**** BEGIN LOGGING AT Sun Apr 25 02:59:56 2021 Apr 25 03:49:32 >KGB-0< https://tests.reproducible-builds.org/openwrt/openwrt_lantiq.html has been updated. (98.2% images and 98.2% packages reproducible in our current test framework.) Apr 25 06:16:30 Build [#67](https://buildbot.openwrt.org/master/images/#builders/64/builds/67) of `realtek/generic` completed successfully. Apr 25 07:25:11 Hauke, or ynezz https://gist.github.com/plntyk/61fe2e7484c7c5f267e89f18d538c195 maybe some change / minor issue with the DSA / upgrade part - as one user had a little trouble in user channel - to make error / docs "more readable" Apr 25 10:31:28 zorun: thanks for updating the release notes Apr 25 10:31:42 I added some more stuff which I could remember yesterday Apr 25 10:32:00 I will check it again and probably publish it in the next hours Apr 25 10:33:02 plntyk: thanks, I will update it Apr 25 11:30:14 Greetings! Apr 25 11:30:42 Is there any method to know from which partition a WRT3200ACM booted to OEM firmware? Apr 25 11:32:11 I'm currently preparing a WRT3200ACM for return and I want to remove all traces of OpenWrt (And especially my configs) from the router ... I already flashed the OEM firmware over the previously active installation of OpenWrt, but I want to make sure it's also gone from the other partition. Apr 25 11:52:19 Ok, I think I could solve it myself :) Apr 25 11:52:38 When the OEM firmware booted and I flash the OEM firmware again, it should go to the other partition ... Apr 25 12:17:19 stintel: lol, i just sent a patch and then noticed you'd already fixed the missing config-5.10 symbol Apr 25 12:17:46 thanks! Apr 25 12:27:37 russell--: welcome :) Apr 25 12:30:37 stintel: i just marked my patch as superseded in patchwork Apr 25 12:30:57 russell--: thanks Apr 25 13:17:38 Hauke: please add this to release notes: the default rootfs patition size changed for targets/devices relying on booting from mass storage (HDD, USB flash, SD card, etc.), so MBR will change and any additional partition will be deleted when sysupgrading. Backup Your data. Apr 25 13:46:36 can a maintainer please approve the running of workflows on https://github.com/openwrt/packages/pull/15510 ? Apr 25 13:46:38 thanks :) Apr 25 13:58:19 Habbie: had a small remark, if you fix that I'll push the button afterwards :) Apr 25 13:58:30 haha Apr 25 13:58:46 i -just- squashed those because I couldn't figure out in which of the two i should bump PKG_RELEASE Apr 25 13:59:30 :D Apr 25 13:59:49 pushed Apr 25 13:59:57 you can actually do it in both I would say Apr 25 13:59:59 oh i messed up the DCO again Apr 25 14:00:02 ok, will do that Apr 25 14:01:27 done and signed off Apr 25 14:01:27 they both change the package so the PKG_RELEASE bump is needed. you could argue that because it's a single PR you can do it only in the last commit of the PR Apr 25 14:01:43 yes, i figure the answer would be 'last' or 'both' or 'squash' Apr 25 14:01:45 *figured Apr 25 14:02:15 hmmm and maybe copy the details from the PR to the relevant commit message Apr 25 14:02:24 makes sense Apr 25 14:03:25 pushed Apr 25 14:03:57 lol, now that approve button disappeared Apr 25 14:04:09 ah Apr 25 14:04:10 it's back Apr 25 14:05:06 interesting that the DCO check runs without approval - a special github feature I guess? Apr 25 14:05:23 magic ;) Apr 25 14:05:43 but the DCO check is cheap, just checks the commit author + message Apr 25 14:06:09 oh, that i get, but apparently github knows this :) Apr 25 14:06:15 the rest of those checks ... take some CPU cycles ;) Apr 25 14:08:37 yep :) Apr 25 14:10:52 i have several more patches and i keep thinking 'i should submit them' but they rely on a release candidate of dnsdist ;) Apr 25 14:37:28 is the pkg maintainer of unbound - Eric Luehrsen - around by any chance? Apr 25 14:39:29 i know a thing or two about unbound - depending on what your question is :) Apr 25 14:41:24 Habbie: i want to forward my queries to a tcp port listening on 127.0.0.1 Apr 25 14:41:59 however everything 127.0.0.1 alike defined in forward zones in /etc/config/unbound doesn't make it into the generated unbound.conf Apr 25 14:42:28 i wonder if it's on purpose or a bug. seems on purpose as everything !=127.0.0.1 /does/ go into the generated unbound.conf Apr 25 14:43:13 it seems about $dns_ast within the uci-parsing/conf-generating scripts within the unbound pkg however i'm a bit lost Apr 25 14:44:54 ah, it explicitly checks whether local_subnet() is true and discards if that's the case Apr 25 14:46:58 but still doesn't make it into the generated conf.., hrmpf Apr 25 14:47:14 is local_subnet true for 127? Apr 25 14:48:35 i think so, but even dropping that check doesn't help Apr 25 14:50:33 without knowing the purpose/intentions it's quite hard to parse and understand the scripts :/ Apr 25 14:50:58 where does 'local_subnet' come from? Apr 25 14:51:47 iptools.sh Apr 25 14:51:53 but found another hint: Apr 25 14:51:53 else Apr 25 14:51:53 case $server in Apr 25 14:51:53 127.*|::0*) Apr 25 14:51:53 # soft brick loop back risk see DNS assist above Apr 25 14:52:17 didn't see it as i was grepping for '127.0.0.1' Apr 25 14:52:20 ah! Apr 25 14:52:56 if it reaches that, it sounds like local_subnet was never true to begin with Apr 25 14:55:05 line 759 may also be in the way later Apr 25 14:55:29 echo " do-not-query-localhost: no" >> $UB_CORE_CONF Apr 25 14:56:47 there's a lot in the way.. still not part of the generated config, but getting lcoser.. Apr 25 14:57:11 just for my curiosity, what is the thing on 127.0.0.1 unbound should talk to? Apr 25 14:59:11 a stunnel instance Apr 25 14:59:36 did you see that unbound has a tls-upstream option? Apr 25 14:59:41 i know, unbound can do tls, however i need mutual authentication and unbound doesn't support tls client auth Apr 25 14:59:46 ack :) Apr 25 15:00:16 ::0* does not match ::1 Apr 25 15:00:56 l 598ff is the deal breaker Apr 25 15:00:56 i wonder what ::0* is supposed to match Apr 25 15:01:15 it drops out with "do nothing", not checking dns_ast anymore Apr 25 15:03:01 unrelated to your problem, there's a typo on line 476 Apr 25 15:04:12 is your dns_ast 1? Apr 25 15:05:05 i see the assist setting does not support 'stunnel' but it supports 'unprotected-loop' Apr 25 15:06:10 oh i found the local_subnet definition in files/iptools.sh Apr 25 15:06:21 it says 'ok' for 127.* and ::1 Apr 25 15:06:34 so if you're reaching that "do nothing" that suggests that dns_ast is not 1 Apr 25 15:07:48 mhm Apr 25 15:07:51 will check Apr 25 15:08:41 so far i parsed it in a way "option dns_assist 1" in UCI sets it to 1 but that didn't help Apr 25 15:08:55 no Apr 25 15:09:00 you need to set it to a supported value Apr 25 15:09:05 bind, dnsmasq, etc. Apr 25 15:09:10 or in your case, literally 'unsupported-loop' Apr 25 15:09:13 sorry Apr 25 15:09:20 'unprotected-loop' Apr 25 15:09:23 line 513 of unbound.sh Apr 25 15:09:41 ah! Apr 25 15:09:53 in the supposed to be forwarded zone or server section? Apr 25 15:10:25 it's in the unbound_zone() function Apr 25 15:10:42 config_foreach unbound_zone zone Apr 25 15:10:45 which is called like that Apr 25 15:10:51 so, forwarded zone i'd say Apr 25 15:11:40 seems like it Apr 25 15:11:44 doesn't work though ;) Apr 25 15:11:47 bah! Apr 25 15:12:09 echo-debugging now Apr 25 15:12:32 'set -x' at the start of unbound.sh may also give some insight Apr 25 15:15:07 "do-not-query-localhost: no" makes it into the generated config now, but the server