Official thread: "zen-sources" - Part III

[dodo1122]

works fine here (on my desktop), although i can't get >rc6 to run on my thinkpad (it locks up/panics after some time).


dodo

[mbar]

Does 2.6.24-zen5 have those reiser4 patches?

[gimpel]

so 2.6.25-rc8-zen0 boots ? no hickups so far ?
is it stable enough for production already?

Did not yet reboot into -rc8, but that -rc7 from 01.04. is still running fine w/o any hickups or any other problems.
Nvidia (patched) and virtualbox (patched) are working fine.

And I still use a full blown config including zen low-latency CFS profile, rcu-preempt, genetic stuff, -march=native, reiser4 on / and /usr etc...
I had the two reiser4 patches applied manually since they were posted on the ml, no probs there too.

There weren't that much changes in -rc8 affecting my setup, so I think that's fine too.

The only thing that seems totally screwed up is fsck'ing a ccreg40 enabled reiser4 partition, that spits out a list of wrong sizes for almost every block on the disk, even though the partition works fine. But that's not a -zen problem. Nightly rdiff backup to spare ext3 partition ftw.

[buubuu]

with 25-rc6-zen0 and now rc8-zen0 i get random freezes at boot...
just now starting my laptop had to reset it twice to be able to start the system

it has always happen in the first boot after i install the kernel.. (both my gentoo and slackware)
is this happening to anybody else? is there a known cause?


EDIT:
gentoo, slackware -> diferent machines

[darklegion]

with 25-rc6-zen0 and now rc8-zen0 i get random freezes at boot...
just now starting my laptop had to reset it twice to be able to start the system

it has always happen in the first boot after i install the kernel.. (both my gentoo and slackware)
is this happening to anybody else? is there a known cause?

Nah, 2.6.25-rc8-zen0 doesn't have any boot issues for me, but I'm not using a laptop.BTW does anyone else have issues with uvesafb when setting CONFIG_NONPROMISC_DEVMEM=Y ? It still boots up but uvesafb fails to load.This is probably because of the way uvesafb is designed but if not, I thought I'd mention it.

[dodo1122]

with 25-rc6-zen0 and now rc8-zen0 i get random freezes at boot...
just now starting my laptop had to reset it twice to be able to start the system

it has always happen in the first boot after i install the kernel.. (both my gentoo and slackware)
is this happening to anybody else? is there a known cause?


EDIT:
gentoo, slackware -> diferent machines

I have that on my thinkpad, as i stated above working fine on my desktop though. I'll have to investigate, although it's pretty hard, as it leaves no backtrace or anything of the kind as far as i know.

dodo

[kernelOfTruth]

Nvidia (patched) and virtualbox (patched) are working fine.

could you please post links to those patches?

or do nvidia-drivers & virtualbox-modules work well without patching meanwhile ?

[gimpel]

Nvidia (patched) and virtualbox (patched) are working fine.

could you please post links to those patches?

or do nvidia-drivers & virtualbox-modules work well without patching meanwhile ?

They don't.

Nvidia: http://www.nvnews.net/vbulletin/showthread.php?t=110088
Vbox: http://www.virtualbox.org/attachment/ti ... -out.patch (or ready-to-go here)

[kernelOfTruth]

Nvidia (patched) and virtualbox (patched) are working fine.

could you please post links to those patches?

or do nvidia-drivers & virtualbox-modules work well without patching meanwhile ?

They don't.

Nvidia: http://www.nvnews.net/vbulletin/showthread.php?t=110088
Vbox: http://www.virtualbox.org/attachment/ti ... -out.patch (or ready-to-go here)

thanks !

unfortunately your server doesn't respond

Fehler: Verbindung fehlgeschlagen

Firefox kann keine Verbindung zu dem Server unter gimpel.ath.cx aufbauen.

[dodo1122]

nvidia drivers compile fine without any patches with master-devel.

[gimpel]

Firefox kann keine Verbindung zu dem Server unter gimpel.ath.cx aufbauen.

Heck, fixed. I was playing with setting up a transparent proxy and somehow messed up iptables.

Haven't tried nVidia w/o patch for quite a while.

[dusanc]

...
The only thing that seems totally screwed up is fsck'ing a ccreg40 enabled reiser4 partition, that spits out a list of wrong sizes for almost every block on the disk, even though the partition works fine. But that's not a -zen problem. Nightly rdiff backup to spare ext3 partition ftw.

