[Intel-gfx] [PATCH 1/3] drm/i915: Check pixel rate for DP to VGA dongle

[Ville Syrjälä]

On Tue, May 03, 2016 at 04:28:00PM +0200, Daniel Vetter wrote:
> On Tue, May 03, 2016 at 04:23:34PM +0300, Ville Syrjälä wrote:
> > On Tue, May 03, 2016 at 02:46:36PM +0300, Mika Kahola wrote:
> > > Prep work to improve DP branch device handling.
> > > 
> > > Filter out a mode that exceeds the max pixel rate setting
> > > for DP to VGA dongle. This is defined in DPCD register 0x81
> > > if detailed cap info i.e. info field is 4 bytes long and
> > > it is available for DP downstream port.
> > > 
> > > The register defines the pixel rate divided by 8 in MP/s.
> > > 
> > > Signed-off-by: Mika Kahola <mika.kahola at intel.com>
> > > ---
> > >  drivers/gpu/drm/i915/intel_dp.c  | 34 ++++++++++++++++++++++++++++++++++
> > >  drivers/gpu/drm/i915/intel_drv.h |  9 +++++++++
> > >  2 files changed, 43 insertions(+)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
> > > index 3633002..74a04ce 100644
> > > --- a/drivers/gpu/drm/i915/intel_dp.c
> > > +++ b/drivers/gpu/drm/i915/intel_dp.c
> > > @@ -201,6 +201,13 @@ intel_dp_mode_valid(struct drm_connector *connector,
> > >  	int max_rate, mode_rate, max_lanes, max_link_clock;
> > >  	int max_dotclk = to_i915(connector->dev)->max_dotclk_freq;
> > >  
> > > +	/* DP to VGA dongle may define max pixel rate in DPCD */
> > > +	if (intel_dp->dfp.present &&
> > > +	    intel_dp->dfp.detailed_cap_info &&
> > > +	    (intel_dp->dfp.type & DP_DS_PORT_TYPE_VGA) &&
> > > +	    (intel_dp->dfp.dot_clk > 0))
> > > +		max_dotclk = min(max_dotclk, intel_dp->dfp.dot_clk);
> > 
> > What's dfp?
> > 
> > Looks like most of this stuff is not really needed. Just storing a max
> > dotclock per downstream port would seem to suffice.
> > 
> > > +
> > >  	if (is_edp(intel_dp) && fixed_mode) {
> > >  		if (mode->hdisplay > fixed_mode->hdisplay)
> > >  			return MODE_PANEL;
> > > @@ -4566,6 +4573,28 @@ static const struct drm_encoder_funcs intel_dp_enc_funcs = {
> > >  	.destroy = intel_dp_encoder_destroy,
> > >  };
> > >  
> > > +static void intel_dp_get_dfp(struct intel_dp *intel_dp)
> > > +{
> > > +	uint8_t dfp_info[4];
> > > +
> > > +	intel_dp->dfp.detailed_cap_info = intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] & DP_DETAILED_CAP_INFO_AVAILABLE;
> > > +
> > > +	if (drm_dp_dpcd_read(&intel_dp->aux, DP_DOWNSTREAM_PORT_0, dfp_info, 4) < 0) {
> > > +		intel_dp->dfp.present = false;
> > > +		intel_dp->dfp.detailed_cap_info = false;
> > > +		return; /* aux transfer failed */
> > > +	}
> > > +
> > > +	intel_dp->dfp.type = dfp_info[0] & DP_DS_PORT_TYPE_MASK;
> > > +
> > > +	if (intel_dp->dfp.detailed_cap_info) {
> > > +		if (intel_dp->dfp.type & DP_DS_PORT_TYPE_VGA) {
> > > +			intel_dp->dfp.dot_clk = dfp_info[1] * 8 * 1000;
> > > +			DRM_DEBUG_KMS("max pixel rate for VGA is %d kHz\n", intel_dp->dfp.dot_clk);
> > > +		}
> > 
> > I would suggest putting this sort of stuff into the dp helper. I once
> > started to hatch something to deal with these downstream port limits,
> > but never finished it. I pushed my WIP stuff (mostly ideas how to parse
> > these port caps) to 
> > 
> > git://github.com/vsyrjala/linux.git dp_downstream_ports
> > 
> > maybe you can to see if there's anything useful for you there.
> 
> Seconded on at least moving the computation into the dp helpers. i915.ko
> really should only ask the helpers for the final result, maybe with an
> intermediate step to cache the dp aux register stuff. There's already some
> structures in the dp helpers to store sink state, we could start using
> those (unfortunately they're not agreeing on what the canonical one should
> be).

Yeah that thing is a bit of mess right now. As usual, I have a branch to
clean some of it up a bit. Mainly aiming to respect the sink HDMI TMDS
clock limits better. But I need to get the DP++ stuff landed before
I continue with that.

-- 
Ville Syrjälä
Intel OTC