[04:20] Bug #1674847 changed: produce package lists for Ubuntu Core versions on the web [06:32] morning [07:07] hey mborzecki ! [07:07] mborzecki: good morning, how are you? [07:07] mvo: hey [07:25] good morning guys [07:25] mvo: last night I realized I have holidays to burn [07:25] zyga: hey hey [07:25] zyga: you can carry over up to 5 days iirc [07:25] mvo: I was thinking about EOWking and just soft-working on few loose ends [07:25] mborzecki: IIRC I have 12-15 ish [07:25] wow [07:28] hey zyga [07:29] zyga: uh, GO [07:29] zyga: like seriously, that is a lot [07:32] * zyga files for a week off [07:37] mvo: we need a trello card for zyga to remind him of taking vacation [07:39] mborzecki: haha - indeed! [07:45] :D [07:50] PR snapd#7909 opened: snap: improve squashfs.ReadFile() error [08:02] morning, mvo: let me know when/if you want to chat about #7908 [08:03] PR #7908: [RFC] boot,devicestate: update modenv and extract kernel [08:03] pedronis: not right now, debugging a issue in the initramfs right now :( [08:04] morning [08:05] hey pstolowski [08:06] pstolowski: hey [08:06] mvo: ok, I'll start looking at its bits in handlers_install.go [08:07] pedronis: thank you! [08:07] pedronis: yeah, let's sync, I think the install steps (the minimal ones) are not that much work [09:02] * zyga goes to check out alternative school for Janek [09:16] PR pc-amd64-gadget#27 opened: grub.cfg-recovery: add run mode chainloading [09:31] mvo: I pushed some improvements to 7908, not quite fully there, we need to chat a bit about /boot layout [09:34] pedronis: ok, should we have a HO in some minutes? [09:35] mvo: yes [09:36] pedronis: ok, I am ready in 3min or so [10:10] mborzecki: hi, can you do a review of https://github.com/snapcore/snapd/pull/7905 when you have a moment , then mvo can do one as well [10:10] PR #7905: cmd/snap-bootstrap: actually parse snapd_recovery_system label [10:11] pedronis: sure, will do [10:11] thx [10:22] PR pc-amd64-gadget#28 opened: grub.cfg-normal: set snapd_recovery_mode= run in run mode [10:27] pstolowski: hi, do we need to chat about services? [10:32] pedronis: hi, yes, that would be great [10:33] * Chipaca steps back from the forum and goes for coffee [10:39] pstolowski: now, in a couple of mins? [10:39] pedronis: in 10 ? [10:39] ok [10:49] pedronis: i'm in the standup ho [10:49] ok [10:49] joining [11:20] PR snapd#7910 opened: boot,image: add skeleton boot.makeBootable20RunMode [11:20] meh, looks like we'll need to pass Device to wrappers [11:20] but probably as a later cleanup [11:24] PR pc-amd64-gadget#27 closed: grub.cfg-recovery: add run mode chainloading [11:33] PR pc-amd64-gadget#28 closed: grub.cfg-normal: set snapd_recovery_mode= run in run mode [11:48] mvo: Q in 7910 [11:49] pedronis: looking [11:50] mvo: updated #7772 [11:50] PR #7772: wrappers: write and undo snapd services on core [11:51] pedronis: of course, I'm a muppet, fixing and adding a test [11:51] mborzecki: nice! [11:58] PR snapd#7911 opened: Uc20 shutdown [12:12] mborzecki, hey [12:13] mborzecki, about lp:1856073 [12:13] you think it is be manually cleaned during tests? [12:15] cachio: hi, yes, left it in the comment https://bugs.launchpad.net/snapd/+bug/1856073/comments/2 [12:15] Bug #1856073: Failed service state recorded by systemd even after snap removal [12:16] mborzecki, yes, I saw that [12:16] I'll make the change in the tests in that case [12:16] ok [12:17] mborzecki, thakns [12:17] zyga, hey, could you please give a second round to #7877 [12:17] PR #7877: tests: avoid mask rsyslog service in case is not enabled on the system [12:17] thanks [12:18] mborzecki, another thing [12:19] mborzecki, I can reproduce this in arch https://paste.ubuntu.com/p/XCwW2VnpXG/ [12:19] mborzecki, I think at some point you reported something similat [12:19] similar [12:26] mvo: reviewed 7910 with some suggestions [12:28] pedronis: cool, thank you [12:31] cachio: no that was antoher test checking whether access to the socket is mediated by apparmor === zyga-laptop is now known as zyga-holidays [12:44] pstolowski: reviewed #7658 [12:44] PR core20#14 closed: run-snapd-from-snap: check for snapd.service existing too [12:45] PR #7658: cmd/snap-preseed: add snap-preseed executable [12:45] PR snapd#7912 opened: boot: write modeenv when creating the run mode [12:45] pedronis: ty! [12:53] * Chipaca gets lunch === ricab is now known as ricab|lunch [13:55] morning folks [14:22] PR snapd#7658 closed: cmd/snap-preseed: add snap-preseed executable [14:38] mborzecki, hey [14:38] pedronis, degville i've sent you an email invitation to edit snap-preseed help doc [14:38] after many tests we left a systemd unit sys-devices-virtual-block-loop1.device [14:38] mborzecki, is it created by snapd? [14:38] any idea? [14:41] pstolowski: thanks, I'll try to get to it today or tomorrow morning [14:41] pedronis: 7912 should be a simple one [14:41] cachio: it's probably a loopback device that's left behind, any chance you could run `losetup -l` there? [14:41] mvo: ok, I need a short break frist though [14:42] PR snapd#7910 closed: boot,image: add skeleton boot.makeBootable20RunMode [14:42] *first [14:42] mborzecki, yes, I'll add more info for next run [14:43] pedronis: yeah, no worries [14:43] cachio: it might be the case that libloop leaves one free device behind as an optimization so that it doesn't need to allocate it again [14:48] mborzecki, thanks, I added a new check for loop devices [14:49] and triggered a new run [14:49] it will be ready on 1 hour aprox [14:49] afer lunch [14:49] * cachio lunch [14:54] need to go to the boys school for a bit, will bbiab [15:17] PR snapd#7852 closed: devicestate: call boot.MakeRunnable() in devicestate <⛔ Blocked> === ricab|lunch is now known as ricab [15:39] hmm MATCH with multiline input behaves somewhat weird [15:40] err MATCH -v actually [15:41] mborzecki: weird how? [15:42] Chipaca: https://paste.ubuntu.com/p/P3hyRbDmcy/ [15:44] Chipaca: we actually have code in tests that does MATCH -v with multiline input [15:47] mborzecki: MATCH -v means "error if there are no lines that don't match" which is a little weird to think about so I wouldn't be surprised if it's used wrong [15:48] Chipaca: heh sooo https://paste.ubuntu.com/p/GD63G4S2sX/ [15:50] hmm that lsmod, snap list, virsh checks are likely wrong [15:50] probably the same in cgroups test :/ [15:58] PR snapd#7913 opened: boot,bootloader: extract kernel in makeBootable20RunMode [16:07] PR snapd#7914 opened: bootstrap: reduce runmode mounts from 5 to 2 steps [16:11] mvo: does 7913 works or we think it works? [16:11] pedronis: I'm testing it now [16:12] pedronis: you can mark it blocked if you want [16:12] pedronis: but it's essentially what I had in mind and tried to describe [16:12] mvo: yes, but I said it makes sense to land only if works [16:12] also it needs more XXX to it :) [16:13] pedronis: heh :) yeah, I need to update the other bits from claudios first [16:13] pedronis: doing this now [16:13] mvo: sounds like bootloader.Options.Recovery needs a different name [16:13] but that's less of an issue [16:14] pedronis: yeah, its a bit too confusing, ideas welcome! [16:14] I need to look again exactly how it is used [16:15] pedronis: \o/ [16:16] mvo: working? [16:17] does it seed again? [16:19] pedronis: yeah, testing if it seeding [16:19] what initramfs is it using? [16:19] pedronis: I mean, unless I miss something (entirely possible) with this PR and the other changes from claudio we should be seeding now [16:19] pedronis: except that snap-bootstrap in run-mode in initramfs is buggy [16:20] pedronis: but xnox is