Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Zen Kernel Discussion/Support Thread - Part 2
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3 ... , 12, 13, 14  Next  
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Unsupported Software
View previous topic :: View next topic  
Author Message
fanthom
n00b
n00b


Joined: 31 May 2007
Posts: 60

PostPosted: Sun Jan 31, 2010 5:58 pm    Post subject: Reply with quote

"simple i/o scheduler" - sounds like a perfect solution for live-usb distros (slax). will try it for sure :)
thanks
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Sun Jan 31, 2010 6:13 pm    Post subject: Reply with quote

thanks for bumping to 2.6.33-rc6-zen1, dodo ! :)

using compress-force lead to space-savings of 43 GB !

df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/btrfs 698G 616G 83G 89% /bak (with compress mount-option)
/dev/mapper/btrfs 698G 573G 125G 83% /bak (with compress-force mount-option)
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
cheater1034
Veteran
Veteran


Joined: 09 Sep 2004
Posts: 1558

PostPosted: Sun Jan 31, 2010 10:14 pm    Post subject: Reply with quote

albright wrote:
Quote:
i've noticed ureadahead was included in the last release of zen sources ( Linux Kernel v2.6.32-zen6 "Forever and never" ).
How do i enable it?
Do i need a some kind daemon ot load it at startup?


I would like to know the answer to these questions as well - there is no
ureadahead ebuild at the moment it seems


You need to compile userspace and have an init script to use it.
https://launchpad.net/ureadahead

I'd strongly suggest opening a bug on bugs.gentoo.org
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.
Back to top
View user's profile Send private message
lordcris
Apprentice
Apprentice


Joined: 09 Jul 2002
Posts: 248

PostPosted: Mon Feb 01, 2010 12:13 am    Post subject: Reply with quote

cheater1034 wrote:
albright wrote:
Quote:
i've noticed ureadahead was included in the last release of zen sources ( Linux Kernel v2.6.32-zen6 "Forever and never" ).
How do i enable it?
Do i need a some kind daemon ot load it at startup?


I would like to know the answer to these questions as well - there is no
ureadahead ebuild at the moment it seems


You need to compile userspace and have an init script to use it.
https://launchpad.net/ureadahead

I'd strongly suggest opening a bug on bugs.gentoo.org


What exactly do you mean by "userspace"?
Is it "FUSE_FS"?
Back to top
View user's profile Send private message
cheater1034
Veteran
Veteran


Joined: 09 Sep 2004
Posts: 1558

PostPosted: Mon Feb 01, 2010 1:28 am    Post subject: Reply with quote

lordcris wrote:
cheater1034 wrote:
albright wrote:
Quote:
i've noticed ureadahead was included in the last release of zen sources ( Linux Kernel v2.6.32-zen6 "Forever and never" ).
How do i enable it?
Do i need a some kind daemon ot load it at startup?


I would like to know the answer to these questions as well - there is no
ureadahead ebuild at the moment it seems


You need to compile userspace and have an init script to use it.
https://launchpad.net/ureadahead

I'd strongly suggest opening a bug on bugs.gentoo.org


What exactly do you mean by "userspace"?
Is it "FUSE_FS"?


No, i mean you need to compile ureadahead stuff outside of the kernel too, in the userspace (but you also need in-kernel stuff)
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.
Back to top
View user's profile Send private message
mantoo
Tux's lil' helper
Tux's lil' helper


Joined: 08 Dec 2007
Posts: 98
Location: Germany

PostPosted: Mon Feb 01, 2010 11:39 am    Post subject: Reply with quote

lordcris wrote:
cheater1034 wrote:
albright wrote:
Quote:
i've noticed ureadahead was included in the last release of zen sources ( Linux Kernel v2.6.32-zen6 "Forever and never" ).
How do i enable it?
Do i need a some kind daemon ot load it at startup?


I would like to know the answer to these questions as well - there is no
ureadahead ebuild at the moment it seems


You need to compile userspace and have an init script to use it.
https://launchpad.net/ureadahead

I'd strongly suggest opening a bug on bugs.gentoo.org


What exactly do you mean by "userspace"?
Is it "FUSE_FS"?


found your bug https://bugs.gentoo.org/show_bug.cgi?id=303044

I tried to fiddle out sth. usable and runnable... result as follows:

dev-libs, libnih-1.0.0.ebuild (without ureadahead didn´t compile)
Code:

# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $

inherit eutils

DESCRIPTION="Library libnih as dependency for ureadahead"
HOMEPAGE="https://launchpad.net/libnih"

SRC_URI="http://launchpad.net/libnih/1.0/1.0.0/+download/${P}.tar.gz"

LICENSE="GPL-2"

SLOT="0"

KEYWORDS="~x86 ~arch"

RDEPEND="${DEPEND}"

src_install() {
        emake DESTDIR="${D}" install || die "emake install failed"
}


and sys-apps, ureadahead-0.100.0.ebuild:
Code:

# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $

inherit eutils

DESCRIPTION="Ureadahead"
HOMEPAGE="https://launchpad.net/ureadahead"

