You are not logged in.
Repo-ck PKGBUILD archive - Looking for an older version of linux-ck or nvidia-ck to build yourself? Look no further.
Owdie that's cool!
Didn't find the sources for the ck-atom though; did I oversee the 'old' repo?
Would like to benchmark through real cases applications scenarios, and laptop's stuck on linux <4.1 (bug report)
You can read the current I/O scheduler from the /s[url=https://bugs.archlinux.org/task/47509ys tree (you can also change it for each device):
$ grep . /sys/block/*/queue/scheduler /sys/block/sda/queue/scheduler:noop deadline cfq [bfq] /sys/block/sr0/queue/scheduler:noop deadline cfq [bfq]
Output's more usefull than with the `cat /sys/block/[sd[a-z]/queue/scheduler` I used till now. Cool!
Seeded last month: Arch 50 gig, derivatives 1 gig
Desktop @3.3GHz 8 gig RAM, linux-ck
laptop #1 Atom 2 gig RAM, Arch linux stock i686 (6H w/ 6yrs old battery ) #2: ARM Tegra K1, 4 gig RAM, ChrOS
Atom Z520 2 gig RAM, OMV (Debian 7) kernel 3.16 bpo on SDHC | PGP Key: 0xFF0157D9
Offline
@kozaki - No, it is the AUR PKGBUILD stuff. If you want the atom flavor, you need to enable the nconfig and then select it in the CPU options. The "-atom" name is really only applicable to the repo. You can build the "linux-ck" package however you want.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Online
Thank you for a fast answer.
So,
1 get your linux-ck-<version> sources
2. Edit pkgname to 'linux-ck-atom', '_makenconfig=yes'
3. nconfig: 'intel atom: Y'
and I'll be good, correct?
Seeded last month: Arch 50 gig, derivatives 1 gig
Desktop @3.3GHz 8 gig RAM, linux-ck
laptop #1 Atom 2 gig RAM, Arch linux stock i686 (6H w/ 6yrs old battery ) #2: ARM Tegra K1, 4 gig RAM, ChrOS
Atom Z520 2 gig RAM, OMV (Debian 7) kernel 3.16 bpo on SDHC | PGP Key: 0xFF0157D9
Offline
No, there is no need to change the pkgname. Just enable the makeconfig and manually edit the config as you build the package. The nconfig is a program that is interactive. You will need to navigate to, Processor type and features>Processor family to select the atom.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Online
Hmm though of that as a way to clearly identify the kernels in the tests logs (e.g. with `uname -rms`). Also when I compile for atom why not calling it atom? Should have added "4. Edit the preset file" to reflect that, and allow for a neat identification of all the kernels in /boot. I do it already as I've got a dual arch Arch install.
nconfig: yeah I know that ha ha thanks Graysky
Seeded last month: Arch 50 gig, derivatives 1 gig
Desktop @3.3GHz 8 gig RAM, linux-ck
laptop #1 Atom 2 gig RAM, Arch linux stock i686 (6H w/ 6yrs old battery ) #2: ARM Tegra K1, 4 gig RAM, ChrOS
Atom Z520 2 gig RAM, OMV (Debian 7) kernel 3.16 bpo on SDHC | PGP Key: 0xFF0157D9
Offline
Hi All,
I am trying to build the older 4.1.6 kernel as it appears my cpu freezes are due to a GPU kernel bug introduced after that kernel ..
However, when I try to build that kernel it fails on a checksum:
==> Validating source files with sha256sums...
linux-4.1.tar.xz ... Passed
linux-4.1.tar.sign ... Skipped
patch-4.1.6.xz ... Passed
patch-4.1.6.sign ... Skipped
config.x86_64 ... Passed
config ... Passed
linux-ck.preset ... Passed
change-default-console-loglevel.patch ... Passed
patch-4.1-ck2.bz2 ... Passed
enable_additional_cpu_optimizations_for_gcc_v4.9+_kernel_v3.15+.patch.gz ... FAILED
0001-block-cgroups-kconfig-build-bits-for-BFQ-v7r8-4.1.patch ... Passed
0002-block-introduce-the-BFQ-v7r8-I-O-sched-for-4.1.patch ... Passed
0003-block-bfq-add-Early-Queue-Merge-EQM-to-BFQ-v7r8-for-4.1.0.patch ... Passed
==> ERROR: One or more files did not pass the validity check!
Anyone have any ideas? Can I get rid of that patch for now?
Offline
I think you can just updpkgsums and accept the new version.
Or you should be able to find the matching file in https://github.com/graysky2/kernel_gcc_patch
Managing AUR repos The Right Way -- aurpublish (now a standalone tool)
Offline
nvidia is giving me troubles here only with the -ck packages. -lts kernel and nvidia-lts module work fine.
pacman -Q | grep '\-ck'
linux-ck-core2 4.3.5-1
linux-ck-headers 4.3.5-1
nvidia-304xx-ck-core2 304.131-2
virtualbox-ck-host-modules-core2 5.0.14-2
dmesg -ck: https://gist.github.com/ead17c159bce7717f103
The error is like this:
[ 11.757674] nvidia: Unknown symbol mtrr_del (err 0)
[ 11.757874] nvidia: Unknown symbol mtrr_add (err 0)
Here is when it works.
dmesg -lts: https://gist.github.com/36c8729a93443968dd47
lspci | grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation G84 [GeForce 8600 GT] (rev a1)
I'm puzzled by what this error means, can you help me? Any other information you need? Should I install the stock arch kernel and nvidia module for a more relevant comparison?
UPDATE: In the meanwhile I found this thread: https://bugs.archlinux.org/task/47092
I'll just follow that, looks like the problem is not -ck relevant.
Last edited by SanskritFritz (2016-02-08 10:17:20)
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
I think you can just updpkgsums and accept the new version.
Or you should be able to find the matching file in https://github.com/graysky2/kernel_gcc_patch
Thanks, did that bit fine .. now failing on:
==> Verifying source file signatures with gpg...
linux-4.1.tar ... FAILED (unknown public key 79BE3E4300411886)
patch-4.1.6 ... FAILED (unknown public key 38DBBDC86092693E)
==> ERROR: One or more PGP signatures could not be verified!
Offline
bmentink, try the makepkg wiki page, it talks about gpg source verification.
Offline
bmentink, try the makepkg wiki page, it talks about gpg source verification.
Yeah .. well tried that, and did a search-keys on the ones returned in the error, and it came back no data ... so can't do a thing ..
Is there some other key URL that I have to plug into dirmngr.conf ... ?
Last edited by bmentink (2016-02-07 23:20:08)
Offline
==> Verifying source file signatures with gpg... linux-4.1.tar ... FAILED (unknown public key 79BE3E4300411886) patch-4.1.6 ... FAILED (unknown public key 38DBBDC86092693E) ==> ERROR: One or more PGP signatures could not be verified!
Shouldn't you have these already?-)
(10 files, 96Kb)─> [0]$ gpg --list-keys|grep -iA2 '411886\|92693E'
pub rsa2048/00411886 2011-09-20 [SC]
uid [ inconnue] Linus Torvalds <torvalds@linux-foundation.org>
..
pub rsa4096/6092693E 2011-09-23 [SC]
uid [ inconnue] Greg Kroah-Hartman (Linux kernel stable release signing key) <greg@kroah.com>
Seeded last month: Arch 50 gig, derivatives 1 gig
Desktop @3.3GHz 8 gig RAM, linux-ck
laptop #1 Atom 2 gig RAM, Arch linux stock i686 (6H w/ 6yrs old battery ) #2: ARM Tegra K1, 4 gig RAM, ChrOS
Atom Z520 2 gig RAM, OMV (Debian 7) kernel 3.16 bpo on SDHC | PGP Key: 0xFF0157D9
Offline
nvidia is giving me troubles here only with the -ck packages. -lts kernel and nvidia-lts module work fine.
UPDATE: In the meanwhile I found this thread: https://bugs.archlinux.org/task/47092
I'll just follow that, looks like the problem is not -ck relevant.
Graysky, would it be possible for you to downgrade the nvidia-304xx package in the repo to the previous working one until this problem gets fixed?
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
@SanskritF - nvidia-304xx-utils-304.131-1 is what is currently in [extra] so me downgrading to the anything but that will break updates for folks. Do any of the patches in the flyspray you referenced resolve the issue for you (manually building the module yourself)?
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Online
@SanskritF - nvidia-304xx-utils-304.131-1 is what is currently in [extra] so me downgrading to the anything but that will break updates for folks. Do any of the patches in the flyspray you referenced resolve the issue for you (manually building the module yourself)?
I will manually build it, OK.
UPDATE: What I dont understand is that nvidia-304xx-lts 304.131-2 works perfectly but nvidia-304xx-ck-core2 304.131-2 doesn't. How is that possible?
Last edited by SanskritFritz (2016-02-09 22:04:24)
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
Does any of the cpu optimized packages work with an AMD A4-5000 ?
$ gcc -c -Q -march=native --help=target | grep march
-march= btver2
In the wiki there is no package mentioned for btver2, will the bobcat (btver1) package work?
Offline
@michis - The march=btver2 is in the patch labeled as "AMD Family 16h (Jaguar)" ... give that a try.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Online
Thanks. So there is no precompiled package for btver2 in repo-ck, right?
Building linux-ck from AUR, do I need to add (e.g.) 'make menuconfig' somewhere in the PKGBUILD and change the config manually to build for the Jaguar architecture?
EDIT: Ah, ok.
_makenconfig=y
Last edited by michis (2016-02-13 21:36:31)
Offline
Compiled over night and seems to work. Thanks for your patches and the PKGBUILD.
