Develop Reference

Cannot send large APDU commands with libnfc using nfc_initiator_transceive_bytes()

I'm trying to build a C++ wrapper around libnfc to make a communication between my Android and the PN532 RFID module.
This helped me a lot: http://nfc-tools.org/index.php/Libnfc:APDU_example
This code is meant to send an APDU command where the body is contained in message (I'm not sending any header bytes etc.) and read the response into response.
Problem: If message exceeds 262 characters then I get a buffer overflow detected error. Otherwise it works perfectly well. I don't even think the error is thrown by the NFC library.
bool send(const std::string &message, std::string &response){
std::vector<uint8_t> apduCmd(message.begin(), message.end());
uint8_t *capdu = &apduCmd[0];
size_t capdulen = apduCmd.size();
uint8_t rapdu[10];
size_t rapdulen = 10;
// BUFFER OVERFLOW HERE
int res = nfc_initiator_transceive_bytes(m_nfcDevice, capdu, capdulen, rapdu, rapdulen, 500);
if (res<0) {
return false;
}
if(res<2 || rapdu[res-2] != 0x90 || rapdu[res-1] != 0x00){
return false;
}
// byteArrayToString omitting the last two bytes
response = byteArrayToString(rapdu, 0, res-2);
return true;
}

The limit of 262 bytes is a hard limit imposed the PN532 NFC chip. This is the maximum size of the raw data that can be sent (and received) in one InDataExchange command. libnfc explicitly enforces this limit for the method nfc_initiator_transceive_bytes() (see the definition of abtCmd in pn53x_initiator_transceive_bytes() and the definition of PN53x_EXTENDED_FRAME__DATA_MAX_LEN).
What you could do to overcome this limit, is to compose your own ISO/IEC 14443-4 blocks (using InCommunicateThru, i.e. nfc_initiator_transceive_bytes() with m_nfcDevice->bEasyFraming turned off. While each frame would still be limited to 263 bytes (the PN532 actually allows 264 bytes for InCommunicateThru but libnfc seems to limit this to 263 bytes), you can then pack your extended length APDUs into multiple ISO/IEC 14443-4 I-blocks. However, you would need to handle the whole ISO/IEC 14443-4 framing on your own (which means that you also have to take care of receive acknowledgements, etc.)
Finally, since the other communication endpoint is an Android device: Many Android devices do not support extended length APDUs. Consequently, even if you send longer APDUs, you might not be able to receive and process them on the Android side. Also, be aware that you should send proper APDUs conforming to the structures defined in ISO/IEC 7816-4 (i.e. APDUs with valid header and lengths fields), otherwise you might run into issues when talking to some devices.

Android NfcV (ISO 15693) tag

Is it possible to write data to specific blocks in memory on the NfcV (ISO 15693) tag? E.g. write data to block# 5 or any specific block#.
I am new to NFC technologies. I am creating an application to read/write NfcV (ISO 15693) tags. I have successfully create the reading portion but the problem is on writing portion. When I want to write some text data into the tag it start from block# 2 to onward and every time doing the same procedure. I have searched lot but I can't find any solution to write data to specific blocks.

The exact details depend on which ISO 15693 compatible chip is inside the tag. The ISO 15693-3 standard lists different write commands. Support for these are all optional, so your tag may support one or more of these or even use a proprietary command for writing data. I would recommend to look up the datasheet of the chip and/or acquire the ISO standard to find out what the right command is.
Once you know what the right command is, you can simply pass the bytes of the command in a byte array to the NfcV.transceive() method. (Usually the command bytes consist of a flag byte, followed by a write command byte, one or more block bytes and the data bytes to be written.)

Tried the following: Getting the "Tag was lost" Exception:
nfc.connect();
byte[] arrByt = new byte[7];
arrByt[0] = 0x40;
arrByt[1] = 0x21;
arrByt[2] = 0x06;
arrByt[3] = 0x00;
arrByt[4] = 0x00;
arrByt[5] = 0x00;
arrByt[6] = 0x00;
byte[] response = nfc.transceive(arrByt);

I guess the android framework does not handle the response from the ISO15693 tags very well. I have been playing with HF-I tags. Few commands work flawlessly and for few other commands the NFC stack throws TAG Lost exception.

