=== thumper is now known as thumper-afk | ||
vthompson | I want to get the Unity 8/Ubuntu Desktop Next session running on an armhf device (RPi 2). Sadly, this means I can not install Unity 8 via the LXC container instructions because there is no armhf iso. If I install the "ubuntu-desktop-next" package, should I assume I might be able to use the Unity 8 session? | 02:06 |
---|---|---|
=== Bl4ckD34Th is now known as zz_Bl4ckD34Th | ||
=== thumper-afk is now known as thumper | ||
pitti | Noskcaj: yes; apt is rather resistant when it comes to uninstalling a package; there must be several new dependencies to the new one, and the new one must at least Provides: the old one, but that's not the case here | 06:30 |
pitti | as I said, it might be that apt became more clever in the last year or so, but an upgrade needs to be tested | 06:30 |
seb128 | good morning desktopers | 06:40 |
pitti | bonjour seb128 | 06:41 |
seb128 | hey pitti :-) | 06:41 |
pitti | so I got up late (we had a concert last night), you got up early, almost the same time now :) | 06:42 |
pitti | convergence! | 06:42 |
seb128 | hehe | 06:46 |
darkxst | so pitti can plug seb128 into the TV and get Ubuntu desktop then ;) | 06:48 |
darkxst | hi all, btw | 06:49 |
didrocks | good morning | 07:25 |
seb128 | hey didrocks | 07:27 |
didrocks | hey seb128 | 07:27 |
pitti | bonjour didrocks | 07:58 |
didrocks | guten morgen pitti | 07:59 |
larsu | hi pitti! | 07:59 |
pitti | larsu: moin moin! | 07:59 |
=== \b is now known as benonsoftware | ||
mlankhorst | morning | 08:24 |
didrocks | hey mlankhorst | 08:26 |
mlankhorst | hey | 08:29 |
seb128 | hey mlankhorst | 08:30 |
TheMuso | Hey willcooke. :) | 08:59 |
willcooke | morning | 08:59 |
Laney | yo | 09:04 |
seb128 | hey TheMuso Laney willcooke | 09:05 |
willcooke | hi all | 09:05 |
willcooke | oh, I just remembered - I'm on holiday tomorrow | 09:06 |
willcooke | and in London on Monday | 09:06 |
willcooke | larsu, cancelled our meeting tomorrow ^^^ | 09:06 |
willcooke | let me know if you need anything | 09:06 |
larsu | willcooke: morning & thanks, will do | 09:07 |
larsu | willcooke: and enjoy you day off ;) | 09:07 |
willcooke | \o/ looking forward to it | 09:08 |
Laney | ooh London, glamorous | 09:08 |
willcooke | I'll send you a postcard of a telephone box Laney | 09:09 |
didrocks | tell us if Doctor Who is around :) | 09:09 |
Laney | go check out Cereal Killer | 09:09 |
davmor2 | Laney: we are all Cereal Killers otherwise they wouldn't sell it in boxes | 09:11 |
Laney | darkxst: thx for uploading | 09:42 |
Laney | seb128: can you check grilo-plugins in vivid new when you've got some minutes today please? | 09:43 |
Laney | (then try totem 3.14 if you have a few more :p) | 09:43 |
seb128 | Laney, sure can | 09:43 |
seb128 | what should I try on totem exactly? | 09:43 |
Laney | random use | 09:43 |
Laney | desrt: thx, saw it, replied on the list | 10:02 |
Laney | I should filter that one out of my main inbox | 10:02 |
seb128 | Laney, totem looks fine to me | 10:09 |
Laney | cool | 10:09 |
Laney | thanks to darkxst and Noskcaj for doing most of the work | 10:10 |
seb128 | Noskcaj, darkxst, good work! | 10:12 |
seb128 | it's a bit weird that the Videos tab is empty by default | 10:12 |
seb128 | would be nice if is was listing the content of XDG_VIDEOS_DIR | 10:13 |
seb128 | also the plugins list in preference is empty | 10:13 |
seb128 | is that normal? | 10:13 |
Laney | has some here | 10:19 |
Laney | is totem-plugins installed? | 10:19 |
seb128 | ups, that got uninstalled | 10:22 |
seb128 | Laney, thanks ;-) | 10:22 |
larsu | Laney: I can't install it because it wants grilo-plugins-0.2-base, which isn't available | 10:22 |
seb128 | Laney, thanks | 10:23 |
Laney | they're coming soon, or you can grab from https://launchpad.net/ubuntu/+source/grilo-plugins/0.2.13-3ubuntu3 | 10:23 |
larsu | thanks | 10:23 |
seb128 | larsu, I just NEWed those binaries, should be available after the next publisher run | 10:23 |
seb128 | Laney, ^ | 10:23 |
Laney | will need to wangle through proposed | 10:23 |
Laney | thanks! | 10:23 |
larsu | seb128: thanks! I'll just wait a bit then | 10:24 |
seb128 | hum | 10:24 |
seb128 | can't connect to people.canonical.com? | 10:24 |
Laney | some kind of maintenance | 10:25 |
seb128 | yeah, saw that in the #is topic | 10:26 |
seb128 | I wanted to look at why glib didn't migrate out of proposed yet | 10:26 |
Laney | gnome-photos failure | 10:27 |
Laney | should indeed be retryable with this NEWing | 10:27 |
Laney | to do with https://launchpad.net/ubuntu/+source/gnome-online-miners/3.14.0-2ubuntu1 | 10:28 |
seb128 | that doesn't seem related to glib? | 10:29 |
seb128 | why do we block on it rather than overriding? | 10:29 |
Laney | was fixing it | 10:30 |
seb128 | k, the logic still feels weird to me, but as long as it's fixed it's all good :-) | 10:31 |
Laney | I always prefer to fix tests if reasonably possible even if they're not directly caused by the thing they are blocking | 10:32 |
seb128 | well that's orthogonal | 10:32 |
seb128 | you can unblock and still fix the test :-) | 10:32 |
flexiondotorg | Morning. There has been a recent release of GTK2 which fixes a nasty bug. Is this where the right channel to progress the fix in Ubuntu? | 10:51 |
Laney | Yeah, we're going to get that soon | 10:53 |
ochosi | flexiondotorg: rly? what bug? | 10:57 |
flexiondotorg | ochosi, https://bugs.launchpad.net/ubuntu/+source/mate-control-center/+bug/1351890 | 10:59 |
ubot5 | Launchpad bug 1351890 in ubuntu-mate "Changing external screen resolution with dual monitors" [Medium,Triaged] | 10:59 |
ochosi | ah that one, ok | 10:59 |
flexiondotorg | ochosi, Although our motivation to prepare the patch was to fix a segfault in MATE, other GTK2 applications are affected. | 10:59 |
flexiondotorg | ochosi, Upstream release now. | 10:59 |
flexiondotorg | Laney, Excellent. Thanks. | 11:00 |
=== MacSlow is now known as MacSlow|lunch | ||
attente_ | seb128: hi, do you have time to look at https://code.launchpad.net/~attente/unity-settings-daemon/fcitx-transition/+merge/230289 and https://code.launchpad.net/~attente/unity-control-center/fcitx-transition/+merge/249523 | 12:12 |
=== alan_g is now known as alan_g|lunch | ||
seb128 | attente_, I can try to have a look today yes | 12:46 |
attente_ | seb128: thanks! | 12:47 |
seb128 | yw! | 12:47 |
seb128 | attente_, btw I did follow up that work recently, did anything change compared to previous cycle? should still be a no-op for ibus users riht? | 12:48 |
attente_ | seb128: yes, should only improve our situation with fcitx | 12:49 |
seb128 | attente_, my ibus is not working, I think it's still like the ubuntu-keyboard thing, do we have a bug open about that? | 12:56 |
attente_ | seb128: yeah, it's because of this: https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1245925 | 12:57 |
ubot5 | Launchpad bug 1245925 in maliit-framework (Ubuntu) "Troublesome export in /etc/profile.d/maliit-framework.sh" [Undecided,Confirmed] | 12:57 |
attente_ | the workaround for now is to comment out the export of QT_IM_MODULES, or remove maliit-framework | 12:58 |
attente_ | sorry, QT_IM_MODULE | 12:58 |
Laney | purge | 12:59 |
attente_ | ah, right. purge, sorry :) | 12:59 |
seb128 | Laney, ? | 12:59 |
seb128 | oh | 12:59 |
seb128 | that's not a solution | 12:59 |
Laney | he called it a workaround | 12:59 |
seb128 | yeah, I'm not interested in a workaround | 12:59 |
seb128 | I was talking about fixing that issue | 13:00 |
seb128 | in the distro | 13:00 |
Laney | yes indeed | 13:01 |
attente_ | one thing we can do is make im-config still set GTK_IM_MODULE and XMODIFIERS, but it doesn't fix the fact that we need QT_IM_MODULES to have both maliitphablet and ibus/fcitx | 13:02 |
Laney | can it have multiple modules? | 13:02 |
Laney | (does stuff work right then?) | 13:02 |
attente_ | don't think so. GTK_IM_MODULES can, but even then, it's just like going through a list of fallbacks until it picks one | 13:03 |
Laney | how does maliit get used for gtk apps? | 13:04 |
Laney | could we kill this file and teach im-config about maliit, then get $phone_place to call it? | 13:05 |
