netdev
[Top] [All Lists]

Re: [RFC] Replacing qdisc tc_stats with generic statistics to make it ex

To: Thomas Graf <tgraf@xxxxxxx>
Subject: Re: [RFC] Replacing qdisc tc_stats with generic statistics to make it extendable
From: "David S. Miller" <davem@xxxxxxxxxxxxx>
Date: Wed, 6 Oct 2004 10:41:43 -0700
Cc: hadi@xxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to: <20041005222149.GA17186@postel.suug.ch>
References: <20041003213124.GG14344@postel.suug.ch> <20041003213954.GI14344@postel.suug.ch> <20041003161436.50293f9a.davem@davemloft.net> <20041003235737.GO14344@postel.suug.ch> <20041005140304.6d0d5f02.davem@davemloft.net> <20041005222149.GA17186@postel.suug.ch>
Sender: netdev-bounce@xxxxxxxxxxx
On Wed, 6 Oct 2004 00:21:49 +0200
Thomas Graf <tgraf@xxxxxxx> wrote:

> Since this is pretty major and will touch a lot of code I'd be interested
> if this is the right way to go and how to do it to make patch submission
> as easy as possible.
> Here would be my scenario on howto do it in several steps not breaking
> anything in between:
> Step 1: Use new generic networking statistics, update all qdiscs and
>         switch to new estimator. Doesn't break anything but must be
>         atomic.
> Step 2-N: Add dump_stats and update qdiscs one after each other.
> 
> Thoughts?

This sounds fine to me.  Don't be concerned about breaking binary-only
qdiscs :-)


<Prev in Thread] Current Thread [Next in Thread>