ijohnson | zyga: when you get in can you take a look at https://bugs.launchpad.net/snapd/+bug/1850720? thanks! | 03:07 |
---|---|---|
mup | Bug #1850720: installing snap with layout on /etc/ld.so.cache results in deleted mount <snapd:New> <https://launchpad.net/bugs/1850720> | 03:07 |
mborzecki | morning | 06:28 |
zyga | ijohnson: done | 07:01 |
zyga | hey mborzecki :) | 07:01 |
mborzecki | zyga: hey hey | 07:01 |
zyga | mborzecki: I think I restored my sleep budget to normal state | 07:01 |
mborzecki | a lot of red and a funch of stuck travis jobs? | 07:02 |
zyga | mborzecki: must be Thursday ;) | 07:02 |
zyga | mborzecki: I'll review the other PR today | 07:08 |
mborzecki | pedronis: mvo: hey | 07:52 |
mborzecki | mvo: fixed the typo in #7701 but the unit tests fail, want me to look into that? | 07:52 |
mup | PR #7701: overlord: add kernel rollback accross reboots manager test and fixes <Created by mvo5> <https://github.com/snapcore/snapd/pull/7701> | 07:52 |
mvo | hey mborzecki | 07:53 |
mvo | mborzecki: oh, hm, hm | 07:53 |
mvo | mborzecki: you are welcome to look | 07:53 |
mborzecki | mvo: this is what's failing there: https://paste.ubuntu.com/p/Hdzh8kbqF9/ | 07:53 |
mvo | mborzecki: oh, fun | 07:54 |
mvo | mborzecki: /o\ yeah, I worked on the manager tests | 07:54 |
mvo | mborzecki: but did not run the whole thing | 07:54 |
mborzecki | mvo: haven't done much around reboots, but i can dive in and check what's happening there | 07:55 |
mvo | mborzecki: is the parallel install and gadget all blocked? if so, yes, a look at this would be great | 07:56 |
mborzecki | mvo: parallel installs is waiting for a review from zyga, jdstrand already +1'ed it | 07:57 |
zyga | mborzecki: yep, on my plate today | 07:57 |
zyga | mborzecki: I think it will land today | 07:57 |
mborzecki | mvo: and gadget remodel needs reviews ;) you're welcome to do a pass there | 07:58 |
mborzecki | zyga: yeah, had to restart the tests | 07:58 |
mvo | mborzecki: aha, nice. I will do have a look | 07:58 |
mvo | mborzecki: cool, more eyes on the boot code are always good | 07:58 |
mborzecki | and wanted to take a look into #7702, see hwther i can move some stuff over from #7193 | 07:58 |
mvo | mborzecki: I suspect its missing fixtures now | 07:58 |
mup | PR #7702: tests: adding fedora 31 to google-unstable backend <Created by sergiocazzolato> <https://github.com/snapcore/snapd/pull/7702> | 07:58 |
mup | PR #7193: [WIP] many: cgroupsv2 spread run <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/7193> | 07:58 |
=== pstolowski|afk is now known as pstolowski | ||
pstolowski | morning | 08:00 |
mvo | mborzecki: yeah, it looks like the mocks are now incomplete in the booted_test. I suspect its just a matter of adding Type: snap.TypeOS etc | 08:12 |
zyga | wow, Tyler got mentioned on https://software.intel.com/security-software-guidance/insights/deep-dive-intel-analysis-speculative-behavior-swapgs-and-segment-registers | 08:15 |
mborzecki | btw https://medium.com/nttlabs/cgroup-v2-596d035be4d7 | 08:17 |
mvo | mborzecki: yeah, its just a missing Type: snap.TypeBase for the snapsetup fixture | 08:18 |
mvo | mborzecki: I push the (trivial) fix | 08:18 |
mborzecki | mvo: ah ok :) | 08:18 |
mvo | mborzecki: thank you still! | 08:19 |
zyga | mborzecki: lol | 08:19 |
zyga | runc, the reference implementation of OCI Runtime Spec, gained the initial support for cgroup v2 just last month (PR: #2113). This is not ready for production, especially because it lacks the implementation for eBPF device controller | 08:19 |
mup | PR #2113: interfaces/builtin: add i2c interface <Created by bergotorino> <Merged by chipaca> <https://github.com/snapcore/snapd/pull/2113> | 08:19 |
zyga | mborzecki: at least they have a PR open | 08:19 |
mborzecki | zyga: yeah, heh | 08:19 |
zyga | anyway, I think we are "state of the art" in the sense that the state is shit anyway | 08:19 |
