Official thread: "zen-sources" - Part 8

[kernelOfTruth]

This is Part 8 of the gentoo-related zen-sources thread,

what is zen-sources ?

- Although not technically a "patchset", zen-sources is a linux kernel variant based on stable linux kernel releases, but primarily maintained through a git repository which tracks linus-2.6.git (linus's latest git repository).

- Open to code/projects not integrated into or accepted into the mainstream linux kernel. Typically features added would be intended to benefit desktop/laptop users, or some unique users (such as a game console).

- Geared towards desktop usage (which an argument can be made that the mainstream linux kernel is not), includes numerous things to benefit performance in desktop situations or features that desktop/laptop users may find useful.

- You may recall some popular kernel patchsets such as "love-sources", "no-sources", "kamikaze-sources", "skunk-sources", "viper-sources", "emission-sources", "nitro-sources", "nicus-sources", "dark-sources", "beyond-sources", or perhaps some more. All of the previous mentioned patchsets were either during or slightly before the infamous "patchset wars", basically when a bunch of different individuals decided to make patches on top of the upstream kernel. Some of the patches were composed in very similar ways and some were not. Zen is different from all of these because of the fact that it is maintained through git, and is maintained co-operatively by multiple developers.

- Zen-Sources is not a patchset, but rather is managed through a live git repository, unlike any of the other patchsets were during their days. Zen is managed through one central branch "master", but all actual code is put in their separate branches, so origin/branch1, origin/branch2, etc. This allows for both the use of patches (like patchsets used), and the merging of git repositories directly. Later, the branches are merged into the "master" branch. This also benefits the user, who does no longer need to use ebuilds but can now update simply by a "git pull", and then recompiling as desired.

- Founded by waninkoko (of kamikaze-sources) and rmh3093 (of skunk-sources)

source http://www.zen-sources.org/content/about

questions and problems with zen-sources should be discussed at its official homepage: http://www.zen-sources.org/ in the section Forums


the installation instructions in this post are provided for your convenience to get you started with the git-provided approach

for a more detailed view on the ways to install zen-sources see: Installation-Instructions on zen-sources.org


in general you should take a look at zen-sources.org for more up-to-date information and instruction


News
*) git changes and some more October 10th 2009 link
*) GIT Changes! Important!!! September 24th 2009 link
--> follow this link if you don't know yet about the recent git changes: there's a zen (latest unstable upstream) and zen-stable [latest stable kernel-releases] repo now !
*) Use KLive to help us track Zen kernel usage stats! December 18th 2008 link


Getting started
checking the sources out

First Time

1) you have 2 choices:
for zen (latest upstream [unstable / testing])
Code:

Code: Select all

cd /usr/src
git clone git://zen-sources.org/kernel/zen.git zen-sources
cd zen-sources
make menuconfig

for zen (stable)
Code:

Code: Select all

cd /usr/src
git clone git://zen-sources.org/kernel/zen-stable.git zen-sources
cd zen-sources
make menuconfig

2) for mmotm ("mmotm" is the "-mm tree of the moment" tree. The -mm patch pile, updated
usually more than once per day.) (see: http://lwn.net/Articles/311677/)
Code:

Code: Select all

cd /usr/src
git clone git://zen-sources.org/zen/mmotm.git mmotm
cd mmotm
make menuconfig

3) for zenrt-sources (doesn't exist yet / anymore)
Code:

Code: Select all

cd /usr/src
git clone git://zen-sources.org/zen/kernel/zenrt.git zen-sources
cd zen-sources
make menuconfig

4) creating a patch
example:

Code: Select all

git diff v2.6.24 v2.6.24-zen1 > 2.6.24-zen1.patch 

5) you can also do most of this stuff graphically (GUI !) with qgit:

Code: Select all

emerge qgit

or

Code: Select all

apt-cache search qgit

then

Code: Select all

sudo apt-get install qgit

(if that's the exact package name on apt/deb-based systems)

6) update sources (inside the directory)

Code: Select all

Update
Code:

cd /usr/src/zen-sources
git pull 

(called "merging")

7) you are in the middle of a conflicted merge:
output could be:

git pull
Auto-merged .gitignore
Auto-merged MAINTAINERS
Auto-merged Makefile
CONFLICT (content): Merge conflict in Makefile
Removed arch/mips/lasat/sysctl.h
Auto-merged drivers/ata/libata-scsi.c
Auto-merged drivers/char/Kconfig
Auto-merged drivers/char/drm/ati_pcigart.c
Auto-merged drivers/char/drm/drm.h
Auto-merged drivers/char/drm/drm_drv.c
CONFLICT (content): Merge conflict in drivers/char/drm/drm_drv.c
...
Automatic merge failed; fix conflicts and then commit the result.

This usually happens when the branch you are trying to update has been rebased. The easiest way to get around this is to checkout a new copy of the branch. For example, if you are currently on the master branch you would do something like the following:

first try:

Code: Select all

git checkout -f && git pull

if that doesn't help, then:

Typically this is on a master or master-2.6.xy branch so the command to fix would be:

Code: Select all

git reset --hard; git branch -M master master-old; git checkout -b master origin/master; git branch -D master-old; git checkout -f 

you need to change to another branch ?
example:
you're on master (on the zen-stable repo) and want to change to master-2.6.30

Code: Select all

# git checkout master-2.6.30

if the branch doesn't exist in your directory (

Code: Select all

git branch

)
e.g. (currently this doesn't reflect the reality since master-light doesn't exist)
master-light, then enter

Code: Select all

git branch master-light origin/master-light && git checkout master-light

9) you want to check out a specific version?

Code: Select all

git checkout v2.6.25-rc5-zen0

Tips and Tricks
T&T 1) speeding up merge (download) of zen-sources (during first time)
1) checkout the linux-2.6 branch of linus (git.eu.kernel.org or git.kernel.org)

using the international server

Code: Select all

git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

or

using the european mirror server

Code: Select all

git clone git://git.eu.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git

2) then merge both (e.g. it's in /usr/src/sources/linux-2.6)

e.g. for the zen-stable branch (for latest stable kernel linux-sources)

Code: Select all

 git clone --reference /usr/src/sources/linux-2.6/ git://zen-sources.org/kernel/zen-stable.git zen-stable

e.g. for the zen(-upstream) branch (for latest stable kernel linux-sources)

Code: Select all

 git clone --reference /usr/src/sources/linux-2.6/ git://zen-sources.org/kernel/zen.git zen-upstream

T&T 2) *click* for a link to the Portage repo for zen kernel ebuilds

T&T 3) Having problems with zen's latest repo-management change (from September 24th 2009) (want to have a stable repo again ?)
do the following:
1) change directory to the root / center of your zen repo, e.g.
cd /usr/src/sources/kernel/zen-sources/

2) modify .git/config file to let it point to git://zen-sources.org/kernel/zen-stable.git instead of git://zen-sources.org/kernel/zen.git
nano -w .git/config

[remote "origin"]
url = git://zen-sources.org/kernel/zen.git

should now be

[remote "origin"]
url = git://zen-sources.org/kernel/zen-stable.git

3) try pulling changes (which most likely in the first 2 or 3 attempts won't work):

Code: Select all

 git pull 

4) reset the state of the repo, delete master and refetch stuff:

Code: Select all

 git fetch; git reset --hard; git branch -M master master-old; git checkout -b master origin/master; git branch -D master-old; git checkout -f 

5) this most likely doesn't work after the first attempt
and probably says:

you are in the middle of a conflicted merge

or

Automatic merge failed; fix conflicts and then commit the result.

so repeat point 3 & 4 another time

6) now it should say:

Code: Select all

 git pull 

Already up-to-date.

what branches of kernel-sources or repositories are there?
currently (september 2009), there are 2 different patchsets and branches used in those are (see):

*) heads (zen[-upstream]) )
*) heads (mmotm) ):
*) heads (zen-stable) )


sometime there's also master-new or similar for testing / rebasing purposes

the other "head"s are branches which are added to the main tree

the project's site can be reached over at:
www.zen-sources.org

the repo is reachable via:
http://git.zen-sources.org/?p=kernel/zen.git;a=summary (zen[-upstream])
and/or
http://git.zen-sources.org/?p=kernel/mm ... ;a=summary (mmotm-sources)
and/or
http://git.zen-sources.org/?p=kernel/ze ... ;a=summary (zen-stable)



for those who are interested there is other (non-kernel related) stuff on the zen git-server:
* THE portage overlay.... random ebuilds usually not found in the Gentoo portage tree...
* A C++ library for communicating with Drupal Services.
* A portage overlay with ebuilds relating to embedded systems
* A desktop agnostic gui for Obex Push Protocol and Obex File Transfer Protocol.
* Cross development toolchain creator for paludis.
* PalmPre Homebrew [currently empty]
* Quassel IRC client contributions [currently empty]
*


update:
removed reiser4-repository links

update2:
added instruction on how to git-clone linus' repository and using that for reference to update to the zen-sources repo

update3:
updated adresses to the new overlay/servers

update4:
updated the git-addresses for zen-sources and mmotm (mm-sources) of the new zen-sources server

update5:
corrected zenmm -> mmotm
since the zenmm doesn't exist right now

update6:
added info about some other repositories

update7:
changed instructions regarding latest changes in zen's git-repo (zen-stable and zen [unstable, Linus' latest upstream])

update8:
added entry in section T&T (tips and tricks) for solving problems with latest zen-sources repo-switch (zen -> zen-stable and new repo zen)

update9: Tuesday, October 6th 2009
updated the addresses for the repo yet again ! ("prefixing" kernel in front of the kernel-repositories)

[kernelOfTruth]

for those who didn't have the chance to test or try BFS extensively due to some BUGs now is the time to do it:

BFS v240
Currently known problems? Do not report these unless you have seen a change in
their behaviour(fixed) or you have a fix for them.

1. FIXED - Intermittent boot failures on some hardware. - much less common now
2. FIXED - Stuck tasks on the same hardware as 1. after extended periods, suggesting
a common problem.
3. Stuck tasks after extensive use of trace functions (ptrace etc.). Note that
some distributions' package managers use trace functions.
4. FIXED - Failure to suspend on some hardware. - should be fixed in 231+
5. Hibernation failures?
6. Fixing recursive fault but reboot is needed error in exit.c
7. Keyboard failures with Xorg

More likely to show up bugs in *other* code due to being much more
aggressive at using multiple CPUs so race conditions will show up more
frequently.

I'm working on all of the above now and have no time frame whatsoever for when
they might be fixed since I don't know what's causing them.

** KNOWN DRIVER INTERACTIONS **

A number of drivers are known to break under BFS and it is still not clear if
BFS is at fault or BFS is exposing underlying race condition bugs in the
driver. More than likely they are BFS bugs.

Drivers that are known to break/hang with BFS currently:
FIXED - ATI Drivers Binary and Free
FIXED - i915 Graphics (possibly better in 231+)
FIXED - ReiserFS 3
NFS. (Last tested 220, buggy)


If you find a newer version of BFS fixes a previous problem, PLEASE REPORT IT
TO ME!

get yourself zen-sources from the zen-stable repo: [url]git://zen-sources.org/zen/zen-stable.git[/url]

and use the default master-branch (no need to check out any other branch):

git branch
* master
master-2.6.30
master-2.6.31

the Makefile should contain the following at the top:

Makefile

VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 31
EXTRAVERSION = -zen2
NAME = Alien Mind Creator

goodie goodie goodie !

a big THANK YOU to Con & of course zen-sources for bringing us these enhanced kernel-sources


now let's see if suspend-to-ram works with this and fglrx & reiserfs

[Ant P.]

Has anyone had a success story with the R700 KMS bits yet? I've been trying to get this to work for ages and I still can't run X on it, I get a nasty crash instead (screen powers off with no signal, sysrq-r + ctrlaltdel works though).

This is on master-2.6.31 btw, just updated today. I haven't updated my x11 stuff in a week so things might have changed in that time, but I can't do it right now because fd.o's servers seem to be down.

