[Xen-devel] backtrace in drivers/net/xen-netback

[Chris Brannon <cmb@xxxxxxxxx>]

I have the following traceback from xen-netback, under kernel 4.9.58.
Does anyone know what might be going on here?

 ------------[ cut here ]------------
kernel BUG at drivers/net/xen-netback/rx.c:325!
invalid opcode: 0000 [#1] SMP
 task: ffff88028f18cec0 task.stack: ffffc90045e08000
RIP: e030:[<ffffffffc03e1436>]  [<ffffffffc03e1436>] 
xenvif_rx_data_slot+0x1e6/0x1f0 [xen_netback]
RSP: e02b:ffffc90045e0bd78  EFLAGS: 00010246
RAX: 00000000000004b8 RBX: ffffc90045e0bdd8 RCX: ffffc90045df3df8
RDX: 0000000000000000 RSI: 0000000000000021 RDI: ffffc90045de93e8
RBP: ffffc90045e0bdc8 R08: 0000000000000056 R09: ffffc90045de93e8
R10: 0000000000327fc7 R11: 0017ffffc0000000 R12: 0000000000000056
R13: 0000000000000056 R14: 0000000000000056 R15: 0000000000001000
FS:  0000000000000000(0000) GS:ffff880301c40000(0000) knlGS:ffff880301c40000
 CS:  e033 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007ff440e29000 CR3: 00000002aa906000 CR4: 0000000000042660
Stack:
 ffff88028f18cf40 ffffc90042b31298 ffffc90045de9000 ffffc90042b31298
 ffffc90000000000 ffffc90045de9000 ffffc90045e0bdd8 ffff8802ac0e98c0
 ffff880166f4bb00 ffff88028f18cec0 ffffc90045e0be38 ffffffffc03e15ac
Call Trace:
 [<ffffffffc03e15ac>] xenvif_rx_skb+0x16c/0x220 [xen_netback]
 [<ffffffffc03e16a1>] xenvif_rx_action+0x41/0x60 [xen_netback]
 [<ffffffffc03e183a>] xenvif_kthread_guest_rx+0x5a/0x16c [xen_netback]
 [<ffffffffc03e17e0>] ? xenvif_queue_carrier_on+0x70/0x70 [xen_netback]
 [<ffffffff810c63f5>] kthread+0xe5/0x100
 [<ffffffff810c6310>] ? __kthread_init_worker+0x40/0x40
 [<ffffffff818de2d5>] ret_from_fork+0x25/0x30
Code: c8 03 eb c1 48 85 c0 74 96 b8 04 00 00 00 eb 91 8b 82 cc 00 00 00 48 8b 
92 d0 00 00 00 48 8b 44 02 08 48 89 43 10 e9 68 ff ff
ff <0f> 0b eb fe 66 0f 1f 44 00 00 55 48 89 e5 41 54 53 48 83 ec 50 
RIP  [<ffffffffc03e1436>] xenvif_rx_data_slot+0x1e6/0x1f0 [xen_netback]
 RSP <ffffc90045e0bd78>
---[ end trace 592a6fbe8ea728a6 ]---

My colleague Sarah made the following observation: there may be a
locking problem of some kind.  At the point that xenvif_kthread_guest_rx
executes, skb_queue_empty is false, but it's empty when
xenvif_rx_dequeue executes.

Any help would be greatly appreciated.

-- Chris

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel