FWIW, I removed the two "megaflow" references in my code and replaced them with "kernel wildcards".
--Justin On Jun 6, 2013, at 5:17 PM, Jesse Gross <[email protected]> wrote: > I mostly meant the kernel code since that's what I've been looking at > but would be good to be consistent throughout the code base. I think > there are just a few instances of the term in the userspace code and > if we think the term in not obvious then it might be better to just > use "wildcard" or some variation. > > On Thu, Jun 6, 2013 at 4:50 PM, Andy Zhou <[email protected]> wrote: >> Thanks, I will address this point in V2 patch. >> >> I agree with you to remove it from the other locations, but want to make >> sure you meant within the scope of kernel code, right? >> >> Andy >> >> >> On Thu, Jun 6, 2013 at 4:41 PM, Jesse Gross <[email protected]> wrote: >>> >>> It looks good - I might also add that if the mask attribute is not >>> specified at all then the flow is exact match. I agree that it is >>> probably clearer to not use the term "mega flow" here but we should >>> then also remove it from the other locations in the code as well. >>> >>> On Thu, Jun 6, 2013 at 3:15 PM, Andy Zhou <[email protected]> wrote: >>>> How about the following change? I choose not to use the term "mega flow" >>>> here to avoid confusing outside people. >>>> >>>> diff --git a/include/linux/openvswitch.h b/include/linux/openvswitch.h >>>> index 6e71b4c..203c7eb 100644 >>>> --- a/include/linux/openvswitch.h >>>> +++ b/include/linux/openvswitch.h >>>> @@ -410,9 +410,11 @@ struct ovs_key_nd { >>>> >>>> * @OVS_FLOW_ATTR_CLEAR: If present in a %OVS_FLOW_CMD_SET request, >>>> clears >>>> the >>>> * last-used time, accumulated TCP flags, and statistics for this flow. >>>> * Otherwise ignored in requests. Never present in notifications. >>>> - * @OVS_FLOW_ATTR_MASK: Optional Nested %OVS_KEY_ATTR_* attributes >>>> specifying >>>> - * the mega flow mask. OVS_KEY_ATTR_* attributes specified only in >>>> OVS_FLOW_ATT >>>> - * but not here will be treated as wildcard field. >>>> + * @OVS_FLOW_ATTR_MASK: Nested %OVS_KEY_ATTR_* attributes specifying >>>> the >>>> + * mask bits for wildcarded flow match. Mask bit value '1' specifies >>>> exact >>>> + * match with corresponding flow key bit, while mask bit value '0' >>>> specifies >>>> + * a wildcarded match. Omitting attribute is treated as wildcarding >>>> + * all corresponding fields. Optional for all requests. >>>> >>>> >>>> On Thu, Jun 6, 2013 at 10:46 AM, Jesse Gross <[email protected]> wrote: >>>>> >>>>> On Wed, Jun 5, 2013 at 10:46 PM, Andy Zhou <[email protected]> wrote: >>>>>> diff --git a/include/linux/openvswitch.h >>>>>> b/include/linux/openvswitch.h >>>>>> index e890fd8..6e71b4c 100644 >>>>>> --- a/include/linux/openvswitch.h >>>>>> +++ b/include/linux/openvswitch.h >>>>>> @@ -410,6 +410,9 @@ struct ovs_key_nd { >>>>>> * @OVS_FLOW_ATTR_CLEAR: If present in a %OVS_FLOW_CMD_SET request, >>>>>> clears the >>>>>> * last-used time, accumulated TCP flags, and statistics for this >>>>>> flow. >>>>>> * Otherwise ignored in requests. Never present in notifications. >>>>>> + * @OVS_FLOW_ATTR_MASK: Optional Nested %OVS_KEY_ATTR_* attributes >>>>>> specifying >>>>>> + * the mega flow mask. OVS_KEY_ATTR_* attributes specified only in >>>>>> OVS_FLOW_ATTR_KEY >>>>>> + * but not here will be treated as wildcard field. >>>>> >>>>> Can you expand this comment a little bit? Outside of people that have >>>>> been working on it, I'm not sure that the term mega flow is really >>>>> self-descriptive. >>>> >>>> >> >> > _______________________________________________ > dev mailing list > [email protected] > http://openvswitch.org/mailman/listinfo/dev _______________________________________________ dev mailing list [email protected] http://openvswitch.org/mailman/listinfo/dev
