Provided by: libarchive-tools_3.2.2-3.1ubuntu0.7_amd64 
NAME
tar — manipulate tape archives
SYNOPSIS
tar [bundled-flags ⟨args⟩] [⟨file⟩ | ⟨pattern⟩ ...]
tar {-c} [options] [files | directories]
tar {-r | -u} -f archive-file [options] [files | directories]
tar {-t | -x} [options] [patterns]
DESCRIPTION
tar creates and manipulates streaming archive files. This implementation can extract from tar, pax,
cpio, zip, jar, ar, xar, rpm, 7-zip, and ISO 9660 cdrom images and can create tar, pax, cpio, ar, zip,
7-zip, and shar archives.
The first synopsis form shows a “bundled” option word. This usage is provided for compatibility with
historical implementations. See COMPATIBILITY below for details.
The other synopsis forms show the preferred usage. The first option to tar is a mode indicator from the
following list:
-c Create a new archive containing the specified items. The long option form is --create.
-r Like -c, but new entries are appended to the archive. Note that this only works on uncompressed
archives stored in regular files. The -f option is required. The long option form is --append.
-t List archive contents to stdout. The long option form is --list.
-u Like -r, but new entries are added only if they have a modification date newer than the
corresponding entry in the archive. Note that this only works on uncompressed archives stored in
regular files. The -f option is required. The long form is --update.
-x Extract to disk from the archive. If a file with the same name appears more than once in the
archive, each copy will be extracted, with later copies overwriting (replacing) earlier copies.
The long option form is --extract.
In -c, -r, or -u mode, each specified file or directory is added to the archive in the order specified on
the command line. By default, the contents of each directory are also archived.
In extract or list mode, the entire command line is read and parsed before the archive is opened. The
pathnames or patterns on the command line indicate which items in the archive should be processed.
Patterns are shell-style globbing patterns as documented in tcsh(1).
OPTIONS
Unless specifically stated otherwise, options are applicable in all operating modes.
@archive
(c and r mode only) The specified archive is opened and the entries in it will be appended to the
current archive. As a simple example,
tar -c -f - newfile @original.tar
writes a new archive to standard output containing a file newfile and all of the entries from
original.tar. In contrast,
tar -c -f - newfile original.tar
creates a new archive with only two entries. Similarly,
tar -czf - --format pax @-
reads an archive from standard input (whose format will be determined automatically) and converts
it into a gzip-compressed pax-format archive on stdout. In this way, tar can be used to convert
archives from one format to another.
-a, --auto-compress
(c mode only) Use the archive suffix to decide a set of the format and the compressions. As a
simple example,
tar -a -cf archive.tgz source.c source.h
creates a new archive with restricted pax format and gzip compression,
tar -a -cf archive.tar.bz2.uu source.c source.h
creates a new archive with restricted pax format and bzip2 compression and uuencode compression,
tar -a -cf archive.zip source.c source.h
creates a new archive with zip format,
tar -a -jcf archive.tgz source.c source.h
ignores the “-j” option, and creates a new archive with restricted pax format and gzip
compression,
tar -a -jcf archive.xxx source.c source.h
if it is unknown suffix or no suffix, creates a new archive with restricted pax format and bzip2
compression.
-B, --read-full-blocks
Ignored for compatibility with other tar(1) implementations.
-b blocksize, --block-size blocksize
Specify the block size, in 512-byte records, for tape drive I/O. As a rule, this argument is
only needed when reading from or writing to tape drives, and usually not even then as the default
block size of 20 records (10240 bytes) is very common.
-C directory, --cd directory, --directory directory
In c and r mode, this changes the directory before adding the following files. In x mode, change
directories after opening the archive but before extracting entries from the archive.
--chroot
(x mode only) chroot() to the current directory after processing any -C options and before
extracting any files.
--clear-nochange-fflags
(x mode only) Before removing file system objects to replace them, clear platform-specific file
flags that might prevent removal.
--disable-copyfile
Mac OS X specific. Disable the use of copyfile(3).
--exclude pattern
Do not process files or directories that match the specified pattern. Note that exclusions take
precedence over patterns or filenames specified on the command line.
--format format
(c, r, u mode only) Use the specified format for the created archive. Supported formats include
“cpio”, “pax”, “shar”, and “ustar”. Other formats may also be supported; see
libarchive-formats(5) for more information about currently-supported formats. In r and u modes,
when extending an existing archive, the format specified here must be compatible with the format
of the existing archive on disk.
