189272ca1SMauro Carvalho Chehab.. SPDX-License-Identifier: GPL-2.0 289272ca1SMauro Carvalho Chehab 389272ca1SMauro Carvalho Chehab========================================== 489272ca1SMauro Carvalho ChehabWHAT IS Flash-Friendly File System (F2FS)? 589272ca1SMauro Carvalho Chehab========================================== 689272ca1SMauro Carvalho Chehab 789272ca1SMauro Carvalho ChehabNAND flash memory-based storage devices, such as SSD, eMMC, and SD cards, have 889272ca1SMauro Carvalho Chehabbeen equipped on a variety systems ranging from mobile to server systems. Since 989272ca1SMauro Carvalho Chehabthey are known to have different characteristics from the conventional rotating 1089272ca1SMauro Carvalho Chehabdisks, a file system, an upper layer to the storage device, should adapt to the 1189272ca1SMauro Carvalho Chehabchanges from the sketch in the design level. 1289272ca1SMauro Carvalho Chehab 1389272ca1SMauro Carvalho ChehabF2FS is a file system exploiting NAND flash memory-based storage devices, which 1489272ca1SMauro Carvalho Chehabis based on Log-structured File System (LFS). The design has been focused on 1589272ca1SMauro Carvalho Chehabaddressing the fundamental issues in LFS, which are snowball effect of wandering 1689272ca1SMauro Carvalho Chehabtree and high cleaning overhead. 1789272ca1SMauro Carvalho Chehab 1889272ca1SMauro Carvalho ChehabSince a NAND flash memory-based storage device shows different characteristic 1989272ca1SMauro Carvalho Chehabaccording to its internal geometry or flash memory management scheme, namely FTL, 2089272ca1SMauro Carvalho ChehabF2FS and its tools support various parameters not only for configuring on-disk 2189272ca1SMauro Carvalho Chehablayout, but also for selecting allocation and cleaning algorithms. 2289272ca1SMauro Carvalho Chehab 2389272ca1SMauro Carvalho ChehabThe following git tree provides the file system formatting tool (mkfs.f2fs), 2489272ca1SMauro Carvalho Chehaba consistency checking tool (fsck.f2fs), and a debugging tool (dump.f2fs). 2589272ca1SMauro Carvalho Chehab 2689272ca1SMauro Carvalho Chehab- git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git 2789272ca1SMauro Carvalho Chehab 2889272ca1SMauro Carvalho ChehabFor reporting bugs and sending patches, please use the following mailing list: 2989272ca1SMauro Carvalho Chehab 3089272ca1SMauro Carvalho Chehab- linux-f2fs-devel@lists.sourceforge.net 3189272ca1SMauro Carvalho Chehab 3289272ca1SMauro Carvalho ChehabBackground and Design issues 3389272ca1SMauro Carvalho Chehab============================ 3489272ca1SMauro Carvalho Chehab 3589272ca1SMauro Carvalho ChehabLog-structured File System (LFS) 3689272ca1SMauro Carvalho Chehab-------------------------------- 3789272ca1SMauro Carvalho Chehab"A log-structured file system writes all modifications to disk sequentially in 3889272ca1SMauro Carvalho Chehaba log-like structure, thereby speeding up both file writing and crash recovery. 3989272ca1SMauro Carvalho ChehabThe log is the only structure on disk; it contains indexing information so that 4089272ca1SMauro Carvalho Chehabfiles can be read back from the log efficiently. In order to maintain large free 4189272ca1SMauro Carvalho Chehabareas on disk for fast writing, we divide the log into segments and use a 4289272ca1SMauro Carvalho Chehabsegment cleaner to compress the live information from heavily fragmented 4389272ca1SMauro Carvalho Chehabsegments." from Rosenblum, M. and Ousterhout, J. K., 1992, "The design and 4489272ca1SMauro Carvalho Chehabimplementation of a log-structured file system", ACM Trans. Computer Systems 4589272ca1SMauro Carvalho Chehab10, 1, 26–52. 4689272ca1SMauro Carvalho Chehab 4789272ca1SMauro Carvalho ChehabWandering Tree Problem 4889272ca1SMauro Carvalho Chehab---------------------- 4989272ca1SMauro Carvalho ChehabIn LFS, when a file data is updated and written to the end of log, its direct 5089272ca1SMauro Carvalho Chehabpointer block is updated due to the changed location. Then the indirect pointer 5189272ca1SMauro Carvalho Chehabblock is also updated due to the direct pointer block update. In this manner, 5289272ca1SMauro Carvalho Chehabthe upper index structures such as inode, inode map, and checkpoint block are 5389272ca1SMauro Carvalho Chehabalso updated recursively. This problem is called as wandering tree problem [1], 5489272ca1SMauro Carvalho Chehaband in order to enhance the performance, it should eliminate or relax the update 5589272ca1SMauro Carvalho Chehabpropagation as much as possible. 5689272ca1SMauro Carvalho Chehab 5789272ca1SMauro Carvalho Chehab[1] Bityutskiy, A. 2005. JFFS3 design issues. http://www.linux-mtd.infradead.org/ 5889272ca1SMauro Carvalho Chehab 5989272ca1SMauro Carvalho ChehabCleaning Overhead 6089272ca1SMauro Carvalho Chehab----------------- 6189272ca1SMauro Carvalho ChehabSince LFS is based on out-of-place writes, it produces so many obsolete blocks 6289272ca1SMauro Carvalho Chehabscattered across the whole storage. In order to serve new empty log space, it 6389272ca1SMauro Carvalho Chehabneeds to reclaim these obsolete blocks seamlessly to users. This job is called 6489272ca1SMauro Carvalho Chehabas a cleaning process. 6589272ca1SMauro Carvalho Chehab 6689272ca1SMauro Carvalho ChehabThe process consists of three operations as follows. 6789272ca1SMauro Carvalho Chehab 6889272ca1SMauro Carvalho Chehab1. A victim segment is selected through referencing segment usage table. 6989272ca1SMauro Carvalho Chehab2. It loads parent index structures of all the data in the victim identified by 7089272ca1SMauro Carvalho Chehab segment summary blocks. 7189272ca1SMauro Carvalho Chehab3. It checks the cross-reference between the data and its parent index structure. 7289272ca1SMauro Carvalho Chehab4. It moves valid data selectively. 7389272ca1SMauro Carvalho Chehab 7489272ca1SMauro Carvalho ChehabThis cleaning job may cause unexpected long delays, so the most important goal 7589272ca1SMauro Carvalho Chehabis to hide the latencies to users. And also definitely, it should reduce the 7689272ca1SMauro Carvalho Chehabamount of valid data to be moved, and move them quickly as well. 7789272ca1SMauro Carvalho Chehab 7889272ca1SMauro Carvalho ChehabKey Features 7989272ca1SMauro Carvalho Chehab============ 8089272ca1SMauro Carvalho Chehab 8189272ca1SMauro Carvalho ChehabFlash Awareness 8289272ca1SMauro Carvalho Chehab--------------- 8389272ca1SMauro Carvalho Chehab- Enlarge the random write area for better performance, but provide the high 8489272ca1SMauro Carvalho Chehab spatial locality 8589272ca1SMauro Carvalho Chehab- Align FS data structures to the operational units in FTL as best efforts 8689272ca1SMauro Carvalho Chehab 8789272ca1SMauro Carvalho ChehabWandering Tree Problem 8889272ca1SMauro Carvalho Chehab---------------------- 8989272ca1SMauro Carvalho Chehab- Use a term, “node”, that represents inodes as well as various pointer blocks 9089272ca1SMauro Carvalho Chehab- Introduce Node Address Table (NAT) containing the locations of all the “node” 9189272ca1SMauro Carvalho Chehab blocks; this will cut off the update propagation. 