2.6.23-kamikaze5: "Vipernicus"

Message

Waninkoko · Post by **Waninkoko** » Wed Oct 31, 2007 5:49 pm

· KAMIKAZE-SOURCES ::.
· 2.6.23-kamikaze5: "Vipernicus"

New version of "kamikaze-sources" !!

Links:

Homepage: http://kamikaze.waninkoko.info

Ebuild: kamikaze-sources-2.6.23-r5.ebuild
Patch: linux-2.6.23-kamikaze5.bz2
Broken-out: broken-out-2.6.23-kamikaze5.tar.bz2
Broken-Out patches:http://kamikaze.waninkoko.info/patches/2.6.23/kamikaze5/broken-out

Also on custom-kernels overlay!

What's new:

This one is a really big patch.

Changes in this version:

"sched-cfs-devel" patch dropped.
"sched-cfs-boost" patch added.
"sched-cfs-tunables" patch added.
"genetic-lib" patch added.
"genetic-io-sched" patch added.
"genetic-as-sched" patch added.
"ckpatches" patch updated.
"improve-relatime" patch added.
"balance-dirty-pages" patch added.
"p5wdh-workaround" patch added.
"b44-power-down" patch added.
"iwlwifi" patch updated.
"silence-deprecated-warnings" patch updated.

Patches:

patch-2.6.23.1

# Completely Fair Scheduler patches
sched-cfs-boost-2.6.23.patch
sched-cfs-tunables-2.6.23.patch

# Genetic patches
genetic-lib-2.6.23.patch
genetic-io-sched-2.6.23.patch
genetic-as-sched-2.6.23.patch

# Con Kolivas patchset
ckpatches-2.6.23-1.patch

# gentoo-sources patchset
genpatches-2.6.23-1.patch

# -mm patches
cpufreq-allow-ondemand-and-conservative-cpufreq-governors-to-be-used-as-default.patch
allow-ondemand-and-conservative-cpufreq-governors-to-be-used-as-default-kconfig-fix.patch

# High Resolution Timers
hrtimers-2.6.23-2.patch

# 432/864 HZ kconfig
hz-432-kconfig-option.patch
hz-864-kconfig-option.patch

# ALSA
alsa-1.0.15-2.6.23.patch

# TuxOnIce
tuxonice-3.0-rc1-for-2.6.23.patch

# Linux-PHC
linux-phc-0.3.1-2.6.23.patch

# edac patches
edac-2.6.23-for-2.6.23.patch

realtime-lsm-2.6.23.patch

# Powersaving patches
export-urb-statistics-for-powertop-2.6.23.patch
powertop-2.6.23.patch
pm_qos-2.6.23.patch

# ACPI patches
acpi-release-20070126-2.6.23-1.patch
acer-acpi-2.6.23.patch
fujitsu-laptop-extras-2.6.23-1.patch
thinkpad-2.6.23-1.patch

# Filesystems patches
ext4-updates-2.6.23-1.patch
reiser4-2.6.23-2.patch
squashfs-lzma-2.6.23.patch
unionfs-2.6.23-1.patch
aufs-cvs-0710115-2.6.23.patch
btrfs-0.8-2.6.23-1.patch
shadow-directories-2.6.23.patch

# Disk I/O patches
improve-relatime-2.6.23.patch
balance-dirty-pages-2.6.23.patch

# ATA/SATA patches
libata-async-scan-2.6.23.patch
sata-an-2.6.23.patch
sata-alpm-2.6.23.patch
sata_nv-sw-ncq-support-2.6.23.patch
p5wdh-workaround-2.6.23.patch

cpu-support-rollup-2.6.23.patch

# Mactel
mactel-patches-2.6.23-1.patch

# Video4Linux
gspca-20070508-2.6.23.patch

# Ethernet patches
b44-power-down-2.6.23.patch

# Wireless patches
mac80211-10.0.0-2.6.23.patch
mac80211-led-support-2.6.23.patch
rfkill-led-support-2.6.23.patch
rfkill-hw-rfkill-button-support-2.6.23.patch
sonics-silicon-backplane-2.6.23.patch

# Wireless drivers
b43-2.6.23-1.patch
ipw3945-1.2.2-2.6.23.patch
iwlwifi-1.1.20-2.6.23.patch
iwlwifi-led-support-2.6.23.patch

# Misc patches
silence-deprecated-warnings-2.6.23.patch
colored-printk-output-2.6.23.patch
enable-4k-stacks-default-2.6.23.patch
build-id-sandbox-violation.patch
gcc-4.3.0-compilation-fix.patch
kamikaze-version.patch

Greetings.

gimpel · Post by **gimpel** » Wed Oct 31, 2007 5:58 pm

Wooo genetics are back.. with IO fingerprinting too

And automatic noatime mounting, nice!