[kernelOfTruth]

Has anyone had a success story with the R700 KMS bits yet? I've been trying to get this to work for ages and I still can't run X on it, I get a nasty crash instead (screen powers off with no signal, sysrq-r + ctrlaltdel works though).

This is on master-2.6.31 btw, just updated today. I haven't updated my x11 stuff in a week so things might have changed in that time, but I can't do it right now because fd.o's servers seem to be down.

yes, but as already posted on phoronix' forums:

this doesn't work when run with user privileges (which more or less is against KMS' intent) and also leads to a black screen which can be overcome by rebooting via Magic SYSRQ Key

I dunno if entering some commands blindly leads to reboot

so ways to get it work:

1) set up /etc/conf.d/xdm to use your favorite desktop manager (kdm, gdm, etc.)

2) start xdm:

Code: Select all

/etc/init.d/xdm restart

3) X should show more than a black screen now

[Ant P.]

Wow... I never would've guessed it'd be that. Thanks

(This would also explain why I couldn't get my Debian laptop to work after uninstalling kdm)

[kernelOfTruth]

Wow... I never would've guessed it'd be that. Thanks

(This would also explain why I couldn't get my Debian laptop to work after uninstalling kdm)

you're welcome


This time's the real deal:

BFS (build 240) and 2.6.31-zen2 running here flawlessly (so far)

I still need much stuff to test but there's (almost) no lagginess during large rsync-jobs anymore -

who says it isn't the CPU scheduler's fault - NOW ?

I've always known that it's a mix of issues caused / included in the VFS (i/o system) and the CPU scheduler


TEST IT NOW ! - you know YOU WANT IT

[Jupiter1TX]

Guys i been using zen since the beginning and i must say
change is usually a good thing but i have never had so
much trouble using zen git as i have now. Trying to get
the latest and greatest zen has become a mind fuck.

There must be an easy/simple way to get the latest zen
unstable that i am just not smart enough to figure out.
The instructions here do not help at all. The instructions
on zen website is poorly organized gibberish.

There must be a better way....

P.S. this is backwards 'i think' re: zen.git zen-stable.git

Code: Select all

1) you have 2 choices:
for zen-sources (stable)
Code:
Code:
cd /usr/src
git clone git://zen-sources.org/zen/zen.git zen-sources
cd zen-sources
make menuconfig


for zen-sources (latest upstream)
Code:
Code:
cd /usr/src
git clone git://zen-sources.org/zen/zen-stable.git zen-sources
cd zen-sources
make menuconfig

[kernelOfTruth]

Guys i been using zen since the beginning and i must say
change is usually a good thing but i have never had so
much trouble using zen git as i have now. Trying to get
the latest and greatest zen has become a mind fuck.

There must be an easy/simple way to get the latest zen
unstable that i am just not smart enough to figure out.
The instructions here do not help at all. The instructions
on zen website is poorly organized gibberish.

There must be a better way....

P.S. this is backwards 'i think' re: zen.git zen-stable.git

Code: Select all

1) you have 2 choices:
for zen-sources (stable)
Code:
Code:
cd /usr/src
git clone git://zen-sources.org/zen/zen.git zen-sources
cd zen-sources
make menuconfig


for zen-sources (latest upstream)
Code:
Code:
cd /usr/src
git clone git://zen-sources.org/zen/zen-stable.git zen-sources
cd zen-sources
make menuconfig

indeed !

sorry & thanks for the find !

[kernelOfTruth]

Jupiter1TX, please try to following if it still doesn't work for you:

I'd added the following to the beginning of this thread:

Having problems with zen's latest repo-management change (from September 24th 2009) (want to have a stable repo again ?)
do the following:
1) change directory to the root / center of your zen repo, e.g.
cd /usr/src/sources/kernel/zen-sources/

2) modify .git/config file to let it point to git://zen-sources.org/zen/zen-stable.git instead of git://zen-sources.org/zen/zen.git
nano -w .git/config

[remote "origin"]
url = git://zen-sources.org/zen/zen.git

should now be

[remote "origin"]
url = git://zen-sources.org/zen/zen-stable.git

3) try pulling changes (which most likely in the first 2 or 3 attempts won't work):

Code: Select all

 git pull 

4) reset the state of the repo, delete master and refetch stuff:

Code: Select all

 git fetch; git reset --hard; git branch -M master master-old; git checkout -b master origin/master; git branch -D master-old; git checkout -f 

5) this most likely doesn't work after the first attempt
and probably says:

you are in the middle of a conflicted merge

or

Automatic merge failed; fix conflicts and then commit the result.

so repeat point 3 & 4 another time

6) now it should say:

Code: Select all

 git pull 

Already up-to-date.