mborzecki | zyga: feels like switching the defaults in systemd is a bit premature, but it's fine that fedora did it, otherwise nobody would care to transition | 08:20 |
zyga | yeah | 08:20 |
zyga | I think it's just like that | 08:20 |
zyga | you have to toggle the experimental feature | 08:21 |
zyga | or it will never really work | 08:21 |
zyga | free systems are the beta testers | 08:21 |
zyga | (in fedora world) | 08:21 |
mborzecki | haha, that sounds familiar tbh :) | 08:21 |
zyga | TBH I find the whole runc/crun/podman/docker/moby stuff a terrible mess | 08:22 |
zyga | as in, all the $$$ behind this wants to land grab | 08:22 |
zyga | probably inevitable but feels like mud with leeches | 08:22 |
mborzecki | zyga: you left out skopeo | 08:24 |
zyga | I'm sure I did | 08:24 |
zyga | it's a zoo of new things that do part of what docker did | 08:24 |
zyga | all with quirky geeky names | 08:24 |
zyga | oh well | 08:24 |
mborzecki | xD | 08:24 |
zyga | brb, cold today, need tea | 08:25 |
=== pedronis_ is now known as pedronis | ||
pedronis | mvo: hi, I made a pass on 7701 (covering also the one is based on I think) | 08:30 |
pedronis | s/made/did/ | 08:30 |
zyga | hey pedronis :) | 08:30 |
* zyga is back with hot tea | 08:30 | |
pedronis | mvo: maybe close 7649 and keep only 7701? I find reasoning about both situations together easier | 08:34 |
pedronis | 7649 is very small | 08:34 |
mvo | pedronis: sure, happy to do that | 08:37 |
mvo | pedronis: thank you! will look at the feedback now | 08:38 |
mvo | fwiw 7651 needs a second review, it will unblock 7652 | 08:38 |
mup | PR snapd#7649 closed: overlord: fix TestRemodelSwitchToDifferentKernel for bootvars <Remodel ๐> <Created by mvo5> <Closed by mvo5> <https://github.com/snapcore/snapd/pull/7649> | 08:40 |
pedronis | mborzecki: did another pass on #7665, looking good | 09:02 |
mup | PR #7665: devicestate: add support for gadget->gadget remodel <Remodel ๐> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/7665> | 09:02 |
mborzecki | pedronis: thanks! | 09:02 |
pedronis | mvo: I looked what that was, but it's my slots-per-plug stuff, I cannot do the 2nd review :) | 09:05 |
pedronis | mvo: thx for the review | 09:05 |
mvo | pedronis: haha, yes | 09:07 |
mvo | pedronis: would love to unblock this today | 09:07 |
mvo | pedronis: I updated the PR based on your feedback (7701) | 09:27 |
pedronis | mvo: thx, will look in a bit | 09:28 |
mvo | Chipaca: I updated 7624 a bit more based on your suggestions, would love to add some more tests, a second look would be great (just to doulbe check) and then your opinion if I should merge and or do some PRs with test improvements first | 09:55 |
Chipaca | mvo: omw | 09:55 |
mvo | Chipaca: no rush | 09:56 |
mvo | mborzecki: you got some feedback on the remodel PR, more to come :) | 10:01 |
mborzecki | mvo: ah ok, i'll wait with the fixes then | 10:01 |
mvo | mborzecki: do the fixes, thats fine, I will not be able to look for at least 1h | 10:02 |
* zyga -> walk | 10:03 | |
Chipaca | mvo: any particular reason you went with syscall.SIGINT instead of os.Interrupt? | 10:05 |
mvo | Chipaca: no reason, we use that elsewhere too | 10:05 |
Chipaca | fair | 10:06 |
Chipaca | drat, there's a chunk of functionality we should abstract somewhere neat | 10:07 |
Chipaca | :) | 10:07 |
Chipaca | mvo: looks good! i'll be building and testing it locally in a bit to see if i catch any other issues | 10:11 |
mvo | Chipaca: this is why I said I'm not super happy yet :/ it feels ok but not great (yet) | 10:15 |
Chipaca | mvo: maybe it's all that 'snap donwload' | 10:16 |
Chipaca | siiigh | 10:18 |
Chipaca | mvo: interrupting the download prints a spurious error message, at least in part because Download() isn't receiving a context form the command so it doesn't know it's been canceled | 10:20 |
Chipaca | mvo: otoh the command could know it's canceled and ignore the error at that point | 10:20 |
Chipaca | mvo: otoยฒh it's third-order nits at this point | 10:20 |
Chipaca | mvo: otoยณh it's the sort of polish we like | 10:20 |
Chipaca | and they're ux regressions | 10:21 |
Chipaca | the --direct version doesn't suffer from this | 10:21 |
Chipaca | :-/ | 10:22 |
Chipaca | mvo: so | 10:22 |
Chipaca | mvo: how about this | 10:22 |
Chipaca | mvo: switch the defaults, land as is, work on the nits until they're where we want them, _then_ switch the defaults | 10:22 |
mvo | Chipaca: nice one! | 10:22 |
Chipaca | mvo: maybe even make the options hidden so we can play with them without breaking stuff | 10:22 |
Chipaca | so don't add --direct, but add a hidden --indirect or something | 10:24 |
Chipaca | mvo: it feels like it's a couple of PRs of refactorings to get the UX where we want it | 10:24 |
Chipaca | that's my gut feeling about it, but my gut is full of shit | 10:25 |
Chipaca | wait | 10:25 |
Chipaca | ยฏ\_(ใ)_/ยฏ | 10:25 |
Chipaca | mvo: disadvantage is that, because people are needing this functionality, they'll start using it even if it's hidden so we'd have to carry it forever | 10:26 |
Chipaca | mvo: but OTOH having a hidden --indirect forever doesn't seem too onerous | 10:26 |
mup | Bug #1574487 changed: possibly unclean shutdown <Snappy:Fix Released> <initramfs-tools-ubuntu-core (Ubuntu):Fix Committed> <https://launchpad.net/bugs/1574487> | 10:30 |
mup | Bug #1606574 changed: SSH Interface is missing <snapd-interface> <Snappy:Fix Released> <https://launchpad.net/bugs/1606574> | 10:31 |
zyga | Unexpected errand, 1.5 hrs | 10:33 |
mup | Bug # changed: 1593450, 1613971, 1624829, 1637611 | 10:34 |
mup | Bug #1641631 changed: Raspberry Pi images do not support boot from USB <Snappy:Fix Released by alfonsosanchezbeato> <https://launchpad.net/bugs/1641631> | 10:37 |
mup | Bug #1642082 changed: Timestamp error when we try to sign a model assertion <snappy-docs> <Snappy:Fix Released by davidc3> <https://launchpad.net/bugs/1642082> | 10:37 |
mup | Bug #1646144 changed: ACLs to devices need to be supported in core <Canonical System Image:Confirmed for pat-mcgowan> <Snappy:Fix Released by ogra> <ubuntu-core-meta (Ubuntu):Fix Committed by ogra> <https://launchpad.net/bugs/1646144> | 10:40 |
mup | Bug #1646333 changed: bind mounts related to content interface plugs remain stale on snap disconnect/connect or snap updates <Snappy:Fix Released> <https://launchpad.net/bugs/1646333> | 10:40 |
mup | Bug # changed: 1647169, 1654588, 1655711, 1656820, 1657751, 1659149 | 10:43 |
mup | Bug # changed: 1659724, 1659744, 1660865, 1673757, 1675054 | 10:46 |
Chipaca | somebody's having fun on their triage day | 10:47 |
mup | Bug # changed: 1663177, 1671266, 1676244, 1680088, 1683368, 1705486, 1737427, 1743301 | 10:49 |
mup | Bug #1747794 changed: cannot resolve host name with avahi interface <Snappy:Fix Released> <https://launchpad.net/bugs/1747794> | 10:52 |
mup | Bug #1758849 changed: Snap not able to enable ssh after core upgrade <Snappy:Fix Released> <https://launchpad.net/bugs/1758849> | 10:52 |
mborzecki | sorry for the spam | 11:01 |
zyga | Errand almost done, grabbing some food now | 11:05 |
zyga | mvo: Iโll make it up tomorrow | 11:05 |
zyga | mborzecki: what is Ubuntu-core-meta? | 11:05 |
Chipaca | mborzecki: don't be sorry! good stuff | 11:06 |
mborzecki | zyga: no clue, better ask ogra | 11:10 |
Chipaca | zyga: it's a meta package | 11:12 |
zyga | I wonder if we could close some weird projects | 11:12 |
zyga | Or disable them from bug reports at least | 11:12 |
Chipaca | zyga: in particular it has nothing to do with us | 11:14 |
Chipaca | ubuntu-core-meta is the source of ubuntu-core-libs | 11:14 |
Chipaca | zyga: where are we getting dragged into it? | 11:15 |
Chipaca | hmmmm | 11:15 |
* Chipaca reads what he wrote | 11:15 | |
Chipaca | ok, in the above, change 'us' or 'we' to 'snappy' | 11:15 |
