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 (obsolete) 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 FAULT_SLAB_ALLOC 0x000008000 200 FAULT_DQUOT_INIT 0x000010000 201 =================== =========== 202mode=%s Control block allocation mode which supports "adaptive" 203 and "lfs". In "lfs" mode, there should be no random 204 writes towards main area. 205 "fragment:segment" and "fragment:block" are newly added here. 206 These are developer options for experiments to simulate filesystem 207 fragmentation/after-GC situation itself. The developers use these 208 modes to understand filesystem fragmentation/after-GC condition well, 209 and eventually get some insights to handle them better. 210 In "fragment:segment", f2fs allocates a new segment in ramdom 211 position. With this, we can simulate the after-GC condition. 212 In "fragment:block", we can scatter block allocation with 213 "max_fragment_chunk" and "max_fragment_hole" sysfs nodes. 214 We added some randomness to both chunk and hole size to make 215 it close to realistic IO pattern. So, in this mode, f2fs will allocate 216 1..<max_fragment_chunk> blocks in a chunk and make a hole in the 217 length of 1..<max_fragment_hole> by turns. With this, the newly 218 allocated blocks will be scattered throughout the whole partition. 219 Note that "fragment:block" implicitly enables "fragment:segment" 220 option for more randomness. 221 Please, use these options for your experiments and we strongly 222 recommend to re-format the filesystem after using these options. 223io_bits=%u Set the bit size of write IO requests. It should be set 224 with "mode=lfs". 225usrquota Enable plain user disk quota accounting. 226grpquota Enable plain group disk quota accounting. 227prjquota Enable plain project quota accounting. 228usrjquota=<file> Appoint specified file and type during mount, so that quota 229grpjquota=<file> information can be properly updated during recovery flow, 230prjjquota=<file> <quota file>: must be in root directory; 231jqfmt=<quota type> <quota type>: [vfsold,vfsv0,vfsv1]. 232offusrjquota Turn off user journalled quota. 233offgrpjquota Turn off group journalled quota. 234offprjjquota Turn off project journalled quota. 235quota Enable plain user disk quota accounting. 236noquota Disable all plain disk quota option. 237whint_mode=%s Control which write hints are passed down to block 238 layer. This supports "off", "user-based", and 239 "fs-based". In "off" mode (default), f2fs does not pass 240 down hints. In "user-based" mode, f2fs tries to pass 241 down hints given by users. And in "fs-based" mode, f2fs 242 passes down hints with its policy. 243alloc_mode=%s Adjust block allocation policy, which supports "reuse" 244 and "default". 245fsync_mode=%s Control the policy of fsync. Currently supports "posix", 246 "strict", and "nobarrier". In "posix" mode, which is 247 default, fsync will follow POSIX semantics and does a 248 light operation to improve the filesystem performance. 249 In "strict" mode, fsync will be heavy and behaves in line 250 with xfs, ext4 and btrfs, where xfstest generic/342 will 251 pass, but the performance will regress. "nobarrier" is 252 based on "posix", but doesn't issue flush command for 253 non-atomic files likewise "nobarrier" mount option. 254test_dummy_encryption 255test_dummy_encryption=%s 256 Enable dummy encryption, which provides a fake fscrypt 257 context. The fake fscrypt context is used by xfstests. 258 The argument may be either "v1" or "v2", in order to 259 select the corresponding fscrypt policy version. 260checkpoint=%s[:%u[%]] Set to "disable" to turn off checkpointing. Set to "enable" 261 to reenable checkpointing. Is enabled by default. While 262 disabled, any unmounting or unexpected shutdowns will cause 263 the filesystem contents to appear as they did when the 264 filesystem was mounted with that option. 265 While mounting with checkpoint=disabled, the filesystem must 266 run garbage collection to ensure that all available space can 267 be used. If this takes too much time, the mount may return 268 EAGAIN. You may optionally add a value to indicate how much 269 of the disk you would be willing to temporarily give up to 270 avoid additional garbage collection. This can be given as a 271 number of blocks, or as a percent. For instance, mounting 272 with checkpoint=disable:100% would always succeed, but it may 273 hide up to all remaining free space. The actual space that 274 would be unusable can be viewed at /sys/fs/f2fs/<disk>/unusable 275 This space is reclaimed once checkpoint=enable. 