attente_ | i guess we could probably do that without teaching im-config about maliit | 13:06 |
* Laney doesn't know how it works | 13:06 | |
attente_ | but it needs to set the environment variable across the session | 13:07 |
Laney | ye, some upstart job | 13:09 |
Laney | well anyway, just throwing out ideas :) | 13:10 |
seb128 | well, I'm not even sure what should happen | 13:15 |
attente_ | is there a way to know when we change from having a physical keyboard to not having one? | 13:16 |
seb128 | it feels like the osk shouldn't be by session | 13:16 |
seb128 | but by input device | 13:16 |
seb128 | like usng a touch screen should display one | 13:16 |
Laney | can the toolkits do that? | 13:16 |
seb128 | if you connect a bt keyboard to your phone it shouldn't display the osk then | 13:16 |
seb128 | not sure | 13:16 |
seb128 | I've feeling we are not going to resolve those usecases under xorg/unity7 though | 13:17 |
seb128 | so meanwhile having a way to turn osk on or off, like we do with onboard, would be something | 13:17 |
Laney | perhaps we just say that we don't use maliit on unity 7 or something :| | 13:19 |
seb128 | that wfm | 13:22 |
seb128 | well, until we figure how we deal better with that | 13:22 |
seb128 | I'm not even sure ubuntu-keyboard works fine on unity7 | 13:22 |
Laney | the upstream one did at least work when I packaged it ages ago | 13:23 |
Laney | seb128: do we do "= unity8-* || = ubuntu-touch" or "!= ubuntu"? | 13:26 |
Laney | attente_: ^ | 13:26 |
seb128 | something is wrong in any case, because it if was taking over ibus, it should display its UI when trying to im? | 13:26 |
Laney | if you fix it like this, it would be good to also make that file a noop if maliit isn't installed | 13:26 |
seb128 | Laney, I would say unity8 | ubuntu-touch | 13:27 |
attente_ | yeah | 13:28 |
Laney | ok, attente_ are you happy to do that? | 13:29 |
attente_ | Laney: sure | 13:29 |
Laney | can review after that | 13:29 |
Laney | sweet | 13:29 |
Laney | lemme know | 13:29 |
seb128 | attente_, can you get desrt to look again to your indicator-keyboard changes and approve it if it's fine, if he is I can do a landing | 13:30 |
attente_ | seb128: sure | 13:30 |
seb128 | thanks | 13:30 |
* Laney pulls ze totem trigger | 13:44 | |
Laney | chuk-chuk boom | 13:44 |
attente_ | this script doesn't even seem to run when i start a unity 8 session... | 13:51 |
seb128 | likely some Xsession.d hackery | 13:53 |
seb128 | or X hacks don't apply to Mir | 13:53 |
attente_ | so how does maliit currently work on the phone? | 13:53 |
attente_ | the phone runs the profile.d scripts but the desktop doesn't? | 13:55 |
seb128 | correct | 13:56 |
seb128 | http://bazaar.launchpad.net/~phablet-team/ubuntu-touch-session/trunk/view/head:/ubuntu-touch-session | 13:56 |
seb128 | that the ubuntu-touch-session script | 13:56 |
seb128 | which is some hackery used on the device | 13:56 |
seb128 | unity8 desktop doesn't have those hacks | 13:57 |
seb128 | the script sources profile.d | 13:57 |
attente_ | maybe this is what we need to disable in u7 | 13:57 |
seb128 | that's one of those things we should deprecate by integrating what it does in the proper packages | 13:57 |
seb128 | you mean? | 13:57 |
seb128 | we are not going to install that script on unity7 if that's what you suggest :-) | 13:58 |
seb128 | it does things like forcing the qt qpa to mir | 13:58 |
seb128 | or has android bits | 13:58 |
seb128 | we should do it the other way around, move things out of this script to the proper packages instead | 13:58 |
attente_ | yeah | 13:59 |
attente_ | but the script is running on u7 right now... | 14:01 |
attente_ | no. i'm wrong, something else is running the /etc/profile.d scripts under u7 | 14:04 |
seb128 | well, profile.d scripts are run through Xsession.d | 14:05 |
attente_ | oh, ok | 14:05 |
=== MacSlow|lunch is now known as MacSlow | ||
seb128 | hum, in fact I don't have that package/script | 14:06 |
seb128 | I wonder why my ibus is not working | 14:07 |
seb128 | cyphermox, hey, can you get https://code.launchpad.net/~mathieu-tl/mtp/lp1421664/+merge/249669 landed? | 14:10 |
cyphermox | sure | 14:10 |
seb128 | thanks | 14:10 |
cyphermox | happy new year btw ;) | 14:10 |
seb128 | cyphermox, I would try to write you that in chinese but my ibus is not working :p | 14:10 |
cyphermox | ahahha | 14:11 |
cyphermox | I could write it in vietnamese but for the same reasons | 14:11 |
attente_ | seb128: what does your ~/.xinputrc say | 14:12 |
cyphermox | there's a bit fewer special accents to add that need ibus | 14:12 |
seb128 | attente_, I deleted that file like 10 minutes ago, but it was empty | 14:12 |
attente_ | can you try 'im-config -n ibus' and restart the session? | 14:12 |
seb128 | why is that needed? | 14:12 |
seb128 | how does it work with new users/installs? | 14:12 |
attente_ | yeah, you're right, it shouldn't be needed | 14:13 |
seb128 | let me restart session in case anyway | 14:13 |
desrt | seb128: i really can't ACK those changes | 14:13 |
seb128 | desrt, why not? | 14:13 |
desrt | because i don't know anything about input methods or indicators :) | 14:14 |
seb128 | are you happy with the code? | 14:14 |
desrt | i can review that part of it again :) | 14:15 |
seb128 | thanks | 14:15 |
seb128 | I can test the IM/indicator side | 14:15 |
attente_ | i can look for another reviewer | 14:15 |
attente_ | or happyaron, do you know someone else? ^ | 14:16 |
seb128 | 你好 | 14:18 |
attente_ | :D | 14:19 |
seb128 | attente_, works after deleting that .xinput rc which had comments only | 14:19 |
seb128 | maybe that was preventing im-config to do his init or something | 14:19 |
attente_ | ah, maybe | 14:19 |
tedg | larsu, Can you look at https://code.launchpad.net/~ted/indicator-messages/lp1385331-unescape-message-ids/+merge/250235 | 14:23 |
attente_ | Laney, seb128, can you guys test https://code.launchpad.net/~attente/maliit/1245925/+merge/250311 on the device? | 14:25 |
attente_ | to make sure osk still works | 14:25 |
attente_ | i can't test it since the behaviour is different on the desktop | 14:25 |
seb128 | attente_, I can do that | 14:25 |
GunnarHj | pitti: Hi Martin! Saw that https://translations.launchpad.net/ubuntu/trusty/+language-packs seems to be ready. Do you have time to get it into -proposed today? | 14:31 |
larsu | tedg: sure, give a few mins | 14:33 |
larsu | *me | 14:33 |
pitti | GunnarHj: yes, it's on my list; I have to wait for the currently running vivid langpack build to finish, though | 14:37 |
pitti | GunnarHj: I'm watching it on a foreground terminal :) | 14:37 |
GunnarHj | pitti: Ok, great! :) | 14:37 |
seb128 | hum | 14:46 |
seb128 | so my desktop doesn't boot anymore | 14:46 |
seb128 | stucked on the plymouth logo | 14:47 |
seb128 | it boots fine if I remove quiet splash though | 14:47 |
seb128 | didrocks, pitti, ^ know of any issue with recent systemd updates? | 14:47 |
didrocks | seb128: didn't upgdate since yesterday, can do if you want | 14:48 |
didrocks | update* | 14:48 |
didrocks | seb128: actually, I'm running latest systemd already, no issue at boot | 14:49 |
didrocks | (did reboot just after latest upgrade) | 14:49 |
pitti | GunnarHj: I'm asking in #ubuntu-release about when to upload them; I figure having them in trusty-proposed should be fine, but I'll double-check | 14:50 |
pitti | seb128: I'm not aware of regressions; could you boot with systemd.debug-shell, and once it's hanging switch to VT9 and check systemctl list-jobs? | 14:50 |
seb128 | pitti, k | 14:50 |
seb128 | looking to the journal those boots have different fails | 14:50 |
seb128 | pitti, http://people.canonical.com/~seb128/log | 14:51 |
GunnarHj | pitti: ok | 14:51 |
seb128 | févr. 19 15:45:12 seb-e6410 systemd[1]: Starting Network Manager... | 14:51 |
seb128 | févr. 19 15:46:02 seb-e6410 systemd[1]: Started Console System Startup Logging. | 14:51 |
seb128 | why that lag? | 14:51 |
seb128 | I guess I need debug as well | 14:52 |
larsu | n conjugating directly for voice, English uses the past participle form of the verb plus an auxiliary verb, either be or get, to indicate passive voice. | 14:56 |
* larsu accidentally pasted. Sorry | 14:56 | |
pitti | darn, I missed the start of the English lesson | 14:56 |
* larsu hates middle-click paste | 14:56 | |
larsu | pitti: reading wikipedia about passive voice :) | 14:56 |
pitti | o_O middle click paste == ♥ | 14:57 |
seb128_ | +1 | 14:57 |
seb128_ | pitti, no vt, including the vt9 one :/ | 14:58 |
pitti | meh | 14:58 |
pitti | the 45 s lag is certainly odd, but it doesn't say anything in that log at that time | 14:59 |
pitti | févr. 19 15:46:27 seb-e6410 dbus[774]: [system] Failed to activate service 'org.freedesktop.PolicyKit1': timed out | 14:59 |
pitti | févr. 19 15:46:27 seb-e6410 dbus[774]: [system] Failed to activate service 'org.freedesktop.systemd1': timed out | 14:59 |
pitti | févr. 19 15:46:27 seb-e6410 dbus[774]: [system] Failed to activate service 'org.freedesktop.ColorManager': timed out | 14:59 |
seb128_ | right, well it boots in 10s without quiet | 14:59 |
pitti | seb128_: that's also weird -- after that it shuts down, did you request that manually, or does it happen automatically? | 15:00 |
seb128_ | I try one with debug to see if it has more | 15:00 |
seb128_ | I do ctrl-alt-f1 and ctrl-alt-del | 15:00 |
pitti | ah, so manual shutdown | 15:03 |
seb128_ | yes | 15:03 |
seb128 | pitti, hum | 15:04 |
seb128 | févr. 19 15:59:09 seb-e6410 systemd[1]: Starting Run Click system-level hooks... | 15:04 |
seb128 | févr. 19 15:59:59 seb-e6410 systemd[1]: Received SIGINT. | 15:04 |
pitti | seb128: urk :) you might not want to install weird stuff | 15:05 |
seb128 | that runs on non debug runs as well and doesn't hang the boot | 15:05 |
seb128 | I'm unsure how to debug that :-/ | 15:05 |
seb128 | systemd.debug-shell doesn't give me a vt9 | 15:06 |
pitti | seb128: so with plymouth is it just taking "long" or actually hanging forever? does it get far enough to ssh in and see what's happening? | 15:06 |
seb128 | it seems to hang forever | 15:06 |
seb128 | but I didn't wait for more than 3 minutes | 15:06 |
desrt | attente_: this patch is pretty great | 15:06 |
pitti | seb128: you mean VT switching doesn't work? or it doesn't even boot far enough after the initramfs? | 15:06 |
seb128 | but a boot without quiet takes 15 seconds | 15:06 |
desrt | step 1) attente copy/pastes something from one part of the code to another part | 15:07 |
seb128 | pitti, ctrl-alt-fn display empty screens | 15:07 |
desrt | step 2) desrt flags in review "this is not necessary" | 15:07 |
seb128 | no command line | 15:07 |
desrt | step 3) attente fixes all of the other instances of same | 15:07 |
seb128 | it just switches away from the plymouth logo | 15:07 |
desrt | step 4) massive net-negative diff | 15:07 |
pitti | seb128: yeah, 90s is the standard timeout, so anything which takes longer can safely count as "forever" | 15:07 |
pitti | desrt: sounds like step 5) hug attente ? | 15:07 |
desrt | srsly. | 15:07 |
willcooke | Sweet5hark1, quick easy way to reset LO settings to default? Deleting something from .config perhaps? | 15:08 |
* Sweet5hark1 is in call | 15:08 | |
willcooke | nw | 15:08 |
pitti | ./.config/libreoffice ? | 15:08 |
Sweet5hark1 | willcooke: "rm -rf ~/.config/libreoffice" should reset your user profile | 15:08 |
willcooke | thanks pitti Sweet5hark1 | 15:08 |
seb128 | pitti, I could try to boot with systemd-bootchart I guess | 15:09 |
pitti | seb128: yeah, good first step | 15:09 |
pitti | seb128: also, do you have plymouth in the initramfs? | 15:09 |
pitti | zcat /initrd.img | cpio -t|grep ply | 15:09 |
seb128 | pitti, how do I tell? | 15:09 |
seb128 | yes | 15:09 |
pitti | seb128: ok, so you have cryptsetup installed probably? | 15:10 |
seb128 | yes | 15:10 |
pitti | seb128: if you don't have encrypted internal partitions, you could try removing it, to compare | 15:10 |
seb128 | but I don't use it | 15:10 |
seb128 | k | 15:10 |
seb128 | let me try | 15:10 |
pitti | (cryptsetup-bin is enough for encrypted USB etc.) | 15:10 |
pitti | seb128: if you still see plymouth, that at least means that it gets far enough into the root system to at least boot | 15:10 |
pitti | seb128: another strategy: | 15:11 |
pitti | - enable persistant journal: sudo mkdir /var/lib/journal | 15:11 |
pitti | reboot with plymouth, wait two mins, reboot | 15:11 |
pitti | - reboot without plymouth, journalctl -b -1 | 15:11 |
seb128_ | pitti, I do have persistant journal, that's how I got you the boot log before | 15:11 |
pitti | ah | 15:11 |
pitti | seb128_: ah, so you boot with bootchart by default? | 15:12 |
seb128_ | pitti, ok, withoyt cryptsetup, stucked on plymouth with animated dots | 15:12 |
pitti | seb128_: so exactly the same? | 15:12 |
seb128_ | pitti, not voluntarily | 15:12 |
seb128_ | yes | 15:12 |
pitti | seb128_: voluntarily> you mean it accidentally boots with bootchart? | 15:13 |
seb128_ | I'm not using bootchart that I know* | 15:13 |
seb128_ | I use init=/lib/systemd | 15:13 |
seb128_ | sorry /bin | 15:13 |
pitti | févr. 19 15:46:34 seb-e6410 umount[1458]: umount: /dev/.bootchart/proc: target is busy | 15:13 |
pitti | ah, perhaps this is normal then, but it looks odd for sure | 15:14 |
seb128_ | I've the bootchart package installed | 15:14 |
pitti | seb128_: oh, you have an /etc/init.d/bootchart | 15:14 |
pitti | seb128_: I don't think I ever tested that with systemd | 15:14 |
seb128_ | guess so | 15:14 |
pitti | but that could very plausibly be the cause | 15:14 |
seb128_ | but that's not new | 15:14 |
seb128_ | ok | 15:14 |
seb128_ | pitti, after 2 minutes the plymouth logo went away | 15:15 |
pitti | seb128_: does that actually work? (or rather, did?) | 15:15 |
seb128_ | I've only an empty screen now | 15:15 |
pitti | hm, so why no VTs and debug shell; darn | 15:15 |
seb128_ | ohh | 15:15 |
seb128_ | I've a vt1 | 15:15 |
seb128_ | shrug | 15:15 |
seb128_ | trying to log in displayed some logind not starting error | 15:15 |
seb128_ | and bounced me to xfailsafe | 15:16 |
seb128_ | ah | 15:16 |
seb128_ | got my vt | 15:16 |
seb128_ | systemctl status hangs :-/ | 15:16 |
seb128_ | systemd is NOT happty | 15:16 |
pitti | ah, bootchart depends: upstart | 15:16 |
pitti | sounds like missing dbus | 15:17 |
seb128_ | "Failed to read server status: ... timeout" | 15:17 |
pitti | *nod* (no dbus) | 15:17 |
* pitti installs bootchart and upstart, uninstalls systemd-sysv, and boots with init=/bin/systemd | 15:18 | |
pitti | hm, that works | 15:18 |
seb128_ | I've virtualbox installed | 15:18 |
seb128_ | the line before the hangs on that box is virtualbox-guest-utils | 15:19 |
* pitti installs that, too | 15:20 | |
seb128_ | I don't understand why it would work without "quiet" though :/ | 15:21 |
pitti | yeah | 15:21 |
pitti | seb128_: the main plymouth related change in 219 was the addition of didrocks's fsckd which tries to connect to plymouth | 15:22 |
Laney | bleh, my system just hard locked | 15:22 |
seb128_ | Laney, one of those days!! | 15:23 |
pitti | seb128_: for experimentation, could you try sudo systemctl mask systemd-fsckd ? | 15:23 |
didrocks | can be the plymouth spam? | 15:23 |
seb128_ | "spam"? | 15:23 |
didrocks | seb128_: if plymouth goes down, it tried to reconnect | 15:23 |
seb128_ | pitti, I purge virtualbox-utils, was not it | 15:23 |
didrocks | tries* | 15:23 |
didrocks | but yeah, masking systemd-fsckd can help as pitti told | 15:24 |
seb128_ | back on plymouth stucked animating dots | 15:24 |
seb128_ | can I do that without dbus? | 15:24 |
ogra_ | "rolex replica now cheap - contact your plymouth daemon !! " | 15:24 |
seb128_ | or is that going to fail like systemctl status? | 15:24 |
pitti | wait | 15:24 |
ogra_ | "your plymouth won the lottery !!" | 15:24 |
* Laney sinbins ogra_ | 15:25 | |
ogra_ | :) | 15:25 |
pitti | seb128_: sudo ln -s /dev/null /etc/systemd/system/systemd-fsckd.service | 15:25 |
didrocks | you can mask the unit manually with the null symlink | 15:25 |
didrocks | yeah | 15:25 |
didrocks | seb128_: also, do you mind checking if fsck runs? | 15:25 |
seb128_ | didrocks, how? I can't access to a vt | 15:25 |
seb128_ | well, not before the timeout that makes plymouth go away | 15:25 |
didrocks | where did you type systemctl then? | 15:26 |
didrocks | ah ok | 15:26 |
didrocks | I doubt fsck would though run at every boot | 15:26 |
didrocks | (and so systemd-fsck should exits without contacting systemd-fsckd) | 15:26 |
didrocks | and thus plymouth | 15:26 |
pitti | didrocks: that might be it | 15:27 |
seb128_ | when plymouth timeout, going to vt7 I've | 15:27 |
pitti | didrocks: I do have it running after a clean boot in my VM | 15:27 |
didrocks | pitti: oh? | 15:28 |
seb128_ | [*** ] (3 of 4) A Start job is running for Wait for Plymouth Boot Screen to quit (2min 43s / no limit) | 15:28 |
seb128_ | where the * are red | 15:28 |
pitti | didrocks: it's still running actually | 15:28 |
seb128_ | the message was there on vt1 and changing | 15:28 |
seb128_ | before I tries to log in | 15:28 |
didrocks | pitti: but only first boot, right? | 15:28 |
didrocks | hum, weird | 15:28 |
didrocks | fsck did run? | 15:28 |
pitti | didrocks: no, I rebooted like 10 times | 15:28 |
seb128_ | no fsck running | 15:28 |
* didrocks is puzzled | 15:28 | |
didrocks | seb128_: systemd-fsckd was running? | 15:28 |
didrocks | (if you didn't mask) | 15:28 |
Laney | oops, I just rebooted to systemd for the lolz | 15:29 |
seb128_ | didrocks, yes, didn't try to mask yet | 15:29 |
seb128_ | doing that now | 15:29 |
pitti | Feb 19 16:27:13 pid1 systemd[1]: Listening on fsck to fsckd communication Socket. | 15:29 |
pitti | Feb 19 16:27:13 pid1 systemd[1]: Starting fsck to fsckd communication Socket. | 15:29 |
pitti | Feb 19 16:27:13 pid1 systemd-fsck[172]: /dev/vda1: clean, 188260/1179648 files, 1021730/4718592 blocks | 15:29 |
Laney | lots of stuff is taking ages to start | 15:29 |
didrocks | pitti: do you still have some systemd-fsck instances running? | 15:30 |
didrocks | fscanf on the fsck pipe should return != 4 | 15:31 |
didrocks | and thus, systemd-fsck closes after the connection | 15:31 |
pitti | $ sudo systemctl --all|grep fsck | 15:31 |
pitti | [sudo] password for martin: | 15:31 |
pitti | systemd-fsck-root.service loaded active exited File System Check on Root Device | 15:31 |
didrocks | and systemd-fsckd doesn't even try to connect to plymouth | 15:31 |
pitti | systemd-fsckd.service loaded active running File System Check Daemon to report status | 15:31 |
pitti | systemd-fsckd.socket loaded active running fsck to fsckd communication Socket | 15:31 |
seb128_ | didrocks, pitti, with systemd-fsck masked I get a purple screen with no plymouth logo and it doesn't boot, get stucked on there | 15:32 |
pitti | didrocks: so, -root was running, seems it didn't properly tell fsckd "I'm done"? | 15:32 |
didrocks | pitti: at boot time, but after a while, (30s) systemd-fsckd.service exited? | 15:32 |
pitti | didrocks: correct | 15:32 |
didrocks | pitti: yeah, so expected, and it didn't even connect to plymouth (normally) | 15:33 |
pitti | didrocks: fsckd to inactive/dead, the socket to active/listening | 15:33 |
pitti | seb128_: I hope that was a typo in IRC only, that you masked fsckd and not fsck :) | 15:33 |
seb128_ | pitti, correct :-) | 15:33 |
didrocks | pitti: yeah, so it's working as expected, we do wake up systemd-fsckd, I should probably do that only if we receive progress | 15:33 |
desrt | attente_: okay.. reviewed again... as always, your patch is fine, but i'm picking on style issues :p | 15:34 |
didrocks | pitti: but that's a noop, basically nothing happens after 30s and systemd-fsckd exits | 15:34 |
attente_ | desrt: thanks, looking at it again. vala did flag that as a warning | 15:34 |
didrocks | (without connecting to plymouth) | 15:34 |
pitti | didrocks: do you have any log_debug() there which could help? | 15:34 |
pitti | didrocks: i. e. if seb128 boots with debug and plymouth, and collects the journal after the next (working) boot? | 15:34 |
desrt | attente_: vala should be reasonable enough to know that inside of a foreach the variable will be non-null | 15:35 |
pitti | seb128_: to be clear, without "splash" everything is fine, yes? | 15:35 |
seb128_ | pitti, correct | 15:35 |
desrt | attente_: but uh... i guess that's why the mode is called "experimental" | 15:35 |
didrocks | pitti: yeah, they are quite a lot of log_debug(), so that could work | 15:35 |
didrocks | pitti: but as seb128_ still has the issue with systemd-fsckd masked… | 15:35 |
desrt | attente_: odd that the indicator has it enabled, to be honest... | 15:35 |
attente_ | desrt: maybe it can't guarantee that for iterables in general.. | 15:35 |
didrocks | let's try to mask as well systemd-fsck@ as well? | 15:35 |
desrt | attente_: it can in this case... | 15:36 |
pitti | didrocks: yeah, I'd still llike to see the debug output; I don't have a better idea how to debug that remotely now | 15:36 |
desrt | attente_: iterators work by ducktyping in a couple of possible ways | 15:36 |
attente_ | desrt: oh. right | 15:36 |
pitti | didrocks: that might cause boot hangs, mounting the root fs etc. requires fsck? | 15:36 |
desrt | attente_: in this way, it works by calling a next_value() function until it returns null | 15:36 |
desrt | so it definitely knows :) | 15:36 |
seb128_ | pitti, sorry, it has the issue without splash as well :-/ | 15:36 |
seb128_ | pitti, it doesn't have the issue if I remove "quiet" though, which doesn't make any sense to me | 15:37 |
didrocks | seb128_: can you try to mask system-fsck as well? | 15:37 |
pitti | seb128_: the long boot and failied dbus? | 15:37 |
desrt | attente_: but good luck getting anyone to care.... it's called --enable-experimental-non-null for a reason | 15:37 |
seb128_ | pitti, no, atm I've no plymouth at all since I disabled fsckd | 15:37 |
pitti | seb128_: oh wait -- if you boot without splash but with quiet, boot is hanging? | 15:37 |
seb128_ | didn't try to wait for 3 minutes | 15:37 |
seb128_ | yes | 15:37 |
pitti | seb128_: with our without $vt_handoff? | 15:38 |
didrocks | seb128_: sudo ln -s /dev/null /etc/systemd/system/systemd-fsck-root.service and sudo ln -s /dev/null /etc/systemd/system/systemd-fsck@.service | 15:38 |
seb128_ | with | 15:38 |
seb128_ | oh | 15:38 |
seb128_ | I stucked on | 15:38 |
desrt | attente_: your countdown/countup loop is.... excessive :) | 15:38 |
seb128_ | A start job is running for File System Check on Root Device (40s / no limit) | 15:38 |
seb128_ | with a bouncing [***] | 15:39 |
seb128_ | this time | 15:39 |
pitti | well, my VM is booting with or without $vt_handoff | 15:39 |
seb128_ | with quiet and no splash | 15:39 |
seb128_ | that job counting | 15:39 |
seb128_ | 1m15 now | 15:39 |
didrocks | I guess it's trying to connect to the socket and wait | 15:39 |
didrocks | (systemd-fsckd socket) | 15:40 |
didrocks | now that systemd-fsckd is masked | 15:40 |
pitti | seb128_: ok, it's a known thing that fsck doesn't report progress with quiet (not sure if that's a bug or a feature); do you think this could actually be a due fsck on your root fs? | 15:40 |
pitti | ah, right | 15:40 |
seb128_ | I doubt it | 15:40 |
seb128_ | it's a 80G ssd | 15:40 |
seb128_ | it usually takes less than a minute | 15:40 |
seb128_ | and it's 3 min now and counting | 15:40 |
didrocks | yeah, I guess it's blocked on the socket | 15:40 |
pitti | I can reproduce a hang | 15:41 |
=== alan_g|lunch is now known as alan_g | ||
didrocks | can you try to mask that unit to ensure it's all due to fsck-root? | 15:41 |
pitti | with masking fsckd and booting with "quiet" but no splash | 15:41 |
seb128_ | pitti, I sent you a screen by email from my phone | 15:41 |
didrocks | pitti: yeah, but not sure the first hangs is the same | 15:41 |
seb128_ | pitti, k, that's what I have as a config | 15:41 |
pitti | so I guess masking fsckd isn't such a great idea after all | 15:41 |
didrocks | this one is expected | 15:41 |
seb128_ | why is removing "quiet" fixing it? | 15:42 |
pitti | but I don't get any other hang | 15:42 |
pitti | ah, the $vt_handoff won't let me see the error message, that's what's causing the empty screen | 15:42 |
pitti | yeah, I get symptoms like seb128 now | 15:42 |
pitti | even the low-graphics mode now after some 5 mins :) | 15:43 |
