linux-image-2.6.24-13-openvz refuses to boot

I have the same problem.

Software: ubuntu-8.04-beta-server-amd64.iso
Package: linux-image-2.6.24-14-openvz
Hardware: Supermicro X7DBE, 1x Xeon Quad Core 1.6GHz, 4GB RAM
Error Message:
---------------CUT---------------
Loading, please wait...
[ 73.310958] BUG: soft lookup - CPU#3 stuck for 11s! [kbd_mode:1063]
[ 85.120779] BUG: soft lookup - CPU#3 stuck for 11s! [kbd_mode:1063]
---------------CUT---------------

Detailed Error Message

Me too, seen on both the -13 and -14.25 openvz kernel. Attempting to run on an Intel Core 2 Duo, Intel chipset.

Here is the same. Tested hardware:
#1: ASUS A7N8X-X / AMD Athlon XP 1500+ / 1.5G RAM
#2: PowerEdge 830 / Intel Celeron / 2G RAM

Oh, and according to http://packages.ubuntu.com/hardy/linux-image-2.6.24-13-openvz, it says "You likely do not want to install this package directly. Instead, install the linux-openvz meta-package, which will ensure that upgrades work correctly, and that supporting packages are also installed."

But there is no linux-openvz in the Hardy repositories.

The same for me with 2.6.24-13-openvz and 2.6.24-14-openvz. Compiled from latest git, openvz works fine ...

Intel Core 2 Duo | Intel Chipset | Ubuntu 8.04 Beta

The linux-openvz meta package is available now with 2.6.24-15, but still fails to boot.

I'm marking this High and milestone it for 8.04, because it appears that it's failing for everybody?!

Very nice that the openvz kernel is available now. Unfortunately, I've been having this problem too, since the openvz config was introduced into the git repository.

Roland, did you use this repository?
http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-hardy.git;a=summary

I'm going to compile it here and see what happens.

Best,

Paulo

I confim the same problem with -14 and -15 on my Intel Core 2 Duo / 3 GB of RAM.

That said, the call trace seems different.

@ ps1234 : How did you manage to get your bootlog ?

Compiling from the latest git didn't work for me.

One of the openvz developers told me that they are aware of this bug, they just have to synchronize the repositories. I gave him a link to this page.

I am also having problems booting the openvz kernel.
My config is a fairly old Medion laptop, Pentium 4, 512 MB.
Kernel version is 2.6.24-15-openvz

Because I cant cut/paste from a non-booting system, I made a picture of the screen and attached it.
These message are coming about every 10 seconds.

For completeness, here is the output of lshw created when booted with the standard kernel :

