1================================================================================ 2WHAT IS Flash-Friendly File System (F2FS)? 3================================================================================ 4 5NAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have 6been equipped on a variety systems ranging from mobile to server systems. Since 7they are known to have different characteristics from the conventional rotating 8disks, a file system, an upper layer to the storage device, should adapt to the 9changes from the sketch in the design level. 10 11F2FS is a file system exploiting NAND flash memory-based storage devices, which 12is based on Log-structured File System (LFS). The design has been focused on 13addressing the fundamental issues in LFS, which are snowball effect of wandering 14tree and high cleaning overhead. 15 16Since a NAND flash memory-based storage device shows different characteristic 17according to its internal geometry or flash memory management scheme, namely FTL, 18F2FS and its tools support various parameters not only for configuring on-disk 19layout, but also for selecting allocation and cleaning algorithms. 20 21The following git tree provides the file system formatting tool (mkfs.f2fs), 22a 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 25For reporting bugs and sending patches, please use the following mailing list: 26>> linux-f2fs-devel@lists.sourceforge.net 27 28================================================================================ 29BACKGROUND AND DESIGN ISSUES 30================================================================================ 31 32Log-structured File System (LFS) 33-------------------------------- 34"A log-structured file system writes all modifications to disk sequentially in 35a log-like structure, thereby speeding up both file writing and crash recovery. 36The log is the only structure on disk; it contains indexing information so that 37files can be read back from the log efficiently. In order to maintain large free 38areas on disk for fast writing, we divide the log into segments and use a 39segment cleaner to compress the live information from heavily fragmented 40segments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and 41implementation of a log-structured file system", ACM Trans. Computer Systems 4210, 1, 26–52. 43 44Wandering Tree Problem 45---------------------- 46In LFS, when a file data is updated and written to the end of log, its direct 47pointer block is updated due to the changed location. Then the indirect pointer 48block is also updated due to the direct pointer block update. In this manner, 49the upper index structures such as inode, inode map, and checkpoint block are 50also updated recursively. This problem is called as wandering tree problem [1], 51and in order to enhance the performance, it should eliminate or relax the update 52propagation as much as possible. 53 54[1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/ 55 56Cleaning Overhead 57----------------- 58Since LFS is based on out-of-place writes, it produces so many obsolete blocks 59scattered across the whole storage. In order to serve new empty log space, it 60needs to reclaim these obsolete blocks seamlessly to users. This job is called 61as a cleaning process. 62 63The process consists of three operations as follows. 641. A victim segment is selected through referencing segment usage table. 652. It loads parent index structures of all the data in the victim identified by 66 segment summary blocks. 673. It checks the cross-reference between the data and its parent index structure. 684. It moves valid data selectively. 69 70This cleaning job may cause unexpected long delays, so the most important goal 71is to hide the latencies to users. And also definitely, it should reduce the 72amount of valid data to be moved, and move them quickly as well. 73 74================================================================================ 75KEY FEATURES 76================================================================================ 77 78Flash 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 84Wandering 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 90Cleaning 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================================================================================ 98MOUNT OPTIONS 99================================================================================ 100 101background_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. 109disable_roll_forward Disable the roll-forward recovery routine 110norecovery Disable the roll-forward recovery routine, mounted read- 111 only (i.e., -o ro,disable_roll_forward) 112discard/nodiscard Enable/disable real-time discard in f2fs, if discard is 113 enabled, f2fs will issue discard/TRIM commands when a 114 segment is cleaned. 115no_heap Disable heap-style segment allocation which finds free 116 segments for data from the beginning of main area, while 117 for node from the end of main area. 118nouser_xattr Disable Extended User Attributes. Note: xattr is enabled 119 by default if CONFIG_F2FS_FS_XATTR is selected. 120noacl Disable POSIX Access Control List. Note: acl is enabled 121 by default if CONFIG_F2FS_FS_POSIX_ACL is selected. 122active_logs=%u Support configuring the number of active logs. In the 123 current design, f2fs supports only 2, 4, and 6 logs. 124 Default number is 6. 125disable_ext_identify Disable the extension list configured by mkfs, so f2fs 126 does not aware of cold files such as media files. 127inline_xattr Enable the inline xattrs feature. 128noinline_xattr Disable the inline xattrs feature. 129inline_data Enable the inline data feature: New created small(<~3.4k) 130 files can be written into inode block. 131inline_dentry Enable the inline dir feature: data in new created 132 directory entries can be written into inode block. The 133 space of inode block which is used to store inline 134 dentries is limited to ~3.4k. 135noinline_dentry Diable the inline dentry feature. 136flush_merge Merge concurrent cache_flush commands as much as possible 137 to eliminate redundant command issues. If the underlying 138 device handles the cache_flush command relatively slowly, 139 recommend to enable this option. 140nobarrier This option can be used if underlying storage guarantees 141 its cached data should be written to the novolatile area. 142 If this option is set, no cache_flush commands are issued 143 but f2fs still guarantees the write ordering of all the 144 data writes. 145fastboot This option is used when a system wants to reduce mount 146 time as much as possible, even though normal performance 147 can be sacrificed. 148extent_cache Enable an extent cache based on rb-tree, it can cache 149 as many as extent which map between contiguous logical 150 address and physical address per inode, resulting in 151 increasing the cache hit ratio. Set by default. 152noextent_cache Disable an extent cache based on rb-tree explicitly, see 153 the above extent_cache mount option. 154noinline_data Disable the inline data feature, inline data feature is 155 enabled by default. 156data_flush Enable data flushing before checkpoint in order to 157 persist data of regular and symlink. 158mode=%s Control block allocation mode which supports "adaptive" 159 and "lfs". In "lfs" mode, there should be no random 160 writes towards main area. 161io_bits=%u Set the bit size of write IO requests. It should be set 162 with "mode=lfs". 163usrquota Enable plain user disk quota accounting. 164grpquota Enable plain group disk quota accounting. 165prjquota Enable plain project quota accounting. 166usrjquota=<file> Appoint specified file and type during mount, so that quota 167grpjquota=<file> information can be properly updated during recovery flow, 168prjjquota=<file> <quota file>: must be in root directory; 169jqfmt=<quota type> <quota type>: [vfsold,vfsv0,vfsv1]. 170offusrjquota Turn off user journelled quota. 171offgrpjquota Turn off group journelled quota. 172offprjjquota Turn off project journelled quota. 173quota Enable plain user disk quota accounting. 174noquota Disable all plain disk quota option. 175whint_mode=%s Control which write hints are passed down to block 176 layer. This supports "off", "user-based", and 177 "fs-based". In "off" mode (default), f2fs does not pass 178 down hints. In "user-based" mode, f2fs tries to pass 179 down hints given by users. And in "fs-based" mode, f2fs 180 passes down hints with its policy. 181alloc_mode=%s Adjust block allocation policy, which supports "reuse" 182 and "default". 183fsync_mode=%s Control the policy of fsync. Currently supports "posix", 184 "strict", and "nobarrier". In "posix" mode, which is 185 default, fsync will follow POSIX semantics and does a 186 light operation to improve the filesystem performance. 187 In "strict" mode, fsync will be heavy and behaves in line 188 with xfs, ext4 and btrfs, where xfstest generic/342 will 189 pass, but the performance will regress. "nobarrier" is 190 based on "posix", but doesn't issue flush command for 191 non-atomic files likewise "nobarrier" mount option. 192test_dummy_encryption Enable dummy encryption, which provides a fake fscrypt 193 context. The fake fscrypt context is used by xfstests. 194 195================================================================================ 196DEBUGFS ENTRIES 197================================================================================ 198 199/sys/kernel/debug/f2fs/ contains information about all the partitions mounted as 200f2fs. Each file shows the whole f2fs information. 201 202/sys/kernel/debug/f2fs/status includes: 203 - major file system information managed by f2fs currently 204 - average SIT information about whole segments 205 - current memory footprint consumed by f2fs. 206 207================================================================================ 208SYSFS ENTRIES 209================================================================================ 210 211Information about mounted f2f2 file systems can be found in 212/sys/fs/f2fs. Each mounted filesystem will have a directory in 213/sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda). 214The files in each per-device directory are shown in table below. 215 216Files in /sys/fs/f2fs/<devname> 217(see also Documentation/ABI/testing/sysfs-fs-f2fs) 218.............................................................................. 219 File Content 220 221 gc_max_sleep_time This tuning parameter controls the maximum sleep 222 time for the garbage collection thread. Time is 223 in milliseconds. 224 225 gc_min_sleep_time This tuning parameter controls the minimum sleep 226 time for the garbage collection thread. Time is 227 in milliseconds. 