Re: [PATCH v2 3/4] drm/dp: Add an EDID quirk for the DPCD register access probe

[Jani Nikula <jani.nikula@linux.intel.com>]

On Tue, 03 Jun 2025, Imre Deak <imre.deak@intel.com> wrote:
> Reading DPCD registers has side-effects and some of these can cause a
> problem for instance during link training. Based on this it's better to
> avoid the probing quirk done before each DPCD register read, limiting
> this to the monitor which requires it. Add an EDID quirk for this. Leave
> the quirk enabled by default, allowing it to be disabled after the
> monitor is detected.
>
> v2: Fix lockdep wrt. drm_dp_aux::hw_mutex when calling
>     drm_dp_dpcd_set_probe_quirk() with a dependent lock already held.
>
> Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
> Cc: Jani Nikula <jani.nikula@linux.intel.com>
> Signed-off-by: Imre Deak <imre.deak@intel.com>
> ---
>  drivers/gpu/drm/display/drm_dp_helper.c | 13 ++++++++++++-
>  drivers/gpu/drm/drm_edid.c              |  3 +++
>  include/drm/display/drm_dp_helper.h     |  6 ++++++
>  include/drm/drm_edid.h                  |  3 ++-
>  4 files changed, 23 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/display/drm_dp_helper.c b/drivers/gpu/drm/display/drm_dp_helper.c
> index dc622c78db9d4..4dad677ac6ebe 100644
> --- a/drivers/gpu/drm/display/drm_dp_helper.c
> +++ b/drivers/gpu/drm/display/drm_dp_helper.c
> @@ -691,6 +691,17 @@ void drm_dp_dpcd_set_powered(struct drm_dp_aux *aux, bool powered)
>  }
>  EXPORT_SYMBOL(drm_dp_dpcd_set_powered);
>  
> +/**
> + * drm_dp_dpcd_set_dpcd_probe_quirk() - Set whether a probing before DPCD access is done
> + * @aux: DisplayPort AUX channel
> + * @enable: Enable the probing if required
> + */
> +void drm_dp_dpcd_set_probe_quirk(struct drm_dp_aux *aux, bool enable)
> +{
> +	WRITE_ONCE(aux->dpcd_probe_disabled, !enable);
> +}
> +EXPORT_SYMBOL(drm_dp_dpcd_set_probe_quirk);

We don't use this yet, which feels a bit odd.

> +
>  /**
>   * drm_dp_dpcd_read() - read a series of bytes from the DPCD
>   * @aux: DisplayPort AUX channel (SST or MST)
> @@ -724,7 +735,7 @@ ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux, unsigned int offset,
>  	 * We just have to do it before any DPCD access and hope that the
>  	 * monitor doesn't power down exactly after the throw away read.
>  	 */
> -	if (!aux->is_remote) {
> +	if (!aux->is_remote && !READ_ONCE(aux->dpcd_probe_disabled)) {

I think it would be worth it to add a small helper to decide whether to
do a dpcd probe. It would include the ->is_remote check as well, and the
big comment could be moved there, simplifying drm_dp_dpcd_read().

>  		ret = drm_dp_dpcd_probe(aux, DP_LANE0_1_STATUS);
>  		if (ret < 0)
>  			return ret;
> diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
> index e867315253493..9250b425ae230 100644
> --- a/drivers/gpu/drm/drm_edid.c
> +++ b/drivers/gpu/drm/drm_edid.c
> @@ -255,6 +255,9 @@ static const struct edid_quirk {
>  };
>  
>  static const struct edid_quirk global_edid_quirk_list[] = {
> +	/* HP ZR24w DP AUX DPCD access requires probing to prevent corruption. */
> +	DRM_EDID_QUIRK(drm_edid_encode_panel_id('H', 'W', 'P', 0x2869), PANEL_NAME_ANY,
> +		       BIT(DRM_EDID_QUIRK_DP_DPCD_PROBE)),

So I think we should keep using EDID_QUIRK(), and maybe add
EDID_QUIRK_NAME() companion to also assign a name != NULL, so we don't
have to pass PANEL_NAME_ANY to all uses. Less verbose that way I think.

>  };
>  
>  /*
> diff --git a/include/drm/display/drm_dp_helper.h b/include/drm/display/drm_dp_helper.h
> index e4ca35143ff96..75fa9548aefa0 100644
> --- a/include/drm/display/drm_dp_helper.h
> +++ b/include/drm/display/drm_dp_helper.h
> @@ -523,10 +523,16 @@ struct drm_dp_aux {
>  	 * @no_zero_sized: If the hw can't use zero sized transfers (NVIDIA)
>  	 */
>  	bool no_zero_sized;
> +
> +	/**
> +	 * @dpcd_probe_disabled: If probing before a DPCD access is disabled.
> +	 */
> +	bool dpcd_probe_disabled;

Is this a negative just so it's false by default?

>  };
>  
>  int drm_dp_dpcd_probe(struct drm_dp_aux *aux, unsigned int offset);
>  void drm_dp_dpcd_set_powered(struct drm_dp_aux *aux, bool powered);
> +void drm_dp_dpcd_set_probe_quirk(struct drm_dp_aux *aux, bool enable);
>  ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux, unsigned int offset,
>  			 void *buffer, size_t size);
>  ssize_t drm_dp_dpcd_write(struct drm_dp_aux *aux, unsigned int offset,
> diff --git a/include/drm/drm_edid.h b/include/drm/drm_edid.h
> index 3d8e168521c82..a878805c81d97 100644
> --- a/include/drm/drm_edid.h
> +++ b/include/drm/drm_edid.h
> @@ -110,7 +110,8 @@ struct detailed_data_string {
>  #define DRM_EDID_CVT_FLAGS_REDUCED_BLANKING  (1 << 4)
>  
>  enum drm_edid_quirk {
> -	DRM_EDID_QUIRK_NONE,
> +	/* Do a dummy read before DPCD accesses, to prevent corruption. */
> +	DRM_EDID_QUIRK_DP_DPCD_PROBE,
>  };
>  
>  struct detailed_data_monitor_range {

-- 
Jani Nikula, Intel