Re: [PATCH] drm/xe/display: read PCON capability only when present

[Imre Deak <imre.deak@intel.com>]

On Wed, Jun 18, 2025 at 05:15:53PM +0530, Chaitanya Kumar Borah wrote:
> Read PCON capabilities only when it is present. This will prevent
> spurious DCPD read failures.
> 
> [drm] *ERROR* Failed to read DPCD register 0x92

The above read failure is due to the sink not being connected during the
read access (in some Chamelium HPD test), not because PCON not being
present in the sink. If sink is connected and its HPD is asserted the
read would not fail, just return all 0s. So the commit message should
just say stg like "avoid reading the PCON capabilities redundantly on
non-branch devices".

> 
> Signed-off-by: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
> ---
>  drivers/gpu/drm/i915/display/intel_dp.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> index 277b40b13948..8a1c2a37a56b 100644
> --- a/drivers/gpu/drm/i915/display/intel_dp.c
> +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> @@ -5681,7 +5681,8 @@ intel_dp_update_dfp(struct intel_dp *intel_dp,
>  		    intel_dp->dfp.max_tmds_clock,
>  		    intel_dp->dfp.pcon_max_frl_bw);
>  
> -	intel_dp_get_pcon_dsc_cap(intel_dp);
> +	if (drm_dp_is_branch(intel_dp->dpcd))
> +		intel_dp_get_pcon_dsc_cap(intel_dp);

Clearing intel_dp->pcon_dsc_dpcd should still happen, you could move the
above check after that happens in intel_dp_get_pcon_dsc_cap().

>  }
>  
>  static bool
> -- 
> 2.25.1
>