You are not logged in.
Pages: 1
Now as the -ck patchset is gone, is there any alternative?
Use the Source, Luke!
Offline
As far as I know, the ck-patchset is obsolete throught the tickless timer in newer kernels (so also the actual one). Someone correct me, if I'm wrong!
now with 80% more sax-appeal!
"I hacked the Phrak, and all I got was this lousy signature"
Offline
I think 'obsolete' is a strong word for a patchset that applies cleanly to the latest stable kernel?
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
A new CPU scheduler is being added in 2.6.23, which is based on the Staircase CPU scheduler from ck. Swap prefetching has been merged into mainline as well, that means the two key components of ck are obsolete.
Offline
A new CPU scheduler is being added in 2.6.23, which is based on the Staircase CPU scheduler from ck. Swap prefetching has been merged into mainline as well, that means the two key components of ck are obsolete.
So in other words the next kernel release will be more responsive than ck ever been?
Use the Source, Luke!
Offline
Con's scheduler is still in debate as it's effectiveness over Igno's CFS, other than that, as of 2.6.22.1-ck1 any new major kernel version update will be better.
From tickless kernel thing, the last ck patchset applies to the 2.6.22.1 so, it also has tickless AFAIK.
From my point of view, CK scheduler has more time of development so, we can't wait for CFS come and reign the world. CK has been shapping this baby for a long time and IMHO can still outperform Igno's scheduler on releases to come.
Offline
The thing is, Con is quitting, and rumour has it CFS is in the mainline tree (2.6.23). So it does not look like there's much debate going on for the moment, except on the outside of the kernel team maybe.
Got Leenucks? :: Arch: Power in simplicity :: Get Counted! Registered Linux User #392717 :: Blog thingy
Offline
Pages: 1