| lool | == Publishing unity-mir (unity8 stack) == | 00:01 |
|---|---|---|
| lool | Mir uploads done, except for the switch | 00:01 |
| lool | and didn't seem to regress SF too much | 00:01 |
| asac | lool: so what will happen you think :)? | 00:11 |
| asac | lool: the default switch ... will that happen today> otherwise we can just kick an image | 00:11 |
| asac | we wills ee the results at least | 00:11 |
| asac | and have super easy instructions how to test it | 00:12 |
| rsalveti | asac: lool: just got mir running on my nexus 7 | 00:12 |
| rsalveti | a bit slow, due the lack of hwcomposer, but working at least | 00:12 |
| lool | rsalveti: no kidding? | 00:12 |
| lool | rsalveti: so should I list it? :-) | 00:13 |
| rsalveti | lool: yup :-) | 00:13 |
| lool | asac: it will happen | 00:13 |
| rsalveti | cyphermox: 580 1002 20 0 1192 256 176 R 99.6 0.0 2:44.92 brcm_patchram_p | 00:13 |
| lool | asac: waiting for builds and tweaking the switch to just change some devices | 00:13 |
| rsalveti | cyphermox: consuming 99% of my cpu | 00:13 |
| cyphermox | I know | 00:13 |
| rsalveti | cyphermox: know why? | 00:15 |
| rsalveti | lool: changes at android + lxc-android-config | 00:15 |
| lool | rsalveti: ah so should I enable it now or later? | 00:17 |
| rsalveti | lool: can land tomorrow | 00:18 |
| lool | rsalveti: mind checking lp:~lool/ubuntu/saucy/lxc-android-config/set-display-mir ? | 00:18 |
| rsalveti | let me just test it a bit more | 00:18 |
| lool | pushed a rev there | 00:18 |
| cyphermox | rsalveti: nothing more than the fact that someone broke my precious with some random change :) | 00:18 |
| rsalveti | cyphermox: hm, ok :-) | 00:18 |
| lool | rsalveti: https://bazaar.launchpad.net/~lool/ubuntu/saucy/lxc-android-config/set-display-mir/revision/111 | 00:18 |
| cyphermox | rsalveti: looks like it's no longer starting in the right time, perhaps | 00:19 |
| cyphermox | rsalveti: I'll tackle this as soon as I know for sure what to do with NM | 00:19 |
| lool | rsalveti: I'll leave out grouper, but that means I'll need another flag for grouper once it's emabled | 00:20 |
| rsalveti | cyphermox: ok, cool | 00:22 |
| rsalveti | lool: yeah, better to leave out grouper for now | 00:23 |
| asac | lool: when do you plan to go off :/ | 00:58 |
| asac | ? | 00:58 |
| lool | asac: All uploaded now, waiting for transition to realese pocket | 00:59 |
| lool | asac: was waiting for an upstart-app-launch review because it's seeded | 00:59 |
| asac | lool: how confident are we that the image is not completely busted? :) | 00:59 |
| lool | asac: pretty confident | 01:00 |
| lool | but I also think the tests will fail | 01:00 |
| lool | and I dont think I'll fix this tonight | 01:00 |
| asac | is there a pretty confident at 3am :)? | 01:00 |
| lool | too many round trips on other things | 01:01 |
| asac | what is missing for the tests? | 01:01 |
| lool | asac: I've been running the packages for hours here | 01:01 |
| lool | I tested and reviewed some music-app changes | 01:01 |
| lool | played with notes-app | 01:01 |
| lool | installed clicks | 01:01 |
| lool | etc. | 01:01 |
| asac | lool: so we missed something for the tests? or just notes-app? | 01:04 |
| lool | to improve the tests? | 01:05 |
| asac | lool: nevermind. sounded like you said that something unexpected with the tests | 01:07 |
| asac | all we need to do is get an image that ollis team can pick up to fix | 01:07 |
| lool | asac: Yup | 01:08 |
| lool | asac: I think I'll cron it in one hour | 01:08 |
| asac | lool: at best with nothing changed beyond the mir stuff | 01:08 |
| asac | :) | 01:08 |
| lool | I'd rather see the britney pass first though | 01:08 |
| lool | asac: Ok, so final: libquvi,lxc-android-config,upstart-app-launch,xterm/arm64 | 01:13 |
| lool | so this is going into release pocket in next run | 01:13 |
| lool | I'm croning a build in 45mn or so | 01:14 |
| lool | done | 01:14 |
| plars | lool, asac: back - and yes I'm aware of the notes-app issues. sergiusens and I discussed farther up | 01:25 |
| lool | cool | 01:25 |
| lool | plars: tests passed for me there now | 01:25 |
| lool | got to go to bed now, sent email out | 01:25 |
| plars | lool: great | 01:26 |
| lool | ah it's in, can kick image early | 01:26 |
| lool | == Building #88 == | 01:27 |
| lool | 'night | 01:35 |
| asac | lool: night!! | 01:39 |
| asac | :() | 01:39 |
| lool | asac: what? | 01:42 |
| * lool really off now | 01:43 | |
| tyhicks | hello! | 02:13 |
| tyhicks | yesterday's dbus upload introduced a regression that causes dbus-daemon to segfault when apparmor mediation is disabled, or unsupported by the kernel, and an application attempts to eavesdrop (LP: #1237059) | 02:13 |
| ubot5 | Launchpad bug 1237059 in dbus (Ubuntu) "dbus-daemon crashed with SIGSEGV" [Critical,Confirmed] https://launchpad.net/bugs/1237059 | 02:13 |
| tyhicks | I've came up with a fix, tested it, and attached the debdiff (which I'll need sponsorship for) to the bug | 02:14 |
| tyhicks | the fix is small and the bug pretty painful for folks running custom kernels so I think it needs to get processed fairly quickly | 02:15 |
| tyhicks | what is my next step from a landing POV? | 02:15 |
| === plars_ is now known as plars | ||
| plars | asac, lool, sergiusens: something still wrong with notes-app: AttributeError: Class 'NotesApp' has no attribute 'get_notes'. | 03:30 |
| plars | http://10.97.0.1:8080/job/saucy-touch_ro-mako-smoke-notes-app-autopilot/157/console is the one in progress with the latest image if you want to see | 03:30 |
| plars | asac, lool: I think there's something wrong with mir enablement on image88, I have a .set-display-mir-done now in /home/phablet (fresh install) but ps still shows surfaceflinger running | 03:50 |
| Mirv | morning | 03:51 |
| robert_ancell | is autolanding turned off? | 04:20 |
| veebers | robert_ancell: I think so, I see an email to ubuntu-phone ML stating that upstream merger is temp. turned off | 04:22 |
| robert_ancell | ah, I see the email from olli | 04:23 |
| rsalveti | lool: will push the grouper specific stuff directly, only affects grouper, and low risk | 04:44 |
| rsalveti | so we can also enable mir on it | 04:44 |
| jibel | Good morning | 05:05 |
| rsalveti | lool: once 20131006-1510-0ubuntu2 migrates to release, just trigger a new image and you should get one which is compatible with mir for nexus 7 | 06:17 |
| rsalveti | there's an issue still when opening apps and moving back to the shell, but this will make it work at least | 06:17 |
| === jasoncwarner_ is now known as jasoncwarner | ||
| sergiusens | plars, it's unmerged | 07:13 |
| ogra_ | lool, hmm ... "start on boot-hooks and starting lxc-android-config" ... doesnt that mean lxc-android-config will be held until we are done ? if so that wont work, without having the container up your getprop call wont work | 07:19 |
| sergiusens | ogra_, hmm, that was added after my review :-/ | 07:35 |
| Mirv | sil2100: hi. are you testing some cu2d stacks before the meeting? I think settings might be one. I just finished with address book. | 07:37 |
| sil2100 | Mirv: my device just finished upgrading - any particular stack you think I could take care of, or it doesn't matter? | 07:39 |
| sil2100 | Mirv: of course, with Mir enabled everything, yes? | 07:39 |
| sil2100 | Mirv: I can test settings if needed, since I anyway work on that code-wise | 07:42 |
| Mirv | sil2100: see the e-mail, it's up to us I guess to ponder what would be useful to have pre-tested. I was just looking at the 'results' page. | 07:44 |
| Mirv | I'm unsure about apps, since I don't know how the click package updates go, plus there is already an item about notes-app in the landing plan. | 07:45 |
| sergiusens | Mirv, if this gets triggered https://code.launchpad.net/~sergiusens/notes-app/click-fixes/+merge/189980 the notes app would work again | 07:46 |
| sil2100 | Mirv: but indeed, I also upgraded and have surfaceflinger by default | 08:00 |
| sil2100 | Mirv: anyway, on my device I'll test both settings packages to check if those are ok, then move on to other stacks | 08:02 |
| ogra_ | lool, sergiusens, http://paste.ubuntu.com/6212771/ ... that would be my recommendation | 08:03 |
| lool | morning | 08:05 |
| lool | ogra_: oh crap | 08:06 |
| lool | ogra_: we need it before lxc-android-config | 08:06 |
| lool | ogra_: so does it break boot? | 08:06 |
| ogra_ | dunno, dashboard looks good and i cant start a download now else i cant use hangouts | 08:07 |
| ogra_ | but i doubt the file will be touched | 08:07 |
| lool | I'll tell you in a sec, I'm upgrading to it | 08:07 |
| lool | it shouldn't stop boot unless getprop is blocking | 08:08 |
| ogra_ | i dont think it will break anything, but if you rely on gerprop before it is available it will likely be a no op | 08:08 |
| ogra_ | *get | 08:08 |
| lool | sergiusens: Yeah, I remembered we didn't want to enable mir outside of mako/maguro | 08:08 |
| lool | sergiusens: but didn't think getprop would need lxc-android-config | 08:09 |
| lool | it does start | 08:09 |
| lool | ogra_: it seems to work | 08:09 |
| lool | ogra_: no idea why ;-) | 08:09 |
| lool | at least no error output in the job | 08:10 |
| lool | let me print this | 08:10 |
| sergiusens | lool, getprop is an android thing talking over hybris | 08:10 |
| lool | sergiusens: problem is I need to start before lxc-android-config which starts SF | 08:11 |
| lool | oh crap | 08:11 |
| lool | Nexus 4 | 08:11 |
| lool | is what I'm getting | 08:11 |
| lool | this is utterly wrong | 08:11 |
| ogra_ | yeah | 08:12 |
| ogra_ | lool, ... device=$(grep ^ro.product.device= /system/build.prop |sed -e 's/.*=//') | 08:13 |
| ogra_ | you can use that if there are no properties | 08:13 |
| lool | ogra_, sergiusens: Oddly, properties seem to work | 08:13 |
| ogra_ | before the container is up ? | 08:14 |
| ogra_ | thats weird | 08:14 |
| ogra_ | unless the container comes up even though you tell upstart to block it | 08:14 |
| lool | doubt it | 08:14 |
| lool | what's the correct one? ro.build.product or ro.product.device? | 08:15 |
| apw | ogra_, could you test the kernel in the ckt PPA to confirm it fixes your readahead issue | 08:15 |
| lool | I guess device | 08:15 |
| ogra_ | apw, will do after the landing team call | 08:15 |
| apw | ogra_, thanks | 08:16 |
| sil2100 | Mirv: btw. is Mir also so 'jitterish' and lagish on your device? | 08:16 |
| sil2100 | And it's utterly crash-prone... | 08:19 |
| sergiusens | lool, I don't get the automerger thing being off to fix tests, does that mean I should manually merge? | 08:22 |
| lool | sergiusens: Yeah | 08:22 |
| sergiusens | lool, ok, so manual merging is allowed? | 08:22 |
| lool | sergiusens: if it fixes tests! ;-) | 08:22 |
| sergiusens | lool, yes, I never merged the notes changes since I wasn't sure what that meant since we are only supposed to merge with jenkins too ;-) | 08:23 |
| lool | plars: Yes, I got it wrong | 08:23 |
| Mirv | sil2100: yeah it's somewhat slow, there are a couple of bugs open like eg. power management slowing down it too much unlike under surfaceflinger etc | 08:24 |
| sil2100 | lool, Mirv: also, I can't completely test the settings at least under Mir, as it is not possible to change the background for instance when Mir is used - once you select and image, the display/shell hangs up and nothing can be done | 08:25 |
