- May 19, 2014
-
-
Lubos Slovak authored
-
Lubos Slovak authored
Mostly wrong test counts. In the pkt tests, the comparison test is failing, but don't know why. It's somehow related to the OPT RR, because when I remove its addition, it doesn't fail anymore. However, I don't know what's the problem yet.
-
Lubos Slovak authored
-
Lubos Slovak authored
- knot_edns_init_from_params() modified to add also empty NSID OPTION. - knot_pkt_add_opt() documented and modified usage not to insert emtpy NSID to reply. - OPTION printing in utils was not implemented.
-
Jan Včelák authored
-
Jan Včelák authored
-
Jan Včelák authored
-
Jan Včelák authored
-
Jan Včelák authored
-
Daniel Salzman authored
-
Jan Kadlec authored
-
Jan Kadlec authored
change.
-
- May 16, 2014
-
-
Jan Kadlec authored
-
Jan Kadlec authored
- remnant of libknot separation I guess.
-
Jan Kadlec authored
-
Jan Kadlec authored
-
Jan Kadlec authored
-
Jan Kadlec authored
-
Marek Vavruša authored
Signed-off-by:
Jan Kadlec <jan.kadlec@nic.cz>
-
Jan Kadlec authored
-
Jan Kadlec authored
-
- May 15, 2014
-
-
Daniel Salzman authored
-
- May 14, 2014
-
-
Lubos Slovak authored
Conflicts: tests/requestor.c
-
Lubos Slovak authored
Was adding the NSID even if it was empty. Note: maybe it would be better not to support adding empty options. TODO: add some checks to unittest, related to this.
-
Lubos Slovak authored
-
Lubos Slovak authored
-
Lubos Slovak authored
- Unittest for setters. - Enabled EDNS unittests. - Fixed unittest messages. - Fixed wrong parameter check in edns_has_option(). - Fixed all EDNS unittests. All tests pass. Hopefully it means something...
-
Lubos Slovak authored
Unittest for requestor.c needs fixing (wrong argument).
-
Lubos Slovak authored
Not even compiled yet.
-
- May 13, 2014
-
-
Marek Vavruša authored
Conflicts: src/knot/server/udp-handler.c
-
Marek Vavruša authored
-
Marek Vavruša authored
-
Lubos Slovak authored
-
Marek Vavruša authored
-
Marek Vavruša authored
protocol-agnostic response, copied socket for deferred cleanup The copy is needed so the TCP worker can close the socket and the connection is held until the UPDATE request is cleared.
-
Marek Vavruša authored
-
Marek Vavruša authored
The MSG_WAITALL doesn't work as expected everywhere and can return partial answer (for example with signal interrupt). Also implemented timeouts for TCP receive.
-
Marek Vavruša authored
-
Marek Vavruša authored
Previously, response was paired to query according to the Message ID and QUESTION. This however doesn't work for zone transfers, as some servers omit the question section for multi-packet answers completely.
-
Marek Vavruša authored
-