Date: Thu, 10 May 2018 16:56:38 +0200 From: "Dries Michiels" <driesm.michiels@gmail.com> To: "'Tom Jones'" <thj@freebsd.org>, <freebsd-net@freebsd.org> Subject: RE: kernel: in6_delayed_cksum: delayed m_pullup Message-ID: <007101d3e86f$196e4fb0$4c4aef10$@gmail.com> In-Reply-To: <20180510134746.GD37534@tom-desk.erg.abdn.ac.uk> References: <5ae60f44.1c69fb81.ee4a5.1043@mx.google.com> <20180429190450.GA97911@tom-desk.erg.abdn.ac.uk> <5ae6cbc7.1c69fb81.ed649.655e@mx.google.com> <20180510134746.GD37534@tom-desk.erg.abdn.ac.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
> -----Original Message-----
> From: Tom Jones <thj@freebsd.org>
> Sent: donderdag 10 mei 2018 15:48
> To: Dries Michiels <driesm.michiels@gmail.com>
> Cc: freebsd-net@freebsd.org
> Subject: Re: kernel: in6_delayed_cksum: delayed m_pullup
>=20
> On Mon, Apr 30, 2018 at 09:54:48AM +0200, Dries Michiels wrote:
> >
> > Tom,
> >
> > I=E2=80=99m using the igb (intel I210) driver for my LAN interface =
and the em
> > (intel I219-V) driver for the WAN interface. I use this FreeBSD box
> > as my home server so I wouldn=E2=80=99t say that it has a heavy ipv6 =
workload.
> >
>=20
> Hi Dries,
>=20
> Does this bug persist on later snapshots? Or have you stayed with the =
same
> one from the original question?
I had updated the day after to see if it was just a regression but the =
error message was not gone:
May 1 22:19:51 vados kernel: in6_delayed_cksum: delayed m_pullup, =
m->len: 48 plen 68 off 56 csum_flags=3D400<CSUM_IP6_TCP>
May 1 22:27:57 vados last message repeated 29 times
May 1 22:36:53 vados last message repeated 17 times
May 1 22:49:41 vados last message repeated 6 times
May 1 23:00:09 vados last message repeated 8 times
May 1 23:01:06 vados last message repeated 7 times
May 1 23:18:05 vados last message repeated 12 times
For this reason I had disabled IPV6 for my LAN clients (disabled =
rtadvd).
My server was still ipv6 reachable and error messages where gone.
> Could you send me the ifconfig flags lines for igb0 and em0?
>=20
> like this
>=20
> $ ifconfig em0
> em0:
> flags=3D8943<UP,BROADCAST,RUNNING,PROMISC,SIMPLEX,MULTICAST>
> metric 0 mtu 1500
> options=3D40098<VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,VL
> AN_HWTSO>
>=20
> $ ifconfig igb0
> igb0: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0
> mtu 1500
>=20
> options=3D6403bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMB
> O_MTU,VLAN_HWCSUM,TSO4,TSO6,VLAN_HWTSO,RXCSUM_IPV6,TXCSUM
> _IPV6>
ifconfig flags of my interfaces shown below.
em0: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu =
1500
description: WAN
=
options=3D209b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,WOL_MAGI=
C>
igb0: flags=3D8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu =
1500
description: LAN
=
options=3D6403bb<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU,VLAN_HWC=
SUM,TSO4,TSO6,VLAN_HWTSO,RXCSUM_IPV6,TXCSUM_IPV6>
> My em interfaces don't have flags for ipv6 checksum offload, I suspect =
the
> issue is from the em side of your network.
>=20
> I have tried moving 100GB of v6 tcp through each interface with iperf3 =
to see
> if I could stimulate the bug, but no luck so far. I will leave the =
test box sitting
> exposed to my network for a while.
>=20
> - [tj]
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?007101d3e86f$196e4fb0$4c4aef10$>