* Chipaca needs to work on that some more | 11:15 | |
Chipaca | ah, i just saw the bug | 11:16 |
Chipaca | so snappy isn't involved at the project level, just that bug that affected it | 11:17 |
Chipaca | ok | 11:17 |
mup | PR snapd#7699 closed: release: 2.42.1 <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/7699> | 11:31 |
mup | PR snapd#7704 opened: snap: extract printInstallHint in cmd_download.go <Simple ๐> <Created by mvo5> <https://github.com/snapcore/snapd/pull/7704> | 11:55 |
mup | PR snapd#7705 opened: o/devicestate: Handle preseed in firstboot [WIP] <Prebaking> <Created by stolowski> <https://github.com/snapcore/snapd/pull/7705> | 11:58 |
pstolowski | pedronis: hey, i marked this ^ WIP because i had to push changes to parts that i intend to propose separately (and that miss tests right now), but i was impossible to avoid these bits | 12:00 |
zyga | Chipaca: :D | 12:01 |
* zyga is back from errand and lunch | 12:01 | |
zyga | and back to coding | 12:01 |
zyga | mborzecki: btw, about those locks, we create locks 0600 | 12:08 |
zyga | so no catastrophy :) | 12:08 |
Chipaca | zyga: https://i.imgur.com/HCe563J.jpg | 12:11 |
* Chipaca goes for lunch | 12:11 | |
zyga | hmm | 12:11 |
zyga | Chipaca: meouch! | 12:11 |
mup | PR snapd#7193 closed: [WIP] many: cgroupsv2 spread run <Created by bboozzoo> <Closed by bboozzoo> <https://github.com/snapcore/snapd/pull/7193> | 12:12 |
pedronis | Chipaca: thanks for the review, I tried to answer your questions, maybe you have further feedback on the first and last | 12:20 |
Chipaca | pedronis: the .More() was to error if there was garbage after the json, fwiw | 12:20 |
pedronis | Chipaca: we could, it's a bit more interesting if it was a network connection | 12:21 |
Chipaca | joeubuntu: wait so you and joedborg _aren't_ the same person? | 12:21 |
Chipaca | :) | 12:21 |
Chipaca | true | 12:21 |
joedborg | Not that I know of ๐ | 12:22 |
Chipaca | joeubuntu: i learned of joeubuntu leading the robotics team at the same time i saw your nick | 12:22 |
joeubuntu | In the singularity all joes are unified in to one... | 12:23 |
pedronis | Chipaca: would it be better if that label was inside an Options struct ? | 12:29 |
pedronis | (I would do that change in one of the follow ups though) | 12:29 |
Chipaca | pedronis: not until we know what we're doing :) | 12:30 |
pedronis | ok | 12:31 |
Chipaca | pedronis: found the bug in the completion pr (i'm having lunch so figured it was a good time to poke at it) | 12:32 |
pedronis | ah, good | 12:32 |
pedronis | Chipaca: the fun with labels happens here: https://github.com/snapcore/snapd/pull/7658/files#diff-eb9825aa18d9bbbcc41ca31728af8157R75 in the later PR | 12:33 |
mup | PR #7658: cmd/snap-preseed: add snap-preseed executable <Prebaking> <Created by stolowski> <https://github.com/snapcore/snapd/pull/7658> | 12:33 |
pedronis | there's just a TODO for now there, it relates to your work (modes and modeenv etc) | 12:33 |
Chipaca | pedronis: i think that's the link to the nil db pr, no the labels one :) | 12:37 |
pedronis | Chipaca: oops | 12:39 |
pedronis | https://github.com/snapcore/snapd/pull/7695/files#diff-0b649265c21137e8fd367d4a16607a82R409 | 12:40 |
mup | PR #7695: o/devicestate: the basics of Core 20 firstboot support with test <Created by pedronis> <https://github.com/snapcore/snapd/pull/7695> | 12:40 |
pedronis | Chipaca: ^ | 12:40 |
Chipaca | pedronis: ta | 12:42 |
Chipaca | i'll get to that eventually (hoping for today!) | 12:42 |
pedronis | Chipaca: any I have a chain with another 2 or 3 PRs on top of the ones there, so fill free to do a rename to OpenWithLabel next week | 13:06 |
pedronis | s/any/anyway/ | 13:06 |
zyga | brb | 13:24 |
Chipaca | pedronis: with the completion tweak we no longer offer 'interfaces'; is that ok? | 13:33 |
pedronis | Chipaca: it's been deprecated since a while | 13:33 |
