[U-Boot] [PATCH v2] spi/kirkwood: add weak functions board_spi_bus_claim/release

[Prafulla Wadaskar]

> -----Original Message-----
> From: Valentin Longchamp [mailto:valentin.longchamp at keymile.com]
> Sent: 30 March 2012 17:45
> To: Prafulla Wadaskar
> Cc: u-boot at lists.denx.de; Gerlando Falauto; Holger Brunck
> Subject: Re: [PATCH v2] spi/kirkwood: add weak functions
> board_spi_bus_claim/release
> 
> Hi Prafulla,
> 
> For the simplicity of the discussion, I have removed everything in the
> discussion that is not relevant for the current open point.
> 
> On 03/30/2012 01:34 PM, Prafulla Wadaskar wrote:
> > In Kirkwood specific claim_bus API, you will backup default
> configuration (which is NF in your case) for these particular pins
> (SPI_SI, SPI_SCK, SPI_CSn, MOSI, MISO).
> 
> But which MPP are these particular pins ?
> 
> Let's have a look at a single signal, SPI_SCK for instance. From the
> 88F6281
> Hardware Spec [1], page 53, SPI_SCK can be MPP[2], MPP[10]. How can
> the generic
> driver know which one actually is wired to the SPI device SCK pin on
> the
> currently running hardware (when none is configured as then, since by
> default
> for us MPP[2] is NF_IO[4] and MPP[10] is UA0_TXD ) ? This is a board
> specific !
> 
> If you tell me how I easily can find this out in the kirkwood driver,

Dear Valentin

Please make a use of CONFIG_SF_DEFAULT_CS for this.
By default this is defined to 0, lets map it to our default use case i.e. using MPP0-3 for default SPI signals

One may pre-define this in his board config as per specific need, and we can use this effectively in Kirkwood_spi driver.
i.e.
 bit0 to be used to configure SPI_CSn
 bit1 to be used to configure SPI_MOSI... and so on

so if my needs are to use
1. MPP7 as SPI_CSn I will define CONFIG_SF_DEFAULT_CS to 0x01
2. MPP6 as SPI_MOSI I will define CONFIG_SF_DEFAULT_CS to 0x02
2. MPP6 as SPI_MOSI and MPP7 as SPI_CSn, I will define CONFIG_SF_DEFAULT_CS to 0x03
.. and so on.

Regards..
Prafulla . . .

> I will be
> happy to implement your proposed solution. Otherwise, I think we
> should stick
> with the board specific function.
> 
> [1]
> http://www.marvell.com/embedded-
> processors/kirkwood/assets/HW_88F6281_OpenSource.pdf
> 
> Regards
> 
> --
> Valentin Longchamp
> Embedded Software Engineer
> Hardware and Chip Integration
> ______________________________________
> KEYMILE AG
> Schwarzenburgstr. 73
> CH-3097 Liebefeld
> Phone +41 31 377 1318
> Fax   +41 31 377 1212
> valentin.longchamp at keymile.com
> www.keymile.com
> ______________________________________
> KEYMILE: A Specialist as a Partner