276checkpoint_merge When checkpoint is enabled, this can be used to create a kernel 277 daemon and make it to merge concurrent checkpoint requests as 278 much as possible to eliminate redundant checkpoint issues. Plus, 279 we can eliminate the sluggish issue caused by slow checkpoint 280 operation when the checkpoint is done in a process context in 281 a cgroup having low i/o budget and cpu shares. To make this 282 do better, we set the default i/o priority of the kernel daemon 283 to "3", to give one higher priority than other kernel threads. 284 This is the same way to give a I/O priority to the jbd2 285 journaling thread of ext4 filesystem. 286nocheckpoint_merge Disable checkpoint merge feature. 287compress_algorithm=%s Control compress algorithm, currently f2fs supports "lzo", 288 "lz4", "zstd" and "lzo-rle" algorithm. 289compress_algorithm=%s:%d Control compress algorithm and its compress level, now, only 290 "lz4" and "zstd" support compress level config. 291 algorithm level range 292 lz4 3 - 16 293 zstd 1 - 22 294compress_log_size=%u Support configuring compress cluster size, the size will 295 be 4KB * (1 << %u), 16KB is minimum size, also it's 296 default size. 297compress_extension=%s Support adding specified extension, so that f2fs can enable 298 compression on those corresponding files, e.g. if all files 299 with '.ext' has high compression rate, we can set the '.ext' 300 on compression extension list and enable compression on 301 these file by default rather than to enable it via ioctl. 302 For other files, we can still enable compression via ioctl. 303 Note that, there is one reserved special extension '*', it 304 can be set to enable compression for all files. 305nocompress_extension=%s Support adding specified extension, so that f2fs can disable 306 compression on those corresponding files, just contrary to compression extension. 307 If you know exactly which files cannot be compressed, you can use this. 308 The same extension name can't appear in both compress and nocompress 309 extension at the same time. 310 If the compress extension specifies all files, the types specified by the 311 nocompress extension will be treated as special cases and will not be compressed. 312 Don't allow use '*' to specifie all file in nocompress extension. 313 After add nocompress_extension, the priority should be: 314 dir_flag < comp_extention,nocompress_extension < comp_file_flag,no_comp_file_flag. 315 See more in compression sections. 316 317compress_chksum Support verifying chksum of raw data in compressed cluster. 318compress_mode=%s Control file compression mode. This supports "fs" and "user" 319 modes. In "fs" mode (default), f2fs does automatic compression 320 on the compression enabled files. In "user" mode, f2fs disables 321 the automaic compression and gives the user discretion of 322 choosing the target file and the timing. The user can do manual 323 compression/decompression on the compression enabled files using 324 ioctls. 325compress_cache Support to use address space of a filesystem managed inode to 326 cache compressed block, in order to improve cache hit ratio of 327 random read. 328inlinecrypt When possible, encrypt/decrypt the contents of encrypted 329 files using the blk-crypto framework rather than 330 filesystem-layer encryption. This allows the use of 331 inline encryption hardware. The on-disk format is 332 unaffected. For more details, see 333 Documentation/block/inline-encryption.rst. 334atgc Enable age-threshold garbage collection, it provides high 335 effectiveness and efficiency on background GC. 336discard_unit=%s Control discard unit, the argument can be "block", "segment" 337 and "section", issued discard command's offset/size will be 338 aligned to the unit, by default, "discard_unit=block" is set, 339 so that small discard functionality is enabled. 340 For blkzoned device, "discard_unit=section" will be set by 341 default, it is helpful for large sized SMR or ZNS devices to 342 reduce memory cost by getting rid of fs metadata supports small 343 discard. 344======================== ============================================================ 345 346Debugfs Entries 347=============== 348 349/sys/kernel/debug/f2fs/ contains information about all the partitions mounted as 350f2fs. Each file shows the whole f2fs information. 351 352/sys/kernel/debug/f2fs/status includes: 353 354 - major file system information managed by f2fs currently 355 - average SIT information about whole segments 356 - current memory footprint consumed by f2fs. 357 358Sysfs Entries 359============= 360 361Information about mounted f2fs file systems can be found in 362/sys/fs/f2fs. Each mounted filesystem will have a directory in 363/sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda). 364The files in each per-device directory are shown in table below. 365 366Files in /sys/fs/f2fs/<devname> 367(see also Documentation/ABI/testing/sysfs-fs-f2fs) 368 369Usage 370===== 371 3721. Download userland tools and compile them. 373 3742. Skip, if f2fs was compiled statically inside kernel. 375 Otherwise, insert the f2fs.ko module:: 376 377 # insmod f2fs.ko 378 3793. Create a directory to use when mounting:: 380 381 # mkdir /mnt/f2fs 382 3834. Format the block device, and then mount as f2fs:: 384 385 # mkfs.f2fs -l label /dev/block_device 386 # mount -t f2fs /dev/block_device /mnt/f2fs 387 388mkfs.f2fs 389--------- 390The mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem, 391which builds a basic on-disk layout. 392 393The quick options consist of: 394 395=============== =========================================================== 396``-l [label]`` Give a volume label, up to 512 unicode name. 397``-a [0 or 1]`` Split start location of each area for heap-based allocation. 398 399 1 is set by default, which performs this. 400``-o [int]`` Set overprovision ratio in percent over volume size. 401 402 5 is set by default. 403``-s [int]`` Set the number of segments per section. 404 405 1 is set by default. 406``-z [int]`` Set the number of sections per zone. 407 408 1 is set by default. 409``-e [str]`` Set basic extension list. e.g. "mp3,gif,mov" 410``-t [0 or 1]`` Disable discard command or not. 411 412 1 is set by default, which conducts discard. 413=============== =========================================================== 414 415Note: please refer to the manpage of mkfs.f2fs(8) to get full option list. 416 417fsck.f2fs 418--------- 419The fsck.f2fs is a tool to check the consistency of an f2fs-formatted 420partition, which examines whether the filesystem metadata and user-made data 421are cross-referenced correctly or not. 422Note that, initial version of the tool does not fix any inconsistency. 423 424The quick options consist of:: 425 426 -d debug level [default:0] 427 428Note: please refer to the manpage of fsck.f2fs(8) to get full option list. 429 430dump.f2fs 431--------- 432The dump.f2fs shows the information of specific inode and dumps SSA and SIT to 433file. Each file is dump_ssa and dump_sit. 434 435The dump.f2fs is used to debug on-disk data structures of the f2fs filesystem. 436It shows on-disk inode information recognized by a given inode number, and is 437able to dump all the SSA and SIT entries into predefined files, ./dump_ssa and 438./dump_sit respectively. 439 440The options consist of:: 441 442 -d debug level [default:0] 443 -i inode no (hex) 444 -s [SIT dump segno from #1~#2 (decimal), for all 0~-1] 445 -a [SSA dump segno from #1~#2 (decimal), for all 0~-1] 446 447Examples:: 448 449 # dump.f2fs -i [ino] /dev/sdx 450 # dump.f2fs -s 0~-1 /dev/sdx (SIT dump) 451 # dump.f2fs -a 0~-1 /dev/sdx (SSA dump) 452 453Note: please refer to the manpage of dump.f2fs(8) to get full option list. 454 455sload.f2fs 456---------- 457The sload.f2fs gives a way to insert files and directories in the exisiting disk 458image. This tool is useful when building f2fs images given compiled files. 459 460Note: please refer to the manpage of sload.f2fs(8) to get full option list. 461 462resize.f2fs 463----------- 464The resize.f2fs lets a user resize the f2fs-formatted disk image, while preserving 465all the files and directories stored in the image. 466 467Note: please refer to the manpage of resize.f2fs(8) to get full option list. 468 469defrag.f2fs 470----------- 471The defrag.f2fs can be used to defragment scattered written data as well as 472filesystem metadata across the disk. This can improve the write speed by giving 473more free consecutive space. 474 475Note: please refer to the manpage of defrag.f2fs(8) to get full option list. 476 477f2fs_io 478------- 479The f2fs_io is a simple tool to issue various filesystem APIs as well as 480f2fs-specific ones, which is very useful for QA tests. 481 482Note: please refer to the manpage of f2fs_io(8) to get full option list. 483 484Design 485====== 486 487On-disk Layout 488-------------- 489 490F2FS divides the whole volume into a number of segments, each of which is fixed 491to 2MB in size. A section is composed of consecutive segments, and a zone 492consists of a set of sections. By default, section and zone sizes are set to one 493segment size identically, but users can easily modify the sizes by mkfs. 494 495F2FS splits the entire volume into six areas, and all the areas except superblock 496consist of multiple segments as described below:: 497 498 align with the zone size <-| 499 |-> align with the segment size 500 _________________________________________________________________________ 501 | | | Segment | Node | Segment | | 502 | Superblock | Checkpoint | Info. | Address | Summary | Main | 503 | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | | 504 |____________|_____2______|______N______|______N______|______N_____|__N___| 505 . . 506 . . 507 . . 508 ._________________________________________. 509 |_Segment_|_..._|_Segment_|_..._|_Segment_| 510 . . 511 ._________._________ 512 |_section_|__...__|_ 513 . . 514 .________. 515 |__zone__| 516 517- Superblock (SB) 518 It is located at the beginning of the partition, and there exist two copies 519 to avoid file system crash. It contains basic partition information and some 520 default parameters of f2fs. 521 522- Checkpoint (CP) 523 It contains file system information, bitmaps for valid NAT/SIT sets, orphan 524 inode lists, and summary entries of current active segments. 525 526- Segment Information Table (SIT) 527 It contains segment information such as valid block count and bitmap for the 528 validity of all the blocks. 529 530- Node Address Table (NAT) 531 It is composed of a block address table for all the node blocks stored in 532 Main area. 533 534- Segment Summary Area (SSA) 535 It contains summary entries which contains the owner information of all the 536 data and node blocks stored in Main area. 537 538- Main Area 539 It contains file and directory data including their indices. 540 541In order to avoid misalignment between file system and flash-based storage, F2FS 542aligns the start block address of CP with the segment size. Also, it aligns the 543start block address of Main area with the zone size by reserving some segments 544in SSA area. 545 546Reference the following survey for additional technical details. 547https://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey 548 549File System Metadata Structure 550------------------------------ 551 552F2FS adopts the checkpointing scheme to maintain file system consistency. At 553mount time, F2FS first tries to find the last valid checkpoint data by scanning 554CP area. In order to reduce the scanning time, F2FS uses only two copies of CP. 555One of them always indicates the last valid data, which is called as shadow copy 556mechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism. 557 558For file system consistency, each CP points to which NAT and SIT copies are 559valid, as shown as below:: 560 561 +--------+----------+---------+ 562 | CP | SIT | NAT | 563 +--------+----------+---------+ 564 . . . . 565 . . . . 566 . . . . 567 +-------+-------+--------+--------+--------+--------+ 568 | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 | 569 +-------+-------+--------+--------+--------+--------+ 570 | ^ ^ 571 | | | 572 `----------------------------------------' 573 574Index Structure 575--------------- 576 577The key data structure to manage the data locations is a "node". Similar to 578traditional file structures, F2FS has three types of node: inode, direct node, 579indirect node. F2FS assigns 4KB to an inode block which contains 923 data block 580indices, two direct node pointers, two indirect node pointers, and one double 581indirect node pointer as described below. One direct node block contains 1018 582data blocks, and one indirect node block contains also 1018 node blocks. Thus, 583one inode block (i.e., a file) covers:: 584 585 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB. 586 587 Inode block (4KB) 588 |- data (923) 589 |- direct node (2) 590 | `- data (1018) 591 |- indirect node (2) 592 | `- direct node (1018) 593 | `- data (1018) 594 `- double indirect node (1) 595 `- indirect node (1018) 596 `- direct node (1018) 597 `- data (1018) 598 599Note that all the node blocks are mapped by NAT which means the location of 600each node is translated by the NAT table. In the consideration of the wandering 601tree problem, F2FS is able to cut off the propagation of node updates caused by 602leaf data writes. 603 604Directory Structure 605------------------- 606 607A directory entry occupies 11 bytes, which consists of the following attributes. 608 609- hash hash value of the file name 610- ino inode number 611- len the length of file name 612- type file type such as directory, symlink, etc 613 614A dentry block consists of 214 dentry slots and file names. Therein a bitmap is 615used to represent whether each dentry is valid or not. A dentry block occupies 6164KB with the following composition. 617 618:: 619 620 Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) + 621 dentries(11 * 214 bytes) + file name (8 * 214 bytes) 622 623 [Bucket] 624 +--------------------------------+ 625 |dentry block 1 | dentry block 2 | 626 +--------------------------------+ 627 . . 628 . . 629 . [Dentry Block Structure: 4KB] . 630 +--------+----------+----------+------------+ 631 | bitmap | reserved | dentries | file names | 632 +--------+----------+----------+------------+ 633 [Dentry Block: 4KB] . . 634 . . 635 . . 636 +------+------+-----+------+ 637 | hash | ino | len | type | 638 +------+------+-----+------+ 639 [Dentry Structure: 11 bytes] 640 641F2FS implements multi-level hash tables for directory structure. Each level has 642a hash table with dedicated number of hash buckets as shown below. Note that 643"A(2B)" means a bucket includes 2 data blocks. 644 645:: 646 647 ---------------------- 648 A : bucket 649 B : block 650 N : MAX_DIR_HASH_DEPTH 651 ---------------------- 652 653 level #0 | A(2B) 654 | 655 level #1 | A(2B) - A(2B) 656 | 657 level #2 | A(2B) - A(2B) - A(2B) - A(2B) 658 . | . . . . 659 level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B) 660 . | . . . . 661 level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B) 662 663The number of blocks and buckets are determined by:: 664 665 ,- 2, if n < MAX_DIR_HASH_DEPTH / 2, 666 # of blocks in level #n = | 667 `- 4, Otherwise 668 669 ,- 2^(n + dir_level), 670 | if n + dir_level < MAX_DIR_HASH_DEPTH / 2, 671 # of buckets in level #n = | 672 `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1), 673 Otherwise 674 675When F2FS finds a file name in a directory, at first a hash value of the file 676name is calculated. Then, F2FS scans the hash table in level #0 to find the 677dentry consisting of the file name and its inode number. If not found, F2FS 678scans the next hash table in level #1. In this way, F2FS scans hash tables in 679each levels incrementally from 1 to N. In each level F2FS needs to scan only 680one bucket determined by the following equation, which shows O(log(# of files)) 681complexity:: 682 683 bucket number to scan in level #n = (hash value) % (# of buckets in level #n) 684 685In the case of file creation, F2FS finds empty consecutive slots that cover the 686file name. F2FS searches the empty slots in the hash tables of whole levels from 6871 to N in the same way as the lookup operation. 688 689The following figure shows an example of two cases holding children:: 690 691 --------------> Dir <-------------- 692 | | 693 child child 694 695 child - child [hole] - child 696 697 child - child - child [hole] - [hole] - child 698 699 Case 1: Case 2: 700 Number of children = 6, Number of children = 3, 701 File size = 7 File size = 7 702 703Default Block Allocation 704------------------------ 705 706At runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node 707and Hot/Warm/Cold data. 708 709- Hot node contains direct node blocks of directories. 710- Warm node contains direct node blocks except hot node blocks. 711- Cold node contains indirect node blocks 712- Hot data contains dentry blocks 713- Warm data contains data blocks except hot and cold data blocks 714- Cold data contains multimedia data or migrated data blocks 715 716LFS has two schemes for free space management: threaded log and copy-and-compac- 717tion. The copy-and-compaction scheme which is known as cleaning, is well-suited 718for devices showing very good sequential write performance, since free segments 719are served all the time for writing new data. However, it suffers from cleaning 720overhead under high utilization. Contrarily, the threaded log scheme suffers 721from random writes, but no cleaning process is needed. F2FS adopts a hybrid 722scheme where the copy-and-compaction scheme is adopted by default, but the 723policy is dynamically changed to the threaded log scheme according to the file 724system status. 725 726In order to align F2FS with underlying flash-based storage, F2FS allocates a 727segment in a unit of section. F2FS expects that the section size would be the 728same as the unit size of garbage collection in FTL. Furthermore, with respect 729to the mapping granularity in FTL, F2FS allocates each section of the active 730logs from different zones as much as possible, since FTL can write the data in 731the active logs into one allocation unit according to its mapping granularity. 732 733Cleaning process 734---------------- 735 736F2FS does cleaning both on demand and in the background. On-demand cleaning is 737triggered when there are not enough free segments to serve VFS calls. Background 738cleaner is operated by a kernel thread, and triggers the cleaning job when the 739system is idle. 740 741F2FS supports two victim selection policies: greedy and cost-benefit algorithms. 