`dd` system ‘dd’ command

Overview

The dd command copies data from an input file or device to an output file or device, optionally performing conversions along the way. It is commonly used for tasks such as backing up and restoring raw disk partitions, creating disk images, and converting data formats.

The name dd stands for “convert and copy” (historically “disk dump”). It reads from stdin by default and writes to stdout by default, making it useful in pipelines and for low-level I/O operations.

Configuration

Enable the command with CONFIG_SYSTEM_DD. This option is enabled by default unless CONFIG_DEFAULT_SMALL is set.

The following configuration options are available:

CONFIG_SYSTEM_DD_PROGNAME: Program name for the dd command. Default: dd.
CONFIG_SYSTEM_DD_PRIORITY: Task priority for the dd command. Default: 100.
CONFIG_SYSTEM_DD_STACKSIZE: Stack size for the dd command. Default: DEFAULT_TASK_STACKSIZE.
CONFIG_SYSTEM_DD_STATS: Enable transfer statistics output. Default: yes.

Usage

dd [if=<infile>] [of=<outfile>] [bs=<sectsize>] [count=<sectors>] [skip=<sectors>] [seek=<sectors>] [verify] [conv=<nocreat,notrunc>]

Options

if=<infile>

Input file. If not specified, stdin is used.

of=<outfile>

Output file. If not specified, stdout is used.

bs=<sectsize>

Block size in bytes. Default: 512 bytes.

count=<sectors>

Number of blocks to copy. Default: copy until end of input.

skip=<sectors>

Skip <sectors> blocks at the start of the input.

seek=<sectors>

Skip <sectors> blocks at the start of the output.

verify

Verify that the output matches the input after copying. Requires both if and of to be specified.

conv=<nocreat,notrunc>

Conversion options:

nocreat: Do not create the output file if it does not exist.
notrunc: Do not truncate the output file before writing.

Multiple conversion options can be separated by commas.

--help

Display usage information and exit.

Examples

Copy from stdin to stdout:

nsh> dd if=/dev/zero of=/tmp/zero.bin bs=1024 count=10
10+0 records in
10+0 records out

Create a disk image:

nsh> dd if=/dev/sda of=/tmp/disk.img bs=512
1024+0 records in
1024+0 records out

Restore a disk image:

nsh> dd if=/tmp/disk.img of=/dev/sda bs=512
1024+0 records in
1024+0 records out

Copy with verification:

nsh> dd if=/dev/sda of=/tmp/sda_backup.img bs=512 verify
1024+0 records in
1024+0 records out

Skip blocks in input:

nsh> dd if=/dev/sda of=/tmp/partition.img bs=512 skip=63 count=1024
1024+0 records in
1024+0 records out

Display usage information:

nsh> dd --help
usage:
  dd [if=<infile>] [of=<outfile>] [bs=<sectsize>] [count=<sectors>] [skip=<sectors>] [seek=<sectors>] [verify] [conv=<nocreat,notrunc>]

Notes

The default block size is 512 bytes, which is the standard sector size for most storage devices.
When count is not specified, dd copies until the end of the input is reached.
The verify option requires both if and of to be specified. It reads back the output and compares it to the input.
Transfer statistics are displayed when CONFIG_SYSTEM_DD_STATS is enabled. The output shows the number of records read/written and the transfer rate.
The conv options can be combined by separating them with commas. For example, conv=nocreat,notrunc.
dd is often used in pipelines with other commands. For example, dd if=/dev/urandom bs=1 count=10 | hexdump.

dd system ‘dd’ command