/srv/irclogs.ubuntu.com/2019/04/15/#snappy.txt

mborzeckimorning05:18
mborzecki60 PRs :/05:53
zygagood morning06:54
zygamborzecki: thank you for the reviews06:55
mborzeckizyga: morning06:56
mborzeckimvo: hey06:56
mvohey mborzecki06:56
mvomborzecki: good morning! what did I miss? anything interessting happend?06:57
mborzeckimvo: not really, just a lot of open PRs06:57
mvomborzecki: ok, let me look at those07:00
mborzeckizyga: mvo: need to go out for a while, sent a message in the forum07:02
zygaack07:02
=== pstolowski|afk is now known as pstolowski
zygahey pstolowski07:09
mvohey zyga and pstolowski07:12
zygahey mvo!07:14
zygamvo: I think today is a review day07:14
zygaAFK with dog07:18
mvozyga: indeed, looking at the PRs I think you are right07:18
pedronismvo: zyga: hi, I'll also focus on reviews, at least the first part of the day07:38
pedronismvo: #6594 needs a quick look from you after the last round of changes and then can land (hopefully)07:49
mupPR #6594: [RFC] tests: run smoke tests on (almost) pristine systems <Created by mvo5> <https://github.com/snapcore/snapd/pull/6594>07:49
pstolowskimborzecki: hey, #6711 can land07:51
mupPR #6711: tests/main/selinux-lxd: make sure LXD from snaps works cleanly with enforcing SELinux <SELinux> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6711>07:51
mupPR snapd#6711 closed: tests/main/selinux-lxd: make sure LXD from snaps works cleanly with enforcing SELinux <SELinux> <Created by bboozzoo> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/6711>07:54
=== chihchun_afk is now known as chihchun
mupPR snapd#6696 closed: image: simplify prefer local logic  and fixes <Created by pedronis> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/6696>08:00
=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
mborzeckire09:04
zygahey mborzecki09:06
zygahow are you feeling?09:06
zygabrb, breakfast09:12
mborzeckihm 2.38 is in fedora updates repo, i might need to update the rpm upgrade triggers in our packaging09:23
mborzeckiheh, and looking at how the last the last 2 merges to master failed in spread i need to :P09:24
pedronispstolowski: I did passes on various of your PRs, some can land after small tweaks and then once merged in the follow ups I can look at those09:25
pstolowskipedronis: ty, yes, i'm addressing your comments09:26
pstolowskipedronis: i've updated https://github.com/snapcore/snapd/pull/6660, you approved it already but please see if the change makes sense09:35
mupPR #6660: cmd/debug: integrate new task timings with "snap debug timings" <Created by stolowski> <https://github.com/snapcore/snapd/pull/6660>09:35
pedronispstolowski: ok, will look in a little bit09:36
pstolowskipedronis: basically, i think Save had a small bug09:37
mborzecki#6713 will unbreak master09:47
mupPR #6713: tests/upgrade/basic: restore SELinux context of /var/cache/fontconfig <SELinux> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6713>09:47
mborzeckizyga: can you take a look at https://github.com/snapcore/snapd/pull/6692 ? super simple and can land quickly09:49
mupPR #6692: interfaces: cleanup internal tool lookup in system-key <Simple 😃> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6692>09:49
zygaabsolutely09:49
pedronispstolowski: looks ok, commented on something else though09:54
pedronispstolowski: sorry, fixed my comment (mistyped a couple of things in it)09:55
pstolowskipedronis: ah, ok, will remove that part of the comment thx09:55
zygamborzecki: https://github.com/snapcore/snapd/pull/6692/files#r275287380 ?09:56
mupPR #6692: interfaces: cleanup internal tool lookup in system-key <Simple 😃> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6692>09:56
mborzeckizyga: there are some units tests for findSnapdPath, maybe i could drop those10:01
zyga+110:03
zygait just seems odd to reduce a function to a call to another while keeping it when both seem to relate to the same topic10:04
sil2100Hey guys! Currently, do we update all the boot-partition contents on gadget snap refreshes?10:25
mvosil2100: we don't, this is part of the work that mborzecki  is doing for the bootloader asset refreshes10:26
sil2100Ok, since I remember this being a work-item but didn't know if it actually landed10:27
sil2100mvo: thanks!10:27
mvosil2100: do you need it for sometihng specific? is there anything we need to update?10:30
mvosil2100: I found the missing validations, sorry for that! will get this fixed10:30
* pstolowski lunch10:31
mupPR snapd#6713 closed: tests/upgrade/basic, packaging/fedoar: restore SELinux context of /var/cache/fontconfig, patch pre-2.39 mount units <SELinux> <Created by bboozzoo> <Merged by bboozzoo> <https://github.com/snapcore/snapd/pull/6713>10:35
mborzeckimaster should be fine now10:40
sil2100mvo: no no, just was wondering because of the new grub2 in the pc core18 gadget, Steve mentioned that existing users will be 'safe' because of snap refresh not updating the contents as of yet and I was wondering if that's really the case10:42
sil2100mvo: as I remembered discussions about making that happen10:42
sil2100mvo: thanks! Just give me a poke once the other buggers are verified10:43
mvosil2100: thanks10:48
* zyga coffee11:11
=== sparkieg` is now known as sparkiegeek
pedronissil2100: notice  that even once implemented it will be opt in from the part of the gadget11:31
sil2100pedronis: ok, good to knog11:31
sil2100*know11:31
* cachio afk 11:47
pedronispstolowski: hi, when you have a little bit of time, could you double check that we have a spread test for this:  https://forum.snapcraft.io/t/how-to-disconnect-a-snap-from-internet/10901/1111:52
pstolowskipedronis: sure, will do11:58
pedronisthx11:59
threshso I have a coredump, and my application doesnt run if it detects being running as root (so snap run --gdb is useless), what do?12:06
threshgdb --core=core /snap/vlc/x1/usr/bin/vlc results in a useless trace (all ??)12:06
zygathresh: it's not easy I'm afraid12:11
zygathresh: what is the base your snap is using? core or core18?12:11
threshzyga, it's core1812:11
zygathresh: so you somehow have to set up gdb to look at debug symbols from ubuntu18.04 -- I don't believe this is documented or explained anywhere12:13
threshcould having ubuntu 18.04 as a host OS help?12:13
zygayes but it's not required12:13
zygayou need the debug symbol packages12:13
zygaand you need to have a debug build of the app as well12:14
zygaI think it's a big endeavour to  explain and document properly though12:14
threshmight be easier to just spin up a VM that trying to do that on my debian12:14
zygayou don't need a vm12:14
threshsure, I've got the idea - will try it, thx12:14
zygajust debootstrap ubuntu 18.04 or get a tarball from cdimage.ubuntu.com12:14
mupPR snapd#6660 closed: cmd/debug: integrate new task timings with "snap debug timings" <Created by stolowski> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/6660>12:14
zygaI have *not* done this myself so anything you come up with is worth sharing12:15
zygagood luck!12:15
mborzeckipedronis: moved selinux cleanup task to the 'done' lane12:18
=== chihchun is now known as chihchun_afk
pedronismborzecki: great!12:20
pstolowskizyga: i'm going to review #6717 in a bit12:47
mupPR #6717: snap: fix interface bindings on implicit hooks <Created by zyga> <https://github.com/snapcore/snapd/pull/6717>12:47
pedronispstolowski: it needs another pass12:47
zygapstolowski: I think it needs a fix again12:47
pstolowskithanks for tackling it12:47
zygapstolowski: look at my comment12:47
zygapstolowski: because I think pedronis was spot on, twice :)12:47
zygapstolowski: and the simplistic fix was insufficient12:47
pedronisto be fair I wasn't spot on, when we tried to fix this first12:48
zygapstolowski: I will add some tests to it today, to show how it fails, then try to fix it, just working on another existing PR12:48
zygapedronis: as an extra idea, I was thinking about changing how we load yaml12:48
pstolowskizyga: ah, i see the new comments, ack12:48
zygapedronis: to allow it to know about the implicit hooks before we read the yaml12:48
zygapedronis: so that we can create them and avoid the late binding complexity12:49
pedroniszyga: maybe, that seems a bigger changes though12:49
