On 2016-11-21 23:29, Ralf Ramsauer wrote:
>>> +/* Bits 8..9 select the MMIO access width */
>>> +#define JAILHOUSE_DBG_FLAG_REG_WIDTH_8     0x0100
>>> +#define JAILHOUSE_DBG_FLAG_REG_WIDTH_32    0x0200
>>
>> Sorry, send you down the wrong road: As said above, we actually specify
>> the register distance here. The access width is specific to the UART and
>> encoded into the driver. So, ...REG_DIST_1 and ...REG_DIST_4 would be an
>> option, encoding things in bytes now as that is more natural for addresses.
> Ok, this clarifies things... I really thought you were talking about
> access width and not register distance.
> 
> Only for my understanding: given a driver with let's say a single 32 bit
> width access in the code, isn't this a naturally given minimum of 4 byte
> register distance?
> In this case, 8 bit distance together with such a driver would not make
> any sense at all, right? But we still want to force-support that?

Well, if a driver is supposed to operate on 8-bit distance devices as
well, it will have to be adjusted to switch its access width as
selected. That implies, of course, that the device has no registers with
more than 8 bits in the first place. The 8250 is a candidate for that,
but I expect us to switch mostly between 8-bit PIO and 32-bit MMIO.

Jan

-- 
Siemens AG, Corporate Technology, CT RDA ITP SES-DE
Corporate Competence Center Embedded Linux

-- 
You received this message because you are subscribed to the Google Groups 
"Jailhouse" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to