Quote of the week

[Posted September 1, 2021 by corbet]

As much as I still prefer running my own kernel on my hardware, I'm having trouble motivating myself after the last 18 months of world madness due to Covid19 and feel that I should really sadly bring this patch-set to a graceful end. My first linux kernel patches stretch back 20 years and with almost no passion for working on it any more, I feel it may be long overdue.
Unfortunately I also do not have faith that there is anyone I can reliably hand the code over to as a successor as well, as almost all forks I've seen on my work have been prone to problems I've tried hard to avoid myself.
There is always the possibility that mainline linux kernel will be so bad that I'll be forced to create a new kernel of my own out of disgust, which is how I got here in the first place, but that looks very unlikely. Many of you would have anticipated this coming after my last motivation blog-post, but unless I can find the motivation to work on it again, or something comes up that gives me a meaningful reason to work on it, I will have to sadly declare 5.12-ck the last of the MuQSS and -ck patches.

— Con Kolivas calls it quits.

A sad day

Posted Sep 2, 2021 8:08 UTC (Thu) by cagrazia (guest, #124754) [Link] (5 responses)

:( Sad thing.

Does anyone know why the changes haven't made mainline in all these years?

A sad day

Posted Sep 2, 2021 14:28 UTC (Thu) by corbet (editor, #1) [Link] (4 responses)

Con's changes haven't been upstreamed for two reasons: he didn't care much about upstreaming (especially in the last decade or so), and his work is obsessively focused on one use case to the detriment of all the others. The CPU scheduler has to work well for everybody, not just interactive desktop users.

A sad day

Posted Sep 2, 2021 20:35 UTC (Thu) by Wol (subscriber, #4433) [Link] (2 responses)

Is the cpu scheduler an optional choice? If so, why isn't it there as an option?

I can understand heavy metal not wanting to use it. But if it's obsessively focussed on the desktop, then a lot of people will want it!

Cheers,
Wol

A sad day

Posted Sep 3, 2021 6:31 UTC (Fri) by flussence (guest, #85566) [Link]

In the distant past there were some third party patches to make the CPU scheduler selectable between a bunch of out-of-tree alternatives, if not at boot time then at least in menuconfig. Needless to say mainline rejected the idea wholesale.

A sad day

Posted Sep 9, 2021 22:40 UTC (Thu) by midol (guest, #25855) [Link]

yes I agree, desktop users are an important user group. What simple means is there to set up the mailine kernel (non-CK) to provide equivalent performans, esp. responsiveness?

A sad day

Posted Sep 6, 2021 2:53 UTC (Mon) by sionescu (subscriber, #59410) [Link]

The kernel developers are deluded in thinking that there can be a scheduler that works well everywhere. The times I tried using the CK kernel, performance was remarkably better than the mainline kernel but since I no longer use Gentoo, I had to give that up.

Quote of the week

Posted Sep 2, 2021 17:27 UTC (Thu) by flussence (guest, #85566) [Link]

MuQSS has been slowly bitrotting since well before the world turned to this; it's never behaved with /proc/pressure/ enabled and trying to use it on my Ryzen just turns every idle into a prolonged hang. If there's a magic combination of settings necessary for stability, it just wasn't documented anywhere.

Mainline CFS used to be really bad up until the Wasted Cores paper came out with hard data (and confirmed I wasn't completely crazy). It's since mostly caught up to the point where the difference doesn't matter. The gradual influx of RT patches is probably helping a lot too.

I guess it's time to let it go. But for a while there, CK's code let me feel like I'd never need another upgrade after 2010.