zygapedronis: yes12:49
=== ricab is now known as ricab|lunch
mborzeckianyone up for a 2nd review of https://github.com/snapcore/snapd/pull/6688 ?13:01
mupPR #6688: gadget: add validation of cross structure overlap and offset writes <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6688>13:01
mupPR snapd#6625 closed: tests: system disable ssh for config defaults in gadget <Created by sergiocazzolato> <Merged by sergiocazzolato> <https://github.com/snapcore/snapd/pull/6625>13:06
mupPR snapcraft#2533 closed: tests: classic confinement spread tests for ant and maven  <Created by sergiusens> <Merged by sergiusens> <https://github.com/snapcore/snapcraft/pull/2533>13:12
zygamborzecki: any ideas, is ausearch failing? https://www.irccloud.com/pastebin/ohLqFsHP/13:24
mborzeckizyga: can you post the debug part of the log?13:25
zygaah, indeed13:26
zygaselinux denials https://www.irccloud.com/pastebin/s5RJWK6u/13:26
cmatsuokakenvandine: hello ken13:29
cmatsuokakenvandine: did you build any kde snap using the frameworks build-snap?13:30
cmatsuokakenvandine: existing material in the internet suggests it should be straightforward, but my tests were not so smooth13:31
kenvandinehey cmatsuoka13:31
kenvandinecmatsuoka: i haven't13:31
mborzeckizyga: which PR is that?13:32
zygaI think it was13:33
zygahttps://github.com/snapcore/snapd/pull/671413:33
cmatsuokakenvandine: I'll check it directly with sitter or the folks in #kde-devel13:33
mupPR #6714: cmd/snap-confine: reject crafted /tmp/snap.$SNAP_NAME <Created by zyga> <https://github.com/snapcore/snapd/pull/6714>13:33
zygamborzecki: no, wait13:33
zygawrong branch13:33
mborzeckizyga: can you check which tests ran before? it looks like /home/test/snap has incorrect type, so it either was created by the upgrade test and not cleaned up, or restorecon did not restore the permission, snapd has all the permissions to poke snappy_home_t but not user_home_t13:33
zygahttps://api.travis-ci.org/v3/job/520237869/log.txt from https://github.com/snapcore/snapd/pull/667313:34
zygachecking13:34
mupPR #6673: cmd,tests: forcibly discard mount namespace when bases change <Created by zyga> <https://github.com/snapcore/snapd/pull/6673>13:34
zygagoogle:fedora-29-64:tests/main/interfaces-broadcom-asic-control google:fedora-29-64:tests/main/snap-connect google:fedora-29-64:tests/main/interfaces-device-buttons google:fedora-29-64:tests/main/debug-sandbox google:fedora-29-64:tests/main/confinement-classic google:fedora-29-64:tests/main/try-with-hooks google:fedora-29-64:tests/main/interfaces-calendar-service google:fedora-29-64:tests/main/refresh:strict_remote13:35
zygagoogle:fedora-29-64:tests/main/install-errors:noreexec google:fedora-29-64:tests/main/snapctl-services google:fedora-29-64:tests/main/nfs-support google:fedora-29-64:tests/main/snap-debug-get-base-declaration google:fedora-29-64:tests/main/snap-service-stop-mode google:fedora-29-64:tests/main/core-snap-not-test-test google:fedora-29-64:tests/main/install-closed-channel google:fedora-29-64:tests/main/snap-disconnect13:35
zygagoogle:fedora-29-64:tests/main/document-portal-activation google:fedora-29-64:tests/main/core18-with-hooks google:fedora-29-64:tests/main/snap-run-symlink-error google:fedora-29-64:tests/main/chattr google:fedora-29-64:tests/main/proxy google:fedora-29-64:tests/main/auto-refresh-private google:fedora-29-64:tests/main/interfaces-netlink-audit google:fedora-29-64:tests/main/auto-aliases13:35
zygagoogle:fedora-29-64:tests/main/command-chain:reexec1 google:fedora-29-64:tests/main/classic-ubuntu-core-transition-two-cores google:fedora-29-64:tests/main/interfaces-content-mkdir-writable:common google:fedora-29-64:tests/main/op-remove google:fedora-29-64:tests/main/snap-service-after-before google:fedora-29-64:tests/main/refresh-delta-from-core google:fedora-29-64:tests/main/interfaces-gpg-public-keys13:35
