Re: [PATCH] serial driver PMC MSP71xx, kernel linux-mips.git mast er

[Marc St-Jean <Marc_St-Jean@pmc-sierra.com>]

Sergei Shtylyov wrote:
> Hello.
> 
> Marc St-Jean wrote:
> 
>  >> >>This I would hope you can hide in the platform specific
>  >> >>serial_in/serial_out functions. If you write the UART_LCR save it in
>  >> >>serial_out(), if you read IER etc.
> 
>  >> > I couldn't find hooks for platform specific serial_in/out functions.
> 
>  >>    It's because there are none. :-)
> 
>  >> > Do you mean using the up->port.iotype's in serial_in/out from 8250.c?
> 
>  >>    Not sure what Alan meant, but this seems the only option for now.
> 
>  > That's the conclusion I came to. I've rewritten the patch to use 
> port.type
>  > instead of iotype since one of the fix is SoC and not UART specific. 
> I guess
> 
>     I failed to folkow your logic. :-)

No longer matters as I can't use port.type. See next comment.

>  > I could use both iotype and type with a test on each for the appropriate
>  > bug, what do you recommend?
> 
>     I think iotype would be enough. You can't pass type for platform 
> devices
> anyway, IIRC (the thing I don't quite like).

I just found that out the hard way, it get's overwritten during autoconfig* and
ends up back at PORT_16550A.

I'm now trying to use my own iotype = UPIO_DWAPB and I've added it to all cases
that check for UPIO_MEM. However at boot time I'm getting:
	"serial8250: ttyS0 at *unknown* (irq = 27) is a 16550A".
It looks like something outside of 8250.c must be checking for UPIO_MEM, I'm
looking into it.

> 
>  >>  >>And we might want to add a void * for board specific insanity to 
> the 8250
>  >> >>structures if we really have to so you can hang your brain damage
>  >> >>privately off that ?
> 
>  >> > Sounds good to me, it would give us a location to store the 
> address of the
>  >> > UART_STATUS_REG required by this UART variant.
> 
>  >>    I doubt we really need to *store* it somewhere. Isn't it an fixed 
> offset
>  >>from UART's base (I haven't seen the header)?
> 
>  > Unfortunately it's not a constant offset from the UART in the SoC 
> register
> 
>     Hm...
> 
>  > space. I've used Alan suggestion and added a classic, on some other 
> OSes %-|,
>  > void "user" pointer.
> 
>     Only do not do it under #ifdef.

Understood, getting rid of them is why I started this thread.

Marc