Bug 483681 - Fedora kernels seem unstable; crash in 2.6.27.12-170.2.5.fc10.x86_64
Summary: Fedora kernels seem unstable; crash in 2.6.27.12-170.2.5.fc10.x86_64
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: 10
Hardware: x86_64
OS: Linux
low
high
Target Milestone: ---
Assignee: Kernel Maintainer List
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2009-02-02 23:17 UTC by Chris Siebenmann
Modified: 2009-02-06 00:47 UTC (History)
3 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2009-02-06 00:47:39 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
serial console log of the panic (8.12 KB, text/plain)
2009-02-02 23:17 UTC, Chris Siebenmann
no flags Details

Description Chris Siebenmann 2009-02-02 23:17:01 UTC
Created attachment 330691 [details]
serial console log of the panic

Description of problem:
Since upgrading from Fedora 8 to Fedora 10, my machine has been relatively
unstable, with frequent kernel crashes and lockups. I have captured the
latest one on a serial console in the hopes that maybe the problems can
be identified and resolved.

Some earlier crashes that I can find logs for appear to be roughly the
same.

Version-Release number of selected component (if applicable):
kernel-2.6.27.12-170.2.5.fc10.x86_64

(for this crash; I have had crashes in kernel-2.6.27.9-159.fc10.x86_64
as well.)

Hardware:
Asus M2N4-SLI motherboard, Athlon X2 4600+, 6 GB of memory, ATI Radeon
X300 PCIE video card. I believe that I've had crashes both with and
without 'nomodeset' as a kernel command line parameter, but this crash
occurred with 'nomodeset'.

(I use 'nomodeset' because if I do not, X eventually slows down and
becomes achingly slow until the system is rebooted; quitting and
restarting the X server makes no difference.)

Call trace taken from the panic log in the attachment:
RIP: 0010:[<ffffffff81020a11>]  [<ffffffff81020a11>] smp_apic_timer_interrupt+0xa6/0xa8
[...]
Call Trace:
 <IRQ>  [<ffffffff810113d8>] ? apic_timer_interrupt+0x88/0x90
 <EOI>  [<ffffffff8102571e>] ? native_safe_halt+0x6/0x8
 [<ffffffff810172cb>] ? need_resched+0x1e/0x28
 [<ffffffff810173b0>] ? default_idle+0x2a/0x4c
 [<ffffffff81017500>] ? c1e_idle+0x120/0x127
 [<ffffffff81336202>] ? atomic_notifier_call_chain+0x13/0x15
 [<ffffffff8100f279>] ? cpu_idle+0xb2/0x10b
 [<ffffffff8132db8c>] ? start_secondary+0x16e/0x173

(I am inlining this bit for easier searching in Bugzilla.)

Comment 1 Chuck Ebbert 2009-02-04 00:40:59 UTC
Your system almost certainly has a hardware problem, probably bad memory but it could also be overheating or have a defective CPU.

It is trying to restore the stack pointer to an illegal address: feff88019f0dffa8

This is exactly one bit different from a valid address, indicating that a bit was flipped.

Comment 2 François Cami 2009-02-06 00:47:39 UTC
Closing as NOTABUG since it is a hardware problem, not a bug in Fedora.


Note You need to log in before you can comment on or make changes to this bug.