[PATCH] powerpc/85xx: Don't report SRAM to L2 cache fallback as error

[PATCH] powerpc/85xx: Don't report SRAM to L2 cache fallback as error

[Claudiu Manoil]

If the SRAM region parameters are missing the SRAM driver
probing exits and the L2 region is configured as L2 cache
entirely.  This is the expected default behaviour, so it
makes no sense to report it as an error.

Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>
---
 arch/powerpc/sysdev/fsl_85xx_l2ctlr.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c b/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c
index 861cebf..15bc867 100644
--- a/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c
+++ b/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c
@@ -91,9 +91,8 @@ static int mpc85xx_l2ctlr_of_probe(struct platform_device *dev)
 	l2cache_size = *prop;
 
 	if (get_cache_sram_params(&sram_params)) {
-		dev_err(&dev->dev,
-			"Entire L2 as cache, provide valid sram offset and size\n");
-		return -EINVAL;
+		dev_info(&dev->dev, "Configure entire L2 as cache\n");
+		return 0;
 	}
 
 
-- 
1.7.11.7

Re: [PATCH] powerpc/85xx: Don't report SRAM to L2 cache fallback as error

[Scott Wood]

On 05/24/2016 10:07 AM, Claudiu Manoil wrote:=0A=
> If the SRAM region parameters are missing the SRAM driver=0A=
> probing exits and the L2 region is configured as L2 cache=0A=
> entirely.  This is the expected default behaviour, so it=0A=
> makes no sense to report it as an error.=0A=
> =0A=
> Signed-off-by: Claudiu Manoil <claudiu.manoil@nxp.com>=0A=
> ---=0A=
>  arch/powerpc/sysdev/fsl_85xx_l2ctlr.c | 5 ++---=0A=
>  1 file changed, 2 insertions(+), 3 deletions(-)=0A=
> =0A=
> diff --git a/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c b/arch/powerpc/sysdev/=
fsl_85xx_l2ctlr.c=0A=
> index 861cebf..15bc867 100644=0A=
> --- a/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c=0A=
> +++ b/arch/powerpc/sysdev/fsl_85xx_l2ctlr.c=0A=
> @@ -91,9 +91,8 @@ static int mpc85xx_l2ctlr_of_probe(struct platform_devi=
ce *dev)=0A=
>  	l2cache_size =3D *prop;=0A=
>  =0A=
>  	if (get_cache_sram_params(&sram_params)) {=0A=
> -		dev_err(&dev->dev,=0A=
> -			"Entire L2 as cache, provide valid sram offset and size\n");=0A=
> -		return -EINVAL;=0A=
> +		dev_info(&dev->dev, "Configure entire L2 as cache\n");=0A=
> +		return 0;=0A=
>  	}=0A=
=0A=
There's no need to report it at all.  Imagine the mess that the boot log=0A=
would be (even more than it already is) if every driver built into the=0A=
kernel reported "no, didn't find this one".=0A=
=0A=
Plus, the message reads as a command to the user rather than a statement=0A=
of fact.=0A=
=0A=
-Scott=0A=
=0A=