=== maclin1 is now known as maclin | ||
tsdgeos | mzanetti: are we landing any more branch on the overlay? or first doing the branch killing? | 07:52 |
---|---|---|
tsdgeos | because we're starting to stockpile again :D | 07:52 |
mzanetti | tsdgeos, next ones will go to trunk again... let me check if the one that sync landed by now | 07:53 |
mzanetti | it was QA'd last night | 07:53 |
mzanetti | oh dear... hangs in proposed somewher | 07:54 |
tsdgeos | :A | 07:57 |
=== davidcalle_ is now known as davidcalle | ||
mzanetti | I fail to understand those adt logs | 08:07 |
mzanetti | seems to fail for qtmir and unity8 | 08:08 |
mzanetti | bootest says it can't do some apparmor things, whatever that means... | 08:08 |
Mirv | Trevinho is great at spamming my bug folder :D | 09:38 |
Trevinho | Mirv: yeah, I had to fire my script to keep things in sync | 09:39 |
Trevinho | Mirv: and.... Well I didn't do that for compiz (yet) as there was 2600 bugs to fix... and it wasn't the case :) | 09:39 |
Mirv | :) | 09:40 |
=== jzheng is now known as jzheng_afk | ||
=== nudtrobert1 is now known as nudtrobert | ||
=== alan_g is now known as alan_g|lunch | ||
=== cwayne-afk is now known as cwayne_ | ||
=== MacSlow is now known as MacSlow|lunch | ||
=== alan_g|lunch is now known as alan_g | ||
mzanetti | dandrader, there's a test failure in that silo 27: https://launchpadlibrarian.net/216844866/buildlog_ubuntu-wily-amd64.qtmir_0.4.6%2B15.10.20150910-0ubuntu1_BUILDING.txt.gz | 13:39 |
dandrader | mzanetti, ok | 13:41 |
mzanetti | let me know when to re-kick the build | 13:42 |
dandrader | mzanetti, can you remove https://code.launchpad.net/~gerboland/qtmir/multimonitor/+merge/269906 from that silo? | 13:43 |
dandrader | mzanetti, so that we can land the critical bug fix it contains | 13:43 |
dandrader | mzanetti, and I guess it wouldn't hurt to add those qtubuntu ones as well: https://code.launchpad.net/~dandrader/qtubuntu/noQtSensors-lp1481389/+merge/267198 and https://code.launchpad.net/~dandrader/qtubuntu/provideMirConnection/+merge/269320 | 13:45 |
dandrader | they're pretty harmless | 13:46 |
mzanetti | tsdgeos, ah, so it's because it's empty ("")? | 13:56 |
=== MacSlow|lunch is now known as MacSlow | ||
tsdgeos | mzanetti: yep | 13:59 |
mzanetti | tsdgeos, right... makes sense. approved | 13:59 |
mzanetti | dandrader, silo building, including those qtubuntu branches | 14:13 |
dandrader | mzanetti, thanks! | 14:13 |
mzanetti | oops. missed the one to remove | 14:14 |
mzanetti | will redo | 14:14 |
=== dandrader_ is now known as dandrader | ||
=== dandrader is now known as dandrader|afk | ||
=== alan_g is now known as alan_g|afk | ||
tsdgeos | hate it when this happens | 15:10 |
tsdgeos | call deleteLater on an objcet and doesn't get deleted | 15:10 |
tsdgeos | :'( | 15:10 |
tsdgeos | found out why! | 15:21 |
tsdgeos | need to be careful when reimplementing ::event | 15:23 |
=== alan_g|afk is now known as alan_g | ||
=== dandrader|afk is now known as dandrader | ||
dandrader | mzanetti, what's the next step for silo 027? | 16:34 |
=== alan_g is now known as alan_g|EOD | ||
=== dandrader is now known as dandrader|afk | ||
mzanetti | dandrader|afk, you walk through the branches in there and verify that the linked bugs are fixed and the promised features are working. | 17:32 |
mzanetti | dandrader|afk, then you walk through the testplans: https://wiki.ubuntu.com/Process/Merges/TestPlans/QtMir | 17:32 |
mzanetti | dandrader|afk, can't find one for QtUbuntu. Make sure related functionality isn't broken | 17:33 |
mzanetti | when you're happy with the results, go here and leave a comment that you've tested it and on what device/image: https://requests.ci-train.ubuntu.com/#/ticket/328 | 17:34 |
mzanetti | dandrader|afk, for example, like this: https://requests.ci-train.ubuntu.com/#/ticket/285 | 17:34 |
mzanetti | dandrader|afk, test at least on rc-proposed, if things could be different for wily, test both | 17:35 |
=== dandrader|afk is now known as dandrader | ||
kgunn | mzanetti: i don't mind helping to test silo27, that's multimonitor right ? | 18:26 |
