mupPR snapd#8927 closed: tests: avoid exit 1 when nested type var is not defined <Created by sergiocazzolato> <Merged by anonymouse64> <https://github.com/snapcore/snapd/pull/8927>00:11
=== jamesh_ is now known as jamesh
mupPR snapd#8748 closed: overlord: refuse to install snaps providing user daemons on Ubuntu 14.04 <Created by jhenstridge> <Merged by anonymouse64> <https://github.com/snapcore/snapd/pull/8748>01:36
jameshijohnson: thanks!01:40
ijohnsonjamesh: np01:40
zygaI’ll start at nine06:07
mborzeckizyga: heya, last day of 'school'06:07
zygaYes. Kids are very happy06:07
zygaToo bad the circumstances are not better06:07
mvohey zyga and mborzecki06:22
seb128hey mvo, thx for the fontconfig and freetype fix ... do you have any hint on the easiest way to test the change?06:30
zygaTook meds now06:30
zygaWill be able to work in 3006:30
mvoseb128: I can build a custom snapd snap for trying it out, will take some minutes06:31
seb128mvo, that would be nice if you could, I'm happy to give it a try today06:32
mvoseb128: sure, building now and once done will ping you06:35
mvogood morning pstolowski07:11
mborzeckipstolowski: heya07:14
mborzeckineed coffee07:14
* zyga starts with reviews07:25
zygamborzecki: btw, for evening reading: https://lwn.net/Articles/824380/07:26
zygaoh and https://github.com/snapcore/snapd/pull/8728 is a low hanging fruit07:26
mupPR #8728: tests: detect stray dbus-daemon <Test Robustness> <Created by zyga> <https://github.com/snapcore/snapd/pull/8728>07:26
mborzeckizyga: oh, nice07:26
mupPR snapd#8928 opened: tests: add spread test for disconnect undo caused by failing disconnect hook <Simple πŸ˜ƒ> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8928>07:28
pstolowskizyga: ^ something easy for a warmup ;)07:28
jameshzyga: if you've got time, could you give this a review? https://github.com/snapcore/snapd/pull/886107:30
mupPR #8861: data,packaging,wrappers: extend D-Bus service activation search path <Created by jhenstridge> <https://github.com/snapcore/snapd/pull/8861>07:30
zygajamesh: I'll do my best to help07:31
jameshzyga: thanks.  You gave it a once over earlier, and now I just need a second sign off07:31
zygajamesh: ah, I really like how github shows what I reviewed07:34
zygaand shows only changes07:34
pstolowskizyga: nb, snap remove --force needs ack from Samuele07:37
zygaI was thinking if it makes sense to fail on interface hooks when removing a snap07:37
pstolowskizyga: indeed07:41
mvoseb128: sorry, took a while, had some unclean checkout and a bug in the cleanup: https://people.canonical.com/~mvo/tmp/snapd_2.45.1+git1648.gf492c71_amd64.snap is the test snaps08:09
zygamborzecki: https://github.com/snapcore/snapd/pull/8925#pullrequestreview-43809466608:13
mupPR #8925: bootloader: allow managed bootloader to update its boot config <Needs Samuele review> <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8925>08:13
mborzeckizyga: thx08:19
mborzeckimvo: can you cherry pick https://github.com/snapcore/snapd/commit/3192141bc8481f58639e71a099b17a608366b7a0#diff-ccde24f4fb0b0adce138f9b26cea79ed to 2.45 in case we do another point release?08:21
mvomborzecki: sure thing08:21
mborzeckimvo: thanks08:22
mvomborzecki: thanks, done08:22
mvomborzecki: nice one!08:22
zygamborzecki: what does NoSlashBoot do?08:29
mborzeckizyga:  haha, so it means you're lookin at the root of the boot fs08:29
mborzeckizyga: instead of the actual rootfs where you'd normally find /boot/..08:30
zygaso /foo/bar/froz becomes just /?08:30
mborzeckiiow no /boot08:30
* zyga looks at the source08:30
mborzeckizyga: not necessarily, we ahve /boot/grub/.. which is a bind mount of /EFI/ubuntu from system-boot iirc08:30
seb128mvo, thx, I will give a try a bit later and let you know if it improves things08:50
mvoseb128: thanks08:53
zygamborzecki: https://github.com/snapcore/snapd/pull/8924#pullrequestreview-43813158709:20
mupPR #8924: gadget, bootloader: preserve managed boot assets during gadget updates <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8924>09:20
zygamore reviews09:32
zygathank you for looking at the tracking PR09:32
zygaI will iterate on it soon09:32
mborzeckizyga: thanks for the reviews too09:37
zygapstolowski: https://github.com/snapcore/snapd/pull/8923#pullrequestreview-43815628309:42
mupPR #8923: wrappers: helper for enabling services (8/9) <Services βš™οΈ> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8923>09:42
zygajdstrand: is https://github.com/snapcore/snapd/pull/8920 something that I should review now or is it better to wait for the final form?09:42
mupPR #8920: interfaces: update cups-control and add cups for providing snaps <β›” Blocked> <Created by jdstrand> <https://github.com/snapcore/snapd/pull/8920>09:42
zygajdstrand: also note that opening a draft vs a blocked PR is probably more idiomatic for GH09:43
pstolowskizyga: thanks!09:43
mupPR snapd#8929 opened: [RFC] many: add new "daemon-startup: inhibit" option <Needs Samuele review> <Created by mvo5> <https://github.com/snapcore/snapd/pull/8929>09:48
zygamvo: ^ quick comment09:53
mvozyga: thanks, interessting idea09:58
zygamvo: it doesn't have to be complex09:58
zygaperhaps just09:58
zygastartup-presets: default: app-name: disabled09:58
zygaif you catch my "language"09:58
zygadefault would be the only supported value now09:58
zygapresets are somewhat different, as they come from the distro09:59
zygabut we could convey the same idea09:59
zygae.g. apache is not started automatically09:59
zygabut the server preset makes that so09:59
zygawhile desktop does not09:59
zyga(just a made up example)10:00
zygaperhaps an app could even have a snapctl level influence over the preset to use10:00
zygaanyway, just an idea10:00
zygahttps://www.freedesktop.org/software/systemd/man/systemd.preset.html has some docs about tihs10:00
pstolowski#8928 needs 2nd review and is a low hanging 🍎10:02
mupPR #8928: tests: add spread test for disconnect undo caused by failing disconnect hook <Simple πŸ˜ƒ> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8928>10:02
zygapstolowski: I see what you did there :D10:03
mupPR snapd#8930 opened: many: managed boot config during run mode setup <UC20> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/8930>10:03
mvozyga: yeah, the idea of having hooks that would enable/disable under certain conditions is part of the idea for this10:05
zygaijohnson: https://github.com/snapcore/snapd/pull/8918#pullrequestreview-43816131810:07
mupPR #8918: many: make nested spread tests more reliable <Test Robustness> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8918>10:07
zygareview days are good10:08
zygaok, only dbus to review10:55
zygaand then back to iteration10:55
zygaor maybe small lunch break and then iteratoin10:56
mupPR snapd#8931 opened: tests/main/install-fontconfig-cache-gen: enhance test by verifying, add fonts to test <Test Robustness> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8931>11:18
zygajamesh: https://github.com/snapcore/snapd/pull/8861#pullrequestreview-43821440711:30
mupPR #8861: data,packaging,wrappers: extend D-Bus service activation search path <Created by jhenstridge> <https://github.com/snapcore/snapd/pull/8861>11:30
zygacc mvo ^11:30
zyganot +1, not sure what to do about this11:30
zygaI think we need to revisit our reexecution strategy with regards to how it should be structured internally11:30
zygawhat's okay and not and if we need any belts and suspenders to make it sane over time11:30
zyga9% of battery, time for a trip downstairs for the charger11:33
zygaoh boy11:33
jdstrandzyga: probably best to wait for cups-control. I wasn't expected it to be blocked for so long but I got pulled away. if you're looking to review things, I suggest https://github.com/snapcore/snapd/pull/8699 and https://github.com/snapcore/snapd/pull/830111:39
mupPR #8699: interfaces/desktop-launch: support confined snaps launching other snaps <Needs Samuele review> <Needs security review> <Created by AlanGriffiths> <https://github.com/snapcore/snapd/pull/8699>11:39
mupPR #8301: interfaces/many: deny arbitrary desktop files and misc from /usr/share <Created by jdstrand> <https://github.com/snapcore/snapd/pull/8301>11:39
zygajdstrand: ack, thanks!11:39
jdstrandbut I am getting back to cups-control today (want to right a spread test)11:39
zygaI'll do both, just need to charge11:39
zygain the office11:54
zygastairs feel wobbly11:54
zygahttps://github.com/snapcore/snapd/pull/8728 needs a 2nd review\12:17
mupPR #8728: tests: detect stray dbus-daemon <Test Robustness> <Created by zyga> <https://github.com/snapcore/snapd/pull/8728>12:17
* zyga takes a break to rest12:25
mupPR snapd#8932 opened: Update security profiles after disconnecting plug-slot in connect und… <Bug> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8932>12:44
mupPR snapd#8933 opened: tests/core/uc20-recovery: apply hack to get gopath in recover mode w/ external backend <Test Robustness> <UC20> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8933>12:59
mborzeckicachio: downloaded the image, trying now13:33
cachiomborzecki, tx13:34
zygaI will resume work in 30 minutes13:37
zygathe schedule for painkillers is somewhat annoying13:38
zygabtw, typing on the faceless keyboard is really much easier than I anticipated; I think the only thing I stil struggle with are arrow keys13:38
mborzeckicachio: in what scenarios does the test fail?13:41
cachiojust need to run the test like this13:41
cachiofirst prepare ssh13:42
cachio./tests/lib/external/prepare-ssh.sh localhost 8022 sergio-j-cazzolato13:42
cachiowith your lp id13:42
cachioexport SPREAD_EXTERNAL_ADDRESS=localhost:802213:42
cachioand run spread -debug external:ubuntu-core-20-64:tests/core/snap-set-core-config13:42
cachiomborzecki, to start the vm I do13:43
cachiosudo qemu-system-x86_64 -smp 2 -m 2048 -snapshot -net nic,model=virtio -net user,hostfwd=tcp::8022-:22 -drive file=/usr/share/OVMF/OVMF_CODE.fd,if=pflash,format=raw,unit=0,readonly=on -drive  file=~/workspace/validator/images/output/pc-amd64-20-edge/pc.img,cache=none,format=raw,id=disk1,if=none -device virtio-blk-pci,drive=disk1,bootindex=1 -machine accel=kvm13:43
mborzeckicachio: hmm, i've built a pc20 image and the test would fail there too from what i can see13:43
mborzeckiwonder why it does not fail when we run it on gce13:43
cachiomborzecki, that's my concern13:43
cachionot sure where the difference is13:44
mupPR snapd#8728 closed: tests: detect stray dbus-daemon <Test Robustness> <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/8728>13:44
cachioI didn't see any special config for console conf in the code13:44
ijohnsoncachio: what's the test that doesn't work ?13:45
ijohnsonsorry I had to drop from SU before you could share details13:45
cachioijohnson, snap-set-core-config13:45
ijohnsonI'm all setup to run external spread tests easily now though :-)13:45
ijohnsoncachio: ack I will give it a try right now13:45
ijohnsoncachio: same pc.img.xz as yesterday ?13:45
ijohnsoncachio: much like the other test, if this is being run with an external backend, it functions differently13:52
ijohnsoncachio: the issue is that this test tries to re-enable console-conf after disabling it, but that only works if console-conf was never run manually by the user13:53
mborzeckiheh, 2nd thunderstorm today13:53
ijohnsoncachio: in our case, we manually run console-conf in order to be able to create the test / ubuntu users13:53
ijohnsoncachio: which means that snapd does the right thing and when we unset the system config, snapd doesn't undo console-conf disabling itself13:54
ijohnsoncachio: the test needs to be adjust to special case external backend, I will prep a PR for oyu13:54
mborzeckiijohnson: thanks for investigating!13:55
cachioijohnson, thanks13:59
cachioijohnson, everything is realted to all the magic we did to prepare snapd to be executed on google14:00
mupPR snapd#8934 opened: overlord: mock timings.DurationThreshold in TestNewWithGoodState <Simple πŸ˜ƒ> <Created by mvo5> <https://github.com/snapcore/snapd/pull/8934>14:04
* zyga finally had lunch and meds and can get back to work :)14:20
ijohnsoncachio: yes essentially all these bugs are related to the magic we do for snapd + UC on google14:57
zygaijohnson: o/15:03
ijohnsonhey zyga15:04
zygaah, you're around - great15:04
zygaI'll push tweaks to comments in 8829 in a moment, wanted to ask if you could look at my explanations there15:04
zygaI'll let you know when they are up15:04
ijohnsonzyga: yes I'll take a look just let me know when they're up15:05
zygathank you15:05
* cachio lunch15:11
zygaijohnson: pushed https://github.com/snapcore/snapd/pull/882915:23
mupPR #8829: sandbox/cgroup: add tracking helpers <Created by zyga> <https://github.com/snapcore/snapd/pull/8829>15:23
