Re: [PATCH v1] xen/serial: Return actual bytes stored in TX FIFO for OMAP

[Julien Grall <julien.grall@citrix.com>]

On 27/10/15 11:05, Jan Beulich wrote:
>>>> On 27.10.15 at 11:54, <julien.grall@citrix.com> wrote:
>> On 27/10/15 10:50, Jan Beulich wrote:
>>>>>> On 27.10.15 at 11:36, <oleksandr.tyshchenko@globallogic.com> wrote:
>>>> --- a/xen/include/xen/8250-uart.h
>>>> +++ b/xen/include/xen/8250-uart.h
>>>> @@ -143,6 +143,15 @@
>>>>  /* Supplementary control register */
>>>>  #define UART_OMAP_SCR     0x10
>>>
>>> I think this one is already misplaced here (as is the one in the context
>>> below the change).
>>
>> I don't think so, the omap UART is based on the 8250 and share some
>> common registers.
> 
> Well, if it's that way then the patch is fine, but I then question why
> we have a separate omap-uart.c.

I've got no clue, the omap is a superset of the 8250 and actually we
already use the 8250 low-level serial for those platforms.

Linux is also having a separate driver for the OMAP UART. Oleksandr, is
there any fundamental difference between the OMAP and ns16550 driver?

Regards,

-- 
Julien Grall