ubuntu
    description: Notebook
    product: DN95
    vendor: MEDION
    serial: SN-1234567890
    width: 32 bits
    capabilities: smbios-2.3 dmi-2.3
    configuration: boot=oem-specific chassis=notebook
  *-core
       description: Motherboard
       vendor: MEDION
       physical id: 0
     *-firmware
          description: BIOS
          vendor: MEDION
          physical id: 0
          version: NOTE BIOS Version/L2850100 (04/04/2002)
          size: 64KB
          capacity: 448KB
          capabilities: isa pci pcmcia pnp apm upgrade shadowing escd cdboot bootselect edd int13floppynec int13floppytoshiba int13floppy1200 int5printscreen int9keyboard int14serial int17printer int10video acpi agp netboot
     *-cpu
          description: CPU
          product: Intel(R) Pentium(R) 4 CPU 2.60GHz
          vendor: Intel Corp.
          physical id: 4
          bus info: cpu@0
          version: 15.2.7
          slot: Socket A
          size: 2600MHz
          width: 32 bits
          clock: 133MHz
          capabilities: fpu fpu_exception wp vme de pse tsc msr pae mce cx8 sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe up pebs bts sync_rdtsc cid
          configuration: id=0
        *-cache:0
             description: L1 cache
             physical id: 8
             slot: Internal
             size: 8KB
             capacity: 32KB
             capabilities: internal write-back
        *-cache:1
             description: L2 cache
             physical id: 9
             slot: Internal
             size: 512KB
             capacity: 512KB
             capabilities: burst internal write-back
     *-memory:0 UNCLAIMED
          description: Flash Memory
          physical id: 1f
          slot: System board or motherboard
          capacity: 512KB
     *-memory:1 UNCLAIMED
          description: Video Memory
          physical id: 20
          slot: System board or motherboard
          capacity: 8MB
     *-memory:2
          description: System Memory
          physical id: 21
          slot: System board or motherboard
          capacity: 512MB
        *-bank:0
             description: DIMM SDRAM Synchronous 266 MHz (3.8 ns)
             physical id: 0
             slot: CN27
             size: 256MB
             width: 64 bits
             clock: 266MHz (3.8ns)
        *-bank:1
             description: DIMM SDRAM Synchronous 266 MHz (3.8 ns)
             physical id: 1
             slot: CN30
             size: 256MB
             width: 64 bits
             clock: 266MHz (3.8ns)
     *-memory:3 UNCLAIMED
          physical id: 1
        *-bank UNCLAIMED
             description: FLASH Non-volatile 100 MHz (10.0 ns)
             physical id: 0
             slot: U32
             size: 512KB
             width: 4 bits
          ...

Hi,

Paulo, i used the git repository from openvz.org, not the ubuntu-reposiory. But i'll try that if i find some time around the next days.

@ ldng: console redirection
/boot/vmlinuz-2.6.24-14-server root=UUID=23cfdfc5-2104-45f2-a329-45e5b8893fba ro console=ttyS1,115200n8 console=tty0

I received word from Alexey that he updated their git repository for the hardy openvz kernel. I just finished compiling it and it works. One of the patches addresses precisely the d_path problem we're all reporting.

http://git.openvz.org/?p=ubuntu-hardy-openvz;a=summary

I hope the ubuntu developers will apply these patches soon.

Merged the latest patches from openvz git (174a23141b10d9ecbe6b2ed9556a54f435c50b77)

I can confirm that the 2.6.24-16 kernel compiled from the ubuntu kernel git, boots and works as expected.
http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-hardy.git;a=summary

$ git-clone git://kernel.ubuntu.com/ubuntu/ubuntu-hardy.git openvz
$ cd openvz
$ AUTOBUILD=1 NOEXTRAS=1 fakeroot debian/rules custom-binary-openvz
$ sudo dpkg -i linux-image-2.6.24-16-openvz_2.6.24-16.30_i386.deb

Thanks for making this happen ;-)
Karoly

Hi,

I just wanted to update this thread with the latest info on the openvz kernel. Both Paulo and myself were able to compile from the git and boot the openvz kernel. Unfortunately something is still not working. The cp and mv commands from coreutils are working fine on the host, but in the VE the cp and mv fail when the target file exists with the following error "/bin/cp: skipping file `%%FILENAME%%', as it was replaced while being copied". Unfortunately this leaves the VE in an unusable state since nothing can be installed by apt.

For reference here is the thread on the openvz forum:
http://forum.openvz.org/index.php?t=msg&th=5912&start=0&

Shall we open an other bug on launchpad for this new bug or just keep going with this one?

Karoly

Hmm,

This morning I dist-upgraded the server and saw that there was a new 2.6.24-16-openvz package. I tried booting it and it failed with the same error: BUG: soft lockup - CPU#0 stuck for 11s!

How come my kernel compiled from the ubuntu git works and the official package doesn't?

Karoly

Same thing here. It's weird that there wasn't much activity in the git repository after the kernel was tagged 2.6.24-16-openvz. But I don't know how the compilation process is automated.

