| mup | PR snapd#6133 opened: tests: fix how pinentry is prepared for new gpg v 2.1 and 2.2 <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6133> | 00:06 |
|---|---|---|
| mup | PR snapd#6134 opened: tests: skip opensuse from interfaces-openvswitch-support test <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6134> | 02:52 |
| mborzecki | morning | 06:14 |
| zyga | good morning | 06:31 |
| mborzecki | zyga: hey, i usually stay away from the computer over the weekends, but i tried 2.36.1 snapd in opensuse TW, seems to be working fine, nvidia, basic AA | 06:34 |
| zyga | that's great, thank you for trying! | 06:34 |
| kjackal | hello snappy people, did we do a core release? | 07:36 |
| mvo | kjackal: we have not made a stable release yet, 2.36.1 is still under QA | 07:38 |
| zyga | hey mvo | 07:45 |
| mvo | zyga: good morning! | 07:45 |
| zyga | hey, just getting into the office | 07:45 |
| mborzecki | mvo: hey, had to restart travis job in 6039 as it timeouted | 07:47 |
| mvo | mborzecki: :( ok | 07:47 |
| mvo | mborzecki: thank you | 07:47 |
| mborzecki | mvo: heh, and another restart, was it the same yday? | 07:48 |
| zyga | mvo: what should I attack first? the bug that a customer found or anything else? | 07:49 |
| mvo | zyga: please check my PR on top of yours, it has a spread test. I'm cleaning the mock gpio a bit more as we speak but the basic idea is the same. but you can also wait for ~10min, then I shall push this small cleanup. either way is fine | 07:49 |
| zyga | super, I'll check it out | 07:50 |
| mvo | zyga: we should probably also switch the other gpio spread tests to the new method (6128). I pushed the updated code I may need to tweak a bit more, spread test is still running but this should allow you to see what is going on | 08:04 |
| pstolowski|afk | morning | 08:05 |
| === pstolowski|afk is now known as pstolowski | ||
| zyga | mvo: thanks, I'm reading reviews and should get to yours in a moment | 08:06 |
| zyga | hey pawel | 08:06 |
| zyga | mvo: https://github.com/snapcore/snapd/pull/6128/commits/dd08e0ed21fe2dd9030887d386f0971391dd0459 hmm | 08:08 |
| mup | PR #6128: overlord/ifacestate: setup security backends phased by backends first <Created by mvo5> <https://github.com/snapcore/snapd/pull/6128> | 08:08 |
| zyga | mvo: the reason I did it that way was to not change the behaviour of other code | 08:08 |
| zyga | that's not something we catch in tests today | 08:08 |
| zyga | but this changes the order of setup in all the interfaces | 08:08 |
| zyga | whereas the initial branch was just changing it for gpio (in practice, because that is the only consumer) | 08:08 |
| zyga | I'd like to postpone that tweak to 2.37 | 08:08 |
| mvo | zyga: my PR is aimed for edge | 08:09 |
| pstolowski | zyga: hey, we need to talk about the mapper and implicit slots, can we have a chat in 10-15 mins? | 08:09 |
| mvo | zyga: we need to find out if the user who is affected is fine with refreshing to edge or if we wants something else | 08:10 |
| mvo | zyga: but unless we know that I think its ok to build something for edge | 08:10 |
| zyga | mvo: I think edge might be okay for them but regardless I think we should do a deeper inspection of the impact of the change in ordering | 08:10 |
| zyga | note that I fused the setup of the principal snap and the affected snaps | 08:10 |
| zyga | so now we do things in quite a different order in practice | 08:11 |
| mup | PR snapd#6135 opened: packaging/arch: fix bash completions path <Simple š> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6135> | 08:11 |
| mvo | zyga: fair enough, I feel slightly uneasy about this too. is gpio the only thing that uses systemd? | 08:11 |
| zyga | yes | 08:11 |