-f file, --file file
Read the archive from or write the archive to the specified file. The filename can be - for
standard input or standard output. The default varies by system; on FreeBSD, the default is
/dev/sa0; on Linux, the default is /dev/st0.
--gid id
Use the provided group id number. On extract, this overrides the group id in the archive; the
group name in the archive will be ignored. On create, this overrides the group id read from
disk; if --gname is not also specified, the group name will be set to match the group id.
--gname name
Use the provided group name. On extract, this overrides the group name in the archive; if the
provided group name does not exist on the system, the group id (from the archive or from the
--gid option) will be used instead. On create, this sets the group name that will be stored in
the archive; the name will not be verified against the system group database.
-H (c and r mode only) Symbolic links named on the command line will be followed; the target of the
link will be archived, not the link itself.
-h (c and r mode only) Synonym for -L.
-I Synonym for -T.
--help Show usage.
--hfsCompression
(x mode only) Mac OS X specific(v10.6 or later). Compress extracted regular files with HFS+
compression.
--ignore-zeros
An alias of --options read_concatenated_archives for compatibility with GNU tar.
--include pattern
Process only files or directories that match the specified pattern. Note that exclusions
specified with --exclude take precedence over inclusions. If no inclusions are explicitly
specified, all entries are processed by default. The --include option is especially useful when
filtering archives. For example, the command
tar -c -f new.tar --include='*foo*' @old.tgz
creates a new archive new.tar containing only the entries from old.tgz containing the string
‘foo’.
-J, --xz
(c mode only) Compress the resulting archive with xz(1). In extract or list modes, this option
is ignored. Note that, unlike other tar implementations, this implementation recognizes XZ
compression automatically when reading archives.
-j, --bzip, --bzip2, --bunzip2
(c mode only) Compress the resulting archive with bzip2(1). In extract or list modes, this
option is ignored. Note that, unlike other tar implementations, this implementation recognizes
bzip2 compression automatically when reading archives.
-k, --keep-old-files
(x mode only) Do not overwrite existing files. In particular, if a file appears more than once
in an archive, later copies will not overwrite earlier copies.
--keep-newer-files
(x mode only) Do not overwrite existing files that are newer than the versions appearing in the
archive being extracted.
-L, --dereference
(c and r mode only) All symbolic links will be followed. Normally, symbolic links are archived
as such. With this option, the target of the link will be archived instead.
-l, --check-links
(c and r modes only) Issue a warning message unless all links to each file are archived.
--lrzip
(c mode only) Compress the resulting archive with lrzip(1). In extract or list modes, this
option is ignored.
--lz4 (c mode only) Compress the archive with lz4-compatible compression before writing it. In input
mode, this option is ignored; lz4 compression is recognized automatically on input.
--lzma (c mode only) Compress the resulting archive with the original LZMA algorithm. Use of this
option is discouraged and new archives should be created with --xz instead. Note that, unlike
other tar implementations, this implementation recognizes LZMA compression automatically when
reading archives.
--lzop (c mode only) Compress the resulting archive with lzop(1). In extract or list modes, this option
is ignored.
-m, --modification-time
(x mode only) Do not extract modification time. By default, the modification time is set to the
time stored in the archive.
-n, --norecurse, --no-recursion
(c, r, u modes only) Do not recursively archive the contents of directories.
--newer date
(c, r, u modes only) Only include files and directories newer than the specified date. This
compares ctime entries.
--newer-mtime date
(c, r, u modes only) Like --newer, except it compares mtime entries instead of ctime entries.
--newer-than file
(c, r, u modes only) Only include files and directories newer than the specified file. This
compares ctime entries.
--newer-mtime-than file
(c, r, u modes only) Like --newer-than, except it compares mtime entries instead of ctime
entries.
--nodump
(c and r modes only) Honor the nodump file flag by skipping this file.
--nopreserveHFSCompression
(x mode only) Mac OS X specific(v10.6 or later). Do not compress extracted regular files which
were compressed with HFS+ compression before archived. By default, compress the regular files
again with HFS+ compression.
--null (use with -I or -T) Filenames or patterns are separated by null characters, not by newlines.
This is often used to read filenames output by the -print0 option to find(1).
