Announcement

Collapse
No announcement yet.

Benchmarking The Linux 2.6.24 Through 2.6.29 Kernels

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

  • #11
    To verify I ran a 32 bit 2.6.28 kernel and only got 35.50 with OpenSSL. So I guess your testsystem was 32 bit before and you tested a 64 bit system against it. Can you provide the PTS link to verify?

    Comment


    • #12
      Originally posted by Kano View Post
      To verify I ran a 32 bit 2.6.28 kernel and only got 35.50 with OpenSSL. So I guess your testsystem was 32 bit before and you tested a 64 bit system against it. Can you provide the PTS link to verify?
      No, the Ubuntu 8.10 installation was x86_64 and the kernels were x86_64.
      Michael Larabel
      http://www.michaellarabel.com/

      Comment


      • #13
        Where are the results with system specs?

        Comment


        • #14
          Hmm. 2/3 here. Loss of 7z speed is nay, but sqlite and imagegick are yay.

          Comment


          • #15
            Kernel configuration

            In past kernel benchmarking articles we have built our vanilla kernels from source, but now that Canonical provides a mainline kernel PPA of each kernel release built in a standard configuration without any extra Ubuntu patches and configuration options, we opted to use this Ubuntu package repository.
            Was the kernel configuration identical between the kernel releases? Ubuntu kernel packages were used, but were the config files compared? This could easily explain some of the highs and lows without even kernel code changes. The quote above would imply they are the same, but I don't think that is good enough.

            Comment


            • #16
              Originally posted by edgan View Post
              Was the kernel configuration identical between the kernel releases?
              Thank you, was just about to ask about this. If you're testing kernels, you really should be building them from source and documenting the configuration changes between them.

              Comment


              • #17
                Configuration differences

                I have confirmed there are configuration differences between Ubuntu mainline ppa kernels. I compared x86_64 versions of 2.6.28 and 2.6.29 with the first command below. I then reduced it down to a more basic form with the second command.

                Some of the new options enabled seem to be debugging options.

                diff -uN /boot/config-2.6.28-020628-generic /boot/config=2.6.29-020629-generic > config.diff

                grep -v "^ " config.diff | grep -v @@ | grep -v '\-\-\-' | grep -v '+++' | grep -v '\#'

