xref: /freebsd/bin/dd/dd.1 (revision 2e3f49888ec8851bafb22011533217487764fdb0)
1.\"-
2.\" Copyright (c) 1990, 1993
3.\"	The Regents of the University of California.  All rights reserved.
4.\"
5.\" This code is derived from software contributed to Berkeley by
6.\" Keith Muller of the University of California, San Diego.
7.\"
8.\" Redistribution and use in source and binary forms, with or without
9.\" modification, are permitted provided that the following conditions
10.\" are met:
11.\" 1. Redistributions of source code must retain the above copyright
12.\"    notice, this list of conditions and the following disclaimer.
13.\" 2. Redistributions in binary form must reproduce the above copyright
14.\"    notice, this list of conditions and the following disclaimer in the
15.\"    documentation and/or other materials provided with the distribution.
16.\" 3. Neither the name of the University nor the names of its contributors
17.\"    may be used to endorse or promote products derived from this software
18.\"    without specific prior written permission.
19.\"
20.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
21.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
22.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
23.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
24.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
25.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
26.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
27.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
28.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
29.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
30.\" SUCH DAMAGE.
31.\"
32.Dd June 4, 2020
33.Dt DD 1
34.Os
35.Sh NAME
36.Nm dd
37.Nd convert and copy a file
38.Sh SYNOPSIS
39.Nm
40.Op Ar operands ...
41.Sh DESCRIPTION
42The
43.Nm
44utility copies the standard input to the standard output.
45Input data is read and written in 512-byte blocks.
46If input reads are short, input from multiple reads are aggregated
47to form the output block.
48When finished,
49.Nm
50displays the number of complete and partial input and output blocks
51and truncated input records to the standard error output.
52.Pp
53The following operands are available:
54.Bl -tag -width "of=file"
55.It Cm bs Ns = Ns Ar n
56Set both input and output block size to
57.Ar n
58bytes, superseding the
59.Cm ibs
60and
61.Cm obs
62operands.
63If no conversion values other than
64.Cm noerror ,
65.Cm notrunc
66or
67.Cm sync
68are specified, then each input block is copied to the output as a
69single block without any aggregation of short blocks.
70.It Cm cbs Ns = Ns Ar n
71Set the conversion record size to
72.Ar n
73bytes.
74The conversion record size is required by the record oriented conversion
75values.
76.It Cm count Ns = Ns Ar n
77Copy only
78.Ar n
79input blocks.
80.It Cm files Ns = Ns Ar n
81Copy
82.Ar n
83input files before terminating.
84This operand is only applicable when the input device is a tape.
85.It Cm fillchar Ns = Ns Ar c
86When padding a block in conversion mode or due to use of
87.Cm noerror
88and
89.Cm sync
90modes, fill with the specified
91.Tn ASCII
92character, rather than using a space or
93.Dv NUL .
94.It Cm ibs Ns = Ns Ar n
95Set the input block size to
96.Ar n
97bytes instead of the default 512.
98.It Cm if Ns = Ns Ar file
99Read input from
100.Ar file
101instead of the standard input.
102.It Cm iflag Ns = Ns Ar value Ns Op , Ns Ar value ...
103Where
104.Cm value
105is one of the symbols from the following list.
106.Bl -tag -width "fullblock"
107.It Cm fullblock
108Reading from the input file may not obtain a full block.
109When a read returns short, continue reading to fill the block.
110Without this flag,
111.Cm count
112limits the number of times
113.Xr read 2
114is called on the input rather than the number of blocks copied in full.
115May not be combined with
116.Cm conv=sync .
117.It Cm direct
118Set the O_DIRECT flag on the input file to make reads bypass any local caching.
119.El
120.It Cm iseek Ns = Ns Ar n
121Seek on the input file
122.Ar n
123blocks.
124This is synonymous with
125.Cm skip Ns = Ns Ar n .
126.It Cm obs Ns = Ns Ar n
127Set the output block size to
128.Ar n
129bytes instead of the default 512.
130.It Cm of Ns = Ns Ar file
131Write output to
132.Ar file
133instead of the standard output.
134Any regular output file is truncated unless the
135.Cm notrunc
136conversion value is specified.
137If an initial portion of the output file is seeked past (see the
138.Cm oseek
139operand),
140the output file is truncated at that point.
141.It Cm oflag Ns = Ns Ar value Ns Op , Ns Ar value ...
142Where
143.Cm value
144is one of the symbols from the following list.
145.Bl -tag -width "direct"
146.It Cm fsync
147Set the O_FSYNC flag on the output file to make writes synchronous.
148.It Cm sync
149Set the O_SYNC flag on the output file to make writes synchronous.
150This is synonymous with the
151.Cm fsync
152value.
153.It Cm direct
154Set the O_DIRECT flag on the output file to make writes bypass any local caching.
155.El
156.It Cm oseek Ns = Ns Ar n
157Seek on the output file
158.Ar n
159blocks.
160This is synonymous with
161.Cm seek Ns = Ns Ar n .
