1.. SPDX-License-Identifier: GPL-2.0 2 3========================================== 4WHAT IS Flash-Friendly File System (F2FS)? 5========================================== 6 7NAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have 8been equipped on a variety systems ranging from mobile to server systems. Since 9they are known to have different characteristics from the conventional rotating 10disks, a file system, an upper layer to the storage device, should adapt to the 11changes from the sketch in the design level. 12 13F2FS is a file system exploiting NAND flash memory-based storage devices, which 14is based on Log-structured File System (LFS). The design has been focused on 15addressing the fundamental issues in LFS, which are snowball effect of wandering 16tree and high cleaning overhead. 17 18Since a NAND flash memory-based storage device shows different characteristic 19according to its internal geometry or flash memory management scheme, namely FTL, 20F2FS and its tools support various parameters not only for configuring on-disk 21layout, but also for selecting allocation and cleaning algorithms. 22 23The following git tree provides the file system formatting tool (mkfs.f2fs), 24a consistency checking tool (fsck.f2fs), and a debugging tool (dump.f2fs). 25 26- git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git 27 28For reporting bugs and sending patches, please use the following mailing list: 29 30- linux-f2fs-devel@lists.sourceforge.net 31 32Background and Design issues 33============================ 34 35Log-structured File System (LFS) 36-------------------------------- 37"A log-structured file system writes all modifications to disk sequentially in 38a log-like structure, thereby speeding up both file writing and crash recovery. 39The log is the only structure on disk; it contains indexing information so that 40files can be read back from the log efficiently. In order to maintain large free 41areas on disk for fast writing, we divide the log into segments and use a 42segment cleaner to compress the live information from heavily fragmented 43segments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and 44implementation of a log-structured file system", ACM Trans. Computer Systems 4510, 1, 26–52. 46 47Wandering Tree Problem 48---------------------- 49In LFS, when a file data is updated and written to the end of log, its direct 50pointer block is updated due to the changed location. Then the indirect pointer 51block is also updated due to the direct pointer block update. In this manner, 52the upper index structures such as inode, inode map, and checkpoint block are 53also updated recursively. This problem is called as wandering tree problem [1], 54and in order to enhance the performance, it should eliminate or relax the update 55propagation as much as possible. 56 57[1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/ 58 59Cleaning Overhead 60----------------- 61Since LFS is based on out-of-place writes, it produces so many obsolete blocks 62scattered across the whole storage. In order to serve new empty log space, it 63needs to reclaim these obsolete blocks seamlessly to users. This job is called 64as a cleaning process. 65 66The process consists of three operations as follows. 67 681. A victim segment is selected through referencing segment usage table. 692. It loads parent index structures of all the data in the victim identified by 70 segment summary blocks. 713. It checks the cross-reference between the data and its parent index structure. 724. It moves valid data selectively. 73 74This cleaning job may cause unexpected long delays, so the most important goal 75is to hide the latencies to users. And also definitely, it should reduce the 76amount of valid data to be moved, and move them quickly as well. 77 78Key Features 79============ 80 81Flash Awareness 82--------------- 83- Enlarge the random write area for better performance, but provide the high 84 spatial locality 85- Align FS data structures to the operational units in FTL as best efforts 86 87Wandering Tree Problem 88---------------------- 89- Use a term, “node”, that represents inodes as well as various pointer blocks 90- Introduce Node Address Table (NAT) containing the locations of all the “node” 91 blocks; this will cut off the update propagation. 