zygagoogle:fedora-29-64:tests/main/interfaces-hardware-random-observe google:fedora-29-64:tests/main/interfaces-network-control-ip-netns google:fedora-29-64:tests/main/auto-refresh:parallel google:fedora-29-64:tests/main/disable-autoconnect google:fedora-29-64:tests/main/interfaces-ssh-keys google:fedora-29-64:tests/main/snap-userd-desktop-app-autostart google:fedora-29-64:tests/main/install-refresh-private13:35
zygagoogle:fedora-29-64:tests/main/interfaces-network-setup-control google:fedora-29-64:tests/main/install-store-laaaarge google:fedora-29-64:tests/main/econnreset google:fedora-29-64:tests/main/refresh-all google:fedora-29-64:tests/main/interfaces-home google:fedora-29-64:tests/main/umask google:fedora-29-64:tests/main/selinux-lxd + echo '# free space'13:35
zygathis ran before13:35
zygamborzecki: ^13:35
abeatozyga, hey, is there any PR/plans for https://bugs.launchpad.net/snapd/+bug/1821023 ?13:40
mupBug #1821023: core18 base on core 16 missing firmware <snapd:Triaged by zyga> <https://launchpad.net/bugs/1821023>13:40
* zyga lunch13:48
mupPR snapd#6700 closed: packaging: disable -buildmode=pie on all arches <â›” Blocked> <Created by mvo5> <Closed by mvo5> <https://github.com/snapcore/snapd/pull/6700>14:05
mvo6418 needs a second review14:06
mupPR snapd#6594 closed: [RFC] tests: run smoke tests on (almost) pristine systems <Created by mvo5> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/6594>14:10
mvo6599 also needs a second review, should be easy and small14:10
=== ricab|lunch is now known as ricab
zygaabeato: no PRs yet, just in the TODO list14:24
abeatozyga, ok, but still up for 2.39?14:25
zygayes14:26
abeatogreat14:30
mvopedronis: I updated 6603, should be good now14:36
mvo6603 also needs a second review14:37
pedronisChipaca: I suppose you were aware that we potentially import state stuff via auth into snap,  seems the go linker is clever that doesn't happen in practice though14:49
pstolowskimvo: +1 to 6599 with a comment14:55
Chipacapedronis: you mean snap → o.auth -> o.state? yeah14:55
pedronisChipaca: yes14:55
Chipacaalso hexchat is silly and doesn't always do the → -> → thing14:56
Chipacalol14:56
pedronisChipaca: I can probably untangle that now, though as I said the linker seems to do the right thing14:56
pedronisChipaca: do you think untangling that would help overall?14:56
Chipacapedronis: not the first thing i worry about wrt untangling tbh14:57
pedronisChipaca: ok, I'm mentioning it because I'm shuffling things around in that area14:57
pedronisanyway14:57
Chipacaah, if you are then maybe yes14:57
=== Girtablulu is now known as Girtablulu|Away
pedronisbut fully untangling that is a bit of extra work14:57
Chipacapedronis: can you nudge it towards it being easier to do, without it being too much extra work?14:58
mupPR snapd#6718 opened: spread, tests: do not leave mislabeled files in restorecon test, attempt to catch similar files <SELinux> <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6718>14:59
mborzeckizyga: ^^14:59
pedronisChipaca: yes, basically I'm moving AuthContext to a storecontext packages, to fully untangle  some bits of auth would need to move to devicestate14:59
mborzeckizyga: hope the safety checks do not catch any thing else14:59
pedronisChipaca: it might be saner anyway15:00
pedronisto do the last bit15:00
Chipacapedronis: move what to devicestate?15:00
pedronisChipaca: Device SetDevice User etc15:00
pedronisall the functions taking state.State15:01
Chipacahah, and DeviceState15:01
pedronisno15:01
pedronisthat needs to stay in auth15:01
pedronisbasically almost only the structs15:01
pedroniswould stay in auth15:01
pedronisand be shared with store etc15:02
pstolowskiChipaca: are you refactoring any store bits (auth?); asking in the context of autorefresh fix that i'm about to start doing15:02
Chipacapstolowski: pedronis is :-)15:02
Chipacamaybe15:02
pedronisChipaca: I am15:02
pedronissorry15:02
pedronispstolowski: I am, but store is actually the package I'm going to touch the least15:03
Chipacapstolowski: I'm not, I'm breaking things in ifacestate and snapstate15:03
pedronispstolowski: so don't worry15:03
Chipacapstolowski: what I heard pedronis say was "race is on"15:03
pstolowskiChipaca: :)15:03
pedronisif I'm touch store a lot I'm doing it wrong15:04