SRC_URI="http://launchpad.net/ureadahead/trunk/0.100.0/+download/${P}.tar.gz"

LICENSE="GPL-2"

SLOT="0"

KEYWORDS="~x86 ~arch"

DEPEND="dev-libs/libnih"
RDEPEND="${DEPEND}"

src_install() {
        emake DESTDIR="${D}" install || die "emake install failed"
}



as init-script i tried to reuse an old readahead-init-script (https://bugs.gentoo.org/attachment.cgi?id=168928) and had a look into ddclient's init-script:
Code:

#!/sbin/runscript
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/net-dns/ddclient/files/ddclient.initd,v 1.2 2007/08/08 03:17:32 uberlord Exp $

# Default location if not specified in /etc/conf.d/ddclient
PIDFILE=${PIDFILE:-/var/run/ureadahead/ureadahead.pid}

depend() {
                need localmount
                # this should start as early as possible
                # we can't do 'before *' as that breaks it
                before bootmisc consolefont keymaps rmnologin serial urandom
}

start() {
        ebegin "Starting ${SVCNAME}"
        start-stop-daemon \
                --start \
                --chuid root \
                --exec /usr/sbin/ureadahead \
                --name ureadahead \
                --pidfile "${PIDFILE}"
        eend $?
}

stop() {
        ebegin "Stopping ${SVCNAME}"
        start-stop-daemon \
                --stop \
                --signal USR1 \
                --pidfile "${PIDFILE}"
        eend $?
}


as far as i can tell at least the two packages do compile an install on my box, ureadahead itself always complains about a file not found (/var/lib/ureadahead/pack if I remember this one correctly).
These are my 1st ebuilds i gave a try; tbh i don´t know much about ebuilds and init-scripts, so pls no blaming for my lousy and n00bish approaches :-)
Back to top
View user's profile Send private message
Bluesquare
n00b
n00b


Joined: 26 Oct 2009
Posts: 24

PostPosted: Tue Feb 02, 2010 6:28 am    Post subject: Reply with quote

You would also need to setup your kernel as per:

Quote:

Just compiled 2.6.32 from kernel.org and got ureadahead working. Ureadahead's sources contain a patch (0001-trace-add-trace-events-for-open-exec-an.patch). Apply it and make sure you have

* Kernel hacking -> Tracers (FTRACE) and

* Kernel hacking -> Tracers -> Trace process context switches and events (ENABLE_DEFAULT_TRACERS)

enabled (this should also select 'Kernel hacking -> Debug Filesystem' (DEBUG_FS)). I'm not entirely sure this is everything you need so report back if that works for you.


https://answers.launchpad.net/ureadahead/+question/91100
Back to top
View user's profile Send private message
tgR10
Apprentice
Apprentice


Joined: 23 Oct 2007
Posts: 262
Location: caly ten ambaras

PostPosted: Tue Feb 02, 2010 7:30 am    Post subject: Reply with quote

@mantoo
i fixed your ebuild and init script a lil:

sys-apps/ureadahead/ureadahead-0.100.0.ebuild

Code:
# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $

inherit eutils

DESCRIPTION="Ureadahead"
HOMEPAGE="https://launchpad.net/ureadahead"

SRC_URI="http://launchpad.net/ureadahead/trunk/0.100.0/+download/${P}.tar.gz"

LICENSE="GPL-2"

SLOT="0"

KEYWORDS="~x86 ~arch"

DEPEND="dev-libs/libnih"
RDEPEND="${DEPEND}"

src_install() {
        emake DESTDIR="${D}" install || die "emake install failed"
   newinitd "${FILESDIR}"/ureadahead.rc ureadahead || die
   dodir /var/lib/ureadahead
   dodir /var/run/ureadahead
}

pkg_postinst() {
   elog "ureadahed to work need some kernel tuning:"
   elog "Kernel hacking -> Tracers (FTRACE)"
   elog "Kernel hacking -> Tracers -> Trace process context switches and events (ENABLE_DEFAULT_TRACERS)"
   elog "which should also select 'Kernel hacking -> Debug Filesystem' (DEBUG_FS))"
}


sys-apps/ureadahead/files/ureadahead.rc
Code:
#!/sbin/runscript
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/net-dns/ddclient/files/ddclient.initd,v 1.2 2007/08/08 03:17:32 uberlord Exp $

# Default location if not specified in /etc/conf.d/ddclient
PIDFILE=${PIDFILE:-/var/run/ureadahead/ureadahead.pid}

depend() {
                need localmount
                # this should start as early as possible
                # we can't do 'before *' as that breaks it
                before bootmisc consolefont keymaps rmnologin serial urandom
}

