/srv/irclogs.ubuntu.com/2019/06/03/#snappy.txt

mborzeckimorning05:01
mborzeckimup_ needs a little nudge05:31
zygahey06:25
zygamborzecki: mvo is offline06:25
zygamborzecki: he asked to relay to the team06:25
zygamborzecki: over weekend I noticed master is broken06:25
zygamborzecki: ineffectual assignment in shlex.go06:25
zygamborzecki: I'll start around ~10 today, moving downtown to work from a cafeteria, it's super hot in the office today, I have no AC here06:26
mborzeckizyga: hey, there's a PR fixing this up already06:27
zygamborzecki: super, looking06:27
zygamborzecki: let's land it and notify chipaca when he's around06:27
mborzeckizyga: ok06:28
zygamborzecki: can you please review https://github.com/snapcore/snapd/pull/6940 -- I will need it today06:28
zygait's just python compatibility for mountinfo-tool06:28
mborzeckizyga: ack, looking06:29
zygaI didn't commit that because I'm not quite sure where to put it06:29
zygabut I have a makefile there that does06:29
zygablack, flak8, mypy -2 and mypy checks06:29
zyga(and shellcheck on the shell script)06:29
zygaand this is 100% green06:29
mborzeckiheh06:31
mborzeckineed more python linters06:31
mborzeckizyga: that's because of 14.04 right?06:33
zyganot even that06:33
zygait's mainly because we need 2.7 for places like amazon06:33
zygaon the upside, now it's compatible with everything06:33
mborzeckizyga: yeah, but the looks of it :/06:34
zygamborzecki: it's not too bad06:34
zygamborzecki: I actually think that function signatures are nicer when they are on the following line06:34
zygaonly python byte/unicode quirks are ugly IMO06:34
zygaok06:35
zygapacking up and going06:35
mborzeckizyga: let's agree to disagree :)06:35
zygathis office is nice but during summer time it's an oven :)06:35
zygamborzecki: don't get me wrong06:36
zygamborzecki: I'd love not to have to do this :D06:36
zygamborzecki: the details we can have opinions on that diverge06:36
mborzeckizyga: makes me wonder, how people actually get python3 on centos06:37
mborzeckizyga: i see 3.6 in epel :)06:37
zygamaybe EPEL06:37
mborzeckion rh it's probably software collections06:38
mborzeckizyga: so centos has python34 & python36 from epel, amzn2 has python 3.7 directly from their repos :P06:40
zygamborzecki: the tool required 3.6 before this patch06:40
mborzeckizyga: right, so there's a quesion of 14.04 then06:41
zygamborzecki: hmm?06:57
zygamborzecki: it works on 14.04, I think06:58
zygaI think it passed all spread locally06:58
pstolowskimorning07:06
zygagood morning pawel07:07
* zyga commutes to work for a change :)07:07
mborzeckipstolowski: hey07:14
mborzeckilooks like it's going to be quite warm today07:17
mborzeckimvo: hey07:32
zygaHey mbo07:36
zygamvo:07:36
zygaWorking from afar today?07:36
zygaokay, arrived and setting up for work07:45
zygaah --version changed across  python versions07:49
zygadrat :)07:49
mvothanks for merging 6928 mborzecki07:54
mvois mup down?07:54
mvomup_: hello, how are you today?07:54
mborzeckimvo: yup, it is07:54
zygamvo: mup is off today :)07:54
* dot-tobias says hi07:55
mborzeckimvo: probably since friday, iirc there was a split during the day07:55
mvohey dot-tobias07:55
mvomborzecki: I cherry picked the ppc build fixes to 2.39 now (just fyi)07:56
mborzeckimvo: can you cherry pick ineffassign fix too?07:56
mvomborzecki: sure, what pr number is/was that?07:56
mborzeckimvo: it's 8d883ae6607:56
mvota07:56
zygamvo: lets bump delta ref to 2.39.107:57
mvomborzecki: thanks for this one! was it needed because of a new go vet?07:57
mborzeckimvo: recent changes in ineffassign07:57
* mvo nods07:58
mborzeckimvo: this commit https://github.com/gordonklaus/ineffassign/commit/8863c8f9d746c0e28d65c1bc46c8d3aa8ab1ee82 :)07:58
mvomborzecki: nice! and nice^2 for tracking the exact change down07:59
