=== chihchun_afk is now known as chihchun === shuduo-afk is now known as shuduo === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === alan_g is now known as alan_g|lunch === dandrader is now known as dandrader|afk === alan_g|lunch is now known as alan_g === dandrader_ is now known as dandrader === dandrader is now known as dandrader|lunch === dandrader|lunch is now known as dandrader [17:05] kdub, ping [17:05] hello mzanetti [17:06] kdub, hi. I'm investigating in the crashers we have on turbo. seems all of them choke on the same boost exception. [17:06] I'd like to test the branch you linked [17:06] do you happen to have a package I could install? [17:07] yes mentioned in tracking bug, https://bugs.launchpad.net/mir/+bug/1554635 [17:07] Launchpad bug 1554635 in Canonical System Image "Importing contacts trigger unity restart - Mir crashed with exception 'failed to add sync point to command buffer'" [Critical,Confirmed] [17:07] mzanetti, still searching for the dmesg and /system/bin/logcat log at the time of the problem, if you have that [17:07] thanks! [17:09] kdub, dmesg: http://paste.ubuntu.com/15408731/ [17:11] kdub, logcat doesn't seem to print anything when this happens [17:11] kdub, well, here it is: http://paste.ubuntu.com/15408756/ [17:12] kdub, the first few lines are before the crash, and then it has stuff when restarting things up [17:12] but immediately before/during the crash, no output happened [17:12] mzanetti, thanks, yeah, doesnt look like anything suspicious [17:13] trying your debs now [17:13] mzanetti, thanks [17:15] kdub, what packages are required? just libmirserver? [17:17] mir-platform-graphics-android8_0.20.3-0ubuntu1_armhf.deb is where the fix is, but the mircommon ones might be needed too === mhall119 is now known as mhall119_canonic === mhall119_canonic is now known as mhall119_cct [17:23] kdub, seems to still crash [17:24] hmm, unfortunate [17:25] it shouldn't crash in the same way though, the exception should be ignored and printed to the log? with an egl error code hopefully [17:25] still seing an exception. let me get a cleaner log === mhall119_cct is now known as mhall119_CC [17:26] kdub, http://paste.ubuntu.com/15408876/ [17:29] failed with EGL_BAD_ALLOC [17:30] ? === mhall119_CC is now known as mhall119_CCT [17:32] mzanetti, well, I'm not sure what that might mean, but I do see how hte exception is still getting out [17:32] mzanetti, let me rebuild quickly so it will be caught and maybe things will just continue [17:33] mzanetti, thanks for the help [17:33] kdub, no prob. please ping me when you have something more to test [17:34] mzanetti, sure, rebuilding packages now === mhall119_CCT is now known as mhall119 === chihchun is now known as chihchun_afk === alan_g is now known as alan_g|EOD [18:09] mzanetti, package built in https://private-fileshare.canonical.com/~kdub/mir-debs-1554635-attempt2.tar.gz with second attempt at catching the exception === dandrader is now known as dandrader|afk [18:25] kdub, crashing, with this message: [18:25] error installing sync point failed to add sync point to command buffer 3003 [18:29] mzanetti, hmm, don't know how the exception is propogating then [18:30] kdub, k, will attach a debugger, gimme a few, just need to flash another deivce [18:30] mzanetti, thanks, and if that doesnt work i'll just log the error and not throw an exception at all === dandrader|afk is now known as dandrader [19:27] mzanetti, take three: https://private-fileshare.canonical.com/~kdub/mir-debs-1554635-attempt3.tar.gz [19:28] should just log without throwing... if it doesn't crash, great, otherwise, its maybe a resource leak [19:36] kdub, logs, not crashing. confirming it "fixes" all the reported unity crashers in that turbo blockers list [19:36] mzanetti, great, are you still seeing a logged error in the unity8 log? I'm interested in if its a loss-of-context problem [19:36] I do see the EGL warning, yes [19:37] is it the 3003 error code? or is it something about having no EGLDisplay [19:39] kdub, hmm... I saw some EGL warning at first, but trying to repro now, I can't see it any more [19:39] lemme restart and try fresh [19:39] it's not crashing any more, neither by switching between between browser/messages, nor the content ub imports. both would trigger it reliable before [19:40] kdub, I get this after a fresh restart and then reproducing: [2016-03-17 19:40:29.234223] egl sync fence: failed to add sync point to command buffer: 0x3003 [19:40] kdub, however, doing it a second time, it won't print this any more [19:41] mzanetti, as long as its not spamming the log with that message, its probabyl an acceptable fix [19:41] ak ok... got it printed a second time now [19:41] although if/when I get a device, i'd be interested to dig further [19:42] it's not spamming tho... it's once per reproducing steps [19:42] kdub, I'll send a mail to victor etc letting him know about this fix, ok? [19:42] will cc you [19:42] right, and as long as its not spamming its probably acceptable to move through [19:42] mzanetti, sure, and thanks for the remote help [19:45] AlbertA, we probably want https://code.launchpad.net/~kdub/mir/fix-1554635/+merge/289419 to ride along in 0.20.3 to unblock that device [19:46] kdub: ok [19:48] thanks AlbertA