162.It Cm seek Ns = Ns Ar n
163Seek
164.Ar n
165blocks from the beginning of the output before copying.
166On non-tape devices, an
167.Xr lseek 2
168operation is used.
169Otherwise, existing blocks are read and the data discarded.
170If the user does not have read permission for the tape, it is positioned
171using the tape
172.Xr ioctl 2
173function calls.
174If the seek operation is past the end of file, space from the current
175end of file to the specified offset is filled with blocks of
176.Dv NUL
177bytes.
178.It Cm skip Ns = Ns Ar n
179Skip
180.Ar n
181blocks from the beginning of the input before copying.
182On input which supports seeks, an
183.Xr lseek 2
184operation is used.
185Otherwise, input data is read and discarded.
186For pipes, the correct number of bytes is read.
187For all other devices, the correct number of blocks is read without
188distinguishing between a partial or complete block being read.
189.It Cm speed Ns = Ns Ar n
190Limit the copying speed to
191.Ar n
192bytes per second.
193.It Cm status Ns = Ns Ar value
194Where
195.Cm value
196is one of the symbols from the following list.
197.Bl -tag -width "progress"
198.It Cm noxfer
199Do not print the transfer statistics as the last line of status output.
200.It Cm none
201Do not print the status output.
202Error messages are shown; informational messages are not.
203.It Cm progress
204Print basic transfer statistics once per second.
205.El
206.It Cm conv Ns = Ns Ar value Ns Op , Ns Ar value ...
207Where
208.Cm value
209is one of the symbols from the following list.
210.Bl -tag -width "unblock"
211.It Cm ascii , oldascii
212The same as the
213.Cm unblock
214value except that characters are translated from
215.Tn EBCDIC
216to
217.Tn ASCII
218before the
219records are converted.
220(These values imply
221.Cm unblock
222if the operand
223.Cm cbs
224is also specified.)
225There are two conversion maps for
226.Tn ASCII .
227The value
228.Cm ascii
229specifies the recommended one which is compatible with
230.At V .
231The value
232.Cm oldascii
233specifies the one used in historic
234.At
235and
236.No pre- Ns Bx 4.3 reno
237systems.
238.It Cm block
239Treats the input as a sequence of newline or end-of-file terminated variable
240length records independent of input and output block boundaries.
241Any trailing newline character is discarded.
242Each input record is converted to a fixed length output record where the
243length is specified by the
244.Cm cbs
245operand.
246Input records shorter than the conversion record size are padded with spaces.
247Input records longer than the conversion record size are truncated.
248The number of truncated input records, if any, are reported to the standard
249error output at the completion of the copy.
250.It Cm ebcdic , ibm , oldebcdic , oldibm
251The same as the
252.Cm block
253value except that characters are translated from
254.Tn ASCII
255to
256.Tn EBCDIC
257after the
258records are converted.
259(These values imply
260.Cm block
261if the operand
262.Cm cbs
263is also specified.)
264There are four conversion maps for
265.Tn EBCDIC .
266The value
267.Cm ebcdic
268specifies the recommended one which is compatible with
269.At V .
270The value
271.Cm ibm
272is a slightly different mapping, which is compatible with the
273.At V
274.Cm ibm
275value.
276The values
277.Cm oldebcdic
278and
279.Cm oldibm
280are maps used in historic
281.At
282and
283.No pre- Ns Bx 4.3 reno
284systems.
285.It Cm fdatasync
286Perform an
287.Xr fdatasync 2
288on the output file before closing it.
289.It Cm fsync
290Perform an
291.Xr fsync 2
292on the output file before closing it.
293.It Cm lcase
294Transform uppercase characters into lowercase characters.
295.It Cm pareven , parnone , parodd , parset
296Output data with the specified parity.
297The parity bit on input is stripped unless
298.Tn EBCDIC
299to
300.Tn ASCII
301conversions is also specified.
302.It Cm noerror
303Do not stop processing on an input error.
304When an input error occurs, a diagnostic message followed by the current
305input and output block counts will be written to the standard error output
306in the same format as the standard completion message.
307If the
308.Cm sync
309conversion is also specified, any missing input data will be replaced
310with
311.Dv NUL
312bytes (or with spaces if a block oriented conversion value was
313specified) and processed as a normal input buffer.
314If the
315.Cm fillchar
316option is specified, the fill character provided on the command line
317will override
318the automatic selection of the fill character.
319If the
320.Cm sync
321conversion is not specified, the input block is omitted from the output.
322On input files which are not tapes or pipes, the file offset
323will be positioned past the block in which the error occurred using
324.Xr lseek 2 .
325.It Cm notrunc
326Do not truncate the output file.
327This will preserve any blocks in the output file not explicitly written
328by
329.Nm .
330The
331.Cm notrunc
332value is not supported for tapes.
333.It Cm osync
334Pad the final output block to the full output block size.
335If the input file is not a multiple of the output block size
336after conversion, this conversion forces the final output block
337to be the same size as preceding blocks for use on devices that require
338regularly sized blocks to be written.