zygait's broken down in commits so you can see it more easily15:23
zygathat's all15:24
zygaif you have cycles today I'd love a review of the counterpart https://github.com/snapcore/snapd/pull/886315:24
mupPR #8863: sandbox/cgroup: allow discovering PIDs of given snap <Created by zyga> <https://github.com/snapcore/snapd/pull/8863>15:24
ijohnsonzyga: thanks, one small comment then LGTM15:26
ijohnsonzyga: I'll take a look at the other one today too15:26
ijohnsonxnox: what is the canonical way to know if console-conf ran? I thought it was the presence of the /var/lib/console-conf/complete file, but it doesn't appear that file is ever written by console-conf itself ? that file simply seems to be used to _disable_ console-conf, not to tell whether console-conf ran or not15:45
zygaijohnson: applied now15:52
xnoxijohnson: pass. I am not sure if it is possible to tell if it ran, or did anything, or if it's state was cleared, or like who/what created the managed user....15:53
ijohnsonxnox: I'm wondering about for our tests in snapd about disable console-conf15:53
ijohnsonxnox: we can hack some things into place, but our test was checking for that file, but it seems that file is not actually created when we expected it to be15:54
pstolowskiijohnson: hey, would you take a look at #8928? it's test only PR15:54
mupPR #8928: tests: add spread test for disconnect undo caused by failing disconnect hook <Simple πŸ˜ƒ> <Created by stolowski> <https://github.com/snapcore/snapd/pull/8928>15:54
ijohnsonxnox: if there's not an "official way" that's fine I can replace it with a working hack to see that the test is checking the right thing15:54
ijohnsonpstolowski: sure, will be my afternoon though at this point15:54
xnoxijohnson: maybe something is buggy? It should be leaving that stamp file at the end... When ran successfully. At least that's what I thought.15:56
mvoseb128: did my updated snapd snap help with the fc-cache issue? no rush, just curious if I can update anything in the PR15:56
ijohnsonxnox: it doesn't leave a stamp file on uc20 or uc18 afaict15:56
ijohnsonxnox: shall I file a bug ?15:56
ijohnsonzyga: thanks for adding that Error(), I'll keep an eye on the PR and merge it for you tonight15:57
seb128mvo, ah, sorry, I meant to comment, I just tested half an hour ago, I wiped the cache, installed your snap and the font was correctly listed in the newly generated cache, so seems to work from that testing15:58
mvoseb128: \o/ yay15:58
zygaonce it lands I need to merge it back into to big helper15:58
seb128mvo, thanks for the quick fix :)15:58
mvoseb128: my pleasure, thanks for all your help with this! (and to jamesh)15:58
ijohnsonmvo: regarding that bug, how will the spread tests get the fix for that? do we build the fc-cache helpers in the snapd deb the way we do for the snapd snap? just wondering how we could properly fix my spread test showing that it works16:01
seb128mvo, I've added a comment on the bug and the pr now16:01
ijohnsoncachio: do you have time today to verify #8933 ?16:02
mupPR #8933: tests/core/uc20-recovery: apply hack to get gopath in recover mode w/ external backend <Test Robustness> <UC20> <Created by anonymouse64> <https://github.com/snapcore/snapd/pull/8933>16:02
mvoijohnson: I think we just merge the fix into edge, rebuild the snap and then re-run your test (that now no longer needed to rebuild the snapd snap). it's not as nice as you test but we can always revert again if needed16:03
mvoijohnson: sorry, it's very low-tech, a real friday evening suggestion :)16:03
ijohnsonmvo: so the test will use the snapd snap ?16:03
ijohnsonmvo: that's fine, just wondering since it's a little annoying in the test to have to revert from the snapd snap to the deb, certainly doable though I think16:03
mvoijohnson: yeah, once fc-cache-builder is in git and the snapd snap is rebuild we should have the right version16:03
ijohnsonmvo: ah actually all the font tests are run on classic, so it's not a problem, the magic spread restore should take care of it16:04
ijohnsonnvm me, thanks mvo16:04
mvoijohnson: aha, ok16:04
mvoijohnson: sorry, I think I was not even grasping the depth of the question :)16:04
mvoijohnson: but happy to help!16:04
