| Summary: | Kernel panic on random yum install or update commands | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Matt Darcy <matt> |
| Component: | kernel | Assignee: | Kernel Maintainer List <kernel-maint> |
| Status: | CLOSED WONTFIX | QA Contact: | Fedora Extras Quality Assurance <extras-qa> |
| Severity: | high | Docs Contact: | |
| Priority: | low | ||
| Version: | 14 | CC: | dekacy10, ffesti, gansalmon, itamar, james.antill, jonathan, kernel-maint, madhu.chinakonda, maxamillion, pmatilai, tla |
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | x86_64 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2011-08-29 20:56:50 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Attachments: | |||
|
Description
Matt Darcy
2011-01-13 21:52:13 UTC
I should also point out this machine is in a standard configuration, no overclocking, fancy cooling, etc. It has passed 4 compelte cycles on memtest without error. I would be happy to provide login credentials to bug resolvers to assist with debugging. (In reply to comment #2) > Actual results: > kernel panic, stack trace and reboot Did you forget to attach the panic messages and stack trace? no, the complete message scrolls off screen quite fast, and the stack trace to screen is quite long and garbled, even disabling the auto reboot I have no method of dumping the message and trace to file as the machine is totally hung. If there is a process to capture this (maybe using netdump ?) I'd be happy to capture the information for you. I've now got the kdump service running so I'll attatch some files on the next crash. beyond a vmcore from kdump is there anything else you'd like ? Created attachment 473872 [details]
Kernel Crash Dump from yum install
this come from a 1.9GB vmcore file dumped on the panic, I can provide this but need a method of getting 1.9GB to you in bugzilla
PANIC: "[183212.409984] Kernel panic - not syncing: Fatal Machine check" This is most likely a real hardware problem, though it could also be a bug in the MCE code. You should be able to get a lot more information from the 'crash' program, e.g. the log command will dump the contents of the kernel log. The last few lines may have additional information about the error. You may want to dump the entire log and attach it to this bug report. Created attachment 474364 [details]
log output from kernel crash file
details of the crash log do seem to reference a hardware problem as suggested.
There are a few references to the video card, and then the panic.
While I agree this looks like a hardware issue, I cannot get this fault to occur on a Fedora 13 machine, and it never happened when this machine ran Fedora 12. I cannot get this issue to happen on Fedora 14 before I do an update to the current kernel, which suggests to me at a high level that there is something in one of the updates to the kernel that does not like a hardware device, rather than a phyiscal hardware failure.
I've attatched the log for viewing and I'll attempt to get more information out the panic.
Can you attach the boot log from a kernel that doesn't hit this problem? I'll need to rebuild this machine as a Fedora 13 or non-updated Fedora 14 machine. this will take a little time. Created attachment 476031 [details]
Boot log from a stable clean Fedora 13 x86_64 base install
this is a clean re-install of Fedora 13 x86_64 on the same machine, no updates, fresh from the install. This is the boot log as requested
Created attachment 476032 [details]
Boot log from a stable updated Fedora 13 x86_64 (2.6.34.7-66.fc13.x86_64)
As requested a boot log from a fully up to date Fedora 13 install updated to the latest current Fedora 13 package set as of the time of writing.
2.6.34.7-66.fc13.x86_64
Fedora 14 logs will be coming shortly.
Created attachment 476044 [details]
Boot log from a stable clean Fedora 14 x86_64 base install
as requested a boot log from a clean Fedora 14 install
Created attachment 476045 [details]
Boot log from an updated Fedora 14 x86_64 (2.6.35.10-74.fc14.x86_64)
a boot look from the updated to current (as of time of posting) Fedora 14 box.
This machine is now back in the configuration where the yum install commands randomly cause the panic
I am wondering is this is related with the problem I faced just know. yum will return the following and then exit: sudo yum update [sudo] password for dimitris: Loaded plugins: fastestmirror, langpacks, presto, refresh-packagekit Adding el_CY to language list Loading mirror speeds from cached hostfile updates/metalink | 6.3 kB 00:00 * fedora: ftp.itu.edu.tr * rpmfusion-free: mirror.yandex.ru * rpmfusion-free-updates: mirror.yandex.ru * rpmfusion-nonfree: mirror.yandex.ru * rpmfusion-nonfree-updates: mirror.yandex.ru * updates: ftp.rhd.ru adobe-linux-i386 | 951 B 00:00 home_Gnurou | 1.7 kB 00:00 rpmfusion-free-updates | 3.3 kB 00:00 rpmfusion-nonfree-updates | 3.3 kB 00:00 updates | 4.7 kB 00:00 Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894492] Oops: 0000 [#3] SMP Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894495] last sysfs file: /sys/devices/system/cpu/cpu2/cache/index2/shared_cpu_map Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894590] Stack: Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894603] Call Trace: Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894641] Code: ff 74 1a 4c 89 06 49 2b 4b 08 eb 05 49 39 d8 72 b4 48 89 0a 31 c0 eb 05 b8 ea ff ff ff 5b 41 5c c9 c3 55 48 89 e5 0f 1f 44 00 00 <8b> 57 08 48 8d 4f 08 eb 02 89 c2 85 d2 74 14 8d 72 01 89 d0 f0 Message from syslogd@innovator at Jan 31 07:05:26 ... kernel:[384473.894681] CR2: ffffea400110fdd8 less than 24 hours after updating to the current Fedora 14 kernel (2.6.35.10-74.fc14.x86_64) a yup update on cifs-tools that was recently released (just the one package was being downloaded) caused a kernel panic and reboot as before (In reply to comment #13) > Created attachment 476045 [details] > Boot log from an updated Fedora 14 x86_64 (2.6.35.10-74.fc14.x86_64) > > a boot look from the updated to current (as of time of posting) Fedora 14 box. > > This machine is now back in the configuration where the yum install commands > randomly cause the panic We need the kernel boot log (/var/log/dmesg). No need to upload the log from the current broken kernel, we have that from kdump, and there's no need for anything from F13 just yet. Just boot the working F14 kernel and then upload /var/log/dmesg from that. Created attachment 476423 [details]
/var/log/dmesg output from the stable F14 kernel
boot log from /var/log/dmesg from the stable F14 kernel.
after more time on the stable F14 kernel I've managed to reproduce this crash with a yum install. I should point out that this is on a Fedora 14 machine that is fully up to date, but using the earlier stable Fedora 14. I assume you'd like this re-installed to a Fedora 14 stock CD install and re-tested. The problem I have is that when I do yum install from the earlier Fedora 14 base install it will update dependencies. is there more information I can provide on this as this machine is still crashing on a regular basis and is now out of use. suggest closing this bu as it doesn't apper to be going anywhere. As the bug raiser I am happy for it to beclosed Closing per comment #20 |