[ovs-dev] [PATCH] Avoid crash in OvS while transmitting fragmented packets over tunnel.

[Darrell Ball]

On Thu, May 10, 2018 at 11:35 PM, Rohith Basavaraja <
rohith.basavaraja at ericsson.com> wrote:

> Hi Darell,
>
> I reproduce the issue by making VM to transmit fragmented packets and in
> OvS if we have corresponding rule
> to transmit the received fragmented packet (from VM) over tunnel then I
> always hit the crash.
>
>
Thanks Rohith

Unfortunately, I don't hit the assert.
I even tried again with the latest master with this change removed.
I used Geneve for the last test, but I am pretty sure I used Vxlan as well
for this kind of test b4, as I use it often as well.

Would you mind describing your test in more detail (packets and flows)?

Thanks Darrell





> Thanks
> Rohith
>
> On 11/05/18, 2:16 AM, "Darrell Ball" <dball at vmware.com> wrote:
>
>     Thanks Rohith
>     I see this patch was applied, but I have one question inline.
>
>
>     On 4/20/18, 1:48 AM, "ovs-dev-bounces at openvswitch.org on behalf of
> Rohith Basavaraja" <ovs-dev-bounces at openvswitch.org on behalf of
> rohith.basavaraja at ericsson.com> wrote:
>
>         Currently when fragmented packets are to be transmitted in to
> tunnel,
>         base_flow->nw_frag which was initially non-zero at reception is not
>         reset to zero when the base_flow and flow are rewritten
>         as part of the emulated tnl_push action in the ofproto-dpif-xlate
>         module.
>
>         Because of this when fragmented packets are transmitted out of
> tunnel,
>         we hit crash caused by the following assert.
>
>         lib/odp-util.c:5654: assertion flow->nw_proto ==
> base_flow->nw_proto &&
>         flow->nw_frag == base_flow->nw_frag failed in
> commit_set_ipv4_action()
>
>     Can you describe how you hit this assertion?
>     I have some testing in and around this code, but have not hit this
> yet, so I was curious?
>
>         With the following change propagate_tunnel_data_to_flow__
>         is modified to reset *nw_frag* to zero.
>
>
>         Also, that currently we don't
>         fragment tunnelled packets, we should reset *nw_frag* to zero in
>         propagate_tunnel_data_to_flow__.
>
>         Signed-off-by: Jan Scheurich <jan.scheurich at ericsson.com>
>         From: Rohith Basavaraja <rohith.basavaraja at ericsson.com>
>         CC: Jan Scheurich <jan.scheurich at ericsson.com>
>
>         ---
>          ofproto/ofproto-dpif-xlate.c | 1 +
>          1 file changed, 1 insertion(+)
>
>         diff --git a/ofproto/ofproto-dpif-xlate.c
> b/ofproto/ofproto-dpif-xlate.c
>         index 94e3ddb..e9ed037 100644
>         --- a/ofproto/ofproto-dpif-xlate.c
>         +++ b/ofproto/ofproto-dpif-xlate.c
>         @@ -3310,6 +3310,7 @@ propagate_tunnel_data_to_flow__(struct flow
> *dst_flow,
>              dst_flow->ipv6_dst = src_flow->tunnel.ipv6_dst;
>              dst_flow->ipv6_src = src_flow->tunnel.ipv6_src;
>
>         +    dst_flow->nw_frag = 0; /* Tunnel packets are unfragmented. */
>
>
>              dst_flow->nw_tos = src_flow->tunnel.ip_tos;
>              dst_flow->nw_ttl = src_flow->tunnel.ip_ttl;
>              dst_flow->tp_dst = src_flow->tunnel.tp_dst;
>         --
>         1.9.1
>
>
>
>
>         _______________________________________________
>         dev mailing list
>         dev at openvswitch.org
>         https://urldefense.proofpoint.com/v2/url?u=https-3A__mail.
> openvswitch.org_mailman_listinfo_ovs-2Ddev&d=DwICAg&c=
> uilaK90D4TOVoH58JNXRgQ&r=BVhFA09CGX7JQ5Ih-uZnsw&m=
> 6ZMZ7J9yNmX0ZQRMQyUfQ8fZrhemcMFiUqpnVD_jN9w&s=lYu98hfGnEvKr7YcK50fxDDY9-
> d0mA3W0yYtpSeeIQo&e=
>
>
>
>
>
>
> _______________________________________________
> dev mailing list
> dev at openvswitch.org
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>