garbage
108 collection is on by default.
109 disable_roll_forward Disable the roll-forward recovery routine
110 norecovery Disable the roll-forward recovery routine, mounted read-
111 only (i.e., -o ro,disable_roll_forward)
112 discard Issue discard/TRIM commands when a segment is cleaned.
113 no_heap Disable heap-style segment allocation which finds free
114 segments for data from the beginning of main area, while
115 for node from the end of main area.
116 nouser_xattr Disable Extended User Attributes. Note: xattr is enabled
117 by default if CONFIG_F2FS_FS_XATTR is selected.
118 noacl Disable POSIX Access Control List. Note: acl is enabled
119 by default if CONFIG_F2FS_FS_POSIX_ACL is selected.
120 active_logs=%u Support configuring the number of active logs. In the
121 current design, f2fs supports only 2, 4, and 6 logs.
122 Default number is 6.
123 disable_ext_identify Disable the extension list configured by mkfs, so f2fs
124 does not aware of cold files such as media files.
125 inline_xattr Enable the inline xattrs feature.
126 inline_data Enable the inline data feature: New created small(<~3.4k)
127 files can be written into inode block.
128 inline_dentry Enable the inline dir feature: data in new created
129 directory entries can be written into inode block. The
130 space of inode block which is used to store inline
131 dentries is limited to ~3.4k.
132 flush_merge Merge concurrent cache_flush commands as much as possible
133 to eliminate redundant command issues. If the underlying
134 device handles the cache_flush command relatively slowly,
135 recommend to enable this option.
136 nobarrier This option can be used if underlying storage guarantees
137 its cached data should be written to the novolatile area.
138 If this option is set, no cache_flush commands are issued
139 but f2fs still guarantees the write ordering of all the
140 data writes.
141 fastboot This option is used when a system wants to reduce mount
142 time as much as possible, even though normal performance
143 can be sacrificed.
144 extent_cache Enable an extent cache based on rb-tree, it can cache
145 as many as extent which map between contiguous logical
146 address and physical address per inode, resulting in
147 increasing the cache hit ratio. Set by default.
148 noextent_cache Disable an extent cache based on rb-tree explicitly, see
149 the above extent_cache mount option.
150 noinline_data Disable the inline data feature, inline data feature is
151 enabled by default.
152 data_flush Enable data flushing before checkpoint in order to
153 persist data of regular and symlink.
154
155 ================================================================================
156 DEBUGFS ENTRIES
157 ================================================================================
158
159 /sys/kernel/debug/f2fs/ contains information about all the partitions mounted as
160 f2fs. Each file shows the whole f2fs information.
161
162 /sys/kernel/debug/f2fs/status includes:
163 - major file system information managed by f2fs currently
164 - average SIT information about whole segments
165 - current memory footprint consumed by f2fs.
166
167 ================================================================================
168 SYSFS ENTRIES
169 ================================================================================
170
171 Information about mounted f2f2 file systems can be found in
172 /sys/fs/f2fs. Each mounted filesystem will have a directory in
173 /sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda).
174 The files in each per-device directory are shown in table below.
175
176 Files in /sys/fs/f2fs/<devname>
177 (see also Documentation/ABI/testing/sysfs-fs-f2fs)
178 ..............................................................................
179 File Content
180
181 gc_max_sleep_time This tuning parameter controls the maximum sleep
182 time for the garbage collection thread. Time is
183 in milliseconds.
184
185 gc_min_sleep_time This tuning parameter controls the minimum sleep
186 time for the garbage collection thread. Time is
187 in milliseconds.
188
189 gc_no_gc_sleep_time This tuning parameter controls the default sleep
190 time for the garbage collection thread. Time is
191 in milliseconds.
192
193 gc_idle This parameter controls the selection of victim
194 policy for garbage collection. Setting gc_idle = 0
195 (default) will disable this option. Setting
196 gc_idle = 1 will select the Cost Benefit approach
197 & setting gc_idle = 2 will select the greedy approach.
198
199 reclaim_segments This parameter controls the number of prefree
200 segments to be reclaimed. If the number of prefree
201 segments is larger than the number of segments
202 in the proportion to the percentage over total
203 volume size, f2fs tries to conduct checkpoint to
204 reclaim the prefree segments to free segments.
205 By default, 5% over total # of segments.
206
207 max_small_discards This parameter controls the number of discard
208 commands that consist small blocks less than 2MB.
209 The candidates to be discarded are cached until
210 checkpoint is triggered, and issued during the
211 checkpoint. By default, it is disabled with 0.