Quote from personal mail:

ok, when checking/repairing compressed partition use fsck with
the option -n to not complaint about wrong bytes
...
Ignore reports about wrong bytes.
This is because cryptcompress plugin doesn't keep a track about bytes,
and fsck calculates and insert correct value. Keeping a track of bytes
by kernel leads to performance drop (even worse then with atime), so
we decided to not support it.

I hope this helps
Dushan

[gimpel]

I hope this helps
Dushan

Indeed, that clarifies it.
Thanks!

[kernelOfTruth]

Firefox kann keine Verbindung zu dem Server unter gimpel.ath.cx aufbauen.

Heck, fixed. I was playing with setting up a transparent proxy and somehow messed up iptables.

Haven't tried nVidia w/o patch for quite a while.

ah, that was the reason,

now it worked, thanks

it compiled fine so the following ebuild should be correct, I guess ?

(just for reference)

Code: Select all

# Copyright 1999-2008 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/app-emulation/virtualbox-modules/virtualbox-modules-1.5.6.ebuild,v 1.1 2008/02/25 21:59:19 flameeyes Exp $

inherit eutils linux-mod

MY_P=vbox-kernel-module-src-${PV}
DESCRIPTION="Modules for Virtualbox OSE"
HOMEPAGE="http://www.virtualbox.org/"
SRC_URI="http://www.virtualbox.org/download/${PV}/${MY_P}.tar.gz"

LICENSE="GPL-2"
SLOT="0"
KEYWORDS="~amd64 ~x86"
IUSE=""

RDEPEND="!=app-emulation/virtualbox-9999"

S=${WORKDIR}/vboxdrv

BUILD_TARGETS="all"
BUILD_TARGET_ARCH="${ARCH}"
MODULE_NAMES="vboxdrv(misc:${S})"

pkg_setup() {
	linux-mod_pkg_setup
#	epatch "${FILESDIR}"/virtualbox-modules-1.5.6-2.6.25.patch
	BUILD_PARAMS="KERN_DIR=${KV_DIR} KERNOUT=${KV_OUT_DIR}"
	enewgroup vboxusers
}

src_unpack() {
        unpack ${A}
        cd "${S}"
        if kernel_is 2 6 25 ; then
                epatch "${FILESDIR}"/virtualbox-modules-1.5.6-2.6.25.patch
        fi
}

src_install() {
	linux-mod_src_install

	# udev rule for vboxdrv
	dodir /etc/udev/rules.d
	echo 'KERNEL=="vboxdrv", GROUP="vboxusers" MODE=660' >> "${D}/etc/udev/rules.d/60-virtualbox.rules"
}

pkg_postinst() {
	linux-mod_pkg_postinst
}

[kernelOfTruth]

lm_sensors fail with 2.6.25-rc8-zen0 / .25 kernels

In file included from kernel/include/i2c-dev.h:27,
from prog/dump/i2cdump.c:29:
/usr/src/linux/include/linux/types.h:203: error: expected specifier-qualifier-list before ‘__kernel_daddr_t’
make: *** [prog/dump/i2cdump.ro] Error 1
make: *** Waiting for unfinished jobs....
In file included from kernel/include/i2c-dev.h:27,
from prog/dump/i2cbusses.c:33:
/usr/src/linux/include/linux/types.h:203: error: expected specifier-qualifier-list before ‘__kernel_daddr_t’
make: *** [prog/dump/i2cbusses.ro] Error 1
In file included from kernel/include/i2c-dev.h:27,
from prog/detect/i2cdetect.c:28:
/usr/src/linux/include/linux/types.h:203: error: expected specifier-qualifier-list before ‘__kernel_daddr_t’
make: *** [prog/detect/i2cdetect.ro] Error 1
*
* ERROR: sys-apps/lm_sensors-2.10.4 failed.
* Call stack:
* ebuild.sh, line 49: Called src_compile
* environment, line 3237: Called die
* The specific snippet of code:
* emake CC=$(tc-getCC) LINUX=${KV_DIR} I2C_HEADERS=${KV_DIR}/include user || die "emake user failed"
* The die message:
* emake user failed
*
* If you need support, post the topmost build error, and the call stack if relevant.
* A complete build log is located at '/var/log/portage/sys-apps:lm_sensors-2.10.4:20080404-110017.log'.
* The ebuild environment file is located at '/var/tmp/portage/sys-apps/lm_sensors-2.10.4/temp/environment'.
*

[kernelOfTruth]

WARNING !!

.25 kernel-series "produces" much higher temperatures than .24 kernels:

Sys Temp: +41°C (high = +123°C, hyst = -65°C) [thermistor]
CPU Temp: +39.5°C (high = +80.0°C, hyst = +75.0°C) [CPU diode ]
AUX Temp: +97.5°C (high = +80.0°C, hyst = +75.0°C) [thermistor] ALARM
vid: +1.338 V

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +73°C (high = +100°C)

coretemp-isa-0001
Adapter: ISA adapter
Core 1: +71°C (high = +100°C)

<-- I never got that high core temperatures, they were always at 56°C (at maximum)

anyone encountered the same ?

processor model: Intel Core 2 Duo 6600

[gimpel]

Code: Select all

k8temp-pci-00c3
Adapter: PCI adapter
Core0 Temp:
             +36°C
Core1 Temp:
             +29°C

Looks fine here, AMD Athlon64 X2

[kernelOfTruth]

Code: Select all

k8temp-pci-00c3
Adapter: PCI adapter
Core0 Temp:
             +36°C
Core1 Temp:
             +29°C

Looks fine here, AMD Athlon64 X2

ok, good news for amd users - bad news for intel users (me )

aren't amd processors famous for low cpu temperatures ?

[darklegion]

WARNING !!

.25 kernel-series "produces" much higher temperatures than .24 kernels:

Sys Temp: +41°C (high = +123°C, hyst = -65°C) [thermistor]
CPU Temp: +39.5°C (high = +80.0°C, hyst = +75.0°C) [CPU diode ]
AUX Temp: +97.5°C (high = +80.0°C, hyst = +75.0°C) [thermistor] ALARM
vid: +1.338 V

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +73°C (high = +100°C)

coretemp-isa-0001
Adapter: ISA adapter
Core 1: +71°C (high = +100°C)

<-- I never got that high core temperatures, they were always at 56°C (at maximum)

anyone encountered the same ?

processor model: Intel Core 2 Duo 6600

Yeah, you're right.I'm getting 78.0C with both cores with a E1260-2.88GHZ.This is with gimps, though, which really pushes the temps but it shouldn't be this high.Could be just the driver misreporting the temps, though.Tommorrow, I'll check and see what temps I get with 2.6.24.

[Undetected]

Also saw much higher core temps (~67C under heavily compilation) with the .25 branch. I switched back to .24, much cooler now (35C on both cores). I'm running an E6600.

[broch]

laptop core duo T2500
no change in cpu tem between kernels (at all)

[FallenWizard]

Code: Select all

k8temp-pci-00c3
Adapter: PCI adapter
Core0 Temp:
             +36°C
Core1 Temp:
             +29°C

Looks fine here, AMD Athlon64 X2

ok, good news for amd users - bad news for intel users (me )

aren't amd processors famous for low cpu temperatures ?

Code: Select all

coretemp-isa-0000
Adapter: ISA adapter
Core 0:      +49.0°C  (high = +78.0°C, crit = +100.0°C)  

coretemp-isa-0001
Adapter: ISA adapter
Core 1:      +50.0°C  (high = +78.0°C, crit = +100.0°C) 

With Core 2 Duo E6320@1,86 ghz. (I am using a better cooler though.)


I have no problem with temperatures anyway. And I am not using a AMD CPU.

[darklegion]

Code: Select all

k8temp-pci-00c3
Adapter: PCI adapter
Core0 Temp:
             +36°C
Core1 Temp:
             +29°C

Looks fine here, AMD Athlon64 X2

ok, good news for amd users - bad news for intel users (me )

aren't amd processors famous for low cpu temperatures ?

Code: Select all

coretemp-isa-0000
Adapter: ISA adapter
Core 0:      +49.0°C  (high = +78.0°C, crit = +100.0°C)  

coretemp-isa-0001
Adapter: ISA adapter
Core 1:      +50.0°C  (high = +78.0°C, crit = +100.0°C) 

With Core 2 Duo E6320@1,86 ghz. (I am using a better cooler though.)


I have no problem with temperatures anyway. And I am not using a AMD CPU.

Under load, or idle?

[FallenWizard]

Under load, or idle?

Under idle.

When I'm compiling glibc:

coretemp-isa-0000
Adapter: ISA adapter
Core 0: +62.0°C (high = +78.0°C, crit = +100.0°C)

coretemp-isa-0001
Adapter: ISA adapter
Core 1: +63.0°C (high = +78.0°C, crit = +100.0°C)

Maybe something is really messed up

[kernelOfTruth]

I think something might have gone into rc8-git3 or so. Check out git-sources.

it seems to be fixed in newer git

please pull !

see

thanks