A Nasty Performance Regression For Some Intel Systems Wound Up In Linux 6.5 Stable

[phoronix]

Phoronix: A Nasty Performance Regression For Some Intel Systems Wound Up In Linux 6.5 Stable

While catching up on my Linux kernel benchmarking across different systems the Intel Core i9 13900K "Raptor Lake" desktop performance has been particularly volatile. Linux 6.5 is running slower for the Core i9 13900K compared to Linux 6.4 for a number of different workloads. Buckle up for a wild ride but the good news is a stable fix is forthcoming.

https://www.phoronix.com/review/linux65-intel-regression

[JEBjames]

Michael

grammar

page 2

"But as more workloads ran, there became other software test cases exhibited with even wider performance regressions on Linux 6.5 stable."

Should probably be something like:

"But as more workloads ran, other software test cases exhibited even wider performance regressions on Linux 6.5 stable.

[chuckula]

My 5950x is also showing a bit of a performance regression since it refuses to boot with a 6.5.3 kernel (it works fine with 6.4.16).

[FelixB]

Food for though... I'm not seeing this regression at all on my 13700KF. Didn't ran the test on 6.4.x since I saw that I had over 8,000,000. It might be compiler related. Both were compiled with GCC 13.2 with -march=raptorlake.
Please see result below.

6.6rc1

https://openbenchmarking.org/result/2309151-NE-WDAWDA65198

6.5.1

https://openbenchmarking.org/result/2309156-NE-REDIS651335

[creative]

Been fairly happy with 5.15.x I can't really tell a difference between it and 6.1.x. Though I am not on intel, I do dislike seeing intel users having to deal with this stuff. Just thought I'd throw that in there.

[GoofyX]

I'm still on 6.4.16 and I saw this article about the performance regresion in the 6.5.x series for Intel users that kept me from upgrading the kernel. I have a 13600K cpu, has this issue been solved in the latest revisions?