[PATCH v11 0/2] NAND ECC engine binding changes

[PATCH v11 0/2] NAND ECC engine binding changes

[Miquel Raynal]

As part of a bigger series, I proposed these changes to facilitate the
insertion of a generic ECC engine abstraction, needed specifically for
new SPI-NAND devices not featuring any on-die engine.

The series is almost ready to be merged but the bindings did not pass
the checks. Now it is solved so I am just waiting for Rob's ack on
these two patches to apply the whole series.

Miquel Raynal (2):
  dt-bindings: mtd: Document nand-ecc-engine
  dt-bindings: mtd: Document boolean NAND ECC properties

 .../bindings/mtd/nand-controller.yaml          | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

-- 
2.20.1

[PATCH v11 1/2] dt-bindings: mtd: Document nand-ecc-engine

[Miquel Raynal]

This property is needed to precisely point to the hardware ECC engine
to use when there are several of them available. Here, hardware also
refers to the on-die possibility.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
---
 .../devicetree/bindings/mtd/nand-controller.yaml     | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/Documentation/devicetree/bindings/mtd/nand-controller.yaml b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
index 4a0798247d2d..0969d2e6720b 100644
--- a/Documentation/devicetree/bindings/mtd/nand-controller.yaml
+++ b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
@@ -56,6 +56,18 @@ patternProperties:
           (Linux will handle the calculations). soft_bch is deprecated
           and should be replaced by soft and nand-ecc-algo.
 
+      nand-ecc-engine:
+        maxItems: 1
+        description:
+	  A phandle on the hardware ECC engine if any. There are
+          basically three possibilities:
+          1/ The ECC engine is part of the NAND controller, in this
+          case the phandle should reference the parent node.
+          2/ The ECC engine is part of the NAND part (on-die), in this
+          case the phandle should reference the node itself.
+          3/ The ECC engine is external, in this case the phandle should
+          reference the specific ECC engine node.
+
       nand-ecc-placement:
         allOf:
           - $ref: /schemas/types.yaml#/definitions/string
-- 
2.20.1

Re: [PATCH v11 1/2] dt-bindings: mtd: Document nand-ecc-engine

[Rob Herring]

On Fri, Jun 26, 2020 at 09:13:56AM +0200, Miquel Raynal wrote:
> This property is needed to precisely point to the hardware ECC engine
> to use when there are several of them available. Here, hardware also
> refers to the on-die possibility.
> 
> Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
> ---
>  .../devicetree/bindings/mtd/nand-controller.yaml     | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/mtd/nand-controller.yaml b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> index 4a0798247d2d..0969d2e6720b 100644
> --- a/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> +++ b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> @@ -56,6 +56,18 @@ patternProperties:
>            (Linux will handle the calculations). soft_bch is deprecated
>            and should be replaced by soft and nand-ecc-algo.
>  
> +      nand-ecc-engine:
> +        maxItems: 1

maxItems is for arrays and this is a phandle. It needs a type reference.

> +        description:

You need '|' at the end if you want the formatting preserved.

> +	  A phandle on the hardware ECC engine if any. There are

A tab is not valid YAML. 

> +          basically three possibilities:
> +          1/ The ECC engine is part of the NAND controller, in this
> +          case the phandle should reference the parent node.
> +          2/ The ECC engine is part of the NAND part (on-die), in this
> +          case the phandle should reference the node itself.
> +          3/ The ECC engine is external, in this case the phandle should
> +          reference the specific ECC engine node.
> +
>        nand-ecc-placement:
>          allOf:
>            - $ref: /schemas/types.yaml#/definitions/string
> -- 
> 2.20.1
> 

[PATCH v11 2/2] dt-bindings: mtd: Document boolean NAND ECC properties

[Miquel Raynal]

Document nand-use-soft-ecc-engine and nand-no-ecc-engine properties.
The former is here to force software correction, the latter prevents
any correction to happen.

These properties (along with nand-ecc-engine) are supposed to be more
accurate than the current nand-ecc-modes wich is very misleading and
very often people think it is mandatory while the core should be
relied upon to decide which correction to handle.

nand-ecc-mode was already inacurate, but it becomes totally
problematic with setups where there are several hardware engines.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
---
 Documentation/devicetree/bindings/mtd/nand-controller.yaml | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/Documentation/devicetree/bindings/mtd/nand-controller.yaml b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
index 0969d2e6720b..a3750978ebb8 100644
--- a/Documentation/devicetree/bindings/mtd/nand-controller.yaml
+++ b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
@@ -68,6 +68,12 @@ patternProperties:
           3/ The ECC engine is external, in this case the phandle should
           reference the specific ECC engine node.
 
+      nand-use-soft-ecc-engine: true
+        description: Use a software ECC engine.
+
+      nand-no-ecc-engine: true
+        description: Do not use any ECC correction.
+
       nand-ecc-placement:
         allOf:
           - $ref: /schemas/types.yaml#/definitions/string
-- 
2.20.1

Re: [PATCH v11 2/2] dt-bindings: mtd: Document boolean NAND ECC properties

[Rob Herring]

On Fri, Jun 26, 2020 at 09:13:57AM +0200, Miquel Raynal wrote:
> Document nand-use-soft-ecc-engine and nand-no-ecc-engine properties.
> The former is here to force software correction, the latter prevents
> any correction to happen.
> 
> These properties (along with nand-ecc-engine) are supposed to be more
> accurate than the current nand-ecc-modes wich is very misleading and
> very often people think it is mandatory while the core should be
> relied upon to decide which correction to handle.
> 
> nand-ecc-mode was already inacurate, but it becomes totally
> problematic with setups where there are several hardware engines.
> 
> Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
> ---
>  Documentation/devicetree/bindings/mtd/nand-controller.yaml | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/mtd/nand-controller.yaml b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> index 0969d2e6720b..a3750978ebb8 100644
> --- a/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> +++ b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> @@ -68,6 +68,12 @@ patternProperties:
>            3/ The ECC engine is external, in this case the phandle should
>            reference the specific ECC engine node.
>  
> +      nand-use-soft-ecc-engine: true
> +        description: Use a software ECC engine.
> +
> +      nand-no-ecc-engine: true
> +        description: Do not use any ECC correction.
> +

Still the same problem as v10.

You want:

nand-no-ecc-engine:
  type: boolean
  description: Do not use any ECC correction.

>        nand-ecc-placement:
>          allOf:
>            - $ref: /schemas/types.yaml#/definitions/string
> -- 
> 2.20.1
> 

Re: [PATCH v11 2/2] dt-bindings: mtd: Document boolean NAND ECC properties

[Miquel Raynal]

Hi Rob,

Rob Herring <robh@kernel.org> wrote on Mon, 29 Jun 2020 17:08:00 -0600:

> On Fri, Jun 26, 2020 at 09:13:57AM +0200, Miquel Raynal wrote:
> > Document nand-use-soft-ecc-engine and nand-no-ecc-engine properties.
> > The former is here to force software correction, the latter prevents
> > any correction to happen.
> > 
> > These properties (along with nand-ecc-engine) are supposed to be more
> > accurate than the current nand-ecc-modes wich is very misleading and
> > very often people think it is mandatory while the core should be
> > relied upon to decide which correction to handle.
> > 
> > nand-ecc-mode was already inacurate, but it becomes totally
> > problematic with setups where there are several hardware engines.
> > 
> > Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
> > ---
> >  Documentation/devicetree/bindings/mtd/nand-controller.yaml | 6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/Documentation/devicetree/bindings/mtd/nand-controller.yaml b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> > index 0969d2e6720b..a3750978ebb8 100644
> > --- a/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> > +++ b/Documentation/devicetree/bindings/mtd/nand-controller.yaml
> > @@ -68,6 +68,12 @@ patternProperties:
> >            3/ The ECC engine is external, in this case the phandle should
> >            reference the specific ECC engine node.
> >  
> > +      nand-use-soft-ecc-engine: true
> > +        description: Use a software ECC engine.
> > +
> > +      nand-no-ecc-engine: true
> > +        description: Do not use any ECC correction.
> > +  
> 
> Still the same problem as v10.

I am very sorry for the error, I fixed both commits with a fixup!
commit at the end of my series and forgot to squash them before
sending...

I'll resend the proper patches. Again, sorry for the time lost.

Thanks,
Miquèl