_abort() {
   rc-abort
   return 1
}

start() {
        ebegin "Starting ${SVCNAME}"
        start-stop-daemon \
                --start \
                --chuid root \
                --exec /usr/sbin/ureadahead \
                --name ureadahead \
                --pidfile "${PIDFILE}"
        eend $?
}

stop() {
        ebegin "Stopping ${SVCNAME}"
   _abort() { return 0; }   
        eend $?
}


funny thing is that i got 2-3 sec longer boot with this ureadahead ...
and before first reboot i had to run it manualy so it could create file list so i woudn't have boot issue
_________________
"bo kto ma racje ? ten kto z bliska zobaczy"
"moge nie wiedziec,wchlaniam niewiedze z malych torebek"
http://i12.tinypic.com/4pow0mu.png
http://userbar.tgr.debil.eu/userbar.jpg
Back to top
View user's profile Send private message
haarp
Guru
Guru


Joined: 31 Oct 2007
Posts: 517

PostPosted: Tue Feb 02, 2010 8:15 am    Post subject: Reply with quote

Could you please add those ebuilds to the bug aswell?
Back to top
View user's profile Send private message
Bluesquare
n00b
n00b


Joined: 26 Oct 2009
Posts: 24

PostPosted: Tue Feb 02, 2010 11:07 am    Post subject: Reply with quote

Typo in the ebuild. Only cosmetics but anyway.

I'm not a native speaker myself so i might be mistaking.

Anyway, thanks.

I don't know if this is off-topic. Perhaps moderators should split the ureadahead discussion off from the zen-sources thread.

Edit: bloody tags doesn't work inside quotes. It's down there anyhow...

tgR10 wrote:
@mantoo
i fixed your ebuild and init script a lil:

sys-apps/ureadahead/ureadahead-0.100.0.ebuild

Code:
# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $

inherit eutils

DESCRIPTION="Ureadahead"
HOMEPAGE="https://launchpad.net/ureadahead"

SRC_URI="http://launchpad.net/ureadahead/trunk/0.100.0/+download/${P}.tar.gz"

LICENSE="GPL-2"

SLOT="0"

KEYWORDS="~x86 ~arch"

DEPEND="dev-libs/libnih"
RDEPEND="${DEPEND}"

src_install() {
        emake DESTDIR="${D}" install || die "emake install failed"
   newinitd "${FILESDIR}"/ureadahead.rc ureadahead || die
   dodir /var/lib/ureadahead
   dodir /var/run/ureadahead
}

pkg_postinst() {
   [B]elog "ureadahed to work need some kernel tuning:"[/B] <--- "ureadahead need some kernel tuning to work:
   elog "Kernel hacking -> Tracers (FTRACE)"
   elog "Kernel hacking -> Tracers -> Trace process context switches and events (ENABLE_DEFAULT_TRACERS)"
   elog "which should also select 'Kernel hacking -> Debug Filesystem' (DEBUG_FS))"
}


sys-apps/ureadahead/files/ureadahead.rc
Code:
#!/sbin/runscript
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/net-dns/ddclient/files/ddclient.initd,v 1.2 2007/08/08 03:17:32 uberlord Exp $

# Default location if not specified in /etc/conf.d/ddclient
PIDFILE=${PIDFILE:-/var/run/ureadahead/ureadahead.pid}

depend() {
                need localmount
                # this should start as early as possible
                # we can't do 'before *' as that breaks it
                before bootmisc consolefont keymaps rmnologin serial urandom
}

_abort() {
   rc-abort
   return 1
}

start() {
        ebegin "Starting ${SVCNAME}"
        start-stop-daemon \
                --start \
                --chuid root \
                --exec /usr/sbin/ureadahead \
                --name ureadahead \
                --pidfile "${PIDFILE}"
        eend $?
}

stop() {
        ebegin "Stopping ${SVCNAME}"
   _abort() { return 0; }   
        eend $?
}


funny thing is that i got 2-3 sec longer boot with this ureadahead ...
and before first reboot i had to run it manualy so it could create file list so i woudn't have boot issue
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Tue Feb 02, 2010 1:03 pm    Post subject: Reply with quote

good news:

reiser4 most likely isn't broken for a long time anymore:

http://old.nabble.com/-patch-0-7--per-bdi-flushing-model-improvements.-reiser4-td27414380.html :)
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
lordcris
Apprentice
Apprentice


Joined: 09 Jul 2002
Posts: 248

PostPosted: Tue Feb 02, 2010 3:23 pm    Post subject: Reply with quote

i've put the two ebuild provided by mantoo and tgR10 on https://bugs.gentoo.org/show_bug.cgi?id=303044
i don't see any speed improvement in my system.
also when i give the command

Code:
rc-status boot


i get


Code:
Runlevel: boot
 hwclock                                                           [  started  ]
 modules                                                           [  started  ]
 device-mapper                                                     [  started  ]
 fsck                                                              [  started  ]
 root                                                              [  started  ]
 mtab                                                              [  started  ]
 hibernate-cleanup                                                 [  started  ]
 localmount                                                        [  started  ]
 hostname                                                          [  started  ]
 sysctl                                                            [  started  ]
 ureadahead                                                        [  crashed  ]
 bootmisc                                                          [  started  ]
 net.lo                                                            [  started  ]
 dnsmasq                                                           [  started  ]
 procfs                                                            [  started  ]
 termencoding                                                      [  started  ]
 audiofase                                                         [  started  ]
 keymaps                                                           [  started  ]
 urandom                                                           [  started  ]
 swap                                                              [  started  ]
 consolefont                                                       [  started  ]


ureadahead [ crashed ]

is there a method to trace ureadahead behavior?
Back to top
View user's profile Send private message
tgR10
Apprentice
Apprentice


