VALE not forwarding on physical interfaces

[Nicholas Bastin]

I'm sure this is a local misconfiguration, but I've been unable to figure
this out.  I have a VALE bridge with 5 physical interfaces connected:

$ sudo ./vale-ctl -l
bdg_ctl [98] bridge:0 port:0 vale0:eth12
bdg_ctl [98] bridge:0 port:1 eth12
bdg_ctl [98] bridge:0 port:2 vale0:eth19
bdg_ctl [98] bridge:0 port:3 eth19
bdg_ctl [98] bridge:0 port:4 vale0:eth20
bdg_ctl [98] bridge:0 port:5 eth20
bdg_ctl [98] bridge:0 port:6 vale0:eth21
bdg_ctl [98] bridge:0 port:7 eth21
bdg_ctl [98] bridge:0 port:8 vale0:eth22
bdg_ctl [98] bridge:0 port:9 eth22

(All of these interfaces were attached with vale-ctl -h vale0:ethXX, as per
the README).  eth12 is the uplink to the rest of the network, and eth19-22
are all directly attached to client servers.  eth12 is an Intel I354
embedded controller, and eth19-22 are 82580EB, all using the igb driver.

There is some learning going on:

$ tail -F syslog|grep learning

Mar  3 07:12:33 bsslnx kernel: [419434.805535] 753.688227 [1086]
netmap_bdg_learning       src c8:2a:14:1c:c5:c0 on port 0
Mar  3 07:12:33 bsslnx kernel: [419434.835356] 753.718022 [1086]
netmap_bdg_learning       src c8:2a:14:1c:c5:c0 on port 0
Mar  3 07:12:36 bsslnx kernel: [419437.297646] 756.178393 [1086]
netmap_bdg_learning       src 00:25:90:4c:35:c9 on port 4
Mar  3 07:12:37 bsslnx kernel: [419438.176860] 757.056920 [1086]
netmap_bdg_learning       src 00:0b:78:66:3a:0b on port 0
Mar  3 07:12:39 bsslnx kernel: [419440.140832] 759.019363 [1086]
netmap_bdg_learning       src 28:cf:e9:17:f6:ad on port 0
Mar  3 07:12:39 bsslnx kernel: [419440.172588] 759.051092 [1086]
netmap_bdg_learning       src 28:cf:e9:17:f6:ad on port 0
Mar  3 07:12:39 bsslnx kernel: [419440.357836] 759.236196 [1086]
netmap_bdg_learning       src 00:25:90:4c:35:c9 on port 4


But there is no reachability and while I haven't put in a tap yet (that is
next), there doesn't appear to be any forwarding going on.  When I ping
from c8:2a:14:1c:c5:c0 (which then causes it to arp for the destination), I
get the following:

Mar  3 07:07:20 bsslnx kernel: [419121.098837] 440.226179 [1086]
netmap_bdg_learning       src c8:2a:14:1c:c5:c0 on port 0
Mar  3 07:07:20 bsslnx kernel: [419121.100163] 440.227506 [1216]
nm_bdg_flush              slot 0 port 0 -> 254
Mar  3 07:07:20 bsslnx kernel: [419121.101520] 440.228860 [1527]
netmap_vp_txsync          vale0:eth12 ring 0 flags 0
Mar  3 07:07:20 bsslnx kernel: [419121.101528] 440.228869 [2505]
netmap_common_irq         received TX queue 0
Mar  3 07:07:20 bsslnx kernel: [419121.101530] 440.228874 [1737]
netmap_bwrap_intr_notify  eth19 TX0 0x0
Mar  3 07:07:20 bsslnx kernel: [419121.101532] 440.228876 [2505]
netmap_common_irq         received RX queue 0
Mar  3 07:07:20 bsslnx kernel: [419121.101534] 440.228878 [1737]
netmap_bwrap_intr_notify  eth19 RX0 0x0
Mar  3 07:07:20 bsslnx kernel: [419121.101537] 440.228880 [1777]
netmap_bwrap_intr_notify  eth19 head 0 cur 0 tail 0 (kring 0 0 0)
Mar  3 07:07:20 bsslnx kernel: [419121.101539] 440.228883 [1796]
netmap_bwrap_intr_notify  how strange, interrupt with no packets on eth19

I'm not sure if that has any relevant information...

Thanks,
Nick