| mlankhorst | morning! | 05:49 |
|---|---|---|
| tjaalton | looks like fullscreen flash is a killer | 07:10 |
| tjaalton | seems to hang my ivb dri | 07:11 |
| mlankhorst | :O | 07:11 |
| tjaalton | 1080p | 07:11 |
| tjaalton | not right away | 07:12 |
| tjaalton | restarting compiz gives just a white screen | 07:13 |
| tjaalton | oh, that was due to flash | 07:13 |
| mlankhorst | I have no 1080p screen to test with :/ | 07:15 |
| tjaalton | nah it was the video resolution | 07:18 |
| mlankhorst | 1080p youtube? | 07:21 |
| mlankhorst | what are you testing on btw, what release/kernel | 07:24 |
| tjaalton | raring | 07:29 |
| tjaalton | yeah youtube | 07:29 |
| tjaalton | I'll try to repro | 07:29 |
| mlankhorst | doing the same atm, no lockup yet | 07:30 |
| tjaalton | so it's not the driver getting stuck, since killing flash & compiz worked | 07:37 |
| tjaalton | probably triggered by the fade out | 07:37 |
| tjaalton | can't reproduce, dunno what happened :) | 07:46 |
| mlankhorst | yeah same | 07:46 |
| mlankhorst | weird, x1.14 broke my laptop, I think I've seen the failure before though, in 1.13 | 09:04 |
| mlankhorst | tjaalton: I uploaded a new xserver to canonical-x ppa, should fix the hybrid case again :) | 09:30 |
| tjaalton | ah, cool | 09:33 |
| mlankhorst | not many bugs that look juicy currently, I think I'll try to play some with mir again | 10:03 |
| zzippy | Anyone knows why xserver-xorg-core 1.14.1 does not work with nvidia319.12/optimus while old version 1.14.0 does work? | 13:41 |
| zzippy | talking about X-staging PPA | 13:41 |
| mlankhorst | try comparing the logs | 13:42 |
| zzippy | mlankhorst: which exactly? Xorg.0.log? | 13:43 |
| mlankhorst | yeah | 13:43 |
| zzippy | I already made a logverbose 6 nvidia bugreport. can't see no difference. Well, it works, but it gives a blank screen | 13:44 |
| zzippy | while 1.14.0 is fine | 13:44 |
| zzippy | ok, will compare again, maybe I overlooked something. | 13:47 |
| zzippy | thanks! | 13:47 |
| zzippy | mlankhorst: maybe you see something, pasted both Xorg.0.log files http://paste.ubuntu.com/5616034/ | 14:12 |
| mlankhorst | zzippy: both look fine to me | 14:43 |
| mlankhorst | oh nm | 14:43 |
| mlankhorst | well I do see xvmc not being initialized on second one, but that was on purpose | 14:45 |
| mlankhorst | and the second one spams a lot of this | 14:46 |
| mlankhorst | [ 37.759] (WW) NVIDIA(GPU-0): Unable to read EDID for display device CRT-0 | 14:46 |
| zzippy | I already fed edid as file, then it stops complaining. | 14:47 |
| zzippy | But now difference. | 14:47 |
| mlankhorst | http://paste.debian.net/1125/ | 14:49 |
| mlankhorst | seems to be the difference between .0 and .1 in applied patches | 14:49 |
| mlankhorst | maybe you can isolate it from there | 14:49 |
| mlankhorst | my guess is one of the 5 patches starting from xf86crtc-don-t-use-display-for-vx-vy-for-gpu-screens.patch | 14:50 |
| zzippy | mlankhorst: erm, ok, so you may want to patch 1.14.1 accordingly some day? Wouldn't be too bad, since whole-desktop-on-nvidia seems to be 3-4x faster than that bumblebee/virtualgl workaround. | 14:59 |
| === chris234 is now known as RAOF | ||
| mlankhorst | zzippy: its the other way around, 1.4.1 has those applied, 1.14.0 doesnt | 16:23 |
| zzippy | mlankhorst: whatever. Thats beyond my horizon. Just wanted somebody to know this who might be in the position to change something, so ubuntu users also can use the new nvidia stuff. | 16:37 |
| mlankhorst | can't you rebuild xserver with some of the new patches in debian/patches/series commented out then? | 16:38 |
| zzippy | mlankhorst: guess I can, theoretically. But without knowing what I am really doing, it would take me a day ;) ...maybe I can ask our forums packaging "guru" to assist. | 16:51 |
| mlankhorst | because adding a # and rebuilding is hard? | 17:46 |
| zzippy | If I should manage to rebuild a working version, randomly outcommenting patches, what then? | 18:20 |
| mlankhorst | well knowing which patch is causing the breakage is enough information for me | 18:25 |
| zzippy | ok. Just reading "How to: Recompiling / Rebuild Debian / Ubuntu" | 18:31 |
| bjsnider | zzippy, install the build dependencies | 18:48 |
| bjsnider | apt-get source your source package | 18:48 |
| bjsnider | make your changes to the patches or whatever | 18:48 |
| zzippy | bjsnider: already have. | 18:49 |
| bjsnider | then run dpkg-buildpackage -us -uc from the source dir | 18:50 |
| bjsnider | that will make debs you can install | 18:50 |
| bjsnider | you may need to install build-essential and whatnot too | 18:51 |
| zzippy | ok, will try. thanks! Will report if I succeed. | 18:54 |
| zzippy | cy | 18:54 |
| wolfgang8741 | I was able to confirm that the bug https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/1170418 affects mesa 9.1.1 and not 9.0.3 via a downgrade. Also attached a backtrace of compiz, but unsure if the trace was successful or useful. | 19:29 |
| ubottu | Launchpad bug 1170418 in compiz (Ubuntu) "Unity Launcher and window borders missing upon login when using 3 monitors" [Undecided,Confirmed] | 19:29 |
| === ripps_ is now known as ripps | ||
| === seb128_ is now known as seb128 | ||
| zzippy | mlankhorst: rebuilt 1.14.1, disabled 7 patches. http://paste.ubuntu.com/5617521/ , works. Where should I start with re-enabling? | 22:08 |
| bryce | zzippy, re-enable the first half of the patches and re-test. | 22:12 |
| zzippy | will do. | 22:16 |
| bjsnider | wow, that guy does whatever you tell him to! | 22:46 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!