Re: [PATCH] parport: parport_pc: PCI SIO access should also depend on SIO option

On Sunday 01 May 2016 03:07 PM, Maciej S. Szmigiero wrote:
> Hi Greg,
> Hi Sudip,
>
> On 01.05.2016 09:45, Sudip Mukherjee wrote:
>> On Sat, Apr 30, 2016 at 01:56:40PM -0700, Greg Kroah-Hartman wrote:
>>> On Wed, Apr 20, 2016 at 01:09:51PM +0530, Sudip Mukherjee wrote:
>>>> From: "Maciej S. Szmigiero" <[email protected]>
>>>>
>>>> CONFIG_PARPORT_PC_SUPERIO toggles Super IO chip support in parport_pc
>>>> code, however only code accessing SIO chip via ISA (or LPC) bus was
>>>> conditional on it.
>>>>
>>>> This patch makes SIO chip accesses via PCI bus also dependent on this
>>>> config option.
>>>>
>>>> It should be noted that Super IO support in parport_pc is needed only when
>>>> firmware has failed to make parallel port available either via PNP or
>>>> on standard I/O ranges and user has one of a few supported SIOs.
>>>>
>>>> Signed-off-by: Maciej S. Szmigiero <[email protected]>
>>>> ---
>>>>
>>>> Hi Greg,
>>>> This patch is not tested on any superio chip based hardware. (I also
>>>> don't have one).
>>>> http://www.spinics.net/lists/kernel/msg2227051.html
>>>>
>>>> I know you dislike adding #ifdef so its upto you.
>>>
>>> It's really a messy patch, surely there is a better solution.
>>
>> yes, might be. But without hardware, should i dare?

<snip>

> Other possibility that comes to my mind would be to split out all such
> probing code to separate file and then either compile it or not
> depending on CONFIG_PARPORT_PC_SUPERIO.

I thought of this one, but the lack of hardware to test is the only 
reason I am hesitating.

regards
sudip