spi: OF module autoloading is still broken

[Javier Martinez Canillas]

Hello Mark,

On 11/16/2015 02:49 PM, Mark Brown wrote:
> On Mon, Nov 16, 2015 at 02:19:27PM -0300, Javier Martinez Canillas wrote:
>> On 11/13/2015 08:48 PM, Brian Norris wrote:
> 
>>> (I believe I avoided this in the first place for mostly-aesthetic
>>> reasons; technically this allows people to put garbage in their DT, like
>>> "garbage,spi-nor". It's unclear whether "garbage" becomes part of the
>>> mythical DT ABI [1].)
> 
>> I don't believe your examples are part of the mythical DT ABI. What I
>> understand is that an ABI is whatever is documented in the DT binding
>> docs but the only document that mentions the m25p80 is:
> 
>> Documentation/devicetree/bindings/mtd/jedec,spi-nor.txt
> 
> Not really, in practice an ABI is something that people notice breaking.
> This means that if enough people ship an undocumented ABI (or it goes
> into important enough products) it's just as good as something that's
> documented, perhaps better than something that's documented and nobody
> ever uses as an ABI.
> 

I see, fair enough. Let's see what Brian say about the spi-nor case and
I'll also post my RFC patch but as a proper patch and adding the comments
you asked me later today.

It would be unfortunate if the SPI drivers would have as a requirement to
always have an SPI device ID table even for OF-only IPs but I don't think
that is that bad either.

Best regards,
-- 
Javier Martinez Canillas
Open Source Group
Samsung Research America