Also, with the kernel I compiled I get kernel messages like:
Apr 9 10:23:34 eclipse kernel: [ 27.454732] snd: disagrees about version of symbol struct_module
Apr 9 10:23:34 eclipse kernel: [ 27.578994] snd_hwdep: disagrees about version of symbol struct_module
Apr 9 10:23:34 eclipse kernel: [ 27.579611] snd_page_alloc: disagrees about version of symbol struct_module

...and, of course, sounds doesn't work.

Stefan, has your merge made it in the current Ubuntu kernel (linux-image-2.6.24-16-openvz 2.6.24-16.30)?

I have the same problem with hardy & newest updates. Can't boot with openvz.
Will this be fixed before hardy-release?

If the git changelog is not misleading those merges should have made it into -16. However there have been config updates later (which changed config options since there has been FTBS errors).

Stefan et al.,

The problem with openvz that Karoly and I were reporting seems to be fixed with the patch at the bottom of this thread (it worked for both of us):
http://forum.openvz.org/index.php?t=msg&&th=5912&goto=29286

It should hit the git repository soon.

Speaking for myself, the compiled openvz kernel still doesn't have sound. With the debian kernel, everything worked fine.

$ dmesg |grep snd
[ 34.494017] snd: disagrees about version of symbol struct_module
[ 34.494835] snd_hwdep: disagrees about version of symbol struct_module
[ 34.496201] snd_page_alloc: disagrees about version of symbol struct_module
[ 34.497541] snd_timer: disagrees about version of symbol struct_module
[ 34.508248] snd_page_alloc: disagrees about version of symbol struct_module
[ 34.518997] snd: disagrees about version of symbol struct_module
[ 34.652625] snd_timer: disagrees about version of symbol struct_module
[ 34.662558] snd_pcm: disagrees about version of symbol struct_module
[ 34.717257] snd_hda_intel: disagrees about version of symbol struct_module

Hi Stefan,

I'd like to let you and others as well know that there is a new patch available which fixes the cp and mv problem I mentioned earlier. Without that the kernel would be useless even if it booted fine.

0049-fgetstat-virtinfo-notifier.patch
it is available here: http://forum.openvz.org/index.php?t=rview&goto=29488#msg_29488

I'm not sure whether this patch has made it to the openvz git or not, so we may need to wait a bit with the merging.

Anyway while it makes it to the offical hardy kernel here's how to build a custom one if one is in a hurry:
$ git-clone git://kernel.ubuntu.com/ubuntu/ubuntu-hardy.git openvz
$ cd openvz
Download the patch: http://forum.openvz.org/index.php?t=getfile&id=528&
Drop it in the openvz/debian/binary-custom.d/openvz/patchset/ folder
$ AUTOBUILD=1 NOEXTRAS=1 fakeroot debian/rules custom-binary-openvz
$ cd ..
$ sudo dpkg -i linux-image-2.6.24-16-openvz_2.6.24-16.30_i386.deb

Cheers,
Karoly

Please disregard my previous message, I found what I was doing wrong. Just in case somebody runs into the same issue, the problem is that I was installing the new kernel without properly removing the previous one. I had to purge the openvz kernel packages, manually delete the corresponding directory in /lib/modules/, and only then install the kernel I compiled.

Now everything seems to be working properly.

Thanks. I hope the patch makes into the kernel soon.

Seeing the same issue on a hp xw6000 workstation.

Will,

Which issue? The booting problem? Or the problem with modules that I had? If it is the booting problem, you can either compile the kernel by yourself (just follow Karoly's instructions, it's that simple) or wait until the new binaries are generated and uploaded to the repositories.

Booting with the current packages, cp/mv with latest git, everything is ok with latest git and most recent patch.

Hi,

I for one am still having problem with latest ubuntu openvz kernel (2.6.24-16-openvz).
Joining a photo showing the bactrace (as I don't have any serial console at hand)
The shot has been reduced to not take so much space.