339This option is incompatible with use of the
340.Cm bs Ns = Ns Ar n
341block size specification.
342.It Cm sparse
343If one or more output blocks would consist solely of
344.Dv NUL
345bytes, try to seek the output file by the required space instead of
346filling them with
347.Dv NUL Ns s ,
348resulting in a sparse file.
349.It Cm swab
350Swap every pair of input bytes.
351If an input buffer has an odd number of bytes, the last byte will be
352ignored during swapping.
353.It Cm sync
354Pad every input block to the input buffer size.
355Spaces are used for pad bytes if a block oriented conversion value is
356specified, otherwise
357.Dv NUL
358bytes are used.
359.It Cm ucase
360Transform lowercase characters into uppercase characters.
361.It Cm unblock
362Treats the input as a sequence of fixed length records independent of input
363and output block boundaries.
364The length of the input records is specified by the
365.Cm cbs
366operand.
367Any trailing space characters are discarded and a newline character is
368appended.
369.El
370.El
371.Pp
372Where sizes or speed are specified, a decimal, octal, or hexadecimal number of
373bytes is expected.
374If the number ends with a
375.Dq Li b ,
376.Dq Li k ,
377.Dq Li m ,
378.Dq Li g ,
379.Dq Li t ,
380.Dq Li p ,
381or
382.Dq Li w ,
383the
384number is multiplied by 512, 1024 (1K), 1048576 (1M), 1073741824 (1G),
3851099511627776 (1T), 1125899906842624 (1P)
386or the number of bytes in an integer, respectively.
387Two or more numbers may be separated by an
388.Dq Li x
389to indicate a product.
390.Pp
391When finished,
392.Nm
393displays the number of complete and partial input and output blocks,
394truncated input records and odd-length byte-swapping blocks to the
395standard error output.
396A partial input block is one where less than the input block size
397was read.
398A partial output block is one where less than the output block size
399was written.
400Partial output blocks to tape devices are considered fatal errors.
401Otherwise, the rest of the block will be written.
402Partial output blocks to character devices will produce a warning message.
403A truncated input block is one where a variable length record oriented
404conversion value was specified and the input line was too long to
405fit in the conversion record or was not newline terminated.
406.Pp
407Normally, data resulting from input or conversion or both are aggregated
408into output blocks of the specified size.
409After the end of input is reached, any remaining output is written as
410a block.
411This means that the final output block may be shorter than the output
412block size.
413.Pp
414If
415.Nm
416receives a
417.Dv SIGINFO
418(see the
419.Cm status
420argument for
421.Xr stty 1 )
422signal, the current input and output block counts will
423be written to the standard error output
424in the same format as the standard completion message.
425If
426.Nm
427receives a
428.Dv SIGINT
429signal, the current input and output block counts will
430be written to the standard error output
431in the same format as the standard completion message and
432.Nm
433will exit.
434.Sh EXIT STATUS
435.Ex -std
436.Sh EXAMPLES
437Check that a disk drive contains no bad blocks:
438.Pp
439.Dl "dd if=/dev/ada0 of=/dev/null bs=1m"
440.Pp
441Do a refresh of a disk drive, in order to prevent presently
442recoverable read errors from progressing into unrecoverable read errors:
443.Pp
444.Dl "dd if=/dev/ada0 of=/dev/ada0 bs=1m"
445.Pp
446Remove parity bit from a file:
447.Pp
448.Dl "dd if=file conv=parnone of=file.txt"
449.Pp
450Check for (even) parity errors on a file:
451.Pp
452.Dl "dd if=file conv=pareven | cmp -x - file"
453.Pp
454To create an image of a Mode-1 CD-ROM, which is a commonly used format
455for data CD-ROM disks, use a block size of 2048 bytes:
456.Pp
457.Dl "dd if=/dev/cd0 of=filename.iso bs=2048"
458.Pp
459Write a filesystem image to a memory stick, padding the end with zeros,
460if necessary, to a 1MiB boundary:
461.Pp
462.Dl "dd if=memstick.img of=/dev/da0 bs=1m conv=noerror,sync"
463.Sh SEE ALSO
464.Xr cp 1 ,
465.Xr mt 1 ,
466.Xr recoverdisk 1 ,
467.Xr tr 1 ,
468.Xr geom 4 ,
469.Xr trim 8
470.Sh STANDARDS
471The
472.Nm
473utility is expected to be a superset of the
474.St -p1003.2
475standard.
476The
477.Cm files
478and
479.Cm status
480operands and the
481.Cm ascii ,
482.Cm ebcdic ,
483.Cm ibm ,
484.Cm oldascii ,
485.Cm oldebcdic
486and
487.Cm oldibm
488values are extensions to the
489.Tn POSIX
490standard.
491.Sh HISTORY
492A
493.Nm
494command appeared in
495.At v5 .
496.Sh BUGS
497Protection mechanisms in the
498.Xr geom 4
499subsystem might prevent the super-user from writing blocks to a disk.
500Instructions for temporarily disabling these protection mechanisms can be
501found in the
502.Xr geom 4
503man page.
504