You are not logged in.
error: failed to prepare transaction (could not satisfy dependencies)
:: Starting full system upgrade...
:: virtualbox-ck-host-modules-core2: requires linux-ck-core2<3.12
I waited a couple of hours figuring that maybe you were still uploading, but the problem persists.
Edit, telling it specifically to install the modules does want to upgrade the kernel, so I'm guessing you forgot to bump the pkgrel? Still can't upgrade, though, getting (invalid or corrupted package (PGP signature))
Edit2, uninstalled the vbox modules and upgraded the kernel just fine, so I don't think the PGP problem is in my local keyring.
Last edited by Scimmia (2013-11-19 18:26:18)
Offline
error: failed to prepare transaction (could not satisfy dependencies) :: Starting full system upgrade... :: virtualbox-ck-host-modules-core2: requires linux-ck-core2<3.12
I waited a couple of hours figuring that maybe you were still uploading, but the problem persists.
Having this problem for i686 in atom mean that probably the problem is general with more archs and both arches
maybe a change in the PKGBUILDes?
Well, I suppose that this is somekind of signature, no?
Offline
Jristz, try removing /var/cache/pacman/pkg/virtualbox-ck-host-modules-*, then tell pacman to reinstall them. See what you get.
Offline
@Scimmia and Jristz - The virtualbox-ck-modules problem was my fault... I didn't bump the pkgver. Fixed now and packages are online. Please update and report back.
@botika - Not sure what's wrong, I can build packages from the AUR just fine.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Installed just fine this time, no PGP error.
Offline
Could the nvidia-ck package dependencies possibly be changed to only require a libgl provider (such as mesa-libgl) rather than the nvidia-libgl specifically?
I'm not sure if this is viable but considering that bumblebee can 'provide' nvidia-libgl while the system uses mesa-libgl , seems to imply that the libgl package can be either.
Last edited by tiberiousr (2013-11-20 10:56:36)
Offline
My strategy for the x-ck packages is to mirror as closely as possible the official Arch packages for consistency and because the devs have more insight into changes affecting other packages than I do. You might wanna open a bug report against the official package if you feel it is important. If they make the change, I will follow suit.
Last edited by graysky (2013-11-20 11:29:21)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Fair enough, I will look into it. Thanks for taking the time to reply.
Offline
graysky, could you make an option when all settings are off and all modules are disabled? the reason is that i'm tired to press "space" to disable settings and modules.
Offline
@pefectg - not sure I understand what you're asking for here.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
graysky, I mean .config with all modules are "is not set", especially specidic modules commonly not used in PC
Last edited by Perfect Gentleman (2013-11-20 14:35:43)
Offline
graysky, could you make an option when all settings are off and all modules are disabled? the reason is that i'm tired to press "space" to disable settings and modules.
I don't see how this is at all relevant to the -ck repo. If you want a different .config, make your own. There is no need to press the space bar repeatedly, just us a sed onliner to change every setting.
"UNIX is simple and coherent" - Dennis Ritchie; "GNU's Not Unix" - Richard Stallman
Offline
I don't see how this is at all relevant to the -ck repo. If you want a different .config, make your own. There is no need to press the space bar repeatedly, just us a sed onliner to change every setting.
Honestly saying, I'm not good in bash and sed, and I don't understand how to change tons of settings with sed.
... and it's not relevant to -ck, so I'm asking, saying it would be great.
I'm using aur-version, or it's not discussing here?
Last edited by Perfect Gentleman (2013-11-20 15:36:07)
Offline
I'm having a problem upgrading linux-ck-core2. It seems the nvidia-304xx-ck-core2 package was modified w/o updating the version:
$pacman -Qi nvidia-304xx-ck-core2
Repository : repo-ck
Name : nvidia-304xx-ck-core2
Version : 304.108-7
...
Depends On : linux-ck-core2>=3.11 linux-ck-core2<3.12 nvidia-304xx-utils=304.108
...
Build Date : Tue 05 Nov 2013 01:48:59 PM MST
$pacman -Si nvidia-304xx-ck-core2
Repository : repo-ck
Name : nvidia-304xx-ck-core2
Version : 304.108-7
...
Depends On : linux-ck-core2>=3.12 linux-ck-core2<3.13 nvidia-304xx-utils=304.108
...
Build Date : Tue 19 Nov 2013 03:46:45 AM MST
Trying to upgrade/reinstall nvidia-304xx-ck-core2 fails during integrity checking stating it's an invalid or corrupted package (PGP signature) .
TIA!
Offline
@perfectg - Please use this thread for repo related issues. I don't know exactly what I could do to the PKGBUILD to make what you're asking for possible
@schmug - Is your sync-db up to date? Try:
pacman -Syy
pacman -Si nvidia-304xx-ck-core2
For me on i686:
% pacman -Si nvidia-304xx-ck-core2
Repository : repo-ck
Name : nvidia-304xx-ck-core2
Version : 304.108-7
...
Depends On : linux-ck-core2>=3.12 linux-ck-core2<3.13 nvidia-304xx-utils=304.108
...
Build Date : Tue 19 Nov 2013 05:47:17 AM EST
And on x86_64:
% pacman -Si nvidia-304xx-ck-core2
Repository : repo-ck
Name : nvidia-304xx-ck-core2
Version : 304.108-7
...
Depends On : linux-ck-core2>=3.12 linux-ck-core2<3.13 nvidia-304xx-utils=304.108
...
