Re: mixed VLAN problems on msk(4)

From: Ian FREISLICH <ianf_at_clue.co.za>
Date: Wed, 10 Jun 2009 12:54:35 +0200
Pyun YongHyeon wrote:
> Does that mean msk(4) cannot see untagged frame if VLAN hardware
> tag stripping is enabled on msk(4)? I don't see suspicious part in

No, it means that untagged frames that arrive around the time that
tagged frames arrive are dropped.

For instance, ping at 10 packets/s on the untagged address and
there's a CARP broadcast recieved from the network on tagged vlan26
every second, then the interface according to ping will get 10%
packet loss.

> code at this moment. Does hardware MAC statistics report something?
> (sysctl dev.msk.0.stats) Or netstat(1) says Ierrs/Drop on vlan/msk0?

dev.msk.0.stats.rx.ucast_frames: 44760
dev.msk.0.stats.rx.bcast_frames: 36924
dev.msk.0.stats.rx.pause_frames: 0
dev.msk.0.stats.rx.mcast_frames: 0
dev.msk.0.stats.rx.crc_errs: 0
dev.msk.0.stats.rx.good_octets: 28970771
dev.msk.0.stats.rx.bad_octets: 0
dev.msk.0.stats.rx.frames_64: 27085
dev.msk.0.stats.rx.frames_65_127: 23167
dev.msk.0.stats.rx.frames_128_255: 9571
dev.msk.0.stats.rx.frames_256_511: 6277
dev.msk.0.stats.rx.frames_512_1023: 2782
dev.msk.0.stats.rx.frames_1024_1518: 12802
dev.msk.0.stats.rx.frames_1519_max: 0
dev.msk.0.stats.rx.frames_too_long: 0
dev.msk.0.stats.rx.jabbers: 0
dev.msk.0.stats.rx.overflows: 0
dev.msk.0.stats.tx.ucast_frames: 54795
dev.msk.0.stats.tx.bcast_frames: 25
dev.msk.0.stats.tx.pause_frames: 0
dev.msk.0.stats.tx.mcast_frames: 2
dev.msk.0.stats.tx.octets: 5900142
dev.msk.0.stats.tx.frames_64: 5228
dev.msk.0.stats.tx.frames_65_127: 45780
dev.msk.0.stats.tx.frames_128_255: 843
dev.msk.0.stats.tx.frames_256_511: 875
dev.msk.0.stats.tx.frames_512_1023: 1964
dev.msk.0.stats.tx.frames_1024_1518: 132
dev.msk.0.stats.tx.frames_1519_max: 0
dev.msk.0.stats.tx.colls: 0
dev.msk.0.stats.tx.late_colls: 0
dev.msk.0.stats.tx.excess_colls: 0
dev.msk.0.stats.tx.multi_colls: 0
dev.msk.0.stats.tx.single_colls: 0
dev.msk.0.stats.tx.underflows: 0

[apple] /usr/home/ianf # netstat -ni
Name    Mtu Network       Address              Ipkts Ierrs    Opkts Oerrs  Coll
msk0   1500 <Link#1>      00:16:cb:9f:bc:09    81858    51    54914     0     0
msk0   1500 41.154.1.0/24 41.154.1.178         55676     -    54885     -     -

Ierrs is climbing.  If I stop the switch from forwarding tagged
frames, then Ierrs stops increasing.  If I move exclusively to
tagged vlans, then the errors stop.

Turning off vlanhwtag isn't really a solution because shortly after
starting a transfer, I get the following:

Jun  9 18:01:32 apple kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Jun  9 18:02:20 apple kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Jun  9 18:02:45 apple kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Jun  9 18:02:52 apple kernel: msk0: link state changed to DOWN
Jun  9 18:02:54 apple kernel: msk0: link state changed to UP
Jun  9 18:03:31 apple kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering
Jun  9 18:03:37 apple kernel: msk0: watchdog timeout (missed Tx interrupts) -- recovering

> I have small patch which corrects bus_dma related bugs but I'm not
> sure it can fix the VLAN issue.

I'll try this patch later today.

Ian

--
Ian Freislich
Received on Wed Jun 10 2009 - 08:54:46 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:49 UTC