You are not logged in.
@enihcam - arch packages = arch linux only.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Repo packages are now signed and my gpg key has been uploaded to the key server. If you are also using pacman 4, please let me know if the repo behaves as expected with package signing etc.
Last edited by graysky (2011-11-08 09:34:08)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Repo packages are now signed and my gpg key has been uploaded to the key server. If you are also using pacman 4, please let me know if the repo behaves as expected with package signing etc.
Will those who still use pacman 3.5.x have problem?
Proud Arch Linux user since 2007.
Offline
@twilight0: shouldn't be a problem. Keys are something that pacman 4 uses, 3.5.x doesn't even know they exist and because of that it shouldn't trouble anything
Last edited by ethail (2011-11-08 12:43:37)
Best Testing Repo Warning: [testing] means it can eat you hamster, catch fire and you should keep it away from children. And I'm serious here, it's not an April 1st joke.
Offline
Repo packages are now signed and my gpg key has been uploaded to the key server. If you are also using pacman 4, please let me know if the repo behaves as expected with package signing etc.
Very cool! But I think it would be usefull if you published your key fingerprint on the wiki or in the first post for this thread (both are even better). And your key id is not known, maybe it should be published too.
Satisfied users don't rant, so you'll never know how many of us there are.
Offline
Repo packages are now signed and my gpg key has been uploaded to the key server. If you are also using pacman 4, please let me know if the repo behaves as expected with package signing etc.
@graysky: Thanks for getting this set up before pacman 4 hits the core repository.
Last edited by dhave (2011-11-08 19:57:28)
Offline
@Graysky
In the linux-ck Wiki, in the section regarding enabling the BFQ I/O scheduler, there is this warning:
Warning: The current package, 3.1.0-2 does NOT have the BFQ available due to upstream having not yet released it. A future release will be published once this has occurred.
In the bfq-iosched user group Paolo has made an announcement that BFQ has been ported (and upgraded along the way) to Kernel 3.1:
ANNOUNCE bfq-v3r1 for .38, .39, 3.0 and 3.1
So can we use bfq in latest packages from your repo now? or do we need to wait for you to rebuild again?
Offline
@denisfalqueto - I will post it both places, but pacman should automatically grab it from the keyserver where it resides.
@Evanlec - building now.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
@graysky: I may be the only one, but, after installing linux-ck-corex 3.1-4 I couldn't launch X ("id 'x' respawning too fast").
To fix this, I rebooted linux-ck-corex 3.1.4 into runlevel 3 and rebuilt nvidia-ck from the AUR, first editing the "depends=" line in the PKGBUILD to read as follows:
depends=('linux-ck-corex>=3.1' 'linux-ck-corex<3.2' "nvidia-utils=${pkgver}")
After installing this newly built nvidia-ck (which required removing nvidia-ck-corex 285.05.09-5), I was able to reboot and launch X with no trouble.
As I say, I may be the only one with this problem, but just in case I'm not, this was my workaround. I imagine if there are others encountering this a new nvidia-ck-corex may be needed (?).
Thanks again for all your work. Recent kernel updates have kept you especially busy of late.
Last edited by dhave (2011-11-09 04:23:10)
Offline
@dhave: you're not the only one. I experienced this problem as well. I've just decided to roll back to 3.1-3 for now…
Offline
Same problem here, rebuilding nvidia-ck from AUR,
My Xorg error :
[71.994] (EE) Failed to load module "nvidia" (module-specific error, 0)
there was a message about bad configuration argument or something in dmesg but didn't jot it down before reboot.
Last edited by Evanlec (2011-11-09 07:12:00)
Offline
@enihcam For server use you wont see much of a difference, because this kernel is optimized for desktop use. Why dont you build your server with archlinux?
Last edited by twilight0 (2011-11-09 09:03:49)
Proud Arch Linux user since 2007.
Offline
Thanks for letting me know about this nvidia issue. Will look into it and fix.
Last edited by graysky (2011-11-09 10:54:59)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Same problem here with the VirtualBox kernel modules (failing to load after successful build), message says "Invalid Argument" when trying to load. Strange...
EDIT:
here's the more verbose output from the kernel log:
(...)
Nov 9 14:47:52 localhost kernel: [ 212.946708] vboxnetflt: Unknown symbol VBoxHost_RTUuidFromStr (err -22)
Nov 9 14:47:52 localhost kernel: [ 212.946710] vboxnetflt: no symbol version for VBoxHost_RTErrConvertToErrno
Nov 9 14:47:52 localhost kernel: [ 212.946712] vboxnetflt: Unknown symbol VBoxHost_RTErrConvertToErrno (err -22)
Nov 9 14:47:52 localhost kernel: [ 212.946717] vboxnetflt: no symbol version for VBoxHost_RTLogRelDefaultInstance
Nov 9 14:47:52 localhost kernel: [ 212.946718] vboxnetflt: Unknown symbol VBoxHost_RTLogRelDefaultInstance (err -22)
And forgot to mention that nvidia is failing to load for me, too.
Last edited by lynix (2011-11-09 13:53:43)
Offline
Update: just caught the upgrade of virtualbox-sources to pkgrel 3 and the bug is gone. Will try to find out what they changed to get that fixed.
Edit: this commit fixes the problem for virtualbox-modules.
Last edited by lynix (2011-11-09 14:17:48)
Offline
I think I have a small error in the build script for the nvidia modules. I will verify later today and rebuild and upload.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Shouldn't the architecture-specific packages "provide" the generic ones? i.e. linux-ck-corex-headers -> linux-ck-headers?
Offline
@darehanl - let me think about that... what are the downsides of that?
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
OK... I believe I tracked the problem down and corrected it. I type this now booted into the corex package set and previously into the generic set. I just pushed the -6 release of nvidia-ck-xxx series for both arches. Please pacman -Syu and let me know if the problem is corrected on your systems as well. As with any nvidia upgrade, you'll need to either reboot your system or drop down to runlevel 3, rmmod nvidia then come back to runlevel 5.
Please let me know if all is well with the -6 release.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
All's well here…much grass*.
(* - "muchas gracias")
Offline
The key stuff all worked out normally/right in pacman 4
Offline
nvidia-ck-corex -6 is working as expected here , thank you!
Offline
I believe all the little bunnies are happy -- at least for now.
Thanks, graysky!
Offline
@graysky Not sure what you mean by "downsides," but if linux-ck-corex-headers "provides" linux-ck-headers, then a package that relies on kernel headers (I'm thinking of tp_smapi-ck, but nvidia-ck, too) can "makedepends" on just linux-ck-headers instead of every one of them. Ditto for linux-ck-corex providing linux-ck.
This isn't something that should need a version bump, but I just thought it might be nice to have in the future.
Offline