[kernelOfTruth]

is it just me or is the network bandwidth / PCI(e) bandwidth significantly reduced with BFS ?

with 2.6.31-zen1 I got around 24 MBit, now with 2.6.31-zen2 and BFS (240) the best I can get is 17 to 18 MBit

this needs attention and fixing

[Jupiter1TX]

Jupiter1TX, please try to following if it still doesn't work for you:

Thanks for the help. I now have 'master' 2.6.32-rc2 with no zen
stuff and can't seem to find my notes on how to get it. It appears
i have reached the limit of my understanding with git and it's COMPLEX
instructions to do such a simple task.

Thanks guys.

[kernelOfTruth]

is it just me or is the network bandwidth / PCI(e) bandwidth significantly reduced with BFS ?

with 2.6.31-zen1 I got around 24 MBit, now with 2.6.31-zen2 and BFS (240) the best I can get is 17 to 18 MBit

this needs attention and fixing

false alarm

it seemed to have been a problem with either:

*) router firmware

*) internet provider

so BFS is working great here (so far)

[cheater1034]

Jupiter1TX, please try to following if it still doesn't work for you:

Thanks for the help. I now have 'master' 2.6.32-rc2 with no zen
stuff and can't seem to find my notes on how to get it. It appears
i have reached the limit of my understanding with git and it's COMPLEX
instructions to do such a simple task.

Thanks guys.

http://zen-sources.org/content/git-changes-important

Those are correct instructions,
if you checkout zen.git you wont have 2.6.32-rc1 yet because it's still in the process of being rebased

*edit*
linus failed so 2.6.32-rc1 says rc2 in the makefile
Anywhooz, bfs 240 is in zen, whoever keeps asking.
and i'm doing zen-sched branch on zen.git now - now it defines BFS_VERSION in sched.h and it printk's as always (except now as KERN_INFO - and in init) - and it prints the bfs version too.

[Jupiter1TX]

Those are correct instructions,
if you checkout zen.git you wont have 2.6.32-rc1 yet because it's still in the process of being rebased

*edit*
linus failed so 2.6.32-rc1 says rc2 in the makefile
Anywhooz, bfs 240 is in zen, whoever keeps asking.
and i'm doing zen-sched branch on zen.git now - now it defines BFS_VERSION in sched.h and it printk's as always (except now as KERN_INFO - and in init) - and it prints the bfs version too.

Thanks for the info cheater. Although i have been losing my mind
for the past 18 months i thought i had this figured out. I went ahead
and updated my 2.6.31-zen1+bfs and reran the PST test...
http://global.phoronix-test-suite.com/i ... 0356-20218

[cheater1034]

Those are correct instructions,
if you checkout zen.git you wont have 2.6.32-rc1 yet because it's still in the process of being rebased

*edit*
linus failed so 2.6.32-rc1 says rc2 in the makefile
Anywhooz, bfs 240 is in zen, whoever keeps asking.
and i'm doing zen-sched branch on zen.git now - now it defines BFS_VERSION in sched.h and it printk's as always (except now as KERN_INFO - and in init) - and it prints the bfs version too.

Thanks for the info cheater. Although i have been losing my mind
for the past 18 months i thought i had this figured out. I went ahead
and updated my 2.6.31-zen1+bfs and reran the PST test...
http://global.phoronix-test-suite.com/i ... 0356-20218

you're still killing me
If you checked out zen-stable.git, you would get it with bfs 240 + 2.6.32 cfs (which shows major improvements, atleast combined with the tunable stuff that has been done)

[pappy_mcfae]

I have bfs-240+2.6.31 running on my dual processor P-4, bam-bam. This is a first. While core-too would boot BFS kernels sometimes, bam-bam was having no part of it. Not only has it booted, it is at 66 out of 137 on an emerge -aev world. It's been doing that for an hour and twenty minutes now. The openGL screensavers have come on and off, and still the machine continues to compile!

