.\" -*- nroff -*-
.\" Copyright 1993, 1994, 1995 by Theodore Ts'o.  All Rights Reserved.
.\" This file may be copied under the terms of the GNU Public License.
.\"
.TH DEBUGFS 8 "@E2FSPROGS_MONTH@ @E2FSPROGS_YEAR@" "E2fsprogs version @E2FSPROGS_VERSION@"
.SH NAME
debugfs \- ext2/ext3/ext4 file system debugger
.SH SYNOPSIS
.B debugfs
[
.B \-DVwci
]
[
.B \-b
blocksize
]
[
.B \-s
superblock
]
[
.B \-f
cmd_file
]
[
.B \-R
request
]
[
.B \-d
data_source_device
]
[
device
]
.SH DESCRIPTION
The
.B debugfs
program is an interactive file system debugger. It can be used to
examine and change the state of an ext2, ext3, or ext4 file system.
.br
.I device
is the special file corresponding to the device containing the
file system (e.g /dev/hdXX).
.SH OPTIONS
.TP
.I \-w
Specifies that the file system should be opened in read-write mode.
Without this option, the file system is opened in read-only mode.
.TP
.I \-c
Specifies that the file system should be opened in catastrophic mode, in
which the inode and group bitmaps are not read initially.  This can be
useful for filesystems with significant corruption, but because of this,
catastrophic mode forces the filesystem to be opened read-only.
.TP
.I \-i
Specifies that
.I device
represents an ext2 image file created by the
.B e2image
program.  Since the ext2 image file only contains the superblock, block
group descriptor, block and inode allocation bitmaps, and
the inode table, many
.B debugfs
commands will not function properly.
.B Warning:
no safety checks are in place, and
.B debugfs
may fail in interesting ways if commands such as
.IR ls ", " dump ", "
etc. are tried without specifying the
.I data_source_device
using the
.I \-d
option.
.B debugfs
is a debugging tool.  It has rough edges!
.TP
.I -d data_source_device
Used with the
.I \-i
option, specifies that
.I data_source_device
should be used when reading blocks not found in the ext2 image file.
This includes data, directory, and indirect blocks.
.TP
.I -b blocksize
Forces the use of the given block size for the file system, rather than
detecting the correct block size as normal.
.TP
.I -s superblock
Causes the file system superblock to be read from the given block
number, instead of using the primary superblock (located at an offset of
1024 bytes from the beginning of the filesystem).  If you specify the
.I -s
option, you must also provide the blocksize of the filesystem via the
.I -b
option.
.TP
.I -f cmd_file
Causes
.B debugfs
to read in commands from
.IR cmd_file ,
and execute them.  When
.B debugfs
is finished executing those commands, it will exit.
.TP
.I -D
Causes
.B debugfs
to open the device using Direct I/O, bypassing the buffer cache.  Note
that some Linux devices, notably device mapper as of this writing, do
not support Direct I/O.
.TP
.I -R request
Causes
.B debugfs
to execute the single command
.IR request ,
and then exit.
.TP
.I -V
print the version number of
.B debugfs
and exit.
.SH SPECIFYING FILES
Many
.B debugfs
commands take a
.I filespec
as an argument to specify an inode (as opposed to a pathname)
in the filesystem which is currently opened by
.BR debugfs .
The
.I filespec
argument may be specified in two forms.  The first form is an inode
number surrounded by angle brackets, e.g.,
.IR <2> .
The second form is a pathname; if the pathname is prefixed by a forward slash
('/'), then it is interpreted relative to the root of the filesystem
which is currently opened by
.BR debugfs .
If not, the pathname is
interpreted relative to the current working directory as maintained by
.BR debugfs .
This may be modified by using the
.B debugfs
command
.IR cd .
.\"
.\"
.\"
.SH COMMANDS
This is a list of the commands which
.B debugfs
supports.
.TP
.BI blocks " filespace"
Print the blocks used by the inode
.I filespec
to stdout.
.TP
.BI bmap " filespec logical_block"
Print the physical block number corresponding to the logical block number
.I logical_block
in the inode
.IR filespec .
.TP
.BI block_dump " [-f filespec] block_num"
Dump the filesystem block given by
.I block_num
in hex and ASCII format to the console.  If the
.I -f
option is specified, the block number is relative to the start of the given
.BR filespec .
.TP
.BI cat " filespec"
Dump the contents of the inode
.I filespec
to stdout.
.TP
.BI cd " filespec"
Change the current working directory to
.IR filespec .
.TP
.BI chroot " filespec"
Change the root directory to be the directory
.IR filespec .
.TP
.BI close " [-a]"
Close the currently open file system.  If the
.I -a
option is specified, write out any changes to the superblock and block
group descriptors to all of the backup superblocks, not just to the
master superblock.
.TP
.BI clri " filespec"
Clear the contents of the inode
.IR filespec .
.TP
.BI dirsearch " filespec filename"
Search the directory
.I filespec
for
.IR filename .
.TP
.B dirty
Mark the filesystem as dirty, so that the superblocks will be written on exit.
.TP
.BI dump " [-p] filespec out_file"
Dump the contents of the inode
.I filespec
to the output file
.IR out_file .
If the
.I -p
option is given set the owner, group and permissions information on
.I out_file
to match
.IR filespec .
.TP
.B dump_mmp
Display the multiple-mount protection (mmp) field values.
.TP
.BI dx_hash " [-h hash_alg] [-s hash_seed] filename"
Calculate the directory hash of
.IR filename .
The hash algorithm specified with
.I -h
may be
.BR legacy , " half_md4" ", or " tea .
The hash seed specified with
.I -s
must be in UUID format.
.TP
.BI dump_extents " [-n] [-l] filespec"
Dump the the extent tree of the inode
.IR filespec .
The
.I -n
flag will cause
.B dump_extents
to only display the interior nodes in the extent tree.   The
.I -l
flag will cause
.B dump_extents
to only display the leaf nodes in the extent tree.
.IP
(Please note that the length and range of blocks for the last extent in
an interior node is an estimate by the extents library functions, and is
not stored in filesystem data structures.   Hence, the values displayed
may not necessarily by accurate and does not indicate a problem or
corruption in the file system.)
.TP
.BI expand_dir " filespec"
Expand the directory
.IR filespec .
.TP
.BI feature " [fs_feature] [-fs_feature] ..."
Set or clear various filesystem features in the superblock.  After setting
or clearing any filesystem features that were requested, print the current
state of the filesystem feature set.
.TP
.BI filefrag " [-dvr] filespec"
Print the number of contiguous extents in
.IR filespec .
If
.I filespec
is a directory and the
.I -d
option is not specified,
.I filefrag
will print the number of contiguous extents for each file in
the directory.  The
.I -v
option will cause
.I filefrag
print a tabular listing of the contiguous extents in the
file.  The
.I -r
option will cause
.I filefrag
to do a recursive listing of the directory.
.TP
.BI find_free_block " [count [goal]]"
Find the first
.I count
free blocks, starting from
.I goal
and allocate it.  Also available as
.BR ffb .
.TP
.BI find_free_inode " [dir [mode]]"
Find a free inode and allocate it.  If present,
.I dir
specifies the inode number of the directory
which the inode is to be located.  The second
optional argument
.I mode
specifies the permissions of the new inode.  (If the directory bit is set
on the mode, the allocation routine will function differently.)  Also
available as
.BR ffi .
.TP
.BI freeb " block [count]"
Mark the block number
.I block
as not allocated.
If the optional argument
.I count
is present, then
.I count
blocks starting at block number
.I block
will be marked as not allocated.
.TP
.BI freefrag " [-c chunk_kb]"
Report free space fragmentation on the currently open file system.
If the
.I \-c
option is specified then the filefrag command will print how many free
chunks of size
.I chunk_kb
can be found in the file system.  The chunk size must be a power of two
and be larger than the file system block size.
.TP
.BI freei " filespec [num]"
Free the inode specified by
.IR filespec .
If
.I num
is specified, also clear num-1 inodes after the specified inode.
.TP
.B help
Print a list of commands understood by
.BR debugfs .
.TP
.BI htree_dump " filespec"
Dump the hash-indexed directory
.IR filespec ,
showing its tree structure.
.TP
.BI icheck " block ..."
Print a listing of the inodes which use the one or more blocks specified
on the command line.
.TP
.BI inode_dump " filespec"
Print the contents of the inode data structure in hex and ASCII format.
.TP
.BI imap " filespec"
Print the location of the inode data structure (in the inode table)
of the inode
.IR filespec .
.TP
.BI init_filesys " device blocksize"
Create an ext2 file system on
.I device
with device size
.IR blocksize .
Note that this does not fully initialize all of the data structures;
to do this, use the
.BR mke2fs (8)
program.  This is just a call to the low-level library, which sets up
the superblock and block descriptors.
.TP
.BI kill_file " filespec"
Deallocate the inode
.I filespec
and its blocks.  Note that this does not remove any directory
entries (if any) to this inode.  See the
.BR rm (1)
command if you wish to unlink a file.
.TP
.BI lcd " directory"
Change the current working directory of the
.B debugfs
process to
.I directory
on the native filesystem.
.TP
.BI ln " filespec dest_file"
Create a link named
.I dest_file
which is a hard link to
.IR filespec .
Note this does not adjust the inode reference counts.
.TP
.BI logdump " [-acs] [-b block] [-i filespec] [-f journal_file] [output_file]"
Dump the contents of the ext3 journal.  By default, dump the journal inode as
specified in the superblock.  However, this can be overridden with the
.I \-i
option, which dumps the journal from the internal inode given by
.IR filespec .
A regular file containing journal data can be specified using the
.I \-f
option.  Finally, the
.I \-s
option utilizes the backup information in the superblock to locate the
journal.
.IP
The
.I \-a
option causes the
.B logdump
program to print the contents of all of the descriptor blocks.
The
.I \-b
option causes
.B logdump
to print all journal records that are refer to the specified block.
The
.I \-c
option will print out the contents of all of the data blocks selected by
the
.I \-a
and
.I \-b
options.
.TP
.BI ls " [-d] [-l] [-p] filespec"
Print a listing of the files in the directory
.IR filespec .
The
.I \-d
flag will list deleted entries in the directory.
The
.I \-l
flag will list files using a more verbose format.
The
.I \-p
flag will list the files in a format which is more easily parsable by
scripts, as well as making it more clear when there are spaces or other
non-printing characters at the end of filenames.
.TP
.BI list_deleted_inodes " [limit]"
List deleted inodes, optionally limited to those deleted within
.I limit
seconds ago.  Also available as
.BR lsdel .
.IP
This command was useful for recovering from accidental file deletions
for ext2 file systems.  Unfortunately, it is not useful for this purpose
if the files were deleted using ext3 or ext4, since the inode's
data blocks are no longer available after the inode is released.
.TP
.BI modify_inode " filespec"
Modify the contents of the inode structure in the inode
.IR filespec .
Also available as
.BR mi .
.TP
.BI mkdir " filespec"
Make a directory.
.TP
.BI mknod " filespec [p|[[c|b] major minor]]"
Create a special device file (a named pipe, character or block device).
If a character or block device is to be made, the
.I major
and
.I minor
device numbers must be specified.
.TP
.BI ncheck " [-c] inode_num ..."
Take the requested list of inode numbers, and print a listing of pathnames
to those inodes.  The
.I -c
flag will enable checking the file type information in the directory
entry to make sure it matches the inode's type.
.TP
.BI open " [-weficD] [-b blocksize] [-s superblock] device"
Open a filesystem for editing.  The
.I -f
flag forces the filesystem to be opened even if there are some unknown
or incompatible filesystem features which would normally
prevent the filesystem from being opened.  The
.I -e
flag causes the filesystem to be opened in exclusive mode.  The
.IR -b ", " -c ", " -i ", " -s ", " -w ", and " -D
options behave the same as the command-line options to
.BR debugfs .
.TP
.BI punch " filespec start_blk [end_blk]"
Delete the blocks in the inode ranging from
.I start_blk
to
.IR end_blk .
If
.I end_blk
is omitted then this command will function as a truncate command; that
is, all of the blocks starting at
.I start_blk
through to the end of the file will be deallocated.
.TP
.BI symlink " filespec target"
Make a symbolic link.
.TP
.B pwd
Print the current working directory.
.TP
.B quit
Quit
.B debugfs
.TP
.BI rdump " directory[...] destination"
Recursively dump
.IR directory ,
or multiple
.IR directories ,
and all its contents (including regular files, symbolic links, and other
directories) into the named
.IR destination ,
which should be an existing directory on the native filesystem.
.TP
.BI rm " pathname"
Unlink
.IR pathname .
If this causes the inode pointed to by
.I pathname
to have no other references, deallocate the file.  This command functions
as the unlink() system call.
.I
.TP
.BI rmdir " filespec"
Remove the directory
.IR filespec .
.TP
.BI setb " block [count]"
Mark the block number
.I block
as allocated.
If the optional argument
.I count
is present, then
.I count
blocks starting at block number
.I block
will be marked as allocated.
.TP
.BI set_block_group " bgnum field value"
Modify the block group descriptor specified by
.I bgnum
so that the block group descriptor field
.I field
has value
.IR value .