slangasek | mzanetti: hi, kgunn pointed me your way regarding this: https://bugs.launchpad.net/canonical-devices-system-image/+bug/1278780/comments/8 | 19:23 |
ubot5 | Ubuntu bug 1278780 in apport (Ubuntu) "apport takes too long to write crash report, appears to lock up phone" [High,Triaged] | 19:23 |
dandrader | mzanetti should be past his EOD by now | 19:28 |
slangasek | dandrader: I imagine so; IRC is a decent messaging service though :) | 19:31 |
kgunn | and he's known to have the problem of not being able to stay away | 19:36 |
dandrader | slangasek, problem is, he will be back only next Tuesday :( | 19:39 |
dandrader | kgunn, and there's that as well :) | 19:39 |
kgunn | ah dang...forgot he was taking some days | 19:43 |
kgunn | slangasek: what is your idea btw | 19:43 |
slangasek | kgunn: so the basic idea is, can we make the crash handler faster for critical components (such as the system compositor or unity8 shell) by having those programs register a custom SIGSEGV handler that unmaps video buffers (for example) and re-raises | 19:50 |
slangasek | kgunn: I don't know if this is safe to do from inside a signal handler, it depends on how the mapping/unmapping is handled; and I don't know if the video memory is a large enough part of the process's address space to make a difference | 19:51 |
slangasek | but I thought it was an idea worth exploring | 19:51 |
kgunn | yeah, the video memory might be the booger | 19:52 |
slangasek | because ultimately, the kind of behavior we've heard described - UI locking up for a minute while the crash handler runs - does sound like it's taking an unreasonably long time for what it /ought/ to be doing | 19:52 |
=== dandrader is now known as dandrader|afk | ||
pmcgowan | kgunn, I just got https://bugs.launchpad.net/canonical-devices-system-image/+bug/1491566 | 20:42 |
ubot5 | Ubuntu bug 1491566 in Canonical System Image "Shell not responsive after an incoming SMS" [Critical,Confirmed] | 20:42 |
pmcgowan | added the logs, its still in the stuck state | 20:42 |
pmcgowan | dandrader|afk, ^ | 20:42 |
=== dandrader|afk is now known as dandrader | ||
dandrader | pmcgowan, still have the device in that state? | 21:12 |
dandrader | pmcgowan, https://bugs.launchpad.net/canonical-devices-system-image/+bug/1491566/comments/26 | 21:14 |
ubot5 | Ubuntu bug 1491566 in Canonical System Image "Shell not responsive after an incoming SMS" [Critical,Confirmed] | 21:14 |
dandrader | so unity8.log shows a log of input events but I can't tell whether those events are *all* from before unity8 became unresponsive or not. What I want to know there is whether it still logs input events *after* it becomes unresponsive | 21:16 |
dandrader | I didn't realize that when I posted the original instructions. Rectified it only in comment #26 | 21:17 |
dandrader | same goes for unity-system-compositor.log | 21:18 |
pmcgowan | dandrader, there was nothing logged after the system froze | 21:19 |
pmcgowan | also see stack trace I added | 21:19 |
pmcgowan | I just restarted lightdm a few mins ago | 21:19 |
dandrader | oh :( | 21:19 |
pmcgowan | well wanted to get my messages :( | 21:19 |
pmcgowan | dandrader, touching the screen I see the touch driver but nothing else get active | 21:20 |
dandrader | alf is the one intereseted in the stack traces. but I can tell they're not all that interesting as you didn't have debug symbols | 21:20 |
dandrader | pmcgowan, what do you mean by "I see the touch driver"? | 21:21 |
pmcgowan | dandrader, a process called mtk-tpd | 21:22 |
dandrader | pmcgowan. so no new messages came to unity8.log or unity-system-compositor.log when you touched the screen while the device was unresponsive?' | 21:22 |
pmcgowan | dandrader, correct, if I tail the logs nothing new | 21:23 |
dandrader | pmcgowan, that's a very important bit of info, it's worth commenting that in the bug report | 21:23 |
pmcgowan | will do | 21:23 |
dandrader | bregma, do you recall what tool we can use to monitor events coming out of evedev files? | 21:33 |
dandrader | evemu-tools | 21:39 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!