ext4 General Information — The Linux Kernel documentation (2025)

Quick usage instructions¶

Note: More extensive information for getting started with ext4 can befound at the ext4 wiki site at the URL:http://ext4.wiki.kernel.org/index.php/Ext4_Howto

• The latest version of e2fsprogs can be found at:

  https://www.kernel.org/pub/linux/kernel/people/tytso/e2fsprogs/

  or

  http://sourceforge.net/project/showfiles.php?group_id=2406

  or grab the latest git repository from:

https://git.kernel.org/pub/scm/fs/ext2/e2fsprogs.git

• Create a new filesystem using the ext4 filesystem type:

  # mke2fs -t ext4 /dev/hda1

  Or to configure an existing ext3 filesystem to support extents:

  # tune2fs -O extents /dev/hda1

  If the filesystem was created with 128 byte inodes, it can beconverted to use 256 byte for greater efficiency via:

  # tune2fs -I 256 /dev/hda1

• Mounting:

  # mount -t ext4 /dev/hda1 /wherever

• When comparing performance with other filesystems, it’s alwaysimportant to try multiple workloads; very often a subtle change in aworkload parameter can completely change the ranking of whichfilesystems do well compared to others. When comparing versus ext3,note that ext4 enables write barriers by default, while ext3 doesnot enable write barriers by default. So it is useful to useexplicitly specify whether barriers are enabled or not when via the‘-o barriers=[0|1]’ mount option for both ext3 and ext4 filesystemsfor a fair comparison. When tuning ext3 for best benchmark numbers,it is often worthwhile to try changing the data journaling mode; ‘-odata=writeback’ can be faster for some workloads. (Note however thatrunning mounted with data=writeback can potentially leave stale dataexposed in recently written files in case of an unclean shutdown,which could be a security exposure in some situations.) Configuringthe filesystem with a large journal can also be helpful formetadata-intensive workloads.

Features¶

case-insensitive file name lookups¶

The case-insensitive file name lookup feature is supported on aper-directory basis, allowing the user to mix case-insensitive andcase-sensitive directories in the same filesystem. It is enabled byflipping the +F inode attribute of an empty directory. Thecase-insensitive string match operation is only defined when we know howtext in encoded in a byte sequence. For that reason, in order to enablecase-insensitive directories, the filesystem must have thecasefold feature, which stores the filesystem-wide encodingmodel used. By default, the charset adopted is the latest version ofUnicode (12.1.0, by the time of this writing), encoded in the UTF-8form. The comparison algorithm is implemented by normalizing thestrings to the Canonical decomposition form, as defined by Unicode,followed by a byte per byte comparison.

The case-awareness is name-preserving on the disk, meaning that the filename provided by userspace is a byte-per-byte match to what is actuallywritten in the disk. The Unicode normalization format used by thekernel is thus an internal representation, and not exposed to theuserspace nor to the disk, with the important exception of disk hashes,used on large case-insensitive directories with DX feature. On DXdirectories, the hash must be calculated using the casefolded version ofthe filename, meaning that the normalization format used actually has animpact on where the directory entry is stored.

When we change from viewing filenames as opaque byte sequences to seeingthem as encoded strings we need to address what happens when a programtries to create a file with an invalid name. The Unicode subsystemwithin the kernel leaves the decision of what to do in this case to thefilesystem, which select its preferred behavior by enabling/disablingthe strict mode. When Ext4 encounters one of those strings and thefilesystem did not require strict mode, it falls back to considering theentire string as an opaque byte sequence, which still allows the user tooperate on that file, but the case-insensitive lookups won’t work.

Options¶

When mounting an ext4 filesystem, the following option are accepted:(*) == default

ro

Mount filesystem read only. Note that ext4 will replay the journal (andthus write to the partition) even when mounted “read only”. The mountoptions “ro,noload” can be used to prevent writes to the filesystem.

journal_checksum

Enable checksumming of the journal transactions. This will allow therecovery code in e2fsck and the kernel to detect corruption in thekernel. It is a compatible change and will be ignored by olderkernels.

journal_async_commit

Commit block can be written to disk without waiting for descriptorblocks. If enabled older kernels cannot mount the device. This willenable ‘journal_checksum’ internally.

