Last edited by psyke83; April 21st, 2009 at 02:02 PM.
Would it make sense to add that repo to the sources.list in order to insure that updates are automagically made available? Should that be done with apt-pinning (of which my understanding is that those repos can be given a lower priority than the Ubuntu repos.) Or would that just open up a bigger can-O-worms?
My understanding is that the need to do this will likely remain until the next version of Ubuntu is released. If that is not correct, then perhaps my concerns are unwarranted.
@oenologist:I just posted to your bug with a workaround that got me super-stable and reasonable performance for everything except games, and allows me to use kwin effects. It's stable under 2.6.28 and 2.6.30
@people asking about "offical" fixes: The consensus on the bug reports seems to be that major fixes to stability or performance regression WILL make it into jaunty-updates. The reason they aren't going into base Jaunty is simply unlucky timing - not likely to get UXA fixed in a supportable way until then. There are quite a few bugs marked with the jaunty-updates milestone, so when those get fixed there will be an approved Jaunty fix for them.
The problem at the moment is that different people get different results when they make the changes suggested here, and some people get the desired performance but then start seeing graphics corruption instead.
So until we've got to the bottom of all the bugs marked critical and found fixes for them that don't introduce NEW critical bugs (!) there won't be an "official" fix... but anything that can be identified and fixed in a stable way WILL be updated.
Looks likely that Jaunty itself will be released with UXA disabled by default because there are still so many unknowns. The lowest common EXA config that -works- for everyone will be deathly slow on some people's hardware, but there doesn't seem to be one config that is stable AND quick at the moment, and stable is definitely better than quick!
At least, that's how I understand it.
Judging by the mix of results, it seems that some testing would be helpful. To that end, I applied the fix as described in the first post (except for the MTRR fix.) It did seem to produce some improvement in frame rates in ppracer. However when I went to load the page with the description for the MTRR fix I found that I had no wireless. My wireless H/W (Eee PC 901) uses the rt2860sta module which is included in linux-image-2.6.28-11-generic but not linux-image-2.6.30-020630rc2-generic. (Someone noticed that the driver is non-free.) I checked the directory where the kernel was found and saw no non-free packages. It looks like I can do limited testing, but w/out wifi this netbook is not so useful.
Edit; suggestions for the command line to run ppracer as a benchmark would be most welcome. I can't seem to get beyond
There's a -f option that I can't seem to get to work.Code:hbarta@bonsai:~$ ppracer -a PPRacer 0.3.1 -- http://racer.planetpenguin.de (c) 2004-2005 The PPRacer team (c) 1999-2001 Jasmin F. Patry<firstname.lastname@example.org> PPRacer comes with ABSOLUTELY NO WARRANTY. This is free software, and you are welcome to redistribute it under certain conditions. See http://www.gnu.org/copyleft/gpl.html for details. get fences failed: -1 param: 6, val: 0 Benchmark error: unable to set course:
Last edited by HankB; April 21st, 2009 at 05:20 PM.
I'm also interested in a repo to keep the intel drivers at the bleeding edge. This fix has solved all regular compiz and video issues (tiling is giving me no problem either), but I notice Google Earth is massively corrupted (not sure if this extends to 3D games since I don't play them). If an update comes along that fixes this and other currently-unknown issues it'd be nice to get it automatically.
I just want to report here that I found a fix for my video problem.
00:02.0 VGA compatible controller : Intel Corporation 82865G Integrated Graphics Controller [8086:2572] (rev 02)
Solution was to revert back to Intel driver 2.4. Go HERE for explanation.
My video now exactly the same as it was using Hardy. Speed is the same. No more choppy avi, vob's. Everything works.
boot_info_script | AMD64 Athlon X2 Dual-Core - nVidia Geforce 6150 SE - 4GB DDR3
Works flawlessly! thank-you I had given up on jaunty because of that.
InstallCode:$ wget -c http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.30-rc2/linux-headers-2.6.30-020630rc2-generic_2.6.30-020630rc2_amd64.deb http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.30-rc2/linux-headers-2.6.30-020630rc2_2.6.30-020630rc2_all.deb http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.30-rc2/linux-image-2.6.30-020630rc2-generic_2.6.30-020630rc2_amd64.deb http://ftp.us.debian.org/debian/pool/main/libd/libdrm/libdrm2_2.4.9-1_amd64.deb http://ftp.us.debian.org/debian/pool/main/libd/libdrm/libdrm-intel1_2.4.9-1_amd64.deb http://http.us.debian.org/debian/pool/main/x/xserver-xorg-video-intel/xserver-xorg-video-intel_2.7.0-1_amd64.deb
Testing now to see how well it works...Code:sudo dpkg -i linux-headers-2.6.30-020630rc2_2.6.30-020630rc2_all.deb linux-headers-2.6.30-020630rc2-generic_2.6.30-020630rc2_amd64.deb linux-image-2.6.30-020630rc2-generic_2.6.30-020630rc2_amd64.deb libdrm-intel1_2.4.9-1_amd64.deb xserver-xorg-video-intel_2.7.0-1_amd64.deb
Disabling tiling is probably not what you actually want. A lot of the complaints about poor 2D performance with 945 are due to tiling being disabled due to GEM not supporting the crazy memory layout needed.