Joined: 23 Oct 2007
Posts: 262
Location: caly ten ambaras

PostPosted: Tue Feb 02, 2010 3:36 pm    Post subject: Reply with quote

Bluesquare wrote:
Typo in the ebuild. Only cosmetics but anyway.
[...]
elog "ureadahed to work need some kernel tuning:" <--- "ureadahead need some kernel tuning to work:

i think both sentence are correct :)

lordcris wrote:
[...]
is there a method to trace ureadahead behavior?


what i noticed is that ureadahead load files list on boot up, then load it to memory and that was it in my case
that's why on stop section in init script there is nothing to do because ureadahead shut down itself(?) after loading that files into memory, but i may be wrong
lordcris wrote:
[...]
i don't see any speed improvement in my system.
[...]

same here .. mine got even slower by 2-3 seconds
Bluesquare wrote:

I don't know if this is off-topic. Perhaps moderators should split the ureadahead discussion off from the zen-sources thread.

this may be a good idea, but afaik that trace kernel patch is included in zen-sources not in gentoo-sources, but still thread could be splited and put into unsuported software
_________________
"bo kto ma racje ? ten kto z bliska zobaczy"
"moge nie wiedziec,wchlaniam niewiedze z malych torebek"
http://i12.tinypic.com/4pow0mu.png
http://userbar.tgr.debil.eu/userbar.jpg
Back to top
View user's profile Send private message
mantoo
Tux's lil' helper
Tux's lil' helper


Joined: 08 Dec 2007
Posts: 98
Location: Germany

PostPosted: Tue Feb 02, 2010 5:59 pm    Post subject: Reply with quote

tgR10 wrote:
@mantoo
i fixed your ebuild and init script a lil:

thanks for fixing/tweaking/tuning :-)

lordcris wrote:
i've put the two ebuild provided by mantoo and tgR10 on https://bugs.gentoo.org/show_bug.cgi?id=303044
i don't see any speed improvement in my system.
is there a method to trace ureadahead behavior?

and as well thank you for adding ebuilds to the opened bugs; would have done that myself if i had a login :-)

tgR10 wrote:
lordcris wrote:
[...]
i don't see any speed improvement in my system.
[...]

same here .. mine got even slower by 2-3 seconds

added tracing and the stuff needed to my kernel-conf, compiled and booted;
1st time booting did hang while/after starting ureadahead;
2nd time booting was exactly 3 secs slower than w/o ureadahead - at least /var/lib/ureadahead/pack isn't empty any more;
3rd time booting now is 4 secs slower than w/o ureadahead.

ureadahead listed 'crashed' in "rc-status boot" here as well

strange... ?!?
Back to top
View user's profile Send private message
haarp
Guru
Guru


Joined: 31 Oct 2007
Posts: 517

PostPosted: Tue Feb 02, 2010 6:26 pm    Post subject: Reply with quote

Why not just try to run ureadahead manually and see what it does? It probably terminates after preloading everything, thus the init system regards it as "crashed"

Last edited by haarp on Wed Feb 03, 2010 8:24 am; edited 1 time in total
Back to top
View user's profile Send private message
niick
Tux's lil' helper
Tux's lil' helper


Joined: 09 Mar 2006
Posts: 88

PostPosted: Tue Feb 02, 2010 7:22 pm    Post subject: Reply with quote

I've got a problem using nouveau in the latest kernel (2.6.33-rc6). When I boot there is a very long pause, ~1min, after grub and before the init scripts start.
I assume this is when the kernel is initializing. All I see when this happening is a blinking cursor. I've tried a kernel with the exact same config except
with no nouveau/DRM and using the VESA driver instead and it boots without a pause.

After the pause everything (inc. nouveau) boots and runs great.

Any advice about a fix or how to find out what is causing the pause is much appreciated.
Back to top
View user's profile Send private message
mantoo
Tux's lil' helper
Tux's lil' helper


Joined: 08 Dec 2007
Posts: 98
Location: Germany

PostPosted: Wed Feb 03, 2010 8:21 am    Post subject: Reply with quote

The 2-3 seconds slowdown disappeard after extendung the init-script by
Code:

start() {
        ebegin "Starting ${SVCNAME}"
        start-stop-daemon \
                --start \
                --chuid root \
                --exec /usr/sbin/ureadahead \
                --name ureadahead \
                --pidfile "${PIDFILE}\
         ==>    -- --daemon"
        eend $?
}


though it´s listed as "crashed" in rc-status...
Back to top
View user's profile Send private message
Waninkoko
Guru
Guru


Joined: 13 May 2005
Posts: 549

PostPosted: Wed Feb 03, 2010 5:33 pm    Post subject: Reply with quote

kernelOfTruth wrote:
good news:

reiser4 most likely isn't broken for a long time anymore:

http://old.nabble.com/-patch-0-7--per-bdi-flushing-model-improvements.-reiser4-td27414380.html :)


I just applied those patches to zen.git
Back to top
View user's profile Send private message
Tyrius
n00b
n00b


Joined: 27 Aug 2004
Posts: 46
Location: Southampton, UK

PostPosted: Wed Feb 03, 2010 6:42 pm    Post subject: Reply with quote

Have done some experiments with ureadahead and have changed the init script which seems to make it work correctly. It has made my bootup sequence around 3-5 seconds faster, in my case from 35 secs to 30 secs (That is from BIOS to GDM login screen).

