[PATCH v8 01/53] media: videobuf2: Rework offset 'cookie' encoding pattern

[Hans Verkuil]

On 27/09/2023 17:35, Benjamin Gaignard wrote:
> Change how offset 'cookie' field value is computed to make possible
> to use more buffers (up to 0x7fff)
> With this encoding pattern we know the maximum number that a queue
> could store so we can check ing at queue init time.
> It also make easier and faster to find buffer and plane from using
> the offset field.
> Change __find_plane_by_offset() prototype to return the video buffer
> itself rather than it index.
> 
> Signed-off-by: Benjamin Gaignard <benjamin.gaignard at collabora.com>
> ---
>  .../media/common/videobuf2/videobuf2-core.c   | 74 +++++++++----------
>  1 file changed, 35 insertions(+), 39 deletions(-)
> 
> diff --git a/drivers/media/common/videobuf2/videobuf2-core.c b/drivers/media/common/videobuf2/videobuf2-core.c
> index cf6727d9c81f..6eeddb3a01c7 100644
> --- a/drivers/media/common/videobuf2/videobuf2-core.c
> +++ b/drivers/media/common/videobuf2/videobuf2-core.c
> @@ -31,6 +31,14 @@
>  
>  #include <trace/events/vb2.h>
>  
> +#define PLANE_INDEX_SHIFT	(PAGE_SHIFT + 3)
> +#define PLANE_INDEX_MASK	0x7
> +#define BUFFER_INDEX_MASK	0x7fff
> +
> +#if PAGE_SHIFT != 12
> +#error Expected PAGE_SHIFT to be 12

So it turns out this can actually be something other than 12.
Search for CONFIG_ARM64_PAGE_SHIFT and CONFIG_PPC_PAGE_SHIFT.

arm64 supports 12, 14, 16 and powerpc64 supports 12, 14, 16, 18.

I think we should calculate the BUFFER_INDEX_MASK based on the
PAGE_SHIFT value. Even with a PAGE_SHIFT of 18, you can still
allocate up to 512 buffers.

Regards,

	Hans