Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Patrick, Also it possible is a kernel stack overflow. Disable CONFIG_4KSTACKS option in your kernel config. -- signature -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Patrick On Fri, May 30, 2008 at 09:33:35AM +0200, Patrick Schoenfeld wrote: Hi Ola, On Fri, May 30, 2008 at 08:10:35AM +0200, Ola Lundqvist wrote: I have not got any response from upstream yet. Good to know that this problem is reproduceable on a newer kernel as well. not really. I would have hoped that the newer patch fixes my problem. :-) Of couse. :) What I ment that it is good to know the status, even if the information is bad in itself. ;) It is a little bit strange that no-one else has seen this. On the other hand few people seem to run openvz on amd64. Agreed. I'm also a little bit puzzled, because I run the patch on several systems and this is the only one were the problems occur. However it is not on amd64. The data in the first report might be wrong, cause I probably filed it on my desktop system. Oh I see. Good information. This makes me suspect hardware related. Not necessarily a fault but maybe something in the patch that is incompatible. Are you running some other system with identical hardware, where you do not have the crash. Or even though it is not identical, is it similar? The system were it runs on is a Intel(R) Celeron(R) CPU 2.40GHz according to /proc/cpuinfo Ok, yes this is good information. I would like to know as much as you can tell about the hardware, related especially to the memory handling. It can very well be a timing issue. * Number of CPU:s * Amount of memory * Memory type and speed * Anything else that you can find that you think can be interesting Tell me, if you need more information and sorry for the confusion. BTW. I suspected first that this is hardware related, but my hoster did an extensive test of the hardware and found no problem. The evidence is still not gone, but I don't have anything to slap the hoster with currently. The panic does not really indicate something hw related. It is indicating something with paging, which relates to memory handling. It do not necessarily need to be a fault, but could be something related. Best regards, // Ola Best Regards, Patrick -- --- Inguza Technology AB --- MSc in Information Technology / [EMAIL PROTECTED]Annebergsslingan 37\ | [EMAIL PROTECTED] 654 65 KARLSTAD| | http://inguza.com/Mobile: +46 (0)70-332 1551 | \ gpg/f.p.: 7090 A92B 18FE 7994 0C36 4FE4 18A1 B1CF 0FE5 3DD9 / --- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Ola, I decode oops -- page fault occurs at tsk-mm dereference in do_page_fault function (arch/i386/mm/fault.c:353) current task pointer is incorrect, perhaps something bad happened with vcpu scheduler. Patrick, May you try boot with our precompiled rhel5 based openvz kernel? I think it almost debian compatible, extract it from rpm (by rpm2cpio) or use alien and generate initrd. I can explain all steps, if needed. http://download.openvz.org/kernel/branches/rhel5-2.6.18/current/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Konstantin, On Fri, May 30, 2008 at 07:45:19PM +0400, Konstantin Khlebnikov wrote: I decode oops -- page fault occurs at tsk-mm dereference in do_page_fault function (arch/i386/mm/fault.c:353) current task pointer is incorrect, perhaps something bad happened with vcpu scheduler. thanks for your input on this issue. Do you think this could be a bug? Or do you think this could be caused by - for example - bad memory? Because I have my memory in suspicioun as system seems to detect only ~ 350 MB memory out of 512M while lshw even says that the RAM module is a 256MB module only. Patrick, May you try boot with our precompiled rhel5 based openvz kernel? I think it almost debian compatible, extract it from rpm (by rpm2cpio) I could do so for testing but I would not feel good doing so for production use. What do you intend to test with that? Why should it be different? Best Regards, Patrick -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Patrick Schoenfeld wrote: Hi Konstantin, On Fri, May 30, 2008 at 07:45:19PM +0400, Konstantin Khlebnikov wrote: I decode oops -- page