92 93Cleaning Overhead 94----------------- 95- Support a background cleaning process 96- Support greedy and cost-benefit algorithms for victim selection policies 97- Support multi-head logs for static/dynamic hot and cold data separation 98- Introduce adaptive logging for efficient block allocation 99 100Mount Options 101============= 102 103 104======================== ============================================================ 105background_gc=%s Turn on/off cleaning operations, namely garbage 106 collection, triggered in background when I/O subsystem is 107 idle. If background_gc=on, it will turn on the garbage 108 collection and if background_gc=off, garbage collection 109 will be turned off. If background_gc=sync, it will turn 110 on synchronous garbage collection running in background. 111 Default value for this option is on. So garbage 112 collection is on by default. 113gc_merge When background_gc is on, this option can be enabled to 114 let background GC thread to handle foreground GC requests, 115 it can eliminate the sluggish issue caused by slow foreground 116 GC operation when GC is triggered from a process with limited 117 I/O and CPU resources. 118nogc_merge Disable GC merge feature. 119disable_roll_forward Disable the roll-forward recovery routine 120norecovery Disable the roll-forward recovery routine, mounted read- 121 only (i.e., -o ro,disable_roll_forward) 122discard/nodiscard Enable/disable real-time discard in f2fs, if discard is 123 enabled, f2fs will issue discard/TRIM commands when a 124 segment is cleaned. 125no_heap Disable heap-style segment allocation which finds free 126 segments for data from the beginning of main area, while 127 for node from the end of main area. 128nouser_xattr Disable Extended User Attributes. Note: xattr is enabled 129 by default if CONFIG_F2FS_FS_XATTR is selected. 130noacl Disable POSIX Access Control List. Note: acl is enabled 131 by default if CONFIG_F2FS_FS_POSIX_ACL is selected. 132active_logs=%u Support configuring the number of active logs. In the 133 current design, f2fs supports only 2, 4, and 6 logs. 134 Default number is 6. 135disable_ext_identify Disable the extension list configured by mkfs, so f2fs 136 is not aware of cold files such as media files. 137inline_xattr Enable the inline xattrs feature. 138noinline_xattr Disable the inline xattrs feature. 139inline_xattr_size=%u Support configuring inline xattr size, it depends on 140 flexible inline xattr feature. 141inline_data Enable the inline data feature: Newly created small (<~3.4k) 142 files can be written into inode block. 143inline_dentry Enable the inline dir feature: data in newly created 144 directory entries can be written into inode block. The 145 space of inode block which is used to store inline 146 dentries is limited to ~3.4k. 147noinline_dentry Disable the inline dentry feature. 148flush_merge Merge concurrent cache_flush commands as much as possible 149 to eliminate redundant command issues. If the underlying 150 device handles the cache_flush command relatively slowly, 151 recommend to enable this option. 152nobarrier This option can be used if underlying storage guarantees 153 its cached data should be written to the novolatile area. 154 If this option is set, no cache_flush commands are issued 155 but f2fs still guarantees the write ordering of all the 156 data writes. 157fastboot This option is used when a system wants to reduce mount 158 time as much as possible, even though normal performance 159 can be sacrificed. 160extent_cache Enable an extent cache based on rb-tree, it can cache 161 as many as extent which map between contiguous logical 162 address and physical address per inode, resulting in 163 increasing the cache hit ratio. Set by default. 164noextent_cache Disable an extent cache based on rb-tree explicitly, see 165 the above extent_cache mount option. 166noinline_data Disable the inline data feature, inline data feature is 167 enabled by default. 168data_flush Enable data flushing before checkpoint in order to 169 persist data of regular and symlink. 170reserve_root=%d Support configuring reserved space which is used for 171 allocation from a privileged user with specified uid or 172 gid, unit: 4KB, the default limit is 0.2% of user blocks. 173resuid=%d The user ID which may use the reserved blocks. 174resgid=%d The group ID which may use the reserved blocks. 