@ldng: Did you observe this issue with a kernel as per Karoly Molnar wrote on 2008-04-15: (https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/25)

I have the exact same problem as reported by ldng on 2008-04-22.
The kernel level is also 2.6.24-16 (binary download with normal apt-get upgrade)
I don't think there is currently a more recent binary download available.

regards,
Harry

Hello Harry and ldng,

The linux-image-2.6.24-16-openvz package in the official repository for some reason was compiled without the latest patches, however if you compiled your own kernel from the official ubuntu kernel git you will end up with a linux-image-2.6.24-16-openvz package which after installation and reboot will give you a fully functional openvz patched system.

You can find the instruction about how to compile the kernel from the git above.

Hope this helps,
Karoly

Karoly, is the cp/mv-fix included in Ubuntu's Git already?

Hi Daniel,

It has not made it into the Ubuntu git. However the patch is available in the ubuntu-hardy-openvz kernel git at openvz:
http://git.openvz.org/?p=ubuntu-hardy-openvz;a=commit;h=b5c04a9caa1ccec03150970499973e761de10c9f

It was committed on April 17th which is far too close to Ubuntu's release date. I'm assuming that we will see these changes in the 2.6.24-17. Till then at least we have a solution.

Cheers,
Karoly

Fingers crossed it's gonna make it to the final release? Or not?
8.04 was touted to have integration with OpenVZ... and if the kernel doesn't even work, that's not very good at all. :(

Experience this very exact problem in the final release.

Would be nice if the fixed .deb (as Karoly Molnar describe how to build) are made available from some ubuntu server and announced here (as this bug report is pretty easy to find).

I've sent a package off for building on my ppa based on the git tree and patch.
https://edge.launchpad.net/~compbrain/+archive

I compiled the kernel from source using Karoly's instructions above (comment on 2008-04-15).
It takes a couple of hours, and installing some additional tools (debhelper, dh-make, build-essential ), but the result is there, it boots now:

metskem@hood:~$ uname -a
Linux hood 2.6.24-17-openvz #1 SMP Fri Apr 25 20:07:50 CEST 2008 i686 GNU/Linux

Thanks a lot, I'll go on building the server and testing .

regards,
Harry

Hello,

For other people who are not sure what packages are required, read this:
https://help.ubuntu.com/community/Kernel/Compile

You need:
$ sudo apt-get install linux-kernel-devel fakeroot build-essential

Also, if you have multiple cores/processors you might be interested in the option:
DEB_BUILD_OPTIONS=parallel=2

Installed Will Nowaks kernel and it boots fine! Thanks Will.

Have not come around to testing cp and mv on guests yet, I'm new to openVZ. I have a few Linux VServer guests laying around and as I understand it they should be possible to run with openVZ - but the config is all different... Have to read some.

I get no sound, witch would not mind much on a server, but on my laptop it is a mild annoyance. Wills kernel is named 2.6.24-17 so it should not be any problem with a previous kernel laying around.

/LaH

Lars,

I didn't have sound either. Did you try what I said in a previous message? In my case I had to delete all existing openvz modules I had in /lib/modules/ and run depmod -a. Perhaps checking the output of dmesg would tell you more about the problem you have.

On migration, check:
http://wiki.openvz.org/Migration_from_Linux-VServer_to_OpenVZ

What are the next steps?
Is there any way to use OpenVZ with Ubuntu in a productive enviroment - while the bug is fixed in a "offical" ubuntu repository?
I mean, building the kernel from source is no problem, but what happens if there are any
security updates for the Ubuntu kernel or the OpenVZ patch?

ps1234,

Just use Will's repositories and install the .17-openvz image.
It worked for me, just that it's not official yet. You can always deploy Will's image first while waiting for the official Ubuntu one to be released. Thereafter, just uninstall Will's image and then reinstall the official Ubuntu one.
Easier than compiling the kernel from scratch.

I'm sure someone is looking into it and I'm optimistic an official fixed kernel will be out soon enough.

Hi,

I updated the OpenVZ Community Documentation and would appreciate any feedback or correction.

https://help.ubuntu.com/community/OpenVZ

Cheers,
Karoly

I took the liberty to make some minor (textual) changes to the mentioned wikipage.

thanks,
Harry

Hi Harry,

Thank you for your support.

Cheers,
Karoly

Will, thank you for the working kernel image and headers. Is there any chance you could also provide linux-source-2.6.24-17-openvz?

Rolly,
Traditionally there is only a linux-source-2.6.24. If you want the source used to build linux-image and linux-headers, you can apt-get source linux-image-2.6.24-17-openvz.

Will, thanks a lot for the fixed kernel image.

I installed it and it booted fine. But when I try to use vzctl, it fails:

Unable to open /dev/vzctl: No such device or address
Please check that vzdev kernel module is loaded and you have sufficient permissions to access the file.

I created /dev/vzctl with "mknode /dev/vzctl c 126 0", but the error remains the same.

I checked if I booted the right kernel:

user@host:~$ uname -a
Linux host 2.6.24-17-openvz #1 SMP Fri Apr 25 04:52:21 UTC 2008 i686 GNU/Linux

When I tried "lsmod | grep vz" it returned nothing. Should there be a openvz kernel module loaded?

Microrechner,

Please try running
$ /etc/init.d/vz status
It should say: OpenVZ is running...

However I'm assuming that in your case it did not start for some reason. If that's the case start it.
$ /etc/init.d/vz start

If there's an error while starting post the result here.

When OpenVZ is running the following modules are loaded into the kernel:
vzethdev, vznetdev, vzdquota, vzmon, vzdev

The /dev/vzctl is created by the init script and you don't need to create it manually.

Hope this helps,
Karoly

linux-image-2.6.24-16-openvz still does have this bug for me!

I tested it twice:
AMD Athlon 2400+ (VIA KM266 chipset).
As a VM inside KVM (on a Intel Core 2 Duo 3.0ghz, Intel Q35 chipset).

Please note using linux-openvz, pull in all kinds of crap like nvidia-kernel-common.

Hi Pascal,

yes we all know that 2.6.24-16-openvz still has this bug, that's why people posted workarounds for this version in the comments above. ;) I guess it will be fixed in 2.6.24-17.

Best regards,
Philipp

SRU Justification:

Impact: The -openvz kernel does not boot.

Fix Description: Pulled updates to openvz custom build.

Patch: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-hardy.git;a=commit;h=174a23141b10d9ecbe6b2ed9556a54f435c50b77

TEST CASE: See first bug description in this report.

Karoly,

thanks a lot - running the init script resolved the problem.

Now I feel kind of stupid for not looking in init.d myself :-)

Thanks again!

Accepted into hardy-proposed.

One question: Will's openvz-kernel has XEN enabled, will this be the case for the official kernel, too? It's unusable on a desktop machine using a nvidia-card and the nvidia-drivers, because they won't build on a xen-enabled kernel ...

It would seem silly to me to have two virtualization solutions enabled simultaneously. For the few folks who actually need this, they can probably maintain their own kernel.

As for the nVidia card. In this case it's silly to have both enabled, but binary drivers are nVidia's problem, you should contact them as well...

It seems like http://git.openvz.org/?p=ubuntu-hardy-openvz;a=commit;h=b5c04a9caa1ccec03150970499973e761de10c9f mentioned in comment #35 has not made it into -17. Was this an accident or intentional? OpenVZ is still rendered useless for me in hardy without that fix.

Yes, Bastian is unfortunately correct: this important fix is not included!

I've created a new bug just for it: https://launchpad.net/bugs/226335

It appears there may be yet another patch required. Using Will Nowak's 2.6.24-17 OpenVZ kernel I am unable to enumerate UDP sockets using netstat. Another user on the OpenVZ forums is apparently reporting the same problem here. I'm not sure if this requires a separate bug to be opened or not but since this bug already has two patches connected to it and I don't see any others on Launchpad for this issue I wanted to mention it here.

