>>  > An additional question from me is, does the removal of the fast-path also 
>>  > necessitate the removal of MAC_CAPAB_MDT?  I was under the impression 
>>  > that we could still make use of MAC_CAPAB_MDT without a fast-path, yet it 
>>  > was ripped out of clearview-uv along with the fast-path.
>>
>> I'm unclear how MAC_CAPAB_MDT would work without a fastpath, unless we
>> taught the GLDv3 framework about MDT in general (which given software LSO
>> seems the wrong direction).
>>     
>
> I tend to agree, but during our last meeting with Crossbow, I seem to 
> remember Cathy saying that she had MDT working in the UV code before the 
> fast-path even existed.  It's not a big deal, I'm just wondering if 
> removing the fast-path was also pulled MAC_CAPAB_MDT out of the gate, or 
> if it was something else.
>   
Right, MAC_CAPAB_MDT is removed as part of the fast-path removal work.

We discussed MDT support in GLDv3 in the context that Nicolas was trying 
to see what options we have if non-fastpath performance result does not 
turn out  well. But if we can prove the new-fast-path works out okay, I 
don't think we will add MDT to GLDv3 which will soon be pulled out when 
the new fast-path code is in.
>> I do recall a separate discussion that the
>> new fastpath should be able to use MDT, but I don't think it needs
>> MAC_CAPAB_LDT to do so.
>>     
That is correct.

Thanks
- Cathy

Reply via email to