Build Date : Tue 19 Nov 2013 05:46:45 AM EST
Last edited by graysky (2013-11-20 19:49:16)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
thx for the quick reply, and repo-ck
Yes, sync-db is up-to-date, though I just updated again to be sure.
I get the same results as you for pacman -Si nvidia-304xx-ck-core2 (x86_64 here), but shouldn't it have a different version? My installed package is the same version 304.108-7 but with a different build date and deps.
If I try to reinstall it fails on a corrupt package, even if I let pacman delete the "corrupted" file and it's downloaded again...hmmm...I even tried manually downloading from a host outside of my current network and I get the same results...well I get identical files, I was trying to rule out corporate proxy server monkey-business.
Offline
@schmug - I am very puzzled by your post. Per my notes, I did not duplicate this package by mistake. What if you
1) comment out repo-ck from pacman.conf
2) pacman -Syy
3) pacman -Sc
(answer N to the first and Y to the 2nd)
4) uncomment repo-ck from pacman.conf
5) pacman -Syy
Then attempt to install again.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Sorry, work keeps getting in the way today...
Still no love after purging repo-ck then re-adding and updating, same corrupt file problem on install. Very puzzling indeed.
I tried turning off integrity checking for repo-ck, but not unexpectedly ended up with a checksum error instead...I must have only skimmed the SigLevel description in the man page.
I am still puzzled though how I have version 304.108-7 installed from repo-ck with a build date of Tue 05 Nov 2013 01:48:59 PM MST, yet repo-ck (just using my browser) has nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz with a date of 2013-11-19 03:46.
This could certainly be a local PC problem/operator-error, but this is the first upgrade prob I've had in the many months since switching to the ck kernel. And haven't had any probs with a different install running linux-ck-atom.
thx again for the quick replies
Offline
Hmm... what if you:
wget http://repo-ck.com/x86_64/nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz && md5sum nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
For me:
6e5a73debd48b466e187ad19675d1b4e nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
If your md5sum matches then
pacman -U nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
We're thinking along the same lines...
I've downloaded on 3 different machines on 3 different networks (home, work, hosting provider) and all produce the same checksum:
$ md5sum nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz*
6e5a73debd48b466e187ad19675d1b4e nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
ba0f65fcebfb291210f3f58e78fcd895 nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz.sig
But, on all those same machines:
$ xz --test nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
xz: nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz: Compressed data is corrupt
and on the core2 machine I'm trying to upgrade:
$ sudo pacman -U nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
loading packages...
error: 'nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz': invalid or corrupted package (PGP signature)
Offline
@schmug - That is interesting.... I too get the same compressed data is corrupt error on the downloaded file; the local copy is fine.
Downloaded file:
md5sum nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
6e5a73debd48b466e187ad19675d1b4e nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
Local copy:
md5sum /mnt/data/repo/x86_64/nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
d2ab9fcf6ef80ae320588a3db2e258ba /mnt/data/repo/x86_64/nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
The problem must be in the upload because the md5sum on the repo-ck.com server matches that of the downloaded one... WTF?!
EDIT: I am reuploading from the local copies via scp rather than via rsync and the md5sums on the server are matching. Please try again (i.e. pacman -S nvidia-304xx-core2).
Last edited by graysky (2013-11-20 23:06:34)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Thanks for that.
Hmmm, still getting corruption complaints from pacman. I did get a newly downloaded package file with a different md5 chksum, but it doesn't match the md5 of your local copy:
md5sum /var/cache/pacman/pkg/nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
0f39358624ac48f401a2a07d4463c4cd /var/cache/pacman/pkg/nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
On my other hosts (home, hosting) I'm still getting the 6e5a73debd48b466e187ad19675d1b4e version when using wget directly even with --no-cache...there's another WTF?!?!
Offline
Something is wrong with the godaddy host. The md5sum on the server (ssh connection) are different from the md5sum on the downloaded package.
On the server:
$ md5sum nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
d2ab9fcf6ef80ae320588a3db2e258ba nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
On my workstation after downloading:
% md5sum nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
6e5a73debd48b466e187ad19675d1b4e nvidia-304xx-ck-core2-304.108-7-x86_64.pkg.tar.xz
Last edited by graysky (2013-11-21 00:56:31)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
OK... I just build 3.12.1-1 and bumped nvidia-304xx to 304.108-8. Please try it. All packages on the repo have the same md5sums as my local versions... if there is something funky going on, blame godaddy.
Last edited by graysky (2013-11-21 01:56:21)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
graysky,
So I remember that recently you had switched a few things to btrfs to try it out. I'm not sure if you are still using it or not, but there is a pretty glaring bug with 'btrfs balance' at the moment. There has been a patch in btrfs-next for quite some time, but it had been forgotten to be added to 3.12. So the 3.13 btrfs pull is taking a bit longer than expected, and until that happens, this patch cannot be backported to the stable releases either. So you might want to think about including this patch until that happens. As far as I know, every other tier 1 distribution has been including this patch with their kernels since the revelation of this bug.
At the moment, a balance can cause csum craziness and file corruption with files that have preallocated extents. So this includes things like the /var/log/journal/* as well as files downloaded via a number of torrent clients.
Heres that link again: http://www.spinics.net/lists/linux-btrfs/msg27836.html
Hopefully, this will make it in soon, but in the mean time, I have been compiling my own kernels with this included and it works great and as expected.
Offline