You are not logged in.
Since you are hitting that problem with official packages (qt4 libreoffice) I don't think it's a repo-ck thing. Best to clear and re-d/l. Enjoy.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
This question might have been answered already, but will the future builds also be able to load microcode early as per the news? (Or, if I am really ignorant: Does it already do this?)
Last edited by runical (2014-10-23 20:29:39)
Offline
runical, it's already done in 3.16.6-3, the databases just haven't been updated yet.
Offline
This question might have been answered already, but will the future builds also be able to load microcode early as per the news?
Already implemented in 3.16.6-3, see the changelog: https://wiki.archlinux.org/index.php/Linux-ck/Changelog... live in the repo a few moments ago.
Last edited by graysky (2014-10-23 20:30:13)
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
Ah, I didn't check the changelog. Thanks!
Offline
@graysky
I frequently receive timeouts when downloading packages from repo-ck. A server issue?
error: failed retrieving file 'linux-ck-core2-3.16.6-3-x86_64.pkg.tar.xz' from repo-ck.com : transfer closed with 44411165 bytes remaining to read------------------------] 22%
warning: failed to retrieve some files
linux-ck-core2-headers-3.16.6-3-x86_64 6.0 MiB 687K/s 00:09 [###############################################################] 100%
error: failed to commit transaction (unexpected error)
Errors occurred, no packages were upgraded.
Offline
@orschiro, yes, been discussed more than once in this thread.
Offline
There's a work-around listed on the repo-ck page.
Sakura:-
Mobo: MSI MAG X570S TORPEDO MAX // Processor: AMD Ryzen 9 5950X @4.9GHz // GFX: AMD Radeon RX 5700 XT // RAM: 32GB (4x 8GB) Corsair DDR4 (@ 3000MHz) // Storage: 1x 3TB HDD, 6x 1TB SSD, 2x 120GB SSD, 1x 275GB M2 SSD
Making lemonade from lemons since 2015.
Online
can anyone confirm or deny that hasswell optimized linux-ck works on haswell generation Celerons (like Celeron 2955U found in many Chromebooks) since they don't have AVX/AVX2 extensions?
A couple months ago I tried on a Chromebox ( same cpu ) but it failed to boot ( blank screen ) Since that I have not tried, please report back how it was going for you.
Offline
runical wrote:This question might have been answered already, but will the future builds also be able to load microcode early as per the news?
Already implemented in 3.16.6-3, see the changelog: https://wiki.archlinux.org/index.php/Linux-ck/Changelog... live in the repo a few moments ago.
In another thread related to the microcode change it was suggested that the "/boot/intel-ucode.img" should be added to the initrd only after the kernel upgrade has occured. Is this the case with linux-ck?
Offline
graysky wrote:runical wrote:This question might have been answered already, but will the future builds also be able to load microcode early as per the news?
Already implemented in 3.16.6-3, see the changelog: https://wiki.archlinux.org/index.php/Linux-ck/Changelog... live in the repo a few moments ago.
In another thread related to the microcode change it was suggested that the "/boot/intel-ucode.img" should be added to the initrd only after the kernel upgrade has occured. Is this the case with linux-ck?
Yes. And it should be same with all the kernels that have implemented the config changes.
Claire is fine.
Problems? I have dysgraphia, so clear and concise please.
My public GPG key for package signing
My x86_64 package repository
Offline
@cl. Thank you. My reading of the wiki didn't explicitly warn about doing the initrd addition only after the kernel upgrade. I will upgrade this evening but before doing so I will remove the initrd ucode.img. I guess it's just a matter of waiting untill the kernels get upgraded.
see post 22 of https://bbs.archlinux.org/viewtopic.php?id=188252
Last edited by czubek (2014-10-24 19:51:52)
Offline
If you omit the manually added line referencing intel-ucode.img nothing bad should happen. No need to uninstall the package, it just won't get loaded. I don't think the converse is true...
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
I haven't used the repo in a while and apart from losing connection (happened before, I can deal with it), I now get ~15kbps download speed. It doesn't have to be 1000 times faster, but if it's an ongoing and widespread issue (or is it just me?), maybe using a torrent would be faster?
Questions:
1. Are there enough users to seed?
2. graysky would have yet another thing to do - or can somebody help him with this?
Offline
Torrent distribution is not practical. The godaddy host shouldn't be that slow. For me, it is nice and fast, maybe try again in a few min?
% wget http://repo-ck.com/x86_64/linux-ck-ivybridge-3.16.6-3-x86_64.pkg.tar.xz
...
