[Xen-devel] x86_64 domU panics on net_rx_action

[Kevin Krsulich <kdk323@xxxxxxx>]

Hi,

I've been seeing frequent PVM domU crashes under xen-3.3.0 on x86_64
when the domU's are seeing a lot of network traffic. Interestingly, the
crashes do not necessarily occur for the domU that is seeing the
traffic. The domU's and dom0 are 2.6.18-xen-r12 on Gentoo.

This seems similar to the issue mentioned here
http://lists.xensource.com/archives/html/xen-devel/2005-10/msg01425.html
and followed up here
http://lists.xensource.com/archives/html/xen-devel/2005-11/msg00056.html

I've opened up a bug
http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1433 with a
handful of dumps and the output of 'xm log' during a few crash events.
I'll tag on an example here:

Unable to handle kernel paging request at 00000000014f8258 RIP:
[<ffffffff803aff3e>] net_rx_action+0x9f/0x1a6 PGD 4681067 PUD 4685067 PMD
0
Oops: 0000 [1] SMP
CPU 0
Modules linked in:
Pid: 0, comm: swapper Not tainted 2.6.18-xen-r12 #1
RIP: e030:[<ffffffff803aff3e>]  [<ffffffff803aff3e>]
net_rx_action+0x9f/0x1a6
RSP: e02b:ffffffff8055bea0  EFLAGS: 00010206
RAX: 0000000000000000 RBX: 00000000014f8100 RCX: ffff8800014279e0
RDX: ffffffffff5fd000 RSI: 0000000000000000 RDI: ffffffff80477740
RBP: ffff8800014279e0 R08: 0000000000000000 R09: 00000000000008f8
R10: 0000000000000000 R11: 0000000000000001 R12: ffff8800014279c0
R13: 000000010002c87f R14: 0000000000000000 R15: ffffffffff5fd000
FS:  00002ade752923e0(0000) GS:ffffffff80513000(0000)
knlGS:0000000000000000
CS:  e033 DS: 0000 ES: 0000
Process swapper (pid: 0, threadinfo ffffffff80526000, task
ffffffff8046ffe0)
Stack:  ffff8800014279c0 ffffffff803a98f7 0000012b00000000
0000000000000001
ffffffff805131b0 000000000000000a 0000000000000000 ffffffff8022eda4
000000000000010b 0000000000000001 ffffffff8055bf18 000000000000000c
Call Trace:
<IRQ> [<ffffffff803a98f7>] kfree_skbmem+0x9/0x75
[<ffffffff8022eda4>] __do_softirq+0x83/0x117
[<ffffffff8020a680>] call_softirq+0x1c/0x28
[<ffffffff8020bb9d>] do_softirq+0x6a/0xed
[<ffffffff8020bb2a>] do_IRQ+0xec/0xf5
[<ffffffff8033075d>] evtchn_do_upcall+0x13c/0x1fb
[<ffffffff8020a1b6>] do_hypervisor_callback+0x1e/0x2c
<EOI> [<ffffffff802063aa>] hypercall_page+0x3aa/0x1000
[<ffffffff802063aa>] hypercall_page+0x3aa/0x1000
[<ffffffff8020cfa6>] raw_safe_halt+0x96/0xaa
[<ffffffff80208d1e>] xen_idle+0x6d/0x7f
[<ffffffff80208a3f>] cpu_idle+0xab/0xce
[<ffffffff8052f784>] start_kernel+0x24b/0x250
[<ffffffff8052f1e5>] _sinittext+0x1e5/0x1eb


Code: 83 bb 58 01 00 00 00 7e 12 48 8d 74 24 14 48 89 df
ff 93 50
RIP  [<ffffffff803aff3e>] net_rx_action+0x9f/0x1a6  RSP
<ffffffff8055bea0>
CR2: 00000000014f8258
<0>Kernel panic - not syncing: Aiee, killing interrupt
handler!



If it's helpful, the network device being used is a Broadcom NetXtreme
BCM5704 using the tg3 driver compiled into the dom0 kernel.

Thanks and let me know if there's any more information I can provide or
somewhere else I should be directing this issue.

Kevin

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