Re: [opensuse-factory] systemd status in Factory : WAIT BEFORE UPGRADING TO CURRENT FACTORY
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
From which state did you upgrade your VM ? -- Frederic Crozat <fcrozat@suse.com> SUSE
-- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 11/07/2012 09:34 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
From which state did you upgrade your VM ?
Just factory before the systemd update: systemd-44. I think it is related to libudev. libudev0 is removed, systemd is reexec'ed => boom. Or maybe other way around. systemd 195 is installed while libudev1 is not installed yet; reexec => boom. Is that a possible scenario? - -- js suse labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.19 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://www.enigmail.net/ iQIcBAEBAgAGBQJQmjC8AAoJEL0lsQQGtHBJY5oP/3dz/Imq2AVxpThN7Z7wHzkS xpQrqsMHV8IDfxBusZtBbEiJDx910BHGq/Qz/6U2bfQDRX5JbA5ZnCxM9JtSbLI3 4fM42DT0YJzAi91TPj5UVgGtWLpajyAi0l6ZmKhr6DQxUoMhrkJf7iuvPbmFoOoa Whyn0Zujt21LsV1HaWHsgnr2BC5gVmPQ1vJqFLhwoZebmnMEgnXj9RMpoq/ywlm7 0jHrkBtv2sBJi8QoAWqq1QdHNzOdBOAJTHMHJiFAj6HNwvjFPHnsVXUOSbFs9ZZJ adqfj0eVS2e8NzELsmgdq9NAl8EDDfTjcVpZ+qMvtSoJvi1nXYQ9tku9Cri7n4pj mxvnlDZVe75AMoeBg953xSFoNEKTD0gjIxUue/Iohoz18Wkq3RK/PzK3TbRVVVIa Dn+0OuyLzwYSjaNMzGK6jzRkg3FPTaDuRM73AWpeK9CRAmtVqaBpRDCAFtPQVWlg yfZEv3jJTfDnd0pFBdAW93Qjtzkl5O0z27Pq7e4x29cSuuSNXVQAbbeT7DnRZbSs g1bXqkPlfXzOR/m34ViMgk9PAxValgsOpnUKSvTIHbTYWKhXrBMLAhjFFvihpNtD dtUR+Cax9oMoPmt0SxTUQpvh8lI59T6Q0NDE1nMZH5VOP3galX0fm8bPdB1i/AqG TO70oAZwA3EqZHsZxyOd =drEh -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 11/07/2012 03:34 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
I encountered the same problem, then after restarting the VM I get: Initramfs unpacking failed: uncompression error Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) Pid: 1, comm: swapper/0 Not tainted 3.6.3-1-desktop #1 Call Trace: dump_trace+0x88/0x300 dump_stack+0x69/0x6f panic+0xc5/0x1d1 mount_block_root+0x1c6/0x266 prepare_namespace+0x138/0x16e kernel_init+0x1cf/0x1d4 kernel_thread_helper+0x4/0x10
From which state did you upgrade your VM ?
From just before teh change to the new systemd package. Later, Robert -- Robert Schweikert MAY THE SOURCE BE WITH YOU SUSE-IBM Software Integration Center LINUX Tech Lead rjschwei@suse.com rschweik@ca.ibm.com 781-464-8147 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Le mercredi 07 novembre 2012 à 10:19 -0500, Robert Schweikert a écrit :
On 11/07/2012 03:34 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
I encountered the same problem, then after restarting the VM I get:
Initramfs unpacking failed: uncompression error Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) Pid: 1, comm: swapper/0 Not tainted 3.6.3-1-desktop #1 Call Trace: dump_trace+0x88/0x300 dump_stack+0x69/0x6f panic+0xc5/0x1d1 mount_block_root+0x1c6/0x266 prepare_namespace+0x138/0x16e kernel_init+0x1cf/0x1d4 kernel_thread_helper+0x4/0x10
You need to reinstall bootloader from a live-cd. I've been able to reproduce the panic on upgrade one time, but now, I need a way to debug it :( -- Frederic Crozat <fcrozat@suse.com> SUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 11/07/2012 10:28 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 10:19 -0500, Robert Schweikert a écrit :
On 11/07/2012 03:34 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
I encountered the same problem, then after restarting the VM I get:
Initramfs unpacking failed: uncompression error Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) Pid: 1, comm: swapper/0 Not tainted 3.6.3-1-desktop #1 Call Trace: dump_trace+0x88/0x300 dump_stack+0x69/0x6f panic+0xc5/0x1d1 mount_block_root+0x1c6/0x266 prepare_namespace+0x138/0x16e kernel_init+0x1cf/0x1d4 kernel_thread_helper+0x4/0x10
You need to reinstall bootloader from a live-cd.
I've been able to reproduce the panic on upgrade one time, but now, I need a way to debug it :(
Well, I fixed up the initrd such that it unpacks properly, but hen I end up with the original kernel issue, i.e. something is trying to kill init. The VM is 12G I can put it somewhere for you to download if that will help. Robert -- Robert Schweikert MAY THE SOURCE BE WITH YOU SUSE-IBM Software Integration Center LINUX Tech Lead rjschwei@suse.com rschweik@ca.ibm.com 781-464-8147 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Le mercredi 07 novembre 2012 à 10:52 -0500, Robert Schweikert a écrit :
On 11/07/2012 10:28 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 10:19 -0500, Robert Schweikert a écrit :
On 11/07/2012 03:34 AM, Frederic Crozat wrote:
Le mercredi 07 novembre 2012 à 09:09 +0100, Jiri Slaby a écrit :
On 11/05/2012 11:42 AM, Frederic Crozat wrote:
Until all those issues are fixed (I'll sent a follow-up mail here), please DON'T UPGRADE YOUR FACTORY INSTALLATION.
Hi, I tried in a VM though. Installing systemd 195 kills process 1 (init) which indeed causes a panic: Installing: systemd-195-4.1 ....................................<100%> Kernel panic - not syncing: Attempted to kill init! exitcode=0x00007f00
And unfortunately, this is something I can't reproduce reliably: got it one time while doing a full upgrade with 195-3.1 and I thought fixing the various udev instances running when upgrading would fix it but apparently, it doesn't.
I encountered the same problem, then after restarting the VM I get:
Initramfs unpacking failed: uncompression error Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(0,0) Pid: 1, comm: swapper/0 Not tainted 3.6.3-1-desktop #1 Call Trace: dump_trace+0x88/0x300 dump_stack+0x69/0x6f panic+0xc5/0x1d1 mount_block_root+0x1c6/0x266 prepare_namespace+0x138/0x16e kernel_init+0x1cf/0x1d4 kernel_thread_helper+0x4/0x10
You need to reinstall bootloader from a live-cd.
I've been able to reproduce the panic on upgrade one time, but now, I need a way to debug it :(
Well, I fixed up the initrd such that it unpacks properly, but hen I end up with the original kernel issue, i.e. something is trying to kill init. The VM is 12G I can put it somewhere for you to download if that will help.
I don't think I need the VM, I just need to finish downloading M0 DVD image to be able to go back and forth between old and new systemd/udev :) -- Frederic Crozat <fcrozat@suse.com> SUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 11/07/2012 04:52 PM, Robert Schweikert wrote:
Well, I fixed up the initrd such that it unpacks properly, but hen I end up with the original kernel issue, i.e. something is trying to kill init.
Yeah I got that too as a result. It told me that libudev0 (yes, 0) is not existing right before dying. Because I had both systemd-44 _and_ system-195 installed, but libudev0 was gone, and systemd-44 not replaced by 195 yet. Well, there was more fun than that. grub2 spitted at me that there is no normal mode available, so I had to set prefix to /usr/lib/grub2 and insmod normal module manually from the grub2 rescue shell (as all the files from the grub2 dir were gone). Then I was not allowed to boot because of partly installed systemd => I fixed that from 11sp2 with - -hdb qemu parameter with the broken factory filesystem. And then I couldn't log on. It was because somebody emptied my /etc/pam.d/common-session file. It took me a while to figure that out because rsyslog was not started, as syslog.service was gone. And the pure fucking systemd journal shit was not actually logging anything (as expected). Some guys from redhat should really pull their heads from their asses. They convinced me at the presentation here at the linuxcon EU that they have no idea what they are doing. One example of that is a statement that all drivers and hardware that pm-utils have quirks for are already fixed. People, they have NO idea what they are doing, do NOT listen to them, neither follow them! kind regards, - -- js suse labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.19 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://www.enigmail.net/ iQIcBAEBAgAGBQJQmwaQAAoJEL0lsQQGtHBJ0rsP/i1w/upuY2vw+ZHVByUmt/Cw gJoR+KIC+4gSCNZpCGqtYQM2lgjrpWZQ+JOiboS/qb9oA8Y3aLF7SveyNFtKBuUX 0FhLwOsEci0f1jUO2ggMLo1JN+FBJQBuUXeena3f0iA0kG4MF5IVN34CW3NPdoi6 +SqVXMFZq4JshUX4udR5ZIGsvwq7/oF7Ke8QQ1vRGHur+k2fb1LZcpajVYg9q2Ug RsFYa1g1u4+o4Lrg01DLSkQ9l1rWn+LvSyIyD5RwmFDCE1zSHyvChcyy6NSLDvOb ST7axe5u5JErZEt3etfhj0b02Jq/fIfrwhbxNP6RNf5dlvmKgOVmqmgZjiyQPJI0 USx9/E6diENIo4QDxpbt3Jp/DVZzio6aamsveBhTlVRYgHEnZi3sAsLcusDOPdWM J0gx/iD0cdCPncSuB2haANu1ZWuoa+dHMxSZQT2igM0WlB78NSE1RqllclLnND0t dBM+/m869LGzLAYX5R9BrEfjysptNlB56IkpkbvKNMH4Qo3ob0UnnAiLjAla/stQ 1sbJ4TPzsrd8XL3NausSHsNT+0UQWDHCF9LEq8UEtSfs9/vYWeOuoE38ID0TLJBT 8WgFSKAayRLhiuGyy89UwV1MwzsYEPykf3xKvXoCbtXhS5YSusZCOJsuObjRcJr1 Ge55pNmKnyzeARgJau4h =w5YJ -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Le jeudi 08 novembre 2012 à 02:10 +0100, Jiri Slaby a écrit :
On 11/07/2012 04:52 PM, Robert Schweikert wrote:
Well, I fixed up the initrd such that it unpacks properly, but hen I end up with the original kernel issue, i.e. something is trying to kill init.
Yeah I got that too as a result. It told me that libudev0 (yes, 0) is not existing right before dying. Because I had both systemd-44 _and_ system-195 installed, but libudev0 was gone, and systemd-44 not replaced by 195 yet.
Well, there was more fun than that. grub2 spitted at me that there is no normal mode available, so I had to set prefix to /usr/lib/grub2 and insmod normal module manually from the grub2 rescue shell (as all the files from the grub2 dir were gone). Then I was not allowed to boot because of partly installed systemd => I fixed that from 11sp2 with -hdb qemu parameter with the broken factory filesystem.
I've been able to reproduce (and understand) the grub2 issue : - zypper is removing libudev0 (due to libudev0 requiring udev of the same version, blame greg KH for this broken dependency, it should have been >= ;) - then udev is updated (before libudev1 is installed) and call mkinitrd which will call update-bootloader which calls grub2 which call grub2-probe which is linked to device-mapper which has a dependency on libudev.so.0 (which, at that time, is no longer present) => crash of grub2 install. I'll see how to fix this (we probably will ressurect libudev0 since some packages like chrome are still linked to it but it might be worth to fix this dependency hell properly) -- Frederic Crozat <fcrozat@suse.com> SUSE -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Thu, Nov 08, 2012 at 01:37:53PM +0100, Frederic Crozat wrote:
Le jeudi 08 novembre 2012 à 02:10 +0100, Jiri Slaby a écrit :
On 11/07/2012 04:52 PM, Robert Schweikert wrote:
Well, I fixed up the initrd such that it unpacks properly, but hen I end up with the original kernel issue, i.e. something is trying to kill init.
Yeah I got that too as a result. It told me that libudev0 (yes, 0) is not existing right before dying. Because I had both systemd-44 _and_ system-195 installed, but libudev0 was gone, and systemd-44 not replaced by 195 yet.
Well, there was more fun than that. grub2 spitted at me that there is no normal mode available, so I had to set prefix to /usr/lib/grub2 and insmod normal module manually from the grub2 rescue shell (as all the files from the grub2 dir were gone). Then I was not allowed to boot because of partly installed systemd => I fixed that from 11sp2 with -hdb qemu parameter with the broken factory filesystem.
I've been able to reproduce (and understand) the grub2 issue : - zypper is removing libudev0 (due to libudev0 requiring udev of the same version, blame greg KH for this broken dependency, it should have been >= ;)
Sorry about that :( greg k-h -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
participants (4)
-
Frederic Crozat -
Greg KH -
Jiri Slaby -
Robert Schweikert