[05:05] good morning [05:10] hey mborzecki [05:11] zyga: morning [05:11] I have some changes to make in https://github.com/snapcore/snapd/pull/7400 [05:11] PR #7400: cmd/snap-update-ns: don't propagate detaching changes [05:11] and then another round ad 2.41 suse package [05:12] zyga: yeah, i saw jstrand's review [05:52] mvo: morning [05:53] mborzecki: goooood morning! how are you? [05:54] mvo: cold ;) +11C outside [05:54] mborzecki: yeah, same here. but much better than the ~40 before :) [05:54] (at least IMO) [05:57] good morning mvo [05:58] mborzecki: +11 wow [05:59] hey zyga [05:59] I'm debugging something that I changed that causes seeding to fail [06:00] it's pretty frustrating because there's no output [06:00] mborzecki: hm, I think we have +9 outside [06:00] I'll try to de-focus and advance some other stuff later, maybe around noon [06:00] mvo: wow, my preference is +20 all day round :P [06:00] mborzecki: check out spain ;) [06:24] mvo: do we have any stats on spread usage? [06:25] zyga: stats in what sense? [06:25] mvo: number of vm/hours used per account [06:25] something like that [06:26] just as trivia [06:26] I wonder what's our record [06:26] zyga: I'm sure gustavo has them, maybe sergio - I don't [06:35] PR snapd#7158 closed: tests: part5 making tests work on ubuntu-core-18 [06:37] off to take my son to school and take some paperwork to my accountant [06:37] bbl [06:51] o/ [07:03] morning [07:03] good morning pawel [07:29] PR snapd#7418 closed: many: pass the rootdir and options to bootloader.Find [07:43] wow, small success [07:43] * zyga tries the other test now [07:47] I'm on the way towards addressing the thing that made MS_SHARED bugfix be more complex in tests [07:47] caused by particular core16 behavior [07:48] re [07:48] mborzecki: hey [07:49] mborzecki: I added syslog to snap-confine [07:49] it helps when things faily mysteriously [07:49] added some draft code to unmap /writable from snaps POV [07:50] that could be the thing to cut the duplicated mount entries on core16 upsetting tests [07:53] zyga: syslog? as in syslog(3)? [07:53] yes [07:53] it shows up even if the invocation is otherwise hidden [07:53] made me realize a mistake earlier [07:56] zyga: got some cleanups in s-c unit tests [07:58] mborzecki: send them over and ping me [08:18] mvo: morning, I looked at #6404 again [08:18] PR #6404: snapstate: auto transition on experimental.snapd-snap=true [08:30] pedronis: thanks, I check it out [08:30] pedronis: I updated the LK one, let me know if you want to go one step further [08:31] mvo: I answered, yes, I think we should, but the change is smaller than what you thought I think [08:31] mvo: we don't use InstallBootConfig at runtime [08:31] I can look myself in a little bit [08:32] pedronis: also 7181 is green, I think we can merge it, yes? [08:32] mvo: you mean 7381 ? [08:33] yes I will merge it and make 7416 smaller [08:34] PR snapd#7381 closed: seed,image,o/devicestate: extract seed loading to seed/seed16.go [08:34] yay [08:34] I identified the core16 + core16 but that duplicated some entries that caused the MS_SLAVE test to fail originally [08:35] the fix is super easy and obviously so [08:35] coming up in 30-60 minutes after some more testing [08:36] pedronis: ok, I will leave the LK one for you then, thank you :) and yes, I did mean 7381 [08:36] (fat fingers) [08:36] it's almost over, the whole thing is aligning :) [08:37] mvo: done, now #7416 is it's proper size (still big, again because tests), also is not the whole thing as I said, more like ~1/3 maybe [08:37] PR #7416: seed/seedwriter: start of Writer and internal policy16/tree16 [08:37] *its [08:38] pedronis: ok, I try to get to it soon [08:46] mborzecki: it passed :D [08:46] mborzecki: mwahaha, the bug is fixed [08:47] running full run across all systems on that test [08:47] I'll start polishing the patch for submission [09:20] mvo: I pushed changed to the LK PR, please have a look when you can [09:20] *changes [09:20] PR snapd#7419 opened: cmd/snap-confine: add unit tests for sc_invocation, cleanup memory leaks in tests [09:21] zyga: ^^ [09:21] zyga: tried to use glib goodies in test_argc_argv() and fix memory leaks there, but gave up eventually as we modify the resulting argv in sc_nonfatal_parse_args() [09:22] pedronis: i'm cleaning up per-revision config stuff a bit and add some more tests [09:23] pstolowski: ok, thx [09:24] ok, heading back home [09:37] PR snapd#7420 opened: cmd/snap-confine: fix /snap duplication in legacy mode [10:04] pedronis: i thought i found a bug in that we only restore revision config on revert but not on refresh; but we explicitely test this behavior in config-revisions spread test. i'm not sure why we wanted that [10:12] pstolowski: you mean a refresh to old revision? [10:13] it makes sense to me though that we don't reapply old config in that case [10:13] pedronis: yep, snap refresh --revision=.. foo [10:14] pstolowski: we don't go find the old data either I think [10:14] revert is really the only op that says please get me back in time [10:14] pedronis: ah i see what you mean [10:15] thanks, all clear [10:15] good i had a spread test; i've added explicit unit test now [10:17] re [10:28] zyga: updated #7419 [10:28] PR #7419: cmd/snap-confine: add unit tests for sc_invocation, cleanup memory leaks in tests [10:31] mborzecki: thank you [10:33] mborzecki: https://github.com/snapcore/snapd/pull/7420 [10:33] PR #7420: cmd/snap-confine: fix /snap duplication in legacy mode [10:40] mborzecki: +1, can you please push a change to include it in the new format [10:40] and reformat [10:40] it will have clang formatting (the -test.c file) so it will be a little bit nicer [10:40] zyga: in the same PR? [10:40] yeah, the -test file is new [10:41] ok [10:42] hmm we should be putting emacs formatting spec in *.[ch] files like systemd does [10:43] fwiw emacs and vim [10:48] mborzecki: +1 [10:49] thoguh I'm not using emacs, not sure if that is picked up universally [10:57] mvo: I need to do an errand downtown today, I will bring my laptop with me but not sure if I will have conditions for the call [10:57] mvo: my update is that I'm zeroing on the re-submission of MS_SHARED bug fix [10:57] mvo: I identified two causes of the failure of the integration test in the original pull request [10:57] mvo: one is under review now [10:58] mvo: the other was sent yesterday and +1'd but -1'd by jamie (code ok, comment tweaks requested) [10:58] mvo: I'll address those and try to propose the fix for MS_SHARED later today if I can [11:07] thanks for the update zyga [11:39] cans omeoenm help me get snap working on kali linux? [11:40] Psil0Cybin: hey, perhaps [11:40] we did this enough times to have some experience [11:41] PR snapd#7421 opened: cmd/snap-confine: unmount /writable from snap view [11:41] mborzecki: one more https://github.com/snapcore/snapd/pull/7421 [11:41] PR #7421: cmd/snap-confine: unmount /writable from snap view [11:41] I made it a draft as it will be probably discussed longer and I want to perform some more analysis [11:47] zyga, tank you becasue i was having problems that my home directory is in /username and not in /home/username [11:47] and it would not let me run any appliations installed etc [11:47] Psil0Cybin: that's a known limitation and it is not supported [11:47] shojld i provide the erorr logs? [11:48] it's not related to the distribution [11:48] so how do i go bout gtting snap.d packages [11:48] Psil0Cybin: no need, we know about this and there is no solution now [11:48] installed [11:48] dang [11:48] you must use the regular location [11:48] so im completely out of luck [11:48] that's the only outcome [11:48] no, just put it in /home/username [11:48] it's not a big deal, is it? [11:48] should not be i just installed kali and created a non root account as per guide [11:49] so what would i do now? [11:49] Psil0Cybin: I don't know much about kali [11:49] I cannot help with that directly [11:49] so what's your actual username? [11:49] and user home directory on the system? [12:10] mvo: I see in the spike code that ubuntu-image is actually installing the recovery bootloader files, is it supposed to work that way in the final code, or prepare-image should do that? === ricab is now known as ricab|lunch [12:11] jdstrand: good day [12:11] jdstrand: I have updated the comment and replied tyo your question on https://github.com/snapcore/snapd/pull/7400 [12:12] PR #7400: cmd/snap-update-ns: don't propagate detaching changes [12:12] jdstrand: I also opened two additional PRs as we discussed [12:12] jdstrand: only one is something I'd like you to review [12:12] jdstrand: it is https://github.com/snapcore/snapd/pull/7420 [12:12] mvo: I'm playing with snap-image to generate an image, after patching prepare-image to generate recovery (mostly from the spike code) [12:12] PR #7420: cmd/snap-confine: fix /snap duplication in legacy mode [12:12] jdstrand: the other one is something that I want to experiment more with, you can see it in https://github.com/snapcore/snapd/pull/7421 if you are curious [12:12] PR #7421: cmd/snap-confine: unmount /writable from snap view [12:13] as soon as 7400 and 7420 land I will propose a v2 of the MS_SHARED fix [12:14] mborzecki: if you can, please do a pass over 7420 [12:14] mvo: (also checking snap-image details with mborzecki) [12:14] mvo: perhaps as well, it's something you should know about [12:24] PR snapd#7422 opened: interfaces: allow reading mutter Xauthority file [12:35] PR snapd#7423 opened: overlord/snapstate: config revision code cleanup and extra tests [12:43] zyga: https://bugs.launchpad.net/snapd/+bug/1828354 didn't have solution, right? or was it fixed elsewhere? [12:43] Bug #1828354: mount event propagation on snapd-mounted tmpfs is incorrect [12:48] PR snapd#7415 closed: tests: fix mountinfo-tool filtering when used with rewriting [12:52] pstolowski: it has a solution but we were unable to land it, my update above (for mvo) was about that [12:52] pstolowski: this is the MS_SHARED bug [12:52] pstolowski: it's close to being fixed now [12:52] pstolowski: I need to land two prerequisites and I can reopen the fix [12:54] zyga: ok, thanks [12:55] zyga: i'm keeping the bug 'in progress' then [13:28] hello! store team here, we've got some slowness issues that are causing timeouts, etc. Info will be here: https://status.snapcraft.io/ [13:28] thank you for sharing tomwardill [13:32] thank you jdstrand!!! [13:35] np [13:36] I'll adjust the git commit message and the comment before landing [13:40] jdstrand: adjusted and squashed for simplicity === ricab|lunch is now known as ricab [13:50] mvo, cachio: thanks for 7288, 7346 and 7350 (system-usernames cleanups)! sorry they were needed [13:50] jdstrand, looking [13:50] jdstrand: no worries, the world is complicated [13:50] mvo: I thought you would stay a bit to discuss snap-image details? [13:50] jdstrand: thank you for making the system-user happen, 2.41 will make a few teams quite happy [13:50] cmatsuoka: yes [13:50] cmatsuoka: sorry, brain [13:51] mvo: 7346 was a particularly nice one [13:51] (validate before create) [13:51] cmatsuoka: I'm back again [13:51] mvo, cachio: hope it didn't cause too much trouble [13:52] cachio: no need to look them up. just you two working through testsuite issues after the feature landed [13:52] jdstrand, everything merged :) snap_daemon did not you [13:52] troubles [14:18] jdstrand: thanks for looking at my apparmor backend PR! [14:20] (k [14:20] ijohnson and pedronis - thank you too! [14:21] i think with the general approval for the direction of this PR i'll work on new tests & the feedback [14:35] hi, out of curiosity, why is that any user can call "snapctl set-health"? [14:48] pstolowski: you're welcome :-) [14:48] ackk: i think that's because any app can set health status of own snap (and doesn't need euid 0 to do that). note you cannot do