Package Details: linux-ck 6.12.1-1

Git Clone URL: https://aur.archlinux.org/linux-ck.git (read-only, click to copy)
Package Base: linux-ck
Description: The Linux kernel and modules with ck's hrtimer patches
Upstream URL: https://wiki.archlinux.org/index.php/Linux-ck
Licenses: GPL-2.0-only
Provides: KSMBD-MODULE, VIRTUALBOX-GUEST-MODULES, WIREGUARD-MODULE
Replaces: virtualbox-guest-modules-arch, wireguard-arch
Submitter: graysky
Maintainer: graysky
Last Packager: graysky
Votes: 459
Popularity: 0.23
First Submitted: 2011-07-22 14:51 (UTC)
Last Updated: 2024-11-23 13:58 (UTC)

Dependencies (14)

Required by (7)

Sources (6)

Latest Comments

« First ‹ Previous 1 .. 211 212 213 214 215 216 217 218 219 220 221 .. 308 Next › Last »

graysky commented on 2013-05-21 07:50 (UTC)

Not a stupid question. Both Sandy and Ivybridge CPUs get a more sophisticated method of managing frequency called pstates. This is true for linux-ck and linux (Arch standard). You aren't using the ondemand any more. Search for pstates in google and on the forums for more.

clayman commented on 2013-05-21 06:58 (UTC)

Hi, probably a stupid question but hey -- with 3.9.3 I've been missing the /sys/devices/system/cpu/cpufreq/ondemand/ignore_nice_load setting. I'm running BOINC and this thing keeps my CPU temps low. Where did it go? I use standard config in this regard.

mantiz commented on 2013-05-20 19:41 (UTC)

@graysky: I know that. ;-) And I already compiled the kernel with the PKGBUILD and patches provided here. But as you see in my last comment, the ck-patch takes care that the options that I think I need can not be enabled at the same time with BFS. As soon as I have some more time that I can spend on this I will contact Con Kolivas why these options are disabled and what the problems are in detail. I don't wanna harm my system or even worse my data. ;-) After that I will give this another try. But to be honest, I don't want a kernel that I have to compile on my own with each update. And since the only reason for me using the ck-kernel are large file copies because with the normal cfq-scheduler they cause many freezes on my machine I will use the "normal" kernel from the core-repo until then. Right now my priority is lxc and large file copies are not that often. ;-) But I am really thankful for your comments, helped me a lot otherwise I would still be trying to get the ck-kernel working and neglect my work that I really have to do now before anything else. :)

graysky commented on 2013-05-20 19:17 (UTC)

@mantiz - Nothing is stopping you from compiling the arch kernel from ABS and enabling whatever patches you like in this PKGBUILD. Or by that same logic, replacing the config/config.x86_64 in this pkg with the stock arch ones, and simply commenting out the line that patches the code with the CK patchset.

mantiz commented on 2013-05-20 16:11 (UTC)

@graysky: I unsuccessfully built this package on my own. Well, the build was successful and I was able to boot this kernel but the required features weren't there. After looking at the ck-patch it is obvious that this couldn't work because these settings are modified by "depends on !SCHED_BFS". So, I think you were right that these settings are conflicting with BFS. I will stick with the core-repo-kernel for now and see if I still need to use the bfq-scheduler or if I can live without it. And I will contact "him" in order to ask if there is any possibility to support these options or any workaround for this. I will post the answer here as soon as I have the time to write this mail and get an answer. Thanks for your help so far.

graysky commented on 2013-05-20 13:55 (UTC)

@mantiz - Thanks for the link. I still recommend that you ask on CK's blog. He knows the code better than I :p

mantiz commented on 2013-05-20 13:10 (UTC)

Ok, I don't think that user sessions are involved ("User namespace" is missing in the core-repo-kernel, too) but I really don't know systemd and cgroups very well, so I don't know which features are related to each other. I will try to add these options manually and see if this works. Thanks for the info, I didn't know that BFS has any problems with cgroups although it is mentioned in the wiki (my fault). You can get informations about lxc at https://wiki.archlinux.org/index.php/Linux_Containers if you are interested. It is also referred to as 'chroot on steriods'. ;-)

graysky commented on 2013-05-20 12:49 (UTC)

Not sure what an archlinux-lxc-container is but if it is related to user sessions which depend on the cgroups you mentioned, they and BFS are mutually exclusive. In other words, BFS can't have them to the best of my knowledge. You might try posting this question to CK's blog to get an official answer from him. http://ck-hack.blogspot.com/2013/05/bfs-0430-ck1-for-linux-39x.html

mantiz commented on 2013-05-20 12:45 (UTC)

I am having problems with starting an archlinux-lxc-container (with systemd) by using the ck-kernel(s). I am using linux-ck-nehalem and linux-ck-ivybridge on my systems. It turned out that the kernel from the core-repo works fine and I think the problem is that the ck-kernels are missing "CONFIG_CGROUP_CPUACCT" and "CONFIG_CGROUP_SCHED" which are enabled by the core-repo-kernel. Would it be possible to enable these options with the ck-kernels? I would really appriciate this. ;-)

« First ‹ Previous 1 .. 211 212 213 214 215 216 217 218 219 220 221 .. 308 Next › Last »