Announcement

Collapse
No announcement yet.

Linux 3.17 To Linux 4.16 Kernel Benchmarks On Intel Gulftown & Haswell Hardware

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

  • Linux 3.17 To Linux 4.16 Kernel Benchmarks On Intel Gulftown & Haswell Hardware

    Phoronix: Linux 3.17 To Linux 4.16 Kernel Benchmarks On Intel Gulftown & Haswell Hardware

    Our latest benchmarking of the near-final Linux 4.16 kernel is checking on the performance of two Intel systems going back to the days of Linux 3.17, the oldest kernel that would successfully boot with the Ubuntu 18.04 user-space. Every major kernel release was tested as we see how the Linux kernel performance has evolved on these Haswell and Gulftown systems since October 2014.

    http://www.phoronix.com/vr.php?view=26150

  • #2
    Typo:

    Originally posted by phoronix View Post
    And the Hackbench results don't see much change going from Linux 4.15 to 4.16 for this kernel scehduler benchmark.

    Comment


    • #3
      Linux has some serious regressions...

      Comment


      • #4
        Originally posted by Steffo View Post
        Linux has some serious regressions...
        Introduced intentionally, to mitigate Intel's screw-ups. It's not just Linux though, other OSes are in the same boat.
        Last edited by Gusar; 03-25-2018, 03:56 PM.

        Comment


        • #5
          I'd be curious to see how this works on more generations. My Westmeres are noticeably laggy and sluggish in 4.15 and Meltdown/Spectre patches. I'm forcing 4.13 for now. I can't imagine the folks on RHEL stuck at 3.10 either?

          Comment


          • #6
            Originally posted by Gusar View Post
            Introduced intentionally, to mitigate Intel's screw-ups. It's not just Linux though, other OSes are in the same boat.
            The Apache benchmark can't be explained by that...

            Comment


            • #7
              Why NO AMD CPU?

              Comment


              • #8
                Michael The vertical graph axis should always start at 0, otherwise the changes can be exaggerated to the point of being misleading.

                Comment


                • #9
                  Originally posted by boeroboy View Post
                  I'd be curious to see how this works on more generations. My Westmeres are noticeably laggy and sluggish in 4.15 and Meltdown/Spectre patches. I'm forcing 4.13 for now. I can't imagine the folks on RHEL stuck at 3.10 either?
                  You must not have the updated microcode yet (not sure if intel will even be releasing updates for Westmere tho). The updated microcode just released last week from intel exposes the hooks that allow for higher performance mitigation (aka Restricted Speculation). Your Westmere probably has speculation disabled entirely in 4.15+ due to the buggy microcode.

                  Boot 4.15 and do a 'dmesg | grep -i spectre' to know for sure. My Ivy Bridge Xeon with latest microcode and kernel 4.15.10 shows this:

                  $ dmesg | grep -i spectre
                  [ 0.014792] Spectre V2 : Mitigation: Full generic retpoline
                  [ 0.014793] Spectre V2 : Spectre v2 mitigation: Enabling Indirect Branch Prediction Barrier
                  [ 0.014793] Spectre V2 : Enabling Restricted Speculation for firmware calls

                  As for RHEL, their kernel is not vanilla, not even close to it. Red Hat back-ports massive amounts of stuff from the newer kernels to their baseline.
                  Last edited by torsionbar28; 03-25-2018, 11:51 PM.

                  Comment


                  • #10
                    Originally posted by nuetzel View Post
                    Why NO AMD CPU?
                    yeah, i think a thread ripper/ryzen 7 benchmark would have been more representative. Not sure if he is compiling all the software or getting it from who ever repos tho.

                    Comment

                    Working...
                    X