Any plans to provide jaguar packages in repo-ck ?
Offline
today i've installed 4.3.6-1 and everything went to hell... not sure what happened but it couldn't mount my disks any longer... this has maybe something to do with the 229 systemd upgrade as well? currently running on arch stable kernel.
Last edited by eNTi (2016-02-20 20:40:36)
Offline
Tried upgrading today and I'm getting
error: failed retrieving file 'virtualbox-ck-host-modules-k10-5.0.14-3-x86_64.pkg.tar.xz' from repo-ck.com : The requested URL returned error: 404
Is it really missing or is something off on my end?
[ Arch x86_64 | linux | Framework 13 | AMD Ryzen™ 5 7640U | 32GB RAM | KDE Plasma Wayland ]
Offline
Tried upgrading today and I'm getting
error: failed retrieving file 'virtualbox-ck-host-modules-k10-5.0.14-3-x86_64.pkg.tar.xz' from repo-ck.com : The requested URL returned error: 404
Is it really missing or is something off on my end?
i got the same when uninstalling linux-ck and moving to linux-ck-piledriver.
as it looks like the guest and host modules are no longer needed when having virtualbox-host-dkms and guest-dkms installed. it still works here!
Offline
Sorry guys, I typed this a few days ago but apparently didn't submit it to the wiki: https://wiki.archlinux.org/index.php?ti … did=420396
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Online
Hi. I can't seem to enable bfq no matter what I do. When I run "grep . /sys/block/*/queue/scheduler" I get "noop deadline [cfq] bfq" every time.
I tried compiling it with that option enabled, also tried the binaries on repo-ck with elevator=bfq but none of those seem to do anything. I don't get any errors either.
dmesg gives me this:
[ 0.190269] io scheduler noop registered
[ 0.190271] io scheduler deadline registered
[ 0.190314] io scheduler cfq registered
[ 0.190350] io scheduler bfq registered (default)
[ 0.190351] BFQ I/O-scheduler: v7r8
[ 0.570823] BFS CPU scheduler v0.467 by Con Kolivas.
so is it enabled or not? i'm confused.
Last edited by lucas7_ (2016-03-17 02:39:48)
Offline
I am having issues intalling the piledriver kernel, the nvidia package failes due to not being able to find the utilites version it wants.
warning: cannot resolve "nvidia-utils=361.28", a dependency of "nvidia-ck-piledriver"
:: The following package cannot be upgraded due to unresolvable dependencies:
error: failed to prepare transaction (could not satisfy dependencies)
:: nvidia-ck-piledriver: requires nvidia-utils=361.28
If i try to just use the kernel without nvidia ck it failes to load modules and i never reach desktop. Am i doing something wrong?
Mar 22 02:49:52 kitsuna-pc kernel: vboxpci: IOMMU found
Mar 22 02:49:52 kitsuna-pc kernel: FS-Cache: Loaded
Mar 22 02:49:52 kitsuna-pc kernel: RPC: Registered named UNIX socket transport module.
Mar 22 02:49:52 kitsuna-pc kernel: RPC: Registered udp transport module.
Mar 22 02:49:52 kitsuna-pc kernel: RPC: Registered tcp transport module.
Mar 22 02:49:52 kitsuna-pc kernel: RPC: Registered tcp NFSv4.1 backchannel transport module.
Mar 22 02:49:52 kitsuna-pc systemd[1]: Started Create Static Device Nodes in /dev.
Mar 22 02:49:52 kitsuna-pc systemd[1]: Reached target Local File Systems (Pre).
Mar 22 02:49:52 kitsuna-pc systemd[1]: Mounting /tmp...
Mar 22 02:49:52 kitsuna-pc systemd[1]: Starting udev Kernel Device Manager...
Mar 22 02:49:52 kitsuna-pc systemd[1]: Mounted /tmp.
Mar 22 02:49:52 kitsuna-pc systemd[1]: Reached target Local File Systems.
Mar 22 02:49:52 kitsuna-pc systemd[1]: Starting Tell Plymouth To Write Out Runtime Data...
Mar 22 02:49:52 kitsuna-pc kernel: FS-Cache: Netfs 'nfs' registered for caching
Mar 22 02:49:52 kitsuna-pc systemd[1]: Started Setup Virtual Console.
Mar 22 02:49:52 kitsuna-pc systemd[1]: systemd-modules-load.service: Main process exited, code=exited, status=1/FAILURE
Mar 22 02:49:52 kitsuna-pc systemd[1]: Failed to start Load Kernel Modules.
Mar 22 02:49:52 kitsuna-pc systemd[1]: systemd-modules-load.service: Unit entered failed state.
Mar 22 02:49:52 kitsuna-pc systemd[1]: systemd-modules-load.service: Failed with result 'exit-code'.
Last edited by kitsuna (2016-03-22 09:53:22)
Offline