anything useful by running snapctl set-health manually because of missing context [14:49] pstolowski, you mean it would only show up in snap info? [14:56] ackk: it would show in snap info, yes (not sure why "only"?) - see https://forum.snapcraft.io/t/health-checks/10605 [14:58] pstolowski, I see thanks [14:58] pstolowski, fwiw I meant, if you run it manually from a snap run --shell, you can still set it so that it shows in the status [15:00] PR snapd#7270 closed: overlord/snapstate: save disabled snap svcs on unlink snap tasks [15:00] pstolowski, what would be the context you're referring to? [15:02] ackk: yes. but that's interesting case, i didn't think about it [15:03] pstolowski, I was wondering a malicious user could just snap set-health error "everything is broken" [15:03] ackk: yes i see what you mean [15:10] * cachio lunch [15:14] re [15:14] * zyga reads backlog [15:14] ackk: snap run --shell gives you a cookie [15:15] 🍪 [15:15] ackk: snap run --shell [15:15] echo $SNAP_COOKIE [15:15] AFAIR [15:16] zyga, indeed [15:17] PR snapd#7420 closed: cmd/snap-confine: fix /snap duplication in legacy mode [15:23] PR snapd#7422 closed: interfaces: allow reading mutter Xauthority file [15:25] master moved by 64 revisions [15:25] man, today is busy [15:30] zyga: oh yes! [15:31] pstolowski: you're welcome! [15:39] code review for https://github.com/snapcore/spread/pull/69 would be great, probably needs some tweaks but a early test would be amazing [15:40] PR spread#69: Manage pagination retrieving images from google backend [15:43] ijohnson: are you are aware of this bit of code: https://github.com/snapcore/snapd/blob/master/overlord/snapstate/backend/link.go#L152 [15:43] ? (I was almost forgetting myself) [15:43] it's called by link-snap === pstolowski is now known as pstolowski|afk [16:33] PR snapcraft#2703 opened: build provider: allow configuration of primary apt mirror [17:25] ijohnson: I left this comment: https://github.com/snapcore/snapd/pull/7270#issuecomment-528941176 [17:25] PR #7270: overlord/snapstate: save disabled snap svcs on unlink snap tasks [18:09] PR snapd#6697 closed: interfaces/daemon_notify: add {net,sys}_admin capabilities, update spread test [18:17] run_checks first executes gofmt and in my local machine each time it returns "formatting wrong in following lines;..." [18:18] Am I doing something wrong, because each time I manually run gofmt to complete run_checks and see results [18:20] I am running it 1.13, is it something because of that? [18:32] ardaguclu_: it's a known flaw [18:33] ardaguclu_: ignore it [18:33] ardaguclu_: you can go test ./... [18:33] and that's just as good [18:33] ardaguclu_: there's a few things it doesn't do but you can check the script to see what's that [18:33] PR snapcraft#2703 closed: build provider: allow configuration of primary apt mirror [18:33] ardaguclu_: go fmt has changed behavior across versions [18:37] zyga: thanks, I am going to check it. [18:39] PR snapcraft#2704 opened: extensions: create the gnome-platform directory [18:39] PR snapcraft#2705 opened: extensions: rename extension classes to known names [18:43] thanks pedronis, sorry was afk for a bit, that makes sense [18:54] PR snapcraft#2706 opened: extensions: improve docsting (used in the cli) [18:54] PR snapd#7400 closed: cmd/snap-update-ns: don't propagate detaching changes [19:04] PR snapd#7424 opened: fixme: move snapfrompid into osutils [19:10] PR snapd#7425 opened: channel: introduce Resolve and ResolveLocked [19:15] * ijohnson reboots [19:40] ijohnson: I went to update trello for the "rainy day list" to add 6697 and recalled from my notes something: https://github.com/snapcore/snapd/pull/6697#issuecomment-528987808 [19:40] PR #6697: interfaces/daemon_notify: add {net,sys}_admin capabilities, update spread test [19:41] thanks jdstrand I'll take a look [20:59] * cachio EOW [22:01] PR snapd#7426 opened: cmd/snap-update-ns: clarify sharing comment [22:20] * zyga goes to sleep