pstolowskipedronis: we still need to find a good name for snapshots.expiration config option (re https://github.com/snapcore/snapd/pull/6669); also your high-level feedback on the "0" semantics would be appreciated (see my self-review comment in the PR)15:09
mupPR #6669: overlord/corecfg: make expiration of automatic snapshots configurable (4/4) <Created by stolowski> <https://github.com/snapcore/snapd/pull/6669>15:09
mborzeckiChipaca: almost forgot, https://github.com/snapcore/snapd/pull/6688 volume/structure overlap validation15:23
mupPR #6688: gadget: add validation of cross structure overlap and offset writes <Created by bboozzoo> <https://github.com/snapcore/snapd/pull/6688>15:23
pedronispstolowski: I know, we cand land 6662 to start?15:26
pstolowskipedronis: oh yes, doing!15:28
mupPR snapd#6662 closed: overlord/snapstate,snapshotstate: create snapshot on snap removal (3/4) <Created by stolowski> <Merged by stolowski> <https://github.com/snapcore/snapd/pull/6662>15:28
mupPR snapd#6693 closed: cmd: tweak internal tool lookup to accept more possible locations <Created by bboozzoo> <Merged by mvo5> <https://github.com/snapcore/snapd/pull/6693>15:34
* cachio lunch15:34
* zyga resumes work on base snap refreshes15:41
mupPR # closed: core-build#11, core-build#22, core-build#26, core-build#3716:10
mupPR # opened: core-build#11, core-build#22, core-build#26, core-build#3716:11
=== pstolowski is now known as pstolowski|afk
zygabreak16:32
zygaabck16:40
mupPR snapd#6719 opened: many: move auth.AuthContext to store.DeviceAndAuthContext, the implemention to a separate storecontext package <Created by pedronis> <https://github.com/snapcore/snapd/pull/6719>16:41
pedronisChipaca:  ^  it's big but mostly boring16:42
pedronis(doesn't do the last bit discussed because it's already big like this, would be a follow up)16:43
mupBug #1824851 opened: snap-update-ns failure on `snap run lxd.activate` <Snappy:New> <https://launchpad.net/bugs/1824851>16:46
zygahmm16:47
roadmrhey folks, does ubuntu-image support any kind of caching or proxy? I'm iterating on a build and it's super painful to download pi-kernel every single time :/16:51
roadmrI have a squid I can point it to, if supported16:51
mupBug #1824851 changed: snap-update-ns failure on `snap run lxd.activate` <snapd:New for zyga> <https://launchpad.net/bugs/1824851>16:52
zygaroadmr: I don't know but I feel the pain as well17:03
roadmrzyga: looks like it'll take local snaps preferently, so a poor man's cache seems to be just downloading the required .snaps and putting them in PWD17:03
roadmrtesting!17:03
* zyga breaks for an hour17:04
mupPR snapcraft#2534 opened: Add SNAPCRAFT_PROJECT_DIR environment variable <Created by tonysimpson> <https://github.com/snapcore/snapcraft/pull/2534>17:16
pedronismvo: cachio: related to good commit messages, we probably shouldn't merge PR levaving [RFC] in their commit17:22
cachiosure, pedronis: which is the PR?17:23
pedroniscachio: 6594  that mvo merged17:25
cachiopedronis: ahhh, ok, your are right17:27
cachiopedronis: you suggest to revert/edit/merge ?17:27
cachioor it is just for next time?17:28
pedronisfor next time17:30
zygaI could use a coffee17:33
roadmrzyga: so - download the snaps that are in your image, name them snap_revision_arch.snap (usual), then point ubuntu-image to them via --extra-snaps=/path/to/foo.snap (use one --extra-snaps per snap). That does a sort of poor man's cache :)17:38
roadmrhttps://blog.ubuntu.com/2017/07/11/ubuntu-core-making-a-factory-image-with-private-snaps has more deets17:38
zygaroadmr: does the image get used correctly17:38
zygaroadmr: as in, with assertions and such17:38
roadmrzyga: hm I didn't test the built image :/18:53
brlinhttps://forum.snapcraft.io/t/anyone-care-to-review-this-article-on-snapd-admin/1093719:56
mupPR snapd#6720 opened: many: move Device/SetDevice to devicestate, start of making them pluggable in storecontext <Created by pedronis> <https://github.com/snapcore/snapd/pull/6720>20:00
=== Girtablulu|Away is now known as Girtablulu

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