On 4 Sep 2024, at 15:54, Simon Horman wrote:

> On Wed, Sep 04, 2024 at 01:40:52PM +0200, Eelco Chaudron wrote:
>>
>>
>> On 4 Sep 2024, at 12:38, Simon Horman wrote:
>>
>>> On Tue, Sep 03, 2024 at 11:37:16AM +0200, Eelco Chaudron wrote:
>>>> The odp_flow_format() function applies a wildcard mask when a
>>>> mask for a given key was not present. However, in the context of
>>>> installing flows in the datapath, the absence of a mask actually
>>>> indicates that the key should be ignored, meaning it should not
>>>> be masked at all.
>>>>
>>>> To address this inconsistency, odp_flow_format() now includes an
>>>> option to skip formatting keys that are missing a mask.
>>>
>>> Hi Eelco,
>>>
>>> Would you be able to provide some example output,
>>> say as part of the commit message.
>>
>> I can add the following example to the commit message:
>>
>> This was found during a debug session of the ‘datapath - ping between two
>> ports on cvlan’ test case. The log message was showing the following:
>>
>>   put[create] ufid:XX recirc_id(0),dp_hash(0/0),skb_priority(0/0),in_port(3),
>>     skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),
>>     eth(src=12:f6:8b:52:f9:75,dst=6e:48:c8:77:d3:8c),eth_type(0x88a8),
>>     vlan(vid=4094,pcp=0/0x0),encap(eth_type(0x8100),
>>     vlan(vid=100/0x0,pcp=0/0x0),encap(eth_type(0x0800),
>>     ipv4(src=10.2.2.2,dst=10.2.2.1,proto=1,tos=0,ttl=64,frag=no),
>>     icmp(type=0,code=0))), actions:2
>>
>> Where it should have shown the below:
>>
>>   put[create] ufid:XX recirc_id(0),dp_hash(0/0),skb_priority(0/0),in_port(3),
>>     skb_mark(0/0),ct_state(0/0),ct_zone(0/0),ct_mark(0/0),ct_label(0/0),
>>     eth(src=12:f6:8b:52:f9:75,dst=6e:48:c8:77:d3:8c),eth_type(0x88a8),
>>     vlan(vid=4094,pcp=0/0x0),encap(eth_type(0x8100)), actions:2
>
> Thanks,
>
> Could you include that when applying the patch, or post a v2,
> whichever you prefer?
>
>>
>>> And, does this warrant a test?
>>
>> Good question, as this was a debug message I thought it would not
>> matter much. But if you feel it’s needed, I can try adding a
>> dp independent unit test.
>
> I think we can pass on the test if it is just a debugging message.
>
>>
>>>>
>>>> Signed-off-by: Eelco Chaudron <[email protected]>
>>>
>>> ...
>
> Acked-by: Simon Horman <[email protected]>

Thanks Simon, applied to main.

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

Reply via email to