[SRU][Trusty][PATCH 0/2] fix "hw csum failure" errors

Jay Vosburgh jay.vosburgh at canonical.com
Fri Jan 16 02:04:06 UTC 2015


BugLink: https://bugs.launchpad.net/bugs/1409123

SRU Justification

Impact: Certain network topologies using VXLAN with certain network
cards will experience "hw csum failure" errors in the kernel log

	"Certain network toplogies" is VXLAN tunneled network between
two containers or VMs on separate hosts.

	"Certain network cards" are at least enic and sky2.  This
probably includes any other device that performs CHECKSUM_COMPLETE
receive checksum offload.

Fix: backport the following two commits from current net-next

commit 17e96834fd35997ca7cdfbf15413bcd5a36ad448
Author: Govindarajulu Varadarajan <_govind at gmx.com>
Date:   Thu Dec 18 15:58:42 2014 +0530

    enic: fix rx skb checksum

commit 2c26d34bbcc0b3f30385d5587aa232289e2eed8e
Author: Jay Vosburgh <jay.vosburgh at canonical.com>
Date:   Fri Dec 19 15:32:00 2014 -0800

    net/core: Handle csum for CHECKSUM_COMPLETE VXLAN forwarding

Testcase:

[ 4297.761899] eth0: hw csum failure
[...]
[ 4297.765223] Call Trace:
[ 4297.765224]  <IRQ>  [<ffffffff8172f026>] dump_stack+0x46/0x58
[ 4297.765235]  [<ffffffff8162ba52>] netdev_rx_csum_fault+0x42/0x50
[ 4297.765238]  [<ffffffff8161c1a0>] ? skb_push+0x40/0x40
[ 4297.765240]  [<ffffffff8162325c>] __skb_checksum_complete+0xbc/0xd0
[ 4297.765243]  [<ffffffff8168c602>] tcp_v4_rcv+0x2e2/0x950
[ 4297.765246]  [<ffffffff81666ca0>] ? ip_rcv_finish+0x360/0x360

	These are reliably reproduced in a network topology of:

container:eth0 == host(OVS VXLAN on VLAN) == bond0 == eth0 (sky2) -> switch

	When VXLAN encapsulated traffic is received from a similarly
configured peer, the above warning is generated in the receive processing
of the encapsulated packet.  Note that the warning is associated with the
container eth0.





More information about the kernel-team mailing list