                -CONFIG_ACPI_SYSTEM=y
                +CONFIG_ACPI_TOSHIBA=m
                +CONFIG_ARCH_WANT_FRAME_POINTERS=y
                -CONFIG_AT24=m
                -CONFIG_BACKLIGHT_CORGI=m
                +CONFIG_BACKLIGHT_GENERIC=y
                +CONFIG_CHELSIO_T3_DEPENDS=y
                +CONFIG_COMPAT_NET_DEV_OPS=y
                +CONFIG_DIRECT_GBPAGES=y
                +CONFIG_DVB_STB0899=m
                +CONFIG_DVB_STB6100=m
                +CONFIG_DVB_TDA8261=m
                +CONFIG_DVB_TUNER_CX24113=m
                -CONFIG_EEPRO100=m
                -CONFIG_FUSE_FS=m
                +CONFIG_FUSE_FS=y
                +CONFIG_GENERIC_BUG_RELATIVE_POINTERS=y
                +CONFIG_GENERIC_FIND_LAST_BIT=y
                +CONFIG_HAVE_FUNCTION_GRAPH_TRACER=y
                +CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST=y
                +CONFIG_HAVE_HW_BRANCH_TRACER=y
                +CONFIG_HAVE_MMIOTRACE_SUPPORT=y
                +CONFIG_HERMES_CACHE_FW_ON_INIT=y
                -CONFIG_HID_BRIGHT=m
                -CONFIG_HID_DELL=m
                +CONFIG_HID_NTRIG=m
                +CONFIG_HID_TOPSEED=m
                -CONFIG_IEEE80211=m
                -CONFIG_IEEE80211_CRYPT_CCMP=m
                -CONFIG_IEEE80211_CRYPT_TKIP=m
                -CONFIG_IEEE80211_CRYPT_WEP=m
                +CONFIG_IOMMU_API=y
                -CONFIG_IP_DCCP_ACKVEC=y
                -CONFIG_IP_DCCP_CCID2=m
                -CONFIG_IP_DCCP_CCID3=m
                +CONFIG_IP_DCCP_CCID3=y
                -CONFIG_IP_DCCP_TFRC_LIB=m
                +CONFIG_IP_DCCP_TFRC_LIB=y
                -CONFIG_KMOD=y
                -CONFIG_LEDS_HP_DISK=m
                +CONFIG_LIB80211=m
                +CONFIG_LIB80211_CRYPT_CCMP=m
                +CONFIG_LIB80211_CRYPT_TKIP=m
                +CONFIG_LIB80211_CRYPT_WEP=m
                +CONFIG_LIBIPW=m
                -CONFIG_MAC80211_RC_DEFAULT="pid"
                -CONFIG_MAC80211_RC_DEFAULT_PID=y
                +CONFIG_MAC80211_RC_DEFAULT="minstrel"
                +CONFIG_MAC80211_RC_DEFAULT_MINSTREL=y
                -CONFIG_MAC80211_RC_PID=y
                +CONFIG_MFD_CORE=m
                +CONFIG_MISC_FILESYSTEMS=y
                -CONFIG_MTD_PHYSMAP_BANKWIDTH=2
                -CONFIG_MTD_PHYSMAP_LEN=0x4000000
                -CONFIG_MTD_PHYSMAP_START=0x8000000
                +CONFIG_QUOTA_TREE=m
                -CONFIG_RESOURCES_64BIT=y
                +CONFIG_SCHED_OMIT_FRAME_POINTER=y
                -CONFIG_SECURITY_DEFAULT_MMAP_MIN_ADDR=0
                +CONFIG_SECURITY_DEFAULT_MMAP_MIN_ADDR=65536
                -CONFIG_SENSORS_EEPROM=m
                +CONFIG_SFC_MTD=y
                +CONFIG_SND_HDA_CODEC_INTELHDMI=y
                +CONFIG_SND_HDA_ELD=y
                +CONFIG_SND_JACK=y
                +CONFIG_SND_SOC_I2C_AND_SPI=m
                +CONFIG_SND_SOC_L3=m
                +CONFIG_SND_SOC_PCM3008=m
                +CONFIG_SND_SOC_UDA134X=m
                +CONFIG_SND_SOC_WM8350=m
                +CONFIG_SND_SOC_WM8728=m
                -CONFIG_SPI_AT25=m
                +CONFIG_SQUASHFS=m
                +CONFIG_SQUASHFS_FRAGMENT_CACHE_SIZE=3
                +CONFIG_TOUCHSCREEN_DA9034=y
                -CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug"
                +CONFIG_UEVENT_HELPER_PATH=""
                +CONFIG_USB_DUMMY_HCD=m
                -CONFIG_USB_EHCI_HCD=m
                +CONFIG_USB_EHCI_HCD=y
                +CONFIG_USB_GADGET_DUMMY_HCD=y
                -CONFIG_USB_GADGET_NET2280=y
                -CONFIG_USB_KBD=m
                -CONFIG_USB_MOUSE=m
                -CONFIG_USB_NET2280=m
                -CONFIG_USB_OHCI_HCD=m
                +CONFIG_USB_OHCI_HCD=y
                -CONFIG_USB_STORAGE_DPCM=y
                -CONFIG_USB_UHCI_HCD=m
                +CONFIG_USB_UHCI_HCD=y
                +CONFIG_USER_STACKTRACE_SUPPORT=y
                -CONFIG_VIDEOBUF_DMA_CONTIG=m
                +CONFIG_VIDEO_BT866=m
                +CONFIG_VIDEO_CX88_MPEG=m
                -CONFIG_VIDEO_SH_MOBILE_CEU=m
                -CONFIG_WIRELESS_OLD_REGULATORY=y
                -CONFIG_X86_ACPI_CPUFREQ=m
                +CONFIG_X86_ACPI_CPUFREQ=y
                +CONFIG_X86_DS=y
                +CONFIG_X86_PLATFORM_DEVICES=y
                -CONFIG_X86_POWERNOW_K8=m
                +CONFIG_X86_POWERNOW_K8=y
                +CONFIG_X86_PTRACE_BTS=y
                -CONFIG_X86_SPEEDSTEP_CENTRINO=m
                +CONFIG_X86_SPEEDSTEP_CENTRINO=y
                +CONFIG_XENFS=y
                +CONFIG_XEN_COMPAT_XENFS=y

                Comment


                • #18
                  The new U kernel have got speedstep active by default even without manageing tool, but thats not really problematic, you could disable EIST in bios or you can use a gui tool (or echo to set the system to performance mode for benchmarking. The rest is not important. Saidly mainline kernels do not provide aufs or squashfs modules (2.6.29 has at least a not backwards compatible squashfs 4.0), so basically you would need squashfs 4 for all older kernels + aufs for all to combine em into a multi kernel live image for better testing.

                  Comment


                  • #19
                    Originally posted by Kano View Post
                    The new U kernel have got speedstep active by default even without manageing tool, but thats not really problematic, you could disable EIST in bios
                    On all Phoronix desktop test machines, EIST is disabled from the BIOS.
                    Michael Larabel
                    http://www.michaellarabel.com/

                    Comment


                    • #20
                      From the GraphicsMagick maintainer:

                      ###


                      As far as Linux kernel improvements go, there is no telling what the cause of the boost is, but the most obvious improvements would be to make best use of the caches by carefully managing thread affinity, and by minimizing unnecessary cache thrashing during synchronization. Other improvements could be related to scheduling floating point operations.

                      Most open source applications are traditional single threaded. People in research labs may develop multi-thread programs but they are not broadly available or easy to use. Applications like GraphicsMagick make OpenMP generally available to test with and easy to use. It is a good test case for Linux kernel developers. Recently the valgrind author was using GraphicsMagick as a test case (GraphicsMagick did good).

                      It shouldn't be surprising that OpenMP performance improves in recent kernels.
                      Michael Larabel
                      http://www.michaellarabel.com/

                      Comment

                      Working...
                      X