| Sarvatt | Ng are you using the 2.6.31-rc3 kernel? | 00:08 |
|---|---|---|
| Ng | yup | 00:08 |
| Sarvatt | input modules werent getting frozen on suspend and they were screwing up on resume beforee | 00:08 |
| Sarvatt | why the heck is this high importance?! | 08:35 |
| Sarvatt | https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/399559 | 08:35 |
| ubottu | Ubuntu bug 399559 in xserver-xorg-video-ati "ATI Radeon Mobility KMS modeset=1 22% slower than vesa" [High,Incomplete] | 08:35 |
| bryce | it could probably be reduced | 10:25 |
| Ng | tjaalton did it an hour ago or so :) | 10:32 |
| hyperair | isn't there something seriously wrong when any driver gets slower than vesa? | 10:33 |
| hyperair | i mean all other drivers are supposed to have GPU-specific features which improve performance among other things, right? | 10:34 |
| Sarvatt | i think the question is more, is there a card made in the past 10 years with a recent ddx that is faster than vesa at gtkperf? | 10:43 |
| hyperair | O_o you mean vesa actually beats them all? | 10:44 |
| hyperair | if that was the case then something should be done shouldn't there? like copying some code from vesa or something =\ | 10:44 |
| Sarvatt | how about, does gtkperf performance mean anything or is it glxgears for 2d? :D | 10:46 |
| hyperair | ah. ._. | 10:47 |
| Ng | glxgears isn't a useful test of anything, gtkperf tests the performance of rendering gtk widgets, which is something that desktops do do, right? :) | 10:52 |
| Ng | it's just not something they do in tight loops | 10:53 |
| jcristau | hyperair: if you accelerate some stuff, but not everything, you sometimes have fallbacks. and falling back once in a while is enough to make performance worse than an all-software thing. | 10:55 |
| hyperair | i see | 10:56 |
| hyperair | that maketh much sense. | 10:56 |
| Sarvatt | http://global.phoronix-test-suite.com/index.php?k=profile&u=robert-14462-8152-19051 | 12:14 |
| Sarvatt | yay my radio button speed is going down! :D | 12:15 |
| apw | Sarvatt, do we expect xrandr to work with kms? | 12:15 |
| Sarvatt | yep, not working for you? | 12:16 |
| apw | not sure. ran tremulous and it left me in 640x480 | 12:17 |
| apw | i uses xrandr to go to 1024x768 | 12:17 |
| apw | and once there i could not change again | 12:17 |
| Sarvatt | intel or radeon? | 12:19 |
| apw | intel | 12:19 |
| Sarvatt | downloading it now to see if it happens to me | 12:21 |
| apw | i guess i should try changing now, see if its a general issue or not | 12:23 |
| Sarvatt | worked fine here, thats odd | 12:24 |
| apw | hrm ... | 12:24 |
| apw | yeah switching back and forth seems to work | 12:24 |
| apw | apw@dm$ xrandr -q | 12:25 |
| apw | xrandr: output LVDS1 cannot use rotation "normal" reflection "none" | 12:25 |
| Sarvatt | i didnt do anything, just started it and it went to 640x480 then joined a game and quit and it went back to 1024x600 | 12:25 |
| apw | what does that mean? | 12:25 |
| jcristau | i think it means the driver is confused | 12:26 |
| jcristau | :) | 12:26 |
| apw | i just ran extreme tux racer and on quit i am in some other mode, not the one i am meant to be in | 12:26 |
| Sarvatt | robert@ubuntu-9{/opt/source/xorg-pkg-tools}:xrandr -q | 12:26 |
| Sarvatt | xrandr: output LVDS1 cannot use rotation "normal" reflection "none" | 12:26 |
| apw | and am getting that poop | 12:26 |
| Sarvatt | dunno but I get it too! lol | 12:26 |
| apw | and now i can't fix it | 12:26 |
| apw | anything i can collect before i logout | 12:27 |
| Sarvatt | apw: just apply the settings in display properties | 12:29 |
| Sarvatt | resets it | 12:29 |
| Sarvatt | i didnt change anything, just applied it and its working right again | 12:29 |
| apw | ok HOW did it do it when xrandr cannot! | 12:30 |
| Sarvatt | just hit another bug like that, xrandr command line tool wasnt in sync with the server changes | 12:30 |
| Sarvatt | guess theres no reflection property in KMS and the game sets the value on exit and confuses xrandr? | 12:38 |
| * Sarvatt has no idea | 12:38 | |
| apw | Sarvatt, must be something like that | 14:13 |
| apw | Sarvatt, when i VT swtich back to X under KMS the screen comes back nice and fast, all there, and then X redraws it all and it all flashes nastily, is that nromal | 14:14 |
| Sarvatt | i get that too, guess its normal | 14:18 |
| Sarvatt | it used to be alot worse, VT>X would change modes in 1.6.0 even under KMS | 14:18 |
| === mvo__ is now known as mvo | ||
| Solarion | any idea why karmic gdm just keeps cycling instead of actually coming up? | 19:28 |
| bryce | Solarion, check your gdm logs | 20:45 |
| fishor | hallo all, did any one has some problem with tadays intel driver on gm915 graphic? xorg on my latop goin to resart loop. workaround is to install oder one. | 21:09 |
| fishor | s/oder/older/ | 21:09 |
| jbarnes | Sarvatt: libdrm 2.4.12? | 22:17 |
| Sarvatt | havent updated it because it was just a version bump with no changes since the last one. will do it in a bit, in the middle of setting up a pc now | 22:18 |
| jbarnes | cool | 22:19 |
| jbarnes | it has a big bo cache fix | 22:19 |
| jbarnes | which should keep gem object usage waaaay lower than it is today | 22:19 |
| Sarvatt | already got that commit in there | 22:20 |
| Sarvatt | and its super nice! | 22:21 |
| Sarvatt | 71MB gem_objects, 3 days uptime with compiz :D | 22:21 |
| jbarnes | cool | 22:24 |
| Sarvatt | phew, took 2.5 hours to set up a second stage debootstrap of arm karmic on my G1 | 22:27 |
| Sarvatt | beat qemu by 2 hours at least | 22:28 |
| Sarvatt | now to compile pixman for no reason | 22:30 |
| bryce | amazingly, I've almost made it through all the -intel bugs for the first time in a long long time | 22:40 |
| bryce | just 15 bugs left to upstream | 22:40 |
| jcristau | can i get a clone of Sarvatt for debian? | 22:40 |
| Sarvatt | what do you mean? i doubt i packaged it right, i am still very new to things but heres the changelog https://edge.launchpad.net/~sarvatt/+archive/xorg-testing/+sourcepub/676434/+listing-archive-extra | 23:03 |
| Sarvatt | dreading packaging all these changes for xserver master in the past few days :( | 23:03 |
| Sarvatt | yuck, looks like it'll take another 2-3 hours to bring in the build deps for it :D | 23:11 |
| jcristau | hehe | 23:12 |
| Sarvatt | you're a monster bryce, thats so many bugs! | 23:37 |
| * bryce wields broom of bug death | 23:42 | |
| Sarvatt | ugh, i dont even know where to start. inputproto libxi libX11 xextproto libXext fixesproto libXtst renderproto and still have to rebuild all video drivers except vesa intel and ati. whee! :) | 23:45 |
| Sarvatt | server 1.6 branch is begging for me to come back to it :) | 23:47 |
| Sarvatt | i should have just done this in a debian chroot, for some reason karmic is super slow unpacking things on arm | 23:59 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!