Announcement

Collapse
No announcement yet.

Sched_ext Merged For Linux 6.12 - Scheduling Policies As BPF Programs

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Sched_ext Merged For Linux 6.12 - Scheduling Policies As BPF Programs

    Phoronix: Sched_ext Merged For Linux 6.12 - Scheduling Policies As BPF Programs

    Linux 6.12 is shaping up to be a heck of a kernel update! Following real-time "PREEMPT_RT" going mainline after twenty years and many other kernel features merged this week, Linus Torvalds just pulled in the much anticipated sched_ext code!..

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Finally, really glad to hear.

    Big thanks to the whole sched-ext team as well as all contributors. Looking forward into the future!

    Comment


    • #3
      Am i dreaming?

      Comment


      • #4
        Time to name it 7.0 ,-)

        Comment


        • #5
          Now all thats left for me would be NTSYNC...

          Comment


          • #6
            With NTsync 6.12 would be one hell of a release

            Comment


            • #7
              Still no Ntsync?
              Still not up to date Zstd code?
              I wished that at least one of these two would come to this kernel!

              Comment


              • #8
                Originally posted by ihatemichael
                What is the advantage of this? Why schedulers need to use BPF?
                General purpose schedulers are almost never optimal for specific tasks, so being able to switch schedulers from userspace at runtime allows using schedulers that are optimized for certain workloads (e.g. a "gaming scheduler" that is optimized for a single foreground processes and the repetitive access patterns from game engines or a "real time pro audio scheduler" which would prioritize audio processing over other stuff (just guessing on this, I have no experience with pro audio)).

                Then there is also the benefit for development: no need to recompile the kernel and reboot when working on developing a scheduler.

                Comment


                • #9
                  Wow whats happening everything coming together in one release smaller file struct, vfs optimisations, more uring goodies, PREMEMT_RT, and sched_ext, they are pushing hard for this one .

                  The PC goes VROOOM.

                  Comment


                  • #10
                    Originally posted by ihatemichael
                    What is the advantage of this? Why schedulers need to use BPF?
                    This is a tool for hyperscalers, just like many other "pro" features in recent years. Have you ever seen any XDP/CPUMAP in the wild, used by some random admin? No? Google was glad having it...
                    Petabyte storages, faster booting with hundreds of CPUs, Wayland (closed-systems display, e.g. automotive)... see the pattern?

                    These things are often sponsored by Big Corps. And to be clear - it's good we have it for free, despite we probably won't ever use them (directly, not as a service...), as they require (re)building some background (e.g. folios, DRI) and there are the real profits for everyone.

                    Comment

                    Working...
                    X