Cool! That's almost all machines running with BFS. Now to see if core-too can be convinced to give it a go.

Blessed be!
Pappy

[Jupiter1TX]

you're still killing me
If you checked out zen-stable.git, you would get it with bfs 240 + 2.6.32 cfs (which shows major improvements, atleast combined with the tunable stuff that has been done)

Well this confusion is killing me too. Yes i have 2.6.31-zen1 with bfs.
Runs fine.
I just now tried dl zen.stable.git and i did NOT get 2.6.32,
i got 2.6.31-zen2 "Alien Mind Creator''

Can this be anymore confusing or am i just getting lucky?

[cheater1034]

you're still killing me
If you checked out zen-stable.git, you would get it with bfs 240 + 2.6.32 cfs (which shows major improvements, atleast combined with the tunable stuff that has been done)

Well this confusion is killing me too. Yes i have 2.6.31-zen1 with bfs.
Runs fine.
I just now tried dl zen.stable.git and i did NOT get 2.6.32,
i got 2.6.31-zen2 "Alien Mind Creator''

Can this be anymore confusing or am i just getting lucky?

I didnt necessarily state it right.
2.6.31-zen1 is outdated:
2.6.31-zen2 is what is found in zen-stable.git, which is what you should be using (until the 2.6.32 rebase is complete and you wanted to use unstable)

2.6.31-zen2 has a backported cfs from 2.6.32 (and tunable/changes) - along with bfs 240 ---- if you ran a test of the cfs in zen-stable.git and bfs in zen-stable.git you can compare the improved cfs and bfs 240 (you could apply the 240-test patch over the top too, which should make the fastest bfs yet)

The improved/2.6.32 cfs that is found in 2.6.31-zen2 has major improvements in the benchmark i ran and others have shown (lattencies) - actually about 50% improvement. Although I haven't found it to be quite on par with bfs 240.

so yes, zen2 is what i want you to run the tests on , with cfs and bfs if at all possible?

Thanks

and if you apply the 240-test.patch, it should apply fine - with the exception of init/main.c (in the update i'm about to push) - but you can ignore the main.c reject, the only changes happen in sched_bfs.c

[kernelOfTruth]

man Con is going strong:

CFS build 300 released !

update:

*) suspend-to-ram (with s2ram) doesn't work -> the keyboard, usb-ports and harddrives obviously are turned off but the computer doesn't do so - it just hangs

*) the same with shutdown the harddrives are halted, (usb-/ps2-)keyboard(s) turned off but it doesn't "shut down" (turn power for the computer off)

[ponciarello]