pstolowskimvo: i've made the measurement before- and after- maxprocs patch, see comments below https://github.com/snapcore/snapd/pull/6933 ; unfortunately no improvement07:59
pstolowskimvo: and hi btw :)07:59
mvopstolowski: goooood morning :)07:59
mvopstolowski: you rock!08:00
mvopstolowski: thank you so much, let me look at the data08:00
mvopstolowski: is there a tl;dr ?08:00
mvopstolowski: aha, looks like zero effect ?08:00
pstolowskimvo: yep, random fluctuations, nothing more08:01
mborzeckimvo: pstolowski: wonder if there'd be an observable change in reaching seeded target08:01
mvopstolowski: a shame - oh well08:02
mvopstolowski: what cmdline did you use to limit the cpu?08:02
pstolowskimvo: setenv snappy_cmdline net.ifnames=0 init=/lib/systemd/systemd ro panic=-1 fixrtc maxcpus=108:04
pstolowskimvo: followed by saveenv08:04
pstolowskimvo: verified with lscpu afterwards08:04
mvopstolowski: cool, thanks!08:08
zygamvo: https://github.com/snapcore/snapd/pull/6937/files08:09
* Chipaca goes for more coffee08:24
pstolowskimvo: https://github.com/snapcore/snapd/pull/6925 can land?10:53
* zyga breaks for lunch10:56
zygaSetting in11:37
cmatsuokamvo, pedronis, zyga: spike successfully installs using writable on tmpfs11:47
zygacmatsuoka: hey11:49
zygacmatsuoka: that's cool :)11:49
zygacmatsuoka: I'm progressing  towards go-only initrd that can hand off to existing stack11:50
* cmatsuoka notices he can't read IRC and code at the same time, he just typed "build_zyga" as a function name11:50
zygahahaha11:51
zygamake -j 211:51
cmatsuokazyga: nice! I ported writable preparation steps from shell script to go for the spike and sometimes it's a bit annoying11:52
cmatsuokaI just exec.Command().Run() in some places just to keep things going11:52
cmatsuokawith proper infrastructure it will be much nicer11:53
zyga+111:54
zygathat's good for the spike IMO11:54
cmatsuokathere are a number of hacks there that must be solved in a better way, such as console-conf telling snapd that it's done11:58
cmatsuokaand when I request a reboot it takes a really long time to reboot11:59
mborzeckioff to pick up the kids12:05
=== ricab is now known as ricab|lunch
mvocmatsuoka: awesome!12:27
ackkhi, I'm trying to get chrony to work in a snap, I added the time-control interface, but getting a permission denied from adjtimex13:15
ackkI'm getting no denials in logs13:18
=== ricab|lunch is now known as ricab
cmatsuokamvo: we were previously injecting stuff into kernel initramfs, I'm now injecting into core and adding a custom snapd from my writable-ramdisk branch13:38
cmatsuokaalso core-build from writable-ramdisk branch13:39
mvocmatsuoka: so core-build is the branch I need to look at?13:40
cmatsuokayou need the writable-ramdisk branch in spike-tools, core-build and snapd13:41
cmatsuokaall of them are in github.com/cmatsuoka13:42
* mvo looks13:44
cmatsuokamvo: hmm, and I think I broke something because I can't log into the installed system anymore13:48
* cmatsuoka investigates13:49
mvook13:50
pstolowski#6935 needs 2nd review13:53
ackkmvo, hi, any idea off the top of your head for the issue above?13:53
Chipacapstolowski: if you have a chance, can you give #6905 a look to see if your LGTM still stands? (and make it an actual +1)13:54
pstolowskiChipaca: ok13:54
mvoackk: just read it, let me think for a minute13:54
* zyga goes to find some shade13:55
mvoackk: it sounds like its a missing syscall in the interface, let me look at the code13:55
ackkmvo, thanks13:56
ackkmvo, I have a simple snap which shows the issue, if it could be useful13:58
mvoackk: see #694314:00
ackkmvo, that doesn't seem like an LP bug# :)14:01
ackkoh you mean https://github.com/snapcore/snapd/pull/694314:02
ackkmvo, thanks!14:02
mvoackk: yes, sorry, usually mup expands this but it seems to be off today :P14:04
ackkheh14:04
mvoackk: lets see if this is enough, if you need more just let me know, hopefully will land in edge soon14:04