/me goes compiling.

buddabrod · Post by **buddabrod** » Wed Oct 31, 2007 6:10 pm

Does it shutdown/suspend now properly? -pre10 didn't and i wonder if this does and i can not test it right now.

dodo1122 · Post by **dodo1122** » Wed Oct 31, 2007 6:21 pm

thanks you. /me compiles

Hope shutdown works now.

dodo

mbar · Post by **mbar** » Wed Oct 31, 2007 6:24 pm

thanks!

mroconnor · Post by **mroconnor** » Wed Oct 31, 2007 6:31 pm

Thanks, I am now addicted to kernel compiling.

/me compiles too!

StifflerStealth · Post by **StifflerStealth** » Wed Oct 31, 2007 6:43 pm

I love the name of this kernel release.

It's the most honorable name for a kernel since Vipe was heavy into kernels. May his memory live on in this release. RIP Vipernicus.

dodo1122 · Post by **dodo1122** » Wed Oct 31, 2007 7:00 pm

StifflerStealth wrote:I love the name of this kernel release. It's the most honorable name for a kernel since Vipe was heavy into kernels. May his memory live on in this release. RIP Vipernicus.

Hear Hear. Vipernicus was a great guy, and i loved his kernel patches; i've been using them since emission-sources

dodo

Dottout · Post by **Dottout** » Wed Oct 31, 2007 7:01 pm

tnx

and.. best choice for the name

edit: uvesafb doesnt work, dmesg:

Code: Select all

uvesafb: Getting VBE info block failed (eax=0x4f00, err=1)
uvesafb: vbe_init() failed with -22
uvesafb: probe of uvesafb.0 failed with error -22

and:

Code: Select all

v86d[942]: segfault at fffffffe eip 0804932d esp af9e37f0 error 6

edit2: fixed, it was my fault, my bzImage was compiled against v86d built with gcc-4.3.I had to re-emerge it and make again bzImage

eremini · Post by **eremini** » Wed Oct 31, 2007 7:43 pm

Does anybody know how well this kernel will perform on a server system (ussual LAMP with heavy mySQL usage). Is it feaseble compared to standard gentoo-sources?

StifflerStealth · Post by **StifflerStealth** » Wed Oct 31, 2007 8:18 pm

uvesafb works for me:
~ # dmesg | grep uvesafb
Kernel command line: quiet root=/dev/ram0 init=/linuxrc real_root=/dev/hda5 udev video=uvesafb:1280x1024-32,mtrr:3,ywrap
uvesafb: NVIDIA Corporation, nv44 Board - p362h5 , Chip Rev , OEM: NVIDIA, VBE v3.0
uvesafb: protected mode interface info at c000:d1f0
uvesafb: pmi: set display start = b00cd226, set palette = b00cd290
uvesafb: pmi: ports = 3b4 3b5 3ba 3c0 3c1 3c4 3c5 3c6 3c7 3c8 3c9 3cc 3ce 3cf 3d0 3d1 3d2 3d3 3d4 3d5 3da
uvesafb: VBIOS/hardware doesn't support DDC transfers
uvesafb: no monitor limits have been set, default refresh rate will be used
uvesafb: scrolling: ywrap using protected mode interface, yres_virtual=2048
uvesafb: framebuffer at 0xe0000000, mapped to 0xf0880000, using 10240k, total 262144k

Did you add the line in the kernel initramfs to point to the correct file? /usr/share/v86d/initramfs I think.

@eremini: It should, just use the deadline scheduler.

I am now running this kernel and it's very fast indeed.

Great work!

gimpel · Post by **gimpel** » Wed Oct 31, 2007 9:36 pm

I'm running with full-blown genetic stuff now, including IO fingerprinting. Let's see o/

This kernel gets a place in my in-memoriam kernel corner, next to last -nitro, -ck and -viper.
R.I.P. Vipernicus!

GenTreK · Post by **GenTreK** » Wed Oct 31, 2007 11:07 pm

Hi guys,

Waninkoko -- Great patch-set.

I was recently trying to separate the Reiser4 patch from Andrew Morton's 2.6.23-mm1 patch set.

I grabbed all the patches mentioning Reiser4 from the broken out archive and tried to patch 2.6.23 with those, but the patches needed other patches.

I installed quilt to handle the hundreds of mm patches and used

quilt graph reiser4.patch

and the like, to figure which other patches are needed, but had no real success.

Would you guys who know what you are doing, tell me how you do such a thing.

Thanks.

Waninkoko · Post by **Waninkoko** » Wed Oct 31, 2007 11:10 pm

GenTreK, I just ported reiser4 2.6.22 patch and then I applied some reiser4 patches from -mm.

GenTreK · Post by **GenTreK** » Wed Oct 31, 2007 11:23 pm

Waninkoko wrote:GenTreK, I just ported reiser4 2.6.22 patch and then I applied some reiser4 patches from -mm.

Yeap. That makes perfect sense.

Generally, how would you go about separating a subset of patches from a large patch-set like mm?

Any ideas on how to tackle such a problem? Or is it just one of those "too hard" problems?

Waninkoko · Post by **Waninkoko** » Wed Oct 31, 2007 11:38 pm

GenTreK wrote:
Waninkoko wrote:GenTreK, I just ported reiser4 2.6.22 patch and then I applied some reiser4 patches from -mm.
Yeap. That makes perfect sense.

Generally, how would you go about separating a subset of patches from a large patch-set like mm?

Any ideas on how to tackle such a problem? Or is it just one of those "too hard" problems?

When I want some patches from -mm I just get them and try if they work. If they don't work then I investigate which extras patches I need. Depending of the patches this can be easy or pretty hard (ie: reiser4).

tranquilcool · Post by **tranquilcool** » Wed Oct 31, 2007 11:42 pm

great set. genetic is rock and roll.
great name to remember a nice guy.

thanks.

Kajan · Post by **Kajan** » Wed Oct 31, 2007 11:51 pm

All is OK, nice

Pozdrawiam

GenTreK · Post by **GenTreK** » Thu Nov 01, 2007 12:04 am

Waninkoko wrote:When I want some patches from -mm I just get them and try if they work. If they don't work then I investigate which extras patches I need. Depending of the patches this can be easy or pretty hard (ie: reiser4).

Do you use quilt, or any other patch handling software?

How did you figure which of the earlier (seemingly non-reiser4) patches you needed for reiser4?

brot · Post by **brot** » Thu Nov 01, 2007 12:10 am

StifflerStealth wrote:I love the name of this kernel release. It's the most honorable name for a kernel since Vipe was heavy into kernels. May his memory live on in this release. RIP Vipernicus.

nice said StifflerStealth.

I am compiling it right now,..
It is very kind of you to name the release like this Waninkoko. Vipernicus really did a good job, and was into all that kernel stuff. And he cared what others said and helped them.

DrAgOnTuX · Post by **DrAgOnTuX** » Thu Nov 01, 2007 12:54 am

compiled and works, thanks

R.I.P. Vipernicus

kernelOfTruth · Post by **kernelOfTruth** » Thu Nov 01, 2007 1:11 am

This kernel gets a place in my in-memoriam kernel corner, next to last -nitro, -ck and -viper.
R.I.P. Vipernicus!

++

definitely! this will get an place of honor in my kernel (-patchset) collection

R.I.P. Vipernicus - may your great memory in our minds never fade away

for me you were a great person - always seeking in helping others, an highly-involved community member and a near-friend

edit:

wooohoo ! genetic scheduling - rock and roll, baby

Kwark · Post by **Kwark** » Thu Nov 01, 2007 8:59 am

I want to build a good gaming and desktop work kernel.
The vanilla (gentoo) kernel just doesn't cut it sometimes with large file transfers and working in the background.

Would the kamikaze kernel be a good option for me?If so, does it need much configuring?
Or is it a matter of applying the patches and setting the correct scheduler?

kernelOfTruth · Post by **kernelOfTruth** » Thu Nov 01, 2007 9:17 am

just saw:

"sched-cfs-devel" patch dropped.

wasn't it that good yet compared to the default in .23 ?

lmmsci · Post by **lmmsci** » Thu Nov 01, 2007 9:49 am

Great! I'm going to try it as quick as it will be possible (in free time).
BTW: What's happened with USB DSL modem support in 2.6.23 kernel?