| mborzecki | trivial PR ^^ | 08:11 |
| zyga | don't get me wrong, I think what you did is _desired_ | 08:11 |
| zyga | it's just that I wanted to do that separately, once the fire is put out | 08:11 |
| zyga | because it has bigger impact | 08:11 |
| mvo | zyga: yeah | 08:11 |
| zyga | and it may not be as fully tested | 08:11 |
| mvo | zyga: ok | 08:11 |
| zyga | since our unit tests use one test backend | 08:11 |
| zyga | they don't exercise those interactions | 08:12 |
| mvo | ok | 08:12 |
| zyga | so we only have spread tests to, hopefully, catch any issues | 08:12 |
| mvo | pstolowski: 6132 should be an easy merge | 08:12 |
| pstolowski | looking | 08:13 |
| mvo | zyga: ok, lets cherry pick the spread test to your PR and we can discuss the approach in the standup | 08:13 |
| zyga | ok | 08:13 |
| mup | PR snapd#6132 closed: many: some small doc comment fixes in recent hotplug code <Created by pedronis> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/6132> | 08:15 |
| zyga | mvo: one thing I don't get about the spread test, where is the GPIO pin definition? | 08:18 |
| zyga | the slot that is | 08:18 |
| mvo | zyga: in prepare.sh | 08:21 |
| mup | PR snapd#6136 opened: tests: run interfaces-openvswitch-support on arch again, the arch bug is fixed <Created by mvo5> <https://github.com/snapcore/snapd/pull/6136> | 08:21 |
| mvo | zyga: it is part of the core snap | 08:21 |
| zyga | ah, I din't notice thatt | 08:21 |
| mvo | zyga: we unpack and modify the core snap during setup to add one | 08:21 |
| mvo | zyga: its quite hidden :( | 08:21 |
| mvo | zyga: also it is not done for core18 so there is work there as well, I want to look at this too | 08:22 |
| mvo | zyga: plus we have some existing gpio tests which can also be simplified | 08:22 |
| mvo | zyga: thanks for the review, I address your points now :) I pushed a small tweak already | 08:26 |
| zyga | mvo: thank *you* for writing this :) | 08:26 |
| zyga | I didn't think about mocking the whole thing this way | 08:26 |
| mborzecki | mvo: one more fix needed in 6039, will push it in a bit | 08:28 |
| mvo | mborzecki: what is the fix ? spread test that checks the error message? thanks for handling this! | 08:28 |
| mborzecki | mvo: yup, the spread test | 08:29 |
| mup | PR snapd#6131 closed: store: remove unused currentSnap and currentSnapJSON <Simple š> <Created by chipaca> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6131> | 08:38 |
| pedronis | pstolowski: hi, I finished reading the already landed hot plug code, thanks for looking at my small doc tweaks. Of that the only bit I wasnt sure about is hotplug/spec.go, it makes it look like a backend but it's not, also not sure why it's organized exactly like it is, but anyway nothing deep, can be reconsidered when I get further. I will start reviewing new code today, when I'm not in meetings :) | 08:41 |
| pstolowski | pedronis: thanks. sure, we can discuss if hotplug spec should be re-organized. it's an API for interfaces, not sure if this is clear from what already landed, but the camera interface is an example of how this is used - https://github.com/snapcore/snapd/pull/6098 ; i've done it like that to follow the general pattern of *.Specifificaton objects that we pass to interfaces, this was from suggestions from early reviews | 08:47 |
| mup | PR #6098: interfaces/builtin: support hotplug for camera interface <Hotplug š> <Created by stolowski> <https://github.com/snapcore/snapd/pull/6098> | 08:47 |
| pedronis | pstolowski: yes, I understand but is used very differently, I'm not sure if the similiraty helps or is confusing | 08:48 |
| pedronis | anyway there is no action atm | 08:48 |
| pedronis | just pointing out that I wondered a bit about this bit | 08:49 |