The OpenVZ forum post with Git patch:

http://forum.openvz.org/index.php?t=msg&goto=30093&

Example output:

root@hardy-pkgbuilder:/# netstat -aun
Active Internet connections (servers and established)
Proto Recv-Q Send-Q Local Address Foreign Address State
netstat: no support for `AF INET (udp)' on this system.

However this does not seem to impact the actual usage of UDP. I am running what appears to be a functioning DNS server in a test environment, I just can't see the activity via netstat.

Ben: it's not included, yes. The last patchset from OpenVZ is from 25-Mar-2008 (2.6.24-ovz004), while the commits for this are from April.
Upstream bug with links to patches: http://bugzilla.openvz.org/show_bug.cgi?id=860
I'm already building the kernel with this patches to verify it and have created bug 230432 for this issue.

I understand that bug #230432 is present for tracking an additional bug with openvz. Can someone confirm that the 2.6.14-17-openvz package in hardy-proposed does fix the specific "soft lockup" boot issue reported in /this/ bug?

Yes, I can confirm that 2.6.14-17-openvz fixes the mentioned soft lockup bug.

Me too, I have successfully run 2.6.24-17 for a few weeks, last weekend I upgraded to 2.6.24-18, this one runs fine too.

regards,
Harry

ok, thank you; marking as verification-done.

linux 2.6.24-17.31 copied to hardy-updates.

I updated the kernel, but now my openvz-container won't work anymore. I used a own build kernel (as described here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/17 ) before and all worked fine.

The error-message inside of a container with the kernel from ubuntu.com is:

cp: skipping file `xxxx', as it was replaced while being copied

for example, when i want to restart postfix, the file is /etc/ssl/certs/ca-certificate.crt .
Now i went back to my own build kernel and all works fine.

Any ideas why the new kernel will not work for me?

This is a known problem in 2.6.24-16. You need at least 2.6.24-17, see comments above (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/18)

regards,
Harry

Thomas,

It's because a patch for inodes didn't make it into the kernel. The bug was discussed in this page [read back and you'll find the same complaint], but there's also another bug report just for it. You can patch it by yourself, it's incredibly easy. The patch file starts with 0049 and you copy it to debian/binary-custom.d/openvz/patchset, then compile. I'm at work now, so I can't give you all the details.

But yes, that's a known bug.

Harry,

I still had this problem with 2.6.24-17-31, I had to recompile. It's fixed inthe git repository, though.

Hi Folks,

I just wanted to mention that if you compiled a kernel on your own you'd better include the following patches also:
http://launchpadlibrarian.net/14574376/0050-ubuntu-make-mmap-POSIX-conform.patch
http://launchpadlibrarian.net/14524536/0051-cherrypick-relevant-ip4_ip6-proc-entries-ovzbug860.patch
http://launchpadlibrarian.net/14740074/0052-8021q-rmmod-ovzbug885.patch

Just drop them in the openvz/debian/binary-custom.d/openvz/patchset/ folder and compile

References:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/231400/ - mmap POSIX
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/230432 - UDP
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/235207 - 8021q

Cheers,
Karoly

Just for reference, OpenVZ upstream has synced their Hardy repository lately.
This includes a lot of bugfixes (all from the OpenVZ patchset and the ones done in Ubuntu since the last released version).

It's available at: http://git.openvz.org/?p=ubuntu-hardy-openvz;a=summary
Instructions for building it are provided above already (https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/17).

