omg! Con Kolivas is back!

Fackamato · 2009-09-07 22:03:12

This is very interesting.

How about some benchmarks from our own Archlinux users?

ugkbunb · 2009-09-07 22:14:17

^^
there is some on the first page

sand_man · 2009-09-07 22:52:07

After playing for a couple of hours last night, I did get that strange, unkillable process. No process in particular. Once it was firefox, once it was gcc while compiling.

brisbin33 · 2009-09-07 23:22:30

i've been booted in bfs for two days now. its been really nice actually. core2quad 2.6; i was just recompiling the kernel as a test, all four cores at 100%, video playing in mplayer, mpd running, six or seven terms running the usual (mutt/ncmcpp/irssi), a few uzbls open, xcompmgr running, xmonad switching workspaces / moving windows, all this at once and not one stutter, glitch, or hiccup.

it was nice to see absolutely 0 lack of responsiveness with 100% usage on all four cores. and i haven't noticed any bugs either. i think i might just stay in this kernel for a while.

just my experience so far, YMMV.

ugkbunb · 2009-09-08 00:18:28

I too have been using kernel26-bfs for a few days now... and although I never felt any stutters before the switch (except with audio when I was using Pulseaudio)... my system seems just as responsive as it was... so it definately not a regression... it perhaps could be faster but I can't tell much of a difference.

Pox · 2009-09-08 13:09:23

I tried this out for an hour or so - things might have been a bit snappier, but nothing drastic, and then I ended up with an unkillable firefox process and went back to the normal kernel.

bangkok_manouel · 2009-09-08 14:58:31

just FTR, not one single problem (and in particular, no unkillable task) with 2.6.31-rc9-sched-bfs-210. using it on 2 machines (home+work), works great.

gonX · 2009-09-08 16:28:54

Perhaps I am missing something - but I tried patching my 2.6.31-rc9 source with the 2.6.31-rc9-sched-bfs-210.patch by doing this in my kernelbuild dir:

patch -p1 < ../2.6.31-rc9-sched-bfs-210.patch

But I can't select it from the menuconfig - is this normal?

Last edited by gonX (2009-09-08 16:29:18)

bangkok_manouel · 2009-09-08 16:42:08

gonX wrote:

Perhaps I am missing something - but I tried patching my 2.6.31-rc9 source with the 2.6.31-rc9-sched-bfs-210.patch by doing this in my kernelbuild dir:
patch -p1 < ../2.6.31-rc9-sched-bfs-210.patch
But I can't select it from the menuconfig - is this normal?

you don't have to select anything actually so yes, that's normal. IIRC, only zen-sources will let you chose between cfs and bfs.

Zariel · 2009-09-08 16:53:46

Hm, anyone running this on x64 ? I get the boot hang on SCSI ..

also theres a fair chunk of discussion over at Zen Thread

bangkok_manouel · 2009-09-08 17:02:02

Zariel wrote:

Hm, anyone running this on x64 ? I get the boot hang on SCSI ..

running a 64bit kernel here (with 32bit userland but that shouldn't make any difference)

gonX · 2009-09-08 17:36:52

Yeah I ended up patching it without problems, but it randomly gets stuck on detecting my harddrives during boot. Also, adding "elevator=bfs" to the kernel line gives me "Scheduler 'bfs' does not exist" or something in the lines of that. I run mdadm on my root device if that makes any difference.

*EDIT*

I did some looking around and it seems that performance counters need to be disabled - these were on, and I'm in the process of recompiling my kernel now

Last edited by gonX (2009-09-08 17:47:41)

bangkok_manouel · 2009-09-08 17:59:47

gonX wrote:

Also, adding "elevator=bfs" to the kernel line gives me "Scheduler 'bfs' does not exist" or something in the lines of that.

this option is for IO schedulers, not CPU schedulers, you don't have to do anything expect patching the source and you'll be using bfs.

gonX · 2009-09-08 18:31:30

bangkok_manouel wrote:

gonX wrote:
Also, adding "elevator=bfs" to the kernel line gives me "Scheduler 'bfs' does not exist" or something in the lines of that.
this option is for IO schedulers, not CPU schedulers, you don't have to do anything expect patching the source and you'll be using bfs.

