[Pkg-xen-devel] Bug#675266: xen-hypervisor-4.0-amd64: Hard reset when starting a DomU on HP DL585 G7 // Opteron 6238

[debian at bugtwister.net]

Package: xen-hypervisor-4.0-amd64
Version: 4.0.1-4
Severity: important

*** Please type your report below this line ***
Hello Debian Team,

I have some strange behavior with a DL585 G7 with only two cpu sockets 
used (Opteron 6238 Interlagos). I
use Debian 6.0 with a XEN Kernel. Every time when I start a DomU, the 
server makes a hard reset without any kernel panic or output.

When I start the same disks in another DL585 G7 with all four cpu sockets
used (Opteron 6174 Magny-Cours), everything works normal!
Is the Opteron Interlagos core not supported in xen debian?

We use many Debian 6.0 Xen Server on different HP Hardware - never seen 
before...


I found some difference in the dmesg output on the two servers:

on the non working DL585:

[    0.000000] No NUMA configuration found
[    0.000000] Faking a node at 0000000000000000-0000000180000000
[    0.000000] Bootmem setup node 0 0000000000000000-0000000180000000
[    0.000000]   NODE_DATA [0000000000008000 - 000000000000ffff]
[    0.000000]   bootmap [0000000000010000 -  000000000003ffff] pages 30
[    0.000000] (9 early reservations) ==> bootmem [0000000000 - 0180000000]
[    0.000000]   #0 [0000000000 - 0000001000]   BIOS data page ==> 
[0000000000 - 0000001000]
[    0.000000]   #1 [00040de000 - 0004103000]   XEN PAGETABLES ==> 
[00040de000 - 0004103000]
[    0.000000]   #2 [0000006000 - 0000008000]       TRAMPOLINE ==> 
[0000006000 - 0000008000]
[    0.000000]   #3 [0001000000 - 00016e7b44]    TEXT DATA BSS ==> 
[0001000000 - 00016e7b44]
[    0.000000]   #4 [0001708000 - 00034dce00]          RAMDISK ==> 
[0001708000 - 00034dce00]
[    0.000000]   #5 [00034dd000 - 00040de000]   XEN START INFO ==> 
[00034dd000 - 00040de000]
[    0.000000]   #6 [00016e8000 - 00016e8180]              BRK ==> 
[00016e8000 - 00016e8180]
[    0.000000]   #7 [0000100000 - 00006cc000]          PGTABLE ==> 
[0000100000 - 00006cc000]
[    0.000000]   #8 [0004103000 - 0004505000]          PGTABLE ==> 
[0004103000 - 0004505000]

on the working:
[    0.000000] Scanning NUMA topology in Northbridge 24
[    0.000000] Number of nodes 8
[    0.000000] Node 0 MemBase 0000000000000000 Limit 000000013f600000
[    0.000000] Node 1 bogus settings 440000000-13f600000.
[    0.000000] Node 2 bogus settings 640000000-13f600000.
[    0.000000] Node 3 bogus settings 840000000-13f600000.
[    0.000000] Node 4 bogus settings a40000000-13f600000.
[    0.000000] Node 5 bogus settings c40000000-13f600000.
[    0.000000] Node 6 bogus settings e40000000-13f600000.
[    0.000000] Node 7 bogus settings 1040000000-13f600000.
[    0.000000] NUMA: Using 63 for the hash shift.
[    0.000000] Using node hash shift of 63

Sorry for the wrong order, but I opened already a Bug at the Xen 
Bugtracker...
http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1823

Please Help!
BR, Juergen Schwarzenecker

-- System Information:
Debian Release: 6.0.4
   APT prefers stable
   APT policy: (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.32-5-xen-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

xen-hypervisor-4.0-amd64 depends on no packages.

Versions of packages xen-hypervisor-4.0-amd64 recommends:
ii  xen-utils-4.0                 4.0.1-4    XEN administrative tools

Versions of packages xen-hypervisor-4.0-amd64 suggests:
pn  xen-docs-4.0 <none>     (no description available)

-- no debconf information