netdev
[Top] [All Lists]

Re: [PATCH 1/1] More ethtool support for sis900 (with locking)

To: Francois Romieu <romieu@xxxxxxxxxxxxx>
Subject: Re: [PATCH 1/1] More ethtool support for sis900 (with locking)
From: Daniele Venzano <webvenza@xxxxxxxxx>
Date: Sun, 6 Mar 2005 11:01:32 +0100
Cc: netdev@xxxxxxxxxxx, Jeff Garzik <jgarzik@xxxxxxxxx>
In-reply-to: <20050305212838.GD25116@xxxxxxxxxxxxxxxxxxxxxxxxxx>
References: <20050305210745.6763.6618@xxxxxxxxxxxxxxxxxxxxx> <20050305212838.GD25116@xxxxxxxxxxxxxxxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx

On 05/mar/05, at 22:28, Francois Romieu wrote:

Daniele Venzano <webvenza@xxxxxxxxx> :
[...]
Index: sis900.c
===================================================================
--- a/drivers/net/sis900.c      (revision 95)
+++ b/drivers/net/sis900.c      (revision 98)
[...]
@@ -1943,10 +1952,47 @@
        sis_priv->msg_enable = value;
 }

+static u32 sis900_get_link(struct net_device *net_dev)
+{
+       struct sis900_private *sis_priv = net_dev->priv;

Any reason to not use netdev_priv() ?
I'm keeping the style of the rest of the driver, when more pressing issues (wol, netconsole, weird behaviour in suspend/resume) are solved I'll convert the whole source to netdev_priv().

I would not mind an empty line between the declaration of variables
and the actual code.
It seemed to me a waste of space, as the functions are so small and obvious.

Thanks for the feedback.

--
Daniele Venzano
http://www.brownhat.org


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