175fault_injection=%d Enable fault injection in all supported types with 176 specified injection rate. 177fault_type=%d Support configuring fault injection type, should be 178 enabled with fault_injection option, fault type value 179 is shown below, it supports single or combined type. 180 181 =================== =========== 182 Type_Name Type_Value 183 =================== =========== 184 FAULT_KMALLOC 0x000000001 185 FAULT_KVMALLOC 0x000000002 186 FAULT_PAGE_ALLOC 0x000000004 187 FAULT_PAGE_GET 0x000000008 188 FAULT_ALLOC_BIO 0x000000010 189 FAULT_ALLOC_NID 0x000000020 190 FAULT_ORPHAN 0x000000040 191 FAULT_BLOCK 0x000000080 192 FAULT_DIR_DEPTH 0x000000100 193 FAULT_EVICT_INODE 0x000000200 194 FAULT_TRUNCATE 0x000000400 195 FAULT_READ_IO 0x000000800 196 FAULT_CHECKPOINT 0x000001000 197 FAULT_DISCARD 0x000002000 198 FAULT_WRITE_IO 0x000004000 199 =================== =========== 200mode=%s Control block allocation mode which supports "adaptive" 201 and "lfs". In "lfs" mode, there should be no random 202 writes towards main area. 203io_bits=%u Set the bit size of write IO requests. It should be set 204 with "mode=lfs". 205usrquota Enable plain user disk quota accounting. 206grpquota Enable plain group disk quota accounting. 207prjquota Enable plain project quota accounting. 208usrjquota=<file> Appoint specified file and type during mount, so that quota 209grpjquota=<file> information can be properly updated during recovery flow, 210prjjquota=<file> <quota file>: must be in root directory; 211jqfmt=<quota type> <quota type>: [vfsold,vfsv0,vfsv1]. 212offusrjquota Turn off user journalled quota. 213offgrpjquota Turn off group journalled quota. 214offprjjquota Turn off project journalled quota. 215quota Enable plain user disk quota accounting. 216noquota Disable all plain disk quota option. 217whint_mode=%s Control which write hints are passed down to block 218 layer. This supports "off", "user-based", and 219 "fs-based". In "off" mode (default), f2fs does not pass 220 down hints. In "user-based" mode, f2fs tries to pass 221 down hints given by users. And in "fs-based" mode, f2fs 222 passes down hints with its policy. 223alloc_mode=%s Adjust block allocation policy, which supports "reuse" 224 and "default". 225fsync_mode=%s Control the policy of fsync. Currently supports "posix", 226 "strict", and "nobarrier". In "posix" mode, which is 227 default, fsync will follow POSIX semantics and does a 228 light operation to improve the filesystem performance. 229 In "strict" mode, fsync will be heavy and behaves in line 230 with xfs, ext4 and btrfs, where xfstest generic/342 will 231 pass, but the performance will regress. "nobarrier" is 232 based on "posix", but doesn't issue flush command for 233 non-atomic files likewise "nobarrier" mount option. 234test_dummy_encryption 235test_dummy_encryption=%s 236 Enable dummy encryption, which provides a fake fscrypt 237 context. The fake fscrypt context is used by xfstests. 238 The argument may be either "v1" or "v2", in order to 239 select the corresponding fscrypt policy version. 240checkpoint=%s[:%u[%]] Set to "disable" to turn off checkpointing. Set to "enable" 241 to reenable checkpointing. Is enabled by default. While 242 disabled, any unmounting or unexpected shutdowns will cause 243 the filesystem contents to appear as they did when the 244 filesystem was mounted with that option. 245 While mounting with checkpoint=disabled, the filesystem must 246 run garbage collection to ensure that all available space can 247 be used. If this takes too much time, the mount may return 248 EAGAIN. You may optionally add a value to indicate how much 249 of the disk you would be willing to temporarily give up to 250 avoid additional garbage collection. This can be given as a 251 number of blocks, or as a percent. For instance, mounting 252 with checkpoint=disable:100% would always succeed, but it may 253 hide up to all remaining free space. The actual space that 254 would be unusable can be viewed at /sys/fs/f2fs/<disk>/unusable 255 This space is reclaimed once checkpoint=enable. 256compress_algorithm=%s Control compress algorithm, currently f2fs supports "lzo", 257 "lz4", "zstd" and "lzo-rle" algorithm. 258compress_log_size=%u Support configuring compress cluster size, the size will 259 be 4KB * (1 << %u), 16KB is minimum size, also it's 260 default size. 261compress_extension=%s Support adding specified extension, so that f2fs can enable 262 compression on those corresponding files, e.g. if all files 263 with '.ext' has high compression rate, we can set the '.ext' 264 on compression extension list and enable compression on 265 these file by default rather than to enable it via ioctl. 