pedronis | but no strong opinion either way | 13:33 |
Chipaca | pedronis: i think it's ok, i do need to fix a test though :) | 13:34 |
Chipaca | very easy fix | 13:34 |
pedronis | use connections instead | 13:34 |
Chipaca | nah, the test is checking the completion of 'snap in<tab>' | 13:34 |
pedronis | ah ok | 13:35 |
Chipaca | :) | 13:35 |
pedronis | we still do have interface | 13:35 |
pedronis | singular | 13:35 |
pstolowski | Chipaca: whom to thank for the nice pre-baking label icon ;) | 13:36 |
Chipaca | yep, that's the fix | 13:36 |
pstolowski | ? | 13:36 |
Chipaca | pstolowski: not me! | 13:36 |
Chipaca | not this time :-D | 13:37 |
pstolowski | hmm | 13:37 |
Chipaca | but whoever did it, thank you for using the actual ๐ and not :bread: | 13:38 |
pedronis | mborzecki: good suggestion in the systemd prebaking PR | 13:44 |
mup | PR snapcraft#2786 opened: cli: add support for 'http-proxy' and 'https-proxy' parameters <Created by cjp256> <https://github.com/snapcore/snapcraft/pull/2786> | 13:54 |
pedronis | pstolowski: will you have time to stay after the standup? | 13:56 |
pstolowski | pedronis: yes, but i may need to drop for a while at around 4pm to pick my daughter from school | 13:57 |
pedronis | pstolowski: ok, let's see what we can, I would like to go over the firstboot stuff with you live | 13:57 |
pedronis | Chipaca: Q to your fix | 13:59 |
Eighth_Doctor | popey: `snapd` is now in EPEL 8, so the website needs instructions for CentOS 8 | 14:01 |
Eighth_Doctor | the instructions should basically be the same as for CentOS 7 | 14:02 |
Eighth_Doctor | so I think only the notes about lack of availablity on RHEL 8 and CentOS 8 need to be removed | 14:03 |
=== ricab is now known as ricab|lunch | ||
Eighth_Doctor | kyrofa: you might want to give me co-maintainership of squashfuse | 14:06 |
Eighth_Doctor | you can do that by going to https://src.fedoraproject.org/rpms/squashfuse, logging in, going to the project settings, go to users & groups, and add "ngompa" as an admin | 14:07 |
Chipaca | pedronis: running spread locally so can't push a fix for the typo, but made a suggestion there if you want to commit it | 14:14 |
degville | Eighth_Doctor: I've been tracking the EPEL updates - I'll update the install docs. | 14:15 |
Eighth_Doctor | degville: thanks | 14:16 |
Eighth_Doctor | `snapd` just synced out to EPEL 8 this morning | 14:16 |
degville | brilliant, thanks! | 14:16 |
pedronis | Chipaca: I committed | 14:59 |
Chipaca | nice OOM running spread locally | 15:07 |
mup | PR snapd#7707 opened: snap: add TestDownloadDirectStoreHappy test <Created by mvo5> <https://github.com/snapcore/snapd/pull/7707> | 15:19 |
=== ricab|lunch is now known as ricab | ||
zyga | re | 15:29 |
zyga | family home now | 15:29 |
zyga | back to work | 15:29 |
roadmr | \o/ | 15:33 |
zyga | mvo, pstolowski, mborzecki: we will likely visit graves tomorrow as it's bound to rain all weekend here | 15:33 |
* cachio lunch | 15:34 | |
mvo | has anyone figured out why the tests are unhappy? | 15:36 |
Chipaca | mvo: because they hate us | 15:37 |
Chipaca | mvo: i've just seen an error in prepare on centos, for example | 15:38 |
Chipaca | + systemctl daemon-reload | 15:38 |
Chipaca | Failed to execute operation: Connection reset by peer | 15:38 |
Chipaca | and on restore, also on centos, | 15:39 |
Chipaca | + systemctl daemon-reload | 15:39 |
Chipaca | Failed to execute operation: Connection timed out | 15:39 |
Chipaca | it's a fractal of WAT | 15:39 |
Chipaca | a WATdelbrot curve | 15:39 |
mvo | hrm, hrm, if its all centos I'm inclined to move it to unstable-systems until its fixed | 15:44 |
mvo | kind of annoying that it blocks landing stuff | 15:44 |
Chipaca | mvo: ah it is in unstable already | 15:44 |
Chipaca | mvo: sorry | 15:44 |
Chipaca | mvo: at least this particular centos is :) | 15:44 |
* mvo nods | 15:45 | |