2014-10-25 12:55:33 (4.86 MB/s) - ‘linux-ck-ivybridge-3.16.6-3-x86_64.pkg.tar.xz’ saved [57049464/57049464]
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
I've been trying for a while and I did finally download it. My internet is nice and fast - apart from the current linux-ck issue.
I'll keep an eye on it. If it's just a once-in-a-blue-moon hiccup, it's not worth trying to work around it.
Offline
Torrent distribution is not practical.
Why do you think so? I would certainly seed.
My problem is mostly that the download gets interrupted many times and I have to start pacman -Syu again. Luckily the download continues from where it was interrupted.
Last edited by SanskritFritz (2014-10-25 17:41:45)
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
graysky wrote:Torrent distribution is not practical.
Why do you think so?
Correct me if I'm wrong but pacman cannot handle distribution via torrents, no?
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
graysky wrote:Torrent distribution is not practical.
Why do you think so? I would certainly seed.
According to stats.pdf (it's linked in the first post) there should be enough users to distribute at least the more popular packages.
My problem is mostly that the download gets interrupted many times and I have to start pacman -Syu again. Luckily the download continues from where it was interrupted.
Are you using the tricks mentioned in https://wiki.archlinux.org/index.php/Re … _regularly ?
I don't need to have it integrated in pacman, I download the file via a torrent and use 'pacman -U'.
Offline
@karol - There are plenty of users, but the logistics of making a .torrent file for each package+sig is too much. Please many folks use repo-ck as a regular repo pulled via pacman. As you pointed out, torrent distribution would be a manual thing.
CPU-optimized Linux-ck packages @ Repo-ck • AUR packages • Zsh and other configs
Offline
SanskritFritz wrote:My problem is mostly that the download gets interrupted many times and I have to start pacman -Syu again. Luckily the download continues from where it was interrupted.
Are you using the tricks mentioned in https://wiki.archlinux.org/index.php/Re … _regularly ?
Thanks, that part skipped my attention.
@graysky I must agree, torrent is not integrated into pacman, this way it would be impractical indeed.
zʇıɹɟʇıɹʞsuɐs AUR || Cycling in Budapest with a helmet camera || Revised log levels proposal: "FYI" "WTF" and "OMG" (John Barnette)
Offline
Linux-ck wiki states that, when compiling from AUR, to enable BFQ I/O scheduler both
_BFQ_enable_="y"
is needed in the pkgbuild and,
Add elevator=bfq to the kernel parameters.
is needed in the boot loader.
I use the pre-compiled linux-ck-atom from graysky's Repo-ck repository. When a new version arrives it automatically installs. I do not make any alterations to the pkgbuild.
I looked at graysky's pkgbuild at https://aur.archlinux.org/packages/li/linux-ck/PKGBUILD and the _BFQ_enable_= flag is blank.
Does this mean that I must intercept the package and revise the BFQ flag manually?
Am I reading the wrong pkgbuild and graysky's pkgbuild already has the BFQ flag enabled?
Offline
Linux-ck wiki states that, when compiling from AUR, to enable BFQ I/O scheduler both
_BFQ_enable_="y"
is needed in the pkgbuild and,
Add elevator=bfq to the kernel parameters.
is needed in the boot loader.
I use the pre-compiled linux-ck-atom from graysky's Repo-ck repository. When a new version arrives it automatically installs. I do not make any alterations to the pkgbuild.
I looked at graysky's pkgbuild at https://aur.archlinux.org/packages/li/linux-ck/PKGBUILD and the _BFQ_enable_= flag is blank.
Does this mean that I must intercept the package and revise the BFQ flag manually?
Am I reading the wrong pkgbuild and graysky's pkgbuild already has the BFQ flag enabled?
The linux-ck packages (including the ones in graysky's repository and the one's I maintain) all come with BFQ compiled in the kernel so that you can use it if you wish.
In graysky's packages, the _BFQ_enable_= flag is blank so you use CFQ (the standard kernel queuing mechanism) by default. Adding elevator=bfq to the kernel parameters tells the OS that you want to use BFQ instead.
If you compile the kernel yourself with _BFQ_enable_="y", then you will be using BFQ all the time with that compiled kernel.
Claire is fine.
Problems? I have dysgraphia, so clear and concise please.
My public GPG key for package signing
My x86_64 package repository
Offline
@ cl. Thank you for that great and thorough explanation. I append the elevator statement in my bootloader but wasn't sure it initiated BFQ. Thank you again.
Offline