[U-Boot] [PATCH v2 2/2] warp7 : run sec_init for CAAM RNG

[Auer, Lukas]

On Fri, 2018-01-26 at 11:32 +0000, Bryan O'Donoghue wrote:
> 
> On 26/01/18 09:09, Auer, Lukas wrote:
> > Hi Bryan,
> > 
> > this fails to apply for me on current HEAD. It seems like you have
> > additional modifications to wrap7.c in your tree (there is no
> > CONFIG_USB_GADGET on master).
> 
> I'm carrying a few patches locally and upstreaming gradually - got 
> caught out here...
> 
> > Regarding the patch, would it make sense to put sec_init()
> > somewhere
> > else, so that it does not have to be duplicated in the board file
> > for
> > all platforms with CAAM?
> 
> It does... to me. Looking at these .. I'd say leave the old 
> powerpc/freescale stuff alone.
> 
> This works for me as an alternative when I tested it
> 
> diff --git a/arch/arm/mach-imx/mx7/soc.c b/arch/arm/mach-
> imx/mx7/soc.c
> index d160e80..d399fd8 100644
> --- a/arch/arm/mach-imx/mx7/soc.c
> +++ b/arch/arm/mach-imx/mx7/soc.c
> @@ -261,6 +261,9 @@ int arch_misc_init(void)
>          else
>                  env_set("soc", "imx7s");
>   #endif
> +       #ifdef CONFIG_FSL_CAAM
> +               sec_init();
> +       #endif
> 
>          return 0;
>   }
> 
> perhaps this would work for other i.mx processors
> 
> diff --git a/arch/arm/mach-imx/mx6/soc.c b/arch/arm/mach-
> imx/mx6/soc.c
> index 43cb581..679c23b 100644
> --- a/arch/arm/mach-imx/mx6/soc.c
> +++ b/arch/arm/mach-imx/mx6/soc.c
> @@ -515,6 +515,10 @@ int board_postclk_init(void)
> 
>          set_ldo_voltage(LDO_SOC, 1175); /* Set VDDSOC to 1.175V */
> 
> +#ifdef CONFIG_FSL_CAAM
> +       sec_init();
> +#endif
> +
>          return 0;
>   }
> 
> diff --git a/arch/arm/mach-imx/mx7ulp/soc.c b/arch/arm/mach-
> imx/mx7ulp/soc.c
> index 454665a..dc3d601 100644
> --- a/arch/arm/mach-imx/mx7ulp/soc.c
> +++ b/arch/arm/mach-imx/mx7ulp/soc.c
> @@ -57,6 +57,11 @@ int arch_cpu_init(void)
>   #ifdef CONFIG_BOARD_POSTCLK_INIT
>   int board_postclk_init(void)
>   {
> +
> +#ifdef CONFIG_FSL_CAAM
> +       sec_init();
> +#endif
> +
>          return 0;
>   }
>   #endif
> 
> I'd say the right thing to do is - fix it for all i.MX7D/S and let 
> others with access to mx6/mx7ulp etc test/patch themselves.
> 
> Anyway I'll send a generic patch for i.mx7s/d in arch_misc_init()
> 
> ---
> bod

I agree, that's a good way to add the initialization code.