228 229 gc_no_gc_sleep_time This tuning parameter controls the default sleep 230 time for the garbage collection thread. Time is 231 in milliseconds. 232 233 gc_idle This parameter controls the selection of victim 234 policy for garbage collection. Setting gc_idle = 0 235 (default) will disable this option. Setting 236 gc_idle = 1 will select the Cost Benefit approach 237 & setting gc_idle = 2 will select the greedy approach. 238 239 gc_urgent This parameter controls triggering background GCs 240 urgently or not. Setting gc_urgent = 0 [default] 241 makes back to default behavior, while if it is set 242 to 1, background thread starts to do GC by given 243 gc_urgent_sleep_time interval. 244 245 gc_urgent_sleep_time This parameter controls sleep time for gc_urgent. 246 500 ms is set by default. See above gc_urgent. 247 248 reclaim_segments This parameter controls the number of prefree 249 segments to be reclaimed. If the number of prefree 250 segments is larger than the number of segments 251 in the proportion to the percentage over total 252 volume size, f2fs tries to conduct checkpoint to 253 reclaim the prefree segments to free segments. 254 By default, 5% over total # of segments. 255 256 max_small_discards This parameter controls the number of discard 257 commands that consist small blocks less than 2MB. 258 The candidates to be discarded are cached until 259 checkpoint is triggered, and issued during the 260 checkpoint. By default, it is disabled with 0. 261 262 trim_sections This parameter controls the number of sections 263 to be trimmed out in batch mode when FITRIM 264 conducts. 32 sections is set by default. 265 266 ipu_policy This parameter controls the policy of in-place 267 updates in f2fs. There are five policies: 268 0x01: F2FS_IPU_FORCE, 0x02: F2FS_IPU_SSR, 269 0x04: F2FS_IPU_UTIL, 0x08: F2FS_IPU_SSR_UTIL, 270 0x10: F2FS_IPU_FSYNC. 271 272 min_ipu_util This parameter controls the threshold to trigger 273 in-place-updates. The number indicates percentage 274 of the filesystem utilization, and used by 275 F2FS_IPU_UTIL and F2FS_IPU_SSR_UTIL policies. 276 277 min_fsync_blocks This parameter controls the threshold to trigger 278 in-place-updates when F2FS_IPU_FSYNC mode is set. 279 The number indicates the number of dirty pages 280 when fsync needs to flush on its call path. If 281 the number is less than this value, it triggers 282 in-place-updates. 283 284 max_victim_search This parameter controls the number of trials to 285 find a victim segment when conducting SSR and 286 cleaning operations. The default value is 4096 287 which covers 8GB block address range. 288 289 dir_level This parameter controls the directory level to 290 support large directory. If a directory has a 291 number of files, it can reduce the file lookup 292 latency by increasing this dir_level value. 293 Otherwise, it needs to decrease this value to 294 reduce the space overhead. The default value is 0. 295 296 ram_thresh This parameter controls the memory footprint used 297 by free nids and cached nat entries. By default, 298 10 is set, which indicates 10 MB / 1 GB RAM. 299 300================================================================================ 301USAGE 302================================================================================ 303 3041. Download userland tools and compile them. 305 3062. Skip, if f2fs was compiled statically inside kernel. 307 Otherwise, insert the f2fs.ko module. 308 # insmod f2fs.ko 309 3103. Create a directory trying to mount 311 # mkdir /mnt/f2fs 312 3134. Format the block device, and then mount as f2fs 314 # mkfs.f2fs -l label /dev/block_device 315 # mount -t f2fs /dev/block_device /mnt/f2fs 316 317mkfs.f2fs 318--------- 319The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem, 320which builds a basic on-disk layout. 321 322The options consist of: 323-l [label] : Give a volume label, up to 512 unicode name. 324-a [0 or 1] : Split start location of each area for heap-based allocation. 325 1 is set by default, which performs this. 326-o [int] : Set overprovision ratio in percent over volume size. 327 5 is set by default. 328-s [int] : Set the number of segments per section. 329 1 is set by default. 330-z [int] : Set the number of sections per zone. 331 1 is set by default. 332-e [str] : Set basic extension list. e.g. "mp3,gif,mov" 333-t [0 or 1] : Disable discard command or not. 334 1 is set by default, which conducts discard. 335 336fsck.f2fs 337--------- 338The fsck.f2fs is a tool to check the consistency of an f2fs-formatted 339partition, which examines whether the filesystem metadata and user-made data 340are cross-referenced correctly or not. 341Note that, initial version of the tool does not fix any inconsistency. 342 343The options consist of: 344 -d debug level [default:0] 345 346dump.f2fs 347--------- 348The dump.f2fs shows the information of specific inode and dumps SSA and SIT to 349file. Each file is dump_ssa and dump_sit. 350 351The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem. 352It shows on-disk inode information recognized by a given inode number, and is 353able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and 354./dump_sit respectively. 