2021/01/19/#snappy.txt

[07:08] <mborzecki> morning
[08:10] <pstolowski> morning
[08:11] <mvo> good morning pstolowski
[08:14] <mborzecki> pstolowski: mvo: morning guys
[08:14] <mvo> good morning mborzecki
[08:18] <pstolowski> mvo: hey, i saw you approved https://github.com/snapcore/snapd/pull/9845 but there is a problem with cla check there and i don't know how and if we can solve it
[08:18] <mup> PR #9845: osutil: update go-udev package <Created by stolowski> <https://github.com/snapcore/snapd/pull/9845>
[08:19] <mvo> pstolowski: well, in an ideal world the cla checker would skip git submodules
[08:19] <mvo> pstolowski: it's not quite clear to me why it is not already doing that but I don't know enough about subtrees
[08:19] <mvo> pstolowski: so if we can fix it somehow that would be good otherwise we need to force-merge I think
[08:20] <pstolowski> mvo: right, i see. i will take a look, not familiar with details either
[08:20] <mup> PR snapd#9843 closed: packaging: fix arch-indep build on debian-sid <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/9843>
[09:03] <zyga> good morning
[09:05] <pstolowski> gpg error again... i think it's has been happening more often recently
[09:05] <pstolowski> hey zyga
[09:06] <zyga> pstolowski, yeah, Ian said he can reporduce that on his laptop
[09:06] <zyga> just by running the test in a loop
[09:06] <zyga> but was unsure how to debug further
[09:06] <pstolowski> yeah i know he has been looking into it
[09:09] <pstolowski> mvo: nb, on git submodules... it's not really submodule in this case, it's subtree, they are different beasts. subtree literally imports another project as a subtree
[09:34] <mvo> pstolowski: hm, ok
[09:35] <mvo> hm, I get LP build errors right now:
[09:35] <mvo> === RUN   TestSnapManager
[09:35] <mvo> fatal error: sync: unlock of unlocked mutex
[09:36] <pstolowski> mvo: no other details?
[09:36] <mvo> pstolowski: lots of details :) https://launchpadlibrarian.net/518145718/buildlog_ubuntu-xenial-i386.snapd_2.48.2+git2653.2672657e~ubuntu16.04.1_BUILDING.txt.gz
[09:36] <pedronis> mvo: that error is probably hinding something else
[09:36] <mvo> pstolowski: strange enough I cannot reproduce on my laptop with master so might be timing
[09:36] <mvo> pedronis: yeah :(
[09:44] <pstolowski> mvo: afaiu assert on s.settle(c) fails so we don't restore the lock with s.state.Lock() and 'defer s.state.Unlock()' kicks in on failure (sorry if i'm stating the obvious)
[09:45] <pstolowski> and settle uses testutil.HostScaledTimeout(5 * time.Second) so maybe it is indeed timing?
[09:46] <mvo> pstolowski: I wonder if we can tweak the code so that we still get the real error. and yeah, timing seems likely but what was unusual is that I got this error from many different LP arches
[09:46] <mvo> pstolowski: but given that I don't get it here locally I strongly suspect timing
[09:47] <mvo> pstolowski: I wish there was a systemd-run --use-cpu-cgroup-to-make-things-really-slow :)
[09:47] <mvo> pstolowski: to run the tests under
[09:48] <zyga> mvo, maybe there is :)
[09:48] <mvo> zyga: I bet there is, I'm just not finding it :)
[09:51] <zyga> mvo, look at systemd.resource-control and look for CPUShares
[09:52] <zyga> there's a lot of options there, like CPUQuota
[09:52] <zyga> you can pass those options to systemd run as well (as properties)
[09:53] <zyga> I'd try with CPUQuota=1%
[09:53] <zyga> and see what happens
[09:55] <pedronis> mvo: pstolowski: we should always pair Lock with defer Unlock etc even in tests, but historically we haven't done that because is a bit annoying/hard to remember
[09:55] <zyga> hey pedronis, how are you :)
[10:28] <pstolowski> oh, duplicated snapshots import logic has a bug, returns empty list of snap names for duplicated import. i think the fix will be slightly annoying
[10:28] <pstolowski> unit tests weren't good enough to catch this
[10:28] <pedronis> zyga: hi, it's mid-cycle week
[10:28] <zyga> pstolowski, oh, super busy times
[10:29] <zyga> er
[10:29]  * zyga is bad at tab completion
[10:29] <zyga> pedronis, good luck with that
[10:39] <pedronis> thx
[11:26] <pstolowski> bbiab
[13:18] <mup> PR snapd#9851 opened: o/snapshotstate: fix returning of snap names when duplicated snapshot is detected <Bug> <Created by stolowski> <https://github.com/snapcore/snapd/pull/9851>
[14:02] <mup> PR snapcraft#3421 opened: cli: remove experimental notice for compression <Created by cjp256> <https://github.com/snapcore/snapcraft/pull/3421>
[15:52] <mup> PR snapcraft#3419 closed: Include SDK snap paths in ACLOCAL_PATH <Created by kenvandine> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3419>
[15:52] <mup> PR snapcraft#3420 closed: plainbox spread tests: set tasks to manual <Created by cjp256> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3420>
[16:02] <mup> PR snapcraft#3422 opened: Revert "cli: allow validation assertions to be revoked (#3417)" <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/3422>
[16:17] <pstolowski> mvo: i cannot  find anything magical to help with cla-check in #9845, i suggest we force-merge it, updates to go-udev are extremely rare anyway
[16:17] <mup> PR #9845: osutil: update go-udev package <Created by stolowski> <https://github.com/snapcore/snapd/pull/9845>
[16:17] <mvo> pstolowski: thanks, I will force merge
[16:19] <pstolowski> ty
[16:24] <mup> PR snapd#9845 closed: osutil: update go-udev package <Created by stolowski> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/9845>
=== azubieta5 is now known as azubieta
[19:43] <mup> PR snapcraft#3422 closed: Revert "cli: allow validation assertions to be revoked (#3417)" <Created by sergiusens> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3422>
[20:38] <mup> PR snapcraft#3423 opened: WIP: gnome-3-38 is no longer experimental <Created by kenvandine> <https://github.com/snapcore/snapcraft/pull/3423>
=== Aavar_ is now known as Aavar
[22:03] <mup> PR snapcraft#3424 opened: Ensure PYTHONPATH is appropriate for building packages with gnome-3-38 <Created by kenvandine> <https://github.com/snapcore/snapcraft/pull/3424>