Here is my /etc/init.d/ureadahead
Code:

#!/sbin/runscript
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2

depend() {
   need localmount
   # this should start as early as possible
   # we can't do 'before *' as that breaks it
   before bootmisc consolefont keymaps rmnologin serial urandom
}

start() {
   ebegin "Starting ${SVCNAME}"
   /usr/sbin/ureadahead --daemon --timeout 30
   eend $?
}


I have not used start-stop-daemon as it doesn't run like a standard daemon, it starts does it thing and then exits. The timeout parameter is needed when it is building the pack files, as it won't exit otherwise. The daemon command places it into the background so the rest of the boot can continue, otherwise the boot waits until the timeout happens.

I hope this helps others who are trying to get this to work.
Back to top
View user's profile Send private message
mantoo
Tux's lil' helper
Tux's lil' helper


Joined: 08 Dec 2007
Posts: 98
Location: Germany

PostPosted: Wed Feb 03, 2010 7:56 pm    Post subject: Reply with quote

Tyrius wrote:
Have done some experiments with ureadahead and have changed the init script which seems to make it work correctly. It has made my bootup sequence around 3-5 seconds faster, in my case from 35 secs to 30 secs (That is from BIOS to GDM login screen).

Here is my /etc/init.d/ureadahead
Code:

#!/sbin/runscript
# Copyright 1999-2006 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2

depend() {
   need localmount
   # this should start as early as possible
   # we can't do 'before *' as that breaks it
   before bootmisc consolefont keymaps rmnologin serial urandom
}

start() {
   ebegin "Starting ${SVCNAME}"
   /usr/sbin/ureadahead --daemon --timeout 30
   eend $?
}


I have not used start-stop-daemon as it doesn't run like a standard daemon, it starts does it thing and then exits. The timeout parameter is needed when it is building the pack files, as it won't exit otherwise. The daemon command places it into the background so the rest of the boot can continue, otherwise the boot waits until the timeout happens.

I hope this helps others who are trying to get this to work.


mmhm... tried that, but as i have just ssh access atm cannot tell if there´s an error msg on bootup. my 19.7 secs until login-prompt are not affected, so i don´t know if it works as it´s proposed to.

i´ll wait for others giving it a try as i cannot afford time for further investigation right now

thanks anyway :-)
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Wed Feb 03, 2010 9:43 pm    Post subject: Reply with quote

Waninkoko wrote:
kernelOfTruth wrote:
good news:

reiser4 most likely isn't broken for a long time anymore:

http://old.nabble.com/-patch-0-7--per-bdi-flushing-model-improvements.-reiser4-td27414380.html :)


I just applied those patches to zen.git


thank you a lot Waninkoko ! :D

I'll probably give it a test run in the weekend
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Thu Feb 04, 2010 10:47 pm    Post subject: Reply with quote

it works better so far - I'll try to emerge openoffice with it tomorrow

with default plugin-profile it still has problems with syncing / umounting (afaik it's a known bug on the reiserfs-mailing list):
Quote:
[ 4095.298379] reiser4: sdb6: found disk format 4.0.0.
[ 4353.446627] ------------[ cut here ]------------
[ 4353.446632] kernel BUG at fs/reiser4/plugin/item/extent_flush_ops.c:551!
[ 4353.446635] invalid opcode: 0000 [#1] PREEMPT SMP
[ 4353.446638] last sysfs file: /sys/devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:12/ATK0110:00/hwmon/hwmon0/temp2_input
[ 4353.446640] CPU 0
[ 4353.446643] Pid: 22916, comm: umount Tainted: P 2.6.33-rc6-zen1_r4 #3 P5W DH Deluxe/P5W DH Deluxe
[ 4353.446645] RIP: 0010:[<ffffffff811af822>] [<ffffffff811af822>] assign_real_blocknrs+0x132/0x150
[ 4353.446652] RSP: 0018:ffff8800a48dfaf8 EFLAGS: 00010246
[ 4353.446654] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff8800a48dffd8
[ 4353.446656] RDX: 0000000000001c0e RSI: 0000000000071c0e RDI: ffff8801af793020
[ 4353.446657] RBP: ffff880109c09190 R08: 0000000000000001 R09: 00000000000004c0
[ 4353.446659] R10: dead000000200200 R11: dead000000100100 R12: 0000000000000005
[ 4353.446661] R13: 0000000000000000 R14: 000000000004839d R15: ffff8801af793020
[ 4353.446663] FS: 00007fbac0d95740(0000) GS:ffff880028200000(0000) knlGS:0000000000000000
[ 4353.446665] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 4353.446667] CR2: 00007f2118253000 CR3: 00000000a4824000 CR4: 00000000000006f0
[ 4353.446669] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 4353.446670] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[ 4353.446673] Process umount (pid: 22916, threadinfo ffff8800a48de000, task ffff8801ae0ac180)
[ 4353.446674] Stack:
[ 4353.446675] 0000000000000010 ffff880109c09200 ffff8800a48dfbd0 0000000000071c0e
[ 4353.446678] <0> 000000000004839d ffff8801aad2d000 ffff880109c09198 ffff8800a48dfbb8
[ 4353.446681] <0> ffff880109c09190 ffff880109c09198 ffff880109c09200 0000000000000001
[ 4353.446685] Call Trace:
[ 4353.446689] [<ffffffff811b0414>] ? reiser4_alloc_extent+0x1b4/0x510
[ 4353.446692] [<ffffffff81172486>] ? move_lh_internal+0xc6/0x1d0
[ 4353.446695] [<ffffffff8117de06>] ? handle_pos_to_twig+0x66/0x140
[ 4353.446698] [<ffffffff8117dfdd>] ? handle_pos_on_twig+0xfd/0x280
[ 4353.446701] [<ffffffff8117ff0f>] ? flush_current_atom+0x73f/0x830
[ 4353.446704] [<ffffffff8117a0fd>] ? reiser4_txn_end+0x36d/0x660
[ 4353.446707] [<ffffffff81179f34>] ? reiser4_txn_end+0x1a4/0x660
[ 4353.446709] [<ffffffff8117a3f9>] ? reiser4_txn_restart+0x9/0x20
[ 4353.446712] [<ffffffff8117acbb>] ? force_commit_atom+0x4b/0x60
[ 4353.446715] [<ffffffff8117ae9c>] ? txnmgr_force_commit_all+0x13c/0x180
[ 4353.446717] [<ffffffff811b33e6>] ? release_format40+0x66/0x110
[ 4353.446720] [<ffffffff81175bf1>] ? reiser4_init_context+0x51/0x90
[ 4353.446723] [<ffffffff81186a59>] ? reiser4_put_super+0x49/0xf0
[ 4353.446726] [<ffffffff810d1a8f>] ? generic_shutdown_super+0x4f/0x100
[ 4353.446728] [<ffffffff810d1b68>] ? kill_block_super+0x28/0x50
[ 4353.446731] [<ffffffff810d2a0e>] ? deactivate_super+0x6e/0xb0
[ 4353.446734] [<ffffffff810e9c20>] ? sys_umount+0x70/0x3b0
[ 4353.446737] [<ffffffff8100276b>] ? system_call_fastpath+0x16/0x1b
[ 4353.446739] Code: ff ff ff 48 8b 54 24 28 8b 02 89 c2 c1 fa 08 38 c2 74 24 48 8b 7c 24 28 48 83 c4 38 5b 5d 41 5c 41 5d 41 5e 41 5f e9 8e 2d 52 00 <0f> 0b eb fe 0f 0b eb fe 0f 0b eb fe 0f 0b eb fe 66 66 66 66 66
[ 4353.446766] RIP [<ffffffff811af822>] assign_real_blocknrs+0x132/0x150
[ 4353.446769] RSP <ffff8800a48dfaf8>
[ 4353.446781] ---[ end trace 692902ea4ec9325b ]---
[ 4353.446783] ------------[ cut here ]------------
[ 4353.446786] WARNING: at kernel/exit.c:899 do_exit+0x762/0x780()
[ 4353.446787] Hardware name: P5W DH Deluxe
[ 4353.446788] Modules linked in: xt_owner xt_iprange xt_CONNMARK fglrx(P) vboxnetadp vboxnetflt vboxdrv btrfs hwmon_vid coretemp rtl8187 mac80211 asus_atk0110 cfg80211 rfkill hwmon sky2 i82975x_edac led_class i2c_i801 eeprom_93cx6 floppy tg3 libphy e1000 scsi_wait_scan sl811_hcd usb_storage [last unloaded: firewire_core]
[ 4353.446807] Pid: 22916, comm: umount Tainted: P D 2.6.33-rc6-zen1_r4 #3
[ 4353.446809] Call Trace:
[ 4353.446811] [<ffffffff81039d73>] ? warn_slowpath_common+0x73/0xb0
[ 4353.446814] [<ffffffff8103d0c2>] ? do_exit+0x762/0x780
[ 4353.446818] [<ffffffff816cf65f>] ? printk+0x40/0x45
[ 4353.446820] [<ffffffff8103b29c>] ? kmsg_dump+0x7c/0x150
[ 4353.446823] [<ffffffff81006eb2>] ? oops_end+0x72/0xa0
[ 4353.446826] [<ffffffff81004114>] ? do_invalid_op+0x84/0xa0
[ 4353.446828] [<ffffffff811af822>] ? assign_real_blocknrs+0x132/0x150
[ 4353.446831] [<ffffffff811b2e21>] ? T.642+0x221/0x230
[ 4353.446833] [<ffffffff81003315>] ? invalid_op+0x15/0x20
[ 4353.446836] [<ffffffff811af822>] ? assign_real_blocknrs+0x132/0x150
[ 4353.446839] [<ffffffff811af77d>] ? assign_real_blocknrs+0x8d/0x150
[ 4353.446842] [<ffffffff811b0414>] ? reiser4_alloc_extent+0x1b4/0x510
[ 4353.446844] [<ffffffff81172486>] ? move_lh_internal+0xc6/0x1d0
[ 4353.446847] [<ffffffff8117de06>] ? handle_pos_to_twig+0x66/0x140
[ 4353.446850] [<ffffffff8117dfdd>] ? handle_pos_on_twig+0xfd/0x280
[ 4353.446853] [<ffffffff8117ff0f>] ? flush_current_atom+0x73f/0x830
[ 4353.446856] [<ffffffff8117a0fd>] ? reiser4_txn_end+0x36d/0x660
[ 4353.446858] [<ffffffff81179f34>] ? reiser4_txn_end+0x1a4/0x660
[ 4353.446861] [<ffffffff8117a3f9>] ? reiser4_txn_restart+0x9/0x20
[ 4353.446864] [<ffffffff8117acbb>] ? force_commit_atom+0x4b/0x60
[ 4353.446866] [<ffffffff8117ae9c>] ? txnmgr_force_commit_all+0x13c/0x180
[ 4353.446869] [<ffffffff811b33e6>] ? release_format40+0x66/0x110
[ 4353.446871] [<ffffffff81175bf1>] ? reiser4_init_context+0x51/0x90
[ 4353.446874] [<ffffffff81186a59>] ? reiser4_put_super+0x49/0xf0
[ 4353.446877] [<ffffffff810d1a8f>] ? generic_shutdown_super+0x4f/0x100
[ 4353.446889] [<ffffffff810d1b68>] ? kill_block_super+0x28/0x50
[ 4353.446891] [<ffffffff810d2a0e>] ? deactivate_super+0x6e/0xb0
[ 4353.446894] [<ffffffff810e9c20>] ? sys_umount+0x70/0x3b0
[ 4353.446896] [<ffffffff8100276b>] ? system_call_fastpath+0x16/0x1b
[ 4353.446898] ---[ end trace 692902ea4ec9325c ]---
[ 4353.446900] note: umount[22916] exited with preempt_count 1
[ 4353.446914] BUG: scheduling while atomic: umount/22916/0x10000002
[ 4353.446915] Modules linked in: xt_owner xt_iprange xt_CONNMARK fglrx(P) vboxnetadp vboxnetflt vboxdrv btrfs hwmon_vid coretemp rtl8187 mac80211 asus_atk0110 cfg80211 rfkill hwmon sky2 i82975x_edac led_class i2c_i801 eeprom_93cx6 floppy tg3 libphy e1000 scsi_wait_scan sl811_hcd usb_storage [last unloaded: firewire_core]
[ 4353.446932] Pid: 22916, comm: umount Tainted: P D W 2.6.33-rc6-zen1_r4 #3
[ 4353.446934] Call Trace:
[ 4353.446936] [<ffffffff816d03d6>] ? schedule+0x816/0x940
[ 4353.446939] [<ffffffff81035ee6>] ? __cond_resched+0x26/0x30
[ 4353.446942] [<ffffffff816d0645>] ? _cond_resched+0x35/0x40
[ 4353.446945] [<ffffffff810af889>] ? unmap_vmas+0x7f9/0x9c0
[ 4353.446949] [<ffffffff810b64ca>] ? exit_mmap+0xda/0x1a0
[ 4353.446951] [<ffffffff81037632>] ? mmput+0x32/0x100
[ 4353.446953] [<ffffffff8103bd8b>] ? exit_mm+0xfb/0x130
[ 4353.446956] [<ffffffff8103d00e>] ? do_exit+0x6ae/0x780
[ 4353.446958] [<ffffffff816cf65f>] ? printk+0x40/0x45
[ 4353.446961] [<ffffffff8103b29c>] ? kmsg_dump+0x7c/0x150
[ 4353.446963] [<ffffffff81006eb2>] ? oops_end+0x72/0xa0
[ 4353.446966] [<ffffffff81004114>] ? do_invalid_op+0x84/0xa0
[ 4353.446969] [<ffffffff811af822>] ? assign_real_blocknrs+0x132/0x150
[ 4353.446971] [<ffffffff811b2e21>] ? T.642+0x221/0x230
[ 4353.446973] [<ffffffff81003315>] ? invalid_op+0x15/0x20
[ 4353.446977] [<ffffffff811af822>] ? assign_real_blocknrs+0x132/0x150
[ 4353.446979] [<ffffffff811af77d>] ? assign_real_blocknrs+0x8d/0x150
[ 4353.446982] [<ffffffff811b0414>] ? reiser4_alloc_extent+0x1b4/0x510
[ 4353.446985] [<ffffffff81172486>] ? move_lh_internal+0xc6/0x1d0
[ 4353.446988] [<ffffffff8117de06>] ? handle_pos_to_twig+0x66/0x140
[ 4353.446991] [<ffffffff8117dfdd>] ? handle_pos_on_twig+0xfd/0x280
[ 4353.446993] [<ffffffff8117ff0f>] ? flush_current_atom+0x73f/0x830
[ 4353.446996] [<ffffffff8117a0fd>] ? reiser4_txn_end+0x36d/0x660
[ 4353.446999] [<ffffffff81179f34>] ? reiser4_txn_end+0x1a4/0x660
[ 4353.447011] [<ffffffff8117a3f9>] ? reiser4_txn_restart+0x9/0x20
[ 4353.447025] [<ffffffff8117acbb>] ? force_commit_atom+0x4b/0x60
[ 4353.447028] [<ffffffff8117ae9c>] ? txnmgr_force_commit_all+0x13c/0x180
[ 4353.447030] [<ffffffff811b33e6>] ? release_format40+0x66/0x110
[ 4353.447033] [<ffffffff81175bf1>] ? reiser4_init_context+0x51/0x90
[ 4353.447036] [<ffffffff81186a59>] ? reiser4_put_super+0x49/0xf0
[ 4353.447038] [<ffffffff810d1a8f>] ? generic_shutdown_super+0x4f/0x100
[ 4353.447041] [<ffffffff810d1b68>] ? kill_block_super+0x28/0x50
[ 4353.447044] [<ffffffff810d2a0e>] ? deactivate_super+0x6e/0xb0
[ 4353.447046] [<ffffffff810e9c20>] ? sys_umount+0x70/0x3b0
[ 4353.447049] [<ffffffff8100276b>] ? system_call_fastpath+0x16/0x1b


[Bug 14915] New: Reiser4 doesn't sync

this BUG should be fixed with latest reiser4-patches - strange :?
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
dusanc
Apprentice
Apprentice


Joined: 19 Sep 2005
Posts: 248
Location: Serbia

PostPosted: Fri Feb 05, 2010 6:33 am    Post subject: Reply with quote

Could you report it as Re: to that bug as you have more info, and maybe it's a new bug?

Thanks :)
_________________
Reiser4 Gentoo FAQ [25Dec2016]
Back to top
View user's profile Send private message
kernelOfTruth
Watchman
Watchman


Joined: 20 Dec 2005
Posts: 6108
Location: Vienna, Austria; Germany; hello world :)

PostPosted: Fri Feb 05, 2010 9:57 am    Post subject: Reply with quote

dusanc wrote:
Could you report it as Re: to that bug as you have more info, and maybe it's a new bug?

Thanks :)


I think Edward will declare it as invalid (like before) since it's related to zen

currently I unfortunately don't have time to try it with a vanilla 2.6.33-rc* and manually patch all the mm-stuff in (system nearly all the time needs to be available) ...


steps were simple:

mkfs.reiser4 of a rather big partition (I wanted to compile openoffice with it)

*) extracting portage-tarball to a big reiser4-partition
*) time sync

*) syncing (updating it) via rsync
*) time sync (it should still work)
*) check in cat /proc/info that "Dirty" is (almost) empty


*) syncing (updating it) via rsync
*) chown -R portage:portage to the overlay-directory part of that partition (e.g. /usr/local/portage <-- there are several other ebuilds, just copy over some from /usr/portage in some directories)
*) time sync
*) check in cat /proc/info that "Dirty" is (almost) empty
*) it should get stuck approximately in half


besides that I've mostly turned away from reiser4 until those kind of problems (not syncing; leading to BUGs during sheer syncing or big load) are mostly solved and it's in a (for me) usable state again ...

edit:

openoffice emerged fine with that partition

so it's still the same (?!):

*) the default filesystem-format has problem with syncing

*) the cryptcompress format works somewhat fine for me
_________________
https://github.com/kernelOfTruth/ZFS-for-SystemRescueCD/tree/ZFS-for-SysRescCD-4.9.0
https://github.com/kernelOfTruth/pulseaudio-equalizer-ladspa

Hardcore Gentoo Linux user since 2004 :D
Back to top
View user's profile Send private message
cheater1034
Veteran
Veteran


Joined: 09 Sep 2004
Posts: 1558

PostPosted: Fri Feb 05, 2010 5:26 pm    Post subject: Reply with quote

kernelOfTruth wrote:
I think Edward will declare it as invalid (like before) since it's related to zen


I think edward is invalid, because zen uses reiser4 directly from upstream so the bugs will occur in both places (zen.git could be screwed up, but I do it upstream in zen-stable)
_________________
IRC!: #zen-sources on irc.rizon.net
zen-kernel.org
--
Lost in android development land.
Back to top
View user's profile Send private message
Display posts from previous:   
This topic is locked: you cannot edit posts or make replies.    Gentoo Forums Forum Index Unsupported Software All times are GMT
Goto page Previous  1, 2, 3 ... , 12, 13, 14  Next
Page 13 of 14

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum