Re: [PATCH 3/3] NET: [SCHED] Qdisc changes and sch_rr added for multiqueue

[Patrick McHardy <kaber@trash.net>]

Waskiewicz Jr, Peter P wrote:
>>> @@ -70,14 +72,28 @@ prio_classify(struct sk_buff *skb, struct Qdisc 
>>> *sch, int *qerr)  #endif
>>>  			if (TC_H_MAJ(band))
>>>  				band = 0;
>>> +			if (q->mq)
>>> +				skb->queue_mapping = 
>>> +						
>>>       
>> q->prio2band[band&TC_PRIO_MAX];
>>     
>>> +			else
>>> +				skb->queue_mapping = 0;
>>>       
>> Might look cleaner if you have one central point where 
>> queue_mapping is set and the band is returned.
>>     
>
> I've taken a stab at this.  I can have one return point, but I'll still
> have multiple assignments of skb->queue_mapping due to the different
> branches for which queue to select in the qdisc.  I suppose we can do a
> rewrite of prio_classify(), but to me that seems beyond the scope of the
> multiqueue patches themselves.  What do you think?
>   

Thats not necessary. I just though you could add one exit point:


...
out:
    skb->queue_mapping = q->mq ? band : 0;
    return q->queues[band];
}

But if that doesn't work don't bother ..