| Mirv | sil2100: can you shed any light on the jibel's lxc tinkering? or most of all, do you understand who has actually built which part of the autopilot machines? I don't basically have the understanding on what is done by the qa team, ci team or just didrocks | 08:25 |
| sergiusens | sil2100, there's an open bug for that | 08:25 |
| Mirv | sil2100: so my e-mails are therefore also a bit shot in the dark | 08:25 |
| sil2100 | Mirv: as for the lxc things - I'm not entirely sure myself, but my understanding was that the autopilot otto machines that we use for daily-release were prepared by jibel + didrocks | 08:27 |
| sil2100 | Mirv: I was not even aware of a #ci channel in the first place | 08:27 |
| jibel | sil2100, Mirv I prepared the machines originally but since the split CI/QA, the CI team is doing maintenance and monitoring of this part of the infrastructure | 08:28 |
| ogra_ | root@ubuntu-phablet:/# ls -l /home/phablet/.display-mir | 08:28 |
| ogra_ | ls: cannot access /home/phablet/.display-mir: No such file or directory | 08:28 |
| ogra_ | root@ubuntu-phablet:/# | 08:28 |
| ogra_ | lool, ^^^ | 08:28 |
| sil2100 | jibel: ACK, it's good to have it all cleared out now | 08:29 |
| Mirv | ogra_: yep, see ubuntu-phone | 08:29 |
| lool | sergiusens: https://code.launchpad.net/~lool/ubuntu/saucy/lxc-android-config/set-display-mir2/+merge/190042 | 08:29 |
| ogra_ | are you sure you didnt have the file before already ? | 08:29 |
| ogra_ | Mirv, right | 08:29 |
| Mirv | jibel: thanks for the clarification. | 08:29 |
| ogra_ | Mirv, lool just tested it and for him it worked (on mako i guess, this was maguro, upgraded from 84) | 08:29 |
| * sil2100 didn't have Mir enabled by default | 08:30 | |
| lool | asac: joining? | 08:31 |
| lool | sil2100: joining? | 08:31 |
| ogra_ | root@ubuntu-phablet:/# grep foo /etc/init/boot-hooks/set-display-mir.conf | 08:38 |
| ogra_ | echo "$(getprop ro.product.model)" >/tmp/foo | 08:38 |
| ogra_ | root@ubuntu-phablet:/# cat /tmp/foo | 08:38 |
| ogra_ | Galaxy Nexus | 08:38 |
| ogra_ | lool, ^^^ ... right, it works (it still shouldnt though) | 08:38 |
| sergiusens | lool, https://code.launchpad.net/~sergiusens/ubuntu-clock-app/confinement_fixes/+merge/190044 | 08:40 |
| asac | seb128: can we drop the goa party from our touch images? | 08:56 |
| asac | http://people.canonical.com/~ogra/touch-image-stats/20131009.changes | 08:56 |
| asac | :-P | 08:56 |
| seb128 | asac, your call, it might makes e-d-s calendar not work correctly | 08:56 |
| asac | seb128: was told that the goa upload was supposed to not change the touch image | 08:57 |
| asac | behaviour | 08:57 |
| seb128 | asac, it's not, but I didn't try to pull the binary out | 08:57 |
| asac | ok | 08:57 |
| seb128 | asac, we replaced the buggy uoa code by putting back goa that we had before | 08:57 |
| seb128 | we never tried to have none of those | 08:57 |
| seb128 | (goa used to be in e-d-s itself and not binary split out) | 08:58 |
| Laney | one package is a party? | 08:59 |
| Laney | I expected to see millions of deps from that complaint | 08:59 |
| Mirv | lool: so does no landings mean also no qtdeclarative update at this point (I added it to landing asks, tsdgeos & co would like a fix in) | 09:13 |
| Mirv | I guess that's like asking 'does no mean no' | 09:13 |
| lool | Mirv: Correct | 09:13 |
| lool | Mirv: no landings except fixing tests | 09:13 |
| lool | and bugs | 09:13 |
| lool | crashes | 09:13 |
| Mirv | yeah, so I'd need to get them to show that it actually fixes some current bug/crash, while it currently fixes an inifnite loop in unity8 but the unity8 change itself was backed up | 09:14 |
| === asac changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: - Type: "@ci" for help | Tree is "closed" to fix MIR-by-default | Known issues: unity8 crashes lp:1228097 unblocked for now, some progress, revisit wed | ||
| Mirv | so, if someone knows an apport usage method that works on device, please tell. I'm running into various problems, have tried running as both phablet and root, apport-retrace is installed, have tried 'examine locally' etc. | 09:20 |
| Mirv | I'm getting eg. No such file or directory: u'/home/phablet/.cache/apport/retrace/Ubuntu 13.10/saucy-Contents-armhf.gz' | 09:21 |
| Mirv | sil2100: tell me about your experiences as well if you're looking at it | 09:21 |
| Mirv | I've started with apport-cli [hud-crash-file.crash] | 09:21 |
| jibel | Mirv, connect your device to a network and run mkdir -p "/home/phablet/.cache/apport/retrace/Ubuntu 13.10/" | 09:23 |
| jibel | Mirv, but pitti fixed that a couple of weeks ago | 09:24 |
| jibel | Mirv, otherwise apport-cli <crash> should work just fine, as long as you're device has web access | 09:26 |
| jibel | your | 09:26 |
| lool | == Building #89 == | 09:44 |
| Mirv | jibel: thanks! | 09:45 |
| lool | I tested the actual hook | 09:45 |
| lool | but will confirm once image is out | 09:46 |
| * lool school & | 09:46 | |
| === alan_g is now known as alan_g|tea | ||
| === alan_g|tea is now known as alan_g | ||
| Mirv | I was able to file a bug on the hud crash via apport but not getting a good retrace https://launchpad.net/bugs/1237322 | 10:16 |
| ubot5 | Ubuntu bug 1233992 in hud (Ubuntu) "duplicate for #1237322 HUD does not support unity8's ApplicationManger" [High,In progress] | 10:16 |
| Mirv | well LP now made it duplicate anyhow | 10:16 |
| Mirv | that mkdir was the key thing to get apport working | 10:17 |
| lool | Images are importing, but should be done shortly | 10:43 |
| lool | could someone test that Mir is default in #89, as intended? | 10:43 |
| * ogra_ is just flashing 69 now for slowness testing | 10:45 | |
| lool | Image is up | 10:51 |
| popey | i can | 10:51 |
| lool | jibel: Mind checking Mir turns on with it | 10:51 |
| lool | popey: ah thanks | 10:51 |
| popey | what's the canonical test to see if mir is enabled by default | 10:52 |
| popey | ? | 10:52 |
| popey | (look for flickering? ㋛ | 10:52 |
| popey | ) | 10:52 |
| lool | == Publishing hud (hud stack) == | 10:52 |
| lool | to get the dbus fix | 10:52 |
| lool | or workaround rather | 10:52 |
| lool | popey: ps fauxww|grep flinger | 10:52 |
| lool | popey: run that now to check perhaps | 10:53 |
| popey | k | 10:53 |
| ogra_ | geez fauxww ? | 10:53 |
| ogra_ | ps ax|grep surf | 10:53 |
| popey | holy cow, 152 desktop updates since yesterday | 10:54 |
| popey | yet, the desktop is being ignored ㋛ | 10:54 |
| lool | well | 10:55 |
| lool | ogra_: one developer, one way to run ps | 10:55 |
| ogra_ | haha, yeah | 10:55 |
| apw | popey, 'de-emphasised' please | 10:56 |
| popey | ah yes. | 10:58 |
| popey | lool: yes, latest proposed image has mir | 11:02 |
| ogra_ | popey, OTA or fresh flash ? | 11:03 |
| popey | fresh | 11:03 |
| ogra_ | (OTA was broken) | 11:03 |
| popey | alan@deep-thought:~$ phablet-flash ubuntu-system --channel saucy-proposed -d mako | 11:03 |
| popey | i did that | 11:03 |
| popey | should have --no-backup i guess | 11:03 |
| ogra_ | yeah, that should have worked in the former one already | 11:03 |
| popey | but I have never enabled mir, so i dont have that file | 11:03 |
| ogra_ | hmm, though --no-backup might indeed be difefrent | 11:04 |
| lool | popey: cool | 11:21 |
| lool | ogra_: OTA was broken? | 11:21 |
| lool | ogra_: you mean in #88? | 11:22 |
| ogra_ | lool, yes | 11:23 |
| asac | lool: any news on webbrowser app and hud service? | 11:29 |
| lool | asac: hud is in unapproved with the dbus fix | 11:30 |
| lool | s/fix/workaround | 11:30 |
| lool | I've poked #ubuntu-release | 11:31 |
| lool | it's seeded on desktop | 11:31 |
| asac | so do we know that 89 fixes things? | 11:43 |
| ogra_ | http://people.canonical.com/~ogra/touch-image-stats/20131009.1.changes | 11:46 |
| ogra_ | what would you expect this to fix ? | 11:46 |
| ogra_ | some address bok stuff perhaps | 11:46 |
| asac | lool: did you check with Saviq on the webbrowser tests? | 11:47 |
| asac | ogra_: that mir starts by default | 11:47 |
| ogra_ | asac, thats verified | 11:47 |
| asac | ogra_: is it? | 11:48 |
| asac | :) | 11:48 |
| ogra_ | works fine | 11:48 |
| asac | like yesterday | 11:48 |
| asac | hehe | 11:48 |
| ogra_ | yes | 11:48 |
| asac | ogra_: it didnt work in #88 :) | 11:48 |
| asac | just double checking that we validated that it really works with 89 | 11:48 |
| ogra_ | asac, thats why 89 has a fix ;) | 11:48 |
| asac | because i just sent a mail telling folks that its fixed there :) | 11:48 |
| asac | thanks | 11:48 |
| ogra_ | it worked for popey, i trust his testing | 11:48 |
| popey | lunatic | 11:49 |
| popey | what works? | 11:49 |
| ogra_ | popey, Mir by default | 11:49 |
| popey | ah, k | 11:49 |
| * ogra_ does a maguro update too | 11:49 | |
| popey | http://paste.ubuntu.com/6213380/ | 11:50 |
| ogra_ | yeah | 11:50 |
| asac | yes, popey has been a flawless machine so far | 11:50 |
| === alan_g is now known as alan_g|lunch | ||
| lool | robru: I've now rejected this very old webapp landing you were on to clear the spreadsheet | 12:09 |
| asac | doanac: do you know if AP tests do something like stopping/restarting unity after they get run? | 12:16 |
| lool | so there are a bunch of staged mir + unity-mir + other changes staged from overmight | 12:21 |
| ogra_ | lool, theer should also be a powerd fix for the "Mir is slow after resume, we need the performance governor" issue | 12:21 |
| ogra_ | <8was a two line fix in powerd btw) | 12:22 |
| psivaa | lool: i have a maguro with 89, do you want met to test anything from ppa? | 12:22 |
| jibel | asac, lool I think bug 1236525 is release critical, I got it on 1rst boot after flashing a mako with build 89 | 12:22 |
| ubot5 | bug 1236525 in unity-mir "unity8 killed/crash then restart can result in mir unable "could not unblank display"" [High,Triaged] https://launchpad.net/bugs/1236525 | 12:22 |
| psivaa | s/met/me | 12:22 |
| jibel | which means no display and adb reboot | 12:23 |
| asac | jibel: did you get a crash file? | 12:24 |
| jibel | asac, only _usr_lib_arm-linux-gnueabihf_upstart-app-launch_desktop-hook.32011.crash | 12:25 |
| lool | ogra_: whereis it? | 12:25 |
| lool | ogra_: I dont see it in trunk | 12:25 |
| lool | psivaa: hud will be coming up | 12:25 |
| lool | psivaa: it's uploaded now, but building | 12:25 |
| psivaa | lool: ack | 12:26 |
| asac | jibel: does it have any useful trace? | 12:26 |
| lool | so for some reason, recent hud changes didn't get picked up by cu2d | 12:26 |
| lool | now building and will get them | 12:26 |
| asac | e.g. backtrace from coredump | 12:26 |
| ogra_ | lool, gimme a sec, i'll dig up the MP ... i guess with upstream merger on manual ricmm_ didnt bother to try to merge | 12:27 |
| ogra_ | lool, https://code.launchpad.net/~ricmm/powerd/use-hal-power-hint/+merge/190027 | 12:27 |
| ogra_ | its actually more then two lines | 12:28 |
| jibel | asac, nope, just 'couldn't unblank display' error in unity8.log http://paste.ubuntu.com/6213507/ | 12:28 |
| ogra_ | but should make waking up a lot faster | 12:28 |
| jibel | *errors even | 12:28 |
| lool | ogra_: it failed to build | 12:28 |
| ogra_ | lool, then lets wait for ricmm_ | 12:29 |
| ogra_ | but it seems to be on a good path | 12:29 |
| ogra_ | (i.e. we should land this today if it gets ready) | 12:29 |
| asac | land what? | 12:30 |
| asac | hud? | 12:30 |
| asac | yes, if it fixes a nice crasher for us | 12:30 |
| asac | at best two :) | 12:30 |