212
213 trim_sections This parameter controls the number of sections
214 to be trimmed out in batch mode when FITRIM
215 conducts. 32 sections is set by default.
216
217 ipu_policy This parameter controls the policy of in-place
218 updates in f2fs. There are five policies:
219 0x01: F2FS_IPU_FORCE, 0x02: F2FS_IPU_SSR,
220 0x04: F2FS_IPU_UTIL, 0x08: F2FS_IPU_SSR_UTIL,
221 0x10: F2FS_IPU_FSYNC.
222
223 min_ipu_util This parameter controls the threshold to trigger
224 in-place-updates. The number indicates percentage
225 of the filesystem utilization, and used by
226 F2FS_IPU_UTIL and F2FS_IPU_SSR_UTIL policies.
227
228 min_fsync_blocks This parameter controls the threshold to trigger
229 in-place-updates when F2FS_IPU_FSYNC mode is set.
230 The number indicates the number of dirty pages
231 when fsync needs to flush on its call path. If
232 the number is less than this value, it triggers
233 in-place-updates.
234
235 max_victim_search This parameter controls the number of trials to
236 find a victim segment when conducting SSR and
237 cleaning operations. The default value is 4096
238 which covers 8GB block address range.
239
240 dir_level This parameter controls the directory level to
241 support large directory. If a directory has a
242 number of files, it can reduce the file lookup
243 latency by increasing this dir_level value.
244 Otherwise, it needs to decrease this value to
245 reduce the space overhead. The default value is 0.
246
247 ram_thresh This parameter controls the memory footprint used
248 by free nids and cached nat entries. By default,
249 10 is set, which indicates 10 MB / 1 GB RAM.
250
251 ================================================================================
252 USAGE
253 ================================================================================
254
255 1. Download userland tools and compile them.
256
257 2. Skip, if f2fs was compiled statically inside kernel.
258 Otherwise, insert the f2fs.ko module.
259 # insmod f2fs.ko
260
261 3. Create a directory trying to mount
262 # mkdir /mnt/f2fs
263
264 4. Format the block device, and then mount as f2fs
265 # mkfs.f2fs -l label /dev/block_device
266 # mount -t f2fs /dev/block_device /mnt/f2fs
267
268 mkfs.f2fs
269 ---------
270 The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem,
271 which builds a basic on-disk layout.
272
273 The options consist of:
274 -l [label] : Give a volume label, up to 512 unicode name.
275 -a [0 or 1] : Split start location of each area for heap-based allocation.
276 1 is set by default, which performs this.
277 -o [int] : Set overprovision ratio in percent over volume size.
278 5 is set by default.
279 -s [int] : Set the number of segments per section.
280 1 is set by default.
281 -z [int] : Set the number of sections per zone.
282 1 is set by default.
283 -e [str] : Set basic extension list. e.g. "mp3,gif,mov"
284 -t [0 or 1] : Disable discard command or not.
285 1 is set by default, which conducts discard.
286
287 fsck.f2fs
288 ---------
289 The fsck.f2fs is a tool to check the consistency of an f2fs-formatted
290 partition, which examines whether the filesystem metadata and user-made data
291 are cross-referenced correctly or not.
292 Note that, initial version of the tool does not fix any inconsistency.
293
294 The options consist of:
295 -d debug level [default:0]
296
297 dump.f2fs
298 ---------
299 The dump.f2fs shows the information of specific inode and dumps SSA and SIT to
300 file. Each file is dump_ssa and dump_sit.
301
302 The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem.
303 It shows on-disk inode information recognized by a given inode number, and is
304 able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and
305 ./dump_sit respectively.
306
307 The options consist of:
308 -d debug level [default:0]
309 -i inode no (hex)
310 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1]
311 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1]
312
313 Examples:
314 # dump.f2fs -i [ino] /dev/sdx
315 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump)
316 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump)
317
318 ================================================================================
319 DESIGN
320 ================================================================================
321
322 On-disk Layout
323 --------------
324
325 F2FS divides the whole volume into a number of segments, each of which is fixed
326 to 2MB in size. A section is composed of consecutive segments, and a zone
327 consists of a set of sections. By default, section and zone sizes are set to one
328 segment size identically, but users can easily modify the sizes by mkfs.
329
330 F2FS splits the entire volume into six areas, and all the areas except superblock
331 consists of multiple segments as described below.
