Package Details: linux-lqx 6.11.7.lqx1-2

Git Clone URL: https://aur.archlinux.org/linux-lqx.git (read-only, click to copy)
Package Base: linux-lqx
Description: The Linux Liquorix kernel and modules
Upstream URL: https://liquorix.net/
Keywords: bbr2 bfq futex pds proton zen
Licenses: GPL-2.0-only
Provides: UKSMD-BUILTIN, VHBA-MODULE, VIRTUALBOX-GUEST-MODULES, WIREGUARD-MODULE
Submitter: akurei
Maintainer: sir_lucjan (damentz)
Last Packager: damentz
Votes: 161
Popularity: 2.41
First Submitted: 2011-08-08 16:08 (UTC)
Last Updated: 2024-11-08 21:09 (UTC)

Dependencies (19)

Required by (11)

Sources (3)

Pinned Comments

damentz commented on 2020-08-31 15:22 (UTC) (edited on 2021-12-21 18:25 (UTC) by damentz)

Official binaries of linux-lqx, linux-lqx-headers, and linux-lqx-docs are now available: https://wiki.archlinux.org/index.php/Unofficial_user_repositories#liquorix

Signing key import instructions: sudo pacman-key --keyserver hkps://keyserver.ubuntu.com --recv-keys 9AE4078033F8024D && sudo pacman-key --lsign-key 9AE4078033F8024D

Latest Comments

« First ‹ Previous 1 .. 82 83 84 85 86 87 88 89 90 91 92 .. 121 Next › Last »

damentz commented on 2014-01-18 01:44 (UTC)

joko, CFS has less functional bugs than BFS. After a few years of going back and forth due to new random bugs, I decided to stick to CFS and work with it instead. One nice feature of CFS that BFS doesn't have is autogrouping, taking advantage of cgroups to share CPU bandwidth more fairly. This is most obvious when building software with multiple threads - the build tool will be grouped and will be treated as a single process, preventing a DoS on your own system. This is important for desktops where increased latency through interactivity is considered a big problem. Lastly, since BFS has been released, many of the bugs that once infested CFS have been fixed. If you read benchmarks, there's no drastic difference between the two schedulers. That leaves behavior, which I don't believe is really an issue anymore. I can be building a kernel, watching youtube, and playing an intensive AAA game at the same time and I can't tell the difference. The only difference with BFS is that you have access to sched_batch, which you can use to lower the priority of a build program that uses multiple threads. In CFS, you have auto grouping, which solves that problem nicely without any action of the user. So there you go, that's the reality of CFS vs BFS. Behavior is mostly identical and throughput is close enough to not matter.

joko commented on 2014-01-18 00:04 (UTC)

Hello, does this kernel use the BFS scheduler? I have compiled this package quickly via yaourt and it seems it doesn't: $ dmesg | grep scheduler [ 0.636725] io scheduler noop registered [ 0.636725] io scheduler deadline registered [ 0.636743] io scheduler cfq registered [ 0.636757] io scheduler bfq registered (default) [ 0.702179] CFS CPU scheduler. Any reason why it is not used?

KairiTech commented on 2014-01-12 21:27 (UTC)

Appreciate your efforts. I'll keep working at it to find an easier way to keep track of the mobo-specific kernels I create. Thanks again for your efforts.

shivik commented on 2014-01-12 20:09 (UTC)

I did a pacman -Rns linux-lqx prior to installing the rebuilt linux-lqx(with the custom name). Maybe this has something to do.I mean if you haven't done that yourself... This is all I can think of for now. Making that work for your case turns out to be beyond my knowledge. Next step is a completely modified PKGBUILD just for your needs.

KairiTech commented on 2014-01-12 18:00 (UTC)

No errors here either but I end up in emergency mode when I re-boot. I've had it work before without any changes to the kernel version but it doesn't seem to work now.

shivik commented on 2014-01-11 17:48 (UTC)

Yeah, two of my experiments must have mixed up 'cause I made a whole new PKGBUILD just for that and probably installed one on top of the other... I just recompiled and I have what you have,but I have a working system and I got no errors during kernel install - initramfs went fine,etc. I use nvidia driver and it is working. I don't use LVM, so no way to test that. The way this is supposed to work is just let you append a string to the kernel version and everything else should be normal - i.e. no need to mess with your bootloader config. Is your system working with this setup?

KairiTech commented on 2014-01-11 14:01 (UTC)

I don't seem to get extramodules-3.12-lqx-M5A99X_EVO or initramfs-linux-lqx-M5A99X_EVO-fallback.img vmlinuz-linux-lqx-M5A99X_EVO initramfs-linux-lqx-M5A99X_EVO.img ls -l /lib/modules 3.12.6-1-ARCH 3.12.6-1-lqx-M5A99X_EVO extramodules-3.12-ARCH extramodules-3.12-lqx ls /boot EFI initramfs-linux-lqx-fallback.img initramfs-linux-lqx.img vmlinuz-linux-lqx

shivik commented on 2014-01-09 10:09 (UTC)

OK, I added an option _append_kernel_custom_string to the PKGBUILD. Change that to your desire.It works on my system,no errors. Changed, rebuilt the kernel and here's my uname -r now :) [shivik@archenemy ~]$ uname -r 3.12.6-1-lqx-M5A99X_EVO My /lib/modules looks like this now: [shivik@archenemy modules]$ ls 3.12.6-1-ARCH 3.12.6-1-lqx-M5A99X_EVO extramodules-3.12-ARCH extramodules-3.12-lqx-M5A99X_EVO Keep in mind that this changes your /boot kernel and initramfs filenames,so you have to edit your bootloader config file accordingly. [shivik@archenemy boot]$ ls initramfs-linux-fallback.img initramfs-linux.img vmlinuz-linux initramfs-linux-lqx-M5A99X_EVO-fallback.img lost+found vmlinuz-linux-lqx-M5A99X_EVO initramfs-linux-lqx-M5A99X_EVO.img syslinux

« First ‹ Previous 1 .. 82 83 84 85 86 87 88 89 90 91 92 .. 121 Next › Last »