CONFIG_PARPORT_PC_SUPERIO

CONFIG_PARPORT_PC_SUPERIO

[Olaf Hering]

Does aynone know if the hardware that is enabled with
CONFIG_PARPORT_PC_SUPERIO is available on PCI devices? The functions
detect_and_report_winbond() and detect_and_report_smsc() poke at random
legacy IO ports. I just wonder if they may also find the hardware if
they come from a PCI device.
If they are i386 only, the config option should probably be i386 (or
whatever) only.

Re: CONFIG_PARPORT_PC_SUPERIO

[David Woodhouse]

On Mon, 2006-10-02 at 11:49 +0200, Olaf Hering wrote:
> Does aynone know if the hardware that is enabled with
> CONFIG_PARPORT_PC_SUPERIO is available on PCI devices? The functions
> detect_and_report_winbond() and detect_and_report_smsc() poke at random
> legacy IO ports. I just wonder if they may also find the hardware if
> they come from a PCI device.
> If they are i386 only, the config option should probably be i386 (or
> whatever) only. 

They're not i386 only. There's plenty of non-i386 machines with SuperIO
chips in. They're the easiest way to bolt floppy/parallel/IrDA/serial
support onto a board.

But we really shouldn't have individual drivers poking at the magic
configuration addresses to find SuperIO chips. There should be support
for those _separately_ -- and then the parport/irda/serial drivers can
all consult it as appropriate, rather than poking directly at the
hardware.

I did make a start on drivers/pnp/superio once, but didn't get very far
before I got distracted. Start with lssuperio.

-- 
dwmw2