[PATCH 4/4] drm/edid: Prep for HDMI VIC aspect ratio (WIP)
Lin, Wayne
Wayne.Lin at amd.com
Mon Oct 21 06:28:18 UTC 2019
> -----Original Message-----
> From: Ville Syrjälä <ville.syrjala at linux.intel.com>
> Sent: Monday, October 14, 2019 10:42 PM
> To: Lin, Wayne <Wayne.Lin at amd.com>
> Cc: dri-devel at lists.freedesktop.org; intel-gfx at lists.freedesktop.org
> Subject: Re: [PATCH 4/4] drm/edid: Prep for HDMI VIC aspect ratio (WIP)
>
> On Mon, Oct 14, 2019 at 09:27:07AM +0000, Lin, Wayne wrote:
> >
> >
> > > -----Original Message-----
> > > From: Ville Syrjala <ville.syrjala at linux.intel.com>
> > > Sent: Friday, October 4, 2019 10:19 PM
> > > To: dri-devel at lists.freedesktop.org
> > > Cc: intel-gfx at lists.freedesktop.org; Lin, Wayne <Wayne.Lin at amd.com>
> > > Subject: [PATCH 4/4] drm/edid: Prep for HDMI VIC aspect ratio (WIP)
> > >
> > > From: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > >
> > > I think this should provide most of necessary logic for adding
> > > aspecr ratios to the HDMI 4k modes.
> > >
> > > Cc: Wayne Lin <waynelin at amd.com>
> > > Signed-off-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > > ---
> > > drivers/gpu/drm/drm_edid.c | 37
> > > +++++++++++++++++++++++++++++++------
> > > 1 file changed, 31 insertions(+), 6 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
> > > index
> > > c7f9f7ca75a2..c76814edc784 100644
> > > --- a/drivers/gpu/drm/drm_edid.c
> > > +++ b/drivers/gpu/drm/drm_edid.c
> > > @@ -3210,6 +3210,11 @@ static enum hdmi_picture_aspect
> > > drm_get_cea_aspect_ratio(const u8 video_code)
> > > return edid_cea_modes[video_code].picture_aspect_ratio;
> > > }
> > >
> > > +static enum hdmi_picture_aspect drm_get_hdmi_aspect_ratio(const u8
> > > +video_code) {
> > > + return edid_4k_modes[video_code].picture_aspect_ratio;
> > > +}
> > > +
> >
> > There are no picture_aspect_ratio attributes defined for modes in
> > edid_4k_modes[] now. Should add on those definitions.
> >
> > > /*
> > > * Calculate the alternate clock for HDMI modes (those from the
> > > HDMI vendor
> > > * specific block).
> > > @@ -3236,6 +3241,9 @@ static u8
> > > drm_match_hdmi_mode_clock_tolerance(const struct drm_display_mode
> *to_
> > > if (!to_match->clock)
> > > return 0;
> > >
> > > + if (to_match->picture_aspect_ratio)
> > > + match_flags |= DRM_MODE_MATCH_ASPECT_RATIO;
> > > +
> > > for (vic = 1; vic < ARRAY_SIZE(edid_4k_modes); vic++) {
> > > const struct drm_display_mode *hdmi_mode =
> &edid_4k_modes[vic];
> > > unsigned int clock1, clock2;
> > > @@ -3271,6 +3279,9 @@ static u8 drm_match_hdmi_mode(const struct
> > > drm_display_mode *to_match)
> > > if (!to_match->clock)
> > > return 0;
> > >
> > > + if (to_match->picture_aspect_ratio)
> > > + match_flags |= DRM_MODE_MATCH_ASPECT_RATIO;
> > > +
> > > for (vic = 1; vic < ARRAY_SIZE(edid_4k_modes); vic++) {
> > > const struct drm_display_mode *hdmi_mode =
> &edid_4k_modes[vic];
> > > unsigned int clock1, clock2;
> >
> > Current code in drm_match_hdmi_mdoe() &
> > drm_match_hdmi_mode_clock_tolerance()
> > use hdmi_mode_alternate_clock() to find alternate clocks.
> > In hdmi_mode_alternate_clock(), it adds an exception for VIC 4 mode
> > (4096x2160 at 24) due to there is no alternate clock defined for that
> > mode in HDMI1.4b. But HDMI2.0 adds 23.98Hz for that mode. Maybe we
> should also revise that part.
>
> I'm tempted to just remove that exception. I have a hard time imagining it
> causing serious problems.
Thanks for your time.
I've run smoke test and CTS to verify these patches (with adding the aspect ratio
attribute to edid_4k_modes[] and removing the exception for23.98Hz). So far it
looks good on my environment. Is there any further modification should be done
on these patches?
>
> >
> > > @@ -5218,6 +5229,7 @@
> > > drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe
> > > *frame,
> > > const struct drm_display_mode *mode) {
> > > enum hdmi_picture_aspect picture_aspect;
> > > + u8 vic, hdmi_vic;
> > > int err;
> > >
> > > if (!frame || !mode)
> > > @@ -5230,7 +5242,8 @@
> > > drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe
> > > *frame,
> > > if (mode->flags & DRM_MODE_FLAG_DBLCLK)
> > > frame->pixel_repeat = 1;
> > >
> > > - frame->video_code = drm_mode_cea_vic(connector, mode);
> > > + vic = drm_mode_cea_vic(connector, mode);
> > > + hdmi_vic = drm_mode_hdmi_vic(connector, mode);
> > >
> > > frame->picture_aspect = HDMI_PICTURE_ASPECT_NONE;
> > >
> > > @@ -5244,11 +5257,15 @@
> > > drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe
> > > *frame,
> > >
> > > /*
> > > * Populate picture aspect ratio from either
> > > - * user input (if specified) or from the CEA mode list.
> > > + * user input (if specified) or from the CEA/HDMI mode lists.
> > > */
> > > picture_aspect = mode->picture_aspect_ratio;
> > > - if (picture_aspect == HDMI_PICTURE_ASPECT_NONE)
> > > - picture_aspect = drm_get_cea_aspect_ratio(frame->video_code);
> > > + if (picture_aspect == HDMI_PICTURE_ASPECT_NONE) {
> > > + if (vic)
> > > + picture_aspect = drm_get_cea_aspect_ratio(vic);
> > > + else if (hdmi_vic)
> > > + picture_aspect = drm_get_hdmi_aspect_ratio(hdmi_vic);
> > > + }
> > >
> > > /*
> > > * The infoframe can't convey anything but none, 4:3 @@ -5256,12
> > > +5273,20 @@ drm_hdmi_avi_infoframe_from_display_mode(struct
> > > hdmi_avi_infoframe *frame,
> > > * we can only satisfy it by specifying the right VIC.
> > > */
> > > if (picture_aspect > HDMI_PICTURE_ASPECT_16_9) {
> > > - if (picture_aspect !=
> > > - drm_get_cea_aspect_ratio(frame->video_code))
> > > + if (vic) {
> > > + if (picture_aspect != drm_get_cea_aspect_ratio(vic))
> > > + return -EINVAL;
> > > + } else if (hdmi_vic) {
> > > + if (picture_aspect !=
> drm_get_hdmi_aspect_ratio(hdmi_vic))
> > > + return -EINVAL;
> > > + } else {
> > > return -EINVAL;
> > > + }
> > > +
> > > picture_aspect = HDMI_PICTURE_ASPECT_NONE;
> > > }
> > >
> > > + frame->video_code = vic;
> > > frame->picture_aspect = picture_aspect;
> > > frame->active_aspect = HDMI_ACTIVE_ASPECT_PICTURE;
> > > frame->scan_mode = HDMI_SCAN_MODE_UNDERSCAN;
> > > --
> > > 2.21.0
> >
> > --
> > Wayne Lin
>
> --
> Ville Syrjälä
> Intel
--
Wayne Lin
More information about the dri-devel
mailing list