You are not logged in.
combuster wrote:It is allowed by default and it was in 2.6.29 also but u need to compile it with support for that option (2.6.29-ARCH wasn't, dunno why)...
uh?
The glxgear non issue http://qa-rockstar.livejournal.com/7869.html
As for KMS not enabled by default, this is true, but you need not recompiling a kernel with whatever options: http://bbs.archlinux.org/viewtopic.php?id=69083&p=1
Hey, Arch Linux is a nice and simple and easy to use distribution. And I just love KMS, but I tend to reboot my laptop instead of hibernating it.
That's why everyone says glxgears is not a good benchmark tool
Offline
http://cworth.org/intel/driver_stability/ quite interestin reading. Makes me more optimistic, anyway I just found out that even with recent combination kernel 2.6.29, xf86-video-intel 2.7.1, GMA 965 ( X3100 ) is openarena running smooth at 1280x800, KMS not enabled.
To developers: I really appreciate your hard work doing such a great distro. It taught me a lot, not only in linux area. And now my question : Is it possible that new kernel 2.6.30 will have KMS enabled by default? ( I know that currently only intel drivers should able to use it )
Thanks a lot
Offline
Currently it is optional only for intel drivers (as i recall enable kms by default is an option in menuconfig that relates to intel driver only). I don't think that developers will enable that, not untill 2.6.31 anyway. But I would like to hear their opinion on that. Anyway we will see when 2.6.30 reach core...
Offline
With the new kernel, my framebuffer console doesn't use the whole screen with KMS.
Some space at the right part and at the bottom (two lines) are not used.
After next reboot everything is well again. And if I reboot again, the problem is back?
< edit/>
Reboot doesn't help anymore.
Last edited by hoschi (2009-06-11 15:03:28)
Offline
Yeah I've got same problem, 2 lines from the bottom is blank, but thats ql compared to kms being enabled by default in kernel and without loading i915 early, first 4 or 5 lines are garbled totally... Only thing is I always get 2 lines blank, even on restart...
Offline
See this thread - http://bbs.archlinux.org/viewtopic.php? … 38#p567438
Offline
I'm using UXA and performance is no better, games like ET are still unplayable on X3100 965GM. What's worse the xf86-video-intel 2.8 beta has some serious performance regressions (tried it today) so we can expect even lower performance in yet another driver release.
You need to install an RTFM interface.
Offline
Happy in my case everything is working wonderfull, kernel from testing, minimal xorg.conf ( only touchpad related ), xf86-video-intel-newest from AUR, X3100 965GM true combat playable at 1280x800
Offline
Like another thread that saw, I got a kernel panic on my 32 bit machine. The 64 bit one is happy.
Offline
error: failed to prepare transaction (could not satisfy dependencies)
:: virtualbox-modules: requires kernel26<2.6.30
Gotta wait...
never trust a toad...
::Grateful ArchDonor::
::Grateful Wikipedia Donor::
Offline
there is something odd. with this kernel i wanted to test the intel drivers from aur (2.8-RC1) if i recall correctly..... now xorg is ignoring accelmethod. i cant use exa anymore. anyone experiencing this?
Offline
Is fastboot enabled in the testing package?
Offline
there is something odd. with this kernel i wanted to test the intel drivers from aur (2.8-RC1) if i recall correctly..... now xorg is ignoring accelmethod. i cant use exa anymore. anyone experiencing this?
Offline
@toad
Weird, my vbox kernel modules compiled without any errors (vbox_build_modules on 2.6.30)...
Offline
eldragon wrote:there is something odd. with this kernel i wanted to test the intel drivers from aur (2.8-RC1) if i recall correctly..... now xorg is ignoring accelmethod. i cant use exa anymore. anyone experiencing this?
thanks, now to go and request a patch for mesa
Offline
Ranguvar wrote:WoW?
Yep, the kernel is getting bloated....
Nope. Do you think so, because entire kernel size is bigger then previous one? Your vmlinuz won't be bigger/bloated, because of new file systems (that's great!) or features. You just gained possibility to enable those extras. It will be bloated (however, only for maintainers) if they will add Xen in its current, crappy form. Thankfully Xen devs will bring smarter patches next time.
Offline
ok, I'll do it guys.
http://www.google.co.th/search?q=WoW
Offline
Allan wrote:Ranguvar wrote:WoW?
Yep, the kernel is getting bloated....
Nope. Do you think so, because entire kernel size is bigger then previous one? Your vmlinuz won't be bigger/bloated, because of new file systems (that's great!) or features. You just gained possibility to enable those extras. It will be bloated (however, only for maintainers) if they will add Xen in its current, crappy form. Thankfully Xen devs will bring smarter patches next time.
I'm pretty sure that his response was a joke in response to the capitalization of "WoW", implying that World of Warcraft has been included in the kernel.
Is ext4 online defragmentation available yet? I haven't been paying attention to its development (still happily using ext3 here).
Last edited by Xyne (2009-06-12 16:07:49)
My Arch Linux Stuff • Forum Etiquette • Community Ethos - Arch is not for everyone
Offline
No ext4 online defragmentation in the kernel yet.
Offline
error: failed to prepare transaction (could not satisfy dependencies) :: virtualbox-modules: requires kernel26<2.6.30
Gotta wait...
yes. you have to wait until is moved to core. i had prepared the package, if you want i can give you
Give what you have. To someone, it may be better than you dare to think.
Offline
This kernel have problem with sound: http://patchwork.kernel.org/patch/29569/. On my integrated AC97 sound is skipping both in tremulous and moc when they are run together.
Offline
This kernel have problem with sound: http://patchwork.kernel.org/patch/29569/. On my integrated AC97 sound is skipping both in tremulous and moc when they are run together.
Same here, really annoying, but I'll be able to live with it until it's fixed (hopefully soon!)
Offline
first impression - slowdown while copying large files on x86_64 seems to be fixed.
copying a multi gigabyte vm harddisk file doesn't result in any system slowdown at all!
edit: i'm using ext4 on all partitions except /boot
Last edited by schuay (2009-06-12 21:13:11)
Offline
copying a multi gigabyte vm harddisk file doesn't result in any system slowdown at all!
Users of ext3 should read "Solving the ext3 latency problem". For ext3 'writeback' mode is now default and it works great, but the default scheduler and ext3 in general got some improvements and I can confirm that with 'ordered', it's important to me because in addition I have encrypted file systems. Writeback has more potential to mess your data, especially in case of a system crash - and thanks to the intel driver we have pleanty of those today.
If you don't want to risk it, to switch back to ordered you can set "data=ordered" in fstab for all paritions but the root file system. Because you will end up with your / mounted as read only if you do, for the root fs you can append "rootflags=data=ordered" to the kernel trough your boot loader configuration.
You need to install an RTFM interface.
Offline
I'm pretty sure that his response was a joke in response to the capitalization of "WoW", implying that World of Warcraft has been included in the kernel.
Right XD. Thanks for pointing this ^^.
@schuay
Great to hear that. While fglrx is no more supported I hope we'll see this kernel soon in stable repo
Offline