Actually, i just tried to install it. And im still getting the following
This kernel require the following features not present on the CPU:
0:6
unable to boot, please use a kernel apprppriate for your CPU

This means you are trying to boot on a CPU without PAE support. The default kernel is compiled with PAE, so it won't run on a non-PAE CPU :-(

Maybe this one can help you : https://answers.launchpad.net/ubuntu/+source/linux/+question/33381

Greetings all - OpenVZ newbie here, trying to get it installed. I was following instructions posted here https://bugs.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/25:

$ git-clone git://kernel.ubuntu.com/ubuntu/ubuntu-hardy.git openvz
$ cd openvz
Download the patch: http://forum.openvz.org/index.php?t=getfile&id=528&
Drop it in the openvz/debian/binary-custom.d/openvz/patchset/ folder
$ AUTOBUILD=1 NOEXTRAS=1 fakeroot debian/rules custom-binary-openvz
$ cd ..
$ sudo dpkg -i linux-image-2.6.24-16-openvz_2.6.24-16.30_i386.deb

When it's patching though, I get this error:

Hunk #1 succeeded at 6 (offset 4 lines).
debian/binary-custom.d/openvz/patchset/0049-fgetstat-virtinfo-notifier.patch
patching file fs/stat.c
debian/binary-custom.d/openvz/patchset/0049-ubuntu-specific-fgetstat-virtinfo-notifier.patch
patching file fs/stat.c
Reversed (or previously applied) patch detected! Assume -R? [n]

Apply a 'y' gives me this:

debian/binary-custom.d/openvz/patchset/0050-ubuntu-make-mmap-POSIX-conform.patch
patching file mm/mmap.c
Hunk #1 succeeded at 930 (offset 20 lines).
patching file arch/ia64/kernel/sys_ia64.c
cat debian/binary-custom.d/openvz/config.i386 > /home/ktulu/openvz/debian/build/custom-build-openvz/.config
make ARCH=i386 EXTRAVERSION=-19-openvz SUBLEVEL=24 -C /home/ktulu/openvz/debian/build/custom-source-openvz O=/home/ktulu/openvz/debian/build/custom-build-openvz silentoldconfig prepare scripts
make[1]: Entering directory `/home/ktulu/openvz/debian/build/custom-source-openvz'
make[3]: Nothing to be done for `/home/ktulu/openvz/debian/build/custom-source-openvz/Makefile'.
  HOSTCC scripts/basic/fixdep
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:107:23: error: sys/types.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:108:22: error: sys/stat.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:109:22: error: sys/mman.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:110:20: error: unistd.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:111:19: error: fcntl.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:112:20: error: string.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:113:20: error: stdlib.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:114:19: error: stdio.h: No such file or directory
In file included from /usr/lib/gcc/i486-linux-gnu/4.2.3/include/syslimits.h:7,
                 from /usr/lib/gcc/i486-linux-gnu/4.2.3/include/limits.h:11,
                 from /home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:115:
/usr/lib/gcc/i486-linux-gnu/4.2.3/include/limits.h:122:61: error: limits.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-openvz/scripts/basic/fixdep.c:116:19: error: ctype.h: No such file or directory
/home/ktulu/openvz/debian/build/custom-source-...

Hi Anthony,

The git now officially contains the 0049 patch:
debian/binary-custom.d/openvz/patchset/0049-ubuntu-specific-fgetstat-virtinfo-notifier.patch

So you don't need to download it and drop it in the folder. Run:
ls -al debian/binary-custom.d/openvz/patchset/
and make sure that there are no duplicate patches. You should see patches from 0001 to 0050 and only one from each. As the Ubuntu kernel team adds these patches to the ubuntu git the old instructions get outdated.

For example in my previous post:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/210672/comments/73
I mentioned to add the 0050-0052 patches, but as of today 0050 is already part of the git, so you only need to add 0051-0052

Hope this helps,
Karoly

linux-image-2.6.24-18-openvz still has the cp bug.

Karoly,

Thanks for the help. I had a feeling that might have been an issue. I removed the extra patch and made sure I had no duplicates, but when I try to run it I get almost the exact same error messages from my previous post. Is there something I am missing? Thank you for all the help!

-Anthony

Hi Anthony,

Could you check that the following packages are installed:

sudo apt-get install linux-kernel-devel fakeroot build-essential

Karoly

Hmm, or maybe cleaning before starting again?