ackkmvo, how do I get that ? via a PPA?14:04
ackk(once it lands of course)14:05
mvoackk: snap install --edge core (or snapd if you use that already) should be enough14:06
mvoackk: once you are happy you can always "snap refresh --beta core" (or snapd) (or candidate/stable) to get off edge of course14:06
ackkmvo, cool, thanks14:08
Chipacapstolowski: does my review make sense?14:29
cmatsuokamvo: during the spike it could be good to add systemd.debug-shell=1 to the gadget grubenv as well14:30
pstolowskiChipaca: absolutely, thank you!14:30
pedronisChipaca: thx for looking at that14:32
mvocmatsuoka: yeah, I think so15:01
* cachio lunch15:42
=== pstolowski is now known as pstolowski|afk
zygaback in the office16:37
zygasorry for the lag, had dinner too16:37
cmatsuokamvo: I'm fixing seed installation on the real writable17:07
=== Greyztar- is now known as Greyztar
jdstrandzyga: hey, do you know who was working on '(cannot find required base "core")'? I thought that was supposed to be fixed in 2.39, but I'm seeing it18:14
* jdstrand has core installed18:14
zygajdstrand: I think eventually that was mvo18:14
zygahow about .1?18:14
zygaoh18:14
zygahmmm18:14
jdstrandsnap version says 2.3918:15
jdstrand$ snap list core18:15
jdstrandName  Version  Rev   Tracking  Publisher   Notes18:15
jdstrandcore  16-2.39  6964  stable    canonical✓  core18:15
jdstrand$ sudo snap refresh review-tools18:15
jdstranderror: cannot perform the following tasks:18:15
jdstrand- Mount snap "review-tools" (857) (cannot find required base "core")18:15
jdstrandhuh18:17
jdstrandso, 847 doesn't have 'base: core', but 857 does have 'base: core'18:17
jdstrandhmmm, remove and install, same thing18:23
zygajdstrand: hmmm18:27
zygajdstrand: can you report it please, I'll check it out tomorrowm orning18:27
jdstrandzyga (cc pedronis): https://forum.snapcraft.io/t/cannot-find-required-base-core-when-installing-refreshing-when-using-base-core/1164118:44
jdstrandzyga: fyi, I refreshed to --beta (2.39.1) and it was the same thing18:45
jdstrandsergiusens[m]: fyi ^ (not sure if you have anything to add to that, so just fyi)18:46
jdstrandsergiusens[m]: hi btw :)18:47
sergiusenshi jdstrand18:54
sergiusensjdstrand: answered on the post18:55
jdstrandsergiusens: thanks! trying that. is that true for snaps that use 'base: core18' too?19:01
=== valentind_ is now known as valentind
jdstrandzyga: fyi, nm, sergiusens fixed it for me (see forum)19:20
sergiusensjdstrand: yes, it is true for core18 as well, let me see if I can find the launchpad bug for this19:35
jdstrandsergiusens: interesting. I have some core18 snaps and didn't see this before19:37
* jdstrand fixes those19:37
ijohnson_hey jdstrand do you think we can merge https://github.com/snapcore/snapd/pull/6805 or do we need to fix that erroneous test?20:09
jdstrandijohnson_: I'm not comfortable doing that. I recommend clicking through to the failing job and doing 'Restart job' until it passes20:15
jdstrandijohnson_: if that doesn't work after a few times, ping mvo or cachio or someone for assistance (they may have insight on the failing test that I don't)20:16
ijohnson_okay, well I don't have permission to restart the jobs, do you have permissions to restart the tests on that PR?20:21
ijohnson_jdstrand: ^20:21
jdstrandI do20:21
jdstrandlet me do it20:21
jdstrandok, done. I'll keep an eye on it20:22
ijohnson_ack, thanks20:22
=== Greyztar- is now known as Greyztar
* cachio afk21:01
ijohnson_hey jdstrand it looks like those tests passed, do you still want another review before merging?21:06
jdstrandijohnson_: there need to be two approvals21:07
jdstrandneeds*21:07
ijohnson_ah okay wasn't aware of that, I will ping someone else tomorrow morning21:07
ijohnson_thanks!21:07
Pharaoh_Atemzyga, looks like we have a problem... https://bugzilla.redhat.com/show_bug.cgi?id=171639721:27

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