On 5 Jan 2023, at 11:05, David Marchand wrote:

> On Thu, Jan 5, 2023 at 10:46 AM Eelco Chaudron <[email protected]> wrote:
>>
>> In some environments, ovs-vswitchd gets shutdown before the pkill of
>> testpmd has been completed, which results in the following error messages:
>>
>>   Removing port 'dpdkvhostuser0' while vhost device still attached.
>>   To restore connectivity after re-adding of port, VM on socket '' must be 
>> restarted.
>>
>> This patch ignore these error messages.
>
> I remember hitting this...
>
> Another way would be to wait for vhost to notice the virtio port disappeared.
> I had a patch that I never finished working on:
>
> diff --git a/tests/system-dpdk.at b/tests/system-dpdk.at
> index 5ef7f8ccdc..e8e404ad70 100644
> --- a/tests/system-dpdk.at
> +++ b/tests/system-dpdk.at
> @@ -154,6 +154,9 @@ AT_CHECK([ip netns exec ns1 ping -c 4 -I tap0
> 172.31.110.12], [], [stdout],
>  dnl Clean up the testpmd now
>  pkill -f -x -9 'tail -f /dev/null'
>
> +dnl Wait for vhost-user handling the socket disconnection
> +OVS_WAIT_UNTIL([grep "vHost Device '$OVS_RUNDIR/dpdkvhostuser0' has
> been removed" ovs-vswitchd.log])
> +

Yes, I also had something like the below, but I thought adding the logs would 
be the easiest:

KILL_PID=$(pgrep -f -x 'tail -f /dev/null’)
kill -9 $KILL_PID; wait $KILL_PID


>  dnl Clean up
>  AT_CHECK([ovs-vsctl del-port br10 dpdkvhostuser0], [], [stdout], [stderr])
>  OVS_VSWITCHD_STOP("m4_join([], [SYSTEM_DPDK_ALLOWED_LOGS], [
>
>
> -- 
> David Marchand

_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to