Intel® TinyCrypt: how to use AES-128/CTR - c++

Question

I need to do AES-128/CTR encryption/decryption using Intel TinyCrypt (written in C). I took a look at the test case provided by the library, but still have a few questions on how to use it. Here's the code of the test case (see the complete source code here):
/*
* NIST SP 800-38a CTR Test for encryption and decryption.
*/
uint32_t test_1_and_2(void)
{
const uint8_t key[16] = {
0x2b, 0x7e, 0x15, 0x16, 0x28, 0xae, 0xd2, 0xa6, 0xab, 0xf7, 0x15, 0x88,
0x09, 0xcf, 0x4f, 0x3c
};
uint8_t ctr[16] = {
0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff
};
const uint8_t plaintext[64] = {
0x6b, 0xc1, 0xbe, 0xe2, 0x2e, 0x40, 0x9f, 0x96, 0xe9, 0x3d, 0x7e, 0x11,
0x73, 0x93, 0x17, 0x2a, 0xae, 0x2d, 0x8a, 0x57, 0x1e, 0x03, 0xac, 0x9c,
0x9e, 0xb7, 0x6f, 0xac, 0x45, 0xaf, 0x8e, 0x51, 0x30, 0xc8, 0x1c, 0x46,
0xa3, 0x5c, 0xe4, 0x11, 0xe5, 0xfb, 0xc1, 0x19, 0x1a, 0x0a, 0x52, 0xef,
0xf6, 0x9f, 0x24, 0x45, 0xdf, 0x4f, 0x9b, 0x17, 0xad, 0x2b, 0x41, 0x7b,
0xe6, 0x6c, 0x37, 0x10
};
const uint8_t ciphertext[80] = {
0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0xfa, 0xfb,
0xfc, 0xfd, 0xfe, 0xff, 0x87, 0x4d, 0x61, 0x91, 0xb6, 0x20, 0xe3, 0x26,
0x1b, 0xef, 0x68, 0x64, 0x99, 0x0d, 0xb6, 0xce, 0x98, 0x06, 0xf6, 0x6b,
0x79, 0x70, 0xfd, 0xff, 0x86, 0x17, 0x18, 0x7b, 0xb9, 0xff, 0xfd, 0xff,
0x5a, 0xe4, 0xdf, 0x3e, 0xdb, 0xd5, 0xd3, 0x5e, 0x5b, 0x4f, 0x09, 0x02,
0x0d, 0xb0, 0x3e, 0xab, 0x1e, 0x03, 0x1d, 0xda, 0x2f, 0xbe, 0x03, 0xd1,
0x79, 0x21, 0x70, 0xa0, 0xf3, 0x00, 0x9c, 0xee
};
struct tc_aes_key_sched_struct sched;
uint8_t out[80];
uint8_t decrypted[64];
uint32_t result = TC_PASS;
TC_PRINT("CTR test #1 (encryption SP 800-38a tests):\n");
(void)tc_aes128_set_encrypt_key(&sched, key);
(void)memcpy(out, ctr, sizeof(ctr));
if (tc_ctr_mode(&out[TC_AES_BLOCK_SIZE], sizeof(plaintext),
plaintext, sizeof(plaintext), ctr, &sched) == 0) {
TC_ERROR("CTR test #1 (encryption SP 800-38a tests) failed in %s.\n", __func__);
result = TC_FAIL;
goto exitTest1;
}
result = check_result(1, ciphertext, sizeof(out), out, sizeof(out));
TC_END_RESULT(result);
TC_PRINT("CTR test #2 (decryption SP 800-38a tests):\n");
(void) memcpy(ctr, out, sizeof(ctr));
if (tc_ctr_mode(decrypted, sizeof(decrypted), &out[TC_AES_BLOCK_SIZE],
sizeof(decrypted), ctr, &sched) == 0) {
TC_ERROR("CTR test #2 (decryption SP 800-38a tests) failed in %s.\n", __func__);
result = TC_FAIL;
goto exitTest1;
}
result = check_result(2, plaintext, sizeof(plaintext),
decrypted, sizeof(plaintext));
exitTest1:
TC_END_RESULT(result);
return result;
}
My input data is a double pointer (2D array) and I'd like to keep it that way:
uint8_t ** mydata = gen_rand_data(const size_t height, const size_t length);
Here's my understanding (correct me if I'm wrong):
AES-128 is a block cipher, with 128-bit key.
CTR mode makes the AES-128 a stream cipher and lets encrypt/decrypt data with arbitrary size.
Having AES-128/CTR, the output (ciphertext) has the same length as plaintext.
To improve the security, it's better to use 128-bit IV (nonce) for encryption.
Now here's my questions:
Can I say ctr (initial counter, right?) is our IV?
Why the ciphertext (expected value) and out (calculated value) are 128-bit bigger than plaintext?
Why the code copies ctr value in the beginning of out?
To encrypt mydata (2D array), I just need to loop over it and simple feed bytes from mydata to tc_ctr_mode, regardless of the dimension of mydata. Is this correct?
Thanks in advance.

Answer 1

The counter is generally supplied as the iv parameter if there is no specific counter parameter.
The output is 16-bytes longer because the counter is prepended, the counter is block sized and AES has a 16-byte block size..
The counter value is not secret and is needed for decryption. Since the same key/counter combination can never be reused if the key is going to be reused (a common situation) the counter must be unique and thus not pre-shared. The general answer is to pre-pend the counter value to the encrypted data.
Yes providing the decryption end knows a priori the array dimensions. But is is probably better to use some form of marshaling to preserve the array shape.