On Tue, Aug 27, 2019 at 12:30:52PM +0300, Jani Nikula wrote:
> On Thu, 22 Aug 2019, Ville Syrjala <ville.syrjala@xxxxxxxxxxxxxxx> wrote:
> > From: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx>
> >
> > Reogranize the HDMI deep color state computation to just
> > loop over possible bpc values. Avoids having to maintain
> > so many variants of the clock etc.
> >
> > The current code also looks confused w.r.t. port_clock vs.
> > bw_constrained. It would happily update port_clock for
> > deep color but then not actually enable deep color due to
> > bw_constrained being set. The new logic handles that case
> > correctly.
>
> Care to elaborate on that please?
Actually I guess it can't happen because:
- bw_constrained is for fdi only atm and no 10bpc hdmi on those
- we can never come here with pipe_bpp==12 + bw_constrained==true anyway
But looks like before commit 3fad10dbb688 ("drm/i915/icl: Fix setting
10 bit deep color mode") this could have been a problem because we could
have legitimately come here with:
pipe_bpp=10*3
bw_constrained=true
deep_color_possible(12) -> true
-> port_clock = clock * 12 / 8;
if (!bw_constained)
pipe_bpp = 12*3;
Ie. we'd have come out with with pipe_bpp=10*3 +
port_clock set up for 12 bpc.
>
> Some nitpicking below.
>
> >
> > Signed-off-by: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx>
> > ---
> > drivers/gpu/drm/i915/display/intel_hdmi.c | 82 ++++++++++-------------
> > 1 file changed, 37 insertions(+), 45 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c
> > index e02f0faecf02..ed1a7afc1ffd 100644
> > --- a/drivers/gpu/drm/i915/display/intel_hdmi.c
> > +++ b/drivers/gpu/drm/i915/display/intel_hdmi.c
> > @@ -2262,8 +2262,7 @@ static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
> > static bool
> > intel_hdmi_ycbcr420_config(struct drm_connector *connector,
> > struct intel_crtc_state *config,
> > - int *clock_12bpc, int *clock_10bpc,
> > - int *clock_8bpc)
> > + int *clock)
> > {
> > struct intel_crtc *intel_crtc = to_intel_crtc(config->base.crtc);
> >
> > @@ -2273,10 +2272,7 @@ intel_hdmi_ycbcr420_config(struct drm_connector *connector,
> > }
> >
> > /* YCBCR420 TMDS rate requirement is half the pixel clock */
> > - config->port_clock /= 2;
> > - *clock_12bpc /= 2;
> > - *clock_10bpc /= 2;
> > - *clock_8bpc /= 2;
> > + *clock /= 2;
> > config->output_format = INTEL_OUTPUT_FORMAT_YCBCR420;
> >
> > /* YCBCR 420 output conversion needs a scaler */
> > @@ -2302,10 +2298,7 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
> > struct drm_scdc *scdc = &connector->display_info.hdmi.scdc;
> > struct intel_digital_connector_state *intel_conn_state =
> > to_intel_digital_connector_state(conn_state);
> > - int clock_8bpc = pipe_config->base.adjusted_mode.crtc_clock;
> > - int clock_10bpc = clock_8bpc * 5 / 4;
> > - int clock_12bpc = clock_8bpc * 3 / 2;
> > - int desired_bpp;
> > + int bpc, clock = pipe_config->base.adjusted_mode.crtc_clock;
> > bool force_dvi = intel_conn_state->force_audio == HDMI_AUDIO_OFF_DVI;
> >
> > if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)
> > @@ -2330,15 +2323,11 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
> >
> > if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK) {
> > pipe_config->pixel_multiplier = 2;
> > - clock_8bpc *= 2;
> > - clock_10bpc *= 2;
> > - clock_12bpc *= 2;
> > + clock *= 2;
> > }
> >
> > if (drm_mode_is_420_only(&connector->display_info, adjusted_mode)) {
> > - if (!intel_hdmi_ycbcr420_config(connector, pipe_config,
> > - &clock_12bpc, &clock_10bpc,
> > - &clock_8bpc)) {
> > + if (!intel_hdmi_ycbcr420_config(connector, pipe_config, &clock)) {
> > DRM_ERROR("Can't support YCBCR420 output\n");
> > return -EINVAL;
> > }
> > @@ -2355,41 +2344,44 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
> > intel_conn_state->force_audio == HDMI_AUDIO_ON;
> > }
> >
> > - /*
> > - * Note that g4x/vlv don't support 12bpc hdmi outputs. We also need
> > - * to check that the higher clock still fits within limits.
> > - */
> > - if (hdmi_deep_color_possible(pipe_config, 12) &&
> > - hdmi_port_clock_valid(intel_hdmi, clock_12bpc,
> > - true, force_dvi) == MODE_OK) {
> > - DRM_DEBUG_KMS("picking bpc to 12 for HDMI output\n");
> > - desired_bpp = 12*3;
> > -
> > - /* Need to adjust the port link by 1.5x for 12bpc. */
> > - pipe_config->port_clock = clock_12bpc;
> > - } else if (hdmi_deep_color_possible(pipe_config, 10) &&
> > - hdmi_port_clock_valid(intel_hdmi, clock_10bpc,
> > - true, force_dvi) == MODE_OK) {
> > - DRM_DEBUG_KMS("picking bpc to 10 for HDMI output\n");
> > - desired_bpp = 10 * 3;
> > -
> > - /* Need to adjust the port link by 1.25x for 10bpc. */
> > - pipe_config->port_clock = clock_10bpc;
> > - } else {
> > - DRM_DEBUG_KMS("picking bpc to 8 for HDMI output\n");
> > - desired_bpp = 8*3;
> > + /* for 8bpc */
> > + pipe_config->port_clock = clock;
> >
> > - pipe_config->port_clock = clock_8bpc;
> > - }
> > + for (bpc = 12; bpc > 8; bpc -= 2) {
>
> So I generally don't like loops like this. You enter the loop with 12
> and 10, don't enter with 8 but assume in later code that's what you got
> if not 12 or 10. And, of course, you need to initialize
> pipe_config->port_clock beforehand for that case too.
>
> Perhaps abstracting the loop to a function would help.
Hmm. I guess. Though maybe I should split even more junk into a few
new functions...
>
> Now, I think this is still an improvement on the status quo, so up to
> you whether to do that now or later.
>
> > + int port_clock;
> > +
> > + if (!hdmi_deep_color_possible(pipe_config, bpc))
> > + continue;
> > +
> > + /*
> > + * Need to adjust the port link by:
> > + * 1.5x for 12bpc
> > + * 1.25x for 10bpc
> > + */
> > + port_clock = clock * bpc / 8;
> > +
> > + if (hdmi_port_clock_valid(intel_hdmi, port_clock,
> > + true, force_dvi) != MODE_OK)
> > + continue;
> >
> > - if (!pipe_config->bw_constrained) {
> > - DRM_DEBUG_KMS("forcing pipe bpp to %i for HDMI\n", desired_bpp);
> > - pipe_config->pipe_bpp = desired_bpp;
> > + pipe_config->port_clock = port_clock;
> > + break;
> > }
> >
> > + /*
> > + * pipe_bpp could already be below 8bpc due to
> > + * FDI bandwidth constraints. We shouldn't bump it
> > + * back up to 8bpc in that case.
> > + */
> > + if (pipe_config->pipe_bpp > bpc * 3)
> > + pipe_config->pipe_bpp = bpc * 3;
> > + DRM_DEBUG_KMS("picking %dbpc for HDMI output (pipe bpp: %d)\n",
> ^
>
> Put a space there?
Maybe. The comments in the code are w/o the space, but I guess we could
stick it in there anyway.
>
> Regardless of the nitpicks,
>
> Reviewed-by: Jani Nikula <jani.nikula@xxxxxxxxx>
>
>
>
> > + bpc, pipe_config->pipe_bpp);
> > +
> > if (hdmi_port_clock_valid(intel_hdmi, pipe_config->port_clock,
> > false, force_dvi) != MODE_OK) {
> > - DRM_DEBUG_KMS("unsupported HDMI clock, rejecting mode\n");
> > + DRM_DEBUG_KMS("unsupported HDMI clock (%d kHz), rejecting mode\n",
> > + pipe_config->port_clock);
> > return -EINVAL;
> > }
>
> --
> Jani Nikula, Intel Open Source Graphics Center
--
Ville Syrjälä
Intel
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