--no-same-owner
(x mode only) Do not extract owner and group IDs. This is the reverse of --same-owner and the
default behavior if tar is run as non-root.
--no-same-permissions
(x mode only) Do not extract full permissions (SGID, SUID, sticky bit, ACLs, extended attributes
or extended file flags). This is the reverse of -p and the default behavior if tar is run as
non-root.
--numeric-owner
This is equivalent to --uname "" --gname "". On extract, it causes user and group names in the
archive to be ignored in favor of the numeric user and group ids. On create, it causes user and
group names to not be stored in the archive.
-O, --to-stdout
(x, t modes only) In extract (-x) mode, files will be written to standard out rather than being
extracted to disk. In list (-t) mode, the file listing will be written to stderr rather than the
usual stdout.
-o (x mode) Use the user and group of the user running the program rather than those specified in
the archive. Note that this has no significance unless -p is specified, and the program is being
run by the root user. In this case, the file modes and flags from the archive will be restored,
but ACLs or owner information in the archive will be discarded.
-o (c, r, u mode) A synonym for --format ustar
--older date
(c, r, u modes only) Only include files and directories older than the specified date. This
compares ctime entries.
--older-mtime date
(c, r, u modes only) Like --older, except it compares mtime entries instead of ctime entries.
--older-than file
(c, r, u modes only) Only include files and directories older than the specified file. This
compares ctime entries.
--older-mtime-than file
(c, r, u modes only) Like --older-than, except it compares mtime entries instead of ctime
entries.
--one-file-system
(c, r, and u modes) Do not cross mount points.
--options options
Select optional behaviors for particular modules. The argument is a text string containing
comma-separated keywords and values. These are passed to the modules that handle particular
formats to control how those formats will behave. Each option has one of the following forms:
key=value
The key will be set to the specified value in every module that supports it. Modules
that do not support this key will ignore it.
key The key will be enabled in every module that supports it. This is equivalent to key=1.
!key The key will be disabled in every module that supports it.
module:key=value, module:key, module:!key
As above, but the corresponding key and value will be provided only to modules whose name
matches module.
The currently supported modules and keys are:
iso9660:joliet
Support Joliet extensions. This is enabled by default, use !joliet or iso9660:!joliet to
disable.
iso9660:rockridge
Support Rock Ridge extensions. This is enabled by default, use !rockridge or
iso9660:!rockridge to disable.
gzip:compression-level
A decimal integer from 1 to 9 specifying the gzip compression level.
gzip:timestamp
Store timestamp. This is enabled by default, use !timestamp or gzip:!timestamp to
disable.
lrzip:compression=type
Use type as compression method. Supported values are bzip2, gzip, lzo (ultra fast), and
zpaq (best, extremely slow).
lrzip:compression-level
A decimal integer from 1 to 9 specifying the lrzip compression level.
lz4:compression-level
A decimal integer from 1 to 9 specifying the lzop compression level.
lz4:stream-checksum
Enable stream checksum. This is by default, use lz4:!stream-checksum to disable.
lz4:block-checksum
Enable block checksum (Disabled by default).
lz4:block-size
A decimal integer from 4 to 7 specifying the lz4 compression block size (7 is set by
default).
lz4:block-dependence
Use the previous block of the block being compressed for a compression dictionary to
improve compression ratio.
lzop:compression-level
A decimal integer from 1 to 9 specifying the lzop compression level.
xz:compression-level
A decimal integer from 0 to 9 specifying the xz compression level.
mtree:keyword
The mtree writer module allows you to specify which mtree keywords will be included in
the output. Supported keywords include: cksum, device, flags, gid, gname, indent, link,
md5, mode, nlink, rmd160, sha1, sha256, sha384, sha512, size, time, uid, uname. The
default is equivalent to: “device, flags, gid, gname, link, mode, nlink, size, time,
type, uid, uname”.
mtree:all
Enables all of the above keywords. You can also use mtree:!all to disable all keywords.
mtree:use-set
Enable generation of /set lines in the output.
mtree:indent
Produce human-readable output by indenting options and splitting lines to fit into 80
columns.
zip:compression=type
Use type as compression method. Supported values are store (uncompressed) and deflate
(gzip algorithm).
zip:encryption
Enable encryption using traditional zip encryption.
zip:encryption=type
Use type as encryption type. Supported values are zipcrypt (traditional zip encryption),
aes128 (WinZip AES-128 encryption) and aes256 (WinZip AES-256 encryption).
read_concatenated_archives
Ignore zeroed blocks in the archive, which occurs when multiple tar archives have been
concatenated together. Without this option, only the contents of the first concatenated
archive would be read. This option is comparable to the -i, --ignore-zeros option of GNU
tar.
If a provided option is not supported by any module, that is a fatal error.
-P, --absolute-paths
Preserve pathnames. By default, absolute pathnames (those that begin with a / character) have
the leading slash removed both when creating archives and extracting from them. Also, tar will
refuse to extract archive entries whose pathnames contain .. or whose target directory would be
altered by a symlink. This option suppresses these behaviors.
-p, --insecure, --preserve-permissions
(x mode only) Preserve file permissions. Attempt to restore the full permissions, including
owner, file modes, file flags and ACLs, if available, for each item extracted from the archive.
This is the default, if tar is being run by root and can be overridden by also specifying
--no-same-owner and --no-same-permissions.
--passphrase passphrase
The passphrase is used to extract or create an encrypted archive. Currently, zip is the only
supported format that supports encryption. You shouldn't use this option unless you realize how
insecure use of this option is.
--posix
(c, r, u mode only) Synonym for --format pax
-q, --fast-read
(x and t mode only) Extract or list only the first archive entry that matches each pattern or
filename operand. Exit as soon as each specified pattern or filename has been matched. By
default, the archive is always read to the very end, since there can be multiple entries with the
same name and, by convention, later entries overwrite earlier entries. This option is provided
as a performance optimization.
-S (x mode only) Extract files as sparse files. For every block on disk, check first if it contains
only NULL bytes and seek over it otherwise. This works similar to the conv=sparse option of dd.
-s pattern
Modify file or archive member names according to pattern. The pattern has the format
/old/new/[ghHprRsS] where old is a basic regular expression, new is the replacement string of the
matched part, and the optional trailing letters modify how the replacement is handled. If old is
not matched, the pattern is skipped. Within new, ~ is substituted with the match, \1 to \9 with
the content of the corresponding captured group. The optional trailing g specifies that matching
should continue after the matched part and stop on the first unmatched pattern. The optional
trailing s specifies that the pattern applies to the value of symbolic links. The optional
trailing p specifies that after a successful substitution the original path name and the new path
name should be printed to standard error. Optional trailing H, R, or S characters suppress
substitutions for hardlink targets, regular filenames, or symlink targets, respectively.
Optional trailing h, r, or s characters enable substitutions for hardlink targets, regular
filenames, or symlink targets, respectively. The default is hrs which applies substitutions to
all names. In particular, it is never necessary to specify h, r, or s.
--same-owner
(x mode only) Extract owner and group IDs. This is the reverse of --no-same-owner and the
default behavior if tar is run as root.
--strip-components count
Remove the specified number of leading path elements. Pathnames with fewer elements will be
silently skipped. Note that the pathname is edited after checking inclusion/exclusion patterns
but before security checks.
-T filename, --files-from filename
In x or t mode, tar will read the list of names to be extracted from filename. In c mode, tar
will read names to be archived from filename. The special name “-C” on a line by itself will
cause the current directory to be changed to the directory specified on the following line.
Names are terminated by newlines unless --null is specified. Note that --null also disables the
special handling of lines containing “-C”. Note: If you are generating lists of files using
find(1), you probably want to use -n as well.
--totals
(c, r, u mode only) After archiving all files, print a summary to stderr.
-U, --unlink, --unlink-first
(x mode only) Unlink files before creating them. This can be a minor performance optimization if
most files already exist, but can make things slower if most files do not already exist. This
flag also causes tar to remove intervening directory symlinks instead of reporting an error. See
the SECURITY section below for more details.
--uid id
Use the provided user id number and ignore the user name from the archive. On create, if --uname
is not also specified, the user name will be set to match the user id.
--uname name
Use the provided user name. On extract, this overrides the user name in the archive; if the
provided user name does not exist on the system, it will be ignored and the user id (from the
archive or from the --uid option) will be used instead. On create, this sets the user name that
will be stored in the archive; the name is not verified against the system user database.
--use-compress-program program
Pipe the input (in x or t mode) or the output (in c mode) through program instead of using the
builtin compression support.