742In the greedy algorithm, F2FS selects a victim segment having the smallest number 743of valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment 744according to the segment age and the number of valid blocks in order to address 745log block thrashing problem in the greedy algorithm. F2FS adopts the greedy 746algorithm for on-demand cleaner, while background cleaner adopts cost-benefit 747algorithm. 748 749In order to identify whether the data in the victim segment are valid or not, 750F2FS manages a bitmap. Each bit represents the validity of a block, and the 751bitmap is composed of a bit stream covering whole blocks in main area. 752 753Write-hint Policy 754----------------- 755 7561) whint_mode=off. F2FS only passes down WRITE_LIFE_NOT_SET. 757 7582) whint_mode=user-based. F2FS tries to pass down hints given by 759users. 760 761===================== ======================== =================== 762User F2FS Block 763===================== ======================== =================== 764N/A META WRITE_LIFE_NOT_SET 765N/A HOT_NODE " 766N/A WARM_NODE " 767N/A COLD_NODE " 768ioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 769extension list " " 770 771-- buffered io 772WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 773WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 774WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 775WRITE_LIFE_NONE " " 776WRITE_LIFE_MEDIUM " " 777WRITE_LIFE_LONG " " 778 779-- direct io 780WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 781WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 782WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 783WRITE_LIFE_NONE " WRITE_LIFE_NONE 784WRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 785WRITE_LIFE_LONG " WRITE_LIFE_LONG 786===================== ======================== =================== 787 7883) whint_mode=fs-based. F2FS passes down hints with its policy. 789 790===================== ======================== =================== 791User F2FS Block 792===================== ======================== =================== 793N/A META WRITE_LIFE_MEDIUM; 794N/A HOT_NODE WRITE_LIFE_NOT_SET 795N/A WARM_NODE " 796N/A COLD_NODE WRITE_LIFE_NONE 797ioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 798extension list " " 799 800-- buffered io 801WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 802WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 803WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_LONG 804WRITE_LIFE_NONE " " 805WRITE_LIFE_MEDIUM " " 806WRITE_LIFE_LONG " " 807 808-- direct io 809WRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 810WRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 811WRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 812WRITE_LIFE_NONE " WRITE_LIFE_NONE 813WRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 814WRITE_LIFE_LONG " WRITE_LIFE_LONG 815===================== ======================== =================== 816 817Fallocate(2) Policy 818------------------- 819 820The default policy follows the below POSIX rule. 821 822Allocating disk space 823 The default operation (i.e., mode is zero) of fallocate() allocates 824 the disk space within the range specified by offset and len. The 825 file size (as reported by stat(2)) will be changed if offset+len is 826 greater than the file size. Any subregion within the range specified 827 by offset and len that did not contain data before the call will be 828 initialized to zero. This default behavior closely resembles the 829 behavior of the posix_fallocate(3) library function, and is intended 830 as a method of optimally implementing that function. 831 832However, once F2FS receives ioctl(fd, F2FS_IOC_SET_PIN_FILE) in prior to 833fallocate(fd, DEFAULT_MODE), it allocates on-disk block addressess having 834zero or random data, which is useful to the below scenario where: 835 836 1. create(fd) 837 2. ioctl(fd, F2FS_IOC_SET_PIN_FILE) 838 3. fallocate(fd, 0, 0, size) 839 4. address = fibmap(fd, offset) 840 5. open(blkdev) 841 6. write(blkdev, address) 842 843Compression implementation 844-------------------------- 845 846- New term named cluster is defined as basic unit of compression, file can 847 be divided into multiple clusters logically. One cluster includes 4 << n 848 (n >= 0) logical pages, compression size is also cluster size, each of 849 cluster can be compressed or not. 850 851- In cluster metadata layout, one special block address is used to indicate 852 a cluster is a compressed one or normal one; for compressed cluster, following 853 metadata maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs 854 stores data including compress header and compressed data. 855 856- In order to eliminate write amplification during overwrite, F2FS only 857 support compression on write-once file, data can be compressed only when 858 all logical blocks in cluster contain valid data and compress ratio of 859 cluster data is lower than specified threshold. 