355 356The options consist of: 357 -d debug level [default:0] 358 -i inode no (hex) 359 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1] 360 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1] 361 362Examples: 363# dump.f2fs -i [ino] /dev/sdx 364# dump.f2fs -s 0~-1 /dev/sdx (SIT dump) 365# dump.f2fs -a 0~-1 /dev/sdx (SSA dump) 366 367================================================================================ 368DESIGN 369================================================================================ 370 371On-disk Layout 372-------------- 373 374F2FS divides the whole volume into a number of segments, each of which is fixed 375to 2MB in size. A section is composed of consecutive segments, and a zone 376consists of a set of sections. By default, section and zone sizes are set to one 377segment size identically, but users can easily modify the sizes by mkfs. 378 379F2FS splits the entire volume into six areas, and all the areas except superblock 380consists of multiple segments as described below. 381 382 align with the zone size <-| 383 |-> align with the segment size 384 _________________________________________________________________________ 385 | | | Segment | Node | Segment | | 386 | Superblock | Checkpoint | Info. | Address | Summary | Main | 387 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | | 388 |____________|_____2______|______N______|______N______|______N_____|__N___| 389 . . 390 . . 391 . . 392 ._________________________________________. 393 |_Segment_|_..._|_Segment_|_..._|_Segment_| 394 . . 395 ._________._________ 396 |_section_|__...__|_ 397 . . 398 .________. 399 |__zone__| 400 401- Superblock (SB) 402 : It is located at the beginning of the partition, and there exist two copies 403 to avoid file system crash. It contains basic partition information and some 404 default parameters of f2fs. 405 406- Checkpoint (CP) 407 : It contains file system information, bitmaps for valid NAT/SIT sets, orphan 408 inode lists, and summary entries of current active segments. 409 410- Segment Information Table (SIT) 411 : It contains segment information such as valid block count and bitmap for the 412 validity of all the blocks. 413 414- Node Address Table (NAT) 415 : It is composed of a block address table for all the node blocks stored in 416 Main area. 417 418- Segment Summary Area (SSA) 419 : It contains summary entries which contains the owner information of all the 420 data and node blocks stored in Main area. 421 422- Main Area 423 : It contains file and directory data including their indices. 424 425In order to avoid misalignment between file system and flash-based storage, F2FS 426aligns the start block address of CP with the segment size. Also, it aligns the 427start block address of Main area with the zone size by reserving some segments 428in SSA area. 429 430Reference the following survey for additional technical details. 431https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey 432 433File System Metadata Structure 434------------------------------ 435 436F2FS adopts the checkpointing scheme to maintain file system consistency. At 437mount time, F2FS first tries to find the last valid checkpoint data by scanning 438CP area. In order to reduce the scanning time, F2FS uses only two copies of CP. 439One of them always indicates the last valid data, which is called as shadow copy 440mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism. 441 442For file system consistency, each CP points to which NAT and SIT copies are 443valid, as shown as below. 444 445 +--------+----------+---------+ 446 | CP | SIT | NAT | 447 +--------+----------+---------+ 448 . . . . 449 . . . . 450 . . . . 451 +-------+-------+--------+--------+--------+--------+ 452 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 | 453 +-------+-------+--------+--------+--------+--------+ 454 | ^ ^ 455 | | | 456 `----------------------------------------' 457 458Index Structure 459--------------- 460 461The key data structure to manage the data locations is a "node". Similar to 462traditional file structures, F2FS has three types of node: inode, direct node, 463indirect node. F2FS assigns 4KB to an inode block which contains 923 data block 464indices, two direct node pointers, two indirect node pointers, and one double 465indirect node pointer as described below. One direct node block contains 1018 466data blocks, and one indirect node block contains also 1018 node blocks. Thus, 467one inode block (i.e., a file) covers: 468 469 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB. 470 471 Inode block (4KB) 472 |- data (923) 473 |- direct node (2) 474 | `- data (1018) 475 |- indirect node (2) 476 | `- direct node (1018) 477 | `- data (1018) 478 `- double indirect node (1) 479 `- indirect node (1018) 480 `- direct node (1018) 481 `- data (1018) 482 483Note that, all the node blocks are mapped by NAT which means the location of 484each node is translated by the NAT table. In the consideration of the wandering 485tree problem, F2FS is able to cut off the propagation of node updates caused by 486leaf data writes. 487 488Directory Structure 489------------------- 490 491A directory entry occupies 11 bytes, which consists of the following attributes. 492 493- hash hash value of the file name 494- ino inode number 495- len the length of file name 496- type file type such as directory, symlink, etc 497 498A dentry block consists of 214 dentry slots and file names. Therein a bitmap is 499used to represent whether each dentry is valid or not. A dentry block occupies 5004KB with the following composition. 501 502 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) + 503 dentries(11 * 214 bytes) + file name (8 * 214 bytes) 504 505 [Bucket] 506 +--------------------------------+ 507 |dentry block 1 | dentry block 2 | 508 +--------------------------------+ 509 . . 510 . . 511 . [Dentry Block Structure: 4KB] . 512 +--------+----------+----------+------------+ 513 | bitmap | reserved | dentries | file names | 514 +--------+----------+----------+------------+ 515 [Dentry Block: 4KB] . . 516 . . 517 . . 518 +------+------+-----+------+ 519 | hash | ino | len | type | 520 +------+------+-----+------+ 521 [Dentry Structure: 11 bytes] 522 523F2FS implements multi-level hash tables for directory structure. Each level has 524a hash table with dedicated number of hash buckets as shown below. Note that 525"A(2B)" means a bucket includes 2 data blocks. 526 527---------------------- 528A : bucket 529B : block 530N : MAX_DIR_HASH_DEPTH 531---------------------- 532 533level #0 | A(2B) 534 | 535level #1 | A(2B) - A(2B) 536 | 537level #2 | A(2B) - A(2B) - A(2B) - A(2B) 538 . | . . . . 539level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B) 540 . | . . . . 541level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B) 542 543The number of blocks and buckets are determined by, 544 545 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2, 546 # of blocks in level #n = | 547 `- 4, Otherwise 548 549 ,- 2^(n + dir_level), 550 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2, 551 # of buckets in level #n = | 552 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1), 553 Otherwise 554 555When F2FS finds a file name in a directory, at first a hash value of the file 556name is calculated. Then, F2FS scans the hash table in level #0 to find the 557dentry consisting of the file name and its inode number. If not found, F2FS 558scans the next hash table in level #1. In this way, F2FS scans hash tables in 559each levels incrementally from 1 to N. In each levels F2FS needs to scan only 560one bucket determined by the following equation, which shows O(log(# of files)) 561complexity. 562 563 bucket number to scan in level #n = (hash value) % (# of buckets in level #n) 564 565In the case of file creation, F2FS finds empty consecutive slots that cover the 566file name. F2FS searches the empty slots in the hash tables of whole levels from 5671 to N in the same way as the lookup operation. 568 569The following figure shows an example of two cases holding children. 570 --------------> Dir <-------------- 571 | | 572 child child 573 574 child - child [hole] - child 575 576 child - child - child [hole] - [hole] - child 577 578 Case 1: Case 2: 579 Number of children = 6, Number of children = 3, 580 File size = 7 File size = 7 581 582Default Block Allocation 583------------------------ 584 585At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node 586and Hot/Warm/Cold data. 587 588- Hot node contains direct node blocks of directories. 589- Warm node contains direct node blocks except hot node blocks. 590- Cold node contains indirect node blocks 591- Hot data contains dentry blocks 592- Warm data contains data blocks except hot and cold data blocks 593- Cold data contains multimedia data or migrated data blocks 594 595LFS has two schemes for free space management: threaded log and copy-and-compac- 596tion. The copy-and-compaction scheme which is known as cleaning, is well-suited 597for devices showing very good sequential write performance, since free segments 598are served all the time for writing new data. However, it suffers from cleaning 599overhead under high utilization. Contrarily, the threaded log scheme suffers 600from random writes, but no cleaning process is needed. F2FS adopts a hybrid 601scheme where the copy-and-compaction scheme is adopted by default, but the 602policy is dynamically changed to the threaded log scheme according to the file 603system status. 604 605In order to align F2FS with underlying flash-based storage, F2FS allocates a 606segment in a unit of section. F2FS expects that the section size would be the 607same as the unit size of garbage collection in FTL. Furthermore, with respect 608to the mapping granularity in FTL, F2FS allocates each section of the active 609logs from different zones as much as possible, since FTL can write the data in 610the active logs into one allocation unit according to its mapping granularity. 611 612Cleaning process 613---------------- 614 615F2FS does cleaning both on demand and in the background. On-demand cleaning is 616triggered when there are not enough free segments to serve VFS calls. Background 617cleaner is operated by a kernel thread, and triggers the cleaning job when the 618system is idle. 619 620F2FS supports two victim selection policies: greedy and cost-benefit algorithms. 621In the greedy algorithm, F2FS selects a victim segment having the smallest number 622of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment 623according to the segment age and the number of valid blocks in order to address 624log block thrashing problem in the greedy algorithm. F2FS adopts the greedy 625algorithm for on-demand cleaner, while background cleaner adopts cost-benefit 626algorithm. 627 628In order to identify whether the data in the victim segment are valid or not, 629F2FS manages a bitmap. Each bit represents the validity of a block, and the 630bitmap is composed of a bit stream covering whole blocks in main area. 631