9289272ca1SMauro Carvalho Chehab 9389272ca1SMauro Carvalho ChehabCleaning Overhead 9489272ca1SMauro Carvalho Chehab----------------- 9589272ca1SMauro Carvalho Chehab- Support a background cleaning process 9689272ca1SMauro Carvalho Chehab- Support greedy and cost-benefit algorithms for victim selection policies 9789272ca1SMauro Carvalho Chehab- Support multi-head logs for static/dynamic hot and cold data separation 9889272ca1SMauro Carvalho Chehab- Introduce adaptive logging for efficient block allocation 9989272ca1SMauro Carvalho Chehab 10089272ca1SMauro Carvalho ChehabMount Options 10189272ca1SMauro Carvalho Chehab============= 10289272ca1SMauro Carvalho Chehab 10389272ca1SMauro Carvalho Chehab 10489272ca1SMauro Carvalho Chehab====================== ============================================================ 10589272ca1SMauro Carvalho Chehabbackground_gc=%s Turn on/off cleaning operations, namely garbage 10689272ca1SMauro Carvalho Chehab collection, triggered in background when I/O subsystem is 10789272ca1SMauro Carvalho Chehab idle. If background_gc=on, it will turn on the garbage 10889272ca1SMauro Carvalho Chehab collection and if background_gc=off, garbage collection 10989272ca1SMauro Carvalho Chehab will be turned off. If background_gc=sync, it will turn 11089272ca1SMauro Carvalho Chehab on synchronous garbage collection running in background. 11189272ca1SMauro Carvalho Chehab Default value for this option is on. So garbage 11289272ca1SMauro Carvalho Chehab collection is on by default. 11389272ca1SMauro Carvalho Chehabdisable_roll_forward Disable the roll-forward recovery routine 11489272ca1SMauro Carvalho Chehabnorecovery Disable the roll-forward recovery routine, mounted read- 11589272ca1SMauro Carvalho Chehab only (i.e., -o ro,disable_roll_forward) 11689272ca1SMauro Carvalho Chehabdiscard/nodiscard Enable/disable real-time discard in f2fs, if discard is 11789272ca1SMauro Carvalho Chehab enabled, f2fs will issue discard/TRIM commands when a 11889272ca1SMauro Carvalho Chehab segment is cleaned. 11989272ca1SMauro Carvalho Chehabno_heap Disable heap-style segment allocation which finds free 12089272ca1SMauro Carvalho Chehab segments for data from the beginning of main area, while 12189272ca1SMauro Carvalho Chehab for node from the end of main area. 12289272ca1SMauro Carvalho Chehabnouser_xattr Disable Extended User Attributes. Note: xattr is enabled 12389272ca1SMauro Carvalho Chehab by default if CONFIG_F2FS_FS_XATTR is selected. 12489272ca1SMauro Carvalho Chehabnoacl Disable POSIX Access Control List. Note: acl is enabled 12589272ca1SMauro Carvalho Chehab by default if CONFIG_F2FS_FS_POSIX_ACL is selected. 12689272ca1SMauro Carvalho Chehabactive_logs=%u Support configuring the number of active logs. In the 12789272ca1SMauro Carvalho Chehab current design, f2fs supports only 2, 4, and 6 logs. 12889272ca1SMauro Carvalho Chehab Default number is 6. 12989272ca1SMauro Carvalho Chehabdisable_ext_identify Disable the extension list configured by mkfs, so f2fs 13089272ca1SMauro Carvalho Chehab does not aware of cold files such as media files. 13189272ca1SMauro Carvalho Chehabinline_xattr Enable the inline xattrs feature. 13289272ca1SMauro Carvalho Chehabnoinline_xattr Disable the inline xattrs feature. 13389272ca1SMauro Carvalho Chehabinline_xattr_size=%u Support configuring inline xattr size, it depends on 13489272ca1SMauro Carvalho Chehab flexible inline xattr feature. 13589272ca1SMauro Carvalho Chehabinline_data Enable the inline data feature: New created small(<~3.4k) 13689272ca1SMauro Carvalho Chehab files can be written into inode block. 13789272ca1SMauro Carvalho Chehabinline_dentry Enable the inline dir feature: data in new created 13889272ca1SMauro Carvalho Chehab directory entries can be written into inode block. The 13989272ca1SMauro Carvalho Chehab space of inode block which is used to store inline 14089272ca1SMauro Carvalho Chehab dentries is limited to ~3.4k. 14189272ca1SMauro Carvalho Chehabnoinline_dentry Disable the inline dentry feature. 14289272ca1SMauro Carvalho Chehabflush_merge Merge concurrent cache_flush commands as much as possible 14389272ca1SMauro Carvalho Chehab to eliminate redundant command issues. If the underlying 14489272ca1SMauro Carvalho Chehab device handles the cache_flush command relatively slowly, 14589272ca1SMauro Carvalho Chehab recommend to enable this option. 14689272ca1SMauro Carvalho Chehabnobarrier This option can be used if underlying storage guarantees 14789272ca1SMauro Carvalho Chehab its cached data should be written to the novolatile area. 14889272ca1SMauro Carvalho Chehab If this option is set, no cache_flush commands are issued 14989272ca1SMauro Carvalho Chehab but f2fs still guarantees the write ordering of all the 15089272ca1SMauro Carvalho Chehab data writes. 15189272ca1SMauro Carvalho Chehabfastboot This option is used when a system wants to reduce mount 15289272ca1SMauro Carvalho Chehab time as much as possible, even though normal performance 15389272ca1SMauro Carvalho Chehab can be sacrificed. 15489272ca1SMauro Carvalho Chehabextent_cache Enable an extent cache based on rb-tree, it can cache 15589272ca1SMauro Carvalho Chehab as many as extent which map between contiguous logical 15689272ca1SMauro Carvalho Chehab address and physical address per inode, resulting in 15789272ca1SMauro Carvalho Chehab increasing the cache hit ratio. Set by default. 15889272ca1SMauro Carvalho Chehabnoextent_cache Disable an extent cache based on rb-tree explicitly, see 15989272ca1SMauro Carvalho Chehab the above extent_cache mount option. 16089272ca1SMauro Carvalho Chehabnoinline_data Disable the inline data feature, inline data feature is 16189272ca1SMauro Carvalho Chehab enabled by default. 16289272ca1SMauro Carvalho Chehabdata_flush Enable data flushing before checkpoint in order to 16389272ca1SMauro Carvalho Chehab persist data of regular and symlink. 16489272ca1SMauro Carvalho Chehabreserve_root=%d Support configuring reserved space which is used for 16589272ca1SMauro Carvalho Chehab allocation from a privileged user with specified uid or 16689272ca1SMauro Carvalho Chehab gid, unit: 4KB, the default limit is 0.2% of user blocks. 16789272ca1SMauro Carvalho Chehabresuid=%d The user ID which may use the reserved blocks. 16889272ca1SMauro Carvalho Chehabresgid=%d The group ID which may use the reserved blocks. 