Kernel panic: Unable to handle kernel paging request

Kernel panic: Unable to handle kernel paging request

[Eric Windisch]

Compiling SBCL in a Debian DomU.
Host is Xen 3.0.2, guest is running Xen 3.0.1 kernel (2.6.12.6-xenU).
Guest has 64mb ram and 128mb swap.

Did the guest run out of memory and swap?  If so, why no OOMs?

I have a lot more log than below, but it is fairly large. Please let me 
know if you need more information.

Thank you,

Eric Windisch

--------------------------------------------------------------------------

garyking login: Unable to handle kernel paging request at
ffff88001c95b450 RIP:
<ffffffff801584a9>{kmem_cache_free+72}
PGD 3dd067 PUD 3de067 PMD 0
Oops: 0002 [1] SMP
CPU 0
Modules linked in: md5 ipv6
Pid: 4607, comm: sbcl Tainted: GF 2.6.12.6-xenU
RIP: e030:[<ffffffff801584a9>] <ffffffff801584a9>{kmem_cache_free+72}
RSP: e02b:ffff880002237db8 EFLAGS: 00010446
RAX: 00000000036f6988 RBX: ffff8800011a6800 RCX: 0000000000000000
RDX: 00000000000001e0 RSI: ffff8800011a6810 RDI: ffff8800011a6630
RBP: 0000000000000000 R08: ffff8800011a6810 R09: 00000000ffffffff
R10: 0000000000000008 R11: ffffffffff5fd000 R12: ffff880000abf2a8
R13: ffff880000abf4b8 R14: ffff8800023964c0 R15: ffff880000cd5fd0
FS: 00002aaaab08a4a0(0000) GS:ffffffff8037d480(0000) knlGS:0000000000000000
CS: e033 DS: 0000 ES: 0000
Process sbcl (pid: 4607, threadinfo ffff880002236000, task ffff8800029a89b0)
Stack: ffff880000abf4b8 ffff880000abf2a8 0000000000000000 ffffffff801639ca
0000000102237e98 0000000000000000 ffff8800009ad280 ffff880001321888
ffff880001321868 000000000000081e
Call Trace:<ffffffff801639ca>{vma_adjust+939}
<ffffffff80163bc7>{vma_merge+282}
<ffffffff8016579b>{sys_mprotect+667}
<ffffffff8011296a>{sys_rt_sigreturn+536}
<ffffffff80112b1a>{system_call+134} <ffffffff80112a94>{system_call+0}


Code: 4c 89 64 c3 10 83 03 01 65 8b 04 25 34 00 00 00 48 98 48 c1
RIP <ffffffff801584a9>{kmem_cache_free+72} RSP <ffff880002237db8>
CR2: ffff88001c95b450
<1>Unable to handle kernel paging request at ffff88001c95b448 RIP:
<ffffffff80157cc7>{kmem_cache_alloc+70}
PGD 3dd067 PUD 3de067 PMD 0
Oops: 0000 [2] SMP
CPU 0
Modules linked in: md5 ipv6
Pid: 300, comm: syslogd Tainted: GF 2.6.12.6-xenU
RIP: e030:[<ffffffff80157cc7>] <ffffffff80157cc7>{kmem_cache_alloc+70}
RSP: e02b:ffff880002519df8 EFLAGS: 00010006
RAX: 00000000036f6987 RBX: ffff8800034e6040 RCX: ffff8800034e6098
RDX: ffff8800011a6800 RSI: 00000000000000d0 RDI: ffff8800011a7c80
RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
R10: ffff880002397dc8 R11: 0000000000000048 R12: ffff8800029a9070
R13: ffff880003a3f358 R14: ffff880002397b40 R15: 0000000000000000
FS: 00002aaaaae00640(0000) GS:ffffffff8037d480(0000) knlGS:0000000000000000
CS: e033 DS: 0000 ES: 0000
Process syslogd (pid: 300, threadinfo ffff880002518000, task
ffff880003074130)
Stack: ffff8800029a9070 ffff8800034e6040 00000000fffffff4 ffffffff8012da54
ffff8800034e6090 ffff880002397ba8 0000000000000000 ffff880002397b48
ffff880002397b40 ffff880000d87540
Call Trace:<ffffffff8012da54>{copy_mm+492}
<ffffffff8012ecd5>{copy_process+2392}
<ffffffff8012f3b6>{do_fork+208} <ffffffff8016fd6c>{get_unused_fd+220}
<ffffffff80112b1a>{system_call+134}
<ffffffff80112ee9>{ptregscall_common+61}

Re: Kernel panic: Unable to handle kernel paging request

[Eric Windisch]

Eric Windisch wrote:
> Compiling SBCL in a Debian DomU.
> Host is Xen 3.0.2, guest is running Xen 3.0.1 kernel (2.6.12.6-xenU).
> Guest has 64mb ram and 128mb swap.

FYI - I increased the ram of the DomU to 256mb and still received a 
kernel panic...

--
Eric Windisch

Re: Kernel panic: Unable to handle kernel paging request

[Eric Windisch]

On Fri, 2006-06-09 at 12:16 -0400, Eric Windisch wrote:
> Eric Windisch wrote:
> > Compiling SBCL in a Debian DomU.
> > Host is Xen 3.0.2, guest is running Xen 3.0.1 kernel (2.6.12.6-xenU).
> > Guest has 64mb ram and 128mb swap.
> 
> FYI - I increased the ram of the DomU to 256mb and still received a 
> kernel panic...

I should also note that I tried 2.6.16-xen (xen 3.0.2-2), and that by
setting the DomU to use extras="noapic apci=no" AND after upgrading from
debian stable to debian testing (etch), I was able to get an OOMs with
64-megs.. which indicates to me that the compile probably got further.
I increased the memory to 512mb and again received a kernel panic.

My guess is that either upgrading to etch, or setting noapic allowed the
compile to go FURTHER, although it still didn't complete.

-- 
Eric Windisch <lists@bwbohh.net>