[05:02] morning [07:06] morning [07:07] good morning pstolowski [07:09] good morning guys :) [07:12] good morning zyga-mbp ! [07:12] * zyga-mbp is semi-trapped in the forest today [07:13] the main road out is closed by lumberjacks and we're not sure our car can go through the service roads [07:15] zyga-mbp: I am very envious of your working environment :) I am just in my house in the suburbs... forest sounds awesome 🌲 [07:16] amurray I went out for an early-morning walk and it's lovely [07:17] nice [07:18] amurray (terrible colors, reality was less contrasty) https://twitter.com/zygoon/status/1437314457567387648 [07:24] oooh pretty [07:26] there's also a lake but I won't mention that ;) [07:32] zyga-mbp: cruel ;) [07:38] PR snapd#10770 opened: systemd: use text.template to generate mount unit [07:52] hi all :-) [07:53] @mvo small PR for spread, https://github.com/snapcore/spread/pull/134 [07:53] PR spread#134: Use greedy regex for $(HOST: ) [07:53] zyga-mbp: we also went in the woods on Saturday, founds quite a few mushrooms -- though we went late, so the best ones had already been collected, it seemed [07:54] I found a few during my morning walk but I left them intact [07:54] the woods should be full of shrooms soon, it's still warm and it rained recently [07:54] too bad I'm at work :P [07:55] zyga-mbp: oh, thanks for this PR! [07:55] zyga-mbp: zelda-style gathetring time? ;) [07:55] hey mardy [07:56] pstolowski less chu-chu jellies and chests ;) [07:56] I've been using that locally but please double check if it makes sense to your eyes [07:56] I didn't do any tests, just a quick change & shared it quickly not to forget [07:58] tests are left to the reader ;-) [07:59] I would not tell *that* to mvo [07:59] but spread has remarkably few tests [08:06] Bug #1943419 opened: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [08:28] zyga-mbp: I quickly looked at the PR, if you could add an example how you noticed it and a "before/after" the change that would be lovely, then I can write a test for you [08:28] yeah, I can even write the test but I didn't get to it yet, it's something I bumped into yesterday [08:28] the example was $(HOST: $(echo example)) [08:29] or maybe [08:29] the example was $(HOST: $(echo example)-foo) [08:29] anyway, I'll update it with unit tests [08:29] zyga-mbp: cool, that is all I need [08:29] the real code was in libzt test suite, I've updated it to support local snap store proxy for caching [08:30] and I've used quite a bit of HOST variables for that [08:30] I can paste the real thing but the point is that the old spread stopped at the first right ")" [08:30] so $(HOST: ...) could never have ) insdie [08:30] *inside [08:39] reviews *very* welcome on this, as I plan to add more tests based on this new EnsureAtLeast() method: https://github.com/snapcore/snapd/pull/10748 [08:39] PR #10748: systemd: add AtLeast() method, add mocking in systemdtest [08:48] re [08:48] hey mborzecki [08:49] zyga-mbp: hey [08:53] mardy: sure, done [09:21] anyone running debian sid or kali? https://forum.snapcraft.io/t/snap-apps-not-running-in-kali-linux/26542 i think it's what mvo mentioned that we need to enable snapd.apparmor.service on that distro [09:32] mborzecki: I did upload a new snapd that enables that late last week, I wonder if the user has the latest version from sid [09:36] mvo: they're running kali which i think pulls in packages from sid? [09:37] mborzecki: if so, maybe they can just test the sid package? [09:37] mvo: asked them about package version, let's see what they come back with [10:00] mvo: thanks, I added the test. Now I need one more +1 from someone else :-) [10:03] PR snapd#10738 closed: o/hookstate/ctlcmd: Implement snapctl refresh --show-lock command [10:04] mardy: i made a remark there about adding a test for EnsureAtLeast [10:21] pstolowski: mmm... I don't see your comment. But indeed, I'll add a test [10:22] mardy: oh dang, it was still pending, I didn't click submit last week [10:22] you should see it now [10:28] yep, thanks === Mirv__ is now known as Mirv [10:55] hm fun with selinux, caps and bpf :/ [11:13] Bug #1943419 changed: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [11:16] Bug #1943419 opened: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [11:19] Bug #1943419 changed: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [11:29] mvo: what shall we do about https://github.com/snapcore/snapd/pull/10540 ? [11:29] PR #10540: cmd/snap-confine: handle CURRENT_TAGS on systems that support it [11:31] Bug #1943419 opened: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [11:34] Bug #1943419 changed: LXD snap auto refresh stuck at copy snap data phase for 16+ hours without making progress [11:35] mborzecki: I think merge - it has a +1 from security and we evaludated the alternatives so dlopen() it is it seems [11:39] ok [11:48] mvo: pushed one more tweak for a test failing on 14.04 and then we should be ready to land it [12:57] mborzecki: nice [12:59] PR snapd#10770 closed: systemd: use text.template to generate mount unit [13:49] PR snapd#10771 opened: DRAFT: Tests reproduce uc20 boot error [13:51] miguelpires: hi! About 10767, other possibilities are "Finished", "Processed", "Completed"... Though I'm fine even with Stopped. [13:52] mvo: can you take a look at https://github.com/snapcore/snapd/pull/10661 ? the tests are looking good, so if they are green we can probably land it, or wait for pedronis? [13:52] PR #10661: cmd/libsnap-confine-private: device cgroup v2 support [13:53] mborzecki: I would say land it, it has no needs-samuele-review and it was carefully reviewed by the team/security [13:55] mardy wanted to avoid implying that the work was actually completed since it may not have been (in the case of Error, Undone and Hold) and Stopped met that criteria I think [14:36] mvo: i think i got it now in https://github.com/snapcore/snapd/pull/10754, let's land it once it's gree and unblock builds in LP [14:36] PR #10754: packaging, tests/lib/prepare-restore: build packages without network access, fix building debs with go modules [14:41] mborzecki: nice [14:41] mborzecki: thanks so much for working on this! did you see https://github.com/snapcore/snapd/pull/10754/files#r705436649 btw? [14:41] PR #10754: packaging, tests/lib/prepare-restore: build packages without network access, fix building debs with go modules [14:42] mvo: heh, missed that, i can push a patch on top, or a followup [14:43] mborzecki: maybe on top is ok, ian had some similar commments afaict [14:43] mborzecki: and 1 more hour for the spread run is probably not too terrible at this point [14:43] ok [14:49] mvo: ok, updated and managed to cancel the previous run [14:55] mborzecki: ta [15:34] PR snapd#10748 closed: systemd: add AtLeast() method, add mocking in systemdtest [15:38] mborzecki: 10661 looks mostly fine, I guess that is ready for merge once spread ran? [15:40] mvo: yeah, i think so [15:40] mardy you gave thumbs down previously, can you take a look again? [15:42] * cachio afk [15:42] lunch [15:45] mborzecki: nice job, things are coming together it seems :) [16:04] @bboozoo one question on 10661 about the security [16:04] err [16:04] @bboozzoo ^ [16:31] ijohnson: replied, makes sense? there's CAP_BPF which in theory can separate bpf operatiosn from CAP_SYS_ADMIN, also selinux can mediate respective operations (map_create, read, write etc.) [16:32] * ijohnson[m] looks === not_phunyguy is now known as phunyguy [21:44] * cachio afk