16989272ca1SMauro Carvalho Chehabfault_injection=%d Enable fault injection in all supported types with 17089272ca1SMauro Carvalho Chehab specified injection rate. 17189272ca1SMauro Carvalho Chehabfault_type=%d Support configuring fault injection type, should be 17289272ca1SMauro Carvalho Chehab enabled with fault_injection option, fault type value 17389272ca1SMauro Carvalho Chehab is shown below, it supports single or combined type. 17489272ca1SMauro Carvalho Chehab 17589272ca1SMauro Carvalho Chehab =================== =========== 17689272ca1SMauro Carvalho Chehab Type_Name Type_Value 17789272ca1SMauro Carvalho Chehab =================== =========== 17889272ca1SMauro Carvalho Chehab FAULT_KMALLOC 0x000000001 17989272ca1SMauro Carvalho Chehab FAULT_KVMALLOC 0x000000002 18089272ca1SMauro Carvalho Chehab FAULT_PAGE_ALLOC 0x000000004 18189272ca1SMauro Carvalho Chehab FAULT_PAGE_GET 0x000000008 18289272ca1SMauro Carvalho Chehab FAULT_ALLOC_BIO 0x000000010 18389272ca1SMauro Carvalho Chehab FAULT_ALLOC_NID 0x000000020 18489272ca1SMauro Carvalho Chehab FAULT_ORPHAN 0x000000040 18589272ca1SMauro Carvalho Chehab FAULT_BLOCK 0x000000080 18689272ca1SMauro Carvalho Chehab FAULT_DIR_DEPTH 0x000000100 18789272ca1SMauro Carvalho Chehab FAULT_EVICT_INODE 0x000000200 18889272ca1SMauro Carvalho Chehab FAULT_TRUNCATE 0x000000400 18989272ca1SMauro Carvalho Chehab FAULT_READ_IO 0x000000800 19089272ca1SMauro Carvalho Chehab FAULT_CHECKPOINT 0x000001000 19189272ca1SMauro Carvalho Chehab FAULT_DISCARD 0x000002000 19289272ca1SMauro Carvalho Chehab FAULT_WRITE_IO 0x000004000 19389272ca1SMauro Carvalho Chehab =================== =========== 19489272ca1SMauro Carvalho Chehabmode=%s Control block allocation mode which supports "adaptive" 19589272ca1SMauro Carvalho Chehab and "lfs". In "lfs" mode, there should be no random 19689272ca1SMauro Carvalho Chehab writes towards main area. 19789272ca1SMauro Carvalho Chehabio_bits=%u Set the bit size of write IO requests. It should be set 19889272ca1SMauro Carvalho Chehab with "mode=lfs". 19989272ca1SMauro Carvalho Chehabusrquota Enable plain user disk quota accounting. 20089272ca1SMauro Carvalho Chehabgrpquota Enable plain group disk quota accounting. 20189272ca1SMauro Carvalho Chehabprjquota Enable plain project quota accounting. 20289272ca1SMauro Carvalho Chehabusrjquota=<file> Appoint specified file and type during mount, so that quota 20389272ca1SMauro Carvalho Chehabgrpjquota=<file> information can be properly updated during recovery flow, 20489272ca1SMauro Carvalho Chehabprjjquota=<file> <quota file>: must be in root directory; 20589272ca1SMauro Carvalho Chehabjqfmt=<quota type> <quota type>: [vfsold,vfsv0,vfsv1]. 20689272ca1SMauro Carvalho Chehaboffusrjquota Turn off user journelled quota. 20789272ca1SMauro Carvalho Chehaboffgrpjquota Turn off group journelled quota. 20889272ca1SMauro Carvalho Chehaboffprjjquota Turn off project journelled quota. 20989272ca1SMauro Carvalho Chehabquota Enable plain user disk quota accounting. 21089272ca1SMauro Carvalho Chehabnoquota Disable all plain disk quota option. 21189272ca1SMauro Carvalho Chehabwhint_mode=%s Control which write hints are passed down to block 21289272ca1SMauro Carvalho Chehab layer. This supports "off", "user-based", and 21389272ca1SMauro Carvalho Chehab "fs-based". In "off" mode (default), f2fs does not pass 21489272ca1SMauro Carvalho Chehab down hints. In "user-based" mode, f2fs tries to pass 21589272ca1SMauro Carvalho Chehab down hints given by users. And in "fs-based" mode, f2fs 21689272ca1SMauro Carvalho Chehab passes down hints with its policy. 21789272ca1SMauro Carvalho Chehaballoc_mode=%s Adjust block allocation policy, which supports "reuse" 21889272ca1SMauro Carvalho Chehab and "default". 21989272ca1SMauro Carvalho Chehabfsync_mode=%s Control the policy of fsync. Currently supports "posix", 22089272ca1SMauro Carvalho Chehab "strict", and "nobarrier". In "posix" mode, which is 22189272ca1SMauro Carvalho Chehab default, fsync will follow POSIX semantics and does a 22289272ca1SMauro Carvalho Chehab light operation to improve the filesystem performance. 22389272ca1SMauro Carvalho Chehab In "strict" mode, fsync will be heavy and behaves in line 22489272ca1SMauro Carvalho Chehab with xfs, ext4 and btrfs, where xfstest generic/342 will 22589272ca1SMauro Carvalho Chehab pass, but the performance will regress. "nobarrier" is 22689272ca1SMauro Carvalho Chehab based on "posix", but doesn't issue flush command for 22789272ca1SMauro Carvalho Chehab non-atomic files likewise "nobarrier" mount option. 22889272ca1SMauro Carvalho Chehabtest_dummy_encryption Enable dummy encryption, which provides a fake fscrypt 22989272ca1SMauro Carvalho Chehab context. The fake fscrypt context is used by xfstests. 23089272ca1SMauro Carvalho Chehabcheckpoint=%s[:%u[%]] Set to "disable" to turn off checkpointing. Set to "enable" 23189272ca1SMauro Carvalho Chehab to reenable checkpointing. Is enabled by default. While 23289272ca1SMauro Carvalho Chehab disabled, any unmounting or unexpected shutdowns will cause 23389272ca1SMauro Carvalho Chehab the filesystem contents to appear as they did when the 23489272ca1SMauro Carvalho Chehab filesystem was mounted with that option. 23589272ca1SMauro Carvalho Chehab While mounting with checkpoint=disabled, the filesystem must 23689272ca1SMauro Carvalho Chehab run garbage collection to ensure that all available space can 23789272ca1SMauro Carvalho Chehab be used. If this takes too much time, the mount may return 23889272ca1SMauro Carvalho Chehab EAGAIN. You may optionally add a value to indicate how much 23989272ca1SMauro Carvalho Chehab of the disk you would be willing to temporarily give up to 24089272ca1SMauro Carvalho Chehab avoid additional garbage collection. This can be given as a 24189272ca1SMauro Carvalho Chehab number of blocks, or as a percent. For instance, mounting 24289272ca1SMauro Carvalho Chehab with checkpoint=disable:100% would always succeed, but it may 24389272ca1SMauro Carvalho Chehab hide up to all remaining free space. The actual space that 24489272ca1SMauro Carvalho Chehab would be unusable can be viewed at /sys/fs/f2fs/<disk>/unusable 24589272ca1SMauro Carvalho Chehab This space is reclaimed once checkpoint=enable. 246f40f31caSLinus Torvaldscompress_algorithm=%s Control compress algorithm, currently f2fs supports "lzo", 247*6d92b201SChao Yu "lz4", "zstd" and "lzo-rle" algorithm. 24889272ca1SMauro Carvalho Chehabcompress_log_size=%u Support configuring compress cluster size, the size will 24989272ca1SMauro Carvalho Chehab be 4KB * (1 << %u), 16KB is minimum size, also it's 25089272ca1SMauro Carvalho Chehab default size. 25189272ca1SMauro Carvalho Chehabcompress_extension=%s Support adding specified extension, so that f2fs can enable 25289272ca1SMauro Carvalho Chehab compression on those corresponding files, e.g. if all files 25389272ca1SMauro Carvalho Chehab with '.ext' has high compression rate, we can set the '.ext' 25489272ca1SMauro Carvalho Chehab on compression extension list and enable compression on 25589272ca1SMauro Carvalho Chehab these file by default rather than to enable it via ioctl. 25689272ca1SMauro Carvalho Chehab For other files, we can still enable compression via ioctl. 25789272ca1SMauro Carvalho Chehab====================== ============================================================ 25889272ca1SMauro Carvalho Chehab 25989272ca1SMauro Carvalho ChehabDebugfs Entries 26089272ca1SMauro Carvalho Chehab=============== 26189272ca1SMauro Carvalho Chehab 26289272ca1SMauro Carvalho Chehab/sys/kernel/debug/f2fs/ contains information about all the partitions mounted as 26389272ca1SMauro Carvalho Chehabf2fs. Each file shows the whole f2fs information. 26489272ca1SMauro Carvalho Chehab 26589272ca1SMauro Carvalho Chehab/sys/kernel/debug/f2fs/status includes: 26689272ca1SMauro Carvalho Chehab 26789272ca1SMauro Carvalho Chehab - major file system information managed by f2fs currently 26889272ca1SMauro Carvalho Chehab - average SIT information about whole segments 26989272ca1SMauro Carvalho Chehab - current memory footprint consumed by f2fs. 27089272ca1SMauro Carvalho Chehab 27189272ca1SMauro Carvalho ChehabSysfs Entries 27289272ca1SMauro Carvalho Chehab============= 27389272ca1SMauro Carvalho Chehab 27489272ca1SMauro Carvalho ChehabInformation about mounted f2fs file systems can be found in 27589272ca1SMauro Carvalho Chehab/sys/fs/f2fs. Each mounted filesystem will have a directory in 27689272ca1SMauro Carvalho Chehab/sys/fs/f2fs based on its device name (i.e., /sys/fs/f2fs/sda). 27789272ca1SMauro Carvalho ChehabThe files in each per-device directory are shown in table below. 27889272ca1SMauro Carvalho Chehab 27989272ca1SMauro Carvalho ChehabFiles in /sys/fs/f2fs/<devname> 28089272ca1SMauro Carvalho Chehab(see also Documentation/ABI/testing/sysfs-fs-f2fs) 28189272ca1SMauro Carvalho Chehab 28289272ca1SMauro Carvalho ChehabUsage 28389272ca1SMauro Carvalho Chehab===== 28489272ca1SMauro Carvalho Chehab 28589272ca1SMauro Carvalho Chehab1. Download userland tools and compile them. 28689272ca1SMauro Carvalho Chehab 28789272ca1SMauro Carvalho Chehab2. Skip, if f2fs was compiled statically inside kernel. 28889272ca1SMauro Carvalho Chehab Otherwise, insert the f2fs.ko module:: 28989272ca1SMauro Carvalho Chehab 29089272ca1SMauro Carvalho Chehab # insmod f2fs.ko 29189272ca1SMauro Carvalho Chehab 29289272ca1SMauro Carvalho Chehab3. Create a directory trying to mount:: 29389272ca1SMauro Carvalho Chehab 29489272ca1SMauro Carvalho Chehab # mkdir /mnt/f2fs 29589272ca1SMauro Carvalho Chehab 29689272ca1SMauro Carvalho Chehab4. Format the block device, and then mount as f2fs:: 29789272ca1SMauro Carvalho Chehab 29889272ca1SMauro Carvalho Chehab # mkfs.f2fs -l label /dev/block_device 29989272ca1SMauro Carvalho Chehab # mount -t f2fs /dev/block_device /mnt/f2fs 30089272ca1SMauro Carvalho Chehab 30189272ca1SMauro Carvalho Chehabmkfs.f2fs 30289272ca1SMauro Carvalho Chehab--------- 30389272ca1SMauro Carvalho ChehabThe mkfs.f2fs is for the use of formatting a partition as the f2fs filesystem, 30489272ca1SMauro Carvalho Chehabwhich builds a basic on-disk layout. 30589272ca1SMauro Carvalho Chehab 30689272ca1SMauro Carvalho ChehabThe options consist of: 30789272ca1SMauro Carvalho Chehab 30889272ca1SMauro Carvalho Chehab=============== =========================================================== 30989272ca1SMauro Carvalho Chehab``-l [label]`` Give a volume label, up to 512 unicode name. 31089272ca1SMauro Carvalho Chehab``-a [0 or 1]`` Split start location of each area for heap-based allocation. 31189272ca1SMauro Carvalho Chehab 31289272ca1SMauro Carvalho Chehab 1 is set by default, which performs this. 31389272ca1SMauro Carvalho Chehab``-o [int]`` Set overprovision ratio in percent over volume size. 31489272ca1SMauro Carvalho Chehab 31589272ca1SMauro Carvalho Chehab 5 is set by default. 31689272ca1SMauro Carvalho Chehab``-s [int]`` Set the number of segments per section. 31789272ca1SMauro Carvalho Chehab 31889272ca1SMauro Carvalho Chehab 1 is set by default. 31989272ca1SMauro Carvalho Chehab``-z [int]`` Set the number of sections per zone. 32089272ca1SMauro Carvalho Chehab 32189272ca1SMauro Carvalho Chehab 1 is set by default. 32289272ca1SMauro Carvalho Chehab``-e [str]`` Set basic extension list. e.g. "mp3,gif,mov" 32389272ca1SMauro Carvalho Chehab``-t [0 or 1]`` Disable discard command or not. 32489272ca1SMauro Carvalho Chehab 32589272ca1SMauro Carvalho Chehab 1 is set by default, which conducts discard. 32689272ca1SMauro Carvalho Chehab=============== =========================================================== 32789272ca1SMauro Carvalho Chehab 32889272ca1SMauro Carvalho Chehabfsck.f2fs 32989272ca1SMauro Carvalho Chehab--------- 33089272ca1SMauro Carvalho ChehabThe fsck.f2fs is a tool to check the consistency of an f2fs-formatted 33189272ca1SMauro Carvalho Chehabpartition, which examines whether the filesystem metadata and user-made data 33289272ca1SMauro Carvalho Chehabare cross-referenced correctly or not. 33389272ca1SMauro Carvalho ChehabNote that, initial version of the tool does not fix any inconsistency. 33489272ca1SMauro Carvalho Chehab 33589272ca1SMauro Carvalho ChehabThe options consist of:: 33689272ca1SMauro Carvalho Chehab 33789272ca1SMauro Carvalho Chehab -d debug level [default:0] 33889272ca1SMauro Carvalho Chehab 33989272ca1SMauro Carvalho Chehabdump.f2fs 34089272ca1SMauro Carvalho Chehab--------- 34189272ca1SMauro Carvalho ChehabThe dump.f2fs shows the information of specific inode and dumps SSA and SIT to 34289272ca1SMauro Carvalho Chehabfile. Each file is dump_ssa and dump_sit. 34389272ca1SMauro Carvalho Chehab 34489272ca1SMauro Carvalho ChehabThe dump.f2fs is used to debug on-disk data structures of the f2fs filesystem. 34589272ca1SMauro Carvalho ChehabIt shows on-disk inode information recognized by a given inode number, and is 34689272ca1SMauro Carvalho Chehabable to dump all the SSA and SIT entries into predefined files, ./dump_ssa and 34789272ca1SMauro Carvalho Chehab./dump_sit respectively. 34889272ca1SMauro Carvalho Chehab 34989272ca1SMauro Carvalho ChehabThe options consist of:: 35089272ca1SMauro Carvalho Chehab 35189272ca1SMauro Carvalho Chehab -d debug level [default:0] 35289272ca1SMauro Carvalho Chehab -i inode no (hex) 35389272ca1SMauro Carvalho Chehab -s [SIT dump segno from #1~#2 (decimal), for all 0~-1] 35489272ca1SMauro Carvalho Chehab -a [SSA dump segno from #1~#2 (decimal), for all 0~-1] 35589272ca1SMauro Carvalho Chehab 35689272ca1SMauro Carvalho ChehabExamples:: 35789272ca1SMauro Carvalho Chehab 35889272ca1SMauro Carvalho Chehab # dump.f2fs -i [ino] /dev/sdx 35989272ca1SMauro Carvalho Chehab # dump.f2fs -s 0~-1 /dev/sdx (SIT dump) 36089272ca1SMauro Carvalho Chehab # dump.f2fs -a 0~-1 /dev/sdx (SSA dump) 36189272ca1SMauro Carvalho Chehab 36289272ca1SMauro Carvalho ChehabDesign 36389272ca1SMauro Carvalho Chehab====== 36489272ca1SMauro Carvalho Chehab 36589272ca1SMauro Carvalho ChehabOn-disk Layout 36689272ca1SMauro Carvalho Chehab-------------- 36789272ca1SMauro Carvalho Chehab 36889272ca1SMauro Carvalho ChehabF2FS divides