Re: [PATCH 1/2] drm/i915/display_wa: Add helpers to check wa

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

On Mon, 30 Jun 2025, Ankit Nautiyal <ankit.k.nautiyal@intel.com> wrote:
> Introduce a generic helper to check display workarounds using an enum.
>
> Convert Wa_16023588340 to use the new interface, simplifying WA checks
> and making future additions easier.
>
> Suggested-by: Jani Nikula <jani.nikula@intel.com>
> Signed-off-by: Ankit Nautiyal <ankit.k.nautiyal@intel.com>
> ---
>  drivers/gpu/drm/i915/display/intel_display_wa.c | 13 +++++++++++++
>  drivers/gpu/drm/i915/display/intel_display_wa.h |  9 +++++++++
>  drivers/gpu/drm/i915/display/intel_fbc.c        |  2 +-
>  3 files changed, 23 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_display_wa.c b/drivers/gpu/drm/i915/display/intel_display_wa.c
> index f57280e9d041..70ba66fc7e26 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_wa.c
> +++ b/drivers/gpu/drm/i915/display/intel_display_wa.c
> @@ -39,3 +39,16 @@ void intel_display_wa_apply(struct intel_display *display)
>  	else if (DISPLAY_VER(display) == 11)
>  		gen11_display_wa_apply(display);
>  }
> +
> +bool __intel_display_wa(struct intel_display *display, enum intel_display_wa wa)
> +{
> +	switch (wa) {
> +	case INTEL_DISPLAY_WA_16023588340:
> +		return intel_display_needs_wa_16023588340(display);
> +	default:
> +		MISSING_CASE(wa);

MISSING_CASE() is a bit of a problem for i915 and display
separation. Please let's not add more. Use drm_WARN() or something.

> +		break;
> +	}
> +
> +	return false;
> +}
> diff --git a/drivers/gpu/drm/i915/display/intel_display_wa.h b/drivers/gpu/drm/i915/display/intel_display_wa.h
> index babd9d16603d..853939ebf1ac 100644
> --- a/drivers/gpu/drm/i915/display/intel_display_wa.h
> +++ b/drivers/gpu/drm/i915/display/intel_display_wa.h
> @@ -7,6 +7,7 @@
>  #define __INTEL_DISPLAY_WA_H__
>  
>  #include <linux/types.h>
> +#include <i915_utils.h>

Please don't put that in the header. And shouldn't be needed anyway if
you drop the MISSING_CASE().

>  
>  struct intel_display;
>  
> @@ -21,4 +22,12 @@ static inline bool intel_display_needs_wa_16023588340(struct intel_display *disp
>  bool intel_display_needs_wa_16023588340(struct intel_display *display);
>  #endif
>  
> +enum intel_display_wa {
> +	INTEL_DISPLAY_WA_16023588340,
> +};
> +
> +bool __intel_display_wa(struct intel_display *display, enum intel_display_wa wa);
> +
> +#define _intel_display_wa_expand(__wa) INTEL_DISPLAY_WA_##__wa

Can't we just do this inline instead of adding another macro?

> +#define intel_display_wa(__display, __wa) __intel_display_wa((__display), _intel_display_wa_expand(__wa))
>  #endif
> diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
> index ec1ef8694c35..f4b7ff549fd4 100644
> --- a/drivers/gpu/drm/i915/display/intel_fbc.c
> +++ b/drivers/gpu/drm/i915/display/intel_fbc.c
> @@ -1464,7 +1464,7 @@ static int intel_fbc_check_plane(struct intel_atomic_state *state,
>  		return 0;
>  	}
>  
> -	if (intel_display_needs_wa_16023588340(display)) {
> +	if (intel_display_wa(display, 16023588340)) {
>  		plane_state->no_fbc_reason = "Wa_16023588340";
>  		return 0;
>  	}

-- 
Jani Nikula, Intel