On 7/1/2019 1:46 PM, Ilya Maximets wrote:
> On 01.07.2019 13:24, Eli Britstein wrote:
>> On 7/1/2019 1:13 PM, Ilya Maximets wrote:
>>> On 30.06.2019 7:47, Eli Britstein wrote:
This patch breaks ovs-dpctl dump-flows, when using TC offloads (kernel).
I added a print in netdev_flow_du
On 01.07.2019 13:24, Eli Britstein wrote:
>
> On 7/1/2019 1:13 PM, Ilya Maximets wrote:
>> On 30.06.2019 7:47, Eli Britstein wrote:
>>> This patch breaks ovs-dpctl dump-flows, when using TC offloads (kernel).
>>>
>>> I added a print in netdev_flow_dump_create, and flow_api is NULL when
>>> invokin
On 7/1/2019 1:13 PM, Ilya Maximets wrote:
> On 30.06.2019 7:47, Eli Britstein wrote:
>> This patch breaks ovs-dpctl dump-flows, when using TC offloads (kernel).
>>
>> I added a print in netdev_flow_dump_create, and flow_api is NULL when
>> invoking ovs-dpctl dump-flows.
>>
>> I think new netdev o
On 30.06.2019 7:47, Eli Britstein wrote:
> This patch breaks ovs-dpctl dump-flows, when using TC offloads (kernel).
>
> I added a print in netdev_flow_dump_create, and flow_api is NULL when
> invoking ovs-dpctl dump-flows.
>
> I think new netdev objects are created to the ports (netdev_open), bu
This patch breaks ovs-dpctl dump-flows, when using TC offloads (kernel).
I added a print in netdev_flow_dump_create, and flow_api is NULL when
invoking ovs-dpctl dump-flows.
I think new netdev objects are created to the ports (netdev_open), but
not properly initialized.
Could you please have a
Current issues with Flow API:
* OVS calls offloading functions regardless of successful
flow API initialization. (ex. on init_flow_api failure)
* Static initilaization of Flow API for a netdev_class forbids
having different offloading types for different instances
of netdev with the same net