Chipaca | the task that says CentOS contains no CentOS | 15:45 |
kyrofa | Eighth_Doctor, are you just looking to get squashfuse updated? Or is there a larger issue that I've missed? | 15:47 |
Chipaca | error: Get https://fastly.cdn.snapcraft.io/download-origin/fastly/b8X2psL1ryVrPt5WEmpYiqfr5emixTd7_1797.snap?token=1572548400_b95c66d9e68375e958b164a800f1b618a0e69653: x509: certificate signed by unknown authority (possibly because of "crypto/rsa: verification error" while trying to verify candidate authority certificate "GlobalSign CloudSSL CA - SHA256 - G3") | 15:50 |
Chipaca | mvo: ^ | 15:50 |
Chipaca | mvo: that might have something to do with the red | 15:51 |
mvo | Chipaca: *grumpf* not cool | 15:56 |
zyga | wahaat? | 15:59 |
zyga | SSL is such a shitshow | 15:59 |
zyga | ijohnson: did my response to the bug related to ldcache make sense? | 16:00 |
ijohnson | zyga: yes sorry for not responding in the bug, but I was able to work around it by having ldconfig work on some other file, then I just copy that file to what's bind-mounted so it never calls unlink on the file | 16:01 |
ijohnson | it is unfortunate that there's not much better we can do about that though | 16:02 |
zyga | ijohnson: I suggested a workaround | 16:02 |
zyga | ijohnson: a symlink would make it easier | 16:02 |
ijohnson | ehh not quite | 16:02 |
ijohnson | I tried the symlink but it still has problems | 16:02 |
zyga | no? is the code sensitive to symlinks? | 16:02 |
zyga | I see, well worth a try | 16:02 |
zyga | I think it will be nicer | 16:02 |
zyga | once we have /etc tmpfs | 16:03 |
ijohnson | I don't remember the issue with symlink, but I tried it and it didn't work for some reason | 16:03 |
zyga | or not tmpfs | 16:03 |
zyga | just a /var/lib/snapd/.../etc | 16:03 |
ijohnson | perhaps it was not an issue with the symlink directly, but anyways I have a workaround | 16:03 |
Eighth_Doctor | kyrofa: 1. squashfuse needs to be updated, it needs to be introduced to epel8, and squashfuse binary needs to be replaced with a symlink to squashfuse_ll binary | 16:03 |
zyga | mborzecki: I have some ideas on how to solve the lock issue | 16:08 |
kyrofa | (1) is a known issue, but squashfuse_ll is news to me. Can you give me some background? | 16:08 |
zyga | mborzecki: the inhibit files can be locks | 16:09 |
zyga | mborzecki: held by snapd | 16:09 |
zyga | mborzecki: and unlocked by snapd | 16:09 |
zyga | mborzecki: and snap run, if the files are present, try to grab a shared lock | 16:10 |
zyga | mborzecki: while snapd holds an exclusive lock | 16:10 |
zyga | mborzecki: I need to think how to create an inhibition file in a way that is race free | 16:10 |
zyga | mborzecki: but I think it can be done by allowing snapd to check for ability to grab the exclusive lock | 16:11 |
zyga | mborzecki: so if the lock is held by snap run already | 16:11 |
mup | PR snapd#7708 opened: parts/plugins: don't xz-compress a deb we're going to discard <Created by chipaca> <https://github.com/snapcore/snapd/pull/7708> | 16:11 |
zyga | mborzecki: we will know in snapd | 16:11 |
zyga | mborzecki: anyway, enjoy your evening | 16:11 |
mborzecki | zyga: we probably need to draw/write this down | 16:11 |
zyga | mborzecki: yeah, I'll write it down in the doc | 16:11 |
Chipaca | zyga: you still here? | 16:21 |
Chipaca | zyga: i think something's change such that google:ubuntu-18.04-64:tests/main/snap-seccomp-syscalls fails | 16:22 |
Chipaca | zyga: https://paste.ubuntu.com/p/FGp43JqGGJ/ | 16:23 |
Chipaca | or maybe that's an mvo ^ not sure | 16:23 |
Chipaca | looks like we got a bunch of new syscalls | 16:23 |
mvo | Chipaca: meh, ok | 16:23 |
Chipaca | most around time64 | 16:23 |
Chipaca | mvo: so that's probably why a lot of things are red | 16:24 |
Chipaca | there's also a fsconfig/fsmount/fsopen/fspick thing we might be interested in :) | 16:24 |
Chipaca | read more about 'em here: https://lwn.net/Articles/759499/ | 16:25 |