860 861- To enable compression on regular inode, there are four ways: 862 863 * chattr +c file 864 * chattr +c dir; touch dir/file 865 * mount w/ -o compress_extension=ext; touch file.ext 866 * mount w/ -o compress_extension=*; touch any_file 867 868- To disable compression on regular inode, there are two ways: 869 870 * chattr -c file 871 * mount w/ -o nocompress_extension=ext; touch file.ext 872 873- Priority in between FS_COMPR_FL, FS_NOCOMP_FS, extensions: 874 875 * compress_extension=so; nocompress_extension=zip; chattr +c dir; touch 876 dir/foo.so; touch dir/bar.zip; touch dir/baz.txt; then foo.so and baz.txt 877 should be compresse, bar.zip should be non-compressed. chattr +c dir/bar.zip 878 can enable compress on bar.zip. 879 * compress_extension=so; nocompress_extension=zip; chattr -c dir; touch 880 dir/foo.so; touch dir/bar.zip; touch dir/baz.txt; then foo.so should be 881 compresse, bar.zip and baz.txt should be non-compressed. 882 chattr+c dir/bar.zip; chattr+c dir/baz.txt; can enable compress on bar.zip 883 and baz.txt. 884 885- At this point, compression feature doesn't expose compressed space to user 886 directly in order to guarantee potential data updates later to the space. 887 Instead, the main goal is to reduce data writes to flash disk as much as 888 possible, resulting in extending disk life time as well as relaxing IO 889 congestion. Alternatively, we've added ioctl(F2FS_IOC_RELEASE_COMPRESS_BLOCKS) 890 interface to reclaim compressed space and show it to user after putting the 891 immutable bit. Immutable bit, after release, it doesn't allow writing/mmaping 892 on the file, until reserving compressed space via 893 ioctl(F2FS_IOC_RESERVE_COMPRESS_BLOCKS) or truncating filesize to zero. 894 895Compress metadata layout:: 896 897 [Dnode Structure] 898 +-----------------------------------------------+ 899 | cluster 1 | cluster 2 | ......... | cluster N | 900 +-----------------------------------------------+ 901 . . . . 902 . . . . 903 . Compressed Cluster . . Normal Cluster . 904 +----------+---------+---------+---------+ +---------+---------+---------+---------+ 905 |compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 | 906 +----------+---------+---------+---------+ +---------+---------+---------+---------+ 907 . . 908 . . 909 . . 910 +-------------+-------------+----------+----------------------------+ 911 | data length | data chksum | reserved | compressed data | 912 +-------------+-------------+----------+----------------------------+ 913 914Compression mode 915-------------------------- 916 917f2fs supports "fs" and "user" compression modes with "compression_mode" mount option. 918With this option, f2fs provides a choice to select the way how to compress the 919compression enabled files (refer to "Compression implementation" section for how to 920enable compression on a regular inode). 921 9221) compress_mode=fs 923This is the default option. f2fs does automatic compression in the writeback of the 924compression enabled files. 925 9262) compress_mode=user 927This disables the automatic compression and gives the user discretion of choosing the 928target file and the timing. The user can do manual compression/decompression on the 929compression enabled files using F2FS_IOC_DECOMPRESS_FILE and F2FS_IOC_COMPRESS_FILE 930ioctls like the below. 931 932To decompress a file, 933 934fd = open(filename, O_WRONLY, 0); 935ret = ioctl(fd, F2FS_IOC_DECOMPRESS_FILE); 936 937To compress a file, 938 939fd = open(filename, O_WRONLY, 0); 940ret = ioctl(fd, F2FS_IOC_COMPRESS_FILE); 941 942NVMe Zoned Namespace devices 943---------------------------- 944 945- ZNS defines a per-zone capacity which can be equal or less than the 946 zone-size. Zone-capacity is the number of usable blocks in the zone. 947 F2FS checks if zone-capacity is less than zone-size, if it is, then any 948 segment which starts after the zone-capacity is marked as not-free in 949 the free segment bitmap at initial mount time. These segments are marked 950 as permanently used so they are not allocated for writes and 951 consequently are not needed to be garbage collected. In case the 952 zone-capacity is not aligned to default segment size(2MB), then a segment 953 can start before the zone-capacity and span across zone-capacity boundary. 954 Such spanning segments are also considered as usable segments. All blocks 955 past the zone-capacity are considered unusable in these segments. 956