| ogra_ | asac, the fix ChickenCutlass mentioned in the other channel | 12:30 |
| ogra_ | asac, https://code.launchpad.net/~ricmm/powerd/use-hal-power-hint/+merge/190027 | 12:30 |
| asac | its not even done yet | 12:30 |
| asac | so... | 12:30 |
| asac | we wait | 12:30 |
| asac | as planned | 12:30 |
| ogra_ | right, i'm just happy we wont need "performance" | 12:31 |
| asac | thats indeed good news | 12:31 |
| lool | psivaa: https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/5087920 done in 5mn or so | 12:35 |
| dbarth | @ci | 12:37 |
| cjohnston | dbarth: what's up | 12:38 |
| asac | Mirv: sil2100: lool: did we get any useful backtraces? | 12:38 |
| psivaa | lool: thanks, will test that | 12:38 |
| lool | cjwatson: hey, there is a new libmirserver6 package queued up in daily PPA; do you need to do some archive admin things before I try to push it later today? | 12:39 |
| cjwatson | Don't think so | 12:39 |
| lool | ok | 12:39 |
| cjwatson | Well, it's possible we might need to manually move to main (depending on exactly which bit of https://bugs.launchpad.net/launchpad/+bug/993120 is tickled), we'll see | 12:40 |
| ubot5 | Ubuntu bug 993120 in Launchpad itself "Copy from PPA with binaries evades NEW and puts new packages into universe" [High,Triaged] | 12:40 |
| cjwatson | But nothing I can do in advance | 12:40 |
| cjwatson | (For clarity, that's evades *binary* NEW; I know we had a previous related conversation but that was about source NEW) | 12:40 |
| cjwatson | Also https://bugs.launchpad.net/launchpad/+bug/1079577 | 12:41 |
| ubot5 | Ubuntu bug 1079577 in Launchpad itself "Copies with binaries don't correctly override NEW binaries" [High,Triaged] | 12:41 |
| lool | cjwatson: ok | 12:41 |
| lool | psivaa: hud is up | 12:42 |
| Mirv | asac: so I didn't get a good backtrace of the hud one, but LP still made my bug report a duplicate of this https://bugs.launchpad.net/ubuntu/+source/hud/+bug/1233992 | 12:42 |
| ubot5 | Ubuntu bug 1233992 in hud (Ubuntu) "HUD does not support unity8's ApplicationManger" [High,In progress] | 12:42 |
| Mirv | and maliit crash was this one https://bugs.launchpad.net/mir/+bug/1233988 | 12:42 |
| ubot5 | Ubuntu bug 1233988 in Mir "With Mir enabled: maliit-server crashed with SIGABRT in __gnu_cxx::__verbose_terminate_handler(), thrown from mir::client::DisplayConfiguration::copy_to_client()" [Medium,Triaged] | 12:42 |
| lool | Mirv, psivaa: Still getting a hud crash with latest hud | 12:45 |
| lool | -rw-r----- 1 phablet whoopsie 1779344 Oct 9 12:44 _usr_lib_arm-linux-gnueabihf_hud_hud-service.32011.crash | 12:45 |
| cjohnston | dbarth: did you need assistance? | 12:45 |
| asac | Mirv: ok... seems like a garbage dupe :) | 12:45 |
| asac | Mirv: you could install the debug packages on the phone and run the AP that triggers the crash | 12:45 |
| asac | Mirv: that has a higher likelyhood to get you something useful - i would think | 12:45 |
| ogra_ | woah | 12:46 |
| ogra_ | phablet 20 0 28812 24m 1136 R 11.2 3.5 1:06.79 init | 12:46 |
| Mirv | asac: I did use apport and let it install .ddebs, but it didn't help in getting a better backtrace, similar to that parent bug | 12:46 |
| ogra_ | i only have run through the intro yet | 12:46 |
| ogra_ | and it has already eaten 24M | 12:46 |
| Mirv | asac: but there's also merge proposals to unity-mir and hud already in there | 12:47 |
| Mirv | both from pete woods | 12:47 |
| asac | lool: question. the blocker issues by rick. are those confirmed to be fixed now? | 12:48 |
| dbarth | cjohnston: should be fine | 12:50 |
| cjohnston | ok | 12:50 |
| dbarth | cjohnston: i was wondering about a bot not merging in my approved branches; but i assume this is the big MIR shutdown? | 12:50 |
| cjohnston | I believe so | 12:50 |
| dbarth | i merged the branch manually | 12:50 |
| lool | == Publishing hud (hud) == | 12:51 |
| lool | with the dbus fix for real | 12:51 |
| Mirv | jibel: so regarding otto problem vila found indications of dbus/apparmor problem, and the timing of the start of the problems (Tue morning UTC time) would co-incide with the dbus and apparmor updates in saucy like https://lists.ubuntu.com/archives/saucy-changes/2013-October/011815.html | 12:54 |
| jdstrand | Mirv: what are the problems? | 12:55 |
| jdstrand | vila: ^ | 12:55 |
| vila | /var/log/syslog: Oct 8 09:16:26 saucy-i386-20131008-0916 dbus[499]: apparmor="DENIED" operation="dbus_method_call" bus="system" path="/org/freedesktop/DBus" interface="org.freedesktop.DBus" member="Hello" mask="send" name="org.freedesktop.DBus" pid=634 profile="lxc-container-default" peer_profile="lxc-container-default" info="Permission denied" | 12:55 |
| jdstrand | vila: that isn't from our change | 12:55 |
| fginther | morning | 12:55 |
| vila | jdstrand: how can we get a better idea of what is causing that and their fallout ? | 12:55 |
| jdstrand | I think that is another bug | 12:56 |
| * jdstrand looks | 12:56 | |
| jdstrand | vila: I think you may be losing the race in bug #1227937 | 12:58 |
| ubot5 | bug 1227937 in lxc (Ubuntu) "lxc-start is unconfined but has a profile defined" [High,Incomplete] https://launchpad.net/bugs/1227937 | 12:58 |
| jdstrand | vila: ie, due to a bug, the profile isn't normally in effect (though it is loaded), but for some reason, it was on your run | 12:58 |
| jdstrand | vila: and because it isn't normally in effect, there might be updates like the one you saw that are missing | 12:59 |
| jdstrand | vila: I suggest talking to stgraber | 12:59 |
| * jdstrand is also commenting in the bug | 12:59 | |
| jdstrand | vila: is that system still running? | 13:03 |
| plars | sergiusens: what's unmerged, the notes fixes? | 13:04 |
| Mirv | lool: note that because merger is offline, also the cu2d release commits don't get merged and cu2d complains about that in the prepare jobs (see eg. unity8 now - unity-mir and unity8) | 13:04 |
| vila | jdstrand: nope, it's part of ci, will try to reproduce locally after lunch | 13:04 |
| === vila is now known as vila-laaaate-lun | ||
| jdstrand | vila-laaaate-lun: I wouldn't try more than 2-3 times. I've never actually seen the profile in effect on Touch | 13:05 |
| jdstrand | vila-laaaate-lun: so think it may be a fluke. I am updating the bug though with all this info | 13:06 |
| === vila-laaaate-lun is now known as vila-late-lunch | ||
| vila-late-lunch | jdstrand: it's on an intel not a touch http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2465/console | 13:14 |
| lool | Mirv: ah that's bad | 13:17 |
| lool | Mirv: can we run these by hand? | 13:17 |
| lool | fginther: Heya | 13:17 |
| lool | fginther: and good morning | 13:17 |
| fginther | lool, mornin' | 13:17 |
| === alan_g|lunch is now known as alan_g | ||
| lool | fginther: can we resurrect all jobs merging archive back into the projects? | 13:17 |
| Mirv | lool: sure, the merge requests are there, I've done that before too. just branch the trunk, merging by hand and pushing. | 13:17 |
| cjohnston | dbarth: what is the branch you manually merged in please | 13:17 |
| fginther | lool, only if we enable all auto merging | 13:18 |
| lool | fginther: we would want to only enable the automerging of the things uploaded in rachive | 13:18 |
| lool | fginther: and selectively enable all automerging for some projects | 13:18 |
| jdstrand | vila-late-lunch: oh, interesting. can you comment in the bug? | 13:19 |
| jdstrand | vila-late-lunch: note, I just filed the bug-- I'm not up on how the container is all setup, etc | 13:20 |
| jdstrand | vila-late-lunch: but I asked hallyn to talk to stgraber about it so you can get some help | 13:20 |
| fginther | lool, there isn't a simple way to do the merge backs from the daily release version bumps | 13:20 |
| fginther | lool, if it isn't a lot, we can manually trigger those | 13:21 |
| fginther | lool, selective enabling can be done | 13:21 |
| fginther | lool, looks like 8 merge backs waiting, that can be done manually | 13:23 |
| lool | fginther: Ok | 13:27 |
| lool | fginther: is there an url where you see them all? | 13:27 |
| lool | fginther: or we ping you regularly, but sucks | 13:27 |
| lool | asac: What would you think of running the unity8 autopilots on SF (just these ones) to reenable autolanding there? | 13:28 |
| fginther | lool, this is the best I have: https://code.launchpad.net/~ps-jenkins | 13:28 |
| asac | lool: i dont think we should start getting more stuff unrelated to | 13:30 |
| asac | mir by default on today | 13:30 |
| lool | asac: this is to help landing unity8 changes | 13:32 |
| dbarth | cjohnston: hang on | 13:32 |
| asac | lool: related to mir by default? | 13:34 |
| === vila-late-lunch is now known as vila | ||
| lool | asac: Yes | 13:34 |
| Mirv | asac: we need those release tag merges in so that we can use cu2d as normal to publish automatically | 13:34 |
| lool | fginther: wow a bunch of branches there | 13:36 |
| lool | fginther: when you say merge by hand, is this kicking a jenkins job or doing the bzr merge? | 13:36 |
| dbarth | cjohnston: lp:~dbarth/cordova-ubuntu-tests/disable-apis-on-armhf | 13:36 |
| lool | fginther: Would you think you could do that | 13:36 |
| cjohnston | ty | 13:36 |
| vila | jdstrand: so to make sure I understand, someone request some resource, gets denied and we end up not being able to start our X server, does that match the bug ? | 13:36 |
| fginther | lool, it also shows all the old merged and rejected ones, only the 8 most recent happened since auto merge was disabled | 13:37 |
| fginther | lool, I'm triggering a jenkins job | 13:37 |
| plars | balloons: have you looked at the test_simple_page.TestSimplePage.test_swipe_delete_calculation_right_left on calculator? There's a bug that sounds like it might be related but it's older than I'd expect | 13:37 |
| lool | fginther: ok, so you're triggering it now for these 8? | 13:37 |
| fginther | lool, yes | 13:37 |
| lool | fginther: will ping you on further landings then | 13:38 |
| fginther | lool, ok, hopefully this won't last long :-) | 13:38 |
| lool | fginther: well jenkins must be quiet with automerger turned off :-) | 13:39 |
| vila | jdstrand: where can I find Serge Hallyn to discuss the issue ? | 13:39 |
| sergiusens | plars, yes | 13:44 |
| sergiusens | plars, not anymore | 13:44 |
| plars | sergiusens: awesome, yeah I see notes is working well on 89 now | 13:45 |
| sergiusens | plars, hopefully calculator and clock as well | 13:45 |
| plars | sergiusens: clock is good now it looks like, calc has one test failure | 13:46 |
| lool | sergiusens: hey | 13:46 |
| plars | sergiusens: a big improvement indeed | 13:46 |
| lool | sergiusens: is the notes-app.click with fixed tests up? | 13:46 |
| lool | sergiusens: Even if non-Mir, this one was tested by plars and me and should go in to help dashboard I think | 13:46 |
| lool | sergiusens: but the other click transitions... deferred til image is more decent | 13:46 |
| plars | lool: notes is working fine in image 89 | 13:47 |
| plars | lool: the email I sent last night was about 88 (and only the beginning part I had seen so far) | 13:47 |
| sergiusens | lool, yeah, you said I could make fixes as long as it fixes current tests, right? | 13:47 |
| lool | sergiusens: you can make fixes to clicks which are clicks or to apps that are apps that improve the tests | 13:48 |
| lool | sergiusens: let's not convert more apps to clicks for now though | 13:48 |