$ fakeroot debian/rules clean

Paulo

Odd things happening ...
I updated at home to 2.6.24-18-openvz (hardy 64 bit, upgraded bit by bit from hardy alpha) and everything seeme to work fine.
Today I updated my server at work (hardy 64bit, fresh install) and I still got the cp-bug ... I'm trying to investigate some more whats happening ...

Thanks for the help! I don't think I had linux-kernel-devel installed, so I got it to compile and install the package OK. However, I'm having the same problem as the other -openvz binary repository kernels - my NIC does not seem to work. It has an IP and the correct module is loaded, you just can't ping anything. I'm going to try debugging it some to see what the exact differences between the kernels... strange though.

I compiled the kernel from git (thnx Molnar) and it's working.

Please forget me previous comment, the cp-problem still exists. I accidentally tried in the wrong shell-window. This mislead me to think it's working, but I was at the host system.

I compiled the kernel from git (2.6.24-19-openvz) and both bugs (cp/mv and the UDP-bug mentioned above) are gone ... hopefully this will make it into the point release ...

Great!!!

So cool to see that all the outstanding patches have been pulled into the official Ubuntu kernel. We are up to 79 now. ;-) I can't wait to recompile and test. Thanks for the great work, hopefully the 2.6.24-19-openvz is going to be a solid version. I'll report my findings.

Thanks again,
Karoly

Has anyone tried the .19-openvz kernel yet?
I'm sorry I can't personally try it as I have production VPS running in the .17 fix by Will Novak.

It works for me on several production servers. The official v19 contains approximately 15 extra patches, quite important ones. So it is really worth the upgrade.

Cheers,
Karoly

I run it for a week now, no problems so far

regards,
Harry

2008/6/20 YorYor <email address hidden>:

> Has anyone tried the .19-openvz kernel yet?
> I'm sorry I can't personally try it as I have production VPS running in the
> .17 fix by Will Novak.
>
> --
> linux-image-2.6.24-13-openvz refuses to boot
> https://bugs.launchpad.net/bugs/210672
> You received this bug notification because you are a direct subscriber
> of the bug.
>

--
met vriendelijke groet,
Harry Metske
Telnr. +31-548-512395
Mobile +31-6-51898081

Thanks Karoly and Harry,

Your prompt responses are much appreciated.

I've put in an urgent maintenance slot and things seem to be running fine as well.

Perhaps any new issues should have their own bug report filed.

This bug should be considered closed since it now boots with an official -openvz kernel.

WFM on amd64

YorYor wrote:
> Has anyone tried the .19-openvz kernel yet?
>

Hi all,

i installed the package linux-image-2.6.24-19-openvz but the kernel still doesn't boot.
The last entry from /var/log/messages is:

Jul 14 17:26:34 blaze kernel: [ 41.681924] Adding 2104504k swap on /dev/sda1. Priority:-1 extents:1 across:2104504k
Jul 14 17:26:34 blaze kernel: [ 41.685913] EXT3 FS on sda2, internal journal
Jul 14 17:26:34 blaze kernel: [ 42.043757] kjournald starting. Commit interval 5 seconds
Jul 14 17:26:34 blaze kernel: [ 42.043987] EXT3 FS on sdb1, internal journal
Jul 14 17:26:34 blaze kernel: [ 42.043990] EXT3-fs: mounted filesystem with ordered data mode.
Jul 14 17:26:38 blaze kernel: [ 46.381603] Clocksource tsc unstable (delta = -321232839 ns)
Jul 14 17:30:27 blaze exiting on signal 15

I don't know what the problem is.
I use hardy (without proposed or backported sources.list entries).

Hi Thomas,

It seems to me that the reason you are unable to boot is some other issue than this bug ticket was originally created for and actually this bug ticket has been "closed", because the bug has been fixed and a new kernel has been released. So I'd advise you to create an another bug with detailed information, hardware, output of a serial redirected kernel boot, etc...

Cheers,
Karoly

hi,

sorry. it was my fault and now the kernel works well!!!

Cheers,

Tom