| pstolowski | sure, i see | 08:49 |
| mvo | zyga: updated 6128 | 08:55 |
| zyga | mmm | 08:55 |
| zyga | neat, thank you! | 08:56 |
| pstolowski | zyga: do you have some time to talk about mapper/implicit slots? | 09:00 |
| pstolowski | #6124 is trivial, needs 2nd review | 09:04 |
| mup | PR #6124: tests: simple reproducer for snap try and hooks bug <Simple š> <Created by stolowski> <https://github.com/snapcore/snapd/pull/6124> | 09:04 |
| zyga | pstolowski: I think so | 09:09 |
| pstolowski | zyga: standup HO? | 09:15 |
| mvo | pstolowski: occupied ;) | 09:15 |
| pstolowski | :) | 09:15 |
| zyga | pstolowski: can we have a call on tg? | 09:16 |
| zyga | pstolowski: just need to print one last invoice | 09:16 |
| zyga | sorry, monthly accounting time | 09:16 |
| pstolowski | zyga: sure. nb, i recommend switching to queterly ;) | 09:17 |
| zyga | queterly? | 09:17 |
| zyga | ah | 09:18 |
| zyga | I get it | 09:18 |
| pstolowski | telegram is fine | 09:18 |
| zyga | one sec, printer didn't respond | 09:19 |
| mup | PR snapd#6130 closed: snap: make Epoch's MarshalJSON not simplify <Simple š> <Created by chipaca> <Merged by chipaca> <https://github.com/snapcore/snapd/pull/6130> | 09:32 |
| zyga | pstolowski: ok, you have my full attention now | 09:32 |
| zyga | pstolowski: is the HO still used? | 09:32 |
| zyga | seems to be | 09:34 |
| pstolowski | zyga: yep | 09:34 |
| zyga | pstolowski: how about https://meet.google.com/fsz-xtpa-xqb?ijlm=1542101675152&adhoc=1&hs=125&authuser=0 | 09:34 |
| mup | PR snapd#6039 closed: snapstate: do not allow classic mode for strict snaps <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/6039> | 09:51 |
| mborzecki | yay! | 09:51 |
| popey | double yay | 09:52 |
| mvo | mborzecki: yay indeed - thank you! | 10:01 |
| mborzecki | mvo: thank you for picking it up in the first place :) | 10:02 |
| mvo | mborzecki: silly question about 6136 - the bug that prevented the openvswitch-support test to run is now fixed. but the test still fails. do we just need to blacklist it or is it something simple (like a missing package dependency)? | 10:07 |
| mborzecki | mvo: i think we may need to start uuidd.socket before the test (at least on arch) | 10:08 |
| mvo | mborzecki: aha, right - this does not happen on package install automatically | 10:13 |
| mup | PR snapd#6137 opened: snap: make Epoch default to {[0],[0]} on load from yaml <Simple š> <Created by chipaca> <https://github.com/snapcore/snapd/pull/6137> | 10:13 |
| zyga | I will need to work from the road one more time today | 10:24 |
| zyga | need to sign the paperwork and say goodbye to my car today | 10:24 |
| zyga | I will focus on reviews while on the road | 10:24 |
| mup | PR core18#88 closed: hooks: make ld-so symlink in /lib64 relative <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/core18/pull/88> | 10:30 |
| zyga | random test failure from reviews: | 10:30 |
| zyga | https://www.irccloud.com/pastebin/6mMG7HK9/ | 10:30 |
| mup | PR core18#83 closed: add swapfile support (ported from ubuntu-core-config) <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/core18/pull/83> | 10:31 |
| zyga | pedronis: question about an error I saw over weekend: | 10:32 |
| zyga | zyga@pi2-2:~$ sudo snap refresh | 10:33 |
| zyga | error: too early for operation, device not yet seeded or device model not acknowledged | 10:33 |
| zyga | pedronis: I updated the timezone, time shifted one hour forward | 10:33 |
| pedronis | that check is not time dependent afaik | 10:33 |
| zyga | pedronis: that device then went into "busted mode" | 10:33 |
| zyga | not sure what happened | 10:34 |
| pedronis | ah | 10:34 |
| zyga | I have it disconnected to inspect | 10:34 |