| sergiusens | plars, I say rerun the cal test, I got 100% | 13:48 |
| lool | sergiusens: but you can prepare them in trunk | 13:48 |
| lool | ah so notes-app.click was correct at the time I've built #89, that's what I wanted to confirm | 13:49 |
| lool | thanks | 13:49 |
| sergiusens | lool, sounds good; filemanager tests fail badly anyways | 13:49 |
| asac | lool: is there a way we can take a sneak preview at the hud changes? are they built? | 13:49 |
| sergiusens | plars, lool oh, wait, cailing calculator test I think is because the sdk added a confirm to delete option | 13:49 |
| lool | asac: Yes, in PPA | 13:49 |
| lool | asac: but there is still a crash with them | 13:50 |
| asac | lool: is there an improvement? | 13:50 |
| asac | lool: hmm. i dont see it in the results page | 13:50 |
| asac | http://people.canonical.com/~platform/cu2d/results | 13:50 |
| lool | asac: because I've pushed it to proposed | 13:50 |
| asac | ah good | 13:50 |
| lool | asac: but it's unapproved | 13:50 |
| lool | the stack needs to be rebuilt once it's in | 13:50 |
| asac | is there an improvement? | 13:50 |
| asac | lool: can you try removing the hud and see if the APs get better? | 13:51 |
| asac | removing == remove the binary forcefull locally i guess :) | 13:51 |
| * asac not sure how to get rid of it | 13:51 | |
| plars | sergiusens: the calc tests work for you though? even with that option? | 13:51 |
| lool | fginther: Can you land https://code.launchpad.net/~mterry/unity-mir/permissions/+merge/189718 ? | 13:52 |
| lool | asac: I'll try to, but right now focusing on landing the mir changes | 13:52 |
| psivaa | lool: just confirming. hud crash is still occurring with even with the version in the ppa | 13:52 |
| asac | lool: kk | 13:53 |
| lool | asac: I rarely have time to dive deeply into this or that between landings; it's only quieter when people get to bed | 13:53 |
| asac | lool: yeah sorry | 13:53 |
| lool | which is when I get stuff done, but then I block on reviews and queues for longer periods | 13:53 |
| asac | ogra_: can you try killing huddy? :) | 13:53 |
| lool | psivaa: Yes, that's what I saw too | 13:53 |
| sergiusens | plars, no, got confused; there's a new feature in ui toolkit to confirm swipe deletes, my bet is it's that | 13:53 |
| lool | psivaa: Might not be the same one though | 13:53 |
| asac | ogra_: i mean locally to see if it makes all the crashes and unity crasehs go away | 13:54 |
| rsalveti | lool: ogra_: if the container is not fully up yet, getprop will parse the file instead | 13:54 |
| psivaa | lool: ack, media scanner is crashing for me as well | 13:54 |
| rsalveti | that's why it works | 13:54 |
| ogra_ | rsalveti, you are awesome ! | 13:54 |
| ogra_ | :) | 13:54 |
| rsalveti | you only need the container for run-time specific properties and to set them | 13:54 |
| asac | so for me the hud crashes and then gallery_app autopilot also fails | 13:54 |
| lool | rsalveti: Good, that's why it worked; glad we have this part of the puzzle | 13:54 |
| asac | ogra_: how can i force disable hud? any idea? | 13:54 |
| lool | rsalveti: BTW mir enabled on grouper in latest lxc-android-config in image #89 | 13:54 |
| ogra_ | asac, sorry, as i said in the other channel, i'm on tracking some other issues too | 13:55 |
| fginther | lool, ack | 13:55 |
| asac | ogra_: sure, just tell me how i can try :) | 13:55 |
| ogra_ | asac, make the binary a link to /bin/true | 13:55 |
| lool | asac: replace it with a link to bin/true | 13:55 |
| rsalveti | lool: cool, were you able to test that? | 13:55 |
| lool | ogra_: :-) | 13:55 |
| * lool hugs ogra_ | 13:55 | |
| ogra_ | :) | 13:55 |
| asac | which binary is that? | 13:55 |
| * ogra_ hugs lool | 13:55 | |
| * asac uses dpkg -S | 13:55 | |
| lool | asac: under usr/lib/triplet | 13:55 |
| lool | asac: see the dbus .service file | 13:55 |
| lool | rsalveti: No | 13:55 |
| lool | rsalveti: I dont have a grouper | 13:55 |
| ogra_ | /usr/lib/arm-linux-gnueabihf/hud/hud-service | 13:55 |
| ogra_ | fi | 13:55 |
| ogra_ | (fi is important :P ) | 13:55 |
| lool | I had a personal one, but I gave it to my in laws to let them video call my son | 13:56 |
| * asac reboots without hud :) | 13:56 | |
| rsalveti | lool: got it, will test then :-) | 13:57 |
| rsalveti | lool: is the image out already? | 13:57 |
| rsalveti | or are you building one now? | 13:57 |
| lool | asac, rsalveti: I think we need to suck in the new kernels now and test them in relative isolation while we can or they will be picked up with the hybris change later | 13:57 |
| lool | rsalveti: #89 is out | 13:57 |
| asac | lool: i was told we dont need kernels for the performance issue | 13:57 |
| lool | rsalveti: mir by default on grouper mako maguro | 13:57 |
| lool | asac: we dont | 13:58 |
| lool | asac: but we will need an android upload that will pick them up | 13:58 |
| asac | lool: why? | 13:58 |
| asac | cant we avoid that? | 13:58 |
| lool | asac: we painfully could | 13:58 |
| ogra_ | asac, the new kernel brigns utreadahead support on mako | 13:58 |
| rsalveti | hm, quite a few apparmor changes | 13:58 |
| lool | asac: but the kernels were all boot tested, on SF and Mir | 13:58 |
| asac | not sure i want that :) | 13:58 |
| asac | noone complained about ureadahead | 13:58 |
| ogra_ | asac, we definitely want it at some point | 13:58 |
| ogra_ | asac, i did :P | 13:58 |
| lool | asac: it's just more work to revert | 13:58 |
| ogra_ | yeah | 13:59 |
| asac | at some point is not necessarily equal to now | 13:59 |
| lool | yep | 13:59 |
| asac | lool: revert what? | 13:59 |
| ogra_ | asac, you want to boot 10sec faster, dont you ? :) | 13:59 |
| lool | asac: the kernel uploads | 13:59 |
| asac | lool: ok... well, get ogasawara involved. her team should help testing stuff then | 13:59 |
| lool | there's a security fix in there too | 13:59 |
| asac | like running autopilots etc. | 13:59 |
| sergiusens | lool, plars confirming that the failing calc test is a new feature change in the sdk | 13:59 |
| lool | asac: they tested it | 13:59 |
| asac | ask her politely | 13:59 |
| lool | asac: yesterday and this morning | 13:59 |
| lool | on all devices | 13:59 |
| asac | lool: what did they test? all autopilots? | 13:59 |
| ogra_ | it was all teted for a few days | 13:59 |
| lool | asac: no boot tested the kernels | 13:59 |
| asac | did we get the test logs? | 13:59 |
| ogra_ | *testet | 14:00 |
| lool | asac: the changes are ureadahead and apparmor | 14:00 |
| plars | sergiusens: saw that... interesting | 14:00 |
| asac | lool: right. autopilots is king | 14:00 |
| plars | balloons: ^ | 14:00 |
| lool | and tyhicks and/or jamie tested the apparmor changes | 14:00 |
| lool | asac: let's build an image with them and look at the results | 14:00 |
| ogra_ | ++ | 14:00 |
| asac | we can... but only before the US wakes up in case our devices get busted :) | 14:00 |
| asac | e.g. not end of day plz | 14:00 |
| lool | well, I think we should do it now | 14:00 |
| ogra_ | lool, hurry up | 14:00 |
| ogra_ | before the US gets up | 14:01 |
| ogra_ | :) | 14:01 |
| plars | ogra_: US is up :) | 14:01 |
| asac | lool: what do we need the android changes for? | 14:01 |
| ogra_ | plars, ssshhhh ! | 14:01 |
| ogra_ | dont let asac see you ! | 14:01 |
| lool | asac: android will need to be uploaded to pick up some upcoming hybris changes | 14:01 |
| ogra_ | asac, to pick up the kernels we need an android rebuild | 14:01 |
| lool | I'm anticipating the kernel landing | 14:01 |
| plars | ogra_: I've been up for a while, but I'd be more than happy to go back to sleep if I wasn't on my 2nd cup of coffee already :) | 14:01 |
| ogra_ | plars, do that then, lool will give you a cal when you are allowed to get up :) | 14:02 |
| plars | haha | 14:02 |
| plars | too much coffee already | 14:02 |
| * lool builds no change android upload | 14:02 | |
| ogra_ | (its all about interpreting magare orders right, isnt it ? :) ) | 14:02 |
| ogra_ | *manager | 14:03 |
| asac | lool: are you coming to standup? | 14:05 |
| lool | oh got | 14:05 |
| lool | asac: my phone is off due to the sim thing | 14:05 |
| lool | thanks for the ping | 14:05 |
| jdstrand | lool: also the kernel team | 14:12 |
| jdstrand | lool: (ie, the security team and the kernel team tested the changes) | 14:12 |
| jdstrand | lool: btw, did you see tyhicks ping in backscroll? | 14:12 |
| lool | jdstrand: no, just reading it now | 14:13 |
| lool | in a HO | 14:13 |
| doanac | asac: the AP tests start/stop unity for the test run. but not in between test cases | 14:19 |
| sergiusens | lool, fwiw, this is today; for the spreadsheet: http://paste.ubuntu.com/6213931/ | 14:20 |
| asac | doanac: between test runs too? | 14:21 |
| asac | doanac: phablet-test-run does that? | 14:21 |
| doanac | asac: no. the "unlock_screen.py" script does it | 14:21 |
| asac | doanac: ok. thanks | 14:22 |
| * ogra_ installs the new hud and reboots | 14:22 | |
| === alan_g is now known as alan_g|tea | ||
| lool | sergiusens: awesome | 14:32 |
| lool | sergiusens: but let's stage it somewhere for next spot | 14:33 |
| lool | uploaded no change android | 14:34 |
| lool | to pick up kernels | 14:34 |
| ogra_ | new hud doesnt fix anything | 14:34 |
| ogra_ | :( | 14:34 |
| lool | ogra_: no | 14:34 |
| sergiusens | lool, problem is, I don't know how to stage this more; or how long until the tests stop working due to changes elsewhere | 14:34 |
| ogra_ | lool, well, and see the backlog in #ubuntu-touch | 14:34 |
| sergiusens | lool, if the answer is: wait for mir; that's fine | 14:34 |
| lool | sergiusens: right now we've stopped the line to land mir related fixes | 14:37 |
| lool | that is, changes for regressions introduced by mir | 14:37 |
| sergiusens | lool, that is fine by me | 14:37 |
| sergiusens | lool, fginther just an fyi; the line hasn't stopped for the community jenkins | 14:37 |
| fginther | lool, ack I missed that one last night. I disabled that about 30 minutes ago | 14:38 |
| rsalveti | lool: grouper working fine with mir | 14:39 |
| rsalveti | only issue with grouper is brcm_patchram consuming 100% of your cpu | 14:39 |
| ogra_ | yeah, thats filed | 14:39 |
| ogra_ | and cyphermox knows about it | 14:39 |
| ogra_ | (not sure he'll have time to fix) | 14:39 |
| cyphermox | working on it.. | 14:39 |
| plars | bfiller, sergiusens: I'm hoping to add the ubuntu-keyboard tests today, but hitting an issue... some of the tests seem to fail without mir, but more importantly, it doesn't seem to work with mir enabled, tells me: | 14:42 |
| plars | ProcessSearchError: ('Search criteria returned no results', 'Unable to find maliit-server dbus object. Has it been started with introspection enabled?') | 14:43 |
| plars | but I've restarted maliit-server with -testability | 14:43 |
| === alan_g|tea is now known as alan_g | ||
| bfiller | plars: no idea, I'd suggest holding off on this until after 13.10. don't need to introduce more troubles at this stage :) | 14:43 |
| psivaa | plars: notes_app in maguro is still timing out with mir, when i run locally it has failures (10) but not timing out | 14:45 |
| psivaa | plars: is it worth changing the device there and re-run the test | 14:45 |
| plars | psivaa: let me try locally | 14:45 |
| psivaa | plars: ack | 14:45 |
| plars | psivaa: took only 6 min. for me on mako, I'm installing on maguro now though | 14:54 |
| psivaa | plars: my failures appear to be related to bug #1237207 | 14:58 |
