| tormod | hmm https://lists.ubuntu.com/archives/ubuntu-devel/2008-May/025471.html indicates 1.2.3ubuntu1 is fine | 00:07 |
|---|---|---|
| tormod | bryce, ok then the version was fine already (was also what dch -i gave me) | 00:10 |
| bryce | ah ok | 00:15 |
| bryce | needs update-maintainer run | 00:28 |
| bryce | tormod, uploaded | 00:30 |
| Ng | anyone else seeing odd resume crashes in karmic atm? | 09:05 |
| Ng | X sort of comes back a bit, but windows don't render properly and it'll lock up shortly after, although sysrq still works | 09:05 |
| d6g | I feel the performance of compiz and cairo-related software is a bit sluggish in karmic, anyone knows where should i start to figure out what is the problem? (i'm using ati mobility x300 with kernel 2.6.31-8-generic & radeon driver) | 14:21 |
| === d6g is now known as d6g|away | ||
| Ng | bah, crash | 15:21 |
| Ng | http://paste2.org/p/410429 | 15:22 |
| Ng | what should I do with that? kernel bug? | 15:22 |
| kees | Ng: that's where I'd file it. it's kernel, but it's intel drivers driver. kernel team should be able to redirect correctly. | 17:05 |
| Ng | kees: yeah that's what I was thinking. I even started filing it, but ubuntu-bug seems to be able to make firefox crash, and I got sidetracked finding the bug for that and subscribing to it ;) | 17:06 |
| kees | urgh, it'd be funny if it weren't so sad. :P | 17:06 |
| Ng | hehe | 17:06 |
| Ng | it's such a weird bug, it spawns a firefox process to take you to LP and most of the time that spawned firefox gets an X error and dies, which triggers an apport request for that | 17:07 |
| Ng | but it's known and mdz reported it, so I feel safe in not doing anything else about that ;) | 17:07 |
| kees | heh | 17:07 |
| mac_v | anyone knows about the radeon driver and the slow FPS? is this the bug related to this issue> http://bugs.freedesktop.org/show_bug.cgi?id=21508 | 17:28 |
| ubottu | Freedesktop bug 21508 in Driver/Radeon "low fps with radeon driver" [Normal,New] | 17:28 |
| mac_v | just had a doubt since comments mention "radeon-rewrite" | 17:29 |
| bryce | mac_v, the radeon-rewrite stuff is in place now in karmic | 18:07 |
| mac_v | bryce: i'v been using the radeon.modeset=1 , kms since the kernel -7, but i get half of the FPS i get without the kms :( | 18:09 |
| mac_v | its ~500 with kms but without its ~1000 | 18:09 |
| mac_v | bryce: is there an upstream bug , you could direct me to regarding this? or is it the above bug ? | 18:11 |
| bryce | mac_v, hard to say, the reporters did not give much solid info | 18:14 |
| bryce | mac_v, but I do know that kms on -ati is known to have a performance regression, but I don't know the bug #. | 18:15 |
| mac_v | oh :( | 18:15 |
| Ng | ok, getting unusable X after every suspend now | 19:47 |
| hyperair | at least you're not getting random panics after every other suspend/hibernate | 19:53 |
| Ng | is that better than having to restart after every suspend? ;) | 19:57 |
| bryce | Ng, bug #? | 19:58 |
| hyperair | Ng: how is your X unusable anyway? all i need to do is restart compiz after every suspend. | 20:00 |
| Ng | bryce: just waiting for the 3g network to upload it | 20:01 |
| bryce | hyperair, did you test the kernel patch proposed for the compiz freeze bug yet? | 20:01 |
| Ng | hyperair: it might be that, I didn't actually try that. it just doesn't seem to draw windows/borders | 20:01 |
| hyperair | bryce: no i didn't. what patch is that? | 20:01 |
| hyperair | Ng: that's the exact problem. restarting compiz should do the trick. just blindly alt+f2 and type compiz ;-) | 20:02 |
| Ng | heh | 20:02 |
| hyperair | that's what i do. ended up sending compiz to a few friends if the run dialog didn't focus properly | 20:02 |
| bryce | hyperair, search lp for the "compiz 100% cpu" bug | 20:03 |
| bryce | hyperair, filed against linux | 20:03 |
| bryce | anyway, I think both of you have that same bug | 20:03 |
| hyperair | bryce: but my compiz doesn't take 100% cpu | 20:03 |
| hyperair | at least, i don't think it does. | 20:03 |
| bryce | hyperair, which is why you ought to test the patch ;-) | 20:03 |
| hyperair | heh yeah, considering i'm already running a patched kernel | 20:04 |
| bryce | 100% cpu is just one symptom, the patch fixes a number of situations where the new mesa causes failures | 20:05 |
| bryce | you can also try downgrading to the x-retro ppa's mesa 7.5, but that's just a workaround | 20:05 |
| * bryce back to bugz | 20:05 | |
| Ng | epic 3g speed fail aside, I'm up for trying a patch | 20:05 |
| hyperair | bryce: i'm running a git kernel and the said patch is already in the kernel i think | 20:09 |
| Ng | hyperair: out of interest, what hardware are you seeing this on? | 20:15 |
| hyperair | Ng: i965 8086:2032 | 20:16 |
| hyperair | 2a03 i mean | 20:16 |
| Ng | I'm assuming bryce meant bug 419264, and it just strikes me a little bit that people there are seeing it all over the place, I'm only seeing it on resume | 20:17 |
| ubottu | Launchpad bug 419264 in linux "Uses 100% CPU with latest mesa/libdrm update" [High,In progress] https://launchpad.net/bugs/419264 | 20:17 |
| hyperair | yeah i saw that too | 20:18 |
| hyperair | the last comment has a commit hash | 20:18 |
| hyperair | which is in my git tree of the kernel | 20:18 |
| bryce | jbarnes, http://people.canonical.com/~bryce/Graphs/drivers.svg | 20:41 |
| jbarnes | woo | 20:42 |
| jbarnes | beating ati finally :) | 20:42 |
| jbarnes | looks like nvidia will be higher soon too | 20:43 |
| bryce | yep | 20:45 |
| ScislaC | 9.04? Who uses /that/? ;) | 20:46 |
| bryce | I wish nouveau would develop faster, I hate putting time into triaging -nvidia bugs, always seems like a waste | 20:46 |
| bryce | ScislaC, ;-) | 20:46 |
| === bryce changed the topic of #ubuntu-x to: https://wiki.ubuntu.com/X | ||
| ScislaC | :P | 20:47 |
| hyperair | what was it before? | 20:47 |
| ScislaC | Ubuntu 9.04 released! | https://wiki.ubuntu.com/X | 20:47 |
| hyperair | ah | 20:48 |
| === d6g|away is now known as d6g | ||
| === ripps_ is now known as ripps | ||
| === d6g is now known as d6g|away | ||
| Ng | hyperair: yeah restarting compiz seems to do the trick | 23:39 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
