=== m_conley_away is now known as m_conley === asac_ is now known as asac === m_conley is now known as m_conley_away === m_conley_away is now known as m_conley === m_conley is now known as m_conley_away [09:29] chrisccoulson: https://bugzilla.mozilla.org/show_bug.cgi?id=649525#c9 can we have that enabled on armel? [09:29] Mozilla bug 649525 in Canvas: WebGL "WebGL layer compositing through the BasicCanvasLayer is very slow (desktop version)" [Normal,New: ] [09:30] or can you make a PPA build available with that patch and with that enabled ;) [09:45] asac, looking === m_conley_away is now known as m_conley [14:16] fta - " there will be a brief disturbance in the ppa universe for maintenence" [14:16] (from #ubuntu-release) [14:17] not sure what that means ;) [14:17] but, just in case they disappear again and you're wondering why... [14:18] oh, i see they're all disabled now anyway, except for the ones which are finishing jobs [14:18] so, too late ;) [14:24] fta: youtube flash in chromium is locking my laptop [14:24] happened 3 times since yesterdat [14:31] chrisccoulson: ping [14:32] hi m_conley, how are you? [14:32] chrisccoulson: I'm alright - how's your day going? [14:32] m_conley, not too bad thanks. just investigating this empty-indicator issue ;) [14:32] how about you? [14:32] chrisccoulson: well, I just merged your patch locally, and I'm getting segfaults anytime I click one of the indicators... [14:33] chrisccoulson: sorry, not seg faulst [14:33] chrisccoulson: assertion failures: Assertion failure: CType::IsSizeDefined(cx, targetType), at /media/Projects/mozilla/thunderbird/mozilla/js/src/ctypes/CTypes.cpp:1730 [14:33] oh, i've not seen that. what version of tbird is that? [14:34] chrisccoulson: trunk. [14:34] chrisccoulson: I guess you've been working with 5.0? [14:34] yeah, i've not tried it on trunk yet [14:34] chrisccoulson: alright, cool [14:37] m_conley, do you run a debug build? [14:38] chrisccoulson: I do - that's probably what's going on [14:38] chrisccoulson: it's just complaining. :p [14:39] interesting [14:39] i'm not sure what the check is actually doing though. it might be worth asking in #jsapi [14:42] m_conley, OOI, where we use g_signal_connect, does it make any difference if you change the data parameter from null to 0? [14:43] chrisccoulson: will check, hang on... [14:43] i'm just guessing here, but the error would suggest we're passing something incorrectly across the C api [14:46] chrisccoulson: no dice [14:46] hmm :/ [14:48] i asked in #jsapi, but i'm not sure they use ctypes in there ;) [14:50] chrisccoulson: mmhmm - finding experts in this domain is a little tricky. ;) [14:50] heh [14:52] i should probably try doing a debug build really [15:00] m_conley, "Some times, like int and int[30] and int *, have a size" [15:00] "But int[] is a type with no size. Therefore you shouldn't be able to construct an object of that type" [15:01] chrisccoulson: where did you read that? Or are you chatting with someone? [15:02] m_conley, that's jorendorff in #jsapi [15:02] gotcha [15:03] chrisccoulson: hrm - not sure what to do with that information, though. [15:03] chrisccoulson: I don't think we're using int arrays anywhere...or am I misunderstanding? [15:03] not sure yet, i need to have a look and see if i'm doing something really silly [15:56] chrisccoulson, ok, thanks for relaying the info. you're more reliable than they are ;) [15:56] BUGabundo, recent regression? [15:57] last two days [15:57] BUGabundo, can you reproduce easily? [15:57] sure [15:57] open youtube [15:57] PUMB [15:57] system locks [15:57] audio plays, video plays [15:57] can you please bisect it? [15:57] but can't access anything , REISUB [15:58] not today [15:58] have to prep last night photo and video sessions [15:58] do you use the h/w acceleration? [15:58] not sure [15:58] let me check the flags [15:58] GPU compositing on all pages is disabled [15:58] here are mine: [15:58] $ grep ^CHROMIUM_FLAGS= /etc/chromium-browser/default [15:58] CHROMIUM_FLAGS="--disable-accelerated-compositing --disable-accelerated-2d-canvas --no-first-run" [15:59] /usr/lib/chromium-browser/chromium-browser --flag-switches-begin --enable-click-to-play --enable-cloud-print-proxy --enable-nacl --experimental-location-features --enable-experimental-extension-apis --ppapi-flash-in-process --preload-instant-search --flag-switches-end [16:00] try with my 2 disable-xx flags [16:00] ok [16:00] restarting [16:06] fta: "Your profile could not be opened correctly. [16:06] Some features may be unavailable. Please check that the profile