lkml.org 
[lkml]   [2018]   [May]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH v1 1/3] drm/tegra: dc: Enable plane scaling filters
On Fri, May 04, 2018 at 02:55:01PM +0300, Dmitry Osipenko wrote:
> On 04.05.2018 14:10, Thierry Reding wrote:
> > On Fri, May 04, 2018 at 03:08:42AM +0300, Dmitry Osipenko wrote:
> >> Currently resized plane produces a "pixelated" image which doesn't look
> >> nice, especially in a case of a video overlay. Enable scaling filters that
> >> significantly improve image quality of a scaled overlay.
> >>
> >> Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
> >> ---
> >> drivers/gpu/drm/tegra/dc.c | 51 ++++++++++++++++++++++++++++++++++++++
> >> drivers/gpu/drm/tegra/dc.h | 7 ++++++
> >> 2 files changed, 58 insertions(+)
> >>
> >> diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
> >> index 9f83a65b5ea9..2e81142281c3 100644
> >> --- a/drivers/gpu/drm/tegra/dc.c
> >> +++ b/drivers/gpu/drm/tegra/dc.c
> >> @@ -361,6 +361,47 @@ static void tegra_dc_setup_window(struct tegra_plane *plane,
> >> if (window->bottom_up)
> >> value |= V_DIRECTION;
> >>
> >> + if (!(dc->soc->has_win_a_without_filter && plane->index == 0) &&
> >> + !(window->src.w == window->dst.w)) {
> >
> > I don't know about you, but I get dizzy trying to parse the above. How
> > about we move this into helpers:
>
> +1
>
> > static bool tegra_plane_has_horizontal_filter(struct tegra_plane *plane)
> > {
> > struct tegra_dc *dc = plane->dc;
> >
> > /* this function should never be called on inactive planes */
> > if (WARN_ON(!dc))
> > return false;
> >
> > if (plane->index == 0 && dc->soc->has_win_a_without_filter)
> > return false;
> >
> > return true;
> > }
> >
> > Then the above becomes:
> >
> > bool scale_x = window->dst.w != window->src.w;
> >
> > if (scale_x && tegra_plane_has_horizontal_filter(plane)) {
> >
> >> + /*
> >> + * Enable horizontal 6-tap filter and set filtering
> >> + * coefficients to the default values defined in TRM.
> >> + */
> >> + tegra_plane_writel(plane, 0x00008000, DC_WIN_H_FILTER_P(0));
> >> + tegra_plane_writel(plane, 0x3e087ce1, DC_WIN_H_FILTER_P(1));
> >> + tegra_plane_writel(plane, 0x3b117ac1, DC_WIN_H_FILTER_P(2));
> >> + tegra_plane_writel(plane, 0x591b73aa, DC_WIN_H_FILTER_P(3));
> >> + tegra_plane_writel(plane, 0x57256d9a, DC_WIN_H_FILTER_P(4));
> >> + tegra_plane_writel(plane, 0x552f668b, DC_WIN_H_FILTER_P(5));
> >> + tegra_plane_writel(plane, 0x73385e8b, DC_WIN_H_FILTER_P(6));
> >> + tegra_plane_writel(plane, 0x72435583, DC_WIN_H_FILTER_P(7));
> >> + tegra_plane_writel(plane, 0x714c4c8b, DC_WIN_H_FILTER_P(8));
> >> + tegra_plane_writel(plane, 0x70554393, DC_WIN_H_FILTER_P(9));
> >> + tegra_plane_writel(plane, 0x715e389b, DC_WIN_H_FILTER_P(10));
> >> + tegra_plane_writel(plane, 0x71662faa, DC_WIN_H_FILTER_P(11));
> >> + tegra_plane_writel(plane, 0x536d25ba, DC_WIN_H_FILTER_P(12));
> >> + tegra_plane_writel(plane, 0x55731bca, DC_WIN_H_FILTER_P(13));
> >> + tegra_plane_writel(plane, 0x387a11d9, DC_WIN_H_FILTER_P(14));
> >> + tegra_plane_writel(plane, 0x3c7c08f1, DC_WIN_H_FILTER_P(15));
> >> +
> >> + value |= H_FILTER;
> >> + }
> >> +
> >> + if (!(dc->soc->has_win_a_without_filter && plane->index == 0) &&
> >> + !(dc->soc->has_win_c_without_vert_filter && plane->index == 2) &&
> >> + !(window->src.h == window->dst.h)) {
> >
> > static bool tegra_plane_has_vertical_filter(struct tegra_plane *plane)
> > {
> > struct tegra_dc *dc = plane->dc;
> >
> > /* this function should never be called on inactive planes */
> > if (WARN_ON(!dc))
> > return false;
> >
> > if (plane->index == 0 && dc->soc->has_win_a_without_filter)
> > return false;
> >
> > if (plane->index == 2 && dc->soc->has_win_c_without_vert_filter)
> > return false;
> >
> > return true;
> > }
> >
> > And the above becomes:
> >
> > bool scale_y = window->dst.h != window->src.h;
> >
> > if (scale_y && tegra_plane_has_vertical_filter(plane)) {
>
> I'll adjust this patch, thanks.
>
> Maybe it also worth to factor filtering out into a custom DRM property? For
> example in a case of libvdpau-tegra we perform the exact same filtering on
> blitting YUV to RGB surface (shown as overlay) by GR2D for displaying video
> players interface on top of video, this results in a double filtering that has
> no visual effect and probably wastes memory bandwidth a tad.
>
> Though I'm not very keen anymore on trying to add custom plane properties after
> being previously NAK'ed. We can wrap filtering into a property later if will be
> desired, should be good to just have it enabled by default for now.

Agreed. Let's keep this enabled by default and then address the corner
cases separately.

Thierry
[unhandled content-type:application/pgp-signature]
\
 
 \ /
  Last update: 2018-05-04 14:10    [W:0.075 / U:0.300 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site