trying latest stable with virtualbox modules setup: dmesg output

Code: Select all

vboxdrv: Trying to deactivate the NMI watchdog permanently...
vboxdrv: Successfully done.
vboxdrv: Found 2 processor cores.
BUG: unable to handle kernel paging request at ffffffffa0b86d20
IP: [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
PGD 1003067 PUD 1007063 PMD 79a7c067 PTE 5fcc2161
Oops: 0003 [#1] PREEMPT SMP
last sysfs file: /sys/class/power_supply/BAT1/charge_full
CPU 1 
Modules linked in: vboxdrv(+) snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ntfs ramzswap xvmalloc lzo_decompress lzo_compress cpufreq_ondemand acpi_cpufreq freq_table fuse option usbserial nvidia(P) joydev arc4 ecb iwlagn iwlcore snd_hda_codec_realtek snd_hda_intel mac80211 snd_hda_codec uvcvideo cfg80211 snd_pcm videodev snd_timer ohci1394 v4l1_compat v4l2_compat_ioctl32 video i2c_i801 psmouse ieee1394 rfkill i2c_core snd_page_alloc output evdev processor thermal ac battery button thermal_sys
Pid: 4693, comm: modprobe Tainted: P           2.6.31-zen2 #1 Satellite X200
RIP: 0010:[<ffffffffa0b86572>]  [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
RSP: 0018:ffff880062d5df10  EFLAGS: 00010246
RAX: ffffffffa0d06d1f RBX: 0000000000000000 RCX: 0000000000000000
RDX: 0000000000180000 RSI: ffffffffa0b86d20 RDI: ffffffffa0d140c8
RBP: 0000000000180000 R08: ffffea0001ba6fb8 R09: 0000000000000001
R10: 0000000000000000 R11: ffffffff81226a00 R12: ffffffffa0b86d20
R13: 00007fff0267ef70 R14: 0000000000000000 R15: 0000000000000000
FS:  00007f754c2576f0(0000) GS:ffff8800015c7000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffffffa0b86d20 CR3: 0000000062f2e000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process modprobe (pid: 4693, threadinfo ffff880062d5c000, task ffff880067706000)
Stack:
 ffffffffa0b7c62b 0000000000000000 0000000000000000 00007f754bafd010
<0> ffffffffa004f0c6 ffffffffa004f000 ffffffff81009044 0000000001c25370
<0> ffffffffa0d13180 00007f754bafd010 ffffffff8105fb45 00007fff0267eb70
Call Trace:
 [<ffffffffa0b7c62b>] ? RTR0MemExecDonate+0x4b/0x70 [vboxdrv]
 [<ffffffffa004f0c6>] ? VBoxDrvLinuxInit+0xc6/0x1b3 [vboxdrv]
 [<ffffffffa004f000>] ? VBoxDrvLinuxInit+0x0/0x1b3 [vboxdrv]
 [<ffffffff81009044>] ? do_one_initcall+0x34/0x1a0
 [<ffffffff8105fb45>] ? sys_init_module+0xe5/0x250
 [<ffffffff8100b3eb>] ? system_call_fastpath+0x16/0x1b
Code: c9 0f 85 9a 00 00 00 48 8d 44 16 ff 48 39 c2 0f 83 94 00 00 00 48 89 f1 83 e1 1f 74 0e b8 20 00 00 00 48 29 c8 48 29 c2 48 01 c6 <48> c7 06 05 01 59 19 48 c7 46 30 ff ff ff ff 48 c7 46 38 ff ff 
RIP  [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
 RSP <ffff880062d5df10>
CR2: ffffffffa0b86d20
---[ end trace 481a788d9b82d0d0 ]---

[kernelOfTruth]

