Linus Torvalds Lands A 2.6% Performance Improvement With Minor Linux Kernel Patch

[phoronix]

Phoronix: Linus Torvalds Lands A 2.6% Performance Improvement With Minor Linux Kernel Patch

Linus Torvalds merged a patch on Wednesday that he authored that with reworking a few lines of code is able to score a 2.6% improvement within Intel's well-exercise "will it scale" per-thread-ops benchmark test case...

https://www.phoronix.com/news/Linus-2.6p-Faster-Scale-Patch

[Danny3]

So nice that it has been done by Linus himself and it will be part of the 6.12 release!
That alleviates a bit the sadness of not having NTSync and up to date Zstd code too!

Good job Linus and many thanks for everything you have done from us!

[Volta]

Good job Linus and many thanks for everything you have done from us!

Yeah, 'thanks' for being corporate bitch with 12 years old mentality.

[JPFSanders]

Yeah, 'thanks' for being corporate bitch with 12 years old mentality.

I'm thankful that in this universe being a corporate bitch with 12 years old mentality is not mutually exclusive with such corporate bitch improving the kernel's performance and its users being grateful for it.

[Britoid]

Don't worry everyone, the next Electron update will use 2.6% more CPU

Didn't think you'd be allowed to have this did you?

[NotMine999]

Don't worry everyone, the next Electron update will use 2.6% more CPU

Didn't think you'd be allowed to have this did you?

The only reasonable post I read in this entire thread.

[damentz]

Yeah, 'thanks' for being corporate bitch with 12 years old mentality.

Was looking for someone that would spin this 2.6% optimization as a negative, not disappointed.

[rmfx]

If Linux can gain 2.6 percent with 10 lines, I wonder how many few lines tweaks = 2 percent optimization are hiding in the kernel leading to a potential x2 performance a few hundred LoC away… :P

[coder]

If Linux can gain 2.6 percent with 10 lines, I wonder how many few lines tweaks = 2 percent optimization are hiding in the kernel leading to a potential x2 performance a few hundred LoC away… :P

There's a reason this is news: because such wins are rare, and especially when they involve such a minor code change.

Lots of people spend a lot of time analyzing kernel performance. For general-purpose stuff, I think there aren't huge gains lurking anywhere. You have to get into some of the corner cases, before you might find potential for big improvements.

[kcrudup]

If i'm reading this right, this doesn't really help you if you run with "mitigations=off" right?