OH, haha! silly me.
Well I got it running, after forcing my system to use PATA, booting a livecd and remaking the initrd - so far I'm not getting any xruns at all in JACK while running BOINC in the background - at the same settings I'd be getting a few xruns on with the standard kernel.

pyther · 2009-09-08 19:30:26

Meh, I just might have to try this
30KB down sucks though

Last edited by pyther (2009-09-08 19:33:09)

Anonymo · 2009-09-08 19:31:55

tomk wrote:

hehe... it's already there - although the comment is not encouraging.

Anyone that has this compiled for i686, can they post a MediaFire link or something. I tried to compile my own but got some errors and don't want to wait another 3 hours on this slow machine.

.:B:. · 2009-09-08 21:56:34

gonX wrote:

Perhaps I am missing something - but I tried patching my 2.6.31-rc9 source with the 2.6.31-rc9-sched-bfs-210.patch by doing this in my kernelbuild dir:
patch -p1 < ../2.6.31-rc9-sched-bfs-210.patch
But I can't select it from the menuconfig - is this normal?

If you look at the patch, you'll see Con happily tosses out almost all the scheduler stuff that is in the mainline tree. So no, it's not like you have any choice .

Check what it says at the start of the patch:

 Documentation/sysctl/kernel.txt |   25 
 fs/pipe.c                       |    4 
 fs/proc/base.c                  |    2 
 include/linux/init_task.h       |   15 
 include/linux/ioprio.h          |    2 
 include/linux/sched.h           |  193 
 init/Kconfig                    |   61 
 kernel/Makefile                 |    4 
 kernel/delayacct.c              |    2 
 kernel/exit.c                   |    6 
 kernel/fork.c                   |    2 
 kernel/kthread.c                |    4 
 kernel/posix-cpu-timers.c       |   12 
 kernel/sched.c                  |10241 ----------------------------------------
 kernel/sched_bfs.c              | 5793 ++++++++++++++++++++++
 kernel/sched_debug.c            |  509 -
 kernel/sched_fair.c             | 1835 -------
 kernel/sched_idletask.c         |  128 
 kernel/sched_rt.c               | 1771 ------
 kernel/sysctl.c                 |  145 
 kernel/trace/trace.c            |    4 
 kernel/workqueue.c              |    2 
 mm/oom_kill.c                   |    2 
 23 files changed, 5896 insertions(+), 14866 deletions(-)

Ranguvar · 2009-09-08 22:39:46

Huh. The old patch just took out everything in CFS (sched.c) and moved it to sched_cfs.c, made a new sched.c and created sched_bfs.c as the BFS scheduler. In other words, a clean approach allowing you to choose between the schedulers. Looks like this approach got old to maintain.

eldragon · 2009-09-08 23:23:15

i just built 2.6.31-rc9-git1 with the patch applied...gnome never got to load. stuck right after the wallpaper. everything else worked ok (i just went to a term and rebooted)..

any clues? PKGBUILD follows

# $Id: PKGBUILD 48125 2009-07-31 18:27:10Z tpowa $
# Maintainer: Tobias Powalowski <tpowa@archlinux.org>
# Maintainer: Thomas Baechler <thomas@archlinux.org>
# pkgname=kernel26                # Build stock -ARCH kernel
pkgname=kernel26-bfs       # Build kernel with a different name
_kernelname=${pkgname#kernel26}
_basekernel=2.6.30
pkgver=2.6.31
pkgrel=1
_patchname="patch-${pkgver}-rc9"
_gitver=1
pkgdesc="The Linux Kernel and modules"
arch=(i686 x86_64)
license=('GPL2')
groups=('base')
url="http://www.kernel.org"
backup=(etc/mkinitcpio.d/${pkgname}.preset)
depends=('coreutils' 'kernel26-firmware>=2.6.30' 'module-init-tools' 'mkinitcpio>=0.5.20')
# pwc, ieee80211 and hostap-driver26 modules are included in kernel26 now
# nforce package support was abandoned by nvidia, kernel modules should cover everything now.
# kernel24 support is dropped since glibc24
replaces=('kernel24' 'kernel24-scsi' 'kernel26-scsi'
          'alsa-driver' 'ieee80211' 'hostap-driver26'
          'pwc' 'nforce' 'squashfs' 'unionfs' 'ivtv'
          'zd1211' 'kvm-modules' 'iwlwifi' 'rt2x00-cvs'
          'gspcav1' 'atl2' 'wlan-ng26' 'aufs' 'rt2500')
install=kernel26.install
source=(ftp://ftp.kernel.org/pub/linux/kernel/v2.6/linux-$_basekernel.tar.bz2
        http://kernel.org/pub/linux/kernel/v2.6/testing/${_patchname}.bz2
    http://kernel.org/pub/linux/kernel/v2.6/snapshots/${_patchname}-git${_gitver}.bz2
    #patches i need
    patch-logo-arch
    2.6.31-rc9-sched-bfs-210.patch
    quirks-h12y.patch
    # the main kernel config files
        config config.x86_64
        # standard config files for mkinitcpio ramdisk
        kernel26.preset)
optdepends=('crda: to set the correct wireless channels of your country')