| zyga | perhaps some tasks didn't run | 10:34 |
| pedronis | it's really busted if it thinks is not seeded anymore | 10:34 |
| pedronis | or it lost its model | 10:34 |
| zyga | it went from "stable released on cdimage" to stable | 10:34 |
| zyga | just flash and refresh | 10:34 |
| zyga | I'll collect more data and report it | 10:34 |
| mvo | doko: hey, could you please click on "Request builds" in https://code.launchpad.net/~ubuntu-core-service/+snap/core18 and do an amd64 build for me? I don't have permissions to do this myself anymore :/ (or someone else from foundations maybe?) | 10:36 |
| Chipaca | zyga: if you're in review-mode, 6137 is nice an' easy :-) | 10:42 |
| Chipaca | it was a tiny part of #6052 but stands on its own | 10:43 |
| mup | PR #6052: snap, store, overlod/snapstate: always send epochs <Created by chipaca> <Closed by chipaca> <https://github.com/snapcore/snapd/pull/6052> | 10:44 |
| Chipaca | and will make the actually-send-epochs pr that much easier to review | 10:44 |
| mborzecki | hmm fake store does not send snap type :/ | 10:53 |
| mup | PR snapd#6124 closed: tests: simple reproducer for snap try and hooks bug <Simple š> <Created by stolowski> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/6124> | 10:54 |
| mvo | mborzecki: uh, I thought we fixed that :( | 10:54 |
| zyga | Iām in review mode but signing papers now | 11:24 |
| Chipaca | mvo: isn't there more than one fake store? | 11:25 |
| mup | PR snapd#6138 opened: overlord/ifacestate: use remapper when checking if system snap is installed <Hotplug š> <Created by stolowski> <https://github.com/snapcore/snapd/pull/6138> | 11:41 |
| pstolowski | zyga: does this make sense? ^ | 11:42 |
| Chipaca | this epoch vs *epoch things gets annoying | 11:51 |
| mup | PR snapd#6134 closed: tests: skip opensuse from interfaces-openvswitch-support test <Created by sergiocazzolato> <Merged by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6134> | 11:51 |
| mup | PR snapd#6129 closed: data/completion: pass documented arguments to completion functions <Created by chipaca> <Merged by chipaca> <https://github.com/snapcore/snapd/pull/6129> | 11:55 |
| zyga | pstolowski|afk: done | 11:56 |
| zyga | Chipaca: 6137 ready | 11:57 |
| mup | PR snapd#6137 closed: snap: make Epoch default to {[0],[0]} on load from yaml <Simple š> <Created by chipaca> <Merged by zyga> <https://github.com/snapcore/snapd/pull/6137> | 11:58 |
| mup | PR snapd#6139 opened: snap, store, overlord/snapshotstate: drop epoch pointers <Created by chipaca> <https://github.com/snapcore/snapd/pull/6139> | 12:01 |
| Chipaca | zyga: you'll like 6139 but it's not simple | 12:10 |
| Chipaca | bah. It's simple. It's also long. | 12:10 |
| Chipaca | relatively | 12:10 |
| ogra | cjwatson, i have a prob with stage-packages when cross building, i have tired three variants yet: https://paste.ubuntu.com/p/XpRF9ND5tf/ the first one gets me a build error telling me i need "dpkg --add-architecture armhf", the latter two are completely ignored, am i doing anything wrong or do i need a pecial part that calls dpkg --add-architecture first ? | 12:24 |
| ogra | *special | 12:25 |
| ogra | note this works fine with a local build | 12:25 |
| ogra | (i had the impression with the recent architecture changes on the buildds such things should be possible without extra hacks) | 12:27 |
| cjwatson | ogra: I really have no idea, I'm sorry. I would not have expected this to be at all affected by recent changes | 12:32 |
| cjwatson | All we did in LP was arrange to understand the syntax enough to dispatch builds to the right architectures | 12:33 |
| zyga | Looking | 12:33 |
| cjwatson | This is mostly on the snapcraft side AFAIK | 12:33 |
