[PATCH] ITE887x SuperIO serial port support

[Michael Smith]

On Mon, 26 Feb 2007, Michael Smith wrote:

> I've cleaned it up a bit and I have something that works for me on 
> 2.6.18.3. But, I would like some advice. In particular:
> 
> * I don't have any ITE887x devices with parallel ports. So I could be 
>   breaking someone's parallel device and not know it.

Ted Wen tested the patch on kernel 2.6.18 with an ITE8871 (1 x parallel) 
and it worked with his HP Laserjet.

> * The patch contains an interrupt handler for ITE887x. No other device in 
>   parport_serial.c or 8250_pci.c requires a custom interrupt handler, so 
>   I'm a little perplexed as to why this device would.
> 
> * I'm using a very ugly way to pass information from parport_serial to 
>   8250_pci. The code in parport_serial determines the number of serial and 
>   parallel ports when called from parport_register(), and sticks the 
>   number of serial ports into the static pci_parport_serial_boards[] 
>   table. Yuck.
> 
> * There must be a better way to store the per-board stuff
>   (struct ite887x_private) rather than putting it in a static array.