the whole volume into a number of segments, each of which is fixed 36989272ca1SMauro Carvalho Chehabto 2MB in size. A section is composed of consecutive segments, and a zone 37089272ca1SMauro Carvalho Chehabconsists of a set of sections. By default, section and zone sizes are set to one 37189272ca1SMauro Carvalho Chehabsegment size identically, but users can easily modify the sizes by mkfs. 37289272ca1SMauro Carvalho Chehab 37389272ca1SMauro Carvalho ChehabF2FS splits the entire volume into six areas, and all the areas except superblock 37489272ca1SMauro Carvalho Chehabconsists of multiple segments as described below:: 37589272ca1SMauro Carvalho Chehab 37689272ca1SMauro Carvalho Chehab align with the zone size <-| 37789272ca1SMauro Carvalho Chehab |-> align with the segment size 37889272ca1SMauro Carvalho Chehab _________________________________________________________________________ 37989272ca1SMauro Carvalho Chehab | | | Segment | Node | Segment | | 38089272ca1SMauro Carvalho Chehab | Superblock | Checkpoint | Info. | Address | Summary | Main | 38189272ca1SMauro Carvalho Chehab | (SB) | (CP) | Table (SIT) | Table (NAT) | Area (SSA) | | 38289272ca1SMauro Carvalho Chehab |____________|_____2______|______N______|______N______|______N_____|__N___| 38389272ca1SMauro Carvalho Chehab . . 38489272ca1SMauro Carvalho Chehab . . 38589272ca1SMauro Carvalho Chehab . . 38689272ca1SMauro Carvalho Chehab ._________________________________________. 38789272ca1SMauro Carvalho Chehab |_Segment_|_..._|_Segment_|_..._|_Segment_| 38889272ca1SMauro Carvalho Chehab . . 38989272ca1SMauro Carvalho Chehab ._________._________ 39089272ca1SMauro Carvalho Chehab |_section_|__...__|_ 39189272ca1SMauro Carvalho Chehab . . 39289272ca1SMauro Carvalho Chehab .________. 39389272ca1SMauro Carvalho Chehab |__zone__| 39489272ca1SMauro Carvalho Chehab 39589272ca1SMauro Carvalho Chehab- Superblock (SB) 39689272ca1SMauro Carvalho Chehab It is located at the beginning of the partition, and there exist two copies 39789272ca1SMauro Carvalho Chehab to avoid file system crash. It contains basic partition information and some 39889272ca1SMauro Carvalho Chehab default parameters of f2fs. 39989272ca1SMauro Carvalho Chehab 40089272ca1SMauro Carvalho Chehab- Checkpoint (CP) 40189272ca1SMauro Carvalho Chehab It contains file system information, bitmaps for valid NAT/SIT sets, orphan 40289272ca1SMauro Carvalho Chehab inode lists, and summary entries of current active segments. 40389272ca1SMauro Carvalho Chehab 40489272ca1SMauro Carvalho Chehab- Segment Information Table (SIT) 40589272ca1SMauro Carvalho Chehab It contains segment information such as valid block count and bitmap for the 40689272ca1SMauro Carvalho Chehab validity of all the blocks. 40789272ca1SMauro Carvalho Chehab 40889272ca1SMauro Carvalho Chehab- Node Address Table (NAT) 40989272ca1SMauro Carvalho Chehab It is composed of a block address table for all the node blocks stored in 41089272ca1SMauro Carvalho Chehab Main area. 41189272ca1SMauro Carvalho Chehab 41289272ca1SMauro Carvalho Chehab- Segment Summary Area (SSA) 41389272ca1SMauro Carvalho Chehab It contains summary entries which contains the owner information of all the 41489272ca1SMauro Carvalho Chehab data and node blocks stored in Main area. 41589272ca1SMauro Carvalho Chehab 41689272ca1SMauro Carvalho Chehab- Main Area 41789272ca1SMauro Carvalho Chehab It contains file and directory data including their indices. 41889272ca1SMauro Carvalho Chehab 41989272ca1SMauro Carvalho ChehabIn order to avoid misalignment between file system and flash-based storage, F2FS 42089272ca1SMauro Carvalho Chehabaligns the start block address of CP with the segment size. Also, it aligns the 42189272ca1SMauro Carvalho Chehabstart block address of Main area with the zone size by reserving some segments 42289272ca1SMauro Carvalho Chehabin SSA area. 42389272ca1SMauro Carvalho Chehab 42489272ca1SMauro Carvalho ChehabReference the following survey for additional technical details. 42589272ca1SMauro Carvalho Chehabhttps://wiki.linaro.org/WorkingGroups/Kernel/Projects/FlashCardSurvey 42689272ca1SMauro Carvalho Chehab 42789272ca1SMauro Carvalho ChehabFile System Metadata Structure 42889272ca1SMauro Carvalho Chehab------------------------------ 42989272ca1SMauro Carvalho Chehab 43089272ca1SMauro Carvalho ChehabF2FS adopts the checkpointing scheme to maintain file system consistency. At 43189272ca1SMauro Carvalho Chehabmount time, F2FS first tries to find the last valid checkpoint data by scanning 43289272ca1SMauro Carvalho ChehabCP area. In order to reduce the scanning time, F2FS uses only two copies of CP. 43389272ca1SMauro Carvalho ChehabOne of them always indicates the last valid data, which is called as shadow copy 43489272ca1SMauro Carvalho Chehabmechanism. In addition to CP, NAT and SIT also adopt the shadow copy mechanism. 43589272ca1SMauro Carvalho Chehab 43689272ca1SMauro Carvalho ChehabFor file system consistency, each CP points to which NAT and SIT copies are 43789272ca1SMauro Carvalho Chehabvalid, as shown as below:: 43889272ca1SMauro Carvalho Chehab 43989272ca1SMauro Carvalho Chehab +--------+----------+---------+ 44089272ca1SMauro Carvalho Chehab | CP | SIT | NAT | 44189272ca1SMauro Carvalho Chehab +--------+----------+---------+ 44289272ca1SMauro Carvalho Chehab . . . . 44389272ca1SMauro Carvalho Chehab . . . . 44489272ca1SMauro Carvalho Chehab . . . . 44589272ca1SMauro Carvalho Chehab +-------+-------+--------+--------+--------+--------+ 44689272ca1SMauro Carvalho Chehab | CP #0 | CP #1 | SIT #0 | SIT #1 | NAT #0 | NAT #1 | 44789272ca1SMauro Carvalho Chehab +-------+-------+--------+--------+--------+--------+ 44889272ca1SMauro Carvalho Chehab | ^ ^ 44989272ca1SMauro Carvalho Chehab | | | 45089272ca1SMauro Carvalho Chehab `----------------------------------------' 45189272ca1SMauro Carvalho Chehab 45289272ca1SMauro Carvalho ChehabIndex Structure 45389272ca1SMauro Carvalho Chehab--------------- 45489272ca1SMauro Carvalho Chehab 45589272ca1SMauro Carvalho ChehabThe key data structure to manage the data locations is a "node". Similar to 45689272ca1SMauro Carvalho Chehabtraditional file structures, F2FS has three types of node: inode, direct node, 45789272ca1SMauro Carvalho Chehabindirect node. F2FS assigns 4KB to an inode block which contains 923 data block 45889272ca1SMauro Carvalho Chehabindices, two direct node pointers, two indirect node pointers, and one double 45989272ca1SMauro Carvalho Chehabindirect node pointer as described below. One direct node block contains 1018 46089272ca1SMauro Carvalho Chehabdata blocks, and one indirect node block contains also 1018 node blocks. Thus, 46189272ca1SMauro Carvalho Chehabone inode block (i.e., a file) covers:: 46289272ca1SMauro Carvalho Chehab 46389272ca1SMauro Carvalho Chehab 4KB * (923 + 2 * 1018 + 2 * 1018 * 1018 + 1018 * 1018 * 1018) := 3.94TB. 46489272ca1SMauro Carvalho Chehab 46589272ca1SMauro Carvalho Chehab Inode block (4KB) 46689272ca1SMauro Carvalho Chehab |- data (923) 46789272ca1SMauro Carvalho Chehab |- direct node (2) 46889272ca1SMauro Carvalho Chehab | `- data (1018) 46989272ca1SMauro Carvalho Chehab |- indirect node (2) 47089272ca1SMauro Carvalho Chehab | `- direct node (1018) 47189272ca1SMauro Carvalho Chehab | `- data (1018) 47289272ca1SMauro Carvalho Chehab `- double indirect node (1) 47389272ca1SMauro Carvalho Chehab `- indirect node (1018) 47489272ca1SMauro Carvalho Chehab `- direct node (1018) 47589272ca1SMauro Carvalho Chehab `- data (1018) 47689272ca1SMauro Carvalho Chehab 47789272ca1SMauro Carvalho ChehabNote that, all the node blocks are mapped by NAT which means the location of 47889272ca1SMauro Carvalho Chehabeach node is translated by the NAT table. In the consideration of the wandering 47989272ca1SMauro Carvalho Chehabtree problem, F2FS is able to cut off the propagation of node updates caused by 48089272ca1SMauro Carvalho Chehableaf data writes. 