Mifare Classic MAD1 CRC8 Generation

[Nika Chkhartishvili]

Hi,

I am trying to format first sector of Mifare Classic as MAD
and ran into a problem of generating correct CRC.

I many implementations of CRC I found on internet and followed datasheet on CRC generation
but every time I get different results from every try and that is not equal to value
that is in datasheet. http://www.nxp.com/documents/application_note/AN10787.pdf

Is there some solution which is tested for Mifare Classic CRC generation ?

There were some variables which wasn't covered in datasheet such as initial value of CRC and using table lookup
which I saw as options in some implementations.

I use Java.

---

+ 0  |  - 0

[Jan Matik]

Hi, it's some time ago, but maybe it will help somebody anyway. I use this code to calculate MAD CRC:

uint8_t CalculateCRC8_MAD(uint8_t *in_pData, uint8_t in_nDataLength)

{

	if (in_pData == nullptr)

		return 0;

	

	if (in_nDataLength == 0)

		return 0;

	

	// init vector defined in MAD appnote is 0xE3. This didn't work, but bit-swapped value 0xC7 works

	boost::crc_optimal crc8;



	uint8_t	*pStart = in_pData;

	uint8_t *pEnd	= in_pData + in_nDataLength;



	crc8 = std::for_each(pStart, pEnd, crc8);



	return crc8.checksum();

}

Regards
Jan

---

+ 0  |  - 0

[Jan Matik]

The most important part in the previous post was cut (boost::crc_optimal), so I will try to post that line again without marking it as code (no idea how could I edit the post):

// init vector defined in MAD appnote is 0xE3. This didn't work, but bit-swapped value 0xC7 works
boost::crc_optimal crc8;

---

+ 0  |  - 0

[Jan Matik]

It looks that everything within angle brackets gets cut, so I will use normal brackets instead to declare crc8 variable from boost::crc_optimal template (I hope it will work finally :-) )

// init vector defined in MAD appnote is 0xE3. This didn't work, but bit-swapped value 0xC7 works
boost::crc_optimal(8, 0x1D, 0xC7, 0, false, false) crc8;

---

+ 0  |  - 0

[Author]

Posts