doesn't really work

[ildar]

Hi! I reviewed both this fork and @socram8888 's pull request. Fixed building with cmake, see my repo Still attempring to run yields disappointing results:

 $ LIBNFC_LOG_LEVEL=3 ./nfc-anticol
info    libnfc.config   Unable to open file: /etc/nfc/libnfc.conf
debug   libnfc.config   key: [device.allow_autoscan], value: [false]
info    libnfc.config   Unknown key in config line: device.allow_autoscan = false
debug   libnfc.config   key: [device.log_level], value: [3]
info    libnfc.config   Unknown key in config line: device.log_level = 3
debug   libnfc.config   key: [device.name], value: [MFRC522]
debug   libnfc.config   key: [device.connstring], value: [rc522_uart:/dev/ttyUSB0]
debug   libnfc.general  log_level is set to 3
debug   libnfc.general  allow_autoscan is set to true
debug   libnfc.general  allow_intrusive_scan is set to false
debug   libnfc.general  1 device(s) defined by user
debug   libnfc.general    #0 name: "MFRC522", connstring: "rc522_uart:/dev/ttyUSB0"
debug   libnfc.driver.rc522_uart    Attempt to open: /dev/ttyUSB0.
debug   libnfc.driver.rc522_uart    Attempting to establish a connection at 9600 bps.
debug   libnfc.bus.uart Serial port speed requested to be set to 9600 baud.
debug   libnfc.bus.uart TX: b7
debug   libnfc.bus.uart RX: 12
debug   libnfc.chip.rc522   RD 37: 12
debug   libnfc.chip.rc522   Aborting self test for unknown version 12.
debug   libnfc.bus.uart TX: 01
debug   libnfc.bus.uart RX: 01
debug   libnfc.bus.uart TX: 2f
debug   libnfc.chip.rc522   WR 01: 2f
debug   libnfc.driver.rc522_uart    Restoring baud rate to default of 9600 bps.
debug   libnfc.bus.uart Serial port speed requested to be set to 9600 baud.
debug   libnfc.bus.uart TX: 81
debug   libnfc.bus.uart RX: 20
debug   libnfc.chip.rc522   RD 01: 20
debug   libnfc.driver.rc522_uart    Upgrading baud rate to user-specified 115200 bps.
debug   libnfc.bus.uart Serial port speed requested to be set to 115200 baud.
debug   libnfc.bus.uart TX: 1f
debug   libnfc.bus.uart RX: 00
debug   libnfc.driver.rc522_uart    rc522_uart_write ack does not match (sent 1F, received 00)
error   libnfc.chip.rc522   Unable to write register 1F!
debug   libnfc.driver.rc522_uart    Attempting to establish a connection at 115200 bps.
debug   libnfc.bus.uart Serial port speed requested to be set to 115200 baud.
debug   libnfc.bus.uart TX: b7
debug   libnfc.bus.uart RX: 00
debug   libnfc.chip.rc522   RD 37: 00
debug   libnfc.chip.rc522   Aborting self test for unknown version 00.
debug   libnfc.bus.uart TX: 01
debug   libnfc.bus.uart RX: 00
debug   libnfc.driver.rc522_uart    rc522_uart_write ack does not match (sent 01, received 00)
error   libnfc.chip.rc522   Unable to write register 01!
debug   libnfc.driver.rc522_uart    Could not connect with RC522 at 9600 or 115200 bps.
debug   libnfc.bus.uart TX: 01
debug   libnfc.bus.uart RX: 00
debug   libnfc.driver.rc522_uart    rc522_uart_write ack does not match (sent 01, received 00)
error   libnfc.chip.rc522   Unable to write register 01!
debug   libnfc.general  Unable to open "rc522_uart:/dev/ttyUSB0".
nfc-anticol: ERROR: Error opening NFC reader

This is the current state of the port, right?

[socram8888]

I cannot talk about this fork, but on mine I got as far as nfc-anticoll working, with 115200 bps working too. I am guessing that you may have run into a hardware issue I am experiencing as well which apparently affects some fake/counterfeit MFRC522 chips.

For the development of the '522 patch, I was using a FM17522 chip, which is a compatible version made by Fudan with version register set to 0x88. Official MFRC522 have two different version codes 0x91 and 0x92 (rc522-internal.h).

Yours seem to have a completely different version code (0x12), which probably has some sorta of hardware bug with anything different than 9600bps.

Seeing this was never going to be accepted upstream, I began developing a MFRC522 driver in userland using Python, and while it worked just fine with the Fudan chip, I am experiencing issues with a fake, completely unbranded chip (like, it doesn't even have anything other than the dot marking pin 1) any time I try to change from 9600 to anything else. So there are odds that it's not related to this port, but more like a hardware issue.

Could you try specifying baud rate as 9600, so it uses the default baud rate, and trying again?

[socram8888]

Wait no. For some reason @ehkom modified the rc522-uart driver and broke baudrate changing: https://github.com/ehkom/libnfc/blob/master/libnfc/drivers/rc522_uart.c#L325-L341

You are supposed to write the new baud rate to the MFRC522, and then change the UART baudrate. Otherwise, the MFRC522 keeps trying to read commands with a baud rate different to the one the PC is sending them at.

Also ignoring errors occured when changing UART speed is not a good idea. If that fails, you'll end up with an UART running at a different speed than the MFRC522, with the rest of the library having no idea and trying to use it in an undefined state.

[ildar]

Thanks, @socram8888 , as I'll base on your branch from now on then issue moves to https://github.com/ildar/libnfc/issues/1