Re: [PATCH net-next-2.6 v4 1/1] can: c_can: Added support for Bosch C_CAN controller

2 messages, 2 authors, 2011-01-12 · open the first message on its own page

Re: [PATCH net-next-2.6 v4 1/1] can: c_can: Added support for Bosch C_CAN controller

From: Wolfgang Grandegger <hidden>
Date: 2011-01-12 09:08:37

On 01/12/2011 09:51 AM, Bhupesh SHARMA wrote:
Hi Marc,

Thanks for your review.
Please see my comments inline:
quoted
-----Original Message-----
From: Marc Kleine-Budde [mailto:mkl-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org]
On 01/11/2011 12:49 PM, Bhupesh Sharma wrote:
...
quoted
quoted
+static int c_can_close(struct net_device *dev) {
+   struct c_can_priv *priv = netdev_priv(dev);
+
+   netif_stop_queue(dev);
+   napi_disable(&priv->napi);
+   c_can_stop(dev);
+   free_irq(dev->irq, dev);
+   close_candev(dev);
+
+   return 0;
+}
+
+struct net_device *alloc_c_can_dev(void)
Please model after alloc_sja1000_dev:

struct net_device *alloc_sja1000dev(int sizeof_priv)

The private for the _user_ of alloc_c_can_dev is behind the sja1000
private, so you can get rid of the void *priv member in the struct
c_can_priv. (see below)
Ok.
But Wolfgang also suggested to use the *priv* member inside c_can_priv struct
for board specific details. In my c_can platform driver I use it to
store the *clk* variable. Do I need to change that as well?
Marc is referring to:

http://lxr.linux.no/#linux+v2.6.37/drivers/net/can/sja1000/sja1000.c#L582

But also there "priv->priv" is used to store a pointer to the board
specific details. Looking to the SJA1000 drivers, only *one* uses
"sizeof_priv > 0", but many other attach a separately allocated
structure to "priv->priv". For that reason, I'm fine with your current
implementation.

Wolfgang.

Re: [PATCH net-next-2.6 v4 1/1] can: c_can: Added support for Bosch C_CAN controller

From: Marc Kleine-Budde <mkl@pengutronix.de>
Date: 2011-01-12 09:30:07

On 01/12/2011 10:08 AM, Wolfgang Grandegger wrote:
On 01/12/2011 09:51 AM, Bhupesh SHARMA wrote:
quoted
Hi Marc,

Thanks for your review.
Please see my comments inline:
quoted
-----Original Message-----
From: Marc Kleine-Budde [mailto:mkl@pengutronix.de]
On 01/11/2011 12:49 PM, Bhupesh Sharma wrote:
...
quoted
quoted
quoted
+static int c_can_close(struct net_device *dev) {
+   struct c_can_priv *priv = netdev_priv(dev);
+
+   netif_stop_queue(dev);
+   napi_disable(&priv->napi);
+   c_can_stop(dev);
+   free_irq(dev->irq, dev);
+   close_candev(dev);
+
+   return 0;
+}
+
+struct net_device *alloc_c_can_dev(void)
Please model after alloc_sja1000_dev:

struct net_device *alloc_sja1000dev(int sizeof_priv)

The private for the _user_ of alloc_c_can_dev is behind the sja1000
private, so you can get rid of the void *priv member in the struct
c_can_priv. (see below)
Ok.
But Wolfgang also suggested to use the *priv* member inside c_can_priv struct
for board specific details. In my c_can platform driver I use it to
store the *clk* variable. Do I need to change that as well?
Marc is referring to:

http://lxr.linux.no/#linux+v2.6.37/drivers/net/can/sja1000/sja1000.c#L582

But also there "priv->priv" is used to store a pointer to the board
specific details. Looking to the SJA1000 drivers, only *one* uses
"sizeof_priv > 0", but many other attach a separately allocated
structure to "priv->priv". For that reason, I'm fine with your current
implementation.
Okay fine with me.

A nice cleanup might be to introduce something like this:

static inline void * give_me_my_priv_from_sja1000_priv
	(struct sja1000_priv *priv)
{
	return (void *)(priv + 1);
}

Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help