build() {
  KARCH=x86

  cd ${srcdir}/linux-$_basekernel
  # Add -ARCH patches
  # See http://projects.archlinux.org/git/?p=linux-2.6-ARCH.git;a=summary
  patch -Np1 -i ${srcdir}/${_patchname} || return 1
  patch -Np1 -i ${srcdir}/${_patchname}-git${_gitver} || return 1
##BRAIN FUCK
  patch -Np1 -i ${srcdir}/2.6.31-rc9-sched-bfs-210.patch
  #  patch -Np1 -i ${srcdir}/patch-logo-arch || return 1
  patch -Np1 -i ${srcdir}/quirks-h12y.patch || return 1


  if [ "$CARCH" = "x86_64" ]; then
    cat ../config.x86_64 >./.config
  else
    cat ../config >./.config
  fi
  if [ "${_kernelname}" != "" ]; then
    sed -i "s|CONFIG_LOCALVERSION=.*|CONFIG_LOCALVERSION=\"${_kernelname}\"|g" ./.config
  fi
  # get kernel version  
  make prepare
  _kernver="$(make kernelrelease)"
  # load configuration
  # Configure the kernel. Replace the line below with one of your choice.
  #make menuconfig # CLI menu for configuration
  #make xconfig # X-based configuration
  #make oldconfig # using old config from previous kernel version
  # ... or manually edit .config
  ####################
  # stop here
  # this is useful to configure the kernel
  #msg "Stopping build"
  #return 1
  ####################
  yes "" | make config
  # build!
  make bzImage modules || return 1
  mkdir -p ${pkgdir}/{lib/modules,boot}
  make INSTALL_MOD_PATH=${pkgdir} modules_install || return 1
  cp System.map ${pkgdir}/boot/System.map26${_kernelname}
  cp arch/$KARCH/boot/bzImage ${pkgdir}/boot/vmlinuz26${_kernelname}
  install -D -m644 Makefile \
    ${pkgdir}/usr/src/linux-${_kernver}/Makefile
  install -D -m644 kernel/Makefile \
    ${pkgdir}/usr/src/linux-${_kernver}/kernel/Makefile
  install -D -m644 .config \
    ${pkgdir}/usr/src/linux-${_kernver}/.config
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/include

  for i in acpi asm-{generic,x86} config linux math-emu media net pcmcia scsi sound trace video; do
    cp -a include/$i ${pkgdir}/usr/src/linux-${_kernver}/include/
  done

  # copy arch includes for external modules
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/arch/x86
  cp -a arch/x86/include ${pkgdir}/usr/src/linux-${_kernver}/arch/x86/

  # copy files necessary for later builds, like nvidia and vmware
  cp Module.symvers ${pkgdir}/usr/src/linux-${_kernver}
  cp -a scripts ${pkgdir}/usr/src/linux-${_kernver}
  # fix permissions on scripts dir
  chmod og-w -R ${pkgdir}/usr/src/linux-${_kernver}/scripts
  #mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/.tmp_versions

  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/arch/$KARCH/kernel

  cp arch/$KARCH/Makefile ${pkgdir}/usr/src/linux-${_kernver}/arch/$KARCH/
  if [ "$CARCH" = "i686" ]; then
    cp arch/$KARCH/Makefile_32.cpu ${pkgdir}/usr/src/linux-${_kernver}/arch/$KARCH/
  fi
  cp arch/$KARCH/kernel/asm-offsets.s ${pkgdir}/usr/src/linux-${_kernver}/arch/$KARCH/kernel/

  # add headers for lirc package
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/video
  cp drivers/media/video/*.h  ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/video/
  for i in bt8xx cpia2 cx25840 cx88 em28xx et61x251 pwc saa7134 sn9c102 usbvideo zc0301; do
   mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/video/$i
   cp -a drivers/media/video/$i/*.h ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/video/$i
  done
  # add docbook makefile
  install -D -m644 Documentation/DocBook/Makefile \
    ${pkgdir}/usr/src/linux-${_kernver}/Documentation/DocBook/Makefile
  # add dm headers
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/drivers/md
  cp drivers/md/*.h  ${pkgdir}/usr/src/linux-${_kernver}/drivers/md
  # add inotify.h
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/include/linux
  cp include/linux/inotify.h ${pkgdir}/usr/src/linux-${_kernver}/include/linux/
  # add wireless headers
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/net/mac80211/
  cp net/mac80211/*.h ${pkgdir}/usr/src/linux-${_kernver}/net/mac80211/
  # add dvb headers for external modules
  # in reference to:
  # http://bugs.archlinux.org/task/9912
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/dvb/dvb-core
  cp drivers/media/dvb/dvb-core/*.h ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/dvb/dvb-core/
  # add dvb headers for external modules
  # in reference to:
  # http://bugs.archlinux.org/task/11194
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/include/config/dvb/
  cp include/config/dvb/*.h ${pkgdir}/usr/src/linux-${_kernver}/include/config/dvb/
  # add dvb headers for http://mcentral.de/hg/~mrec/em28xx-new
  # in reference to:
  # http://bugs.archlinux.org/task/13146
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/dvb/frontends/
  cp drivers/media/dvb/frontends/lgdt330x.h ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/dvb/frontends/
  cp drivers/media/video/msp3400-driver.h ${pkgdir}/usr/src/linux-${_kernver}/drivers/media/dvb/frontends/
  # add xfs and shmem for aufs building
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/fs/xfs
  mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/mm
  cp fs/xfs/xfs_sb.h ${pkgdir}/usr/src/linux-${_kernver}/fs/xfs/xfs_sb.h
  # add headers vor virtualbox
  # in reference to:
  # http://bugs.archlinux.org/task/14568
  cp -a include/drm $pkgdir/usr/src/linux-${_kernver}/include/
  # add headers for broadcom wl
  # in reference to:
  # http://bugs.archlinux.org/task/14568
  cp -a include/trace $pkgdir/usr/src/linux-${_kernver}/include/
  # add vmlinux
  cp vmlinux ${pkgdir}/usr/src/linux-${_kernver}
  # copy in Kconfig files
  for i in `find . -name "Kconfig*"`; do 
    mkdir -p ${pkgdir}/usr/src/linux-${_kernver}/`echo $i | sed 's|/Kconfig.*||'`
    cp $i ${pkgdir}/usr/src/linux-${_kernver}/$i
  done

  cd ${pkgdir}/usr/src/linux-${_kernver}/include && ln -s asm-$KARCH asm

  chown -R root.root ${pkgdir}/usr/src/linux-${_kernver}
  find ${pkgdir}/usr/src/linux-${_kernver} -type d -exec chmod 755 {} \;
  cd ${pkgdir}/lib/modules/${_kernver} && \
    (rm -f source build; ln -sf ../../../usr/src/linux-${_kernver} build)
  # install fallback mkinitcpio.conf file and preset file for kernel
  install -m644 -D ${srcdir}/kernel26.preset ${pkgdir}/etc/mkinitcpio.d/${pkgname}.preset || return 1
  # set correct depmod command for install
  sed \
    -e  "s/KERNEL_NAME=.*/KERNEL_NAME=${_kernelname}/g" \
    -e  "s/KERNEL_VERSION=.*/KERNEL_VERSION=${_kernver}/g" \
    -i $startdir/kernel26.install
  sed \
    -e "s|source .*|source /etc/mkinitcpio.d/kernel26${_kernelname}.kver|g" \
    -e "s|default_image=.*|default_image=\"/boot/${pkgname}.img\"|g" \
    -e "s|fallback_image=.*|fallback_image=\"/boot/${pkgname}-fallback.img\"|g" \
    -i ${pkgdir}/etc/mkinitcpio.d/${pkgname}.preset

  echo -e "# DO NOT EDIT THIS FILE\nALL_kver='${_kernver}'" > ${startdir}/pkg/etc/mkinitcpio.d/${pkgname}.kver
  # remove unneeded architectures
  rm -rf ${pkgdir}/usr/src/linux-${_kernver}/arch/{alpha,arm,arm26,avr32,blackfin,cris,frv,h8300,ia64,m32r,m68k,m68knommu,mips,microblaze,mn10300,parisc,powerpc,ppc,s390,sh,sh64,sparc,sparc64,um,v850,xtensa}
  # remove the firmware
  rm -rf ${pkgdir}/lib/firmware
}
md5sums=('7a80058a6382e5108cdb5554d1609615'
         '6bb612cbd006be9bdfaa6ff29ce090ea'
         '157d7877a9425f5effbb4a150121ddd8'
         '8ab65785c335b7cd8d7d5d829fc215c3'
         '3f462de28e625ce858cca4514e235f6a'
         '759ea1da1c1697c507a2c7c8779ca1c5'
         'bac951d734bbfb9e4f57cf26d24fefbe'
         'cdf9b2696a101d2d69bcb975ab805051'
         '25584700a0a679542929c4bed31433b6')

pyther · 2009-09-09 01:56:24

  CC      kernel/sched_bfs.o
kernel/sched_bfs.c: In function 'init_numa_sched_groups_power':
kernel/sched_bfs.c:4844: error: implicit declaration of function 'group_first_cpu'
kernel/sched_bfs.c: At top level:
kernel/sched_bfs.c:4905: error: conflicting types for 'group_first_cpu'
kernel/sched_bfs.c:4844: note: previous implicit declaration of 'group_first_cpu' was here
make[1]: *** [kernel/sched_bfs.o] Error 1
make: *** [kernel] Error 2

Any one have any ideas what might be causing this?

bangkok_manouel · 2009-09-09 02:32:19

could you please pastebin your kernel config?

iphitus · 2009-09-09 02:58:32

Whoa, Ingo completely missed the point. His tests all benchmark performance attributes, throughput, computation, etc. BFS is not designed for performance.

That's like claiming my 93 Toyota Camry is useless because it lacks the acceleration of a Porsche. Despite it having better range, reliability, fuel economy, lower running costs, self servicable, cheaper parts, etc.

Last edited by iphitus (2009-09-09 03:11:22)

eldragon · 2009-09-09 03:08:53

i *think* BFS was designed to spam the LKML

ngoonee · 2009-09-09 04:04:14

Very possible

raf_kig · 2009-09-09 08:02:43

Testing it right now and it does seem to make a differnence here. Runnnig HD youtube videos + opening a few tabs in firefox is something that used to reliably produce stutter, with bfs it's perfectly smooth. Still, i'll have to think about some more standardized tests.

Arch Linux

#51 2009-09-07 22:03:12

Re: omg! Con Kolivas is back!

#52 2009-09-07 22:14:17

Re: omg! Con Kolivas is back!

#53 2009-09-07 22:52:07

Re: omg! Con Kolivas is back!

#54 2009-09-07 23:22:30

Re: omg! Con Kolivas is back!

#55 2009-09-08 00:18:28

Re: omg! Con Kolivas is back!

#56 2009-09-08 13:09:23

Re: omg! Con Kolivas is back!

#57 2009-09-08 14:58:31

Re: omg! Con Kolivas is back!

#58 2009-09-08 16:28:54

Re: omg! Con Kolivas is back!

#59 2009-09-08 16:42:08

Re: omg! Con Kolivas is back!

#60 2009-09-08 16:53:46

Re: omg! Con Kolivas is back!

#61 2009-09-08 17:02:02

Re: omg! Con Kolivas is back!

#62 2009-09-08 17:36:52

Re: omg! Con Kolivas is back!

#63 2009-09-08 17:59:47

Re: omg! Con Kolivas is back!

#64 2009-09-08 18:31:30

Re: omg! Con Kolivas is back!

#65 2009-09-08 19:30:26

Re: omg! Con Kolivas is back!

#66 2009-09-08 19:31:55

Re: omg! Con Kolivas is back!

#67 2009-09-08 21:56:34

Re: omg! Con Kolivas is back!

#68 2009-09-08 22:39:46

Re: omg! Con Kolivas is back!

#69 2009-09-08 23:23:15

Re: omg! Con Kolivas is back!

#70 2009-09-09 01:56:24

Re: omg! Con Kolivas is back!

#71 2009-09-09 02:32:19

Re: omg! Con Kolivas is back!

#72 2009-09-09 02:58:32

Re: omg! Con Kolivas is back!

#73 2009-09-09 03:08:53

Re: omg! Con Kolivas is back!

#74 2009-09-09 04:04:14

Re: omg! Con Kolivas is back!

#75 2009-09-09 08:02:43

Re: omg! Con Kolivas is back!

Board footer