| ogra | ok, but we dont automatically add all possible arches to the sources.lists | 12:33 |
| cjwatson | We do not, at present | 12:33 |
| * Chipaca ~> lunch | 12:34 | |
| ogra | i'll try a part that call "dpkg --add-architecture armhf" then, lets see | 12:34 |
| ogra | *calls | 12:34 |
| cjwatson | Maybe also apt update, not sure | 12:34 |
| ogra | yep | 12:34 |
| ogra | (was to lazy to type that out :) ) | 12:34 |
| pstolowski | pedronis: can you take a look at https://github.com/snapcore/snapd/pull/6138 (it's trivial)? | 12:35 |
| mup | PR #6138: overlord/ifacestate: use remapper when checking if system snap is installed <Hotplug š> <Created by stolowski> <https://github.com/snapcore/snapd/pull/6138> | 12:35 |
| pedronis | pstolowski: in a bit, about to go to my dentist appt | 12:36 |
| pstolowski | sure | 12:36 |
| zyga | Chipaca: reviewed | 12:40 |
| Chipaca | who's best to pick up a custom store contact? | 12:40 |
| zyga | Dunno | 12:41 |
| Chipaca | JamieBennett: ^? | 12:41 |
| Chipaca | zyga: thanks | 12:41 |
| JamieBennett | Chipaca: feel free to pass them on to me | 12:41 |
| Chipaca | JamieBennett: @'ed you on the forum | 12:43 |
| Chipaca | JamieBennett: thanks | 12:43 |
| ogra | cjwatson, FYI ... seemss to have worked | 12:43 |
| Chipaca | now yes, lunch | 12:43 |
| ogra | ARGH ! | 12:44 |
| ogra | Store upload failed: | 12:44 |
| ogra | ('Connection aborted.', gaierror(-3, 'Temporary failure in name resolution')) | 12:44 |
| ogra | i hate when that happens !! | 12:44 |
| cjwatson | You and me both | 12:45 |
| cjwatson | I've asked for the haywire worker to be killed | 12:47 |
| mup | PR snapd#6140 opened: tests, fakestore: extend refresh tests with parallel installed snaps <Parallel installs ā> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6140> | 12:54 |
| zyga | mborzecki: will you have time to look at https://github.com/snapcore/snapd/pull/6111 | 12:57 |
| mup | PR #6111: packaging/opensuse: move most logic to snapd.mk <Created by zyga> <https://github.com/snapcore/snapd/pull/6111> | 12:57 |
| zyga | Nothing urgent | 12:57 |
| mborzecki | zyga: i'm going out to pick up the kids in a bit, i can do that later or during the standup | 12:57 |
| zyga | Sure | 12:58 |
| zyga | Looking at https://github.com/snapcore/snapd/pull/6104 now | 12:58 |
| mup | PR #6104: snapctl: add "services" <Created by chipaca> <https://github.com/snapcore/snapd/pull/6104> | 12:58 |
| Chipaca | zyga: I need to tweak that pr | 12:59 |
| Chipaca | also i need to actually go do something about lunch | 13:00 |
| Chipaca | drat | 13:00 |
| * Chipaca rips himself away | 13:00 | |
| mup | PR snapd#6136 closed: tests: run interfaces-openvswitch-support on arch again, the arch bug is fixed <Created by mvo5> <Closed by mvo5> <https://github.com/snapcore/snapd/pull/6136> | 13:24 |
| cachio | plars, cwayne, hey, could you please confirm if 2.36.1 is ready for candidate? | 13:41 |
| cachio | thanks | 13:41 |
| zyga | Chipaca: sure, let me know | 13:56 |
| mborzecki | Chipaca: standup | 14:01 |
| popey | cjwatson: got a developer reporting the name resolution error on build.snapcraft.io | 14:20 |
| popey | apolitech is the developer account | 14:21 |
| popey | hello APolihron :) | 14:26 |
| cjwatson | popey: how long ago? | 14:28 |
| mup | PR snapcraft#2403 opened: project_loader: add git to build-packages for version: git <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2403> | 14:28 |
| popey | APolihron: ^ When did it happen? | 14:28 |
| cjwatson | never mind | 14:28 |
| pedronis | Chipaca: did you see my pm with the link? | 14:39 |
| APolihron_ | hey all! so i've change the yalm file from my github projects to build only for a amd64 , bump up the release number and the grade to stable but i get build, fail to release when using snapcraft.io build serv | 14:40 |
| APolihron_ | this is one of the project that will not release | 14:40 |
| APolihron_ | https://github.com/apolitech/Udemy/blob/master/snap/snapcraft.yaml | 14:40 |
| APolihron_ | and this is the error Error:('Connection aborted.', gaierror(-3, 'Temporary failure in name resolution')) | 14:41 |
| cjwatson | APolihron_: This is https://bugs.launchpad.net/launchpad/+bug/1792920 - I've asked for the worker to be killed | 14:42 |
| mup | Bug #1792920: celery workers sometimes end up cursed and produce OOPSes for all SnapStoreUploadJobs <oops> <Launchpad itself:Triaged> <https://launchpad.net/bugs/1792920> | 14:42 |
| APolihron_ | so how can i help or what i need to do? | 14:50 |
| cjwatson | APolihron_: wait until I tell you to try again, then try again | 14:51 |
| cjwatson | (I'm waiting for a sysadmin to get back to me) | 14:52 |
| APolihron_ | ok | 14:54 |
| APolihron_ | thanks | 14:54 |
| cjwatson | APolihron_: try again now | 15:04 |
| mborzecki | heh, our file.sh helpers are not eally working well with char devices, the checks all use `if [ -f ... ]` which is onviously not true for paths that are not regular files | 15:11 |
| mup | PR snapd#6141 opened: interfaces: export gpio pin in AppArmorConnectedPlug <Created by mvo5> <https://github.com/snapcore/snapd/pull/6141> | 15:18 |
| pstolowski | mborzecki: oh wow, nice finding! | 15:18 |
| mvo | pedronis: sorry to put more burden on your shoulders - I pushed 6141 which contains an alternative fix for the gpio problem. your input on this would be most welcome | 15:19 |
| mvo | pedronis: it sidesteps the problem entirely by avoiding the dependency between appamror and systemd on the expense of some extra code in the gpio interface | 15:19 |
| pedronis | mvo: so we understand we are both exporting it permanently (across reboots) via systemd units, and explicit in the app armor backend? | 15:24 |
| pedronis | s/we/I/ | 15:24 |
| mvo | pedronis: correct | 15:28 |
| mvo | pedronis: this way we avoid the dependency that the systemd backend needs to start the unit before the apparmor backend can stat the symlink | 15:29 |
| pedronis | mvo: ok, I left a question there | 15:29 |
| APolihron_ | cjwatson working, thanks | 15:30 |
| mvo | pedronis: thanks, I will look at this and reply inline there is a tocttou race here I think, both places check. otoh we don't run this is parallel yet so it should be fine | 15:30 |
| pedronis | mvo: as I said we let go of the lock for backends, so whether is run in parallel or not is a bit unclear | 15:31 |
| mvo | pedronis: so the race would be if two tasks both want to export gpio30 ? | 15:32 |
| pedronis | or we might have the unit | 15:32 |
| pedronis | and systemd is doing strange things | 15:32 |
| pedronis | or the user | 15:32 |
| pedronis | mvo: it seems at least that if you fail to write, you need to check if by chance the link is there | 15:33 |
| pedronis | if it is ignore the error | 15:33 |
| pedronis | that's my 2cts | 15:33 |
| mvo | pedronis: indeed, that is a good suggestion | 15:33 |
| mvo | pedronis: we have the same problem with the unit, there is also a test || write in there which might race | 15:33 |
| mvo | pedronis: I look into it, thanks! | 15:33 |
| pedronis | np | 15:34 |
| cachio | niemeyer, hey, I need to save the images we use for beta/edge validation | 15:34 |
| * pedronis break | 15:34 | |
| mvo | pedronis: on a meta level (not urgent) we need some suggestion which of the approaches we should take to fix the actual bug. zyga and I are not of the same opinion about what approach is the one to pick :) | 15:34 |
| mvo | pedronis: enjoy your break, not urgent | 15:34 |
| cachio | niemeyer, do you think we can store them in the gcloud filestorage? | 15:34 |
| cachio | niemeyer, we are not gonna use more than 5GB | 15:35 |
| micah | hi, i need to symlink /snap to /rw/snap on my system, but if I do that, then the core mount job fails because of: snap-core-5742.mount: Mount path /snap/core/5742 is not canonical (contains a symlink) -- is there some way I can override that? | 15:41 |
| * cachio lunch | 15:41 | |
| micah | starting with systemd 238, mount units will fail if the mount point needs to follow a symbolic link | 15:45 |
| ppisati | ogra: for the wifi issue you were mentioning some time ago, could you test this kernel snap? -> https://code.launchpad.net/~p-pisati/+snap/raspi2-test/+build/377679/+files/pi2-kernel_4.4.0-1101.109_armhf.snap | 15:49 |
| ogra | ppisati, will do | 15:50 |
| pedronis | mvo: what are main zyga objections about? the duplication, the fact that is new unproven code, the fact that is on the plug side? | 15:56 |
| zyga | That it feels like layering violation. No other interface code modified the system | 16:12 |
| pedronis | zyga: modified in which sense? | 16:12 |
| zyga | That it feels like layering violation. No other interface code modified the system | 16:13 |
| pedronis | ? | 16:13 |
| pedronis | zyga: modify in which sense? they all touch files on the system? and for sure the systemd code is creating a unit that modifies the system? | 16:14 |
| zyga | Writing to sysfs | 16:14 |
| zyga | Exporting GPIO from the kernel to userspace in a method that computes apparmor snippet | 16:15 |
| pedronis | ok, that's clearer | 16:15 |
| zyga | Systems unit is just defined | 16:16 |
| zyga | Not on disk | 16:16 |
| zyga | All changes are done by backend | 16:16 |
| mup | PR snapd#6139 closed: snap, store, overlord/snapshotstate: drop epoch pointers <Created by chipaca> <Merged by chipaca> <https://github.com/snapcore/snapd/pull/6139> | 16:16 |
| zyga | Systemd specification holds units in memory | 16:17 |
| pedronis | zyga: I understood | 16:17 |
| mup | PR snapd#6142 opened: overlord/snapstate, store: always send epochs <Created by chipaca> <https://github.com/snapcore/snapd/pull/6142> | 16:17 |
| pedronis | zyga: mvo: at the moment all considered #6128 seems my favorite fix, but it depends whether we think we need this also for 2.36.x | 16:34 |
| mup | PR #6128: overlord/ifacestate: setup security backends phased by backends first <Created by mvo5> <https://github.com/snapcore/snapd/pull/6128> | 16:35 |
| zyga | pedronis: that's fine though this implementation has slightly bigger impact than the first one I pushed (with the special casing of systemd backend) | 16:38 |
| zyga | pedronis: after a QA pass I don't mind landing that at all | 16:38 |
| pedronis | zyga: that's why I said it depends whether we need something to backport | 16:40 |
| pedronis | zyga: anyway I think mvo is offline changing location | 16:40 |
| pedronis | atm | 16:40 |
| zyga | ack | 16:40 |
| zyga | I'll get onto other things | 16:40 |
| zyga | I'm happy to discuss this when needed | 16:41 |
| pedronis | zyga: what are you working on atm? back to user namespaces? | 16:48 |
| zyga | yes | 16:48 |
| pedronis | ok, thanks | 16:48 |
| zyga | I will do some reviews of pawel's work as well | 16:48 |
| pstolowski | zyga: ty! | 16:52 |
| === chihchun is now known as chihchun_afk | ||
| === pstolowski is now known as pstolowski|afk | ||
| Chipaca | error: cannot install snap file: cannot refresh snap "basic" as new epoch (2) can't read old epoch (0) | 17:39 |
| Chipaca | looks like it might work | 17:39 |
| roadmr | epochs ftwwwww | 17:40 |
| Chipaca | that's from a local install | 17:41 |
| pedronis | roadmr: yes, Chipaca is progressing on them | 17:42 |
| roadmr | yya :) | 17:42 |
| zyga | Chipaca: _neat_ | 17:43 |
| zyga | snaps get serious š | 17:43 |
| * Chipaca mostly EODs | 17:54 | |
| Chipaca | a little early because the chorizo + split lentil stew smell is distracting me too much | 17:57 |
| === zarcade_droid is now known as Guest59711 | ||
| zyga | o/ | 18:49 |
| * cachio afk | 18:59 | |
| mvo | zyga: you had a chat about gpio? can you point me to a pastebin or give me a tl;dr summary? | 19:33 |
| zyga | with Samuele? | 19:34 |
| mvo | zyga: yes | 19:34 |
| mvo | zyga: i.e. should I stop messing around with 6141 ? | 19:34 |
| zyga | I only said what my objections are | 19:34 |
| pedronis | I wrote this: <pedronis> zyga: mvo: at the moment all considered #6128 seems my favorite fix, but it depends whether we think we need this also for 2.36.x | 19:35 |
| pedronis | <mup> PR #6128: overlord/ifacestate: setup security backends phased by backends first <Created by mvo5> <https://github.com/snapcore/snapd/pull/6128> | 19:35 |
| pedronis | <zyga> pedronis: that's fine though this implementation has slightly bigger impact than the first one I pushed (with the special casing of systemd backend) | 19:35 |
| pedronis | <zyga> pedronis: after a QA pass I don't mind landing that at all | 19:35 |
| mup | PR #6128: overlord/ifacestate: setup security backends phased by backends first <Created by mvo5> <https://github.com/snapcore/snapd/pull/6128> | 19:35 |
| pedronis | <pedronis> zyga: that's why I said it depends whether we need something to backport | 19:35 |
| pedronis | <pedronis> zyga: anyway I think mvo is offline changing location | 19:35 |
| zyga | we did not discuss more than that | 19:35 |
| pedronis | we can chat more in the morning, but we do need to try to converge, and also decide about 2.36.x | 19:35 |
| zyga | I agree | 19:36 |
| mvo | zyga: ok, thanks | 19:38 |
| mvo | pedronis: thank you as well | 19:39 |
| mvo | it looks like they all need reviews anyway, so lets sync up in the morning | 19:39 |
| mvo | ackk: core18 with the ld-so symlink fix is in edge now, please let me know if you it is not sufficient | 19:54 |
| === zarcade_droid is now known as arcade_droid | ||
| jdstrand | roadmr: hi! whenever its convenient, can you pull 1155 of the review-tools? | 21:01 |
| mup | PR #1155: integration-tests: remove the integration daemon tests <Created by elopio> <Merged by niemeyer> <https://github.com/snapcore/snapd/pull/1155> | 21:01 |
| roadmr | jdstrand: sure thing! | 21:01 |
| jdstrand | roadmr: thanks | 21:02 |
| Caelum | zyga: thank you so much for finally doing the snapd update, to be honest I've been using flatpak spotify in the meantime | 21:03 |
| Caelum | zyga: if you ever want help with anything, just ping me | 21:05 |
| mup | PR snapcraft#2404 opened: cli: handle yaml errors for cleanbuild <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2404> | 21:47 |
| mup | PR snapd#6143 opened: cmd: install snap-discard-ns in "make hack" <Simple š> <Created by zyga> <https://github.com/snapcore/snapd/pull/6143> | 23:25 |
| mup | PR snapd#6144 opened: cmd: handle tumbleweed and leap in autogen.sh <Simple š> <Created by zyga> <https://github.com/snapcore/snapd/pull/6144> | 23:28 |
| ijohnson | zyga: sorry I forgot but I just filed https://bugs.launchpad.net/snapd/+bug/1803210 for you to look at | 23:29 |
| mup | PR snapd#6145 opened: cmd/snap-confine: capture initialized per-user mount ns <Created by zyga> <https://github.com/snapcore/snapd/pull/6145> | 23:29 |
| mup | Bug #1803210: snap's device cgroup is not discarded upon uninstall <snapd:New> <https://launchpad.net/bugs/1803210> | 23:30 |
| ijohnson | no rush it's a low priority | 23:30 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!