| ubot5 | bug 1237207 in Unity 8 "Autopilot test applications get killed" [Undecided,New] https://launchpad.net/bugs/1237207 | 14:58 |
| psivaa | but on maguro | 14:58 |
| lool | psivaa: are you running with click tests? | 15:01 |
| psivaa | lool: yes | 15:01 |
| lool | psivaa: this is what you have to run: | 15:01 |
| lool | ./phablet-config autopilot --dbus-probe enable | 15:01 |
| lool | ./phablet-click-test-setup --click com.ubuntu.notes | 15:01 |
| lool | ./phablet-test-run notes_app | 15:01 |
| lool | haven't tried on #89 though | 15:01 |
| lool | psivaa: is it what you're running? | 15:01 |
| psivaa | lool: i use http://pastebin.ubuntu.com/6213423/ | 15:02 |
| psivaa | lool: it's a combination of commands that i got from the automated tests | 15:02 |
| lool | psivaa: Ok, not sure about yours | 15:03 |
| lool | psivaa: the apparmor stuff looks similar | 15:03 |
| lool | psivaa: But I'd trust my set better | 15:03 |
| lool | psivaa: cause a) I got it from sergiusens b) it's all wrapped up nicely in phablet-tools c) it worked :-) | 15:04 |
| psivaa | lool: ack, will use that from now on :) | 15:04 |
| lool | upstart (1.10-0ubuntu7) saucy; urgency=low | 15:06 |
| lool | ogra_: upstart upload? | 15:06 |
| ogra_ | lool, ++ | 15:07 |
| ogra_ | yeah :) | 15:07 |
| lool | ogra_: err | 15:07 |
| ogra_ | lool, will help maguro | 15:07 |
| lool | ogra_: yes, but where is this tracked? | 15:07 |
| ogra_ | it isnt, but it will be held in proposed anyway | 15:07 |
| lool | how so? | 15:07 |
| ogra_ | gimme a sec, in meeting | 15:09 |
| plars | psivaa: webbrowser is still getting some strange issue on maguro also | 15:10 |
| lool | fginther: hey, I need some upstream merges, mind running these: | 15:10 |
| dobey | lool: is there a way to see what exactly is in the devel-proposed image? i'm pretty sure the ubuntuone-credentials and unity-scope-click fixes aren't in it | 15:10 |
| lool | fginther: https://code.launchpad.net/~kgunn72/platform-api/bump-mir-dep14/+merge/190011 | 15:10 |
| lool | fginther: https://code.launchpad.net/~kgunn72/unity-system-compositor/bump-mir-dep14/+merge/190012 | 15:10 |
| psivaa | plars: ohh yea, wanted to try that locally too, but missed. i'll try that in a bit | 15:10 |
| fginther | lool, ack | 15:10 |
| lool | fginther: https://code.launchpad.net/~kgunn72/unity-mir/bump-mir-dep14/+merge/190010 but not top approved yet (can't) but can go in | 15:10 |
| lool | this is all for now | 15:10 |
| ogra_ | lool, bug 1234743 ... and as a sideefect bug 1235649 ... for the upstart upload | 15:10 |
| ubot5 | bug 1234743 in linux (Ubuntu) "omapfb module floods system with udev events on samsung galaxy nexus when playing mp4" [Undecided,Incomplete] https://launchpad.net/bugs/1234743 | 15:11 |
| ubot5 | bug 1235649 in upstart (Ubuntu Saucy) "uevent spam causes session upstart to consume massive amounts of memory on Ubuntu Touch" [Critical,Confirmed] https://launchpad.net/bugs/1235649 | 15:11 |
| lool | dobey: http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/pending/saucy-preinstalled-touch-armhf.manifest | 15:11 |
| lool | dobey: ubuntuone-credentials-common:armhf13.10-0ubuntu1 | 15:11 |
| dobey | lool: right, that doesn't include the fix | 15:11 |
| cjwatson | heads-up, we may have to upload eglibc to help with a widespread set of blockers for the arm64 port | 15:13 |
| dobey | lool: i'm about to prepare a package with the fix, and another trivial focus fix for 2fa entry, though; can we get it landed today? | 15:13 |
| lool | cjwatson: is that going to break something while it's building / in proposed? | 15:13 |
| lool | dobey: ah I thought the fix for my issue (empty email address was in), but that's a different issue then? | 15:14 |
| dobey | lool: yes, this is for bug #1233756 | 15:14 |
| ubot5 | bug 1233756 in Unity Click Scope "Possible infinite loop in scope search" [Critical,New] https://launchpad.net/bugs/1233756 | 15:14 |
| lool | dobey: it will have to be tomorrow though | 15:14 |
| cjwatson | there's a dlopen segfault on statically linked programs that doesn't generally matter on most arches, but the problem is that our arm64 hardware is randomly faulting in a few other places and so we've had to fail any build that faults to avoid misbuilds; and this segfault happens (normally harmlessly) in the middle of typical configure runs and causes widespread confusion | 15:14 |
| jdstrand | lool: trivial bug fix: http://paste.ubuntu.com/6214169/ | 15:15 |
| lool | cjwatson: but no risk for other arches, just a heads up that you're working on this package and that we will see it updated on the image? thanks for the heads up then | 15:15 |
| dobey | lool: ooh, no, that's the click scope bug, sorry. | 15:15 |
| jdstrand | not sure why that is coming up now, but cking reported it to me with his testing | 15:15 |
| dobey | lool: u1-credentials is bug #1237006 | 15:15 |
| ubot5 | bug 1237006 in ubuntuone-credentials (Ubuntu) "'back' from existing account page deletes account" [Critical,Confirmed] https://launchpad.net/bugs/1237006 | 15:15 |
| cjwatson | lool: don't believe it breaks anything while building, I checked and there's no arch all/any skew involved | 15:16 |
| lool | jdstrand: Ok, please upload; I'll pretend it's Mir fix for grouper | 15:16 |
| jdstrand | it might actually be one :) | 15:16 |
| psivaa | plars: i used the method that lool said above for notes tests and still getting 9 failures on maguro | 15:16 |
| cjwatson | lool: yep. I don't believe so but am build-testing on amd64 while waiting for Adam to be around to review | 15:16 |
| lool | dobey: Ok; please add a landing ask for it | 15:17 |
| lool | dobey: will be tomorrow | 15:17 |
| lool | not critical for thursday | 15:17 |
| lool | might be friday or so | 15:17 |
| dobey | lool: when should i upload it to archive then? | 15:18 |
| lool | fginther: third MP approved | 15:18 |
| lool | dobey: rather not | 15:18 |
| lool | dobey: sorry, I meant no | 15:19 |
| lool | dobey: when we're done with fixing the top mir issues; we will likely resume upsream merger at around that time too | 15:19 |
| lool | dobey: I'm guesstimating end of thursday or friday | 15:19 |
| dobey | ok | 15:20 |
| jdstrand | lool: thanks, uploaded | 15:21 |
| lool | rsalveti: what's in the libhybris upload? don't see it in lp:libhybris/ubuntu | 15:23 |
| rsalveti | lool: sorry, let me push that, adding power.h | 15:23 |
| rsalveti | done | 15:23 |
| bfiller | fginther: can you re-run the maguro test on this MR: https://code.launchpad.net/~schwann/camera-app/camera-vf-centered/+merge/190143 | 15:24 |
| bfiller | lool: for MR's that have been approved and top approved will they get automatically merged when upstream merger turned back on? | 15:25 |
| lool | bfiller: I guess | 15:25 |
| lool | bfiller: if anything helps current tests on Mir, we're interested in helping merge it | 15:25 |
| bfiller | fginther: ^^^ can you confirm this? | 15:25 |
| fginther | bfiller, can rerun and yes, once the auto merging is turned on, those MPs will be processed without any further action | 15:26 |
| bfiller | fginther: cool | 15:26 |
| bfiller | lool: is the MIR image available now? | 15:28 |
| lool | bfiller: Yes, #89 is Mir by default | 15:29 |
| bfiller | lool: thanks | 15:29 |
| lool | fginther: bzr: ERROR: Error parsing trunk.recipe:2:2: Indent not a multiple of two spaces. | 15:30 |
| lool | https://code.launchpad.net/~kgunn72/unity-mir/bump-mir-dep14/+merge/190010 | 15:30 |
| cjohnston | lool: I think that was my fault | 15:30 |
| cjohnston | http://10.97.0.26:8080/job/unity-mir-saucy-amd64-autolanding/23/console | 15:30 |
| kgunn | lool: not trying to create a mess...but was the test fmwk going to take care of the stale mir socket ? | 15:30 |
| kgunn | just trying to understand why unity8 is 0 test runs | 15:31 |
| lool | kgunn: you mean the stop powerd workaround? | 15:31 |
| cjohnston | lool: does that failure look better? (even though it's still a failure) | 15:32 |
| cjohnston | fginther: ^ | 15:32 |
| kgunn | lool: yeah | 15:33 |
| fginther | cjohnston, that's a different issue, missing dependency | 15:33 |
| fginther | cjohnston, Lool, looking into it | 15:33 |
| cjohnston | fginther: right.. the first failure was my fault... this one was actually valid (i.e. better) | 15:33 |
| lool | fginther: thanks | 15:38 |
| lool | kgunn: not applied yet AFAIK | 15:39 |
| lool | plars, doanac: Not sure you got the thing about the flag | 15:39 |
| lool | plars, doanac: Could we add a flag to phablet-test-run to "stop powerd" before running the tests and start it again afterwards? we'd use this for unity8 mir tests | 15:39 |
| lool | temporary workaround | 15:39 |
| plars | lool: sure, but I thought we needed it active? | 15:40 |
| lool | plars: powerd? | 15:40 |
| plars | lool: also, how temporary | 15:40 |
| lool | plars: temporary some days until mir is fixed | 15:40 |
| lool | hopefully removed before release | 15:40 |
| lool | plars: we have a critical bug to fix this anyway since first boot is too slow | 15:40 |
| doanac | plars: you want me to look at this powerd change, or can you do it? | 15:41 |
| plars | doanac: I can do it, I have the latest image installed on both devices, just need a few for some tests to finish | 15:42 |
| lool | doanac, plars: Goal is to get unity8 results showing up for mir | 15:42 |
| doanac | cool. you plan to just alter the setup.sh for that test? | 15:42 |
| lool | doanac, plars: Expected result is one pass rest failures | 15:42 |
| plars | lool: that's not a huge improvement :) | 15:42 |
| doanac | 0-1 is an infinite improvement :) | 15:43 |
| plars | heh | 15:43 |
| lool | plars: it will allow seeing the improvement in unity8 when it happens though :-) | 15:43 |
| plars | lool: this is true, but is someone looking at why that's still a far cry from what we see with sf? | 15:43 |
| lool | plars: Oh yes, Saviq is AFAIK | 15:45 |
| lool | Saviq: Do you see the end of the tunnel for the unity8 AP input issue? | 15:45 |
| kgunn | lool: asac jfunk ....can we do this, can we turn CI back on for unity8 minus the AP tests? this would at least be more QA safe than where we are with manual merge (w/ no tests run) | 15:47 |
| lool | kgunn: there is more to turning upstream merger on for unity8 | 15:49 |
| fginther | lool, kgunn, http://10.97.0.26:8080/job/unity-mir-saucy-i386-autolanding/59/console is failing due to a build dependency. | 15:49 |
| lool | kgunn: we need to fix the unity8 AP tests at the least | 15:49 |
| kgunn | lool: asac jfunk ....we have over 250 tests run outside of AP (of which there are only like 12) | 15:50 |
| fginther | lool, kgunn, I think something justs needs to be rebuilt and tried again | 15:50 |
| lool | fginther: it just landed in PPA for armhf | 15:50 |
| lool | not amd64 though | 15:50 |
| lool | fginther: it should be there on i386 though | 15:51 |
| lool | fginther: failed 20 ago, mir finished building 17 minutes ago | 15:52 |
| lool | fginther: mind giving it back? | 15:52 |
| lool | https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/5088266 | 15:52 |
| fginther | lool, "give it back" I don't follow | 15:53 |
| lool | fginther: http://10.97.0.26:8080/job/unity-mir-saucy-i386-autolanding/59/console | 15:54 |
| lool | fginther: can you retry on i386 | 15:55 |
| fginther | lool, yes, already in progress | 15:55 |
| lool | fginther: mir i386 went in minutes after this failed | 15:55 |
