Patches Provide For Much Faster In-Kernel Zstd Due To Embarrassingly Bad Inefficiency

https://www.phoronix.com/news/In-Kernel-Zstd-BMI2-Probing

83 points · 7 comments · view on lemmy.world

7 Comments

Linus has already chimed in, and wants Zstd to just use the standard way of checking for CPU features instead of rolling their own. Apparently, that would be even faster, compiling to a single jump instruction.

yuman@programming.dev · 8 pts · 3d (1 reply)

this concerns both btrfs and zram, so this would be a huge boost?

aBundleOfFerrets@sh.itjust.works · 2 pts · 2d

Only in a VM guest

django@discuss.tchncs.de · 6 pts · 3d

I can't wait for this fix to drop.

cyrl@lemmy.world · 5 pts · 3d

Ace, its great to see so many significant performance improvements in the kernel lately

Lemmchen@feddit.org · 2 pts · 3d (1 reply)

Is this relevant when using built-in zstd compression on ZFS?

aBundleOfFerrets@sh.itjust.works · 1 pts · 2d

Only in a VM guest