freebsd stable - Jan 2017 - FreeBSD 11 Stable on a Xen :: bridge0 crashing server

On Fri, 20 Jan 2017 14:26-0000, SK wrote:
> Dear List members
> 
> Apologies for cross posting to two other lists -- was not sure which one is
> more appropriate for this, as it may be related to stable, xen or
> virtualization.
> 
> Here is the situation:
> I have a VPS server from a well reputed provider (and they deserve the
> reputation), running FreeBSD 11 stable x64 under Xen Full Virtualization
> (HVM). I have the xn0 interface which is working fine. I intend to use
VIMAGE,
> so I compiled the kernel, added cloned_interface="bridge0" and
restarted the
> server. But as soon as I am attaching the xn0 to bridge0, the kernel is
> panicking and the server restarting.
> 
> The provider confirmed that they found similar behaviour on a freshly
> installed/provisioned FreeBSD 11 x64 system.
> 
> Further to that, the provider ran some tests, and what they are telling me
is
> if they install/provision FreeBSD 10 x64, there is no problem with
attaching
> bridge0 to xn0. Then they updated/upgraded that box to FreeBSD 11 x64, and
> that too worked when they attached bridge0 to xn0. However, they are yet to
> come back with my request to build a custom kernel with options VIMAGE and
try
> it out.
> 
> In the meanwhile, I took a crash dump and followed the FreeBSD handbook
(not
> that it gave me much idea, I'm not a programmer nor am at that level of
> skillsets). The dump/debug is available at http://pastebin.com/XuWvJHUZ for
> anyone who is interested.
> 
> Just wondering how to go about it. I do need to have bridge0 with xn0 and
all
> the epairs I'm going to use for my jails.
> 
> Any suggestion/pointer/test-instruction is highly appreciated.
The code crashes at line 427 of sys/netinet/if_ether.c:

  ARPSTAT_INC(txrequests);

See 
https://svnweb.freebsd.org/base/stable/11/sys/netinet/if_ether.c?view=annotate#l427

stable/11 has problems accounting the outgoing octets of any xn 
interface, although this isn't connected to your case.

Just to rule out any uncertainty, try this patch:

https://svnweb.freebsd.org/base?view=revision&revision=308126

See PR 213439, 
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=213439

Note, I'm not a developer nor a committer, just a humble sysadmin.

-- 
+-------------------------------+------------------------------------+
| Vennlig hilsen,               | Best regards,                      |
| Trond Endrest?l,              | Trond Endrest?l,                   |
| IT-ansvarlig,                 | System administrator,              |
| Fagskolen Innlandet,          | Gj?vik Technical College, Norway,  |
| tlf. mob.   952 62 567,       | Cellular...: +47 952 62 567,       |
| sentralbord 61 14 54 00.      | Switchboard: +47 61 14 54 00.      |
+-------------------------------+------------------------------------+

On 20.01.2017 18:57, Trond Endrest?l wrote:
>> Here is the situation:
>> I have a VPS server from a well reputed provider (and they deserve the
>> reputation), running FreeBSD 11 stable x64 under Xen Full
Virtualization
>> (HVM). I have the xn0 interface which is working fine. I intend to use
VIMAGE,
>> so I compiled the kernel, added cloned_interface="bridge0"
and restarted the
>> server. But as soon as I am attaching the xn0 to bridge0, the kernel is
>> panicking and the server restarting.
>> Any suggestion/pointer/test-instruction is highly appreciated.
>
> The code crashes at line 427 of sys/netinet/if_ether.c:
>
>   ARPSTAT_INC(txrequests);
>
> See
>
https://svnweb.freebsd.org/base/stable/11/sys/netinet/if_ether.c?view=annotate#l427
>
> stable/11 has problems accounting the outgoing octets of any xn
> interface, although this isn't connected to your case.
>
> Just to rule out any uncertainty, try this patch:
>
> https://svnweb.freebsd.org/base?view=revision&revision=308126
>
> See PR 213439,
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=213439
>
> Note, I'm not a developer nor a committer, just a humble sysadmin.
This problem is unrelated. ARP statistics is global and isn't related to 
some specific interface. IMHO, the kernel panics due to missing VNET 
context. As I see from the code in sys/dev/xen, it is not capable with 
VIMAGE.

-- 
WBR, Andrey V. Elsukov