fault occurs at tsk-mm dereference in do_page_fault function (arch/i386/mm/fault.c:353) current task pointer is incorrect, perhaps something bad happened with vcpu scheduler. thanks for your input on this issue. Do you think this could be a bug? Or do you think this could be caused by - for example - bad memory? Because I have my memory in suspicioun as system seems to detect only ~ 350 MB memory out of 512M while lshw even says that the RAM module is a 256MB module only. Maybe. Test it with memtest and optionally combined with cpuburn. http://wiki.openvz.org/Hardware_testing Patrick, May you try boot with our precompiled rhel5 based openvz kernel? I think it almost debian compatible, extract it from rpm (by rpm2cpio) I could do so for testing but I would not feel good doing so for production use. What do you intend to test with that? Why should it be different? That is different patches, maybe something was broken while porting to debian kernel. Best Regards, Patrick -- To unsubscribe, send mail to [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: Fwd: Re: Bug#482657: kernel-patch-openvz causes a kernel panic
Forgotten to CC the bugtracker. ---BeginMessage--- Hi Ola, On Tue, May 27, 2008 at 10:38:24PM +0200, Ola Lundqvist wrote: Ok, good to know that so far. Tell me if you see the problem again. it happened faster as expected... unfortune. Here is the new kernel panic log. Have you gotten any feedback from upstream? For me this is very critical, because currently the panics happen almost every day, blocking the server for some time, till I reboot the system. And I'm not very lucky with the need to cold reset the system cause I fear potential data loss. Best Regards, Patrick _conntrack_amanda vzethdev vznetdev simfs vzrst ip_nat vzcpt ip_conntrack tun vzdquota vzmon vzdev af_packet xt_tcpudp xt_length ipt_ttl xt_tcpmss ipt_TCPMSS iptable_mangle iptable_filter xt_multiport xt_limit ipt_tos ipt_REJECT ip_tables x_tables dummy i2c_i801 shpchp i2c_core xfs e100 mii thermal processor fan CPU:0, VCPU: 1768697699.1819113570 EIP:0060:[c011464a]Not tainted VLI EFLAGS: 00010046 (2.6.18+openvz #2 028stab053.5) EIP is at do_page_fault+0x5a/0x5a5 eax: cde72000 ebx: ecx: 007b edx: esi: 00030001 edi: 0094 ebp: 0006 esp: cde72050 ds: 007b es: 007b ss: 0068 BUG: unable to handle kernel paging request at virtual address 7261687b printing eip: c010422d *pde = Recursive die() failure, output suppressed Kernel panic - not syncing: Attempted to kill the idle task! BUG: unable to handle kernel paging request at virtual address 7261716f printing eip: c01183a3 *pde = Oops: [#3] Modules linked in: ipt_LOG xt_state iptable_nat ip_nat_tftp ip_nat_irc ip_nat_ftp ip_nat_amanda xt_conntrack ip_conntrack_tftp ip_conntrack_irc ip_conntrack_ftp ts_kmp ip_conntrack_amanda vzethdev vznetdev simfs vzrst ip_nat vzcpt ip_conntrack tun vzdquota vzmon vzdev af_packet xt_tcpudp xt_length ipt_ttl xt_tcpmss ipt_TCPMSS iptable_mangle iptable_filter xt_multiport xt_limit ipt_tos ipt_REJECT ip_tables x_tables dummy i2c_i801 shpchp i2c_core xfs e100 mii thermal processor fan CPU:0, VCPU: 1768697699.1819113570 EIP:0060:[c01183a3]Not tainted VLI EFLAGS: 00010002 (2.6.18+openvz #2 028stab053.5) EIP is at account_system_time+0x23/0xc0 eax: 72616873 ebx: 11425ae0 ecx: edx: 0001 esi: 11425ae0 edi: ebp: c058bf90 esp: c058bf88 ds: 007b es: 007b ss: 0068 BUG: unable to handle kernel paging request at virtual address 7261687b printing eip: c010422d *pde = Oops: [#4] Modules linked in: ipt_LOG xt_state iptable_nat ip_nat_tftp ip_nat_irc ip_nat_ftp ip_nat_amanda xt_conntrack ip_conntrack_tftp ip_conntrack_irc ip_conntrack_ftp ts_kmp ip_conntrack_amanda vzethdev vznetdev simfs vzrst ip_nat vzcpt ip_conntrack tun vzdquota vzmon vzdev af_packet xt_tcpudp xt_length ipt_ttl xt_tcpmss ipt_TCPMSS iptable_mangle iptable_filter xt_multiport xt_limit ipt_tos ipt_REJECT ip_tables x_tables dummy i2c_i801 shpchp i2c_core xfs e100 mii thermal processor fan CPU:0, VCPU: 1768697699.1819113570 EIP:0060:[c010422d]Not tainted VLI EFLAGS: 00010086 (2.6.18+openvz #2 028stab053.5) EIP is at show_registers+0x15d/0x260 eax: 11425ae0 ebx: 0002 ecx: c058b000 edx: 72616873 esi: 00010002 edi: c058bf54 ebp: c058bf88 esp: c058be9c ds: 007b es: 007b ss: 0068 BUG: unable to handle kernel paging request at virtual address 7261687b printing eip: c010422d *pde = Recursive die() failure, output suppressed Kernel panic - not syncing: Fatal exception in interrupt ---End Message---
Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Patrick Ok, good to know that so far. Tell me if you see the problem again. Best regards, // Ola Quoting Patrick Schoenfeld [EMAIL PROTECTED]: Hi, On Sun, May 25, 2008 at 02:09:44PM +0200, Ola Lundqvist wrote: It would be good to know if you have the same problem with the kernel version in unstable. The reason is that, that version has really a lot of bugs fixed. It will apply fine to the kernel in stable (as far as I have tested at least). I built the kernel with the latest unstable patch applied and it builts fine. However to see if the error happens again needs some time, because it occurs sporadic and not in a regular fashion. Best Regards, Patrick -- --- Ola Lundqvist systemkonsult --- M Sc in IT Engineering / [EMAIL PROTECTED] Annebergsslingan 37\ | [EMAIL PROTECTED] 654 65 KARLSTAD| | http://opalsys.net/ Mobile: +46 (0)70-332 1551 | \ gpg/f.p.: 7090 A92B 18FE 7994 0C36 4FE4 18A1 B1CF 0FE5 3DD9 / --- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Hi Patrick It would be good to know if you have the same problem with the kernel version in unstable. The reason is that, that version has really a lot of bugs fixed. It will apply fine to the kernel in stable (as far as I have tested at least). Best regards, // Ola On Sat, May 24, 2008 at 11:40:03AM +0200, Patrick Schoenfeld wrote: Package: kernel-patch-openvz Version: 028.18.1+etch6 Severity: critical Hi, the OpenVZ kernel patch causes random kernel panics on a production server. Unfortunately I couldn't capture the whole output of the panic ofter the serial console (because screen does not seem to capture if the screen is not attached :(). The file that I will attach in a few minutes is all I got. I set this to severity serious because it totally breaks my system (it panics and is unresponsive after that, till I reset it) and also is responsible for potential data loss (no syncing happens, kernel needs to be reset without flushing buffers). There is another issue with IPv6 beeing enabled and the OpenVZ patches. It then panics on reboot, causing the reboot to fail. But that just as an unrelated sidenote (the current .config has IPv6 disabled). The issue did not happen with a stock kernel. Any ideas? Best Regards, Patrick -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 2.6.24-1-686 (SMP w/1 CPU core) Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash -- - Ola Lundqvist --- / [EMAIL PROTECTED] Annebergsslingan 37 \ | [EMAIL PROTECTED] 654 65 KARLSTAD | | http://inguza.com/ +46 (0)70-332 1551 | \ gpg/f.p.: 7090 A92B 18FE 7994 0C36 4FE4 18A1 B1CF 0FE5 3DD9 / --- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#482657: kernel-patch-openvz causes a kernel panic
Package: kernel-patch-openvz Version: 028.18.1+etch6 Severity: critical Hi, the OpenVZ kernel patch causes random kernel panics on a production server. Unfortunately I couldn't capture the whole output of the panic ofter the serial console (because screen does not seem to capture if the screen is not attached :(). The file that I will attach in a few minutes is all I got. I set this to severity serious because it totally breaks my system (it panics and is unresponsive after that, till I reset it) and also is responsible for potential data loss (no syncing happens, kernel needs to be reset without flushing buffers). There is another issue with IPv6 beeing enabled and the OpenVZ patches. It then panics on reboot, causing the reboot to fail. But that just as an unrelated sidenote (the current .config has IPv6 disabled). The issue did not happen with a stock kernel. Any ideas? Best Regards, Patrick -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'stable') Architecture: i386 (i686) Kernel: Linux 2.6.24-1-686 (SMP w/1 CPU core) Locale: LANG=de_DE.UTF-8, LC_CTYPE=de_DE.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/dash -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
