Re: [PATCH v10 1/2] dt-bindings: mtd: fixed-partitions: Add alignment properties

[Miquel Raynal]

Hi Simon,

sjg@xxxxxxxxxxxx wrote on Tue, 26 Mar 2024 14:06:44 -0600:

> Add three properties for controlling alignment of partitions, aka
> 'entries' in fixed-partition.
> 
> For now there is no explicit mention of hierarchy, so a 'section' is
> just the 'fixed-partitions' node.
> 
> These new properties are inputs to the Binman packaging process, but are
> also needed if the firmware is repacked, to ensure that alignment
> constraints are not violated. Therefore they are provided as part of
> the schema.
> 
> Signed-off-by: Simon Glass <sjg@xxxxxxxxxxxx>
> Reviewed-by: Rob Herring <robh@xxxxxxxxxx>
> ---
> 
> Changes in v10:
> - Update the minimum to 2
> 
> Changes in v9:
> - Move binding example to next batch to avoid build error
> 
> Changes in v7:
> - Drop patch 'Add binman compatible'
> - Put the alignment properties into the fixed-partition binding
> 
> Changes in v6:
> - Correct schema-validation errors missed due to older dt-schema
>   (enum fix and reg addition)
> 
> Changes in v5:
> - Add value ranges
> - Consistently mention alignment must be power-of-2
> - Mention that alignment refers to bytes
> 
> Changes in v2:
> - Fix 'a' typo in commit message
> 
>  .../bindings/mtd/partitions/partition.yaml    | 51 +++++++++++++++++++
>  1 file changed, 51 insertions(+)
> 
> diff --git a/Documentation/devicetree/bindings/mtd/partitions/partition.yaml b/Documentation/devicetree/bindings/mtd/partitions/partition.yaml
> index 1ebe9e2347ea..656ca3db1762 100644
> --- a/Documentation/devicetree/bindings/mtd/partitions/partition.yaml
> +++ b/Documentation/devicetree/bindings/mtd/partitions/partition.yaml
> @@ -57,6 +57,57 @@ properties:
>        user space from
>      type: boolean
>  
> +  align:
> +    $ref: /schemas/types.yaml#/definitions/uint32
> +    minimum: 2
> +    maximum: 0x80000000
> +    multipleOf: 2
> +    description:
> +      This sets the alignment of the entry in bytes.
> +
> +      The entry offset is adjusted so that the entry starts on an aligned
> +      boundary within the containing section or image. For example ‘align =
> +      <16>’ means that the entry will start on a 16-byte boundary. This may
> +      mean that padding is added before the entry. The padding is part of
> +      the containing section but is not included in the entry, meaning that
> +      an empty space may be created before the entry starts. Alignment
> +      must be a power of 2. If ‘align’ is not provided, no alignment is
> +      performed.
> +
> +  align-size:
> +    $ref: /schemas/types.yaml#/definitions/uint32
> +    minimum: 2
> +    maximum: 0x80000000
> +    multipleOf: 2
> +    description:
> +      This sets the alignment of the entry size in bytes. It must be a power
> +      of 2.
> +
> +      For example, to ensure that the size of an entry is a multiple of 64
> +      bytes, set this to 64. While this does not affect the contents of the
> +      entry within binman itself (the padding is performed only when its
> +      parent section is assembled), the end result is that the entry ends
> +      with the padding bytes, so may grow. If ‘align-size’ is not provided,
> +      no alignment is performed.

I don't think we should mention binman here. Can we have a software
agnostic description? This should be understandable from anyone playing
with mtd partitions I guess.

> +
> +  align-end:
> +    $ref: /schemas/types.yaml#/definitions/uint32
> +    minimum: 2
> +    maximum: 0x80000000
> +    multipleOf: 2

seems not to perfectly match the constraint, but I don't know if there
is a powerOf keyword? (same above)

> +    description:
> +      This sets the alignment (in bytes) of the end of an entry with respect
> +      to the containing section. It must be a power of 2.
> +
> +      Some entries require that they end on an alignment boundary,
> +      regardless of where they start. This does not move the start of the
> +      entry, so the contents of the entry will still start at the beginning.
> +      But there may be padding at the end. While this does not affect the
> +      contents of the entry within binman itself (the padding is performed

content?				same comment about binman?

> +      only when its parent section is assembled), the end result is that the
> +      entry ends with the padding bytes, so may grow. If ‘align-end’ is not
> +      provided, no alignment is performed.
> +
>  if:
>    not:
>      required: [ reg ]


Thanks,
Miquèl