Image tool
The Python program scripts/imgtool.py
can be used to perform the
operations that are necessary to manage keys and sign images. Using
this script should be preferred to the manual steps described in
doc/signed_images.md
.
This program is written for Python3, and has several dependencies on Python libraries. These can be installed using ‘pip3’:
pip3 install --user -r scripts/requirements.txt
Managing keys
This tool currently supports rsa-2048, rsa-3072, ecdsa-p256 and ed25519 keys. You can generate a keypair for one of these types using the ‘keygen’ command:
./scripts/imgtool.py keygen -k filename.pem -t rsa-2048
or use rsa-3072, ecdsa-p256, or ed25519 for the type. The key type used should match what MCUboot is configured to verify.
This key file is what is used to sign images, this file should be protected, and not widely distributed.
You can add the -p
argument to keygen
, which will cause it to
prompt for a password. You will need to enter this password in every
time you use the private key.
Incorporating the public key into the code
There is a development key distributed with MCUboot that can be used for testing. Since this private key is widely distributed, it should never be used for production. Once you have generated a production key, as described above, you should replace the public key in the bootloader with the generated one.
For Zephyr, the keys live in the file boot/zephyr/keys.c
. For
mynewt, follow the instructions in doc/signed_images.md
to generate
the key file.
./scripts/imgtool.py getpub -k filename.pem
will extract the public key from the given private key file, and
output it as a C data structure. You can replace or insert this code
into the key file. However, when the MCUBOOT_HW_KEY
config option is
enabled, this last step is unnecessary and can be skipped.
Signing images
Image signing takes an image in binary or Intel Hex format intended for the primary slot and adds a header and trailer that the bootloader is expecting:
Usage: imgtool.py sign [OPTIONS] INFILE OUTFILE
Create a signed or unsigned image
INFILE and OUTFILE are parsed as Intel HEX if the params have .hex
extension, otherwise binary format is used
Options:
-k, --key filename
--public-key-format [hash|full]
--align [1|2|4|8|16|32] Alignment used by swap update modes.
-v, --version TEXT [required]
-s, --security-counter TEXT Specify the value of security counter. Use
the `auto` keyword to automatically generate
it from the image version.
-d, --dependencies TEXT
--pad-sig Add 0-2 bytes of padding to ECDSA signature
(for MCUboot <1.5)
-H, --header-size INTEGER [required]
--pad-header Add --header-size zeroed bytes at the
beginning of the image
-S, --slot-size INTEGER Size of the slot where the image will be
written [required]
--pad Pad image to --slot-size bytes, adding
trailer magic
--confirm When padding the image, mark it as confirmed
-M, --max-sectors INTEGER When padding allow for this amount of
sectors (defaults to 128)
--boot-record sw_type Create CBOR encoded boot record TLV. The
sw_type represents the role of the software
component (e.g. CoFM for coprocessor
firmware). [max. 12 characters]
--overwrite-only Use overwrite-only instead of swap upgrades
-e, --endian [little|big] Select little or big endian
-E, --encrypt filename Encrypt image using the provided public key
--save-enctlv When upgrading, save encrypted key TLVs
instead of plain keys. Enable when
BOOT_SWAP_SAVE_ENCTLV config option was set.
-L, --load-addr INTEGER Load address for image when it should run
from RAM.
-x, --hex-addr INTEGER Adjust address in hex output file.
-R, --erased-val [0|0xff] The value that is read back from erased
flash.
-h, --help Show this message and exit.
The main arguments given are the key file generated above, a version field to place in the header (1.2.3 for example), the alignment of the flash device in question, and the header size.
The header size depends on the operating system and the particular
flash device. For Zephyr, it will be configured as part of the build,
and will be a small power of two. By default, the Zephyr build system will
already prepended a zeroed header to the image. If another build system is
in use that does not automatically add this zeroed header, --pad-header
can
be passed and the --header-size
will be added by imgtool. If --pad-header
is used with an Intel Hex file, --header-size
bytes will be subtracted from
the load address (in Intel Hex terms, the Extended Linear Address record) to
adjust for the new bytes prepended to the file. The load address of all data
existing in the file should not change.
The --slot-size
argument is required and used to check that the firmware
does not overflow into the swap status area (metadata). If swap upgrades are
not being used, --overwrite-only
can be passed to avoid adding the swap
status area size when calculating overflow.
The optional --pad
argument will place a trailer on the image that
indicates that the image should be considered an upgrade. Writing this image
in the secondary slot will then cause the bootloader to upgrade to it.
A dependency can be specified in the following way:
-d "(image_id, image_version)"
. The image_id
is the number of the image
which the current image depends on. The image_version
is the minimum version
of that image to satisfy compliance. For example -d "(1, 1.2.3+0)"
means this
image depends on Image 1 which version has to be at least 1.2.3+0.
The --public-key-format
argument can be used to distinguish where the public
key is stored for image authentication. The hash
option is used by default, in
which case only the hash of the public key is added to the TLV area (the full
public key is incorporated into the bootloader). When the full
option is used
instead, the TLV area will contain the whole public key and thus the bootloader
can be independent from the key(s). For more information on the additional
requirements of this option, see the design document.