didrocks | without masking? | 15:43 |
seb128_ | :-) | 15:43 |
* seb128_ removes the masking | 15:43 | |
didrocks | seb128_: no, please adds the masking on the others I asked | 15:43 |
didrocks | add* | 15:43 |
pitti | didrocks: no, with masking | 15:44 |
seb128_ | didrocks, let me read backlog | 15:44 |
pitti | so if I mask fsckd *and* boot with "quiet $vt_handoff", I get the trouble | 15:44 |
didrocks | pitti: yeah, as said, this is expected as the socket is blocking | 15:44 |
didrocks | but that doesn't explain the first issue seb128_ had | 15:44 |
pitti | if I boot without all options, I at least see the messages, but still blocking | 15:45 |
pitti | right | 15:45 |
seb128_ | shrug, I don't manage to boot anymore now | 15:45 |
pitti | actually, no -- without any options it doesn't hang (very long) | 15:45 |
pitti | seb128_: boot with upstart (under advanced options in grub) | 15:45 |
seb128_ | pitti, yeah, I was going to do that by editing the init= :-) | 15:46 |
didrocks | seb128_: did the option for you man :) | 15:46 |
seb128_ | didrocks, let me try | 15:46 |
seb128_ | shrug, you can't even ctrl-alt-del reboot when it's stucked waiting on thge socket | 15:47 |
didrocks | pitti: I guess removing the non blocking socket wasn't a good idea after all :p | 15:47 |
pitti | seb128_: press it 7 times in 2 s | 15:48 |
pitti | err, "more than 7x" | 15:48 |
seb128_ | k | 15:48 |
pitti | (that's an emergency fallback; haven't tested it yet) | 15:48 |
seb128_ | do I need to hang a chicken as well? | 15:48 |
pitti | seb128_: ok, sorry for the bad advice with masking | 15:48 |
pitti | seb128_: on the bright side, with that I reproduced something which is veeeery close to your symptoms, so it's useful after all | 15:49 |
didrocks | pitti: actually, that's a good point that seb128_ masked it, I think I should reintroduce non waiting socket for this | 15:49 |
seb128_ | didrocks, ok, boots fine after masking fsck-root and @ | 15:50 |
didrocks | ok, so clearly related to this… what happens :/ | 15:51 |
seb128_ | I also had a fsck run when I booted in recovery | 15:51 |
seb128_ | it took like 30 seconds | 15:51 |
pitti | we don't have a debug journal for what actually happens, right? | 15:51 |
seb128_ | so it's not fsck holding boot for 3 minutes | 15:51 |
pitti | i. e. without any masking, and plymouth and quiet and stuff | 15:51 |
seb128_ | pitti, the pastebin from earlier is a debug journal | 15:52 |
pitti | yeah, it's clearly hanging at trying to talk to the socket | 15:52 |
didrocks | there are quite a log of debug in system-fsckd, less then systemd-fsck | 15:52 |
didrocks | but yeah, would be interesting to get the systemd-fsckd and systemd-fsck-root debugs | 15:52 |
pitti | seb128_: ah, ok | 15:52 |
pitti | http://people.canonical.com/~seb128/log ? | 15:52 |
pitti | that doesn't have any fsckd messages | 15:53 |
didrocks | nor fsck… | 15:53 |
seb128_ | let me try again | 15:53 |
seb128_ | removing all the masks and doing a debug boot | 15:53 |
seb128_ | bah | 15:57 |
seb128_ | my system works now | 15:57 |
seb128_ | I wonder if there was a fsck needed and it was never completing | 15:58 |
didrocks | argh | 15:58 |
seb128_ | which it did when I booted in recovery | 15:58 |
didrocks | do you mind forcing a fsck? | 15:58 |
pitti | seb128_: thanks for the picture | 15:58 |
seb128_ | there is a flag to force it | 15:58 |
seb128_ | pitti, yw :-) | 15:58 |
seb128_ | didrocks, what's the file to touch again? | 15:58 |
pitti | seb128_: "failed to listen on fsck to fsckd communication Socket" | 15:58 |
pitti | didrocks: ^ does that tell you anything? | 15:58 |
didrocks | pitti: yeah, it did write this, one sec | 15:59 |
pitti | seb128_: sudo tune2fs -C 50 /dev/sdXX | 15:59 |
seb128_ | pitti, the picture was from fsckd masked | 15:59 |
pitti | seb128_: ah, ok; that's expected then | 15:59 |
pitti | seb128_: but I doubt that the actual fscking is related to that | 15:59 |
didrocks | pitti: hum, no… | 16:00 |
didrocks | so yeah :) | 16:00 |
seb128_ | ok, so forced fsck | 16:00 |
didrocks | so, at least, it means the bailing out works | 16:01 |
didrocks | now, let's see with your manual fsck forcing | 16:01 |
seb128_ | but that works, it indicates the % on plymouth | 16:01 |
didrocks | :/ | 16:01 |
seb128_ | and boots then* | 16:01 |
pitti | yay heisenbug | 16:01 |
seb128_ | :-/ | 16:01 |
seb128_ | thanks guys | 16:01 |
seb128_ | I can ping you again if it ever comes back | 16:01 |
pitti | so I still think when it hung for you there was some problem to talk to fsckd | 16:02 |
didrocks | yeah, especially if you can reproduce it in loop | 16:02 |
didrocks | right, probably | 16:02 |
didrocks | would be nice to know exactly why… | 16:02 |
seb128_ | bah | 16:02 |
seb128_ | doing it again, on the reboot after the fsck | 16:02 |
seb128_ | I didn't boot with debug though | 16:02 |
seb128_ | let me try to reboot | 16:02 |
seb128_ | I can journalctl -b -n anyway to get that log | 16:03 |
didrocks | that's weird… as the next one was working as expected | 16:03 |
didrocks | yeah | 16:03 |
pitti | also, in http://people.canonical.com/~seb128/log there was no actual fsck | 16:04 |
Laney | sooooooooo systemd takes ages to boot my system | 16:04 |
pitti | févr. 19 15:45:10 seb-e6410 systemd-fsck[272]: /dev/sda1 : propre, 937151/4685824 fichiers, 16702765/18730240 blocs | 16:05 |
didrocks | maybe Laney got the same issue… | 16:05 |
pitti | I know what "prope" means :) | 16:05 |
Laney | well it eventually timed out and booted | 16:05 |
seb128_ | sooo | 16:05 |
Laney | kernel is 76 seconds, then networking is $ages | 16:05 |
Laney | from a systemd-analyze plot | 16:05 |
Laney | you probably can't be blamed for the kernel part, eh :) | 16:05 |
pitti | Laney: journalctl | pastebinit ? | 16:05 |
didrocks | Laney: yeah :p | 16:06 |
pitti | well, 76 s is absurdly long | 16:06 |
Laney | indeed | 16:06 |
pitti | I figure that doesn't happen under upstart? | 16:06 |
Laney | don't know | 16:06 |
Laney | oh I did have bootchart installed, maybe I have some of those lying about | 16:06 |
pitti | ah, don't :) | 16:06 |
pitti | (I mean, uninstall it) | 16:06 |
pitti | init=/lib/systemd/systemd-bootchart if you actually want one | 16:07 |
Laney | oh yeah LOADS! | 16:07 |
Laney | I mean from my previous upstart boots | 16:07 |
seb128_ | didrocks, pitti, you have email | 16:07 |
pitti | Startup finished in 4.583s (firmware) + 4.202s (loader) + 2.719s (kernel) + 4.569s (userspace) = 16.075s | 16:07 |
seb128_ | that's my vt7 after the plymouth timeout | 16:07 |
pitti | that can most certainly be improved, but it's not too bad, given that we did exactly zero work to optimize it | 16:08 |
seb128_ | A start job is running for oFono Mobile telephony stack (17s / 1min30) | 16:08 |
didrocks | seb128_: not what I was expected… | 16:08 |
seb128_ | is that sort of thing expected? | 16:08 |
didrocks | so, the first hang would have been oFono? | 16:09 |
Laney | pitti: ah, I think that's lies | 16:09 |
seb128_ | dunno | 16:09 |
Laney | at least part of it | 16:09 |
* didrocks isn't familiar with that screen | 16:09 | |
pitti | I didn't get seb128's second mail yet | 16:10 |
Laney | it was still trying to initialise some network iface even after the system was up | 16:10 |
pitti | (btw, welcome to #ubuntu-bootwoes) | 16:10 |
didrocks | pitti: 344Kb, too much for your mailbox :p | 16:10 |
* seb128_ pets good old upstart | 16:10 | |
Laney | http://paste.ubuntu.com/10309940/ | 16:10 |
seb128_ | didrocks, pitti, well at least I can easy reproduce again :/ doing a debug boot atm, waiting for plymouth to timeout so I go to a vt | 16:11 |
didrocks | seb128_: always oFono? | 16:11 |
seb128_ | dunno yet, I can't get to anything until plymouth timeouts | 16:11 |
pitti | Laney: hm, that's a mere 7 seconds | 16:11 |
pitti | from first message to graphical.target | 16:11 |
Laney | definitely did take longer | 16:12 |
Laney | where does it get the 1min15.966 seconds from? | 16:12 |
didrocks | ok, at least, fsck/fsckd doesn't hang Laney's laptop | 16:12 |
Laney | desktop | 16:12 |
Laney | laptop/systemd is super fast! | 16:12 |