332
333 align with the zone size <-|
334 |-> align with the segment size
335 _________________________________________________________________________
336 | | | Segment | Node | Segment | |
337 | Superblock | Checkpoint | Info. | Address | Summary | Main |
338 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | |
339 |____________|_____2______|______N______|______N______|______N_____|__N___|
340 . .
341 . .
342 . .
343 ._________________________________________.
344 |_Segment_|_..._|_Segment_|_..._|_Segment_|
345 . .
346 ._________._________
347 |_section_|__...__|_
348 . .
349 .________.
350 |__zone__|
351
352 - Superblock (SB)
353 : It is located at the beginning of the partition, and there exist two copies
354 to avoid file system crash. It contains basic partition information and some
355 default parameters of f2fs.
356
357 - Checkpoint (CP)
358 : It contains file system information, bitmaps for valid NAT/SIT sets, orphan
359 inode lists, and summary entries of current active segments.
360
361 - Segment Information Table (SIT)
362 : It contains segment information such as valid block count and bitmap for the
363 validity of all the blocks.
364
365 - Node Address Table (NAT)
366 : It is composed of a block address table for all the node blocks stored in
367 Main area.
368
369 - Segment Summary Area (SSA)
370 : It contains summary entries which contains the owner information of all the
371 data and node blocks stored in Main area.
372
373 - Main Area
374 : It contains file and directory data including their indices.
375
376 In order to avoid misalignment between file system and flash-based storage, F2FS
377 aligns the start block address of CP with the segment size. Also, it aligns the
378 start block address of Main area with the zone size by reserving some segments
379 in SSA area.
380
381 Reference the following survey for additional technical details.
382 https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey
383
384 File System Metadata Structure
385 ------------------------------
386
387 F2FS adopts the checkpointing scheme to maintain file system consistency. At
388 mount time, F2FS first tries to find the last valid checkpoint data by scanning
389 CP area. In order to reduce the scanning time, F2FS uses only two copies of CP.
390 One of them always indicates the last valid data, which is called as shadow copy
391 mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism.
392
393 For file system consistency, each CP points to which NAT and SIT copies are
394 valid, as shown as below.
395
396 +--------+----------+---------+
397 | CP | SIT | NAT |
398 +--------+----------+---------+
399 . . . .
400 . . . .
401 . . . .
402 +-------+-------+--------+--------+--------+--------+
403 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 |
404 +-------+-------+--------+--------+--------+--------+
405 | ^ ^
406 | | |
407 `----------------------------------------'
408
409 Index Structure
410 ---------------
411
412 The key data structure to manage the data locations is a "node". Similar to
413 traditional file structures, F2FS has three types of node: inode, direct node,
414 indirect node. F2FS assigns 4KB to an inode block which contains 923 data block
415 indices, two direct node pointers, two indirect node pointers, and one double
416 indirect node pointer as described below. One direct node block contains 1018
417 data blocks, and one indirect node block contains also 1018 node blocks. Thus,
418 one inode block (i.e., a file) covers:
419
420 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB.
421
422 Inode block (4KB)
423 |- data (923)
424 |- direct node (2)
425 | `- data (1018)
426 |- indirect node (2)
427 | `- direct node (1018)
428 | `- data (1018)
429 `- double indirect node (1)
430 `- indirect node (1018)
431 `- direct node (1018)
432 `- data (1018)
433
434 Note that, all the node blocks are mapped by NAT which means the location of
435 each node is translated by the NAT table. In the consideration of the wandering
436 tree problem, F2FS is able to cut off the propagation of node updates caused by
437 leaf data writes.
438
439 Directory Structure
440 -------------------
441
442 A directory entry occupies 11 bytes, which consists of the following attributes.
443
444 - hash hash value of the file name
445 - ino inode number
446 - len the length of file name
447 - type file type such as directory, symlink, etc
448
449 A dentry block consists of 214 dentry slots and file names. Therein a bitmap is
450 used to represent whether each dentry is valid or not. A dentry block occupies
451 4KB with the following composition.
452
453 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) +
454 dentries(11 * 214 bytes) + file name (8 * 214 bytes)
455
456 [Bucket]
457 +--------------------------------+
458 |dentry block 1 | dentry block 2 |
459 +--------------------------------+
460 . .
461 . .
462 . [Dentry Block Structure: 4KB] .
463 +--------+----------+----------+------------+
464 | bitmap | reserved | dentries | file names |
465 +--------+----------+----------+------------+
466 [Dentry Block: 4KB] . .
467 . .
