Re: [knot-dns-users] problem with Knot 2.4.0 on FreeBSD 10.3 and 11.0 - i386

From: "Leo Vandewoestijne" <knot(a)dns.company> To: "knot-dns-users" <knot-dns-users(a)lists.nic.cz> Sent: Monday, 26 June, 2017 12:23:36 Subject: Re: [knot-dns-users] problem with Knot 2.4.0 on FreeBSD 10.3 and 11.0 - i386

Hello, [atomic]: I'm still having the problem: http://beefy10.nyi.freebsd.org/data/110i386-default/444089/logs/knot2-2.4.3… I tried to solve the problem, but when upgrading to 2.5.1 I marked it broken for i386. Meanwhile, after this problem: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=220241 I've upgraded the port to 2.5.2. Now when I test explicit with gcc or clang I still get atomic issues on i386. [libknot]: The cascading result is knot-resolver also doesn't install (on i386), since it's depending on knot2 to install libknot. So, I in my dev env I made a seperate port for libknot that would use knot2 distro; that installs on i386 without atomic issues, and so i386 users would be able install knot-resolver again. But it's a bit complicated to avoid collisions with these two (or actually three) when using both knot2 and knot-resolver (or removing knot2 and keeping knot-resolver). So before I continue in making a meta-port or the knot2 port, just for the sake of libknot, I'm wondering; wouldn't other OS'es packaging systems not benefit from having libknot seperate from knot? -- Met vriendelijke groet, With kind regards, Leo Vandewoestijne &lt;leo(a)dns.company&gt; <www.dns.company> On Tue, 07 Feb 2017, daniel.salzman(a)nic.cz wrote: ____________________________________________ >8 etc. _______________________________________________ knot-dns-users mailing list knot-dns-users(a)lists.nic.cz https://lists.nic.cz/cgi-bin/mailman/listinfo/knot-dns-users