forward is still dropped Apr 25 15:16:57 but i messed around already quite a bit within the scripts, will just clean up and try again Apr 25 15:18:34 ah, looks good now Apr 25 15:18:53 cool :) Apr 25 15:19:28 'unprotected-loop' was the crucial point - thanks a lot! Apr 25 15:20:16 np! Apr 25 15:20:19 also, https://github.com/openwrt/packages/pull/15511 Apr 25 15:20:21 :) Apr 25 15:21:14 stintel, i see $(AUTORELEASE) would also have caused two bumps on my two commits - if the port used it :) Apr 25 15:21:39 nice catch Apr 25 15:21:47 ah I haven't used that before Apr 25 15:22:31 stintel, i was about to push my unbound PR, then i realised i forgot to bump it, and there was no number there Apr 25 15:23:06 Habbie: so you're involved in powerdns? Apr 25 15:23:12 mirko, yep Apr 25 15:24:09 interesting, using it heavily with a maintenance frontend from like 10 years ago (poweradmin) - upgrading/re-doing on my todo list for years and coming indeed closer Apr 25 15:24:19 ah Apr 25 15:24:24 changing frontends too i bet? Apr 25 15:25:13 Habbie: it's protected via tls client cert auth and users having certs considered trusted, but yes, at least evaluation what's out there Apr 25 15:25:18 migration will be pain in the ass though Apr 25 15:25:28 did you see this list? https://github.com/PowerDNS/pdns/wiki/WebFrontends Apr 25 15:25:54 not yet, it's on my todo list without having started any evaluation yet Apr 25 15:25:59 ack :) Apr 25 15:26:15 but thanks, will put the list right next to my powerdns-todo-list-item :> Apr 25 15:26:19 hehe Apr 25 15:26:33 heh poweradmin. also using that one still :) Apr 25 15:26:50 oh, there's frontend directly using the powerdns API nowadays Apr 25 15:26:57 several in fact Apr 25 15:26:58 that's neat, so you can interchange frontends? Apr 25 15:27:14 well, you could also interchange db-based frontends, depending on how much extra stuff they store Apr 25 15:27:23 and some API-based frontends still have their own extra db for user management etc. Apr 25 15:27:45 but if you ignore all that, then yes, you can :) Apr 25 15:27:46 ok, that helps a lot in not having to put too much thought into which frontend, as i figured i have to stick with what i choose fr the next 10 years as custom database schema Apr 25 15:28:08 PowerDNS-Admin is the community favourite, followed by Opera DNS UI, I believe Apr 25 15:28:12 (i have not tried any of them myself) Apr 25 15:28:47 poweradmin "Last Commit 2019-02-21" - not as bad as i feared.. maybe i could just get away with updating.. Apr 25 15:29:09 but once on it, doing it right i guess - so looking will at PowerDNS-Admin first. thanks! Apr 25 15:30:33 i use poweradmin's features pretty excessively.. from user management over its templates for new zones, etc. Apr 25 15:30:51 so migration will be quite some effort either way Apr 25 15:31:00 right Apr 25 16:00:10 stintel, can you hit the button on https://github.com/openwrt/packages/pull/15511 too? :) Apr 25 16:12:28 Habbie: i can, too - let me at least check for syntax issues before doing so Apr 25 16:12:36 thanks! Apr 25 16:14:43 just taking sensitivies into account, eric wouldn't mind, as he's the maintainer? Apr 25 16:15:12 i don't know Eric - I pinged him because he seems like the best person to verify this Apr 25 16:15:17 (and also the template asked me to do so :) ) Apr 25 16:15:53 yeah, well, it's an obvious typo - i guess it'll be fine Apr 25 16:16:53 done Apr 25 16:17:05 ta :) Apr 25 17:20:38 hi family! Apr 25 17:25:34 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html Apr 25 18:59:51 tmn505: thanks added it Apr 25 20:20:04 I managed to update hostapd package for todays src and I created a PR to update it **** ENDING LOGGING AT Mon Apr 26 03:00:18 2021