[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-Id: <20090106.104344.130715147.davem@davemloft.net>
Date: Tue, 06 Jan 2009 10:43:44 -0800 (PST)
From: David Miller <davem@...emloft.net>
To: buytenh@...vell.com
Cc: shemminger@...tta.com, netdev@...r.kernel.org
Subject: Re: [PATCH 4/8] dsa: convert to net_device_ops
From: Lennert Buytenhek <buytenh@...vell.com>
Date: Mon, 5 Jan 2009 22:07:05 -0800
> On Mon, Jan 05, 2009 at 12:14:46PM -0800, Stephen Hemminger wrote:
>
> > - slave_dev->open = dsa_slave_open;
> > - slave_dev->stop = dsa_slave_close;
> > - slave_dev->change_rx_flags = dsa_slave_change_rx_flags;
> > - slave_dev->set_rx_mode = dsa_slave_set_rx_mode;
> > - slave_dev->set_multicast_list = dsa_slave_set_rx_mode;
> > - slave_dev->set_mac_address = dsa_slave_set_mac_address;
> > - slave_dev->do_ioctl = dsa_slave_ioctl;
> > + slave_dev->netdev_ops = &dsa_netdev_ops;
> > SET_NETDEV_DEV(slave_dev, parent);
> > slave_dev->vlan_features = master->vlan_features;
>
> Can you make three separate sets of netdevice ops, one for each of the
> tagging formats? On the platforms this driver is used, packet forwarding
> performance matters a _lot_, and I'm pretty sure that adding an extra
> function call indirection in the TX path will translate to a directly
> measurable performance loss.
Agreed, that's probably the way this should be handled.
I'll drop this patch while this issue is being worked out.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists