Re: [PATCH][Updated] New AT91 MCI Driver that supports both MCI slots used at the same time

[Rob Emanuele <poorarm@shoreis.com>]

Haavard,

I'll revise my previous thought about integrating my changes back into
the Atmel driver.  I'll start that today.  I figure this way, the
changes I make can get in front of a wider audience.

Thanks,

Rob

On Mon, Jun 8, 2009 at 9:58 AM, Rob Emanuele<poorarm@shoreis.com> wrote:
> Hi Haavard,
>
> I'd be happy to reintegrate my changes to this back into the Atmel
> driver once I get it working correct on the at91.  Do you have the
> equipment to give my changes a try?
>
> Converting my #defines for the the features to cpu_is_xxx is easy
> enough.  I'll take care of that.
>
> Thanks,
>
> Rob
>
> On Mon, Jun 8, 2009 at 3:47 AM, Haavard
> Skinnemoen<haavard.skinnemoen@atmel.com> wrote:
>> Rob Emanuele wrote:
>>> This driver is a port the Atmel AVR32 MCI driver which uses similar silicon.
>>
>> While I do understand the motivation for this patch, I was kind of
>> hoping we could reduce the number of drivers for the Atmel MCI
>> hardware, not increase it...
>>
>> So I think it would be much better if your changes were integrated into
>> one of the existing drivers.
>>
>>> Joey Oravec has mentioned that he has had better reliability with the chip
>>> he is using (AT91SAM9261) if the MMC controller is reset before each
>>> command.  There is a compile time option to do just that.
>>
>> This could be a useful option for atmel-mci too.
>>
>>> Joey Oravec has stated that the Read and Write Proof functionality that
>>> prevents buffer overflows is not available on the AT91SAM9261.  This was
>>> confirmed by Nicolas Ferre and also affects the AT91RM9200.  This new
>>> patch excludes those register config bits.
>>
>> Preferably use the cpu_is_xxx() functions to achieve that. Perhaps even
>> wrap those tests inside a mci_has_xxproof() function to make it
>> clearer what you're testing for, and easier to change if we decide to
>> do the test based on version registers, etc. later.
>>
>> Haavard
>>
>