From mboxrd@z Thu Jan  1 00:00:00 1970
From: Andrew Lunn <andrew@lunn.ch>
Subject: Re: [PATCH 1/2] eth phy: add mdio bus char device interface
Date: Mon, 18 Jun 2018 09:25:38 +0200
Message-ID: <20180618072538.GA31732@lunn.ch>
References: <1529292733-13243-1-git-send-email-liwei1412@163.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: f.fainelli@gmail.com, netdev@vger.kernel.org
To: Wei Li <liwei1412@163.com>
Return-path: <netdev-owner@vger.kernel.org>
Received: from vps0.lunn.ch ([185.16.172.187]:34896 "EHLO vps0.lunn.ch"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1754976AbeFRHZr (ORCPT <rfc822;netdev@vger.kernel.org>);
        Mon, 18 Jun 2018 03:25:47 -0400
Content-Disposition: inline
In-Reply-To: <1529292733-13243-1-git-send-email-liwei1412@163.com>
Sender: netdev-owner@vger.kernel.org
List-ID: <netdev.vger.kernel.org>

1;5202;0cOn Mon, Jun 18, 2018 at 11:32:12AM +0800, Wei Li wrote:
> Add the char device interface of mdio bus, like what i2c-dev or spidev do.
> They make it possible for user-space programs to access the bus directly.
> 


Hi Wei

What is your real use case here? 

What we don't want to do is add an easy to use API to allow vendors to
write user space binary blob propriatary Ethernet switch or PHY
drivers.

Also, it has been agreed that all new API's to the network stack need
to use netlink sockets, not IOCTLs.

The IOCTL interface you provide is very similar to what the stack
already has. Why cannot use you use that?

	Andrew