Re: [PATCH][2.6] Add command function to struct i2c_adapter

[Michael Hunold <hunold-ml@web.de>]

Hi,

On 21.09.2004 17:41, Greg KH wrote:
> On Mon, Sep 20, 2004 at 07:19:24PM +0200, Michael Hunold wrote:
> 
>> 
>>+	/* a ioctl like command that can be used to perform specific functions
>>+	 * with the adapter.
>>+	 */
>>+	int (*command)(struct i2c_adapter *adapter, unsigned int cmd, void *arg);

> Ick ick ick.  We don't like ioctls for the very reason they aren't type
> safe, and you can pretty much stick anything in there you want.  So
> let's not try to add the same type of interface to another subsystem.

Ok.

> How about we add the exact explicit functionality that you want, one
> function per "type" of operation, like all other subsystems have.

Hm, but the functionality depends heavily on the types of clients and 
adapters.

For the dvb subsystem, for example, if we know that the i2c adapter is 
some sort of dvb device, we might need to set the pll from the frontend 
i2c client if the user wants to tune to some frequency. The pll settings 
are very h/w specific, so they should be in the driver implementing the 
i2c adapter. So from the dvb frontend dvb i2c client we would call 
adapter->command(adapter, DVB_FE_SET_PLL, &arg).

You don't mean to add a int(*dvb_fe_set_pll)(...) function to the struct 
i2c_adapter instead, don't you?

Isn't this is a general problem? Isn't there the need to have some 
abstraction to allow message passing between i2c clients and i2c 
adapters in both ways, because i2c clients are never that independend 
from the i2c adapter and have always some h/w dependend parts inside?

> thanks,
> greg k-h

Regards
Michael.