jamal wrote:
Apologies for latency - was busy at (my real) work.
you don't have to apologize for that :) it's something
everybody can easily understand :)
Let me break this email into several ones since it is getting long.
On Tue, 2004-08-17 at 09:40, sandr8 wrote:
something like enqueue(dev) that will indirectly call dev->qdisc->enqueue
and handle in that single place that stuff that does not fit well in
net/core/dev.c
Enqueue of _root_ qdisc is the place to do it.
Maybe even dev.c calls to it. Lets defer this to the next email.
then i should add some code in every qdisc that can
be root and execute it only if it is actually root?
isn't it more complex?
Let me say this:
I am happy with Haralds billing patch which is already in as is.
In other words, although there is an accounting discrepancy it is not
that big.
What does that mean? unbilling is not something to rush in and patch in
if its going to have an impact on other pieces. It doesnt matter whether
it goes in in 2.6.20 or doesnt even go in as far as i am concerned.
However this shouldnt dicourage you because you have actually opened an
issue we need to resolve. So please keep up the discussions.
sure, for the time being i thought to simply concentrate
on the rest and maintain it as indipendent as possible
from the choice that will be taken.
cheers
alessandro
|