| lool | fginther: great thanks | 15:55 |
| kgunn | lool: asac jfunk ....right now we're working without a net, i'm actually asking for tests that are passing as part of regular ci/autolanding on unity8 to be re-enabled ....so that we don't break them! | 15:56 |
| fginther | lool, same problem | 15:56 |
| fginther | I can't figure out whey it's using libmirserver5 | 15:56 |
| lool | fginther: it's the new bdep | 15:57 |
| kgunn | lool: asac jfunk wih the goal that we would make unity8 AP test suite green...to eventually re-enable it | 15:57 |
| lool | fginther: it's odd, I see it there https://launchpad.net/~ubuntu-unity/+archive/daily-build/+build/5088266 | 15:58 |
| fginther | lool, amd64 is done now, just waiting for it to publish in the ppa | 15:58 |
| lool | kgunn: yes, so doanac and plars are adding the powerd workaround IIUC | 15:58 |
| lool | kgunn: will only get you one test passing | 15:59 |
| lool | kgunn: I proposed running the unity8 tests on SF for now, but asac didn't want that | 15:59 |
| kgunn | Saviq: ^ | 15:59 |
| lool | I dont remember the rationale | 15:59 |
| plars | lool, kgunn, doanac: yes, I'm trying it locally right now | 15:59 |
| kgunn | probably because we're full steam ahead on mir (SF is so yesterday) | 15:59 |
| lool | yeah | 16:00 |
| plars | lool: well, we *do* run them on sf and on mir | 16:00 |
| lool | still prefer having a testsuite be running for other code changes | 16:00 |
| plars | lool: but I don't think we should have a mixed image where we run part of the tests on mir and part on sf | 16:00 |
| lool | plars: at image build time | 16:00 |
| lool | plars: we're speaking of upstream merger though | 16:00 |
| plars | ah | 16:00 |
| plars | gotcha | 16:00 |
| Saviq | lool, why will it get us one test passing? | 16:02 |
| Saviq | lool, the reason why the unity8 tests are failing (other than the first one) | 16:02 |
| Saviq | lool, is that the socket is left over | 16:02 |
| Saviq | lool, so if you clear the socket between tests | 16:02 |
| Saviq | lool, most should pass | 16:02 |
| lool | Saviq: do you have a way to do this? | 16:03 |
| lool | Saviq: or is a fix already in the mir pipe for this? | 16:03 |
| kgunn | rm /tmp/mir_socket | 16:03 |
| lool | kgunn: yeah, but we need this in the AP infra | 16:03 |
| Saviq | lool, it would get cleared but unity8 crashes on exit | 16:03 |
| lool | "easy" to develop I guess, just asking in case it's somewhere already | 16:03 |
| Saviq | lool, kgunn we can do it in our ap suite | 16:03 |
| Saviq | lool, kgunn before starting unity8 | 16:04 |
| lool | Saviq: let's do it then | 16:04 |
| ogra_ | asac, hangout ? | 16:06 |
| asac | ogra_: one sec | 16:06 |
| asac | ogra_: please go ahead unless you need me. will be there soon | 16:07 |
| ogra_ | https://code.launchpad.net/~ricmm/powerd/use-hal-power-hint/+merge/190027 | 16:13 |
| ogra_ | lool, ^^ | 16:13 |
| plars | lool, kgunn, Saviq: I'm trying to run the unity8-autopilot tests with powerd stopped, but I'm hitting a problem at the end when it tries to start unity8 again, it just hangs there. This is what's prevented us from completing the runs before though, so it's not unique to running with powerd stopped | 16:20 |
| plars | in the unity8 upstart log I see: | 16:20 |
| plars | terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<std::runtime_error> >' | 16:20 |
| plars | what(): Could not unblank display | 16:20 |
| Saviq | plars, that ↑ says display is blanked | 16:21 |
| Saviq | plars, sure powerd was stopped - and that's when screen was *not* blank | 16:21 |
| Saviq | ? | 16:21 |
| plars | Saviq: could have been blanked already, so we need to ensure that the screen is ON, and *then* stop powerd? | 16:22 |
| Saviq | plars, yes, I'm afraid | 16:22 |
| plars | hmm | 16:22 |
| Saviq | plars, can't we use powerd-cli to keep it on? | 16:22 |
| plars | Saviq: well, for other things we use powerd-cli active | 16:23 |
| Saviq | lool, kgunn https://code.launchpad.net/~saviq/unity8/remove-socket-in-ap-tests/+merge/190175 | 16:23 |
| balloons | fginther, the core apps status spreadsheet should be all up to date. 2 apps left basically, and they are dependency issues more or less. | 16:57 |
| fginther | balloons, awesome | 16:58 |
| balloons | fginther, the core apps ppa is a part of the testing vm, but for instance terminal needs a dependency found in there | 17:00 |
| seb128 | hum | 17:01 |
| seb128 | is the upstream merger being down limited to packages impacting touch? | 17:01 |
| seb128 | or is that for everything under auto landing? | 17:01 |
| seb128 | fginther, ^ | 17:03 |
| fginther | seb128, everything | 17:03 |
| seb128 | :-( | 17:03 |
| fginther | seb128, everything saucy | 17:03 |
| lool | asac: ready when you are | 17:04 |
| fginther | balloons, please be careful when rebuilding jobs, the parameters for terminal-app hav changed to pick up that dependency | 17:04 |
| balloons | fginther, :-) | 17:05 |
| fginther | balloons, I started this one: http://91.189.93.70:8080/job/ubuntu-terminal-app-ci/12/ | 17:05 |
| balloons | I see it now.. you are right, I should do full build now's | 17:05 |
| robru | fginther, so, what was decided for autolanding? enabled for some stacks only? can it be enabled for webapps? | 17:06 |
| lool | fginther: forgot one thing | 17:06 |
| lool | fginther: how do you run the upstream merger tests, without actually merging (before top approve) | 17:06 |
| lool | seb128: fginther is looking at reenabling a whitelist of projects for which APs will pass | 17:07 |
| fginther | lool, do the exact same thing, except use the "${project}-ci" job instead of "${project}-autolanding" | 17:07 |
| === alan_g is now known as alan_g|EOD | ||
| seb128 | lool, what about desktop specific projects? | 17:07 |
| lool | fginther: ok thanks | 17:07 |
| seb128 | lool, why were those shutdown to start? | 17:08 |
| fginther | lool, also, those ci jobs are still be automatically ran | 17:08 |
| fginther | robru, we have to assess any impact on touch first | 17:08 |
| lool | seb128: too many failures because of mir switch, stopping the line to deal with mir issues, wanting a quiet environment to land only mir changes and focus on only that today and until fixed | 17:09 |
| robru | fginther, hmmm, from my point of view, webapps (at least cordova) is broken, and having to manually merge branches is impeding my ability to fix things. | 17:09 |
| seb128 | lool, well, I fail to see how e.g gnome-control-center-unity has anything to do with Mir or mir switch | 17:09 |
| lool | asac: I think Ill go have dinner now, waiting for mir to build in PPA | 17:09 |
| seb128 | lool, it's a GTK based component that doesn't run on the phone and has no tests | 17:09 |
| seb128 | lool, just to take an exemple | 17:10 |
| lool | seb128: we didn't have finer grained switches, but we're getting them now | 17:10 |
| seb128 | lool, btw that sort of shutdown ought to be announced on our devel lists... | 17:10 |
| lool | seb128: it was announced on the ubuntu-phone list; I'm sorry if you missed it there | 17:10 |
| seb128 | lool, not everyone is on the phone list, you triggered shutdown for desktop components and impacted Ubuntu community members that don't follow phone work | 17:11 |
| seb128 | lool, it would have been nice to keep them updated as well | 17:11 |
| lool | seb128: _I_ didn't trigger shutdown | 17:11 |
| seb128 | lool, well, whoever did, you are the one commented so I'm replying to you, sorry about that ;-) | 17:11 |
| lool | seb128: that's ok | 17:11 |
| fginther | robru, I'll get back to you | 17:12 |
| robru | fginther, thanks | 17:12 |
| lool | Building unity8 stack to get unity-mir snapshot against new mir | 17:22 |
| * lool => dinner & | 17:28 | |
| jdstrand | lool: whenever you get back (this isn't urgent), I'd like to upload a trivial fix in ntp: http://paste.ubuntu.com/6214702/. ntpdate (from the ntp package) is on the image, but the patch doesn't touch ntpdate | 17:37 |
| jdstrand | dang it | 17:41 |
| jdstrand | can people who upload hybris changes check higher up the stack? | 17:41 |
| jdstrand | Oct 9 17:41:03 ubuntu-phablet kernel: [ 920.495251] type=1400 audit(1381340463.439:75): apparmor="DENIED" operation="open" parent=1494 profile="com.ubuntu.developer.webapps.webapp-twitter_webapp-twitter_1.0.3" name="/run/shm/hybris_shm_data" pid=2473 comm="webbrowser-app" requested_mask="rwc" denied_mask="rwc" fsuid=32011 ouid=32011 | 17:42 |
| * jdstrand does yet another apparmor-easyprof-ubuntu update | 17:42 | |
| seb128 | jdstrand, you need an integration/autopilot test there ;-) | 17:45 |
| jdstrand | seb128: I would think that preinstalled webapps tests would catch it | 17:46 |
| jdstrand | I'm not sure those have landed yet | 17:46 |
| kgunn | lool: did you alredy take in the new mir and bump api ? (its ok if you did, just going to have another right behind it) | 17:47 |
| lool | kgunn: it's now all staged in PPA | 17:57 |
| lool | kgunn: took a while to merge and upload everything there | 17:57 |
| lool | kgunn: now about to test this | 17:58 |
| lool | jdstrand: this was meant to be a header only change | 17:59 |
| jdstrand | lool: I'll have the fix in a moment | 17:59 |
| kgunn | lool: thanks...no prob...i'll queue another for tomorrow morn your time (repeat the process unfortunately) | 17:59 |
| lool | jdstrand: which real change caused this? | 17:59 |
| lool | jdstrand: libhybris r87 seems it will add a .h, must be something else | 18:00 |
| lool | kgunn: With ABI? | 18:00 |
| jdstrand | lool: I have no idea, but libhybris is what creates that file (see bug #1226569) | 18:00 |
| ubot5 | bug 1226569 in libhybris (Ubuntu Saucy) "libhybris should use an app-specific path for shared memory files" [High,New] https://launchpad.net/bugs/1226569 | 18:00 |
| kgunn | lool: yeah | 18:01 |
| lool | jdstrand: since when do you see this? | 18:01 |
| lool | jdstrand: anyway, thanks for catching that, adding to plan | 18:01 |
| jdstrand | lool: before, access was only needed in the camera policy group. today I installed 89 and this is breaking at least twitter | 18:01 |
| lool | jdstrand: could this be "thanks to" the gstreamer fix? | 18:02 |
| lool | jdstrand: remember, we changed it so that we could run the hybris media scanner | 18:02 |
| jdstrand | lool: I had to do several gstreamer fixes in 1.0.36 because of that | 18:03 |
| jdstrand | lool: I really don't know when it was introduced-- I tested a lot yesterday and so did the kernel team, but it was only today that confined click apps fail to launch | 18:04 |
| lool | jdstrand: maybe with mir switch? | 18:06 |
| jdstrand | *shrug* | 18:07 |
| jdstrand | I'm surprised our click app tests didn't catch this | 18:07 |
| jdstrand | sudoku and twitter both fail to start | 18:08 |
| jdstrand | lool: http://paste.ubuntu.com/6214815/ | 18:13 |
| jdstrand | verified to fix it on grouper | 18:13 |
| jdstrand | (ie, apps start again) | 18:13 |
| lool | jdstrand: twitter does start for me | 18:15 |
| jdstrand | lool: you installed --channel=devel-proposed? | 18:16 |
| lool | jdstrand: maybe it's grouper specific, there was a change to enable mir on it this morning | 18:16 |
| lool | yes, I'm on #89 | 18:16 |
| lool | w/ Mir | 18:16 |
| jdstrand | maybe. I see surfaceflinger on grouper is not running | 18:16 |
| jdstrand | I don't know why this wouldn't also be a problem on mako/manta though | 18:17 |
| jdstrand | maybe they use less of hybris | 18:17 |
| jdstrand | again, I don't know-- but the fix works | 18:18 |
| jdstrand | lool: I see you put INARCHIVE. I haven't uploaded yet, but will | 18:19 |
| lool | jdstrand: that was the previous one | 18:20 |
| lool | jdstrand: fixed, thanks | 18:20 |
| jdstrand | the previous one wasn't a hybris file though | 18:20 |
| jdstrand | oh, I see there is a critical conjunction in there :) | 18:20 |
| jdstrand | lool: may I upload? | 18:21 |
| ogra_ | protocol failure | 18:24 |
| ogra_ | ERROR:phablet-flash:Command 'adb push /home/ogra/Downloads/phablet-flash/imageupdates/pool/ubuntu-4eaaaaf4b35f7feed61e8f100909763a87abf4dc93cdeeb56079e45a14484d6e.tar.xz /cache/recovery/' returned non-zero exit status 1 | 18:24 |
| * ogra_ goes crazy | 18:24 | |
| ogra_ | lool, i suspect it will take rarher longer for me to test the kernel ... | 18:24 |
| ogra_ | my first try to replace the modules in the image trashed it and now i cant re-flash | 18:25 |
| jdstrand | does INFLIGHT mean I'm allowed to upload? | 18:28 |
| * jdstrand is going to be happy when t opens | 18:28 | |
| ogra_ | jdstrand, you expect anything to change in T ? | 18:28 |
| kgunn | jdstrand: worth mentioning, alan_g's fix to change the mir_socket location is now in the ubuntu-unity/daily-build packages | 18:29 |
| ogra_ | (we'll surely have a more convenient landing process then, but stuff wont just go in either) | 18:29 |
| kgunn | jdstrand: altho...if you pull the libmir*'s you'll also need unity-mir & platform-api (those got rebuilt/api build dep bump) | 18:29 |
| jdstrand | ogra_: I would hope so-- is that not the case? | 18:32 |
| jdstrand | kgunn: awesome! :) | 18:33 |
| jdstrand | I figured the super-strict gating was for stabilization | 18:33 |
| ogra_ | jdstrand, i dont think touch will change, stuff that lands on the image will go through a gatekeeping process ... i dream of doing the gatekeeping through -proposed and have the spreadsheet replaced by a mixture of FFe/SRU bug processes though | 18:35 |
| jdstrand | that gives me a sad face. this has been a difficult process. hopefully the changes you mention will make it less so | 18:36 |
| rsalveti | jdstrand: actually, nothing changed in hybris, that was always required | 18:36 |
| jdstrand | rsalveti: it only just showed up today though | 18:37 |
| rsalveti | maybe because of mir | 18:37 |
| rsalveti | not sure | 18:37 |
| jdstrand | it is weird cause I run mir on mako it wasn't needed | 18:37 |
| jdstrand | (though I am not on 89 there) | 18:37 |
| jdstrand | it was only ever needed for the camera | 18:38 |
| lool | jdstrand: yes please | 18:38 |
| lool | jdstrand: (upload) | 18:38 |
| lool | ogra_: did you manage to test the kenrel in the end? | 18:38 |
| lool | ogra_: I know how now | 18:38 |
| ogra_ | lool, for the modules ? | 18:39 |
| ogra_ | i manage to replace the bootimg | 18:39 |
| lool | ogra_: for system.img + boot.img | 18:39 |
| lool | ogra_: http://paste.ubuntu.com/6214921/ | 18:39 |
| jdstrand | rsalveti: the policy update has been uploaded now, so it doesn't matter (though, bug #1226569 remains) | 18:40 |
| ubot5 | bug 1226569 in libhybris (Ubuntu Saucy) "libhybris should use an app-specific path for shared memory files" [High,New] https://launchpad.net/bugs/1226569 | 18:40 |
| rsalveti | jdstrand: thanks | 18:40 |
| rsalveti | jdstrand: and /dev/knvmap is because of mir | 18:41 |
| rsalveti | on grouper | 18:41 |
| * jdstrand nods | 18:41 | |
| jdstrand | rsalveti: thanks | 18:41 |
| rsalveti | so thanks for the fix :-) | 18:42 |
| jdstrand | :) | 18:44 |
| lool | jdstrand: ntp >> please upload | 18:44 |
| jdstrand | lool: ack, thanks | 18:44 |
| lool | tiredness is showing today | 18:44 |
| jdstrand | lool: seems you are getting a lot of pings indeed-- all the ones from me is probably tiring enough | 18:45 |
| ogra_ | ++ | 18:45 |
| ogra_ | and + | 18:45 |
| jdstrand | lool: sorry about that :\ | 18:45 |
| ogra_ | jdstrand, not your fault :) | 18:45 |
| jdstrand | :) | 18:45 |
| ogra_ | we agreed to this insane process ... so we have to carry our own burnout we cause :) | 18:45 |
| jdstrand | heh | 18:46 |
| rsalveti | yeah, we asked for it | 18:47 |
| rsalveti | now we take it | 18:47 |
| ogra_ | well, i meant we as in the landing team ... :) | 18:48 |
| jdstrand | tyhicks, lool: dbus uploaded | 18:48 |
| ogra_ | the rest is just annoyed :) | 18:48 |
| jdstrand | lool: ntp uploaded | 18:48 |
| tyhicks | thanks! | 18:48 |
| lool | tyhicks, jdstrand: Thanks | 18:50 |
| lool | ogra_: testing this | 18:56 |
| lool | http://paste.ubuntu.com/6214995/ | 18:56 |
| lool | modules are from 9th of october | 18:56 |
| lool | err 8th | 18:56 |
| lool | drwxr-xr-x 5 root root 4096 Oct 8 13:34 net | 18:56 |
| lool | which seems consistent | 18:56 |
| ogra_ | lool, yeah, that looks about right | 18:57 |
| ogra_ | root@ubuntu-phablet:/# uname -a | 19:07 |
| ogra_ | Linux ubuntu-phablet 3.0.0-3-maguro #18-Ubuntu SMP PREEMPT Tue Oct 8 13:07:25 UTC 2013 armv7l armv7l armv7l GNU/Linux | 19:07 |
| ogra_ | phew | 19:07 |
| ogra_ | lool, looks all fine to me on maguro ... apps still and behave as expected | 19:08 |
| ogra_ | *still run | 19:08 |
| lool | ogra_: cool, did you do as I noted to replace system.img? | 19:09 |
| ogra_ | yeah | 19:09 |
| ogra_ | works fine | 19:09 |
| ogra_ | we should have a Touch/LandingTeam namespace on the wiki where such stuff lives :) | 19:10 |
| balloons | fginther, can you add pyflakes and pep8 to music-app requirements now? Just pushed a merge getting everything running and conforming | 19:14 |
| lool | if someone wants to test mir in PPA, apt-get install apt-get install libmirclient3 libmirplatform libmirprotobuf0 libunity-mir1 libplatform-api1-hybris libubuntu-application-api-mirclient1:armhf libubuntu-application-api-mirserver1:armhf libubuntu-application-api1:armhf libubuntu-platform-hardware-api1:armhf | 19:14 |
| lool | balloons: is python used just during build/ | 19:15 |
| lool | ? | 19:15 |
| balloons | lool, the autopilot tests are python :-) | 19:15 |
| balloons | so yes | 19:15 |
| fginther | balloons, I should get to it soon | 19:18 |
| balloons | no rush at all | 19:18 |
| fginther | kgunn, Saviq, are the indicators tested as part of the unity8 tests on touch? | 19:19 |
| Saviq | fginther, not really | 19:19 |
| Saviq | fginther, alesage was working on indicator tests | 19:19 |
| fginther | Saviq, ok, that's fine | 19:19 |
| Saviq | fginther, we had a few (like 4 to start with) | 19:20 |
| Saviq | fginther, but they needed a rework with the new indicators and all | 19:20 |
| lool | balloons: haha right | 19:23 |
| lool | balloons: stupid me | 19:23 |
| balloons | lool, wondering how I had managed to make pyflakes relevant to qml eh? | 19:26 |
| lool | balloons: fearing someone was trying to bundle python in a click for the first time! ;-) | 19:27 |
| bfiller | fginther: looks like the otto tests are failing on all recent address-book-app MR's. here are a few https://code.launchpad.net/~renatofilho/address-book-app/fix-1230404/+merge/188473 | 19:41 |
| bfiller | fginther: https://code.launchpad.net/~renatofilho/address-book-app/preview/+merge/190204 | 19:42 |
| fginther | bfiller, I see the problem, a prior run didn't cleanup all the way (for both). I'll re-run these | 19:47 |
| bfiller | fginther: thanks | 19:47 |
| rsalveti | lool: just added a new entry for the hybris/gst related changes to add support for software rendering | 20:22 |
| rsalveti | aka video playback using the browser and initial support for thumbnail | 20:22 |
| rsalveti | if possible would like to land this today (in a few hours) | 20:23 |
| lool | rsalveti: I might have to defer this one | 20:31 |
| lool | I have a super long list of landings to do now | 20:31 |
| Saviq | lool, https://code.launchpad.net/~dandrader/unity-mir/fix_lp1234600/+merge/190190 and https://code.launchpad.net/~dandrader/ubuntu-keyboard/fix_lp1234600/+merge/190189 | 20:37 |
| Saviq | lool, will bring up webbrowser tests to 100% | 20:37 |
| lool | Saviq: just heard that in the Mir standup, cool! | 20:39 |
| lool | Saviq: I'm landing all of this this night I hope | 20:39 |
| lool | rsalveti: how does one test the powerd change? | 20:40 |
| lool | rsalveti: binary is in PPA in case you want to test | 20:40 |
| rsalveti | lool: I tested it already, you need to have a n4, suspend for a few seconds and then try to resume it again | 20:41 |
| rsalveti | the issue was happening when resuming it | 20:41 |
| rsalveti | causing a huge slow down | 20:41 |
| rsalveti | that package fixes it | 20:41 |
| rsalveti | but I tested it already | 20:41 |
| rsalveti | lool: I'll keep testing the gst changes, but let me know before going to bed | 20:42 |
| lool | rsalveti: ok | 20:43 |
| lool | rsalveti: BTW the powerd change, does it help with resuming after a long time, resuming immediately, or resuming with a finger on the screen types of bugs? | 20:45 |
| lool | rsalveti: we weren't sure which one it fixed | 20:46 |
| rsalveti | lool: resuming with a finger on the screen after a few minutes, so you know it was in the deep suspend state | 20:46 |
| sergiusens | lool, how's the mir stuff progressing? | 20:47 |
| lool | ok | 20:47 |
| lool | sergiusens: tons of fixes landing now | 20:47 |
| sergiusens | lool, ack, will you ping me when we can start again? | 20:48 |
| plars | lool, Saviq: so I've had some limited success from turning on the display and disabling powerd, in once case I even got 7 passes from the unity8 tests, but 2 of the times I've run it hung midway through the tests | 20:48 |
| Saviq | plars, https://code.launchpad.net/~saviq/unity8/remove-socket-in-ap-tests/+merge/190175 should help | 20:49 |
| Saviq | plars, assuming it hangs due to the socket being there | 20:49 |
| Saviq | plars, it's in review right now | 20:49 |
| lool | sergiusens, rsalveti: I'll try, but list is long | 20:49 |
| lool | it might be just tomorrow | 20:49 |
| plars | Saviq: will killing powerd still be necessary with that fix in too? | 20:50 |
| sergiusens | lool, ack; my biggest concern is that no one is communicating what happens after tomorrow | 20:50 |
| vila | Mirv: seems like the issue withhttp://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2465/console was indeed linked to dbus/apparmor, the fix is in lxc on its way to land for saucy | 20:52 |
| rsalveti | lool: you could either test it yourself or trust at my test results :-) | 20:52 |
| Saviq | plars, yes | 20:52 |
| vila | now, if someone can give me a crash course on how to read http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html | 20:52 |
| Saviq | plars, that only deals with the stale socket being left if unity8 dies | 20:52 |
| vila | specifically the lxc part ;) | 20:52 |
| plars | Saviq: this is pretty small, I'll patch mine locally and see if it helps | 20:53 |
| Saviq | plars, gimme a sec, need to push a fix | 20:54 |
| lool | sergiusens: tomorrow is like in a super long time | 20:55 |
| plars | Saviq: it's very encouraging that the first line of the fix is "+ # FIXME: we shouldn't be doing this" :) | 20:55 |
| lool | sergiusens: Basically, until we have until tomorrow evening to promote an image which is shipable as a base image where you can pick up calls and do OS updates; then we have one week to update it for the worst bugs | 20:55 |
| plars | Saviq: I think from what I've heard before though, that leaving that stale socket behind isn't limited to the autopilot tests right? | 20:56 |
| lool | sergiusens: right now we've stopped the line to get Mir working better; mainly input issues | 20:56 |
| lool | but also crashes | 20:56 |
| lool | rsalveti: for powerd? if you look at spreadsheet I've noted "rsalveti tested" and will land it like that | 20:57 |
| lool | rsalveti: it's just a note to blame you if it's bad ;-) | 20:57 |
| Saviq | plars, of course | 20:57 |
| rsalveti | lool: that's fine :-) | 20:57 |
| Saviq | plars, but "real life" shouldn't see it | 20:57 |
| Saviq | plars, and a reboot helps as a last resort | 20:57 |
| Saviq | plars, but of course - we need to fix it for real - that's just a workaround - hence the FIXME | 20:58 |
| plars | right | 20:58 |
| fginther | bfiller, those two address-book-app MPs passed otto on the re-run | 20:58 |
| fginther | MPs were updated | 20:58 |
| fginther | and the problem that caused them to fail has been fixed | 20:58 |
| sergiusens | lool, ok, so gallery app, mediaplayer and camera not being click; are those bugs? | 20:59 |
| lool | sergiusens: In the grand scheme of things, they are lower priority ones | 20:59 |
| sergiusens | lool, I'm guessing that using the ppa for apps is a huge bug though :-) | 20:59 |
| lool | sergiusens: yes | 20:59 |
| lool | sergiusens: I personally care to drop the PPA, I'd also like to easily update them via store after release, and I also think there's a preview bug on .debs | 20:59 |
| bfiller | fginther: thanks | 21:00 |
| sergiusens | plars, what's this? http://jenkins.qa.ubuntu.com/job/saucy-touch_ro-maguro-smoke-ubuntu-rssreader-app-autopilot/117/artifact/clientlogs/application-click-com.ubuntu.shorts_shorts_0.2.log/*view*/ | 21:00 |
| plars | sergiusens: what do you mean? | 21:03 |
| sergiusens | plars, I get a garbaged file | 21:03 |
| plars | sergiusens: you mean the errors at the end? not familiar enough with rssreader tests, balloons might have an idea | 21:04 |
| plars | sergiusens: oh? mine has a couple of odd characters in it it looks like, but it very readable | 21:04 |
| sergiusens | plars, might be my firefox being crappy | 21:06 |
| sergiusens | plars, sorry for the noise | 21:06 |
| plars | sergiusens: odd, I'm on firefox too | 21:07 |
| plars | wow, that time I got 19 passes with unit8, let me try it with the patch enabled now | 21:07 |
| cjwatson | lool: landing ask 167 is the click change you asked about the other day | 21:08 |
| lool | cjwatson: Thanks | 21:09 |
| cjwatson | which reminds me, better get running ap tests on that | 21:10 |
| lool | cjwatson: if we had some for click scope! :-) | 21:11 |
| lool | I think we have for updater perhaps | 21:11 |
| cjwatson | well I'm under prior instruction to mindlessly run at least the unity8 tests | 21:11 |
| cjwatson | not seen any click-UI-related ap tests, indeed | 21:12 |
| lool | == Publishing powerd == | 21:14 |
| plars | Saviq: still seems to be getting stuck for me | 21:16 |
| Saviq | plars, got log output? | 21:17 |
| lool | fginther: http://10.97.0.26:8080/job/generic-land/18188/console | 21:20 |
| lool | fginther: AttributeError: 'NoneType' object has no attribute 'commit_message' | 21:20 |
| lool | but the branch has one | 21:20 |
| lool | I must have run this wrong | 21:20 |
| plars | Saviq: no, let me restart | 21:20 |
| Saviq | plars, it's interesting what unity8 says when stuck | 21:20 |
| Saviq | if anything | 21:20 |
| fginther | lool, yeah, check the parameters | 21:20 |
| plars | Saviq: from where? | 21:21 |
| fginther | lool, packaging_branch has the MP url | 21:21 |
| fginther | lool, this will almost all be over soon | 21:21 |
| Saviq | plars, autopilot run -v should show it | 21:21 |
| fginther | lool, will be enabling auto merge for just about everything | 21:21 |
| lool | ok, sorry | 21:22 |
| plars | Saviq: I can probably get the last bit at least, sec | 21:22 |
| plars | Saviq: http://paste.ubuntu.com/6215593/ | 21:22 |
| Saviq | plars, ah, that looks like it's stuck on exit | 21:23 |
| Saviq | plars, if apport is not collecting the core, it must be looping somewhere | 21:23 |
| kgunn | lool: if you're still on ...ask sheet now has update for unity8 ap test to delete the stale mir socket | 21:30 |
| kgunn | merged on unity8 trunk | 21:31 |
| lool | kgunn: 10 revisions landed in unity8 trunk though | 21:33 |
| kgunn | lool but trunk is sacred to us...so they are all good | 21:36 |
| lool | == Publishing qtubuntu == | 21:38 |
| lool | kgunn: many dont relate to mir at all and we've stopped the line to fix mir stuff :-/ | 21:38 |
| lool | == Building services stack (to pickup ubuntu-keyboard) == | 21:41 |
| plars | Saviq: so I'm able to reproduce this scenario most of the time I run it, the process it's trying to kill is unity8 of course, and there is no apport running | 21:43 |
| plars | Saviq: but at least this time, it looks like maliit-server is eating 95% of the cpu | 21:43 |
| plars | 95+ | 21:43 |
| kgunn | lool: so this is nuts, we don't have a branch to put all this on to | 21:43 |
| Saviq | plars, right, I expect this to actually be the case every time - forgot about that | 21:44 |
| plars | Saviq: is there some fix coming? or a workaround at least? | 21:44 |
| kgunn | lool: team was told this morning things can keep moving per our asac discussion | 21:44 |
| sergiusens | lool, fginther would it be safe for these? https://code.launchpad.net/~sergiusens/dropping-letters/click/+merge/190228 https://code.launchpad.net/~sergiusens/sudoku-app/click/+merge/190231 https://code.launchpad.net/~sergiusens/stock-ticker-mobile-app/click/+merge/190238 ? | 21:44 |
| sergiusens | they fix apparmor issues | 21:44 |
| kgunn | lool: consider this...indicator for sim pin entry isn't mir related....you wouldn't take that fix ??? | 21:45 |
| kgunn | lool: only fixes are coming in | 21:45 |
| lool | kgunn: Yes | 21:45 |
| Saviq | plars, not yet, we'll need to see where it's getting stuck | 21:46 |
| fginther | sergiusens, those should be automerged in a moment, I had the list wrong | 21:48 |
| fginther | sergiusens, lool, those projects are all ok to merge now | 21:49 |
| lool | kgunn: yeah you're right | 21:50 |
| lool | sergiusens: did you do the AP runs with the click setup? | 21:51 |
| lool | fginther: \o/ | 21:51 |
| sergiusens | lool, these apps don't have ap test; oh wait dropping letters does, give me a sec | 21:52 |
| lool | sergiusens: oh that's good, no change I get slapped for regressed APs then! ;-) | 21:55 |
| lool | sergiusens: so if you could give them a quick run test, that would be good | 21:55 |
| sergiusens | lool, I manually ran them, but give me a sec for drop letters | 21:55 |
| Saviq | plars, can you file a bug against maliit for that? | 21:57 |
| Saviq | plars, I'll try and get more info into it | 21:57 |
| plars | Saviq: is it related to https://bugs.launchpad.net/mir/+bug/1233988 | 21:57 |
| ubot5 | Ubuntu bug 1233988 in Mir "With Mir enabled: maliit-server crashed with SIGABRT in __gnu_cxx::__verbose_terminate_handler(), thrown from mir::client::DisplayConfiguration::copy_to_client()" [Medium,Triaged] | 21:57 |
| lool | plars: we have an upcoming keyboard | 21:57 |
| lool | plars: update | 21:57 |
| lool | plars: it has changes, notably crash fixes; not sure about CPU usage | 21:58 |
| plars | Saviq: I actually have to run away right now, let me try it after the keyboard fix that lool is talking about (assuming that's in the next image right?) when I get back | 21:58 |
| Saviq | plars, that's a crash, not a loop | 21:58 |
| Saviq | plars, sure, take care | 21:58 |
| lool | plars: it will be in PPA in some hours | 21:59 |
| lool | then in image htis night | 21:59 |
| sergiusens | lool, fginther I added the test results in the description of https://code.launchpad.net/~sergiusens/dropping-letters/click/+merge/190228 | 22:03 |
| lool | sergiusens: cool | 22:04 |
| lool | sergiusens: adding landing slot | 22:05 |
| sergiusens | fginther, lool, also a merge from three hours ago removed my manifest.json (merge issue), this MR just readds it https://code.launchpad.net/~sergiusens/music-app/manifest/+merge/190250 | 22:05 |
| sergiusens | @ci | 22:05 |
| * sergiusens sees no help | 22:05 | |
| lool | sergiusens: but music-app doesn't need to land | 22:06 |
| lool | it might land as a .deb with some unrelated changes from our pov | 22:07 |
| sergiusens | lool, yeah, but this is a regression (in trunk albeit) still a regression | 22:07 |
| sergiusens | lool, I'm still building the music app in parallel and would be completely broken without this | 22:08 |
| lool | sergiusens: it's only a regression to make it a click, right | 22:08 |
| sergiusens | lool, yes | 22:08 |
| lool | sergiusens: it's ok, just clarifying it will be visible but wont impact us | 22:08 |
| sergiusens | lool, for a deb you can consider it a readme ;-) | 22:08 |
| sergiusens | in source readme | 22:09 |
| lool | right | 22:11 |
| lool | cjwatson: Hey, would you mind scoring this amd64 rebuild of mir in my PPA up? https://launchpad.net/~lool/+archive/ppa/+build/5089423 retrying the build in case the tests are flaky as upstream said it might pass | 22:37 |
| lool | it failed in the daily-build ppa but I'm not member of the team, so can't bump it there | 22:37 |
| cjwatson | lool: done | 22:38 |
| lool | cjwatson: thanks | 22:38 |
| cjwatson | though shouldn't we just retry it in the daily-build ppa directly? | 22:38 |
| cjwatson | (I can) | 22:38 |
| lool | cjwatson: that would be nice too | 22:39 |
| lool | cjwatson: I'm not a member so couldn't do that | 22:39 |
| cjwatson | your ppa is probably virtualised too, which is different | 22:39 |
| lool | cjwatson: I'll have to ping didrocks to get added tomorrow | 22:39 |
| lool | cjwatson: ah true | 22:39 |
| cjwatson | I'm probably not either but I have buildd-admin superpowers :) | 22:39 |
| lool | :-) | 22:40 |
| cjwatson | it's building now | 22:40 |
| lool | plars: do you know how to run the ubuntu-keyboard autopilot? | 22:50 |
| lool | cjwatson: it indeed built | 22:51 |
| sergiusens | fginther, weren't those MRs I put out supposed to be merging by now? | 22:51 |
| thomi | lool: I believe the person to ask is veebers - he wrote those tests | 22:55 |
| lool | thomi: thanks, it seems to be a good time too | 22:56 |
| lool | lol, cu2d is waiting on arm64 builds | 23:24 |
| lool | cjwatson: do you have a way to disable arm64 in that PPA? | 23:24 |
| lool | seems like a waste | 23:25 |
| cjwatson | lool: we're building arm64 in the primary archive, and that PPA copies stuff into the primary archive | 23:30 |
| cjwatson | cu2d should skip it though | 23:30 |
| cjwatson | and no, I don't, you'd need an admin if it's actually the right thing to do | 23:30 |
| cjwatson | I think you're mistaken as to the cause though - https://launchpad.net/~ubuntu-unity/+archive/daily-build/+builds shows no arm64 builds | 23:31 |
| cjwatson | maybe cu2d is waiting incorrectly | 23:31 |
| lool | cjwatson: ah yes, it's probably waiting indefinitely on arm64 builds | 23:31 |
| lool | on one of them | 23:32 |
| lool | indicator-messages it was it was saying | 23:32 |
| lool | doesn't seem to have a build record | 23:32 |
| cjwatson | it would surprise me if any ppas were enabled for arm64 yet, certainly | 23:33 |
| lool | Bah, entangled in upstart-app-launch now | 23:35 |
| fginther | sergiusens, fixed it | 23:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!