netdev
[Top] [All Lists]

Re: [PATCH 2/3] NET: Generic rate estimator

To: Thomas Graf <tgraf@xxxxxxx>
Subject: Re: [PATCH 2/3] NET: Generic rate estimator
From: jamal <hadi@xxxxxxxxxx>
Date: 04 Oct 2004 09:24:16 -0400
Cc: "David S. Miller" <davem@xxxxxxxxxxxxx>, netdev@xxxxxxxxxxx
In-reply-to: <20041004125300.GA15898@xxxxxxxxxxxxxx>
Organization: jamalopolous
References: <20041003213124.GG14344@xxxxxxxxxxxxxx> <20041003213954.GI14344@xxxxxxxxxxxxxx> <20041003161436.50293f9a.davem@xxxxxxxxxxxxx> <20041003233647.GN14344@xxxxxxxxxxxxxx> <1096852582.1046.378.camel@xxxxxxxxxxxxxxxx> <20041004125300.GA15898@xxxxxxxxxxxxxx>
Reply-to: hadi@xxxxxxxxxx
Sender: netdev-bounce@xxxxxxxxxxx
On Mon, 2004-10-04 at 08:53, Thomas Graf wrote:

> I don't think this is possible:
> 
> octave:1> gcd([1000,1024,1200,200,128,100,32,50,122,24])
> ans = 2
> 
> Why not let userspace provide it? ticks/usec and usec/ticks are
> exported via /proc/net/psched.

Easier to define a compile time constant in the kernel.
What you need is to replace the HZ/4 in the creation and execution
of the timer with a constant. And that the value of said constant would
be preferably in the 250ms range.

If you fix this also would be worth fixing the ones in
net/sched/estimator.c


cheers,
jamal


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