-v, --verbose
Produce verbose output. In create and extract modes, tar will list each file name as it is read
from or written to the archive. In list mode, tar will produce output similar to that of ls(1).
An additional -v option will also provide ls-like details in create and extract mode.
--version
Print version of tar and libarchive, and exit.
-w, --confirmation, --interactive
Ask for confirmation for every action.
-X filename, --exclude-from filename
Read a list of exclusion patterns from the specified file. See --exclude for more information
about the handling of exclusions.
-y (c mode only) Compress the resulting archive with bzip2(1). In extract or list modes, this
option is ignored. Note that, unlike other tar implementations, this implementation recognizes
bzip2 compression automatically when reading archives.
-Z, --compress, --uncompress
(c mode only) Compress the resulting archive with compress(1). In extract or list modes, this
option is ignored. Note that, unlike other tar implementations, this implementation recognizes
compress compression automatically when reading archives.
-z, --gunzip, --gzip
(c mode only) Compress the resulting archive with gzip(1). In extract or list modes, this option
is ignored. Note that, unlike other tar implementations, this implementation recognizes gzip
compression automatically when reading archives.
ENVIRONMENT
The following environment variables affect the execution of tar:
TAR_READER_OPTIONS
The default options for format readers and compression readers. The --options option
overrides this.
TAR_WRITER_OPTIONS
The default options for format writers and compression writers. The --options option
overrides this.
LANG The locale to use. See environ(7) for more information.
TAPE The default device. The -f option overrides this. Please see the description of the -f
option above for more details.
TZ The timezone to use when displaying dates. See environ(7) for more information.
EXIT STATUS
The tar utility exits 0 on success, and >0 if an error occurs.
EXAMPLES
The following creates a new archive called file.tar.gz that contains two files source.c and source.h:
tar -czf file.tar.gz source.c source.h
To view a detailed table of contents for this archive:
tar -tvf file.tar.gz
To extract all entries from the archive on the default tape drive:
tar -x
To examine the contents of an ISO 9660 cdrom image:
tar -tf image.iso
To move file hierarchies, invoke tar as
tar -cf - -C srcdir . | tar -xpf - -C destdir
or more traditionally
cd srcdir ; tar -cf - . | (cd destdir ; tar -xpf -)
In create mode, the list of files and directories to be archived can also include directory change
instructions of the form -Cfoo/baz and archive inclusions of the form @archive-file. For example, the
command line
tar -c -f new.tar foo1 @old.tgz -C/tmp foo2
will create a new archive new.tar. tar will read the file foo1 from the current directory and add it to
the output archive. It will then read each entry from old.tgz and add those entries to the output
archive. Finally, it will switch to the /tmp directory and add foo2 to the output archive.
An input file in mtree(5) format can be used to create an output archive with arbitrary ownership,
permissions, or names that differ from existing data on disk:
$ cat input.mtree
#mtree
usr/bin uid=0 gid=0 mode=0755 type=dir
usr/bin/ls uid=0 gid=0 mode=0755 type=file content=myls
$ tar -cvf output.tar @input.mtree
The --newer and --newer-mtime switches accept a variety of common date and time specifications, including
“12 Mar 2005 7:14:29pm”, “2005-03-12 19:14”, “5 minutes ago”, and “19:14 PST May 1”.
The --options argument can be used to control various details of archive generation or reading. For
example, you can generate mtree output which only contains type, time, and uid keywords:
tar -cf file.tar --format=mtree --options='!all,type,time,uid' dir
or you can set the compression level used by gzip or xz compression:
tar -czf file.tar --options='compression-level=9'.
For more details, see the explanation of the archive_read_set_options() and archive_write_set_options()
API calls that are described in archive_read(3) and archive_write(3).
COMPATIBILITY
The bundled-arguments format is supported for compatibility with historic implementations. It consists
of an initial word (with no leading - character) in which each character indicates an option. Arguments
follow as separate words. The order of the arguments must match the order of the corresponding
characters in the bundled command word. For example,
tar tbf 32 file.tar
specifies three flags t, b, and f. The b and f flags both require arguments, so there must be two
additional items on the command line. The 32 is the argument to the b flag, and file.tar is the argument
to the f flag.
The mode options c, r, t, u, and x and the options b, f, l, m, o, v, and w comply with SUSv2.
For maximum portability, scripts that invoke tar should use the bundled-argument format above, should
limit themselves to the c, t, and x modes, and the b, f, m, v, and w options.
Additional long options are provided to improve compatibility with other tar implementations.
SECURITY
Certain security issues are common to many archiving programs, including tar. In particular, carefully-
crafted archives can request that tar extract files to locations outside of the target directory. This
can potentially be used to cause unwitting users to overwrite files they did not intend to overwrite. If
the archive is being extracted by the superuser, any file on the system can potentially be overwritten.
There are three ways this can happen. Although tar has mechanisms to protect against each one, savvy
users should be aware of the implications:
• Archive entries can have absolute pathnames. By default, tar removes the leading / character
from filenames before restoring them to guard against this problem.
• Archive entries can have pathnames that include .. components. By default, tar will not extract
files containing .. components in their pathname.
• Archive entries can exploit symbolic links to restore files to other directories. An archive can
restore a symbolic link to another directory, then use that link to restore a file into that
directory. To guard against this, tar checks each extracted path for symlinks. If the final
path element is a symlink, it will be removed and replaced with the archive entry. If -U is
specified, any intermediate symlink will also be unconditionally removed. If neither -U nor -P
is specified, tar will refuse to extract the entry.
To protect yourself, you should be wary of any archives that come from untrusted sources. You should
examine the contents of an archive with
tar -tf filename
before extraction. You should use the -k option to ensure that tar will not overwrite any existing files
or the -U option to remove any pre-existing files. You should generally not extract archives while
running with super-user privileges. Note that the -P option to tar disables the security checks above
and allows you to extract an archive while preserving any absolute pathnames, .. components, or symlinks
to other directories.
SEE ALSO
bzip2(1), compress(1), cpio(1), gzip(1), mt(1), pax(1), shar(1), xz(1), libarchive(3),
libarchive-formats(5), tar(5)
STANDARDS
There is no current POSIX standard for the tar command; it appeared in ISO/IEC 9945-1:1996 (“POSIX.1”)
but was dropped from IEEE Std 1003.1-2001 (“POSIX.1”). The options supported by this implementation were
developed by surveying a number of existing tar implementations as well as the old POSIX specification
for tar and the current POSIX specification for pax.
The ustar and pax interchange file formats are defined by IEEE Std 1003.1-2001 (“POSIX.1”) for the pax
command.
HISTORY
A tar command appeared in Seventh Edition Unix, which was released in January, 1979. There have been
numerous other implementations, many of which extended the file format. John Gilmore's pdtar public-
domain implementation (circa November, 1987) was quite influential, and formed the basis of GNU tar. GNU
tar was included as the standard system tar in FreeBSD beginning with FreeBSD 1.0.
This is a complete re-implementation based on the libarchive(3) library. It was first released with
FreeBSD 5.4 in May, 2005.
BUGS
This program follows ISO/IEC 9945-1:1996 (“POSIX.1”) for the definition of the -l option. Note that GNU
tar prior to version 1.15 treated -l as a synonym for the --one-file-system option.
The -C dir option may differ from historic implementations.
All archive output is written in correctly-sized blocks, even if the output is being compressed. Whether
or not the last output block is padded to a full block size varies depending on the format and the output
device. For tar and cpio formats, the last block of output is padded to a full block size if the output
is being written to standard output or to a character or block device such as a tape drive. If the
output is being written to a regular file, the last block will not be padded. Many compressors,
including gzip(1) and bzip2(1), complain about the null padding when decompressing an archive created by
tar, although they still extract it correctly.
The compression and decompression is implemented internally, so there may be insignificant differences
between the compressed output generated by
tar -czf - file
and that generated by
tar -cf - file | gzip
The default should be to read and write archives to the standard I/O paths, but tradition (and POSIX)
dictates otherwise.
The r and u modes require that the archive be uncompressed and located in a regular file on disk. Other
archives can be modified using c mode with the @archive-file extension.
To archive a file called @foo or -foo you must specify it as ./@foo or ./-foo, respectively.
In create mode, a leading ./ is always removed. A leading / is stripped unless the -P option is
specified.
There needs to be better support for file selection on both create and extract.
There is not yet any support for multi-volume archives.
Converting between dissimilar archive formats (such as tar and cpio) using the @- convention can cause
hard link information to be lost. (This is a consequence of the incompatible ways that different archive
formats store hardlink information.)
Debian September 16, 2014 TAR(1)