trying latest stable with virtualbox modules setup: dmesg output

Code: Select all

vboxdrv: Trying to deactivate the NMI watchdog permanently...
vboxdrv: Successfully done.
vboxdrv: Found 2 processor cores.
BUG: unable to handle kernel paging request at ffffffffa0b86d20
IP: [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
PGD 1003067 PUD 1007063 PMD 79a7c067 PTE 5fcc2161
Oops: 0003 [#1] PREEMPT SMP
last sysfs file: /sys/class/power_supply/BAT1/charge_full
CPU 1 
Modules linked in: vboxdrv(+) snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ntfs ramzswap xvmalloc lzo_decompress lzo_compress cpufreq_ondemand acpi_cpufreq freq_table fuse option usbserial nvidia(P) joydev arc4 ecb iwlagn iwlcore snd_hda_codec_realtek snd_hda_intel mac80211 snd_hda_codec uvcvideo cfg80211 snd_pcm videodev snd_timer ohci1394 v4l1_compat v4l2_compat_ioctl32 video i2c_i801 psmouse ieee1394 rfkill i2c_core snd_page_alloc output evdev processor thermal ac battery button thermal_sys
Pid: 4693, comm: modprobe Tainted: P           2.6.31-zen2 #1 Satellite X200
RIP: 0010:[<ffffffffa0b86572>]  [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
RSP: 0018:ffff880062d5df10  EFLAGS: 00010246
RAX: ffffffffa0d06d1f RBX: 0000000000000000 RCX: 0000000000000000
RDX: 0000000000180000 RSI: ffffffffa0b86d20 RDI: ffffffffa0d140c8
RBP: 0000000000180000 R08: ffffea0001ba6fb8 R09: 0000000000000001
R10: 0000000000000000 R11: ffffffff81226a00 R12: ffffffffa0b86d20
R13: 00007fff0267ef70 R14: 0000000000000000 R15: 0000000000000000
FS:  00007f754c2576f0(0000) GS:ffff8800015c7000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffffffa0b86d20 CR3: 0000000062f2e000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process modprobe (pid: 4693, threadinfo ffff880062d5c000, task ffff880067706000)
Stack:
 ffffffffa0b7c62b 0000000000000000 0000000000000000 00007f754bafd010
<0> ffffffffa004f0c6 ffffffffa004f000 ffffffff81009044 0000000001c25370
<0> ffffffffa0d13180 00007f754bafd010 ffffffff8105fb45 00007fff0267eb70
Call Trace:
 [<ffffffffa0b7c62b>] ? RTR0MemExecDonate+0x4b/0x70 [vboxdrv]
 [<ffffffffa004f0c6>] ? VBoxDrvLinuxInit+0xc6/0x1b3 [vboxdrv]
 [<ffffffffa004f000>] ? VBoxDrvLinuxInit+0x0/0x1b3 [vboxdrv]
 [<ffffffff81009044>] ? do_one_initcall+0x34/0x1a0
 [<ffffffff8105fb45>] ? sys_init_module+0xe5/0x250
 [<ffffffff8100b3eb>] ? system_call_fastpath+0x16/0x1b
Code: c9 0f 85 9a 00 00 00 48 8d 44 16 ff 48 39 c2 0f 83 94 00 00 00 48 89 f1 83 e1 1f 74 0e b8 20 00 00 00 48 29 c8 48 29 c2 48 01 c6 <48> c7 06 05 01 59 19 48 c7 46 30 ff ff ff ff 48 c7 46 38 ff ff 
RIP  [<ffffffffa0b86572>] RTHeapSimpleInit+0x62/0xf0 [vboxdrv]
 RSP <ffff880062d5df10>
CR2: ffffffffa0b86d20
---[ end trace 481a788d9b82d0d0 ]---

I got the exact same error before some "fixes" were commited instantly after loading the vboxdrv kernel-module

now VirtualBox works for me (only tested it for a short time) - running BFS here, so it could be a CFS-related problem (if you're running CFS)

after which period of usage does this appear ?

[ponciarello]

i'm running bfs: it happens when the setup script loads the just-build modules
it can be that bfs300 will fix this

[broch]

I can confirm that with 2.6.31-zen2
suspend to RAM fails
in contrast patched with bfs vanilla 2.6.31.1 does not have problems with suspend-to-ram, so maybe this is not a bfs fault?


additionally, when shutting down kernel generates error regarding "mounting/read only"
system shuts down but error is a little bit worrying.

again, this is specific to zen, bfs patched vanilla *.31.1 works

[tranquilcool]

I can confirm that with 2.6.31-zen2
suspend to RAM fails
in contrast patched with bfs vanilla 2.6.31.1 does not have problems with suspend-to-ram, so maybe this is not a bfs fault?


additionally, when shutting down kernel generates error regarding "mounting/read only"
system shuts down but error is a little bit worrying.

again, this is specific to zen, bfs patched vanilla *.31.1 works

shutdown errors here too.

[wrc1944]

I installed zen2 last night on another box, but today keep getting this:

Code: Select all

amd64 src # git clone git://zen-sources.org/zen/zen-stable.git zen-sources
Initialized empty Git repository in /usr/src/zen-sources/.git/
fatal: Unable to look up zen-sources.org (port 9418) (Temporary failure in name resolution)
amd64 src # git clone git://zen-sources.org/zen/zen-stable.git zen-sources
Initialized empty Git repository in /usr/src/zen-sources/.git/
fatal: read error: Connection reset by peer
amd64 src #

Is it just something on your end- like updating to a new version, or a server problem? Nothing seems wrong on this system.