266 For other files, we can still enable compression via ioctl. 267 Note that, there is one reserved special extension '*', it 268 can be set to enable compression for all files. 269inlinecrypt When possible, encrypt/decrypt the contents of encrypted 270 files using the blk-crypto framework rather than 271 filesystem-layer encryption. This allows the use of 272 inline encryption hardware. The on-disk format is 273 unaffected. For more details, see 274 Documentation/block/inline-encryption.rst. 275atgc Enable age-threshold garbage collection, it provides high 276 effectiveness and efficiency on background GC. 277======================== ============================================================ 278 279Debugfs Entries 280=============== 281 282/sys/kernel/debug/f2fs/ contains information about all the partitions mounted as 283f2fs. Each file shows the whole f2fs information. 284 285/sys/kernel/debug/f2fs/status includes: 286 287 - major file system information managed by f2fs currently 288 - average SIT information about whole segments 289 - current memory footprint consumed by f2fs. 290 291Sysfs Entries 292============= 293 294Information about mounted f2fs file systems can be found in 295/sys/fs/f2fs. Each mounted filesystem will have a directory in 296/sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda). 297The files in each per-device directory are shown in table below. 298 299Files in /sys/fs/f2fs/<devname> 300(see also Documentation/ABI/testing/sysfs-fs-f2fs) 301 302Usage 303===== 304 3051. Download userland tools and compile them. 306 3072. Skip, if f2fs was compiled statically inside kernel. 308 Otherwise, insert the f2fs.ko module:: 309 310 # insmod f2fs.ko 311 3123. Create a directory to use when mounting:: 313 314 # mkdir /mnt/f2fs 315 3164. Format the block device, and then mount as f2fs:: 317 318 # mkfs.f2fs -l label /dev/block_device 319 # mount -t f2fs /dev/block_device /mnt/f2fs 320 321mkfs.f2fs 322--------- 323The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem, 324which builds a basic on-disk layout. 325 326The quick options consist of: 327 328=============== =========================================================== 329``-l [label]`` Give a volume label, up to 512 unicode name. 330``-a [0 or 1]`` Split start location of each area for heap-based allocation. 331 332 1 is set by default, which performs this. 333``-o [int]`` Set overprovision ratio in percent over volume size. 334 335 5 is set by default. 336``-s [int]`` Set the number of segments per section. 337 338 1 is set by default. 339``-z [int]`` Set the number of sections per zone. 340 341 1 is set by default. 342``-e [str]`` Set basic extension list. e.g. "mp3,gif,mov" 343``-t [0 or 1]`` Disable discard command or not. 344 345 1 is set by default, which conducts discard. 346=============== =========================================================== 347 348Note: please refer to the manpage of mkfs.f2fs(8) to get full option list. 349 350fsck.f2fs 351--------- 352The fsck.f2fs is a tool to check the consistency of an f2fs-formatted 353partition, which examines whether the filesystem metadata and user-made data 354are cross-referenced correctly or not. 355Note that, initial version of the tool does not fix any inconsistency. 356 357The quick options consist of:: 358 359 -d debug level [default:0] 360 361Note: please refer to the manpage of fsck.f2fs(8) to get full option list. 362 363dump.f2fs 364--------- 365The dump.f2fs shows the information of specific inode and dumps SSA and SIT to 366file. Each file is dump_ssa and dump_sit. 367 368The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem. 369It shows on-disk inode information recognized by a given inode number, and is 370able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and 371./dump_sit respectively. 372 373The options consist of:: 374 375 -d debug level [default:0] 376 -i inode no (hex) 377 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1] 378 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1] 379 380Examples:: 381 382 # dump.f2fs -i [ino] /dev/sdx 383 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump) 384 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump) 385 386Note: please refer to the manpage of dump.f2fs(8) to get full option list. 387 388sload.f2fs 389---------- 390The sload.f2fs gives a way to insert files and directories in the exisiting disk 391image. This tool is useful when building f2fs images given compiled files. 392 393Note: please refer to the manpage of sload.f2fs(8) to get full option list. 394 395resize.f2fs 396----------- 397The resize.f2fs lets a user resize the f2fs-formatted disk image, while preserving 398all the files and directories stored in the image. 399 400Note: please refer to the manpage of resize.f2fs(8) to get full option list. 401 402defrag.f2fs 403----------- 404The defrag.f2fs can be used to defragment scattered written data as well as 405filesystem metadata across the disk. This can improve the write speed by giving 406more free consecutive space. 407 408Note: please refer to the manpage of defrag.f2fs(8) to get full option list. 409 410f2fs_io 411------- 412The f2fs_io is a simple tool to issue various filesystem APIs as well as 413f2fs-specific ones, which is very useful for QA tests. 414 415Note: please refer to the manpage of f2fs_io(8) to get full option list. 416 417Design 418====== 419 420On-disk Layout 421-------------- 422 423F2FS divides the whole volume into a number of segments, each of which is fixed 424to 2MB in size. A section is composed of consecutive segments, and a zone 425consists of a set of sections. By default, section and zone sizes are set to one 426segment size identically, but users can easily modify the sizes by mkfs. 427 428F2FS splits the entire volume into six areas, and all the areas except superblock 429consist of multiple segments as described below:: 430 431 align with the zone size <-| 432 |-> align with the segment size 433 _________________________________________________________________________ 434 | | | Segment | Node | Segment | | 435 | Superblock | Checkpoint | Info. | Address | Summary | Main | 436 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | | 437 |____________|_____2______|______N______|______N______|______N_____|__N___| 438 . . 439 . . 440 . . 441 ._________________________________________. 442 |_Segment_|_..._|_Segment_|_..._|_Segment_| 443 . . 444 ._________._________ 445 |_section_|__...__|_ 446 . . 447 .________. 448 |__zone__| 449 450- Superblock (SB) 451 It is located at the beginning of the partition, and there exist two copies 452 to avoid file system crash. It contains basic partition information and some 453 default parameters of f2fs. 454 455- Checkpoint (CP) 456 It contains file system information, bitmaps for valid NAT/SIT sets, orphan 457 inode lists, and summary entries of current active segments. 458 459- Segment Information Table (SIT) 460 It contains segment information such as valid block count and bitmap for the 461 validity of all the blocks. 462 463- Node Address Table (NAT) 464 It is composed of a block address table for all the node blocks stored in 465 Main area. 466 467- Segment Summary Area (SSA) 468 It contains summary entries which contains the owner information of all the 469 data and node blocks stored in Main area. 470 471- Main Area 472 It contains file and directory data including their indices. 473 474In order to avoid misalignment between file system and flash-based storage, F2FS 475aligns the start block address of CP with the segment size. Also, it aligns the 476start block address of Main area with the zone size by reserving some segments 477in SSA area. 478 479Reference the following survey for additional technical details. 480https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey 481 482File System Metadata Structure 483------------------------------ 484 485F2FS adopts the checkpointing scheme to maintain file system consistency. At 486mount time, F2FS first tries to find the last valid checkpoint data by scanning 487CP area. In order to reduce the scanning time, F2FS uses only two copies of CP. 488One of them always indicates the last valid data, which is called as shadow copy 489mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism. 490 491For file system consistency, each CP points to which NAT and SIT copies are 492valid, as shown as below:: 493 494 +--------+----------+---------+ 495 | CP | SIT | NAT | 496 +--------+----------+---------+ 497 . . . . 498 . . . . 499 . . . . 500 +-------+-------+--------+--------+--------+--------+ 501 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 | 502 +-------+-------+--------+--------+--------+--------+ 503 | ^ ^ 504 | | | 505 `----------------------------------------' 506 507Index Structure 508--------------- 509 510The key data structure to manage the data locations is a "node". Similar to 511traditional file structures, F2FS has three types of node: inode, direct node, 512indirect node. F2FS assigns 4KB to an inode block which contains 923 data block 513indices, two direct node pointers, two indirect node pointers, and one double 514indirect node pointer as described below. One direct node block contains 1018 515data blocks, and one indirect node block contains also 1018 node blocks. Thus, 516one inode block (i.e., a file) covers:: 517 518 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB. 519 520 Inode block (4KB) 521 |- data (923) 522 |- direct node (2) 523 | `- data (1018) 524 |- indirect node (2) 525 | `- direct node (1018) 526 | `- data (1018) 527 `- double indirect node (1) 528 `- indirect node (1018) 529 `- direct node (1018) 530 `- data (1018) 531 532Note that all the node blocks are mapped by NAT which means the location of 533each node is translated by the NAT table. In the consideration of the wandering 534tree problem, F2FS is able to cut off the propagation of node updates caused by 535leaf data writes. 536 537Directory Structure 538------------------- 539 540A directory entry occupies 11 bytes, which consists of the following attributes. 541 542- hash hash value of the file name 543- ino inode number 544- len the length of file name 545- type file type such as directory, symlink, etc 546 547A dentry block consists of 214 dentry slots and file names. Therein a bitmap is 548used to represent whether each dentry is valid or not. A dentry block occupies 5494KB with the following composition. 550 551:: 552 553 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) + 554 dentries(11 * 214 bytes) + file name (8 * 214 bytes) 555 556 [Bucket] 557 +--------------------------------+ 558 |dentry block 1 | dentry block 2 | 559 +--------------------------------+ 560 . . 561 . . 562 . [Dentry Block Structure: 4KB] . 563 +--------+----------+----------+------------+ 564 | bitmap | reserved | dentries | file names | 565 +--------+----------+----------+------------+ 566 [Dentry Block: 4KB] . . 567 . . 568 . . 569 +------+------+-----+------+ 570 | hash | ino | len | type | 571 +------+------+-----+------+ 572 [Dentry Structure: 11 bytes] 573 574F2FS implements multi-level hash tables for directory structure. Each level has 575a hash table with dedicated number of hash buckets as shown below. Note that 576"A(2B)" means a bucket includes 2 data blocks. 577 578:: 579 580 ---------------------- 581 A : bucket 582 B : block 583 N : MAX_DIR_HASH_DEPTH 584 ---------------------- 585 586 level #0 | A(2B) 587 | 588 level #1 | A(2B) - A(2B) 589 | 590 level #2 | A(2B) - A(2B) - A(2B) - A(2B) 591 . | . . . . 592 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B) 593 . | . . . . 594 level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B) 595 596The number of blocks and buckets are determined by:: 597 598 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2, 599 # of blocks in level #n = | 600 `- 4, Otherwise 601 602 ,- 2^(n + dir_level), 603 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2, 604 # of buckets in level #n = | 605 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1), 606 Otherwise 607 608When F2FS finds a file name in a directory, at first a hash value of the file 609name is calculated. Then, F2FS scans the hash table in level #0 to find the 610dentry consisting of the file name and its inode number. If not found, F2FS 611scans the next hash table in level #1. In this way, F2FS scans hash tables in 612each levels incrementally from 1 to N. In each level F2FS needs to scan only 613one bucket determined by the following equation, which shows O(log(# of files)) 614complexity:: 615 616 bucket number to scan in level #n = (hash value) % (# of buckets in level #n) 617 618In the case of file creation, F2FS finds empty consecutive slots that cover the 619file name. F2FS searches the empty slots in the hash tables of whole levels from 6201 to N in the same way as the lookup operation. 621 622The following figure shows an example of two cases holding children:: 623 624 --------------> Dir <-------------- 625 | | 626 child child 627 628 child - child [hole] - child 629 630 child - child - child [hole] - [hole] - child 631 632 Case 1: Case 2: 633 Number of children = 6, Number of children = 3, 634 File size = 7 File size = 7 635 636Default Block Allocation 637------------------------ 638 639At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node 640and Hot/Warm/Cold data. 641 642- Hot node contains direct node blocks of directories. 643- Warm node contains direct node blocks except hot node blocks. 644- Cold node contains indirect node blocks 645- Hot data contains dentry blocks 646- Warm data contains data blocks except hot and cold data blocks 647- Cold data contains multimedia data or migrated data blocks 648 649LFS has two schemes for free space management: threaded log and copy-and-compac- 650tion. The copy-and-compaction scheme which is known as cleaning, is well-suited 651for devices showing very good sequential write performance, since free segments 652are served all the time for writing new data. However, it suffers from cleaning 653overhead under high utilization. Contrarily, the threaded log scheme suffers 654from random writes, but no cleaning process is needed. F2FS adopts a hybrid 655scheme where the copy-and-compaction scheme is adopted by default, but the 656policy is dynamically changed to the threaded log scheme according to the file 657system status. 658 659In order to align F2FS with underlying flash-based storage, F2FS allocates a 660segment in a unit of section. F2FS expects that the section size would be the 661same as the unit size of garbage collection in FTL. Furthermore, with respect 662to the mapping granularity in FTL, F2FS allocates each section of the active 663logs from different zones as much as possible, since FTL can write the data in 664the active logs into one allocation unit according to its mapping granularity. 665 666Cleaning process 667---------------- 668 669F2FS does cleaning both on demand and in the background. On-demand cleaning is 670triggered when there are not enough free segments to serve VFS calls. Background 671cleaner is operated by a kernel thread, and triggers the cleaning job when the 672system is idle. 673 674F2FS supports two victim selection policies: greedy and cost-benefit algorithms. 675In the greedy algorithm, F2FS selects a victim segment having the smallest number 676of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment 677according to the segment age and the number of valid blocks in order to address 678log block thrashing problem in the greedy algorithm. F2FS adopts the greedy 679algorithm for on-demand cleaner, while background cleaner adopts cost-benefit 680algorithm. 681 682In order to identify whether the data in the victim segment are valid or not, 683F2FS manages a bitmap. Each bit represents the validity of a block, and the 684bitmap is composed of a bit stream covering whole blocks in main area. 685 686Write-hint Policy 687----------------- 688 6891) whint_mode=off. F2FS only passes down WRITE_LIFE_NOT_SET. 690 6912) whint_mode=user-based. F2FS tries to pass down hints given by 692users. 693 694===================== ======================== =================== 695User F2FS Block 696===================== ======================== =================== 697N/A META WRITE_LIFE_NOT_SET 698N/A HOT_NODE " 699N/A WARM_NODE " 700N/A COLD_NODE " 701ioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 702extension list " " 703 704-- buffered io 705WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 706WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 707WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 708WRITE_LIFE_NONE " " 709WRITE_LIFE_MEDIUM " " 710WRITE_LIFE_LONG " " 711 712-- direct io 713WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 714WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 715WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 716WRITE_LIFE_NONE " WRITE_LIFE_NONE 717WRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 718WRITE_LIFE_LONG " WRITE_LIFE_LONG 719===================== ======================== =================== 720 7213) whint_mode=fs-based. F2FS passes down hints with its policy. 