| View previous topic :: View next topic |
| Author |
Message |
seppe
Guru
Joined: 01 Sep 2003
Posts: 431
Location: Hove, Antwerp, Belgium
|
 Posted: Thu Sep 16, 2004 6:36 pm Post subject: 2.6.9-rc2-nitro1 "There's no future in time travel" |
 |
|
Tadaa! A new nitro-sources for 2.6.9-rc2!
Applied patches
| Code: |
2.6.9-rc2-nitro1 "There's no future in time travel"
***************************************************
# Start of CK base
from_2.6.9-rc2_to_staircase8.3 | A complete scheduler policy rewrite
schedrange.diff | Infrastructure for more policies
schedbatch2.4.diff | Batch scheduling
schediso2.6.diff | Isochronous scheduling
mapped_watermark4.diff | Lighter caching, very unlikely to swap
1g_lowmem2_i386.diff | Allows 1G ram without enabling highmem
defaultcfq.diff | Enables the CFQ (completely fair queueing) I/O scheduler by default
cfq_iosched_v2.patch | Completely Fair Queueing v2
cfq_v2_20040909.patch | CFQ update
akpm-latency-fix1.patch | Minor latency improvement hack
9000-SuSE-117-writeback-lat.patch | Writeback latency fix
cddvd-cmdfilter-drop.patch | cdrecord fix
supermount-ng205.diff | automaticly mount removable media
invalidate_inodes-speedup.patch | add seperate list for searching in the inode lists
make-tree_lock-an-rwlock.patch | Convert tree_lock to an rwlock, improves performance at Oracle
# End of CK base (pre1)
reiser4-for-2.6.9-rc2-nitroX | The Reiser4 filesystem
ipw2100-0.54 | Intel Pro Wireless 2100 drivers
acpi-dsdt-initrd-patch-v0.6-2.6.9.patch | Custom acpi dsdt
via-v4l-1.4a-drm.patch | VIA Video4Linux
cpu-vendor-select.diff | select more than 1 CPU vendor
lirc-2.6.5-mm1-20040406 | Linux InfraRed Control support
menuconfig-NAME-v2.1-dev5.patch | Show kernel name in menuconfig
squashfs2.0-patch | SquashFS v2.0, a squashed read-only filesystem for Linux
gcloop-2.6-20040527.patch | Gentoo Compressed loopback support for 2.6
lufs-0.9.7-2.6.0-test9.patch | Linux Userland FileSystem (mount ftp connections, etc ..)
omnibook-2.6.8-rc2-bk2.diff | omnibook support
config-nr-tty-devices.diff | config /dev/tty* count for a cleaner /dev
cdfs-2.6.3a.diff | exports all tracks and boot images on a CD as normal files
acx100-2.6.8-rc2-bk2.diff | ACX WLAN drivers
acerhk.patch | Acer HotKeys support
iteraid_1.44.diff | Giga Raid
configurable-hid-mouse-polling-2.6.9-rc1.patch | usb 500hz mouse hack
packet-2.6.8-2.patch | packet writing support for CD/DVD RW's
journal_clean_checkpoint_list-latency-fix.patch | minor latency improvement hack
pty_write-latency-fix.patch | minor latency improvement hack
igxb-speedup.patch | speed up interrupt routine call
kallsyms-data-size-reduction--lookup-speedup.patch | speedup kallsyms
get_user_pages-latency-fix.patch | minor latency improvement hack
config_hz.diff | Set the internal clock frequency
back_journal_clean_checkpoint_list-latency-fix.patch | Minor latency improvement patch
fbsplash-0.9-r8-2.6.9-rc1.patch | Gensplash, a bootsplash replacement for Gentoo
vesafb-tng-0.9-rc4-r3-2.6.9-rc2.patch | vesafb-tng-0.9-rc4-r2-2.6.8.1.patch | a new and more functional version of the vesafb Linux driver
vesafb_change_config.diff | change default fb
change_reiser4_config.diff | Do not allow 4k stacks with Reiser4
orinoco-0.13e-SN-5 | Orinocco drivers with support for scanning and monitor mode
# Start of Ingo Molnar's latency fixes
preempt-smp.patch
preempt-cleanup.patch
preempt-cleanup-fix.patch
add-lock_need_resched.patch
sched-add-cond_resched_softirq.patch
sched-fix-latency-in-random-driver.patch
sched-ext3-fix-scheduling-latencies-in-ext3.patch
sched-vfs-fix-scheduling-latencies-in-invalidate_inodes.patch
sched-vfs-fix-scheduling-latencies-in-prune_dcache-and-select_parent.patch
sched-net-fix-scheduling-latencies-in-netstat.patch
sched-net-fix-scheduling-latencies-in-__release_sock.patch
sched-mm-fix-scheduling-latencies-in-copy_page_range.patch
sched-mm-fix-scheduling-latencies-in-unmap_vmas.patch
sched-mm-fix-scheduling-latencies-in-get_user_pages.patch
fix-keventd-execution-dependency.patch
sched-fix-scheduling-latencies-in-mttrc.patch
sched-fix-scheduling-latencies-in-vgaconc.patch
sched-fix-scheduling-latencies-for-preempt-kernels.patch
# End of Ingo Molnar's latency fixes
reiser4-cond_resched-build-fix.patch | reiser4 tweak
kernel-events-rml-2.6.9-rc1-1.patch | A simple sysfs change notifier over netlink
|
Download
Ebuild
bz2
What's new?
- based on 2.6.9-rc2
- CK's patches are updated
- Ingo Molnar's latency patches are in. It are the same latency patches which are also in his big voluntary-preempt patch
- kernel-events patch is back
- orinocco is back (but not tested)
- vesafb-tng is updated
Warning, read this!
#1. Gensplash is broken. It doesn't work at my machine. I think it's broken because fbsplash_init(); is misplaced in drivers/video/console/fbcon.c, I tried to replace it etc .. but it didn't helped. Vesafb-tng works without problems though. If anyone knows why fbsplash is broken, please let me know (or make a fix for it  )
#2. Nitro-sources isn't well tested at all, so things could always go wrong. Always keep a second kernel image ready in your boot manager.
More info at
IRC: irc.freenode.net #nitro-sources
WWW: http://sepi.be/nitro.php
RSS: http://sepi.be/nitro-rss.php
Have fun with it 
_________________
nitro-sources, because between stable and experimental there exists only speed
Latest release I made: 2.6.13.2-nitro1 |
|
| Back to top |
|
 |
discomfitor
l33t
Joined: 21 Feb 2003
Posts: 927
Location: None
|
| Posted: Thu Sep 16, 2004 6:43 pm Post subject: |
|
|
Compiling now. I'll edit this to let you know how it runs, but so far the ebuild has worked, and nothing critical has happened. ...I think.
EDIT: Critical things happened when I tried to boot. It didn't boot. What it DID do was infinitely scroll the message "scheduling while atomic". Any ideas?
_________________
There is no substitute for experience.
Imperfection indicates a lack of effort.
Last edited by discomfitor on Thu Sep 16, 2004 7:10 pm; edited 1 time in total |
|
| Back to top |
|
|
miseiler
Tux's lil' helper
Joined: 17 Mar 2003
Posts: 118
|
| Posted: Thu Sep 16, 2004 6:49 pm Post subject: |
|
|
Hmm...no sched-adjust-p4gain? Oh well, not terribly important, I guess.
Will compile this in a sec, hope the reiser4 tweaks don't b0rk my system.
_________________
Cthulhu for president. |
|
| Back to top |
|
|
codergeek42
Bodhisattva
Joined: 05 Apr 2004
Posts: 5142
Location: Anaheim, CA (USA)
|
| Posted: Thu Sep 16, 2004 6:54 pm Post subject: |
|
|
Excellent. Will emerge when I get home later. Thanx for a wonderful kernel patchset!
EDIT: Okey dokey. Source tarballs and patches are downloading. I'll let you know how it goes.
_________________
~~ Peter: Programmer, Mathematician, STEM & Free Software Advocate, Enlightened Agent, Transhumanist, Fedora contributor
Who am I? :: EFF & FSF
Last edited by codergeek42 on Thu Sep 16, 2004 11:29 pm; edited 1 time in total |
|
| Back to top |
|
|
sibov
Tux's lil' helper
Joined: 09 Feb 2004
Posts: 147
Location: Germany
|
| Posted: Thu Sep 16, 2004 10:36 pm Post subject: |
|
|
| Darckness wrote: | Compiling now. I'll edit this to let you know how it runs, but so far the ebuild has worked, and nothing critical has happened. ...I think.
EDIT: Critical things happened when I tried to boot. It didn't boot. What it DID do was infinitely scroll the message "scheduling while atomic". Any ideas? |
ME 2
| Code: | | Now, it runs. I disabled "kernel irq balancing" and "Preemptible Kernel" |
_________________
Gentoo 2.6.17-gentoo-r5 with dmraid Raid0 Dualboot
AMD64 X2 3800+, 4x512MB DDR PC3200Pro Corsair Geforce EN6600GT silencer @ 256MB
4x250GB SATAII Samsung SP120, Asus A8N Premium, Plextor PX712SA DVDRW
Logitech Dinovo, SB Audigy4
Last edited by sibov on Thu Sep 16, 2004 11:15 pm; edited 2 times in total |
|
| Back to top |
|
|
Moloch
Apprentice
Joined: 17 Mar 2003
Posts: 293
Location: Albuquerque, NM, US
|
| Posted: Thu Sep 16, 2004 10:37 pm Post subject: |
|
|
Thanks!
Emerging as I type. Hope it runs as good as the previous releases. |
|
| Back to top |
|
|
discomfitor
l33t
Joined: 21 Feb 2003
Posts: 927
Location: None
|
| Posted: Fri Sep 17, 2004 12:12 am Post subject: |
|
|
| sibov wrote: | | Darckness wrote: | Compiling now. I'll edit this to let you know how it runs, but so far the ebuild has worked, and nothing critical has happened. ...I think.
EDIT: Critical things happened when I tried to boot. It didn't boot. What it DID do was infinitely scroll the message "scheduling while atomic". Any ideas? |
ME 2
| Code: | | Now, it runs. I disabled "kernel irq balancing" and "Preemptible Kernel" |
|
Unfortunately, I don't want to disable preempting in my kernel. Guess I'll wait for a patch/new release.
_________________
There is no substitute for experience.
Imperfection indicates a lack of effort. |
|
| Back to top |
|
|
MrApples
Guru
Joined: 13 Dec 2002
Posts: 511
|
| Posted: Fri Sep 17, 2004 12:27 am Post subject: |
|
|
are any of us running this successfully with preempting?
_________________
http://www.whatsinyourbox.org -- Technology discussion, news, and more. |
|
| Back to top |
|
|
rafelbev
n00b
Joined: 15 Jul 2003
Posts: 53
|
| Posted: Fri Sep 17, 2004 12:54 am Post subject: Trouble with ebuild. Fix needed |
|
|
I just tried emerging and it fails on the patching. I guess the ebuild needs fixing.
| Code: | >>> emerge (1 of 1) sys-kernel/nitro-sources-2.6.9_rc2-r1 to /
>>> md5 src_uri ;-) patch-2.6.9-rc2.bz2
>>> md5 src_uri ;-) linux-2.6.8.tar.bz2
>>> md5 src_uri ;-) patch-2.6.9-rc2-nitro1.bz2
>>> Preparing to unpack...
>>> Unpacking source...
>>> Unpacking linux-2.6.8.tar.bz2 to /var/tmp/portage/nitro-sources-2.6.9_rc2-r1/work
* Applying patch-2.6.9-rc2.patch (-p0+)...
patch: **** read error : No such file or directory
patch: pch.c:614: intuit_diff_type: Assertion `i0 != NONE' failed. [ !! ]
* Please attach /var/tmp/portage/nitro-sources-2.6.9_rc2-r1/temp/patch-2.6.9-rc2.err to any bug you may post.
!!! ERROR: sys-kernel/nitro-sources-2.6.9_rc2-r1 failed.
!!! Function unipatch, Line 534, Exitcode 0
!!! Unable to dry-run patch. |
A quick
| Code: | | head /var/tmp/portage/nitro-sources-2.6.9_rc2-r1/temp/patch-2.6.9-rc2.err |
gives me
| Code: | Attempting Dry-run:
cmd: patch -p0 --dry-run -f < /var/tmp/portage/nitro-sources-2.6.9_rc2-r1/work/p
atches/1/patch-2.6.9-rc2.patch
=======================================================
can't find file to patch at input line 4
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|diff -Nru a/CREDITS b/CREDITS
|--- a/CREDITS 2004-09-12 22:34:41 -07:00
|+++ b/CREDITS 2004-09-12 22:34:41 -07:00
|
The rest is the same. |
|
| Back to top |
|
|
d.roe
n00b
Joined: 12 May 2004
Posts: 27
|
| Posted: Fri Sep 17, 2004 2:34 am Post subject: |
|
|
This nitro is very pretty indeed. It installed perfectly and now I'm compiling two progs, running glxgears at 2800 fps, and listening to music via gstreamer. Best performance I've ever gotten. No trouble compiling nvidia, ipw2100 works straight, the new alsa version is good (was working off 1.0.4 before). I didn't experience any patching problems.
Many thanks for this excellent patch set.
regards |
|
| Back to top |
|
|
Moloch
Apprentice
Joined: 17 Mar 2003
Posts: 293
Location: Albuquerque, NM, US
|
| Posted: Fri Sep 17, 2004 2:37 am Post subject: |
|
|
Yup, works great!
In terms of preempt issues... I thought preempt was never actually a good idea... or is it different with the ck patchset? |
|
| Back to top |
|
|
miseiler
Tux's lil' helper
Joined: 17 Mar 2003
Posts: 118
|
| Posted: Fri Sep 17, 2004 3:00 am Post subject: |
|
|
| Darckness wrote: | | sibov wrote: | | Darckness wrote: | Compiling now. I'll edit this to let you know how it runs, but so far the ebuild has worked, and nothing critical has happened. ...I think.
EDIT: Critical things happened when I tried to boot. It didn't boot. What it DID do was infinitely scroll the message "scheduling while atomic". Any ideas? |
ME 2
| Code: | | Now, it runs. I disabled "kernel irq balancing" and "Preemptible Kernel" |
|
Unfortunately, I don't want to disable preempting in my kernel. Guess I'll wait for a patch/new release. |
| Quote: | Yup, works great!
In terms of preempt issues... I thought preempt was never actually a good idea... or is it different with the ck patchset? |
Preempting is not a good idea, and CK personally recommends against it.
From his site:
| Code: | | I recommend disabling preemption in the kernel configuration unless you have a specific low latency requirement! It causes poorly written applications to misbehave far more. The gains in lateny with in-kernel preemption in 2.6 would not be noticable to a human, unlike 2.4 which had much larger latencies in the kernel. For serious audio work, video capture etc, you should enable preempt. |
I've never noticed a difference, personally.
_________________
Cthulhu for president. |
|
| Back to top |
|
|
codergeek42
Bodhisattva
Joined: 05 Apr 2004
Posts: 5142
Location: Anaheim, CA (USA)
|
| Posted: Fri Sep 17, 2004 3:20 am Post subject: |
|
|
Um...is anyone else getting this?
| Code: | peter-computer /usr/src/linux # uname -a
Linux peter-computer 2.6.9-rc2-nitro1y #1 Thu Sep 16 19:55:57 PDT 2004 i686 AMD Athlon(tm) processor AuthenticAMD GNU/Linux |
Why is there a 'y'? This is messing up my nvidia-kernel modprobing goodness...
_________________
~~ Peter: Programmer, Mathematician, STEM & Free Software Advocate, Enlightened Agent, Transhumanist, Fedora contributor
Who am I? :: EFF & FSF |
|
| Back to top |
|
|
d.roe
n00b
Joined: 12 May 2004
Posts: 27
|
| Posted: Fri Sep 17, 2004 3:22 am Post subject: |
|
|
codergeek42: CONFIG_LOCALVERSION takes a string that it appends at the end of your version, so you can personalise the kernel. You must have typed "y" when making oldconfig. Just make sure CONFIG_LOCALVERSION="" and you'll be back as you were before.
regards, |
|
| Back to top |
|
|
codergeek42
Bodhisattva
Joined: 05 Apr 2004
Posts: 5142
Location: Anaheim, CA (USA)
|
| Posted: Fri Sep 17, 2004 3:29 am Post subject: |
|
|
| d.roe wrote: | codergeek42: CONFIG_LOCALVERSION takes a string that it appends at the end of your version, so you can personalise the kernel. You must have typed "y" when making oldconfig. Just make sure CONFIG_LOCALVERSION="" and you'll be back as you were before.
regards, |
That did it. Thanx. Another excellent kernel patchset. Thanx again, seppe!
_________________
~~ Peter: Programmer, Mathematician, STEM & Free Software Advocate, Enlightened Agent, Transhumanist, Fedora contributor
Who am I? :: EFF & FSF |
|
| Back to top |
|
|
Isaiah
Guru
Joined: 25 Feb 2003
Posts: 359
|
| Posted: Fri Sep 17, 2004 5:03 am Post subject: |
|
|
| MrApples wrote: | | are any of us running this successfully with preempting? |
Yes, we are still doing that here - will add we have never tried "kernel irq balancing" thing  |
|
| Back to top |
|
|
Fire-Reiher
n00b
Joined: 20 Dec 2002
Posts: 28
|
|
| Back to top |
|
|
crazy-bee
Apprentice
Joined: 03 Jan 2003
Posts: 170
|
| Posted: Fri Sep 17, 2004 10:45 am Post subject: Re: 2.6.9-rc2-nitro1 "There's no future in time travel& |
|
|
| seppe wrote: |
- Ingo Molnar's latency patches are in. It are the same latency patches which are also in his big voluntary-preempt patch
|
Nice, been waiting for it. Are they enabled by default with no kernel option, or do I have to enable the "old" preemptive option to get the new voluntary-preempt patch? |
|
| Back to top |
|
|
Rainmaker
Veteran
Joined: 12 Feb 2004
Posts: 1650
Location: /home/NL/ehv/
|
| Posted: Fri Sep 17, 2004 11:30 am Post subject: |
|
|
wow, I didn't even have mm1, and you already had a patchset out 
Kernel runs... not great but good...
Especially switching terminals to X and visa versa is very slow ans slugish.
also this kernel breaks with subfs.
I'm gonna see if this has to do with mm, or your own patches.
Load balance is OK though. I can do a lot of things at the same time, all running pretty quick.
_________________
If you can't dazzle them with brilliance, baffle them with bullshit. |
|
| Back to top |
|
|
Pink
Veteran
Joined: 24 Jul 2003
Posts: 1062
|
| Posted: Fri Sep 17, 2004 12:09 pm Post subject: |
|
|
| Rainmaker wrote: | wow, I didn't even have mm1, and you already had a patchset out
Kernel runs... not great but good...
Especially switching terminals to X and visa versa is very slow ans slugish.
also this kernel breaks with subfs.
I'm gonna see if this has to do with mm, or your own patches.
Load balance is OK though. I can do a lot of things at the same time, all running pretty quick. |
There is no mm in this kernel. As the initial post states, it is based in 2.6.9-rc2. The main 'base' if you like are the ck patches for 2.6.9-rc2.
HTH |
|
| Back to top |
|
|
sobers_2002
Veteran
Joined: 16 Mar 2004
Posts: 1128
|
| Posted: Fri Sep 17, 2004 12:22 pm Post subject: |
|
|
about the fbsplash thing........i saw that spock had released new patches for rc2.......i think they should be working........seppe plz check out.
_________________
Pdict - dockable dictionary client for linux
FREE97WIN: Use this code on Dreamhost and you get $97 off !! |
|
| Back to top |
|
|
Realmaker
l33t
Joined: 17 Sep 2003
Posts: 683
Location: Bremen, Germany
|
| Posted: Fri Sep 17, 2004 12:23 pm Post subject: |
|
|
When i'm trying to boot my new kernel, i get lots of "bad: scheduling while atomic"-errors till nothing happens anymore, mostly when trying to bring up eth0. it is possible to restart with ctrl+alt+del, but it seems like it's taking ages to finish booting 
_________________
Athlon 2600+ Barton, Asus A7N8x-X, nForce2, Asus Geforce FX5700 256mb, TwinMos 512ddr-sdram cl 2.5
Wer andern eine Bratwurst brät, der hat ein Bratwurstbratgerät. |
|
| Back to top |
|
|
Rainmaker
Veteran
Joined: 12 Feb 2004
Posts: 1650
Location: /home/NL/ehv/
|
| Posted: Fri Sep 17, 2004 12:52 pm Post subject: |
|
|
| PickledOnion wrote: | | Rainmaker wrote: | wow, I didn't even have mm1, and you already had a patchset out
Kernel runs... not great but good...
Especially switching terminals to X and visa versa is very slow ans slugish.
also this kernel breaks with subfs.
I'm gonna see if this has to do with mm, or your own patches.
Load balance is OK though. I can do a lot of things at the same time, all running pretty quick. |
There is no mm in this kernel. As the initial post states, it is based in 2.6.9-rc2. The main 'base' if you like are the ck patches for 2.6.9-rc2.
HTH |
Sorry... Same problem with rc2-mm1 and submount though.
No problems with 2.6.8-r3
_________________
If you can't dazzle them with brilliance, baffle them with bullshit. |
|
| Back to top |
|
|
seppe
Guru
Joined: 01 Sep 2003
Posts: 431
Location: Hove, Antwerp, Belgium
|
| Posted: Fri Sep 17, 2004 1:17 pm Post subject: |
|
|
I think Ingo Molnar's patches are causing those "bad: scheduling while atomic" errors. But it looks like you only get those errors when you enable preemptible kernel. Oh and not all his patches only have effect when you've enabled preemptible kernel.
I have problems with this kernel too, I just saw that I couldn't emerge ati-drivers
edit: ATI drivers work now, I was so stupid to emerge ati-drivers while /usr/src/linux was still clean 
_________________
nitro-sources, because between stable and experimental there exists only speed
Latest release I made: 2.6.13.2-nitro1
Last edited by seppe on Fri Sep 17, 2004 2:03 pm; edited 1 time in total |
|
| Back to top |
|
|
woodwizzle
Apprentice
Joined: 30 Nov 2003
Posts: 225
|
| Posted: Fri Sep 17, 2004 1:56 pm Post subject: |
|
|
| does this kernel have a fix for the cdrecord bug? |
|
| Back to top |
|
|
|
Kernel & Hardware
