netdev
[Top] [All Lists]

Re: [RFC] enhanced version of net_random()

To: Andi Kleen <ak@xxxxxxx>
Subject: Re: [RFC] enhanced version of net_random()
From: "David S. Miller" <davem@xxxxxxxxxx>
Date: Sun, 15 Aug 2004 23:27:54 -0700
Cc: shemminger@xxxxxxxx, alan@xxxxxxxxxxxxxxxxxxx, tytso@xxxxxxx, netdev@xxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, greearb@xxxxxxxxxxxxxxx
In-reply-to: <20040813212857.7dd50320.ak@xxxxxxx>
References: <20040812104835.3b179f5a@xxxxxxxxxxxxxxxxxxxxx> <20040812124854.646f1936.davem@xxxxxxxxxx> <20040813115140.0f09d889@xxxxxxxxxxxxxxxxxxxxx> <20040813212857.7dd50320.ak@xxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
On Fri, 13 Aug 2004 21:28:57 +0200
Andi Kleen <ak@xxxxxxx> wrote:

> On Fri, 13 Aug 2004 11:51:40 -0700
> Stephen Hemminger <shemminger@xxxxxxxx> wrote:
> 
> > Here is another alternative, using tansworthe generator.  It uses percpu
> > state. The one small semantic change is the net_srandom() only affects
> > the current cpu's seed.  The problem was that having it change all cpu's
> > seed would mean adding locking 
> 
> I would just update the other CPUs without locking. Taking
> a random number from a partially updated state shouldn't be a big 
> issue.

I personally don't think we need to touch the other cpus
at all, and that having a different current seed on each
cpu might actually be a good thing.

Stephen, I like this one a lot, especially compared to
what we had before.  I'm going to add this to my tree for
the time being.

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