on this :) [16:20] mvo: but what is putting the kernel and the base in the right place? [16:21] pedronis: maybe more is missing and I'm missing this - the kernel will be put in place byhttps://github.com/snapcore/snapd/pull/7913/files#diff-06592349bc3e8eaeecc7bc0cafcb3cacR471 and the snap-bootstrap should open the right base in run mode and make it available in /run/mnt/base [16:21] PR #7913: boot,bootloader: extract kernel in makeBootable20RunMode <⛔ Blocked> [16:21] pedronis: we also have the seed [16:22] I probably don't remember what snap-bootstrap run mode is doing then [16:22] is probably wrong [16:22] somehow [16:23] mvo: to be clear, I'm talking about ubuntu-data/var/lib/snapd/snaps [16:23] mvo: so for runmode I will do two things, reduce runmode steps from 5 to 2 => see my PR. And actually make it just step 1, as initrd will try to mount the "basic" things meaning in runmode, snap-bootstrap is done and dusted in one step. [16:23] pedronis: we setup the uc20 seed there and go through normal seeding iirc, no? [16:23] mvo: no, something is confusing me [16:23] or maybe is right for the first boot but not the other [16:24] I need to stare at snap-bootstrap code [16:24] I suppose [16:25] let me look at it again [16:28] mvo: snap-bootstrap expects the base and kernel to be in filepath.Join(dataDir, "system-data", dirs.SnapBlobDir, modeEnv.Base) [16:28] that's the bit I don't undersand [16:28] when do we put them there [16:28] ?? [16:30] extracting the kernel doesn't do that [16:31] pedronis: oh, right - now I see what you mean [16:31] so I'm not sure how it works [16:31] except magic :) ? [16:31] pedronis: well, we are not there yet :/ [16:31] you said it worked ? [16:32] work != work [16:33] pedronis: sorry, I think I miswrote, I wanted to say that I'm trying to see if it works [16:33] pedronis: and that I need bits from cladio first and that it's a bit of work to get to the point [16:33] mvo: heh, it's ok, I got very confused [16:36] mvo: I gave some comments in 7913 [16:36] pedronis: but it does sound like another wrench in the gears [16:36] pedronis: oh well! so much things to fix still [16:36] mvo: well, it's easy to fix, we need to copy those snaps [16:36] that we knew [16:36] pedronis: in makebootable [16:36] yes [16:36] * mvo nods [16:37] like the UC16 makes symlinkes, this one need to do copies [16:37] right [16:37] pedronis: sorry, brain is a bit fried - I added a todo for myself here [16:41] mvo: anything I can help with? [16:47] pedronis: hm, maybe (if you haven't done already) point the extra xxx in 7913, other than that I think it's mostly working through stuff now [16:47] pedronis: I mean, I try to work through the stuff from claudio [16:48] mvo: sorry, "point the extra" ? address the extra? add an extra ? [16:55] pedronis: sorry again! I mean, you mentioned earlier that 7913 will need a bunch of extra XXX beyond the ones I added there [16:56] pedronis: if you haven't done so already, pointing them out in the PR would be great [17:01] is ubuntu-image expected to build partitions from gadget.yaml for core20 pc gadget yet? [17:02] I just tried building a uc20 image with ubuntu-image with 20/edge snaps and what seems to be a correct dangerous model and I only got the ubuntu-seed partition, the ubuntu-boot partition seems to be missing [17:09] ijohnson: that's expected [17:09] the images have only seed [17:09] boot is added at installation [17:11] pedronis: ah ok so for my purposes then I'll manually create the ubuntu-boot partition since it doesn't seem like trying to get install mode to work just to create that partition [19:05] * cachio afk 10 min [19:21] PR snapd#7915 opened: lkenv.go: adjust for new location of include file [20:03] PR snapd#7916 opened: interfaces/browser-support: add more product/vendor paths