mvo | Chipaca: hm, thats on 18.04? | 16:25 |
Chipaca | they look like a change in the right direction, aiui, given some of our woes | 16:26 |
Chipaca | mvo: ye | 16:26 |
Chipaca | so i suspect this is a blocker, righ tnow | 16:28 |
mvo | Chipaca: I can fix this probably | 16:29 |
zyga | Chipaca: yes I'm here | 16:29 |
zyga | Chipaca: looking now | 16:29 |
Chipaca | p > 0.9? | 16:29 |
zyga | oh | 16:29 |
zyga | lots new | 16:29 |
mvo | Chipaca: yeah | 16:29 |
zyga | we have some problems | 16:30 |
zyga | clone3 :/ | 16:30 |
zyga | it may bite us | 16:30 |
zyga | as soon as glibc uses it | 16:30 |
zyga | and we respond with EPERM | 16:30 |
zyga | and glibc doesn't use older clone | 16:30 |
zyga | Chipaca: 180.04 though is very surprising | 16:31 |
zyga | Chipaca: is this new seccomp in 18.04 | 16:31 |
zyga | ? | 16:31 |
zyga | jdstrand: ^ do you know if we have a backported libseccomp in 18.04 | 16:31 |
Chipaca | zyga: this is in 18.04, as above | 16:31 |
zyga | jdstrand: or a new kernel and libseccomp now reports new syscalls it knew about | 16:31 |
Chipaca | zyga: ie this is failing google:ubuntu-18.04-64:tests/main/snap-seccomp-syscalls | 16:31 |
zyga | Chipaca: I'll park my work and look | 16:32 |
pstolowski | pedronis: +1 on #7694 | 16:45 |
mup | PR #7694: many: load/consume Core 20 seeds (aka recovery systems) <Created by pedronis> <https://github.com/snapcore/snapd/pull/7694> | 16:45 |
zyga | Chipaca: testing locally | 16:48 |
zyga | I'll send a PR soon | 16:48 |
zyga | brb | 16:48 |
Chipaca | zyga: FWIW 4.15.0-47-generic landed in 18.04 recently | 16:50 |
Chipaca | zyga: no seccomp update since august afaik | 16:51 |
Chipaca | (that's according to the dpkg.log in this kvm 1804, fwiw) | 16:51 |
pedronis | pstolowski: thanks, tried to answer your question | 16:52 |
mup | PR snapd#7694 closed: many: load/consume Core 20 seeds (aka recovery systems) <Created by pedronis> <Merged by pedronis> <https://github.com/snapcore/snapd/pull/7694> | 16:53 |
pedronis | Chipaca: pstolowski: I'm going now to rebase the two follow ups | 16:55 |
Chipaca | pedronis: thank you | 16:55 |
pstolowski | great | 16:55 |
pedronis | Chipaca: pstolowski: mvo: I rebased #7695 (firstboot stuff) and #7696 (prepare image stuff) and they are ready for review | 16:59 |
mup | PR #7695: o/devicestate: the basics of Core 20 firstboot support with test <Created by pedronis> <https://github.com/snapcore/snapd/pull/7695> | 16:59 |
mup | PR #7696: cmd/snap,image: initial support for Core 20 in prepare-image with test <Created by pedronis> <https://github.com/snapcore/snapd/pull/7696> | 16:59 |
pedronis | hopefully that latter doesn't break spread tests | 17:00 |
jdstrand | Chipaca, zyga: I need to step away in a minute, but did the gke kernel get an update? (I didn't look at the failure either) | 17:02 |
pstolowski | pedronis: ty | 17:02 |
Chipaca | jdstrand: plain ol' 18.04 at least did | 17:03 |
jdstrand | Chipaca: I have like 3 minutes. can you give a url so I can see the failure? | 17:03 |
Chipaca | jdstrand: the failure is in https://api.travis-ci.org/v3/job/605555475/log.txt the diff expected vs found syscalls is https://paste.ubuntu.com/p/FGp43JqGGJ/ | 17:04 |
Chipaca | jdstrand: dunno if you saw zyga say he's got a fix that he's testing locally | 17:05 |
Chipaca | i assume you did and you're wanting to look at what the new syscalls are to be in the loop, but thought i'd mention it just in case :) | 17:05 |
mborzecki | Chipaca: zyga: we need to update the list of syscalls in cmd/snap-seccomp/syscalls/syscalls.go, apparently upstream was updated | 17:06 |
jdstrand | Chipaca: I was curious, yes. also curious cause this shouldn't have happened unless the kernel changed iirc how the test was written | 17:06 |
jdstrand | but I may be forgetting the test | 17:07 |
jdstrand | oh, 3 minutes up. back in a little while | 17:07 |
mvo | pedronis: \o/ | 17:09 |
jdstrand | Chipaca: thanks for all the info btw | 17:20 |
* jdstrand is now really gone | 17:20 | |
mborzecki | Chipaca: zyga: https://github.com/snapcore/snapd/pull/7709 | 17:21 |
mup | PR #7709: cmd/snap-seccomp/syscalls: update the list of known syscalls <Simple ๐> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/7709> | 17:21 |
mup | PR snapd#7709 opened: cmd/snap-seccomp/syscalls: update the list of known syscalls <Simple ๐> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/7709> | 17:21 |
jdstrand | oh right, we git clone libseccomp and go from there. nothing to do with the host | 17:24 |
jdstrand | mborzecki: thanks | 17:24 |
jdstrand | ok, really, really gone | 17:24 |
mup | PR snapd#7710 opened: snap-seccomp: add new syscalls after libseccomp update <Created by mvo5> <https://github.com/snapcore/snapd/pull/7710> | 17:25 |
mup | PR snapcraft#2775 closed: remote-build: add clean flag <Created by cjp256> <Closed by cjp256> <https://github.com/snapcore/snapcraft/pull/2775> | 17:27 |
kenvandine | jdstrand: any thoughts on bug 1661626 | 17:31 |
mup | Bug #1661626: GSettings/dconf reports incorrect values on setting change under confinement <desktop> <snapd:Triaged by marcustomlinson> <https://launchpad.net/bugs/1661626> | 17:31 |
mup | PR snapd#7710 closed: snap-seccomp: add new syscalls after libseccomp update <Created by mvo5> <Closed by mvo5> <https://github.com/snapcore/snapd/pull/7710> | 17:38 |
zyga | AFK with kids now | 18:10 |
cachio | zyga, hey | 18:24 |
cachio | zyga, snap-seccomp-syscalls test is failing and reading the comments in the test it sayts | 18:24 |
cachio | # both lists should be identical, otherwise we need an update in snap-seccomp | 18:25 |
cachio | any idea on how to update snap-seccomp? | 18:25 |
Chipaca | cachio: PR already up | 18:27 |
cachio | Chipaca, nice, thanks | 18:27 |
Chipaca | cachio: https://github.com/snapcore/snapd/pull/7709 | 18:27 |
mup | PR #7709: cmd/snap-seccomp/syscalls: update the list of known syscalls <Simple ๐> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/7709> | 18:27 |
cachio | Chipaca, I'll merge it once the tests pass | 18:28 |
cachio | it has 2 +1 | 18:28 |
Chipaca | cachio: only if i don't merge it first :-p | 18:28 |
cachio | hhehee | 18:30 |
zyga | Iโm upstairs with now sleeping Lucy | 18:38 |
zyga | Maciek made the same patch but found the upstream commit so kudos to that | 18:39 |
zyga | Letโs mere it | 18:39 |
* pstolowski afk | 18:40 | |
Chipaca | zyga: just got my first trick-or-treaters *ever* | 18:57 |
Chipaca | and, i must admit i did not tell them to go away | 18:57 |
noise][ | nice, did you give them snapcraft stickers orโฆ ? | 19:01 |
Chipaca | noise][: i gave them support tickets that were older than they were | 19:15 |
Chipaca | nah, i panicked and gave them some m&ms | 19:15 |
Chipaca | i've been in this house 3? maybe 4 years and this is the first time they've knocked | 19:16 |
noise][ | m&ms was the right call | 19:19 |
mup | PR snapcraft#2787 opened: Safe grade <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/2787> | 19:46 |
mup | PR snapd#7711 opened: seed: test and improve Core 20 seed handling errors <Created by pedronis> <https://github.com/snapcore/snapd/pull/7711> | 19:52 |
mup | PR snapd#7712 opened: seed: support in Core 20 seeds local unasserted snaps for model snaps <Created by pedronis> <https://github.com/snapcore/snapd/pull/7712> | 20:01 |
* cachio EoD | 21:03 | |
mup | PR snapd#7713 opened: seed: Core 20 seeds channel overrides support for grade dangerous <Created by pedronis> <https://github.com/snapcore/snapd/pull/7713> | 21:16 |
mup | PR snapd#7709 closed: cmd/snap-seccomp/syscalls: update the list of known syscalls <Simple ๐> <Created by bboozzoo> <Merged by chipaca> <https://github.com/snapcore/snapd/pull/7709> | 23:28 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!