From mboxrd@z Thu Jan 1 00:00:00 1970 From: Erwan Velu Subject: Re: [PATCH] e100: expose broadcast_disabled as a module option Date: Mon, 26 Apr 2010 10:49:55 +0200 Message-ID: References: <20100423135816.23f5861f@nehalam> <20100423150207.7969c9f6@nehalam> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Cc: Jeff Kirsher , netdev , David Miller , linux-kernel@vger.kernel.org, jesse.brandeburg@intel.com, bruce.w.allan@intel.com, alexander.h.duyck@intel.com, peter.p.waskiewicz.jr@intel.com, john.ronciak@intel.com To: Stephen Hemminger Return-path: Received: from mail-bw0-f219.google.com ([209.85.218.219]:42814 "EHLO mail-bw0-f219.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750804Ab0DZIt5 (ORCPT ); Mon, 26 Apr 2010 04:49:57 -0400 In-Reply-To: <20100423150207.7969c9f6@nehalam> Sender: netdev-owner@vger.kernel.org List-ID: Agreed that will be a better implementation. Changes of IFF_BROADCAST could play with this "broadcast_disabled" configuration switch to increase the cpu efficiency. I'm not a kernel expert and I don't really figure how the changes of the IFF_BROADCAST should be forwarded to the interfaces and how it can be manipulated. I saw that ifconfig have a '-broadcast' option but looks like none of my drivers are compatible with that. Does any one you could help me understanding what should be the good way to 1- enabled/disabled IFF_BROADCAST for a given interface 2- populate this changes to the driver Once I'll be able to have a proper patch using that, I'll post it again to the list. Cheers, Erwan 2010/4/24 Stephen Hemminger : > On Fri, 23 Apr 2010 23:03:59 +0200 [...] > The point is that the driver can look at IFF_BROADCAST rather than having > module parameter. Module parameters are device driver specific and should > be avoid as much as possible in favor of general mechanism. This is a repeated > problem where users and vendors make special hooks that only work with their > driver, which makes life hard for other users and distribution providers. >