Re: [PATCH v2] ARM: OMAP: Clear GPMC bits when applying new setting.

[Jon Hunter <jon-hunter@ti.com>]

On 03/05/2013 04:13 AM, Mark Jackson wrote:
> When setting the GPMC device type, make sure any previous
> bits are cleared down, before applying the new setting.
> 
> Signed-off-by: Mark Jackson <mpfj@newflow.co.uk>
> ---
> Changes in v2:
> - Change mux type to 2 bits
> - Add extra mux types in gpmc.h
> 
>  arch/arm/mach-omap2/gpmc.c |    4 ++++
>  arch/arm/mach-omap2/gpmc.h |    5 ++++-
>  2 files changed, 8 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/mach-omap2/gpmc.c b/arch/arm/mach-omap2/gpmc.c
> index 410e1ba..479369c 100644
> --- a/arch/arm/mach-omap2/gpmc.c
> +++ b/arch/arm/mach-omap2/gpmc.c
> @@ -613,6 +613,10 @@ int gpmc_cs_configure(int cs, int cmd, int wval)
>  
>  	case GPMC_CONFIG_DEV_TYPE:
>  		regval  = gpmc_cs_read_reg(cs, GPMC_CS_CONFIG1);
> +		/* clear 4 target bits */
> +		regval &= ~(GPMC_CONFIG1_DEVICETYPE(3) |
> +			    GPMC_CONFIG1_MUXTYPE(3));
> +		/* set the proper value */
>  		regval |= GPMC_CONFIG1_DEVICETYPE(wval);
>  		if (wval == GPMC_DEVICETYPE_NOR)
>  			regval |= GPMC_CONFIG1_MUXADDDATA;
> diff --git a/arch/arm/mach-omap2/gpmc.h b/arch/arm/mach-omap2/gpmc.h
> index fe0a844..f79cbde 100644
> --- a/arch/arm/mach-omap2/gpmc.h
> +++ b/arch/arm/mach-omap2/gpmc.h
> @@ -58,7 +58,10 @@
>  #define GPMC_CONFIG1_DEVICESIZE_16      GPMC_CONFIG1_DEVICESIZE(1)
>  #define GPMC_CONFIG1_DEVICETYPE(val)    ((val & 3) << 10)
>  #define GPMC_CONFIG1_DEVICETYPE_NOR     GPMC_CONFIG1_DEVICETYPE(0)
> -#define GPMC_CONFIG1_MUXADDDATA         (1 << 9)
> +#define GPMC_CONFIG1_MUXTYPE(val)       ((val & 3) << 8)
> +#define GPMC_CONFIG1_MUXNONMUX          GPMC_CONFIG1_MUXTYPE(0)
> +#define GPMC_CONFIG1_MUXAAD             GPMC_CONFIG1_MUXTYPE(1)
> +#define GPMC_CONFIG1_MUXADDDATA         GPMC_CONFIG1_MUXTYPE(2)
>  #define GPMC_CONFIG1_TIME_PARA_GRAN     (1 << 4)
>  #define GPMC_CONFIG1_FCLK_DIV(val)      (val & 3)
>  #define GPMC_CONFIG1_FCLK_DIV2          (GPMC_CONFIG1_FCLK_DIV(1))

Thanks for updating. I do plan to get rid of the part of the
gpmc_cs_configure() function [1]. However, I will pull in and based my
series on top of this.

Jon

[1] https://patchwork.kernel.org/patch/2186631/