No response from MIFARE CLASSIC

I'm trying to authenticate any sector of a MIFARE classic card. I'm using a twinlinx mymax sticker (which makes almost any bluetooth device NFC enabled). It sends commands to a connected NFC tag. I've already made a connection and sent and recieved data with a Ultralight C tag, but so far I had no success on accessing a Mifare Classic. Here is my authentication code:
private boolean authenticate(int sector, byte[] key, boolean keyA) {
byte[] cmd = new byte[12];
// First byte is the command
if (keyA) {
cmd[0] = 0x60; // phHal_eMifareAuthentA
} else {
cmd[0] = 0x61; // phHal_eMifareAuthentB
}
// Second byte is block address
cmd[1] = (byte) 0x03;
// Next 6 bytes are the key
System.arraycopy(key, 0, cmd, 2, 6);
// Next 4 bytes is the UID
System.arraycopy(Answer, 3, cmd, 8,4);
byte[] test = null;
//this makes a connection to the NFC tag (and this works)
TR.ConnectToExternalCard(AUTH, (byte)0x00);
//checking if the tag is still connected
if (TR.isCardPresent() == true){
//sending command to authenticate
test = TR.SendCommandPropAndWaitResponse(cmd, (byte) 0x00);
}
try {
if (test != null) {
return true;
}
}
I'm using standard MIFARE Classic keys, the tags are fresh from the factory. The complete command (in bytes) which is sent to the tag is:
[0x60, 0x3, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xf3, 0xf4, 0xa9, 0xfb]
Any Ideas? The tag seems to be non-responsive... tried accessing other Classic tags but also had no success. Thanks!

It is hard to say what you are doing wrong using an SDK that is not publicly available. However, the API looks familiar enough, so I will give it a try anyway. I can think of a number of things you may try (in decreasing order of likeliness):
The UID bytes may be in the wrong order, so try reversing them.
Perhaps Answer does not only contain the UID, but other bytes, too (e.g. SAK), and you are copying the wrong bytes from it.
The MIFARE Classic tags you have may have a 7-byte UID and you are not using the correct 4 bytes from that.
May be TR.SendCommandPropAndWaitResponse() is the wrong method to use. Perhaps there is a dedicated method for MIFARE Classic.
The MyMax sticker may not support MIFARE Classic. I don't see explicit confirmation on their website that they do. However, indications are that their solution is based on NXP hardware, which always supports MIFARE Classic.

Making Mifare Ulralight Write protected in Android

i write some data (NDEF) into Mifare Ultralight Tag. Then the tag is set to read only with Ndef.MakeReadOnly(). The Tag seems to be read only afterwards.
But reading the Tag content shows, that only pages (4-15, Lx bytes are set to 1) are locked, the three BLx Bytes are left 0. When i got it right, this means that Pages Locks could be altered again. Page 2, Bytes 2 = 11111000, Page 2 Byte 3 = 11111111 - (see http://www.nxp.com/documents/data_sheet/MF0ICU1.pdf , page 12).
Is there a way to make the tag permanently write protected?
Thanks for your time.

To make the whole tag read only, you could get the tag again as a MifareUltralight instead of as a Ndef. Then you do
MifareUltralightTag.writePage(2,new byte[] {0x00, 0x00, 0xFF, 0xFF});
followed by
MifareUltralightTag.transceive(new byte[] {0x26});
to "activate the new locking configuration."

How do you send an extended-ascii AT-command (CCh) from Android bluetooth to a serial device?

This one really has me banging my head. I'm sending alphanumeric data from an Android app, through the BluetoothChatService, to a serial bluetooth adaptor connected to the serial input of a radio transceiver.
Everything works fine except when I try to configure the radio on-the-fly with its AT-commands. The AT+++ (enter command mode) is received OK, but the problem comes with the extended-ascii characters in the next two commands: Changing the radio destination address (which is what I'm trying to do) requires CCh 10h (plus 3 hex radio address bytes), and exiting the command mode requires CCh ATO.
I know the radio can be configured OK because I've done it on an earlier prototype with the serial commands from PIC basic, and it also can be configured by entering the commands directly from hyperterm. Both these methods somehow convert that pesky CCh into a form the radio understands.
I've have tried just about everything an Android noob could possibly come up with to finagle the encoding such as:
private void command_address() {
byte[] addrArray = {(byte) 0xCC, 16, 36, 65, 21, 13};
CharSequence addrvalues = EncodingUtils.getString(addrArray, "UTF-8");
sendMessage((String) addrvalues);
}
but no matter what, I can't seem to get that high-order byte (CCh/204/-52) to behave as it should. All other (< 127) bytes, command or data, transmit with no problem. Any help here would be greatly appreciated.
-Dave

Welll ... turns out the BluetoothChat code re-creates the byte array with message.getBytes() before sending to the service. (after all, being chat code it would normally source only regular ascii strings) As others on this site have pointed out, getBytes() can create encoding issues in some cases. So, for purposes of sending these extended-ascii commands, I don't mess with strings and just send the byte array to the service with
private void sendCommand(byte[] cmd) {
mChatService.write(cmd);
}
The so-called command array is first initialized with placeholders for the hex radio address elements
byte[] addrArray = {(byte) 0xCC, 16, 0, 0, 0, 13};
and then filled in with the help of the conversion method
radioArray = HexStringToByteArray(radioAddr1);
which can be found here: HexStringToByteArray#stackoverflow