> Would you please post a filename and line range(s) for me to look at? You need the draft RFC in one hand. You also need the NTP extension specs - RFC 7822. They are similar, but different in the way the length is interpreted.
git log -p 7fb3de6ee9a88cda7910f697a6002dc3e78f5a7b will show you the details of the fix. I think the hostname is the only text string in NTS-KE. The problem was not length but missing NUL when turning it into a c string. The low level worker routines are in nts.c. The names dealing with the length differences have ke+_ or ex_ prefixes. Maybe we should add a low level routine to handle strings. There are 3 pairs of pack/unpack code. The high level code that packs a KE request is in nts_client.c, the unpack is in nts_server.c The pack/unpack roles are reversed for the KE response. The packing/unpacking for cookies is in nts_cookie.c all in ntpd/ At the top of the loop that scans the KE-response, there is code inside an if (0) that prints out the type and length of each chunk. It's superverbose, but occasionally very handy for debugging. -- These are my opinions. I hate spam. _______________________________________________ devel mailing list devel@ntpsec.org http://lists.ntpsec.org/mailman/listinfo/devel
