On Thu, Feb 07, 2019 at 05:28:33AM -0800, Hal Murray via devel wrote: > > I pushed the start of NTS-KE-client code, partly in order to find things like > this. > > > Job #157857979 ( > https://secure-web.cisco.com/16UbTIDf3-JpOVrQQQf2Lji3hOcnSngcm8aSRfJb9Y7vqNRqMTOrDAM-dEeUuENnnKgsbBlt5T1kxk0tJuU8h_W8-d2fZSmTlM2hoD3x_u1kHHBA7Q7TaQGtlR9KM5u0qGlIVLqSHmPl5sbG9-ILahM1cUTDX3UPj7ae9KmVqic5zdrjGSM6HmMWnm61LcXEdqKRLd26rIBrF0FbmV-2n2U5J6ItN3NW4LvWoRK4RvgAYdncb93uZx4RWHOzsxz5GvrE6Z4zLIWxVM8G5n2ieGg/https%3A%2F%2Fgitlab.com%2FNTPsec%2Fntpsec%2F-%2Fjobs%2F157857979 > ) > > Stage: build > Name: macos-basic > Trace: "_res_9_init", referenced from: > _open_TCP_socket in nts_client.c.1.o > ld: symbol(s) not found for architecture x86_64 > clang: error: linker command failed with exit code 1 (use -v to see > invocation) > > Looks like it needs another library.
Yes, or we just skip the res_init() like we do in libntp/ntp_dns.c I submitted https://gitlab.com/NTPsec/ntpsec/merge_requests/936 to fix this. macOS does have res_init(), but our detection of it doesn't work. The impact is that we don't re-initialize the resolver functions. That's probably OK. We really only need this on network changes that would have updated resolv.conf > Has anybody tried tests/option-tester.sh on macos? I haven't, but that wouldn't have helped here. The CI system caught it in the first commit (yours) that broke it. I think this worked out the way that we intended. Thanks, -Matt _______________________________________________ devel mailing list devel@ntpsec.org http://lists.ntpsec.org/mailman/listinfo/devel
