From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [PATCH net-next-2.6] net/ethtool: Add support for the ethtool feature to flash firmware image from a specified file. Date: Wed, 02 Sep 2009 22:43:59 -0700 (PDT) Message-ID: <20090902.224359.256829926.davem@davemloft.net> References: <20090903030241.GA19390@serverengines.com> <4A9F550D.1090105@pobox.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit Cc: ajitk@serverengines.com, netdev@vger.kernel.org To: jgarzik@pobox.com Return-path: Received: from 74-93-104-97-Washington.hfc.comcastbusiness.net ([74.93.104.97]:34348 "EHLO sunset.davemloft.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754225AbZICFnp (ORCPT ); Thu, 3 Sep 2009 01:43:45 -0400 In-Reply-To: <4A9F550D.1090105@pobox.com> Sender: netdev-owner@vger.kernel.org List-ID: From: Jeff Garzik Date: Thu, 03 Sep 2009 01:33:01 -0400 > It passes a _filepath_ directly into the kernel? That seems quite > wrong... > > we fixed kernel drivers to _not_ think they could read directly from > the userland filesystem many years ago. > > Seems like passing in an mmap'd fd would be far preferable. Jeff, it's passing in something that's going to requested via the firmware loading infrastructure in the kernel. It is the most appropriate way to do this.