Re: [PATCH RFC v3 2/6] drm: drm_fourcc: Introduce macro-pixel info to drm_format_info

[Daniel Vetter]

On Fri, Feb 09, 2018 at 05:35:52PM -0800, Hyun Kwon wrote:
> Multiple pixels can be grouped as a single unit and form a 'macro-pixel'.
> This is to model formats where multiple non-byte aligned pixels are stored
> together in a byte-aligned way. For example, if 3 - 10 bit
> pixels are stored in 32 bit, the 32 bit stroage can be treated as
> a single macro-pixel with 3 pixels. This aligns non-byte addressable
> formats with drm core where each pixel / component is expected to be
> byte aligned.
> 
> Add 'pixels_per_macro' to note how many pixels are in a macro-pixel.
> 'bytes_per_macro' specifies the size of a macro-pixel in bytes.
> 
> Signed-off-by: Hyun Kwon 
> ---
> v3
> - Rename members and rephrase descriptions
> - Rephrase the commit message
> - Use in-line style comments
> v2
> - Introduce macro-pixel over scaling factors
> ---
> ---
>  include/drm/drm_fourcc.h | 24 +++-
>  1 file changed, 23 insertions(+), 1 deletion(-)
> 
> diff --git a/include/drm/drm_fourcc.h b/include/drm/drm_fourcc.h
> index b00bae4..ce59329 100644
> --- a/include/drm/drm_fourcc.h
> +++ b/include/drm/drm_fourcc.h
> @@ -58,11 +58,33 @@ struct drm_format_info {
>   /**
>* @cpp:
>*
> -  * Number of bytes per pixel (per plane)
> +  * Number of bytes per pixel (per plane). @cpp shouldn't be used when
> +  * @pixels_per_macropixel and @bytes_per_macropixel are used.
>*/
>   u8 cpp[3];
>  
>   /**
> +  * @pixels_per_macropixel:
> +  *
> +  * Number of pixels per macro-pixel (per plane). A macro-pixel is
> +  * composed of multiple pixels, and there can be extra bits between
> +  * pixels. This must be used along with @bytes_per_macropixel, only
> +  * when single pixel size is not byte-aligned. In this case, @cpp
> +  * is not valid and should be 0.
> +  */
> + u8 pixels_per_macropixel[3];
> +
> + /*
> +  * @bytes_per_macropixel:
> +  *
> +  * Number of bytes per macro-pixel (per plane). A macro-pixel is
> +  * composed of multiple pixels. The size of single macro-pixel should
> +  * be byte-aligned. This should be used with @pixels_per_macropixel,
> +  * and @cpp should be 0.
> +  */
> + u8 bytes_per_macropixel[3];

Reviewed-by: Daniel Vetter 

> +
> + /**
>* @hsub:
>*
>* Horizontal chroma subsampling factor
> -- 
> 2.7.4
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch
___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel

[PATCH RFC v3 2/6] drm: drm_fourcc: Introduce macro-pixel info to drm_format_info

[Hyun Kwon]

Multiple pixels can be grouped as a single unit and form a 'macro-pixel'.
This is to model formats where multiple non-byte aligned pixels are stored
together in a byte-aligned way. For example, if 3 - 10 bit
pixels are stored in 32 bit, the 32 bit stroage can be treated as
a single macro-pixel with 3 pixels. This aligns non-byte addressable
formats with drm core where each pixel / component is expected to be
byte aligned.

Add 'pixels_per_macro' to note how many pixels are in a macro-pixel.
'bytes_per_macro' specifies the size of a macro-pixel in bytes.

Signed-off-by: Hyun Kwon 
---
v3
- Rename members and rephrase descriptions
- Rephrase the commit message
- Use in-line style comments
v2
- Introduce macro-pixel over scaling factors
---
---
 include/drm/drm_fourcc.h | 24 +++-
 1 file changed, 23 insertions(+), 1 deletion(-)

diff --git a/include/drm/drm_fourcc.h b/include/drm/drm_fourcc.h
index b00bae4..ce59329 100644
--- a/include/drm/drm_fourcc.h
+++ b/include/drm/drm_fourcc.h
@@ -58,11 +58,33 @@ struct drm_format_info {
/**
 * @cpp:
 *
-* Number of bytes per pixel (per plane)
+* Number of bytes per pixel (per plane). @cpp shouldn't be used when
+* @pixels_per_macropixel and @bytes_per_macropixel are used.
 */
u8 cpp[3];
 
/**
+* @pixels_per_macropixel:
+*
+* Number of pixels per macro-pixel (per plane). A macro-pixel is
+* composed of multiple pixels, and there can be extra bits between
+* pixels. This must be used along with @bytes_per_macropixel, only
+* when single pixel size is not byte-aligned. In this case, @cpp
+* is not valid and should be 0.
+*/
+   u8 pixels_per_macropixel[3];
+
+   /*
+* @bytes_per_macropixel:
+*
+* Number of bytes per macro-pixel (per plane). A macro-pixel is
+* composed of multiple pixels. The size of single macro-pixel should
+* be byte-aligned. This should be used with @pixels_per_macropixel,
+* and @cpp should be 0.
+*/
+   u8 bytes_per_macropixel[3];
+
+   /**
 * @hsub:
 *
 * Horizontal chroma subsampling factor
-- 
2.7.4

___
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel