Michal Soltys wrote:

One question regarding size tables:

Should I assume that subsequent qdiscs attached to HFSC leaves will not 
use [different] size tables ?

This is currently the case, but from the perspective of a generic 
solution - if some inner qdisc changes pkt_len stored in cb area during 
qdisc_enqueue, then HFSC might get wrong size during peek operation.

As leaves will usually have "simple" work-conserving qdiscs attached, 
this is probably not an issue, but ...

Using different size tables for inner qdiscs is supposed to work.
What is the "wrong size"? ->peek() should return a packet with the
size the same as ->dequeue() would, which I think it does.