topic MMCAU SHA256 in Kinetis Microcontrollers

MMCAU SHA256

pietrodicastri — Mon, 25 Feb 2019 13:29:02 GMT

Good morning

I run the example mmcau_api. I test the SHA256 with the proposed test vector. Here it is

static uint8_t g_testSha[TEST_LENGTH] = {0x54, 0x68, 0x65, 0x20, 0x71, 0x75, 0x69, 0x63, 0x6b, 0x20, 0x62, 0x72, 0x6f,
0x77, 0x6e, 0x20, 0x66, 0x6f, 0x78, 0x20, 0x6a, 0x75, 0x6d, 0x70, 0x73, 0x20,
0x6f, 0x76, 0x65, 0x72, 0x20, 0x74, 0x68, 0x65, 0x20, 0x6c, 0x61, 0x7a, 0x79,
0x20, 0x64, 0x6f, 0x67, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x58};

The result I get is

{3618175923, 131563668, 1774885564, 2953326159, 2371244516, 1832704886, 755159231, 935978386}

I test the same vector on line here https://www.fileformat.info/tool/hash.htm?hex=0x54%2C+++0x68%2C+0x65%2C+0x20%2C+0x71%2C+0x75%2C+0x69%2C+0x63%2C+0x6b%2C+…

and I get the result

d4b8527d7a673b096ba19aabfb458c357a43f008d214aac413ee18cf29485fe2

It does not match. So what is wrong???

Thank You

Pietro

Re: MMCAU SHA256

mjbcswitzerland — Mon, 25 Feb 2019 19:17:16 GMT

Hi Pietro

I checked in the uTasker project (supporting SW, mmCAN and LTC SHA256):

static const ALIGNED_BUFFER text = { 0,{0x54, 0x68, 0x65, 0x20, 0x71, 0x75, 0x69, 0x63, 0x6b, 0x20, 0x62, 0x72, 0x6f,
0x77, 0x6e, 0x20, 0x66, 0x6f, 0x78, 0x20, 0x6a, 0x75, 0x6d, 0x70, 0x73, 0x20,
0x6f, 0x76, 0x65, 0x72, 0x20, 0x74, 0x68, 0x65, 0x20, 0x6c, 0x61, 0x7a, 0x79,
0x20, 0x64, 0x6f, 0x67, 0x80, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x58} };

int i;
fnSHA256(text.ucData, recovered.ucData, 64, SHA_START_CALCULATE_TERMINATE);
for (i = 0; i < 32; i++) {                         // the HASH is always 32 bytes in length
    fnDebugHex(recovered.ucData[i], (WITH_SPACE | sizeof(unsigned char) | WITH_LEADIN));
}
‍‍‍‍‍‍‍‍‍‍‍

The output it gives is:

0xd4 0xb8 0x52 0x7d 0x7a 0x67 0x3b 0x09 0x6b 0xa1 0x9a 0xab 0xfb 0x45 0x8c 0x35 0x7a 0x43 0xf0 0x08 0xd2 0x14 0xaa 0xc4 0x13 0xee 0x18 0xcf 0x29 0x48 0x5f 0xe2

which matches with the one from your link.

1. It looks like you are displaying the result in decimal form - try displaying in hex instead.
2. Also ensure that you are displaying as a string of binary values rather than risking little-/big endian confusion.
3. If using mmCAU ensure that your buffers are long word aligned since the crypto accelerators tend to require this.

Regards

Mark

Complete Kinetis solutions, training and support: http://www.utasker.com/kinetis.html

Re: MMCAU SHA256

pietrodicastri — Tue, 26 Feb 2019 08:01:18 GMT

Hei Mark

Good to find You here also.

I have tested the source code as is in the example and I expect it should work. I have here the FRDMK64 and the TOWER for the KV58.

Both of them give the same wrong result.

The HEX counterpart is here and it does not help. All of the buffers are correctly aligned.

That's the dump of the hex

If you have suggestionns are welcome.

Thank You

Pietro

Re: MMCAU SHA256

pietrodicastri — Tue, 26 Feb 2019 08:42:14 GMT

I am also wandering what is the LTC SHA ???

Re: MMCAU SHA256

mjbcswitzerland — Tue, 26 Feb 2019 10:52:50 GMT

Pietro

LTC is the Low power trusted cryptography module found in some of Kinetis parts (it may be instead of mmCAU or it may be in addition to mmCAU, but it is only in a few parts).

I also had some difficulties with the SHA256 in mmCAU and needed to do additional research to get it working:

See SHA-256 with mmCAU in K64

If you have access to the professional version of the uTasker project you can find the interface in crypto.c which handles all of the nuances (alignment, padding, endian-ness) needed by the various crypto accelerators.

Regards

Mark

Re: MMCAU SHA256

pietrodicastri — Tue, 26 Feb 2019 11:46:53 GMT

Hello Mark

There is a good new now. I tried to use the wrapper of the wolfssl and it is working in this way. I will try to catch what is problem.

Thank You