**** BEGIN LOGGING AT Sun Jul 14 03:00:21 2019 Jul 14 09:36:01 was urngd not supposed to get rid of [ 9.809363] random: ubusd: uninitialized urandom read (4 bytes read, 104 bits of entropy available) Jul 14 11:53:55 Hi. As we discussed during the week, I was going to make an attempt at adding minimal nftables-support to firewall3 (firewall4). Before I started on that task, I wanted to get a feel for nftables and manually converted the filter table. I don't know if they will be of use to anyone, but I put my notes here: https://pastebin.com/spdvjuix Jul 14 11:54:29 notes + rules + config, that is :) For the nftables rules to work, the patch I just submitted to the mailing list needs to applied (on 4.19 at least) Jul 14 15:22:46 blogic: did you see the changes I made to the JSON buildroot patch? Jul 14 15:28:00 aparcar[m]: yes Jul 14 15:30:10 blogic: what do you think? Jul 14 15:31:35 none to be honest Jul 14 15:33:02 blogic: no thoughts or what do you mean? Jul 14 15:47:27 aparcar[m]: did you have a moment to try my sysupgrade change? Jul 14 15:48:30 apropos sysupgrade: happy if someone could have a look at that: https://pb.nanl.de/show.php?id=44738 Jul 14 15:50:36 mirko: ack Jul 14 15:50:41 mirko: looks sane Jul 14 15:51:04 it's untested, but the supposed-to-be-fixed issue is definitely there (found out the hard way) Jul 14 15:51:26 i didn't commit for ages, thus i'm happy if someone would just pick it up Jul 14 15:52:47 mirko: i think you should do it to see if it still works Jul 14 15:52:51 i did that recently aswell Jul 14 15:53:52 ok, it's less the technical setup, rather than figuring out what to do before committing to not get blamed / yelled at :) Jul 14 15:54:14 am i supposed to create a merge request beforehand or sth. like that? Jul 14 15:54:33 am i supposed to contact the current maintainer of sysupgrade before committing? Jul 14 15:54:51 didn't follow policy changes Jul 14 15:55:42 no Jul 14 15:55:55 just commit it to the tree Jul 14 15:56:06 or send a patch to the ml and wait for a response beforehand Jul 14 15:56:18 ack, thanks Jul 14 17:56:43 Hauke: sound good Jul 14 18:11:52 ynezz: ping Jul 14 18:13:13 ynezz: looking at openwrt.git/packages I see various packages beeing uuntouched for 2 years, are these candidates to move to packges.git? Jul 14 18:16:51 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Jul 14 18:46:20 Who does review assignments for openwrt (and not "packages")? Jul 14 18:49:41 aparcar[m]: I've started looking into it last week, here's my wip of the mail https://pad.riseup.net/p/openwrt-packages-draft Jul 14 18:57:00 aparcar[m]: as discussed during the meeting (if I still remember it correctly), we should probably try to somehow define package qualification criteria for being maintained in the master and then do the transfer Jul 14 19:04:04 blogic: I've tried to get rid of "random: ubusd: uninitialized urandom read" with urngd as well, but it would mean starting urngd before ubus AND blocking the system until there is enough entropy available Jul 14 19:04:21 it wasn't simply enough to start urngd before ubusd Jul 14 19:05:21 and for example on mt7620 urngd fails to initialize due to absent high res timer Jul 14 19:22:48 mirko: re that sysupgrade patch, could/should that script just be run with set -e? is that the only place the exit code isn't/wasn't checked? Jul 14 20:20:48 it does get rid of those warnings on multi-core units Jul 14 20:21:20 on mt7621, I get none Jul 14 20:27:29 mangix: BTW I've just tested `sudo ./scripts/qemustart x86 64 --smp cpus=2 --enable-kvm --rootfs openwrt-x86-64-combined-squashfs.img` and it still has the warnings Jul 14 20:39:41 rmilecki: BTW what's the reason for having libroxml in the master tree? I don't see any user of this library Jul 14 20:40:33 rmilecki: background https://pad.riseup.net/p/openwrt-packages-draft (that package is on the list, so that's why I'm asking) Jul 14 20:45:07 ynezz: we can move it Jul 14 20:46:08 ynezz: we probably won't be moving it for the 19.07 release Jul 14 20:46:27 ynezz: so for the small bump with regression fix, I'll push that commit anyway, at least to 19.07 Jul 14 20:47:46 yeah, sure, just checking if I didn't missed something while exploring the package interdependencies Jul 14 20:49:33 and 19.07 is a good point as well, as backporting of the fixes is going to be more work as well Jul 14 20:49:50 I mean once the package is moved Jul 14 20:52:12 ynezz: sure. But on read hardware I don't see it Jul 14 20:53:03 https://gist.github.com/neheb/56516a699a90f012d277c572d09e4198 Jul 14 20:53:09 **real Jul 14 20:53:54 The situation is the same with my Turris Omnia Jul 14 20:54:12 on my ath79 devices with a single core, I get them Jul 14 21:00:05 karlp: i was thinking about the same, but got scared of breaking too much at one :/ Jul 14 22:07:01 jow: any idea why there is no kernel-debug.tar.bz2 in the release builds, like for example http://downloads.openwrt.org/releases/18.06.4/targets/ipq806x/generic/ ? **** ENDING LOGGING AT Mon Jul 15 03:01:15 2019