Re: [PATCH 06/13] dev->priv to netdev_priv(dev), for drivers/net/ibm_emac

[Yoann Padioleau <padator@wanadoo.fr>]

Eugene Surovegin <ebs@ebshome.net> writes:

> On Fri, Aug 03, 2007 at 07:34:19PM +0200, Yoann Padioleau wrote:
>> 
>> Replacing accesses to dev->priv to netdev_priv(dev). The replacment
>> is safe when netdev_priv is used to access a private structure that is
>> right next to the net_device structure in memory. Cf
>> http://groups.google.com/group/comp.os.linux.development.system/browse_thread/thread/de19321bcd94dbb8/0d74a4adcd6177bd
>> This is the case when the net_device structure was allocated with
>> a call to alloc_netdev or one of its derivative.
>
> NAK.
>
> While that assumption is correct for the actual emac net device, it's 
> not for MAL poll one.

I forgot to add another condition in my semantic patch. 

I was looking only for xxx->priv = yyy; 
with this rule:

@ danger @
struct net_device *dev;
expression E;
@@
 dev->priv = E

whereas sometimes it's written xxx.priv = yyy; 
as in ibm_emac_mal ( mal->poll_dev.priv = mal; )

By adding the following rule in my semantic patch I correctly
_dont_ modify anything under drivers/net/ibm_emac

@ danger @
struct net_device dev;
expression E;
@@
 dev.priv = E


It's also the case for  drivers/net/e1000/e1000_main.c.


>
> You patch breaks a working driver.
>
> -- 
> Eugene