[U-Boot] [PATCH] powerpc/mpc85xx: Enabling CPC conditionally based on hwconfig options

[Scott Wood]

On 06/13/2013 01:59:51 AM, Sandeep Singh wrote:
> If hwconfig does not contains "en_cpc" then by default all cpcs are  
> enabled
> If this config is defined then only those individual cpcs which are  
> defined
> in the subargument of "en_cpc" will be enabled e.g en_cpc:cpc1,cpc2;  
> (this
> will enable cpc1 and cpc2) or en_cpc:cpc2; (this enables just cpc2)
> 
> Signed-off-by: Sandeep Singh <Sandeep at freescale.com>
> ---
>  arch/powerpc/cpu/mpc85xx/cpu_init.c |   32  
> ++++++++++++++++++++++++++++++++
>  1 files changed, 32 insertions(+), 0 deletions(-)
> 
> diff --git a/arch/powerpc/cpu/mpc85xx/cpu_init.c  
> b/arch/powerpc/cpu/mpc85xx/cpu_init.c
> index 185e0d5..ea75ce5 100644
> --- a/arch/powerpc/cpu/mpc85xx/cpu_init.c
> +++ b/arch/powerpc/cpu/mpc85xx/cpu_init.c
> @@ -159,11 +159,43 @@ void config_8560_ioports (volatile ccsr_cpm_t *  
> cpm)
>  static void enable_cpc(void)
>  {
>  	int i;
> +	int arglen;
> +	int ret;
>  	u32 size = 0;
> 
>  	cpc_corenet_t *cpc = (cpc_corenet_t *)CONFIG_SYS_FSL_CPC_ADDR;
> +	char buffer[HWCONFIG_BUFFER_SIZE];
> +	char cpc_subarg[16];
> +	bool have_hwconfig = 0;

= false

> +	const char *cpc_args = NULL;
> +
> +	/*
> +	 * Extract hwconfig from environment since environment
> +	 * is not setup properly yet
> +	 */
> +	ret = getenv_f("hwconfig", buffer, sizeof(buffer));
> +	if (ret == -1) {
> +		printf("Error getting hwconfig\n");
> +		return;
> +	}

It is not an error for hwconfig to be missing.  That just means that no  
options are set.

> +	/*
> +	 * If "en_cpc" is not defined in hwconfig then by default all
> +	 * cpcs are enable. If this config is defined then individual
> +	 * cpcs which have to be enabled should also be defined.
> +	 * e.g en_cpc:cpc1,cpc2;
> +	 */
> +	if (hwconfig_f("en_cpc", buffer))
> +		have_hwconfig = 1;

= true

have_hwconfig should be set based on the return of getenv_f.  Is there  
really any benefit to checking whether en_cpc is present, separately  
from the hwconfig_sub_f call?

-Scott