journal_path=path, journal_dev=devnum

When the external journal device’s major/minor numbers have changed,these options allow the user to specify the new journal location. Thejournal device is identified through either its new major/minor numbersencoded in devnum, or via a path to the device.

norecovery, noload

Don’t load the journal on mounting. Note that if the filesystem wasnot unmounted cleanly, skipping the journal replay will lead to thefilesystem containing inconsistencies that can lead to any number ofproblems.

data=journal

All data are committed into the journal prior to being written into themain file system. Enabling this mode will disable delayed allocationand O_DIRECT support.

data=ordered (*)

All data are forced directly out to the main file system prior to itsmetadata being committed to the journal.

data=writeback

Data ordering is not preserved, data may be written into the main filesystem after its metadata has been committed to the journal.

commit=nrsec (*)

This setting limits the maximum age of the running transaction to‘nrsec’ seconds. The default value is 5 seconds. This means that ifyou lose your power, you will lose as much as the latest 5 seconds ofmetadata changes (your filesystem will not be damaged though, thanksto the journaling). This default value (or any low value) will hurtperformance, but it’s good for data-safety. Setting it to 0 will havethe same effect as leaving it at the default (5 seconds). Setting itto very large values will improve performance. Note that due todelayed allocation even older data can be lost on power failure sincewriteback of those data begins only after time set in/proc/sys/vm/dirty_expire_centisecs.

barrier=<0|1(*)>, barrier(*), nobarrier

This enables/disables the use of write barriers in the jbd code.barrier=0 disables, barrier=1 enables. This also requires an IO stackwhich can support barriers, and if jbd gets an error on a barrierwrite, it will disable again with a warning. Write barriers enforceproper on-disk ordering of journal commits, making volatile disk writecaches safe to use, at some performance penalty. If your disks arebattery-backed in one way or another, disabling barriers may safelyimprove performance. The mount options “barrier” and “nobarrier” canalso be used to enable or disable barriers, for consistency with otherext4 mount options.

inode_readahead_blks=n

This tuning parameter controls the maximum number of inode table blocksthat ext4’s inode table readahead algorithm will pre-read into thebuffer cache. The default value is 32 blocks.

nouser_xattr

Disables Extended User Attributes. See the attr(5) manual page formore information about extended attributes.

noacl

This option disables POSIX Access Control List support. If ACL supportis enabled in the kernel configuration (CONFIG_EXT4_FS_POSIX_ACL), ACLis enabled by default on mount. See the acl(5) manual page for moreinformation about acl.

bsddf (*)

Make ‘df’ act like BSD.

minixdf

Make ‘df’ act like Minix.

debug

Extra debugging information is sent to syslog.

abort

Simulate the effects of calling ext4_abort() for debugging purposes.This is normally used while remounting a filesystem which is alreadymounted.

errors=remount-ro

Remount the filesystem read-only on an error.

errors=continue

Keep going on a filesystem error.

errors=panic

Panic and halt the machine if an error occurs. (These mount optionsoverride the errors behavior specified in the superblock, which can beconfigured using tune2fs)

data_err=ignore(*)

Just print an error message if an error occurs in a file data buffer inordered mode.

data_err=abort

Abort the journal if an error occurs in a file data buffer in orderedmode.

grpid | bsdgroups

New objects have the group ID of their parent.

nogrpid (*) | sysvgroups

New objects have the group ID of their creator.

resgid=n

The group ID which may use the reserved blocks.

resuid=n

The user ID which may use the reserved blocks.

sb=

Use alternate superblock at this location.

quota, noquota, grpquota, usrquota

These options are ignored by the filesystem. They are used only byquota tools to recognize volumes where quota should be turned on. Seedocumentation in the quota-tools package for more details(http://sourceforge.net/projects/linuxquota).

jqfmt=<quota type>, usrjquota=<file>, grpjquota=<file>

These options tell filesystem details about quota so that quotainformation can be properly updated during journal replay. They replacethe above quota options. See documentation in the quota-tools packagefor more details (http://sourceforge.net/projects/linuxquota).

stripe=n

Number of filesystem blocks that mballoc will try to use for allocationsize and alignment. For RAID5/6 systems this should be the number ofdata disks * RAID chunk size in file system blocks.

delalloc (*)

Defer block allocation until just before ext4 writes out the block(s)in question. This allows ext4 to better allocation decisions moreefficiently.

nodelalloc

Disable delayed allocation. Blocks are allocated when the data iscopied from userspace to the page cache, either via the write(2) systemcall or when an mmap’ed page which was previously unallocated iswritten for the first time.

max_batch_time=usec

Maximum amount of time ext4 should wait for additional filesystemoperations to be batch together with a synchronous write operation.Since a synchronous write operation is going to force a commit and thena wait for the I/O complete, it doesn’t cost much, and can be a hugethroughput win, we wait for a small amount of time to see if any othertransactions can piggyback on the synchronous write. The algorithmused is designed to automatically tune for the speed of the disk, bymeasuring the amount of time (on average) that it takes to finishcommitting a transaction. Call this time the “commit time”. If thetime that the transaction has been running is less than the committime, ext4 will try sleeping for the commit time to see if otheroperations will join the transaction. The commit time is capped bythe max_batch_time, which defaults to 15000us (15ms). Thisoptimization can be turned off entirely by setting max_batch_time to 0.

min_batch_time=usec

This parameter sets the commit time (as described above) to be at leastmin_batch_time. It defaults to zero microseconds. Increasing thisparameter may improve the throughput of multi-threaded, synchronousworkloads on very fast disks, at the cost of increasing latency.

journal_ioprio=prio

The I/O priority (from 0 to 7, where 0 is the highest priority) whichshould be used for I/O operations submitted by kjournald2 during acommit operation. This defaults to 3, which is a slightly higherpriority than the default I/O priority.

auto_da_alloc(*), noauto_da_alloc

Many broken applications don’t use fsync() when replacing existingfiles via patterns such as fd = open(“foo.new”)/write(fd,..)/close(fd)/rename(“foo.new”, “foo”), or worse yet, fd = open(“foo”,O_TRUNC)/write(fd,..)/close(fd). If auto_da_alloc is enabled, ext4will detect the replace-via-rename and replace-via-truncate patternsand force that any delayed allocation blocks are allocated such that atthe next journal commit, in the default data=ordered mode, the datablocks of the new file are forced to disk before the rename() operationis committed. This provides roughly the same level of guarantees asext3, and avoids the “zero-length” problem that can happen when asystem crashes before the delayed allocation blocks are forced to disk.

noinit_itable

Do not initialize any uninitialized inode table blocks in thebackground. This feature may be used by installation CD’s so that theinstall process can complete as quickly as possible; the inode tableinitialization process would then be deferred until the next time thefile system is unmounted.

init_itable=n

The lazy itable init code will wait n times the number of millisecondsit took to zero out the previous block group’s inode table. Thisminimizes the impact on the system performance while file system’sinode table is being initialized.

discard, nodiscard(*)

Controls whether ext4 should issue discard/TRIM commands to theunderlying block device when blocks are freed. This is useful for SSDdevices and sparse/thinly-provisioned LUNs, but it is off by defaultuntil sufficient testing has been done.

nouid32

Disables 32-bit UIDs and GIDs. This is for interoperability witholder kernels which only store and expect 16-bit values.

block_validity(*), noblock_validity

These options enable or disable the in-kernel facility for trackingfilesystem metadata blocks within internal data structures. Thisallows multi- block allocator and other routines to notice bugs orcorrupted allocation bitmaps which cause blocks to be allocated whichoverlap with filesystem metadata blocks.

dioread_lock, dioread_nolock

Controls whether or not ext4 should use the DIO read locking. If thedioread_nolock option is specified ext4 will allocate uninitializedextent before buffer write and convert the extent to initialized afterIO completes. This approach allows ext4 code to avoid using inodemutex, which improves scalability on high speed storages. However thisdoes not work with data journaling and dioread_nolock option will beignored with kernel warning. Note that dioread_nolock code path is onlyused for extent-based files. Because of the restrictions this optionscomprises it is off by default (e.g. dioread_lock).

max_dir_size_kb=n

This limits the size of directories so that any attempt to expand thembeyond the specified limit in kilobytes will cause an ENOSPC error.This is useful in memory constrained environments, where a very largedirectory can cause severe performance problems or even provoke the OutOf Memory killer. (For example, if there is only 512mb memoryavailable, a 176mb directory may seriously cramp the system’s style.)

i_version

Enable 64-bit inode version support. This option is off by default.

dax

Use direct access (no page cache). SeeDirect Access for files. Note that this option isincompatible with data=journal.

inlinecrypt

When possible, encrypt/decrypt the contents of encrypted files using theblk-crypto framework rather than filesystem-layer encryption. Thisallows the use of inline encryption hardware. The on-disk format isunaffected. For more details, seeInline Encryption.

Data Mode¶

There are 3 different data modes:

• writeback mode

  In data=writeback mode, ext4 does not journal data at all. This mode providesa similar level of journaling as that of XFS, JFS, and ReiserFS in its defaultmode - metadata journaling. A crash+recovery can cause incorrect data toappear in files which were written shortly before the crash. This mode willtypically provide the best ext4 performance.

• ordered mode

  In data=ordered mode, ext4 only officially journals metadata, but it logicallygroups metadata information related to data changes with the data blocks intoa single unit called a transaction. When it’s time to write the new metadataout to disk, the associated data blocks are written first. In general, thismode performs slightly slower than writeback but significantly faster thanjournal mode.

• journal mode

  data=journal mode provides full data and metadata journaling. All new data iswritten to the journal first, and then to its final location. In the event ofa crash, the journal can be replayed, bringing both data and metadata into aconsistent state. This mode is the slowest except when data needs to be readfrom and written to disk at the same time where it outperforms all othersmodes. Enabling this mode will disable delayed allocation and O_DIRECTsupport.

/proc entries¶

Information about mounted ext4 file systems can be found in/proc/fs/ext4. Each mounted filesystem will have a directory in/proc/fs/ext4 based on its device name (i.e., /proc/fs/ext4/hdc or/proc/fs/ext4/dm-0). The files in each per-device directory are shownin table below.

Files in /proc/fs/ext4/<devname>

mb_groups

details of multiblock allocator buddy cache of free blocks

/sys entries¶

Information about mounted ext4 file systems can be found in/sys/fs/ext4. Each mounted filesystem will have a directory in/sys/fs/ext4 based on its device name (i.e., /sys/fs/ext4/hdc or/sys/fs/ext4/dm-0). The files in each per-device directory are shownin table below.

Files in /sys/fs/ext4/<devname>:

(see also Documentation/ABI/testing/sysfs-fs-ext4)

delayed_allocation_blocks

This file is read-only and shows the number of blocks that are dirty inthe page cache, but which do not have their location in the filesystemallocated yet.

inode_goal

Tuning parameter which (if non-zero) controls the goal inode used bythe inode allocator in preference to all other allocation heuristics.This is intended for debugging use only, and should be 0 on productionsystems.

inode_readahead_blks

Tuning parameter which controls the maximum number of inode tableblocks that ext4’s inode table readahead algorithm will pre-read intothe buffer cache.

lifetime_write_kbytes

This file is read-only and shows the number of kilobytes of data thathave been written to this filesystem since it was created.

max_writeback_mb_bump

The maximum number of megabytes the writeback code will try to writeout before move on to another inode.

mb_group_prealloc

The multiblock allocator will round up allocation requests to amultiple of this tuning parameter if the stripe size is not set in theext4 superblock

mb_max_to_scan

The maximum number of extents the multiblock allocator will search tofind the best extent.

mb_min_to_scan

The minimum number of extents the multiblock allocator will search tofind the best extent.

mb_order2_req

Tuning parameter which controls the minimum size for requests (as apower of 2) where the buddy cache is used.

mb_stats

Controls whether the multiblock allocator should collect statistics,which are shown during the unmount. 1 means to collect statistics, 0means not to collect statistics.

mb_stream_req

Files which have fewer blocks than this tunable parameter will havetheir blocks allocated out of a block group specific preallocationpool, so that small files are packed closely together. Each large filewill have its blocks allocated out of its own unique preallocationpool.

session_write_kbytes

This file is read-only and shows the number of kilobytes of data thathave been written to this filesystem since it was mounted.

reserved_clusters

This is RW file and contains number of reserved clusters in the filesystem which will be used in the specific situations to avoid costlyzeroout, unexpected ENOSPC, or possible data loss. The default is 2% or4096 clusters, whichever is smaller and this can be changed however itcan never exceed number of clusters in the file system. If there is notenough space for the reserved space when mounting the file mount will_not_ fail.