48189272ca1SMauro Carvalho Chehab 48289272ca1SMauro Carvalho ChehabDirectory Structure 48389272ca1SMauro Carvalho Chehab------------------- 48489272ca1SMauro Carvalho Chehab 48589272ca1SMauro Carvalho ChehabA directory entry occupies 11 bytes, which consists of the following attributes. 48689272ca1SMauro Carvalho Chehab 48789272ca1SMauro Carvalho Chehab- hash hash value of the file name 48889272ca1SMauro Carvalho Chehab- ino inode number 48989272ca1SMauro Carvalho Chehab- len the length of file name 49089272ca1SMauro Carvalho Chehab- type file type such as directory, symlink, etc 49189272ca1SMauro Carvalho Chehab 49289272ca1SMauro Carvalho ChehabA dentry block consists of 214 dentry slots and file names. Therein a bitmap is 49389272ca1SMauro Carvalho Chehabused to represent whether each dentry is valid or not. A dentry block occupies 49489272ca1SMauro Carvalho Chehab4KB with the following composition. 49589272ca1SMauro Carvalho Chehab 49689272ca1SMauro Carvalho Chehab:: 49789272ca1SMauro Carvalho Chehab 49889272ca1SMauro Carvalho Chehab Dentry Block(4 K) = bitmap (27 bytes) + reserved (3 bytes) + 49989272ca1SMauro Carvalho Chehab dentries(11 * 214 bytes) + file name (8 * 214 bytes) 50089272ca1SMauro Carvalho Chehab 50189272ca1SMauro Carvalho Chehab [Bucket] 50289272ca1SMauro Carvalho Chehab +--------------------------------+ 50389272ca1SMauro Carvalho Chehab |dentry block 1 | dentry block 2 | 50489272ca1SMauro Carvalho Chehab +--------------------------------+ 50589272ca1SMauro Carvalho Chehab . . 50689272ca1SMauro Carvalho Chehab . . 50789272ca1SMauro Carvalho Chehab . [Dentry Block Structure: 4KB] . 50889272ca1SMauro Carvalho Chehab +--------+----------+----------+------------+ 50989272ca1SMauro Carvalho Chehab | bitmap | reserved | dentries | file names | 51089272ca1SMauro Carvalho Chehab +--------+----------+----------+------------+ 51189272ca1SMauro Carvalho Chehab [Dentry Block: 4KB] . . 51289272ca1SMauro Carvalho Chehab . . 51389272ca1SMauro Carvalho Chehab . . 51489272ca1SMauro Carvalho Chehab +------+------+-----+------+ 51589272ca1SMauro Carvalho Chehab | hash | ino | len | type | 51689272ca1SMauro Carvalho Chehab +------+------+-----+------+ 51789272ca1SMauro Carvalho Chehab [Dentry Structure: 11 bytes] 51889272ca1SMauro Carvalho Chehab 51989272ca1SMauro Carvalho ChehabF2FS implements multi-level hash tables for directory structure. Each level has 52089272ca1SMauro Carvalho Chehaba hash table with dedicated number of hash buckets as shown below. Note that 52189272ca1SMauro Carvalho Chehab"A(2B)" means a bucket includes 2 data blocks. 52289272ca1SMauro Carvalho Chehab 52389272ca1SMauro Carvalho Chehab:: 52489272ca1SMauro Carvalho Chehab 52589272ca1SMauro Carvalho Chehab ---------------------- 52689272ca1SMauro Carvalho Chehab A : bucket 52789272ca1SMauro Carvalho Chehab B : block 52889272ca1SMauro Carvalho Chehab N : MAX_DIR_HASH_DEPTH 52989272ca1SMauro Carvalho Chehab ---------------------- 53089272ca1SMauro Carvalho Chehab 53189272ca1SMauro Carvalho Chehab level #0 | A(2B) 53289272ca1SMauro Carvalho Chehab | 53389272ca1SMauro Carvalho Chehab level #1 | A(2B) - A(2B) 53489272ca1SMauro Carvalho Chehab | 53589272ca1SMauro Carvalho Chehab level #2 | A(2B) - A(2B) - A(2B) - A(2B) 53689272ca1SMauro Carvalho Chehab . | . . . . 53789272ca1SMauro Carvalho Chehab level #N/2 | A(2B) - A(2B) - A(2B) - A(2B) - A(2B) - ... - A(2B) 53889272ca1SMauro Carvalho Chehab . | . . . . 53989272ca1SMauro Carvalho Chehab level #N | A(4B) - A(4B) - A(4B) - A(4B) - A(4B) - ... - A(4B) 54089272ca1SMauro Carvalho Chehab 54189272ca1SMauro Carvalho ChehabThe number of blocks and buckets are determined by:: 54289272ca1SMauro Carvalho Chehab 54389272ca1SMauro Carvalho Chehab ,- 2, if n < MAX_DIR_HASH_DEPTH / 2, 54489272ca1SMauro Carvalho Chehab # of blocks in level #n = | 54589272ca1SMauro Carvalho Chehab `- 4, Otherwise 54689272ca1SMauro Carvalho Chehab 54789272ca1SMauro Carvalho Chehab ,- 2^(n + dir_level), 54889272ca1SMauro Carvalho Chehab | if n + dir_level < MAX_DIR_HASH_DEPTH / 2, 54989272ca1SMauro Carvalho Chehab # of buckets in level #n = | 55089272ca1SMauro Carvalho Chehab `- 2^((MAX_DIR_HASH_DEPTH / 2) - 1), 55189272ca1SMauro Carvalho Chehab Otherwise 55289272ca1SMauro Carvalho Chehab 55389272ca1SMauro Carvalho ChehabWhen F2FS finds a file name in a directory, at first a hash value of the file 55489272ca1SMauro Carvalho Chehabname is calculated. Then, F2FS scans the hash table in level #0 to find the 55589272ca1SMauro Carvalho Chehabdentry consisting of the file name and its inode number. If not found, F2FS 55689272ca1SMauro Carvalho Chehabscans the next hash table in level #1. In this way, F2FS scans hash tables in 55789272ca1SMauro Carvalho Chehabeach levels incrementally from 1 to N. In each levels F2FS needs to scan only 55889272ca1SMauro Carvalho Chehabone bucket determined by the following equation, which shows O(log(# of files)) 55989272ca1SMauro Carvalho Chehabcomplexity:: 56089272ca1SMauro Carvalho Chehab 56189272ca1SMauro Carvalho Chehab bucket number to scan in level #n = (hash value) % (# of buckets in level #n) 56289272ca1SMauro Carvalho Chehab 56389272ca1SMauro Carvalho ChehabIn the case of file creation, F2FS finds empty consecutive slots that cover the 56489272ca1SMauro Carvalho Chehabfile name. F2FS searches the empty slots in the hash tables of whole levels from 56589272ca1SMauro Carvalho Chehab1 to N in the same way as the lookup operation. 56689272ca1SMauro Carvalho Chehab 56789272ca1SMauro Carvalho ChehabThe following figure shows an example of two cases holding children:: 56889272ca1SMauro Carvalho Chehab 56989272ca1SMauro Carvalho Chehab --------------> Dir <-------------- 57089272ca1SMauro Carvalho Chehab | | 57189272ca1SMauro Carvalho Chehab child child 57289272ca1SMauro Carvalho Chehab 57389272ca1SMauro Carvalho Chehab child - child [hole] - child 57489272ca1SMauro Carvalho Chehab 57589272ca1SMauro Carvalho Chehab child - child - child [hole] - [hole] - child 57689272ca1SMauro Carvalho Chehab 57789272ca1SMauro Carvalho Chehab Case 1: Case 2: 57889272ca1SMauro Carvalho Chehab Number of children = 6, Number of children = 3, 57989272ca1SMauro Carvalho Chehab File size = 7 File size = 7 58089272ca1SMauro Carvalho Chehab 58189272ca1SMauro Carvalho ChehabDefault Block Allocation 58289272ca1SMauro Carvalho Chehab------------------------ 58389272ca1SMauro Carvalho Chehab 58489272ca1SMauro Carvalho ChehabAt runtime, F2FS manages six active logs inside "Main" area: Hot/Warm/Cold node 58589272ca1SMauro Carvalho Chehaband Hot/Warm/Cold data. 58689272ca1SMauro Carvalho Chehab 58789272ca1SMauro Carvalho Chehab- Hot node contains direct node blocks of directories. 58889272ca1SMauro Carvalho Chehab- Warm node contains direct node blocks except hot node blocks. 58989272ca1SMauro Carvalho Chehab- Cold node contains indirect node blocks 59089272ca1SMauro Carvalho Chehab- Hot data contains dentry blocks 59189272ca1SMauro Carvalho Chehab- Warm data contains data blocks except hot and cold data blocks 59289272ca1SMauro Carvalho Chehab- Cold data contains multimedia data or migrated data blocks 59389272ca1SMauro Carvalho Chehab 59489272ca1SMauro Carvalho ChehabLFS has two schemes for free space management: threaded log and copy-and-compac- 59589272ca1SMauro Carvalho Chehabtion. The copy-and-compaction scheme which is known as cleaning, is well-suited 59689272ca1SMauro Carvalho Chehabfor devices showing very good sequential write performance, since free segments 59789272ca1SMauro Carvalho Chehabare served all the time for writing new data. However, it suffers from cleaning 59889272ca1SMauro Carvalho Chehaboverhead under high utilization. Contrarily, the threaded log scheme suffers 59989272ca1SMauro Carvalho Chehabfrom random writes, but no cleaning process is needed. F2FS adopts a hybrid 60089272ca1SMauro Carvalho Chehabscheme where the copy-and-compaction scheme is adopted by default, but the 60189272ca1SMauro Carvalho Chehabpolicy is dynamically changed to the threaded log scheme according to the file 60289272ca1SMauro Carvalho Chehabsystem status. 60389272ca1SMauro Carvalho Chehab 60489272ca1SMauro Carvalho ChehabIn order to align F2FS with underlying flash-based storage, F2FS allocates a 60589272ca1SMauro Carvalho Chehabsegment in a unit of section. F2FS expects that the section size would be the 60689272ca1SMauro Carvalho Chehabsame as the unit size of garbage collection in FTL. Furthermore, with respect 60789272ca1SMauro Carvalho Chehabto the mapping granularity in FTL, F2FS allocates each section of the active 60889272ca1SMauro Carvalho Chehablogs from different zones as much as possible, since FTL can write the data in 60989272ca1SMauro Carvalho Chehabthe active logs into one allocation unit according to its mapping granularity. 61089272ca1SMauro Carvalho Chehab 61189272ca1SMauro Carvalho ChehabCleaning process 61289272ca1SMauro Carvalho Chehab---------------- 61389272ca1SMauro Carvalho Chehab 61489272ca1SMauro Carvalho ChehabF2FS does cleaning both on demand and in the background. On-demand cleaning is 61589272ca1SMauro Carvalho Chehabtriggered when there are not enough free segments to serve VFS calls. Background 61689272ca1SMauro Carvalho Chehabcleaner is operated by a kernel thread, and triggers the cleaning job when the 61789272ca1SMauro Carvalho Chehabsystem is idle. 61889272ca1SMauro Carvalho Chehab 61989272ca1SMauro Carvalho ChehabF2FS supports two victim selection policies: greedy and cost-benefit algorithms. 62089272ca1SMauro Carvalho ChehabIn the greedy algorithm, F2FS selects a victim segment having the smallest number 62189272ca1SMauro Carvalho Chehabof valid blocks. In the cost-benefit algorithm, F2FS selects a victim segment 62289272ca1SMauro Carvalho Chehabaccording to the segment age and the number of valid blocks in order to address 62389272ca1SMauro Carvalho Chehablog block thrashing problem in the greedy algorithm. F2FS adopts the greedy 62489272ca1SMauro Carvalho Chehabalgorithm for on-demand cleaner, while background cleaner adopts cost-benefit 62589272ca1SMauro Carvalho Chehabalgorithm. 62689272ca1SMauro Carvalho Chehab 62789272ca1SMauro Carvalho ChehabIn order to identify whether the data in the victim segment are valid or not, 62889272ca1SMauro Carvalho ChehabF2FS manages a bitmap. Each bit represents the validity of a block, and the 62989272ca1SMauro Carvalho Chehabbitmap is composed of a bit stream covering whole blocks in main area. 63089272ca1SMauro Carvalho Chehab 63189272ca1SMauro Carvalho ChehabWrite-hint Policy 63289272ca1SMauro Carvalho Chehab----------------- 63389272ca1SMauro Carvalho Chehab 63489272ca1SMauro Carvalho Chehab1) whint_mode=off. F2FS only passes down WRITE_LIFE_NOT_SET. 63589272ca1SMauro Carvalho Chehab 63689272ca1SMauro Carvalho Chehab2) whint_mode=user-based. F2FS tries to pass down hints given by 63789272ca1SMauro Carvalho Chehabusers. 63889272ca1SMauro Carvalho Chehab 63989272ca1SMauro Carvalho Chehab===================== ======================== =================== 64089272ca1SMauro Carvalho ChehabUser F2FS Block 64189272ca1SMauro Carvalho Chehab===================== ======================== =================== 64289272ca1SMauro Carvalho Chehab META WRITE_LIFE_NOT_SET 64389272ca1SMauro Carvalho Chehab HOT_NODE " 64489272ca1SMauro Carvalho Chehab WARM_NODE " 64589272ca1SMauro Carvalho Chehab COLD_NODE " 64689272ca1SMauro Carvalho Chehabioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 64789272ca1SMauro Carvalho Chehabextension list " " 64889272ca1SMauro Carvalho Chehab 64989272ca1SMauro Carvalho Chehab-- buffered io 65089272ca1SMauro Carvalho ChehabWRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 65189272ca1SMauro Carvalho ChehabWRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 65289272ca1SMauro Carvalho ChehabWRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 65389272ca1SMauro Carvalho ChehabWRITE_LIFE_NONE " " 65489272ca1SMauro Carvalho ChehabWRITE_LIFE_MEDIUM " " 65589272ca1SMauro Carvalho ChehabWRITE_LIFE_LONG " " 65689272ca1SMauro Carvalho Chehab 65789272ca1SMauro Carvalho Chehab-- direct io 65889272ca1SMauro Carvalho ChehabWRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 65989272ca1SMauro Carvalho ChehabWRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 66089272ca1SMauro Carvalho ChehabWRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 66189272ca1SMauro Carvalho ChehabWRITE_LIFE_NONE " WRITE_LIFE_NONE 66289272ca1SMauro Carvalho ChehabWRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 66389272ca1SMauro Carvalho ChehabWRITE_LIFE_LONG " WRITE_LIFE_LONG 66489272ca1SMauro Carvalho Chehab===================== ======================== =================== 66589272ca1SMauro Carvalho Chehab 66689272ca1SMauro Carvalho Chehab3) whint_mode=fs-based. F2FS passes down hints with its policy. 