[Xen-users] netback.c Bug

[IIC Support <support@xxxxxxxxxxxxxxx>]

When we run more than 1 guest domain on a number of different systems  
AMD, INTEL P4 Xeon Etc. we have seen the error below.  This happens  
with and without hyper threading and in both single and dual CPUS.   
There have been a few other posts I have seen on this but there was  
no resolution.  This did not seem to happen on the testing network,  
however once it was placed on the (very busy) live network it started  
to show up.  We are lucky to get 10 mins of uptime with 6 Guests  
running.  The guests are not doing anything, they are just running.   
Also we have the same error on FC4 using RPMS, and DEBIAN using the  
binary tar ball.

Here is the output

kernel BUG at drivers/xen/netback/netback.c:297 (net_rx_action)!
[<c02485aa>] net_rx_action+0x40a/0x510
[<c02fc7e8>] fib_lookup+0x118/0x160
[<c02c52d8>] ip_route_input_slow+0x198/0x990
[<f48e8dc0>] br_forward_finish+0x0/0x70 [bridge]
[<f48eea5c>] br_nf_forward_arp+0x17c/0x1c0 [bridge]
[<c02a3dd4>] kfree_skbmem+0x24/0x30
[<c0149660>] prep_new_page+0x60/0x70
[<c02af488>] neigh_lookup+0x88/0xd0
[<c02a3dd4>] kfree_skbmem+0x24/0x30
[<c02ed5fe>] arp_process+0x8e/0x5a0
[<f4880898>] e100_intr+0x78/0x1a0 [e100]
[<c0143450>] handle_IRQ_event+0x50/0xb0
[<c01435a5>] __do_IRQ+0xf5/0x120
[<c010e5b2>] do_IRQ+0x22/0x30
[<c01049e6>] evtchn_do_upcall+0xa6/0xe0
[<c01280af>] tasklet_action+0x8f/0x160
[<c0127c3b>] __do_softirq+0x10b/0x130
[<c0127cf5>] do_softirq+0x95/0xa0
[<c010e5b2>] do_IRQ+0x22/0x30
[<c01049e6>] evtchn_do_upcall+0xa6/0xe0
[<c0109c64>] hypervisor_callback+0x2c/0x34
[<c010762b>] xen_idle+0x4b/0xa0
[<c01076f0>] cpu_idle+0x70/0xb0
[<c040a9f7>] start_kernel+0x1b7/0x200
[<c040a380>] unknown_bootoption+0x0/0x210
Kernel panic - not syncing: BUG!

Anyone have any ideas?

Thanks
Chris

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users