Ioctls¶

Ext4 implements various ioctls which can be used by applications to accessext4-specific functionality. An incomplete list of these ioctls is shown in thetable below. This list includes truly ext4-specific ioctls (EXT4_IOC_*) aswell as ioctls that may have been ext4-specific originally but are now supportedby some other filesystem(s) too (FS_IOC_*).

Table of Ext4 ioctls

FS_IOC_GETFLAGS

Get additional attributes associated with inode. The ioctl argument isan integer bitfield, with bit values described in ext4.h.

FS_IOC_SETFLAGS

Set additional attributes associated with inode. The ioctl argument isan integer bitfield, with bit values described in ext4.h.

EXT4_IOC_GETVERSION, EXT4_IOC_GETVERSION_OLD

Get the inode i_generation number stored for each inode. Thei_generation number is normally changed only when new inode is createdand it is particularly useful for network filesystems. The ‘_OLD’version of this ioctl is an alias for FS_IOC_GETVERSION.

EXT4_IOC_SETVERSION, EXT4_IOC_SETVERSION_OLD

Set the inode i_generation number stored for each inode. The ‘_OLD’version of this ioctl is an alias for FS_IOC_SETVERSION.

EXT4_IOC_GROUP_EXTEND

This ioctl has the same purpose as the resize mount option. It allowsto resize filesystem to the end of the last existing block group,further resize has to be done with resize2fs, either online, oroffline. The argument points to the unsigned logn number representingthe filesystem new block count.

EXT4_IOC_MOVE_EXT

Move the block extents from orig_fd (the one this ioctl is pointing to)to the donor_fd (the one specified in move_extent structure passed asan argument to this ioctl). Then, exchange inode metadata betweenorig_fd and donor_fd. This is especially useful for onlinedefragmentation, because the allocator has the opportunity to allocatemoved blocks better, ideally into one contiguous extent.

EXT4_IOC_GROUP_ADD

Add a new group descriptor to an existing or new group descriptorblock. The new group descriptor is described by ext4_new_group_inputstructure, which is passed as an argument to this ioctl. This isespecially useful in conjunction with EXT4_IOC_GROUP_EXTEND, whichallows online resize of the filesystem to the end of the last existingblock group. Those two ioctls combined is used in userspace onlineresize tool (e.g. resize2fs).

EXT4_IOC_MIGRATE

This ioctl operates on the filesystem itself. It converts (migrates)ext3 indirect block mapped inode to ext4 extent mapped inode by walkingthrough indirect block mapping of the original inode and convertingcontiguous block ranges into ext4 extents of the temporary inode. Then,inodes are swapped. This ioctl might help, when migrating from ext3 toext4 filesystem, however suggestion is to create fresh ext4 filesystemand copy data from the backup. Note, that filesystem has to supportextents for this ioctl to work.

EXT4_IOC_ALLOC_DA_BLKS

Force all of the delay allocated blocks to be allocated to preserveapplication-expected ext3 behaviour. Note that this will also starttriggering a write of the data blocks, but this behaviour may change inthe future as it is not necessary and has been done this way only forsake of simplicity.

EXT4_IOC_RESIZE_FS

Resize the filesystem to a new size. The number of blocks of resizedfilesystem is passed in via 64 bit integer argument. The kernelallocates bitmaps and inode table, the userspace tool thus just passesthe new number of blocks.

EXT4_IOC_SWAP_BOOT

Swap i_blocks and associated attributes (like i_blocks, i_size,i_flags, ...) from the specified inode with inode EXT4_BOOT_LOADER_INO(#5). This is typically used to store a boot loader in a secure part ofthe filesystem, where it can’t be changed by a normal user by accident.The data blocks of the previous boot loader will be associated with thegiven inode.

References¶

kernel source: <file:fs/ext4/>

<file:fs/jbd2/>

programs: http://e2fsprogs.sourceforge.net/

useful links: https://fedoraproject.org/wiki/ext3-devel

http://www.bullopensource.org/ext4/http://ext4.wiki.kernel.org/index.php/Main_Pagehttps://fedoraproject.org/wiki/Features/Ext4