blob: b195db69a94daec2cd9daed92e156ec612231fc2 (
plain) (
tree)
|
|
# Testing the implementation
After you've [built](https://github.com/egor-tensin/aesni#building) the basic utilities,
you can verify the implementation either manually or automatically using scripts.
## Manually
The basic utilities have uniform interfaces.
For the ECB mode of operation, the usage is follows:
aesNNNecb_encrypt.exe KEY [PLAIN...]
and
aesNNNecb_decrypt.exe KEY [CIPHER...]
For the modes of operation involving initialization vectors (CBC, CFB, OFB, CTR, etc.),
use the utilities like this:
aesNNNxxx_encrypt.exe KEY INIT_VECTOR [PLAIN...]
and
aesNNNxxx_decrypt.exe KEY INIT_VECTOR [CIPHER...]
For example,
> aes128ecb_encrypt.exe 000102030405060708090a0b0c0d0e0f 00112233445566778899aabbccddeeff
69c4e0d86a7b0430d8cdb78070b4c55a
> aes192cbc_encrypt.exe 000102030405060708090a0b0c0d0e0f1011121314151617 1032547698badcfe1032547698badcfe 00112233445566778899aabbccddeeff 00112233445566778899aabbccddeeff 00112233445566778899aabbccddeeff
92c01276b27eb8baaa3cabe2c661d4a8
d42bdf90c1a48221a92a5137c1445418
96248fca82fbefa31345ae7d8fb7933e
On older CPUs, you can run the executables
[using Intel SDE](https://github.com/egor-tensin/aesni#running-on-older-cpus).
## Using test vectors
### From NIST 800-38A
You can test the implementation against the vectors from
[NIST Special Publication 800-38A](http://csrc.nist.gov/publications/nistpubs/800-38a/sp800-38a.pdf)
using `800-32a.py`.
The script is written in Python 3, so you need to be able to run Python 3 scripts prior to testing.
Then you can run the script, passing the path to the directory with the required `*_encrypt.exe` and `*_decrypt.exe` files like this:
python 800-32a.py -r C:\build\test\Debug
On older CPUs, you can make the script run the executables
[using Intel SDE](https://github.com/egor-tensin/aesni#running-on-older-cpus)
using
python 800-32a.py -r C:\build\test\Debug -e
You can investigate what went wrong (if it did) by logging calls to the `*_encrypt.exe` and `*_decrypt.exe` utilities by passing the `--log/-l FILENAME` option to the script.
The tests are really Python's `unittest` `TestCase`s, so the output of the script should be self-explanatory and familiary to users with prior Python testing experience.
|