Package Base Details: linux-amd-znver3

Git Clone URL: https://aur.archlinux.org/linux-amd-znver3.git (read-only, click to copy)
Submitter: None
Maintainer: bebna
Last Packager: bebna
Votes: 11
Popularity: 1.72
First Submitted: 2023-05-04 15:47 (UTC)
Last Updated: 2024-05-03 10:20 (UTC)

Pinned Comments

<deleted-account> commented on 2023-05-07 09:41 (UTC)

Tired of compiling? Use this binary repo instead! Add this at the end of /etc/pacman.conf :

[linuxkernels]
Server = http://nhameh.ovh/$repo/$arch
SigLevel = Optional TrustAll

Latest Comments

« First ‹ Previous 1 2 3 4 5 6 7 8 9 10 Next › Last »

<deleted-account> commented on 2024-01-11 09:59 (UTC)

ok, done, ty for telling!

yamai commented on 2024-01-11 00:25 (UTC)

Would it be possible to add the module drivetemp?

<deleted-account> commented on 2024-01-02 06:54 (UTC)

I mean, yeah, you're right, it isn't a big diffrence. Infact, most things I do in these kernels dont really make that big of a diffrence (mtune, disable undeeded mods,...). It depends where you draw the line.

But in my timed tests lzo4 was actually faster in zstd, maybe it was a fluke, because I also saw tests that showed bigger performance with zstd. It's just not what I saw on my system.

onurmercury commented on 2024-01-01 21:55 (UTC)

well, i don't really expect to see big differences between decompression times of initrds (except xz) for any znver3 system anyway.

and i know this test doesn't really matter (because conditions aren't same with kernel boot factors) but zstd decompresses faster than others in userspace.

i want to test this someday with kernel boot-time decompression & non-kvm qemu system to see real results.

<deleted-account> commented on 2024-01-01 20:03 (UTC)

Im sorry but im really not impressed by zstd reading performance. I know you would rather keep everything compliant and generic, but at that point, I would really just advice you to use the vanilla arch kernel instead of this one ;-)

onurmercury commented on 2024-01-01 19:50 (UTC) (edited on 2024-01-01 19:52 (UTC) by onurmercury)

ofc it's all up to you and not so big deal but i think forcing lzop via .preset is not ideal. it just takes away the global compression control of mkinicpio.conf.

Mario156090 commented on 2024-01-01 19:40 (UTC)

Aparently only were a installation problem because now works perfectly.

[mario@asus-arch ~]$ uname -r 6.6.9-AMD-znver3 [mario@asus-arch ~]$ ^C [mario@asus-arch ~]$

<deleted-account> commented on 2024-01-01 19:37 (UTC)

@mario good to hear because I was all out of ideas..

Mario156090 commented on 2024-01-01 19:36 (UTC)

Well, after reinstall it's booted correctly.

<deleted-account> commented on 2024-01-01 19:19 (UTC)

I will keep the preset file since it forces lzop compression, which is sometimes faster than zstd, especially in reading ==> (just like I do in the internal kernel)

CONFIG_KERNEL_LZO=y

these kernels were using lzop far before the arch kernels switched to zstd, and I havent been convinced by their performance after some testing to do some switching myself.

The other two are indeed interesting (although the move of the systemmap to headers seems a bit trivial to me), ill have a look at it soon.

« First ‹ Previous 1 2 3 4 5 6 7 8 9 10 Next › Last »