Also available as
.BR set_bg .
.TP
.BI seti " filespec [num]"
Mark inode
.I filespec
as in use in the inode bitmap.  If
.I num
is specified, also set num-1 inodes after the specified inode.
.TP
.BI set_inode_field " filespec field value"
Modify the inode specified by
.I filespec
so that the inode field
.I field
has value
.I value.
The list of valid inode fields which can be set via this command
can be displayed by using the command:
.B set_inode_field -l
Also available as
.BR sif .
.TP
.BI set_mmp_value " field value"
Modify the multiple-mount protection (MMP) data so that the MMP field
.I field
has value
.I value.
The list of valid MMP fields which can be set via this command
can be displayed by using the command:
.B set_mmp_value -l
Also available as
.BR smmp .
.TP
.BI set_super_value " field value"
Set the superblock field
.I field
to
.I value.
The list of valid superblock fields which can be set via this command
can be displayed by using the command:
.B set_super_value -l
Also available as
.BR ssv .
.TP
.BI show_super_stats " [-h]"
List the contents of the super block and the block group descriptors.  If the
.I -h
flag is given, only print out the superblock contents. Also available as
.BR stats .
.TP
.BI stat " filespec"
Display the contents of the inode structure of the inode
.IR filespec .
.TP
.BI testb " block [count]"
Test if the block number
.I block
is marked as allocated in the block bitmap.
If the optional argument
.I count
is present, then
.I count
blocks starting at block number
.I block
will be tested.
.TP
.BI testi " filespec"
Test if the inode
.I filespec
is marked as allocated in the inode bitmap.
.TP
.BI undel " <inode_number> [pathname]"
Undelete the specified inode number (which must be surrounded by angle
brackets) so that it and its blocks are marked in use, and optionally
link the recovered inode to the specified pathname.  The
.B e2fsck
command should always be run after using the
.B undel
command to recover deleted files.
.IP
Note that if you are recovering a large number of deleted files, linking
the inode to a directory may require the directory to be expanded, which
could allocate a block that had been used by one of the
yet-to-be-undeleted files.  So it is safer to undelete all of the
inodes without specifying a destination pathname, and then in a separate
pass, use the debugfs
.B link
command to link the inode to the destination pathname, or use
.B e2fsck
to check the filesystem and link all of the recovered inodes to the
lost+found directory.
.TP
.BI unlink " pathname"
Remove the link specified by
.I pathname
to an inode.  Note this does not adjust the inode reference counts.
.TP
.BI write " source_file out_file"
Copy the contents of
.I source_file
into a newly-created file in the filesystem named
.IR out_file .
.TP
.BI zap_block " [-f filespec] [-o offset] [-l length] [-p pattern] block_num"
.TP
Overwrite the block specified by
.I block_num
with zero (NUL) bytes, or if
.I -p
is given use the byte specified by
.IR pattern .
If
.I -f
is given then
.I block_num
is relative to the start of the file given by
.IR filespec .
The
.I -o
and
.I -l
options limit the range of bytes to zap to the specified
.I offset
and
.I length
relative to the start of the block.
.TP
.BI zap_block " [-f filespec] [-b bit] block_num"
Bit-flip portions of the physical
.IR block_num .
If
.I -f
is given, then
.I block_num
is a logical block relative to the start of
.IR filespec .
.SH ENVIRONMENT VARIABLES
.TP
.B DEBUGFS_PAGER, PAGER
The
.B debugfs
program always pipes the output of the some commands through a
pager program.  These commands include:
.IR show_super_stats " (" stats ),
.IR list_directory " (" ls ),
.IR show_inode_info " (" stat ),
.IR list_deleted_inodes " (" lsdel ),
and
.IR htree_dump .
The specific pager can explicitly specified by the
.B DEBUGFS_PAGER
environment variable, and if it is not set, by the
.B PAGER
environment variable.
.IP
Note that since a pager is always used, the
.BR less (1)
pager is not particularly appropriate, since it clears the screen before
displaying the output of the command and clears the output the screen
when the pager is exited.  Many users prefer to use the
.BR less (1)
pager for most purposes, which is why the
.B DEBUGFS_PAGER
environment variable is available to override the more general
.B PAGER
environment variable.
.SH AUTHOR
.B debugfs
was written by Theodore Ts'o <tytso@mit.edu>.
.SH SEE ALSO
.BR dumpe2fs (8),
.BR tune2fs (8),
.BR e2fsck (8),
.BR mke2fs (8),
.BR ext4 (5)