468 . .
469 +------+------+-----+------+
470 | hash | ino | len | type |
471 +------+------+-----+------+
472 [Dentry Structure: 11 bytes]
473
474 F2FS implements multi-level hash tables for directory structure. Each level has
475 a hash table with dedicated number of hash buckets as shown below. Note that
476 "A(2B)" means a bucket includes 2 data blocks.
477
478 ----------------------
479 A : bucket
480 B : block
481 N : MAX_DIR_HASH_DEPTH
482 ----------------------
483
484 level #0 | A(2B)
485 |
486 level #1 | A(2B) - A(2B)
487 |
488 level #2 | A(2B) - A(2B) - A(2B) - A(2B)
489 . | . . . .
490 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B)
491 . | . . . .
492 level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B)
493
494 The number of blocks and buckets are determined by,
495
496 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2,
497 # of blocks in level #n = |
498 `- 4, Otherwise
499
500 ,- 2^(n + dir_level),
501 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2,
502 # of buckets in level #n = |
503 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1),
504 Otherwise
505
506 When F2FS finds a file name in a directory, at first a hash value of the file
507 name is calculated. Then, F2FS scans the hash table in level #0 to find the
508 dentry consisting of the file name and its inode number. If not found, F2FS
509 scans the next hash table in level #1. In this way, F2FS scans hash tables in
510 each levels incrementally from 1 to N. In each levels F2FS needs to scan only
511 one bucket determined by the following equation, which shows O(log(# of files))
512 complexity.
513
514 bucket number to scan in level #n = (hash value) % (# of buckets in level #n)
515
516 In the case of file creation, F2FS finds empty consecutive slots that cover the
517 file name. F2FS searches the empty slots in the hash tables of whole levels from
518 1 to N in the same way as the lookup operation.
519
520 The following figure shows an example of two cases holding children.
521 --------------> Dir <--------------
522 | |
523 child child
524
525 child - child [hole] - child
526
527 child - child - child [hole] - [hole] - child
528
529 Case 1: Case 2:
530 Number of children = 6, Number of children = 3,
531 File size = 7 File size = 7
532
533 Default Block Allocation
534 ------------------------
535
536 At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node
537 and Hot/Warm/Cold data.
538
539 - Hot node contains direct node blocks of directories.
540 - Warm node contains direct node blocks except hot node blocks.
541 - Cold node contains indirect node blocks
542 - Hot data contains dentry blocks
543 - Warm data contains data blocks except hot and cold data blocks
544 - Cold data contains multimedia data or migrated data blocks
545
546 LFS has two schemes for free space management: threaded log and copy-and-compac-
547 tion. The copy-and-compaction scheme which is known as cleaning, is well-suited
548 for devices showing very good sequential write performance, since free segments
549 are served all the time for writing new data. However, it suffers from cleaning
550 overhead under high utilization. Contrarily, the threaded log scheme suffers
551 from random writes, but no cleaning process is needed. F2FS adopts a hybrid
552 scheme where the copy-and-compaction scheme is adopted by default, but the
553 policy is dynamically changed to the threaded log scheme according to the file
554 system status.
555
556 In order to align F2FS with underlying flash-based storage, F2FS allocates a
557 segment in a unit of section. F2FS expects that the section size would be the
558 same as the unit size of garbage collection in FTL. Furthermore, with respect
559 to the mapping granularity in FTL, F2FS allocates each section of the active
560 logs from different zones as much as possible, since FTL can write the data in
561 the active logs into one allocation unit according to its mapping granularity.
562
563 Cleaning process
564 ----------------
565
566 F2FS does cleaning both on demand and in the background. On-demand cleaning is
567 triggered when there are not enough free segments to serve VFS calls. Background
568 cleaner is operated by a kernel thread, and triggers the cleaning job when the
569 system is idle.
570
571 F2FS supports two victim selection policies: greedy and cost-benefit algorithms.
572 In the greedy algorithm, F2FS selects a victim segment having the smallest number
573 of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment
574 according to the segment age and the number of valid blocks in order to address
575 log block thrashing problem in the greedy algorithm. F2FS adopts the greedy
576 algorithm for on-demand cleaner, while background cleaner adopts cost-benefit
577 algorithm.
578
579 In order to identify whether the data in the victim segment are valid or not,
580 F2FS manages a bitmap. Each bit represents the validity of a block, and the
581 bitmap is composed of a bit stream covering whole blocks in main area.
This page took 0.043868 seconds and 5 git commands to generate.