66789272ca1SMauro Carvalho Chehab 66889272ca1SMauro Carvalho Chehab===================== ======================== =================== 66989272ca1SMauro Carvalho ChehabUser F2FS Block 67089272ca1SMauro Carvalho Chehab===================== ======================== =================== 67189272ca1SMauro Carvalho Chehab META WRITE_LIFE_MEDIUM; 67289272ca1SMauro Carvalho Chehab HOT_NODE WRITE_LIFE_NOT_SET 67389272ca1SMauro Carvalho Chehab WARM_NODE " 67489272ca1SMauro Carvalho Chehab COLD_NODE WRITE_LIFE_NONE 67589272ca1SMauro Carvalho Chehabioctl(COLD) COLD_DATA WRITE_LIFE_EXTREME 67689272ca1SMauro Carvalho Chehabextension list " " 67789272ca1SMauro Carvalho Chehab 67889272ca1SMauro Carvalho Chehab-- buffered io 67989272ca1SMauro Carvalho ChehabWRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 68089272ca1SMauro Carvalho ChehabWRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 68189272ca1SMauro Carvalho ChehabWRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_LONG 68289272ca1SMauro Carvalho ChehabWRITE_LIFE_NONE " " 68389272ca1SMauro Carvalho ChehabWRITE_LIFE_MEDIUM " " 68489272ca1SMauro Carvalho ChehabWRITE_LIFE_LONG " " 68589272ca1SMauro Carvalho Chehab 68689272ca1SMauro Carvalho Chehab-- direct io 68789272ca1SMauro Carvalho ChehabWRITE_LIFE_EXTREME COLD_DATA WRITE_LIFE_EXTREME 68889272ca1SMauro Carvalho ChehabWRITE_LIFE_SHORT HOT_DATA WRITE_LIFE_SHORT 68989272ca1SMauro Carvalho ChehabWRITE_LIFE_NOT_SET WARM_DATA WRITE_LIFE_NOT_SET 69089272ca1SMauro Carvalho ChehabWRITE_LIFE_NONE " WRITE_LIFE_NONE 69189272ca1SMauro Carvalho ChehabWRITE_LIFE_MEDIUM " WRITE_LIFE_MEDIUM 69289272ca1SMauro Carvalho ChehabWRITE_LIFE_LONG " WRITE_LIFE_LONG 69389272ca1SMauro Carvalho Chehab===================== ======================== =================== 69489272ca1SMauro Carvalho Chehab 69589272ca1SMauro Carvalho ChehabFallocate(2) Policy 69689272ca1SMauro Carvalho Chehab------------------- 69789272ca1SMauro Carvalho Chehab 69889272ca1SMauro Carvalho ChehabThe default policy follows the below posix rule. 69989272ca1SMauro Carvalho Chehab 70089272ca1SMauro Carvalho ChehabAllocating disk space 70189272ca1SMauro Carvalho Chehab The default operation (i.e., mode is zero) of fallocate() allocates 70289272ca1SMauro Carvalho Chehab the disk space within the range specified by offset and len. The 70389272ca1SMauro Carvalho Chehab file size (as reported by stat(2)) will be changed if offset+len is 70489272ca1SMauro Carvalho Chehab greater than the file size. Any subregion within the range specified 70589272ca1SMauro Carvalho Chehab by offset and len that did not contain data before the call will be 70689272ca1SMauro Carvalho Chehab initialized to zero. This default behavior closely resembles the 70789272ca1SMauro Carvalho Chehab behavior of the posix_fallocate(3) library function, and is intended 70889272ca1SMauro Carvalho Chehab as a method of optimally implementing that function. 70989272ca1SMauro Carvalho Chehab 71089272ca1SMauro Carvalho ChehabHowever, once F2FS receives ioctl(fd, F2FS_IOC_SET_PIN_FILE) in prior to 71189272ca1SMauro Carvalho Chehabfallocate(fd, DEFAULT_MODE), it allocates on-disk blocks addressess having 71289272ca1SMauro Carvalho Chehabzero or random data, which is useful to the below scenario where: 71389272ca1SMauro Carvalho Chehab 71489272ca1SMauro Carvalho Chehab 1. create(fd) 71589272ca1SMauro Carvalho Chehab 2. ioctl(fd, F2FS_IOC_SET_PIN_FILE) 71689272ca1SMauro Carvalho Chehab 3. fallocate(fd, 0, 0, size) 71789272ca1SMauro Carvalho Chehab 4. address = fibmap(fd, offset) 71889272ca1SMauro Carvalho Chehab 5. open(blkdev) 71989272ca1SMauro Carvalho Chehab 6. write(blkdev, address) 72089272ca1SMauro Carvalho Chehab 72189272ca1SMauro Carvalho ChehabCompression implementation 72289272ca1SMauro Carvalho Chehab-------------------------- 72389272ca1SMauro Carvalho Chehab 72489272ca1SMauro Carvalho Chehab- New term named cluster is defined as basic unit of compression, file can 72589272ca1SMauro Carvalho Chehab be divided into multiple clusters logically. One cluster includes 4 << n 72689272ca1SMauro Carvalho Chehab (n >= 0) logical pages, compression size is also cluster size, each of 72789272ca1SMauro Carvalho Chehab cluster can be compressed or not. 72889272ca1SMauro Carvalho Chehab 72989272ca1SMauro Carvalho Chehab- In cluster metadata layout, one special block address is used to indicate 73089272ca1SMauro Carvalho Chehab cluster is compressed one or normal one, for compressed cluster, following 73189272ca1SMauro Carvalho Chehab metadata maps cluster to [1, 4 << n - 1] physical blocks, in where f2fs 73289272ca1SMauro Carvalho Chehab stores data including compress header and compressed data. 73389272ca1SMauro Carvalho Chehab 73489272ca1SMauro Carvalho Chehab- In order to eliminate write amplification during overwrite, F2FS only 73589272ca1SMauro Carvalho Chehab support compression on write-once file, data can be compressed only when 73689272ca1SMauro Carvalho Chehab all logical blocks in file are valid and cluster compress ratio is lower 73789272ca1SMauro Carvalho Chehab than specified threshold. 73889272ca1SMauro Carvalho Chehab 73989272ca1SMauro Carvalho Chehab- To enable compression on regular inode, there are three ways: 74089272ca1SMauro Carvalho Chehab 74189272ca1SMauro Carvalho Chehab * chattr +c file 74289272ca1SMauro Carvalho Chehab * chattr +c dir; touch dir/file 74389272ca1SMauro Carvalho Chehab * mount w/ -o compress_extension=ext; touch file.ext 74489272ca1SMauro Carvalho Chehab 74589272ca1SMauro Carvalho ChehabCompress metadata layout:: 74689272ca1SMauro Carvalho Chehab 74789272ca1SMauro Carvalho Chehab [Dnode Structure] 74889272ca1SMauro Carvalho Chehab +-----------------------------------------------+ 74989272ca1SMauro Carvalho Chehab | cluster 1 | cluster 2 | ......... | cluster N | 75089272ca1SMauro Carvalho Chehab +-----------------------------------------------+ 75189272ca1SMauro Carvalho Chehab . . . . 75289272ca1SMauro Carvalho Chehab . . . . 75389272ca1SMauro Carvalho Chehab . Compressed Cluster . . Normal Cluster . 75489272ca1SMauro Carvalho Chehab +----------+---------+---------+---------+ +---------+---------+---------+---------+ 75589272ca1SMauro Carvalho Chehab |compr flag| block 1 | block 2 | block 3 | | block 1 | block 2 | block 3 | block 4 | 75689272ca1SMauro Carvalho Chehab +----------+---------+---------+---------+ +---------+---------+---------+---------+ 75789272ca1SMauro Carvalho Chehab . . 75889272ca1SMauro Carvalho Chehab . . 75989272ca1SMauro Carvalho Chehab . . 76089272ca1SMauro Carvalho Chehab +-------------+-------------+----------+----------------------------+ 76189272ca1SMauro Carvalho Chehab | data length | data chksum | reserved | compressed data | 76289272ca1SMauro Carvalho Chehab +-------------+-------------+----------+----------------------------+ 763