exists and you have permission to read and write its contents." [16:06] should I be getting this ? [16:07] no [16:07] well, not crashing now [16:07] I'm still here [16:08] trying "old" profile (if I still have it) [16:08] bah [16:08] now I always get that :( [16:08] and lost historic [16:08] [28240:28240:6462179688:ERROR:extension_prefs.cc(1010)] Bad or missing pref 'state' for extension 'ahfgeienlihckogmohjhadlkjgocpleb' [16:08] [28240:28240:6463559002:ERROR:profile_impl.cc(1085)] Could not initialize login database. [16:08] probably a wrong perm in your profile.. hold on [16:09] [28240:28240:6474209270:ERROR:profile_sync_service.cc(518)] Unrecoverable error detected -- ProfileSyncService unusable.Sync Configuration failed with 3 [16:09] but was fine, till I tried those flags [16:09] :( [16:11] find ~/.config/chromium \! -perm -u=r -ls [16:11] should return nothing [16:11] empty [16:13] find ~/.config/chromium -user root -ls [16:13] nada [16:13] hm [16:16] BUGabundo, file ~/.config/chromium/Default/* | grep SQLite | cut -d: -f1 | xargs -n1 -d '\n' sh -c 'echo $0 >&2;sqlite3 "$0" .schema' >/dev/null [16:16] BUGabundo, anything reported as locked? [16:17] Error: database is locked [16:17] uhuh, which one? [16:17] a bunch [16:17] pastesbin [16:17] http://paste.ubuntu.com/631847/ [16:18] BUGabundo, are you sure no chromium process was running when you did that? [16:19] checking [16:19] $ psx chrome [16:19] 1000 29834 0.0 0.0 17268 904 pts/3 S+ 16:19 0:00 grep --color=auto chrome [16:19] I had two windows opened [16:19] closed it all [16:19] run from cli with flags [16:19] ofc, it crashed before [16:19] but I opened it after reboot, just fine [16:20] not chrome, chromium [16:20] doh [16:20] 1000 3606 0.0 0.4 296004 16584 ? S 14:28 0:00 /usr/lib/chromium-browser/chromium-browser --type=zygote [16:21] close it, or kill it [16:21] ... as cleanly as possible :P [16:22] but why is it left running ? [16:22] stupid chromium OS virus alike [16:22] lol [16:22] after two pkills its closed [16:22] no locks reported [16:22] trying the flags again [16:23] no, the zygote is the master process, opening all the files & dbs, before entering the sandbox where nothing is visible [16:23] testing no flag [16:23] I'm still here ? [16:24] damn, don't you hate non reproduble bigs? [16:24] *bugs [16:24] did you check the "background Apps" flags? [16:24] [30602:30602:7378557193:ERROR:content_settings_pref_provider.cc(763)] Invalid pattern strings: [*.]www.youtube.com, [16:24] no idea what that is [16:24] that's ugly , lots of line reporting... probably bad addon [16:25] I turned off all my flags [16:25] trying one by one now [16:25] fuuuuu [16:26] click to play, seems safe. now location [16:28] webaudio ok [16:29] heck [16:29] now everything works [16:29] :( [16:30] well, if you can reproduce, bisect it. you can use my helper for that: http://people.ubuntu.com/~fta/chromium/bisect-chromium-with-continuous-builds.sh [16:30] now I can't reproduce it anymore [16:31] maybe it was fixed on the last update [16:31] nvidia? [16:31] and when i booted I still had the one from yesterday [16:31] yes [16:31] do you use my flags now? [16:31] no [16:32] just the ones I had before [16:32] more or less, I think [16:38] BUGabundo, maybe my bug 718858 [16:38] Launchpad bug 718858 in nvidia-graphics-drivers "GPU hang using GL on x86_64" [High,Triaged] https://launchpad.net/bugs/718858 [16:38] maybe [16:38] but I'm using nouveua [16:39] or bug 755099 [16:39] Launchpad bug 755099 in unity "X lockup caused by compiz/unity" [Undecided,Incomplete] https://launchpad.net/bugs/755099 [16:39] oh, i know nothing about nouveau [16:39] it doesn't even start here [16:40] no compiz here , either [16:41] BUGabundo, you don't use unity? [16:41] gnome shell? [16:42] gnome3 [16:42] "classic" [16:49] BUGabundo, is your box still reachable over ssh when you freeze? [16:50] no idea [16:51] it's likely a driver issue locking X, so unless you can get a X trace, there's zero chance anyone would care [16:51] ok [16:51] ill try to SSH from my tablet [16:52] ..and even so, there's little chance they'll care, like for my bugs [16:52] last time, i caught a trace from my tablet too ;) [16:53] :) [17:04] BUGabundo, instructions are here: https://wiki.ubuntu.com/X/Backtracing [17:04] I know them [17:04] thanks [19:49] fta: crashed, but couldn't SSH into [21:37] chrisccoulson: about xulrunner removal. xiphos is in process of being ported to webkit & gtk3. I can make it build against gtkhtml to drop xulrunner dependency. === m_conley is now known as m_conley_away