Re: [PATCH] drm/vkms: Fix RGB565 pixel conversion

[Melissa Wen <mwen@xxxxxxxxxx>]

On 04/25, Maíra Canal wrote:
> Perform the correct casting of the intermediate coefficients of the
> RGB565 pixel conversion. Currently, the pixel conversion is using s64
> for the intermediate coefficients, which is causing the IGT pixel-format
> tests to fail. So, cast the operands to s32 in order to improve the
> vkms' test coverage.
>
> Tested with igt@kms_plane@pixel-format and igt@kms_plane@pixel-format-source-clamping.
>
> Fixes: 89b03aeaef16 ("drm/vkms: fix 32bit compilation error by replacing macros")
> Signed-off-by: Maíra Canal <mcanal@xxxxxxxxxx>
> ---
>  drivers/gpu/drm/vkms/vkms_formats.c | 20 ++++++++++----------
>  1 file changed, 10 insertions(+), 10 deletions(-)
>
> diff --git a/drivers/gpu/drm/vkms/vkms_formats.c b/drivers/gpu/drm/vkms/vkms_formats.c
> index 8d948c73741e..f6aeaea81902 100644
> --- a/drivers/gpu/drm/vkms/vkms_formats.c
> +++ b/drivers/gpu/drm/vkms/vkms_formats.c
> @@ -88,8 +88,8 @@ static void RGB565_to_argb_u16(u8 *src_pixels, struct pixel_argb_u16 *out_pixel)
>  {
>  	u16 *pixels = (u16 *)src_pixels;
>  
> -	s64 fp_rb_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(31));
> -	s64 fp_g_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(63));
> +	s32 fp_rb_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(31));
> +	s32 fp_g_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(63));
>  
>  	u16 rgb_565 = le16_to_cpu(*pixels);
>  	s64 fp_r = drm_int2fixp((rgb_565 >> 11) & 0x1f);
> @@ -97,9 +97,9 @@ static void RGB565_to_argb_u16(u8 *src_pixels, struct pixel_argb_u16 *out_pixel)
>  	s64 fp_b = drm_int2fixp(rgb_565 & 0x1f);
>  
>  	out_pixel->a = (u16)0xffff;
> -	out_pixel->r = drm_fixp2int(drm_fixp_mul(fp_r, fp_rb_ratio));
> -	out_pixel->g = drm_fixp2int(drm_fixp_mul(fp_g, fp_g_ratio));
> -	out_pixel->b = drm_fixp2int(drm_fixp_mul(fp_b, fp_rb_ratio));
> +	out_pixel->r = drm_fixp2int((s32)drm_fixp_mul(fp_r, fp_rb_ratio));
> +	out_pixel->g = drm_fixp2int((s32)drm_fixp_mul(fp_g, fp_g_ratio));
> +	out_pixel->b = drm_fixp2int((s32)drm_fixp_mul(fp_b, fp_rb_ratio));

Hi Maíra,

First, good catch!

Instead of all casting, wouldn't it be better to implement a rounding
operation or something like drm_fixp2int_round()?

Thanks,

Melissa

>  }
>  
>  void vkms_compose_row(struct line_buffer *stage_buffer, struct vkms_plane_state *plane, int y)
> @@ -208,17 +208,17 @@ static void argb_u16_to_RGB565(struct vkms_frame_info *frame_info,
>  	int x_limit = min_t(size_t, drm_rect_width(&frame_info->dst),
>  			    src_buffer->n_pixels);
>  
> -	s64 fp_rb_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(31));
> -	s64 fp_g_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(63));
> +	s32 fp_rb_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(31));
> +	s32 fp_g_ratio = drm_fixp_div(drm_int2fixp(65535), drm_int2fixp(63));
>  
>  	for (size_t x = 0; x < x_limit; x++, dst_pixels++) {
>  		s64 fp_r = drm_int2fixp(in_pixels[x].r);
>  		s64 fp_g = drm_int2fixp(in_pixels[x].g);
>  		s64 fp_b = drm_int2fixp(in_pixels[x].b);
>  
> -		u16 r = drm_fixp2int(drm_fixp_div(fp_r, fp_rb_ratio));
> -		u16 g = drm_fixp2int(drm_fixp_div(fp_g, fp_g_ratio));
> -		u16 b = drm_fixp2int(drm_fixp_div(fp_b, fp_rb_ratio));
> +		u16 r = drm_fixp2int((s32)drm_fixp_div(fp_r, fp_rb_ratio));
> +		u16 g = drm_fixp2int((s32)drm_fixp_div(fp_g, fp_g_ratio));
> +		u16 b = drm_fixp2int((s32)drm_fixp_div(fp_b, fp_rb_ratio));
>  
>  		*dst_pixels = cpu_to_le16(r << 11 | g << 5 | b);
>  	}
> --
> 2.40.0