Re: [PATCH] drm/rockchip: Filter out alpha formats for primary plane

[Sean Paul <seanpaul@chromium.org>]

On Mon, Apr 16, 2018 at 12:01:03PM -0700, Kristian H. Kristensen wrote:
> Rockchip doesn't support per-pixel alpha blending for the lowest
> window in the stack. While the hw supports restacking the windows, we
> don't expose that in KMS, so just filter out alpha formats for the
> primary plane.
> 
> Change-Id: I7828d5bf0f0b5c6dd23f9e52aa01a5a683131d2f

Can you please strip this from upstream commit msgs?

> Signed-off-by: Kristian H. Kristensen <hoegsberg@chromium.org>
> ---
>  drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 28 ++++++++++++++++++++++++++--
>  1 file changed, 26 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> index fae37b1cd691..2efe78600fd0 100644
> --- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> +++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
> @@ -1547,6 +1547,28 @@ static irqreturn_t vop_isr(int irq, void *data)
>  	return ret;
>  }
>  
> +static uint32_t formats_for_win(const struct vop_win_phy *phy, uint32_t win,
> +				uint32_t *formats, uint32_t max_formats)
> +{
> +	uint32_t i, j;
> +
> +	BUG_ON(max_formats < phy->nformats);
> +
> +	j = 0;
> +	for (i = 0; i < phy->nformats; i++) {
> +		/*
> +		 * win0 doesn't support per-pixel alpha, so don't add any RGBA
> +		 * formats for that plane.
> +		 */
> +		if (win == 0 && is_alpha_support(phy->data_formats[i]))
> +			continue;
> +
> +		formats[j++] = phy->data_formats[i];
> +	}
> +
> +	return j;
> +}
> +
>  static int vop_create_crtc(struct vop *vop)
>  {
>  	const struct vop_data *vop_data = vop->data;
> @@ -1568,15 +1590,17 @@ static int vop_create_crtc(struct vop *vop)
>  	for (i = 0; i < vop_data->win_size; i++) {
>  		struct vop_win *vop_win = &vop->win[i];
>  		const struct vop_win_data *win_data = vop_win->data;
> +		uint32_t nformats, formats[16];

I guess 16 is probably a safe bet, but perhaps either pull it out into a define,
or just devm_kzalloc the format to phy->nformats. If you choose the former, the
loop in formats_for_win should ensure it doesn't overrun formats.

Sean

>  
>  		if (win_data->type != DRM_PLANE_TYPE_PRIMARY &&
>  		    win_data->type != DRM_PLANE_TYPE_CURSOR)
>  			continue;
>  
> +		nformats = formats_for_win(win_data->phy, i,
> +					   formats, ARRAY_SIZE(formats));
>  		ret = drm_universal_plane_init(vop->drm_dev, &vop_win->base,
>  					       0, &vop_plane_funcs,
> -					       win_data->phy->data_formats,
> -					       win_data->phy->nformats,
> +					       formats, nformats,
>  					       win_data->phy->format_modifiers,
>  					       win_data->type, NULL);
>  		if (ret) {
> -- 
> 2.13.5
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

-- 
Sean Paul, Software Engineer, Google / Chromium OS
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel