[Pkg-xen-devel] Bug#665433: xen hypervisor FATAL PAGE FAULT after linux kernel BUG: unable to handle kernel paging request

[Ian Campbell]

Control: tag -1 +moreinfo

Hi Daniel,

You say you weren't able to reproduce back then, so I suspect I know what the
answer will be, but are you able to reproduce this with the Xen packages from
Wheezy and/or Jessie?

If this was just a one-off with the Squeeze version of the hypervisor then I'd
be inclined to close it by now.

Thanks,
Ian.


On Fri, Mar 23, 2012 at 11:12:53PM -0400, Daniel Kahn Gillmor wrote:
> Package: xen-hypervisor-4.0-i386
> Version: 4.0.1-4
> 
> This is the same machine referenced in http://bugs.debian.org/665413.
> It is an HP d530 SFF workstation (model DG784A) with 4GiB of RAM.  It
> has run for years on the lenny xen + linux stack.  If my notes are
> correct, it also ran smoothly for about a year with the lenny hypervisor
> + kernel and the squeeze userland.
> 
> Rebooting into squeeze xen + linux makes the machine very crashy.  The
> dom0 kernel often crashes, and the xen hypervisor also crashes.
> 
> This is a report of a hypervisor crash (i have not been able to
> reproduce it, because the machine does not actually reboot in 5 seconds
> as claimed, and physical access to hard-reset the box is more
> time-consuming than i'd like):
> 
> [  471.094798] BUG: unable to handle kernel paging request at c1194b18
> [  471.094827] IP: [<c104fe70>] prepare_creds+0x39/0x82
> [  471.094850] *pdpt = 000000001dd86001 *pde = 0000000002e31067 *pte = 0000000001194061 
> [  471.094875] Oops: 0003 [#2] SMP 
> [  471.094890] last sysfs file: /sys/module/mii/initstate
> (XEN) ----[ Xen-4.0.1  x86_32p  debug=n  Not tainted ]----
> (XEN) CPU:    0
> (XEN) EIP:    e008:[<0fff0000>] ???
> (XEN) EFLAGS: 00010092   CONTEXT: hypervisor
> (XEN) eax: ffb80280   ebx: 00000004   ecx: 00000000   edx: 00000004
> (XEN) esi: ff32bf8c   edi: ff32bfb8   ebp: ff32bf88   esp: ff32becc
> (XEN) cr0: 8005003b   cr4: 000006f0   cr3: d1480000   cr2: 0fff0000
> (XEN) ds: e010   es: e010   fs: 00d8   gs: 00e0   ss: e010   cs: e008
> (XEN) Xen stack trace from esp=ff32becc:
> (XEN)    ff15004d 00000004 00000004 00000000 ffbd8000 00000000 feda8970 00000000
> (XEN)    0000000b ffb80280 ff32bfb4 ff32bfb4 ffb80280 ff14e079 00000004 2f69696d
> (XEN)    6cb90e00 0000006e c3246050 39194000 00000000 f7bbda34 ff111a16 00000001
> (XEN)    00000001 00000000 00007ff0 00000000 00000004 c3246050 00000020 00000000
> (XEN)    ffb8029c ffbd8000 0972eb40 ff226000 ff32bfb4 c1194b18 ff32bfb4 c1194b18
> (XEN)    ff16963b 00003ffc 0000e010 0000e010 0000ef29 ff147642 ff32bf88 00003ffc
> (XEN)    0000e010 c14800fc 0000007b 0000007b 0000ef29 00000011 00f10000 ff1ccc93
> (XEN)    0000e008 00000203 00003ffc 00000000 c14800fc c1489fbc 0000ef29 0000ef29
> (XEN)    00000011 000e0003 c1002227 00000061 00000246 de6b5cfc 00000069 0000007b
> (XEN)    0000007b 000000d8 000000e0 00000000 ffbd8000
> (XEN) Xen call trace:
> (XEN)    [<0fff0000>] ???
> (XEN)    [<ff15004d>] move_masked_irq+0x9d/0xc0
> (XEN)    [<ff14e079>] do_IRQ+0x89/0x650
> (XEN)    [<ff111a16>] do_multicall+0x156/0x2c0
> (XEN)    [<ff16963b>] do_page_fault+0x10b/0x320
> (XEN)    [<ff147642>] common_interrupt+0x52/0x60
> (XEN)    [<ff1ccc93>] hypercall+0x53/0x9b
> (XEN)    
> (XEN) Pagetable walk from 0fff0000:
> (XEN)  L3[0x000] = 00000000d147f001 0001f01f
> (XEN)  L2[0x07f] = 0000000000000000 ffffffff 
> (XEN) 
> (XEN) ****************************************
> (XEN) Panic on CPU 0:
> (XEN) FATAL PAGE FAULT
> (XEN) [error_code=0000]
> (XEN) Faulting linear address: 0fff0000
> (XEN) ****************************************
> (XEN) 
> (XEN) Reboot in five seconds...
> 
> 
> I'll be happy to report other details when i can get around to rebooting
> the machine; feel free to request any data you need that might be
> helpful in debugging.
> 
> Thanks for maintaining xen in debian!
> 
> Regards,
> 
>         --dkg
> 
> 
>