ijohnsonmvo: haha no worries it's probably EOW time for you anyways :-)16:04
mvoijohnson: it is - plus it's hot here :)16:05
ijohnsonyah I hear that16:05
* mvo is good at making excuses16:05
* mvo will just write one or two more mails and then call it a day16:05
ijohnsonxnox: seems that /run/console-conf/login-details* is the best indicator if console-conf ran successfully / normally on the system16:10
xnoxijohnson: but that could also be generated if the system got provisioned elsehow too I think16:13
ijohnsonxnox: I don't see it, I tried both with cloud-init and with system user assertions and don't see that dir16:13
ijohnsonxnox: maybe I checked in bad ways16:13
xnoxAnd after reboot it's still there?16:13
ijohnsonxnox: but I can make my test pass now, so that's all that's important :-)16:13
ijohnsonxnox: afaict yes16:13
xnoxCool. Roll with it.16:14
xnoxNext time I touch consoleconf I will try to understand how it works.16:14
ijohnsonhaha sounds good16:14
=== benfrancis1 is now known as benfrancis
cachioijohnson, sure16:45
cachioijohnson, https://paste.ubuntu.com/p/RvVKN4978X/17:50
ijohnsoncachio: very interesting, was this when the system was in recover mode or run mode ?17:51
cachiono space error17:51
ijohnsoncachio: ah I see that is in recover mode17:51
=== benfrancis0 is now known as benfrancis
=== benfrancis4 is now known as benfrancis
ijohnsoncachio: also re: the snap-set-core-config, the test exposed a real bug, so I will need to fix that bug before we can fix the test18:16
cachioijohnson, hehe18:33
cachiogood news in that case18:34
ijohnsoncachio: yeah :-)18:34
ijohnsoncachio: also for your uc20-recovery is that snapd panic reproducible for you ?18:34
ijohnsonor did it just happen once, as I don't see that with my VM18:34
cachiothe once I sent before?18:36
ijohnsoncachio: the one from https://paste.ubuntu.com/p/hsNtwZNYsw/18:36
ijohnsoncachio: what arguments are you running the external VM with ?18:37
cachioI have a debug session opened18:37
ijohnsoncachio: I use this:18:37
ijohnsonmaybe I need to provide less RAM to the VM18:38
cachioijohnson, I use https://paste.ubuntu.com/p/mCyynrpCpQ/18:38
cachiobig diff18:38
ijohnsoncachio: ack I'll try with 2048 to see if I can reproduce it18:39
ijohnsoncause yeah we might have run out of space, but we shouldn't panic in that case I'm pretty sure18:39
cachioI'll try with 3gb18:39
mupPR snapd#8829 closed: sandbox/cgroup: add tracking helpers <Created by zyga> <Merged by zyga> <https://github.com/snapcore/snapd/pull/8829>18:40
cachioijohnson, with 3GB didn't work19:25
cachiotrying with 4GB now19:25
cachiogot stuck after reboot19:25
ijohnsoncachio: you mean with 3 GB it didn't panic but it did get stuck after reboot?19:28
cachiodont know because after 20 minutes I killed spread19:28
cachioijohnson, this is the output19:29
ijohnsonah cachio that seems to be the bug that I thought I fixed by deleting the ~/.ssh/rc file19:32
ijohnsoncachio: interesting, I need to step out for a while now, but I will try again to see if I can reproduce that issue19:32
* ijohnson -> afk for bit19:32
cachioijohnson, np, see you then19:34
cachioijohnson, with 4GB workerd well19:38
cachioijohnson, https://paste.ubuntu.com/p/PDt2jk3W2f/19:39
cachiothe problem is that in the lab devices we don't have 4gb19:53
mupPR snapcraft#3190 opened: extensions: export content snap egl vendor dir <maintenance> <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/3190>19:59
ijohnsoncachio: yes I will see what to do about using 2G, also won't this test be run on devices like rpi that might only have 1G of RAM?20:01
cachioijohnson, yes20:25
cachioit runs in pi3 with 1 or 2 gb20:25
cachioand pi4 with 4/8gb20:26
cachiowhich is fine20:26
mupPR snapcraft#3174 closed: link_or_copy: do not try to create hardlinks to symlinks <bug> <Created by hpoul> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3174>21:09
mupPR snapcraft#3190 closed: extensions: export content snap egl vendor dir <maintenance> <Created by sergiusens> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/3190>21:09
mupPR snapcraft#3191 opened: plugins: introduce v1.FlutterPlugin <Created by sergiusens> <https://github.com/snapcore/snapcraft/pull/3191>22:04

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!