Hi, Here's the summary of the IRC meeting.
--- COMMUNITY MEETING Place: #openvpn-meeting on irc.freenode.net Date: Wed 24th February 2021 Time: 11:30 CET (10:30 UTC) Planned meeting topics for this meeting were here: <https://community.openvpn.net/openvpn/wiki/Topics-2021-02-24> Your local meeting time is easy to check from services such as <http://www.timeanddate.com/worldclock> SUMMARY cron2, dazo, d12fk, lev, mattock and plaisthos participated in this meeting. --- Talked about OpenVPN 2.5.1 release. The release was tagged today morning. Packages were being built during the meeting and the release was pushed out after the meeting. -- Plaishos is planning on writing an email announcing the alpha version of openvpn2+dco. It was noted that we need to pick up slack on "master/2.6" stuff or merging these DCO bits will become extremely painful. -- Lev is working on ovpn-dco-win stability - added missing locks, SAL annotations, running under driver verifier / KMDF verifier etc. Right now he's adding IPv6 support. -- Decided to move the rest of the discussion topics to next week as cron2 needed to split early and mattock wanted to push out the 2.5.1 release. --- Full chatlog attached
(12:30:04) cron2_: hah! (12:30:05) cron2_: made it! (12:31:19) d12fk: hiho (12:31:56) ***plaisthos hides under a rock (12:32:09) d12fk: you hide too loud (12:32:29) cron2_: hiho d12fk :) (12:35:25) lev__: hello (12:35:25) dazo: hey! (12:35:28) mattock: hi (12:37:13) mattock: msi builds just finished (12:37:16) cron2_: so. I do not have very much time today. My wife needs milk for her coffee... I will be pained otherwise... :-) (12:37:18) mattock: so, topics for today? (12:37:23) cron2_: 2.5.1 release! (12:37:29) mattock: it is ongoing, end of story :D (12:37:37) mattock: the less we babble here, the faster :D (12:37:52) mattock: I'm about to push the windows msi packages for testing (12:37:53) mattock: then testing (12:37:54) cron2_: yeah. For the others: 2.5.1 has been tagged and pushed this morning. (12:37:56) mattock: then release notes etc. (12:38:00) dazo: I'm running test builds of 2.5.1 for Fedora 34 now ... kicking of the Copr builds soon after that (12:38:20) cron2_: cool. (12:38:52) plaisthos: I will probably write a email announcing the alpha version of openvpn2+dco (12:39:02) plaisthos: this week or early next week (12:39:06) cron2_: cool! (12:39:37) cron2_: we need to pick up slack on "master/2.6" stuff, otherwise merging these bits will be extremely painful (12:39:41) plaisthos: draft for the announcement so far: https://github.com/schwabe/openvpn/blob/dco/Readme.dco.md (12:40:18) plaisthos: relax, it is only 51 commits ahead of master ;) (12:40:36) dazo: :-D (12:40:46) cron2_: I'm worried about the SRV patch from themiron, which I assume to be conflict prone (12:40:58) plaisthos: should be too bad (12:41:05) plaisthos: I barely touch that part (12:41:16) cron2_: "not" missing, I hope :-) (12:41:29) dazo: "shouldn't be too bad" ... or ... "would be too bad" ;-( (12:41:30) dazo: ;-) (12:41:34) mattock: fyi: https://build.openvpn.net/downloads/releases/OpenVPN-2.5.1-I601-amd64.msi (12:41:37) mattock: I'll smoke-test that one (12:41:43) cron2_: and I want to get rid of the heap of half-acked delayed-auth patches :-) (12:42:21) plaisthos: yeah, there is also the patch that fixed a bug for jjk but I never heard back from him (12:43:01) lev__: I was working on ovpn-dco-win stability - added missing locks, SAL annotations, running under driver verifier / KMDF verifier etc (12:43:05) dazo: plaisthos: If you get the last outstanding bits of the 4 last delayed-auth patches updated, I can have a quick look at those (12:43:08) lev__: now adding IPv6 support (12:43:18) cron2_: the announcement sounds good. I wonder about the "IPv6 mapped IPv4 addresses", but this is maybe better discussed this afternoon, outside the meeting (12:43:28) cron2_: lev__: cool (12:43:55) plaisthos: cron2_: currently bug/limitation in ovpn-dco itself. (12:44:51) cron2_: plaisthos: but what is the limitation? "receiving an IPv4 connection on an IPv6 socket, and passing the v4-mapped v6 socket to the kernel"? (12:45:02) cron2_: or "inside"? (12:45:30) cron2_: payload should never ever see v4-mapped addresses - they are illegal to be "put on the wire" (12:45:46) plaisthos: cron2_: the first thing (12:46:18) plaisthos: for inside the the tunnel that is something we leave to the linux kernel :) (12:46:26) cron2_: okay. These ugly code paths... - but as long as we have no dual-listen-sockets, we'll have to make this work (12:46:51) cron2_: I would be totally OK with "we have no dual-stack sockets anymore", but that requires "dual listen sockets" (12:46:52) plaisthos: Yeah multiple sockets might came later since I understand those code paths better now but one step at a time (12:47:21) cron2_: whatever is the more sane path forward for dual-stacked servers (12:47:35) plaisthos: the other strange multi ip option is also not supported (12:47:48) cron2_: which one? (12:47:51) plaisthos: that uses recvfrom/sendto to allow multiple ips on one socket (12:48:02) cron2_: oh, --multihome (12:48:09) plaisthos: yes (12:48:31) cron2_: that needs to go to the limitations section (12:48:46) cron2_: (because it breaks UDP-mode on hosts with multiple IP addresses) (12:49:03) plaisthos: yeah (12:49:11) cron2_: UDP server, that is, client works (12:50:26) cron2_: ah, dafuq. Need to go over and fix my ISP... something exploded. Sorry. (12:50:46) mattock: windows smoke testing passed (12:53:59) mattock: anything we need to discuss? I'd love to wrap up the release if there's nothing (12:56:40) mattock2 [~mattock@openvpn/corp/admin/mattock] è entrato nella stanza. (12:56:40) modalità (+o mattock2) da ChanServ (13:02:39) cron2_: let's move the remaining stuff from *last* week to *next* week :-) (13:12:41) becm [~b...@port-92-196-82-167.dynamic.as20676.net] è entrato nella stanza. (13:14:40) mattock2: +1
_______________________________________________ Openvpn-devel mailing list Openvpn-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openvpn-devel
