Package Details: linux-ck 6.11.7-1

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: 0.32
First Submitted: 2011-07-22 14:51 (UTC)
Last Updated: 2024-11-09 16:45 (UTC)

Dependencies (14)

Required by (6)

Sources (6)

Latest Comments

« First ‹ Previous 1 .. 99 100 101 102 103 104 105 106 107 108 109 .. 307 Next › Last »

mrkline commented on 2016-10-29 00:18 (UTC)

I'm late to the party, but I've run with NUMA disabled since I started using this package (4+ months?). Zero issues on all my machines (Skylake, Haswell, and Ivy Bridge).

graysky commented on 2016-10-28 18:08 (UTC)

Bump to v4.8.5-1 Changelog: https://www.kernel.org/pub/linux/kernel/v4.x/ChangeLog-4.8.5 Commit: https://wiki.archlinux.org/index.php/Linux-ck/Changelog

graysky commented on 2016-10-27 19:15 (UTC)

Thanks for the comments, all.

Saren commented on 2016-10-27 13:58 (UTC)

Dual haswell xeon cpu. 4.8.4-4, NUMA and MuQSS 116 enabled. 0 issue so far and great performance.

artafinde commented on 2016-10-27 07:04 (UTC) (edited on 2016-10-27 07:04 (UTC) by artafinde)

I was one of the persons who had issues with NUMA on an AMD FX 8100. About 6 months ago I changed to i7 6700 and disabled NUMA. No issues so far. I did had btrfs corruption with linux-ck early MuQSS versions (v110 I think) so I was a late adopter but now running pretty stable for days. I think the NUMA issue were related to AM3+ socket.

<deleted-account> commented on 2016-10-27 04:28 (UTC)

I'm running linux-ck with NUMA disabled since about two weeks. I haven't yet experienced any problems with my Skylake CPU.

graysky commented on 2016-10-26 21:16 (UTC) (edited on 2016-10-27 19:14 (UTC) by graysky)

@QD - It's a small increase using a make endpoint in my experience (see the flyspray I note in the PKGBUILD comments). It could be that other endpoints reveal more substantial gains as well. In any case, NUMA is really for servers with multiple sockets; it has no point on a single socket motherboard is my understanding. Linux-ck had it disabled for a long time until some combination of upstream/BFS + NUMA disabled was believed to be responsible for problems which is when I disabled the code to disable it. Now that MuQSS has replaced BFS, this may not be the case any more. Several users have posted to the AUR reporting stability with it disabled as is my experience as well.

QuartzDragon commented on 2016-10-26 20:50 (UTC)

How much of a speed increase does disabling NUMA actually give, anyway? I've never really felt the difference.

graysky commented on 2016-10-26 19:17 (UTC)

Greg just tagged 4.8.5-rc1 and has it scheduled to release on Friday/noon UTC[1]. We can give disabling NUMA a whirl in 4.8.5-1 if there are no reports linking this setting to bad behavior with MuQSS (I am not experiencing anything bad on my Haswell). 1. https://lkml.org/lkml/2016/10/26/646

metaphorex0 commented on 2016-10-26 10:45 (UTC)

I've been running linux-ck with NUMA disabled since you added that option back to PKGBUILD. I've had 0 issues so far.

« First ‹ Previous 1 .. 99 100 101 102 103 104 105 106 107 108 109 .. 307 Next › Last »