Evgeny Vereshchagin
Evgeny Vereshchagin
Looks like it works. `_http._sub._qotd._tcp.local.` was advertised by resolved and the mDNS responder on another machine saw it. The "Deregister" method deregistered the service too in the sense that goodbye...
Looks like it violates the RFC (a bit) in the sense that `resolved` advertises RRs pointing `_services._dns-sd._udp.local` to `_http._sub._qotd._tcp.local.`: ```sh -bash-5.2# dig @192.168.124.17 -p 5353 _services._dns-sd._udp.local ANY +notcp +noall +ans...
I have to say that these options are so weird that the only thing I've broken so far is my packet generator :-) I'll get to it once I fix...
I pulled the latest version of `dnr_v4` (from the "fixup" commit) and the option seems to be there: ```sh "dhcp.option.type": "162", "dhcp.option.type_tree": { "dhcp.option.length": "54", "dhcp.option.value": "00:19:00:01:09:03:61:62:63:03:78:79:7a:00:04:01:02:03:04:00:01:00:04:03:64:6f:74:00:19:00:01:09:03:78:79:7a:03:61:62:63:00:04:05:06:07:08:00:01:00:04:03:64:6f:74" } ``` so...
The ND option parser was updated in https://github.com/systemd/systemd/pull/31584. Looks like the router advertisement part should be updated here too.
The CIFuzz MSan failure is unrelated. It should be fixed in https://github.com/systemd/systemd/pull/31802
The good news is that it seems that valid encrypted DNS options can get past the ND option parser now. The bad news is that it's possible to bring down...
Looks like now it can be crashed a bit differently: ```sh Assertion 'r > 0' failed at src/network/networkd-ndisc.c:1432, function ndisc_router_process_encrypted_dns(). Aborting. Program received signal SIGABRT, Aborted. 0x00007f1e254b0884 in __pthread_kill_implementation ()...
Let's wait for @yuwata here. I have to admit I don't fully grok what's going on with the network-next branch and where the "IPv6 compliance" part (which heavily touches the...
> I guess we can get the fuzzer to look at the ndisc options now? The ndisc fuzz target fuzzed options up to a certain point. https://github.com/systemd/systemd/pull/31584 took it to...