722 723===================== ======================== =================== 724User F2FS Block 725===================== ======================== =================== 726N/A META WRITE_LIFE_MEDIUM; 727N/A HOT_NODE WRITE_LIFE_NOT_SET 728N/A WARM_NODE " 729N/A COLD_NODE WRITE_LIFE_NONE 730ioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 731extension list " " 732 733-- buffered io 734WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 735WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 736WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_LONG 737WRITE_LIFE_NONE " " 738WRITE_LIFE_MEDIUM " " 739WRITE_LIFE_LONG " " 740 741-- direct io 742WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 743WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 744WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 745WRITE_LIFE_NONE " WRITE_LIFE_NONE 746WRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 747WRITE_LIFE_LONG " WRITE_LIFE_LONG 748===================== ======================== =================== 749 750Fallocate(2) Policy 751------------------- 752 753The default policy follows the below POSIX rule. 754 755Allocating disk space 756 The default operation (i.e., mode is zero) of fallocate() allocates 757 the disk space within the range specified by offset and len. The 758 file size (as reported by stat(2)) will be changed if offset+len is 759 greater than the file size. Any subregion within the range specified 760 by offset and len that did not contain data before the call will be 761 initialized to zero. This default behavior closely resembles the 762 behavior of the posix_fallocate(3) library function, and is intended 763 as a method of optimally implementing that function. 764 765However, once F2FS receives ioctl(fd, F2FS_IOC_SET_PIN_FILE) in prior to 766fallocate(fd, DEFAULT_MODE), it allocates on-disk block addressess having 767zero or random data, which is useful to the below scenario where: 768 769 1. create(fd) 770 2. ioctl(fd, F2FS_IOC_SET_PIN_FILE) 771 3. fallocate(fd, 0, 0, size) 772 4. address = fibmap(fd, offset) 773 5. open(blkdev) 774 6. write(blkdev, address) 775 776Compression implementation 777-------------------------- 778 779- New term named cluster is defined as basic unit of compression, file can 780 be divided into multiple clusters logically. One cluster includes 4 << n 781 (n >= 0) logical pages, compression size is also cluster size, each of 782 cluster can be compressed or not. 783 784- In cluster metadata layout, one special block address is used to indicate 785 a cluster is a compressed one or normal one; for compressed cluster, following 786 metadata maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs 787 stores data including compress header and compressed data. 788 789- In order to eliminate write amplification during overwrite, F2FS only 790 support compression on write-once file, data can be compressed only when 791 all logical blocks in cluster contain valid data and compress ratio of 792 cluster data is lower than specified threshold. 793 794- To enable compression on regular inode, there are three ways: 795 796 * chattr +c file 797 * chattr +c dir; touch dir/file 798 * mount w/ -o compress_extension=ext; touch file.ext 799 800Compress metadata layout:: 801 802 [Dnode Structure] 803 +-----------------------------------------------+ 804 | cluster 1 | cluster 2 | ......... | cluster N | 805 +-----------------------------------------------+ 806 . . . . 807 . . . . 808 . Compressed Cluster . . Normal Cluster . 809 +----------+---------+---------+---------+ +---------+---------+---------+---------+ 810 |compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 | 811 +----------+---------+---------+---------+ +---------+---------+---------+---------+ 812 . . 813 . . 814 . . 815 +-------------+-------------+----------+----------------------------+ 816 | data length | data chksum | reserved | compressed data | 817 +-------------+-------------+----------+----------------------------+ 818 819NVMe Zoned Namespace devices 820---------------------------- 821 822- ZNS defines a per-zone capacity which can be equal or less than the 823 zone-size. Zone-capacity is the number of usable blocks in the zone. 824 F2FS checks if zone-capacity is less than zone-size, if it is, then any 825 segment which starts after the zone-capacity is marked as not-free in 826 the free segment bitmap at initial mount time. These segments are marked 827 as permanently used so they are not allocated for writes and 828 consequently are not needed to be garbage collected. In case the 829 zone-capacity is not aligned to default segment size(2MB), then a segment 830 can start before the zone-capacity and span across zone-capacity boundary. 831 Such spanning segments are also considered as usable segments. All blocks 832 past the zone-capacity are considered unusable in these segments. 833