didrocks | seb128_: just for my understanding, so you boot, plymouth timeouts and drop you into that vt? | 16:12 |
Laney | didn't try with 219 yet though | 16:13 |
seb128_ | didrocks, no, just goes away, then I can vt switch | 16:13 |
seb128_ | didrocks, vt7 is the screenshot I tool | 16:13 |
seb128_ | took | 16:13 |
seb128_ | I can log in on vt1 | 16:13 |
seb128_ | well, "log in", then xfailsafe kicks in | 16:13 |
didrocks | seb128_: ah ok, vt7… | 16:13 |
didrocks | seb128_: if you notice some bugs in xfailsafe, it's on me :) | 16:13 |
seb128_ | :-) | 16:14 |
* didrocks forgot that was this cycle as well | 16:14 | |
didrocks | seb128_: always, can locate ofono | grep service, I don't have any here | 16:15 |
* didrocks smells a touch thingy… | 16:15 | |
pitti | Laney: ok, I'm afraid I don't see anything slow in that log; it starts at 15:40:38, eth0 is up at 15:40:42, lightdm is running at 15:40:45 | 16:15 |
* Laney reboots again | 16:16 | |
seb128 | didrocks, ofono: /lib/systemd/system/ofono.service | 16:17 |
* Laney uses irc-as-a-stopwatch | 16:17 | |
pitti | Laney: oh! I see you are affected by the "spontaneously unmounts my partitions" issue | 16:17 |
didrocks | seb128: maybe try to hem… mask… it :) | 16:17 |
Laney | grub... | 16:17 |
Laney | hitting enter now | 16:17 |
pitti | Laney: you want to upgrade your system, there was a badly broken lxcfs which messed up stuff | 16:17 |
Laney | black screen | 16:17 |
pitti | Laney: current vivid has that fixed, and systemd 219 has a robustification | 16:17 |
seb128 | didrocks, pitti, http://people.canonical.com/~seb128/log ... debug log from buggy boot | 16:18 |
Laney | hmm, I thought I did dist-upgrade, when did that land? | 16:18 |
pitti | Laney: bug 1419623 | 16:18 |
ubot5 | bug 1419623 in lxcfs (Ubuntu) "systemd unmounts mounted filesystems when lxcfs is installed" [Critical,Fix released] https://launchpad.net/bugs/1419623 | 16:18 |
pitti | Laney: systemd 218-10ubuntu1, and lxcfs 0.5-0ubuntu2, but already a week ago | 16:18 |
didrocks | seb128: pitti: yeah, the hang is way too late to be systemd-fsck* related (as it's blocking other phases and you won't be at the network stage there) | 16:18 |
Laney | surely I have those | 16:19 |
Laney | will see in a second | 16:19 |
Laney | (still black screen) | 16:19 |
didrocks | seeing it's just after some nm activity, it's probably this ofono… | 16:19 |
didrocks | seb128: so try masking (I don't think anything else is depending on it) and reboot? | 16:19 |
Laney | bleh | 16:20 |
Laney | I hit escape | 16:20 |
Laney | some stuff is still starting | 16:20 |
Laney | ofono, plymouth, logind, virbr0 | 16:21 |
seb128_ | didrocks, same issue with ofono uninstalled :-/ | 16:22 |
Laney | oh, xfallback | 16:22 |
didrocks | seb128_: hum, and this time it's telling that it's waiting on… ? | 16:22 |
seb128_ | didrocks, waiting on the timeout ... | 16:23 |
didrocks | seb128_: don't tell me that masking ofono was a bad idea, I don't want we create that meme :p | 16:23 |
seb128_ | lol | 16:23 |
seb128_ | it has 3 stuff he was cycle through | 16:25 |
seb128_ | login service | 16:25 |
seb128_ | waiting for plymouth to exit | 16:25 |
seb128_ | and some other | 16:25 |
Laney | I think I'm in the same place as seb128_ now | 16:25 |
Laney | \o/ | 16:25 |
seb128_ | :-/ | 16:25 |
seb128_ | oh, hey xfailsafe | 16:25 |
didrocks | if only we could have a vt… | 16:25 |
Laney | hahaha | 16:26 |
Laney | I have a vt | 16:26 |
seb128_ | me too | 16:26 |
didrocks | yeah, but not when it's hanging | 16:26 |
Laney | oh right | 16:26 |
didrocks | or while* | 16:26 |
seb128_ | just takes 3 minutes to get to it | 16:26 |
pitti | GunnarHj: fresh trusty-proposed langpacks uploaded and accepted, they are building now | 16:27 |
didrocks | so, maybe waiting on plymouth to exit would be due to fsckd in some form? | 16:27 |
didrocks | (but the timeout is supposively 30s) | 16:27 |
seb128_ | could be | 16:27 |
seb128_ | masking fsck | 16:27 |
seb128_ | masking fsck* makes my system boot | 16:27 |
GunnarHj | pitti: Great, Martin. I'm going to notify the translators later tonight. | 16:27 |
didrocks | let's try to reinstall an older systemd-fsck, which doesn't talk to fsck | 16:27 |
seb128_ | didrocks, is that just the binary to copy? | 16:28 |
didrocks | seb128_: yeah, /lib/systemd/systemd-fsck from https://launchpad.net/ubuntu/+source/systemd/218-8ubuntu2 | 16:28 |
didrocks | seb128_: you didn't upgrade this machine to 218-10ubuntu1 since Friday? | 16:29 |
didrocks | (this was the first one with systemd-fsckd) | 16:29 |
seb128_ | I think I did | 16:29 |
seb128_ | I did daily upgrades | 16:29 |
didrocks | I guess you would have got this one though | 16:30 |
didrocks | but let's bail it out anyway | 16:30 |
Laney | http://paste.ubuntu.com/10310208/ | 16:30 |
seb128_ | maybe I didn't reboot | 16:30 |
Laney | this troublesome boot starts at 16:19:49 | 16:30 |
seb128_ | Laney, no such timestamp in your log? | 16:30 |
didrocks | I was going to say… | 16:31 |
Laney | there is ... | 16:31 |
didrocks | seb128_: it's in the systemd binary btw (sorry, didn't told you) | 16:31 |
didrocks | oh multiple boots | 16:31 |
Laney | yeah it's just all of syslog | 16:31 |
didrocks | so, fsck didn't block | 16:32 |
pitti | From sebastien.bacher@laposte.net Thu Feb 19 17:07:36 2015 | 16:32 |
didrocks | my only guess at this point (if fsckd is to blame) is that it hangs plymouth | 16:32 |
pitti | Subject: Screenshot | 16:32 |
pitti | Folder: spam | 16:32 |
pitti | seb128_: ^ sorry | 16:32 |
pitti | seb128_: what kind of pictures do you send me? :-) | 16:32 |
seb128_ | lol | 16:33 |
didrocks | pitti: next time he will send with "Get a billion dollars in 30 minutes" :) | 16:33 |
pitti | didrocks: yeah, sounds like it; sorry, got diverted in other channels -- did you ask Laney to mask the magic stuff and try with that already? | 16:35 |
Laney | systemd-fsckd ofono ? | 16:35 |
pitti | systemd-fsck-root also, I believe? | 16:36 |
pitti | (didrocks knows that better, sorry) | 16:36 |
didrocks | pitti: seb128_ is trying to downgrade systemd-fsck binary first | 16:36 |
didrocks | Laney: systemd-fsck-root.service | 16:36 |
Laney | okey dokey | 16:36 |
pitti | Laney: ok, so perhaps let's wait until didrocks is done with debugging that with seb, and then we'll try the same on your box? | 16:37 |
Laney | I'm suspicous that it booted properly the previous time | 16:37 |
Laney | so could indeed be fsck | 16:37 |
didrocks | yeah, sounds good :) | 16:37 |
Laney | 'properly' but slow | 16:37 |
pitti | I'm sure that using btrfs magically saves me from that, or it's just a weird timing difference | 16:37 |
pitti | although my test VM does have ext4 | 16:37 |
didrocks | well, I never had any issue with ext4 here | 16:37 |
seb128_ | didrocks, pitti, sorry, I don't have the issue anymore now :/ | 16:37 |
Laney | how do I see if it wants to fsck? | 16:38 |
didrocks | seb128_: try harder! :) | 16:38 |
didrocks | Laney: you should have progress reported on plymouth and /dev/console | 16:38 |
didrocks | seb128_: that was before downgrading the binary? | 16:38 |
seb128_ | didrocks, yes, didn't try that yet, I just downloaded the deb | 16:39 |
didrocks | argh :/ | 16:39 |
didrocks | Laney: do you still get it? | 16:39 |
seb128_ | I wanted to restore the masked units first | 16:39 |
didrocks | like every boot? | 16:39 |
seb128_ | bah | 16:39 |
seb128_ | I forced fsck and I get no plymouth | 16:39 |
seb128_ | frozen on a purple screen | 16:39 |
Laney | didrocks: I wanted to see if the drive thinks it needs fscking | 16:40 |
* seb128_ power down and reboot | 16:40 | |
pitti | didrocks: what kind of death message does fsckd send to plymouth :) | 16:40 |
seb128_ | back at having the bug | 16:40 |
didrocks | pitti: a random one apparently :p | 16:40 |
pitti | didrocks: could that be any 32 bit vs. 64 bit thing? | 16:40 |
didrocks | I doubt Laney is using 32 bits | 16:40 |
seb128_ | didrocks, pitti, it seems to start happening after a fsck run, well the boot after that | 16:40 |
didrocks | that's… weird | 16:41 |
didrocks | seb128_: so replace the binary | 16:41 |
