From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263125AbTJPRdV (ORCPT ); Thu, 16 Oct 2003 13:33:21 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263129AbTJPRdU (ORCPT ); Thu, 16 Oct 2003 13:33:20 -0400 Received: from waste.org ([209.173.204.2]:8863 "EHLO waste.org") by vger.kernel.org with ESMTP id S263125AbTJPRdU (ORCPT ); Thu, 16 Oct 2003 13:33:20 -0400 Date: Thu, 16 Oct 2003 12:31:35 -0500 From: Matt Mackall To: Jeff Garzik Cc: Eli Billauer , linux-kernel@vger.kernel.org, Nick Piggin Subject: Re: [RFC] frandom - fast random generator module Message-ID: <20031016173135.GL5725@waste.org> References: <3F8E552B.3010507@users.sf.net> <3F8E58A9.20005@cyberone.com.au> <3F8E70E0.7070000@users.sf.net> <3F8E8101.70009@pobox.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3F8E8101.70009@pobox.com> User-Agent: Mutt/1.3.28i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Oct 16, 2003 at 07:29:05AM -0400, Jeff Garzik wrote: > > So, given that trend and also given the existing /dev/[u]random, I > disagree completely: /dev/frandom is the perfect example of something > that should _not_ be in the kernel. If you want /dev/urandom faster, > then solve _that_ problem. Don't try to solve a /dev/urandom problem by > creating something totally new. I have some performance fixes for /dev/urandom, but there's a fair amount of other cleanup that has to go in first. -- Matt Mackall : http://www.selenic.com : Linux development and consulting