Re: Xen3 domU crashes

Manuel Bouyer kirjoitti:

> Adding mem=3072M on the xen.gz boot line worked around
it for me ...



ROOT xen3:~> xm list
Name                                      ID   Mem VCPUs    
 State   Time(s)
Domain-0                                   0   256     1    
r-----     11.9
netbsd1                                    1   255     1    
-b----      0.7

ROOT xen3:~> cat /grub/menu.lst
default=0
timeout=10
title           NetBSD/xen with RAID-1
root            (hd0,a)
kernel          (hd0,a)/xen.gz dom0_mem=262144 mem=3072M
module          (hd0,a)/netbsd root=/dev/hda1 ro
console=tty0

title           NetBSD/i386 GENERIC
root            (hd0,a)
kernel          /netbsd.GENERIC


This makes me wonder, what could be causing this? BIOS bug?
Does this affect 
Linux dom0 as well?

Martti

Re: Number of interfaces in Xen 3

So I'm running NetBSD 4.0_BETA2 and I tried to add 16
interfaces to my domU but 
it crashed. It works with 7 interfaces but more than 7
caused the domU to crash 
like this.

Martti


xenbus0 at hypervisor0: Xen Virtual Bus Interface
xencons0 at hypervisor0: Xen Virtual Console Driver
xencons0: console major 143, unit 0
xencons0: using event channel 2
npx0 at hypervisor0: using exception 16
Xen clock: using event channel 4
Kernelized RAIDframe activated
md0: internal 5000 KB image area
xenbus0: using event channel 1
xennet0 at xenbus0 id 0: Xen Virtual Network Interface
xennet0: MAC address 00:16:3e:00:00:12
xennet0: using event channel 5
xennet1 at xenbus0 id 1: Xen Virtual Network Interface
xennet1: MAC address 00:16:3e:00:00:05
xennet1: using event channel 6
xennet2 at xenbus0 id 2: Xen Virtual Network Interface
xennet2: MAC address 00:16:3e:00:00:08
xennet2: using event channel 7
xennet3 at xenbus0 id 3: Xen Virtual Network Interface
xennet3: MAC address 00:16:3e:00:00:06
xennet3: using event channel 8
xennet4 at xenbus0 id 4: Xen Virtual Network Interface
xennet4: MAC address 00:16:3e:00:00:01
xennet4: using event channel 9
xennet5 at xenbus0 id 5: Xen Virtual Network Interface
xennet5: MAC address 00:16:3e:00:00:16
xennet5: using event channel 10
xennet6 at xenbus0 id 6: Xen Virtual Network Interface
xennet6: MAC address 00:16:3e:00:00:13
xennet6: using event channel 11
xennet7 at xenbus0 id 7: Xen Virtual Network Interface
xennet7: MAC address 00:16:3e:00:00:11
xennet7: using event channel 12
xennet8 at xenbus0 id 8: Xen Virtual Network Interface
xennet8: MAC address 00:16:3e:00:00:03
xennet9 at xenbus0 id 9: Xen Virtual Network Interface
xennet9: MAC address 00:16:3e:00:00:14
xennet10 at xenbus0 id 10: Xen Virtual Network Interface
xennet10: MAC address 00:16:3e:00:00:09
xennet11 at xenbus0 id 11: Xen Virtual Network Interface
xennet11: MAC address 00:16:3e:00:00:07
xennet12 at xenbus0 id 12: Xen Virtual Network Interface
xennet12: MAC address 00:16:3e:00:00:02
xennet13 at xenbus0 id 13: Xen Virtual Network Interface
xennet13: MAC address 00:16:3e:00:00:15
xennet14 at xenbus0 id 14: Xen Virtual Network Interface
xennet14: MAC address 00:16:3e:00:00:04
xennet15 at xenbus0 id 15: Xen Virtual Network Interface
xennet15: MAC address 00:16:3e:00:00:10
xbd0 at xenbus0 id 1: Xen Virtual Block Device Interface
uvm_fault(0xc0730140, 0xe0750000, 2) -> 0xe
kernel: supervisor trap page fault, code=0
Stopped in pid 3.1 (xenwatch) at        netbsdengnt
_free_entry+0x2b:  orl     %
edx,netbsd:gnt_entries_bitmask(,%ebx,4)
db>

On Thu, May 10, 2007 at 08:18:07PM +0300, Martti Kuparinen
wrote:
> Manuel Bouyer kirjoitti:
> 
> >Adding mem=3072M on the xen.gz boot line worked
around it for me ...
> 
> 
> 
> ROOT xen3:~> xm list
> Name                                      ID   Mem
VCPUs      State   
> Time(s)
> Domain-0                                   0   256    
1     r-----     11.9
> netbsd1                                    1   255    
1     -b----      0.7
> 
> ROOT xen3:~> cat /grub/menu.lst
> default=0
> timeout=10
> title           NetBSD/xen with RAID-1
> root            (hd0,a)
> kernel          (hd0,a)/xen.gz dom0_mem=262144
mem=3072M
> module          (hd0,a)/netbsd root=/dev/hda1 ro
console=tty0
> 
> title           NetBSD/i386 GENERIC
> root            (hd0,a)
> kernel          /netbsd.GENERIC
> 
> 
> This makes me wonder, what could be causing this? BIOS
bug? Does this 
> affect Linux dom0 as well?

Initially I though it was a bios bug, but after more reports
and though
it's probably a problem somewhere else. Brian Marcotte did
more research
on this (he has the problem with a supermicro board), you
can see his mails
on port-xen.

-- 
Manuel Bouyer <bouyerantioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la
difference
--