Laney | tune2fs is hanging | 16:41 |
Laney | beh | 16:41 |
pitti | Laney: wut? | 16:41 |
Laney | the system is in some kind of disturbing state | 16:42 |
didrocks | pitti: it's really easy to crash plymouth for the record with the wrong protocole | 16:42 |
didrocks | so that can be it, but we both tested this I guess in multiple boots… | 16:42 |
Laney | oh I think it's sudo that is hanging | 16:44 |
Laney | PAM woes probably | 16:44 |
pitti | Laney: D-BUS not yet running, I figure? | 16:44 |
seb128_ | Laney, in my case logind and dbus were unhappy | 16:44 |
Laney | yes this went wrong for me | 16:45 |
pitti | ps aux|grep message.*dbus-daemon ? | 16:45 |
seb128_ | didrocks, boots fine with old systemd-fsck | 16:45 |
seb128_ | let me restore the new one and see if that restore the bug | 16:45 |
kenvandine | bregma, i got my yoga 2 pro, beautiful laptop! | 16:46 |
kenvandine | bregma, but chrome sucks in high dpi :/ | 16:46 |
kenvandine | bregma, is there anyway to make chrome suck less? | 16:46 |
pitti | Laney: could you try replacing /lib/systemd/systemd-fsck with http://people.canonical.com/~pitti/tmp/systemd-fsck ? | 16:46 |
seb128_ | kenvandine, talk to qengho | 16:46 |
Laney | pitti: it's running but services which want to use it didn't start up | 16:46 |
pitti | Laney: that's the amd64 binary from 218-8ubuntu2 | 16:46 |
kenvandine | bregma, and wow... what a nice display! | 16:46 |
seb128_ | pitti, is that an old one or a new with debug one? | 16:47 |
Laney | pitti: okay | 16:47 |
kenvandine | qengho, any tips for making chrome work better with high dpi? | 16:47 |
seb128_ | pitti, Laney, old one works fine for me | 16:47 |
didrocks | pitti: this is the one seb128_ installed FYI, let's see if reboots work for me | 16:47 |
kenvandine | chromium seems to support it | 16:47 |
didrocks | him* | 16:47 |
kenvandine | but the menus are terrible in chromium | 16:47 |
Laney | pitti: I pasted a syslog http://paste.ubuntu.com/10310208/ (search for 16:19:49) in case you can see anything useful | 16:47 |
seb128_ | didrocks, can you force a fsck and try to reboot twice? | 16:47 |
didrocks | seb128_: so, interested into your second boot with the upgraded :p | 16:47 |
didrocks | seb128_: you mean rebooting twice with fsck each time? | 16:48 |
qengho | kenvandine: I may have something that works in a week or so, with release of 40.0.2214.115. | 16:48 |
seb128_ | didrocks, no, fsck once, then reboot | 16:48 |
qengho | kenvandine: menus might be better for that. | 16:48 |
qengho | kenvandine: for GOOG Chrome, i have no tips. | 16:48 |
kenvandine | cool, i changed the resolution to 1080p for now :) | 16:48 |
didrocks | seb128_: well, this is what you do with tune2fs, right? You get the fsck the first time, let it go. and then reboot? | 16:49 |
kenvandine | which i shouldn't complain about | 16:49 |
seb128_ | didrocks, yes | 16:49 |
didrocks | so you shouldn't get fsck the second time | 16:49 |
seb128_ | shrug, no plymouth on the fsck boot | 16:49 |
didrocks | well fsck is running | 16:49 |
didrocks | and telling "nothing to do" | 16:49 |
kenvandine | but wow... at 3200x1800... beautiful! | 16:49 |
kenvandine | just not the browser :/ | 16:49 |
kenvandine | firefox isn't much better | 16:50 |
qengho | kenvandine: you can fix like four times the number of bugs, with that much screen! | 16:50 |
kenvandine | qengho, exactly! | 16:50 |
didrocks | seb128_: with the old version? | 16:50 |
didrocks | or upgraded to the new one? | 16:50 |
seb128_ | new one | 16:50 |
kenvandine | bregma, the touch pad annoys me more than the keyboard, it's really sensitive to taps | 16:51 |
kenvandine | turned that off :) | 16:51 |
didrocks | I don't understand, I don't send anything to plymouth if there is no fsck to do… | 16:51 |
didrocks | so, why the second time wouldn't work well :/ | 16:51 |
seb128_ | maybe it's nothing to do with it | 16:52 |
didrocks | well, it sounds like it though :/ | 16:52 |
seb128_ | in fact I had the run where fsck is needed not displaying plymouth on some boot | 16:52 |
pitti | Laney: looks like plymouth indeed; so, things to try: (1) test with the older sytemd-fsck, or (2) test booting without "splash"? | 16:52 |
seb128_ | so I sit on power down | 16:52 |
seb128_ | because it seemed stucked | 16:52 |
seb128_ | maybe it made fsck unhappy | 16:53 |
seb128_ | and send something that plymouth doesn't handle | 16:53 |
bregma | kenvandine, yeah, a little annoying until you get used to it | 16:53 |
didrocks | seb128_: ah ok, so you would still have fsck on the second time | 16:54 |
didrocks | and I send something to plymouth that it doesn't like | 16:54 |
didrocks | but sometimes, for you, it's working | 16:54 |
Laney | pitti: ack, trying without splash | 16:54 |
didrocks | and here, every boot is :/ | 16:54 |
seb128_ | didrocks, that's my guess | 16:54 |
kenvandine | bregma, the tap to click is so touchy, i can't really use the indicators... but that could be me using only thinkpad's for 15 years :) | 16:55 |
didrocks | ok, so the thing that changed last minute is the Control+C | 16:55 |
kenvandine | and always disabling the touchpad completely :) | 16:55 |
pitti | so, bug report; fsckd sends ping of death to plymouth? | 16:55 |
didrocks | as plymouth boot and plymouth x11 have different protocoles | 16:55 |
Laney | how come it makes dbus fail to work properly? | 16:56 |
kenvandine | this is my first non-thinkpad in probably 15 years | 16:56 |
didrocks | seb128_: it's on an amd64 machine? | 16:56 |
seb128_ | didrocks, 32bits for me | 16:56 |
didrocks | hum, mind if I send you a patch then? and rebuild systemd? (just extract systemd-fsckd from it) | 16:57 |
didrocks | I want to remove the "cancel" option | 16:57 |
Laney | pitti: no 'splash' -> still broken | 16:57 |
Laney | Still seeing 'Wait for Plymouth to quit' | 16:57 |
bregma | kenvandine, mine is a clickpad, I never tap-to-click because of the lack of tactile feedback | 16:57 |
* bregma needs clicks or does not compute | 16:58 | |
Laney | Plymouth Boot Screen | 16:58 |
kenvandine | bregma, yeah, i'm happier with that | 16:59 |
didrocks | seb128_: that should be enough: http://paste.ubuntu.com/10310639/ | 16:59 |
kenvandine | qengho, is there a PPA for chromium that you land stuff in before vivid? | 17:00 |
qengho | kenvandine: Yes. ppa:canonical-chromium-builds/stage | 17:00 |
kenvandine | i'll enable that and hope to see some improvements :) | 17:01 |
qengho | kenvandine: Nothing new there for another day or so. | 17:01 |
kenvandine | qengho, mind pinging me when i should try testing high-dpi again? | 17:01 |
qengho | kenvandine: sure. | 17:01 |
kenvandine | fwiw... windows 8.1 isn't much better at that resolution :) | 17:01 |
kenvandine | IE handles it fine | 17:02 |
kenvandine | but quite a few apps i played with was terrible | 17:02 |
kenvandine | the browsers are the only apps i had problems with on vivid | 17:02 |
kenvandine | all the other apps i've tried worked well | 17:02 |
kenvandine | so great job guys! | 17:02 |
kenvandine | bregma, ^^ i suspect you had something to do with that :) | 17:02 |
Laney | doesn't look like the old systemd-fsck is helping | 17:07 |
seb128_ | in fact I just had the issue with it as well I think | 17:08 |
seb128_ | waiting for the system to give me back a login to check that I correctly copied the binary | 17:08 |
didrocks | seb128_: with the old systemd-fsck? | 17:08 |
seb128_ | yes | 17:08 |
didrocks | pitti: this is becoming crazy ^ :/ | 17:09 |
seb128_ | fsck is maybe just a redherring | 17:09 |
seb128_ | could be another bug in new systemd | 17:09 |
didrocks | yeah | 17:09 |
didrocks | Laney: you told you were not on latest systemd though? | 17:09 |
=== ralsina is now known as jamestiberiuskir | ||
didrocks | and did my memory/reading/english/french betrayed me? | 17:10 |
Laney | should be, let me check | 17:10 |
=== jamestiberiuskir is now known as ralsina | ||
Laney | ya, 219 | 17:10 |
pitti | Laney does have 219 | 17:10 |
=== ralsina is now known as bonesmcoy | ||
Laney | how come I get waiting for plymouth even without 'splash', btw? | 17:10 |
seb128_ | didrocks, pitti, so yeah, issue is there as well with old systemd-dsck | 17:11 |
=== bonesmcoy is now known as ralsina | ||
pitti | seb128_: that is, boot hangs on ofono and other stuff? | 17:11 |
didrocks | ConditionKernelCommandLine=splash | 17:11 |
* pitti lost track a bit, sorry (getting pinged in other channels) | 17:11 | |
seb128_ | pitti, I removed ofono, so not it | 17:11 |
didrocks | Laney: and you can't get a status of which job exactly it is waiting on? | 17:12 |
Laney | there's some general problem with dbus | 17:12 |
Laney | I think ofono is probably an instance of that | 17:12 |
seb128_ | yes, things timeout | 17:12 |
didrocks | Laney: shouldn't be plymouth-start.service due to this ConditionKernelCommandLine | 17:12 |
seb128_ | likely | 17:12 |
Laney | didrocks: grep splash /proc/cmdline -> nothing | 17:12 |
* didrocks backlog to see the exact line that Laney is seeing | 17:13 | |
didrocks | "Plymouth Boot Screen" | 17:13 |
Laney | I typed that | 17:13 |
didrocks | Laney: do you have a "Show" before? | 17:13 |
didrocks | Wait for | 17:13 |
didrocks | or Terminate ? | 17:13 |
Laney | terminate IIRC | 17:14 |
didrocks | ExecStart=-/bin/plymouth quit | 17:14 |
didrocks | unconditionally | 17:14 |
didrocks | so we always tries to quit plymouth | 17:14 |
didrocks | which is a noop I guess if not running | 17:14 |
Laney | seems so | 17:14 |
didrocks | then, maybe the dbus issues makes systemd puzzle or it's really that plymouth quit hangs | 17:15 |
didrocks | Laney: the timeout is 20s, doesn't seem what you see though, right? | 17:15 |
Laney | it has no timeout | 17:15 |
Laney | just counts up forever | 17:15 |
seb128 | worry, I'm stopping debugging for today, need to go in less than 1 hour and I've stuff I wanted to get uploaded before feature freeze today | 17:15 |
didrocks | seb128: yeah, I guess we won't solve it today TBH | 17:15 |
Laney | some other jobs did get their timeouts though | 17:15 |
didrocks | seb128: next step (tomorrow?) would be to downgrade to -10ubuntu1 | 17:15 |
didrocks | or 2 | 17:16 |
seb128 | I've faith in Laney | 17:16 |
didrocks | ahah :) | 17:16 |
seb128 | he can resolve it :-) | 17:16 |
* didrocks should have stopped already btw :p | 17:16 | |
didrocks | Laney: the timeout is once the "quit" job fires | 17:16 |
Laney | you're funny | 17:16 |
didrocks | so if everything else is delayed | 17:16 |
didrocks | I guess this one isn't accurate | 17:17 |
didrocks | Laney: ok, let's try a bigger hammer, downgrade to -10ubuntu2? | 17:17 |
Laney | ok | 17:17 |
Laney | do you have the binaries on hand? | 17:17 |
Laney | oh maybe ssh is working | 17:17 |
Laney | don't fancy navigating LP at the console :p | 17:17 |
didrocks | uno momento! | 17:18 |
Laney | it's ok, ssh is working, I can scp them | 17:18 |
didrocks | ah good: https://launchpad.net/ubuntu/+source/systemd/218-10ubuntu2/+build/6984026/+files/udev_218-10ubuntu2_amd64.deb | 17:18 |
didrocks | (so that you can win 20s :p) | 17:18 |
Laney | so many packages | 17:19 |
* didrocks gives some TCP ones to Laney :p | 17:21 | |
Laney | wget --recursive \o/ | 17:21 |
Laney | come on udev postinst | 17:22 |
Laney | you can do it | 17:22 |
Laney | ... | 17:23 |
Laney | can you? | 17:23 |
* didrocks trusts in udev postinst :p | 17:24 | |
seb128 | attente_, your osk .sh change seems to work fine on the device for me | 17:24 |
Laney | wait what | 17:24 |
Laney | I didn't have new udev stuff before | 17:24 |
didrocks | oh? | 17:25 |
attente_ | seb128: great, thanks! | 17:26 |
Laney | didrocks: I'd done apt-get install systemd-sysv but not dist-upgraded | 17:26 |
Laney | let me try rebooting with 218 anyway | 17:26 |
didrocks | yeah… let's see | 17:27 |
Laney | didrocks: lightdm! | 17:28 |
didrocks | Laney: you see lightdm, you mean, or you did spot it was lightdm? :) | 17:29 |
Laney | I see lightdm now | 17:29 |
Laney | i.e. it works | 17:29 |
didrocks | "long time no see" :) | 17:29 |
didrocks | mind reupgrading, just to confirm? | 17:30 |
didrocks | oh before that | 17:30 |
didrocks | please copy /lib/systemd/system-fsckd and /lib/systemd/system-fsck somewhere | 17:30 |
Laney | k | 17:30 |
didrocks | so that you can restore them | 17:30 |
didrocks | just to check if it's the protocole change that would be in cause :p | 17:30 |
Laney | erm | 17:31 |
Laney | dist-upgrade wants to remove ubuntu-desktop | 17:31 |
Laney | and all of xorg | 17:31 |
willcooke | erm | 17:32 |
didrocks | Laney: apt-get install <binary-packages>/vivid? :/ | 17:33 |
willcooke | Laney, you didnt add mlankhorst's ppa did you? | 17:33 |
Laney | no | 17:33 |
Laney | didrocks: can't on a dist-upgrade | 17:33 |
Laney | aptitude points at xserver-xorg-core | 17:33 |
Laney | meh, one thing at a time | 17:33 |
* Laney leaves that behind for now | 17:33 | |
Laney | willcooke: oh wait, yes I did, ha | 17:34 |
Laney | I just checked some package which wasn't in it | 17:35 |
* Laney fail | 17:35 | |
willcooke | This is an epic day for me | 17:35 |
* willcooke <-- l33t | 17:35 | |
willcooke | seb128, getting that power thing out the way has fixed 99% of my provlems | 17:36 |
seb128 | willcooke, great | 17:36 |
willcooke | the 1% being my ability to type | 17:37 |
seb128 | lol | 17:37 |
Laney | ok, systemd 219: take II | 17:38 |
Laney | didrocks: seems bad again - try replacing those binaries? | 17:41 |
=== zz_Bl4ckD34Th is now known as Bl4ckD34Th | ||
didrocks | Laney: yes please | 17:43 |
Laney | super ack | 17:45 |
Laney | no good, looks like you are innocent | 17:48 |
didrocks | |o| | 17:49 |
didrocks | pitti: ok, bad news is that it seems there is a real regression in 219… :/ | 17:49 |
Laney | did you try it? | 17:50 |
didrocks | Laney: I'm running it, no issue here | 17:53 |
didrocks | (did reboot twice since yesterday) | 17:53 |
didrocks | and I'm sure Martin did as well… | 17:53 |
Laney | nod | 17:53 |
Laney | always the way | 17:53 |
didrocks | of course :/ | 17:55 |
mlankhorst | Laney: x-staging is needed for xorg-server 1.17 | 17:59 |
Laney | it's okay | 17:59 |
Laney | I didn't really want to run it | 17:59 |
didrocks | Laney: let's dig that more tomorrow I guess, until then, you can reboot with upstart with the grub menu | 17:59 |
Laney | didrocks: ya, no worries, I gtg in a minute anyway | 18:00 |
Laney | btw 'busctl' works | 18:00 |
Laney | so the system bus isn't completely broken | 18:00 |
didrocks | interesting | 18:00 |
didrocks | well, we'll see :) | 18:00 |
Laney | fun! | 18:00 |
* didrocks will go as well, playing with unity3d support in ubuntu! | 18:01 | |
didrocks | see you guys :) | 18:01 |
Laney | bye! | 18:01 |
=== vrruiz_ is now known as rvr | ||
=== alan_g is now known as alan_g|EOD | ||
seb128 | calling it a day, have a nice evening | 18:11 |
larsu | seb128: enjoy! | 18:11 |
seb128 | larsu, thanks | 18:11 |
Laney | pitti: go to go now, but for info | 18:14 |
Laney | I removed BusName from polkitd.service and SystemdService from /u/s/dbus-1/system-services/org.freedesktop.PolicyKit1.service and now I have lightdm again | 18:15 |
Laney | tried this because journalctl -u dbus was showing polkit failing to get the name as its first action, which I guessed probably caused cascading failures | 18:15 |
* Laney waves \o | 18:17 | |
larsu | bye Laney! | 18:17 |
dagerian | anyone mind helping me out | 18:43 |
pitti | hm, so I missed didrocks and Laney | 18:52 |
pitti | and seb128 too | 18:52 |
=== Bl4ckD34Th is now known as zz_Bl4ckD34Th | ||
=== Zdra is now known as xclaesse | ||
andrzejr | Hi, what is the status of libraries like libindicator or libido? I'm using the first one in xfce4-indicator-plugin for xubuntu and was planning to use the other one (not limited to xubuntu) but there are no releases newer than 12.10. | 21:13 |
andrzejr | There are quite a few new versions in bzr so it looks as if these libraries were now only intended for use in Ubuntu. Is that correct? | 21:14 |
=== zz_Bl4ckD34Th is now known as Bl4ckD34Th | ||
* Bl4ckD34Th Bl4ckD34Th return to take your soul! You own to Bl4ckD34Th!!! | 22:38 | |
dobey | nope | 22:41 |
Bl4ckD34Th | ok | 22:41 |
Bl4ckD34Th | sorry | 22:41 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!