[PATCH 1/4] Add Reliable Asynchronous Transfer Protocol

[Peter Korsgaard]

>>>>> "Sascha" == Sascha Hauer <s.hauer at pengutronix.de> writes:

Hi,

>> > +	crc_expect = cyg_crc16(data, len);
 >> 
 >> Why crc16? RFC916 states that the checksum is the inverted 1s complement
 >> 16bit sum, E.G. like TCP/UDP/IP header checksums:

 > Indeed CRC16 is not correct here according to RFC916. We can implement
 > the original checksum algorithm, I'm not sure if that's a good idea
 > though. Plain checksumming does not detect lost 0 characters.

True, but missing/extra 0 characters would normally get caught by the
length mismatch, so I'm not sure it is a real problem.

-- 
Venlig hilsen,
Peter Korsgaard