[Intel-gfx] [PATCH 1/2] drm/i915/mtl: Initial display workarounds

Lucas De Marchi lucas.demarchi at intel.com
Fri Dec 2 01:18:38 UTC 2022


On Thu, Dec 01, 2022 at 03:44:07PM -0800, Matt Roper wrote:
>On Thu, Dec 01, 2022 at 03:27:25PM -0800, Lucas De Marchi wrote:
>> On Thu, Dec 01, 2022 at 03:01:05PM -0800, Matt Roper wrote:
>> > On Wed, Nov 30, 2022 at 03:17:08PM -0800, Matt Atwood wrote:
>> > > From: Jouni Högander <jouni.hogander at intel.com>
>> > >
>> > > This patch introduces initial workarounds for mtl platform
>> > >
>> > > Bspec: 66624
>> > >
>> > > Signed-off-by: Matt Atwood <matthew.s.atwood at intel.com>
>> > > Signed-off-by: Jouni Högander <jouni.hogander at intel.com>
>> > > ---
>> > >  drivers/gpu/drm/i915/display/intel_fbc.c  |  4 +++-
>> > >  drivers/gpu/drm/i915/display/intel_hdmi.c |  3 ++-
>> > >  drivers/gpu/drm/i915/display/intel_psr.c  | 27 ++++++++++++++++-------
>> > >  drivers/gpu/drm/i915/i915_drv.h           |  4 ++++
>> > >  4 files changed, 28 insertions(+), 10 deletions(-)
>> > >
>> > > diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
>> > > index b5ee5ea0d010..8f269553d826 100644
>> > > --- a/drivers/gpu/drm/i915/display/intel_fbc.c
>> > > +++ b/drivers/gpu/drm/i915/display/intel_fbc.c
>> > > @@ -1095,7 +1095,9 @@ static int intel_fbc_check_plane(struct intel_atomic_state *state,
>> > >  	}
>> > >
>> > >  	/* Wa_14016291713 */
>> > > -	if (IS_DISPLAY_VER(i915, 12, 13) && crtc_state->has_psr) {
>> > > +	if ((IS_DISPLAY_VER(i915, 12, 13) ||
>> > > +	     IS_MTL_DISPLAY_STEP(i915, STEP_A0, STEP_C0)) &&
>> > > +	    crtc_state->has_psr) {
>> > >  		plane_state->no_fbc_reason = "PSR1 enabled (Wa_14016291713)";
>> > >  		return 0;
>> > >  	}
>> > > diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c
>> > > index e82f8a07e2b0..efa2da080f62 100644
>> > > --- a/drivers/gpu/drm/i915/display/intel_hdmi.c
>> > > +++ b/drivers/gpu/drm/i915/display/intel_hdmi.c
>> > > @@ -537,7 +537,8 @@ void hsw_write_infoframe(struct intel_encoder *encoder,
>> > >  			       0);
>> > >
>> > >  	/* Wa_14013475917 */
>> > > -	if (DISPLAY_VER(dev_priv) == 13 && crtc_state->has_psr &&
>> > > +	if ((DISPLAY_VER(dev_priv) == 13 ||
>> > > +	     IS_MTL_DISPLAY_STEP(dev_priv, STEP_A0, STEP_B0)) && crtc_state->has_psr &&
>> > >  	    type == DP_SDP_VSC)
>> > >  		return;
>> > >
>> > > diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c
>> > > index 5b678916e6db..7982157fb1ff 100644
>> > > --- a/drivers/gpu/drm/i915/display/intel_psr.c
>> > > +++ b/drivers/gpu/drm/i915/display/intel_psr.c
>> > > @@ -797,7 +797,7 @@ static bool psr2_granularity_check(struct intel_dp *intel_dp,
>> > >  		return intel_dp->psr.su_y_granularity == 4;
>> > >
>> > >  	/*
>> > > -	 * adl_p and display 14+ platforms has 1 line granularity.
>> > > +	 * adl_p and mtl platforms has 1 line granularity.
>> > >  	 * For other platforms with SW tracking we can adjust the y coordinates
>> > >  	 * to match sink requirement if multiple of 4.
>> > >  	 */
>> > > @@ -1170,11 +1170,14 @@ static void intel_psr_enable_source(struct intel_dp *intel_dp,
>> > >  				     PSR2_ADD_VERTICAL_LINE_COUNT);
>> > >
>> > >  		/*
>> > > -		 * Wa_16014451276:adlp
>> > > +		 * Wa_16014451276:adlp,mtl[a0,b0]
>> >
>> > These days we don't really add steppings in these comments; at best
>> > they're just reiterating information that can be easily seen from the
>> > code below, at worst they get out of sync and cause confusion.  I'd drop
>> > the "[a0,b0]" part (which also isn't accurate anyway if you're using
>> > range notation...it would be "[a0..b0)" in that case).
>> >
>> > Honestly even the list of platform names on workarounds is of
>> > questionable value and I'm thinking about writing a patch that just
>> > drops all of them throughout i915, leaving just the workaround lineage
>> > number.  Maybe I'd keep the platform names in the few cases where we
>> > have multiple workaround numbers (with different sets of platforms) all
>> > requesting the same programming of a register...
>>
>> I'd be happy to ack such patch :)
>>
>> >
>> > >  		 * All supported adlp panels have 1-based X granularity, this may
>> > >  		 * cause issues if non-supported panels are used.
>> > >  		 */
>> > > -		if (IS_ALDERLAKE_P(dev_priv))
>> > > +		if (IS_MTL_DISPLAY_STEP(dev_priv, STEP_A0, STEP_B0))
>> > > +			intel_de_rmw(dev_priv, MTL_CHICKEN_TRANS(cpu_transcoder), 0,
>> > > +				     ADLP_1_BASED_X_GRANULARITY);
>> > > +		else if (IS_ALDERLAKE_P(dev_priv))
>> > >  			intel_de_rmw(dev_priv, CHICKEN_TRANS(cpu_transcoder), 0,
>> > >  				     ADLP_1_BASED_X_GRANULARITY);
>> > >
>> > > @@ -1185,8 +1188,12 @@ static void intel_psr_enable_source(struct intel_dp *intel_dp,
>> > >  				     TRANS_SET_CONTEXT_LATENCY_MASK,
>> > >  				     TRANS_SET_CONTEXT_LATENCY_VALUE(1));
>> > >
>> > > -		/* Wa_16012604467:adlp */
>> > > -		if (IS_ALDERLAKE_P(dev_priv))
>> > > +		/* Wa_16012604467:adlp,mtl[a0,b0] */
>> > > +		if (IS_MTL_DISPLAY_STEP(dev_priv, STEP_A0, STEP_B0))
>> > > +			intel_de_rmw(dev_priv,
>> > > +				     MTL_CLKGATE_DIS_TRANS(cpu_transcoder), 0,
>> > > +				     MTL_CLKGATE_DIS_TRANS_DMASC_GATING_DIS);
>> > > +		else if (IS_ALDERLAKE_P(dev_priv))
>> > >  			intel_de_rmw(dev_priv, CLKGATE_DIS_MISC, 0,
>> > >  				     CLKGATE_DIS_MISC_DMASC_GATING_DIS);
>> > >
>> > > @@ -1362,8 +1369,12 @@ static void intel_psr_disable_locked(struct intel_dp *intel_dp)
>> > >  				     TRANS_SET_CONTEXT_LATENCY(intel_dp->psr.transcoder),
>> > >  				     TRANS_SET_CONTEXT_LATENCY_MASK, 0);
>> > >
>> > > -		/* Wa_16012604467:adlp */
>> > > -		if (IS_ALDERLAKE_P(dev_priv))
>> > > +		/* Wa_16012604467:adlp,mtl[a0,b0] */
>> > > +		if (IS_MTL_DISPLAY_STEP(dev_priv, STEP_A0, STEP_B0))
>> > > +			intel_de_rmw(dev_priv,
>> > > +				     MTL_CLKGATE_DIS_TRANS(intel_dp->psr.transcoder),
>> > > +				     MTL_CLKGATE_DIS_TRANS_DMASC_GATING_DIS, 0);
>> > > +		else if (IS_ALDERLAKE_P(dev_priv))
>> > >  			intel_de_rmw(dev_priv, CLKGATE_DIS_MISC,
>> > >  				     CLKGATE_DIS_MISC_DMASC_GATING_DIS, 0);
>> > >
>> > > @@ -1625,7 +1636,7 @@ static void psr2_man_trk_ctl_calc(struct intel_crtc_state *crtc_state,
>> > >
>> > >  	if (full_update) {
>> > >  		/*
>> > > -		 * Not applying Wa_14014971508:adlp as we do not support the
>> > > +		 * Not applying Wa_14014971508:adlp,mtl as we do not support the
>> > >  		 * feature that requires this workaround.
>> > >  		 */
>> > >  		val |= man_trk_ctl_single_full_frame_bit_get(dev_priv);
>> > > diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
>> > > index a8a5bd426e78..ecb027626a21 100644
>> > > --- a/drivers/gpu/drm/i915/i915_drv.h
>> > > +++ b/drivers/gpu/drm/i915/i915_drv.h
>> > > @@ -727,6 +727,10 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
>> > >  	(IS_SUBPLATFORM(__i915, INTEL_METEORLAKE, INTEL_SUBPLATFORM_##variant) && \
>> > >  	 IS_GRAPHICS_STEP(__i915, since, until))
>> > >
>> > > +#define IS_MTL_DISPLAY_STEP(__i915, since, until) \
>> > > +	(DISPLAY_VER(__i915) == 14 && \
>> >
>> > As Tvrtko noted, this could come back to bite us in the future if
>> > another platform shows up with 14.10, 14.50, etc.  MTL has exactly
>> > version 14.0, so best to make this
>> >
>> >        DISPLAY_VER_FULL(__i915) == IP_VER(14, 0)
>> >
>> > so that it won't automatically capture future platforms by accident.
>>
>> I think it's better to do a platform check as the other platforms are
>> doing. See DG2 for example:
>>
>> #define IS_DG2_DISPLAY_STEP(__i915, since, until) \
>>         (IS_DG2(__i915) && \
>>          IS_DISPLAY_STEP(__i915, since, until))
>
>I guess that's fine in the short term, but in the long term that's the
>kind of thing we need to move away from.  We're really not supposed to
>be using platform checks (which are derived from PCI ID) anymore going
>forward.  At some point, even things like IS_MTL_DISPLAY_STEP() will get
>replaced with something more along the lines of
>
>   IS_GMDID_DISPLAY_STEP(12, 70, STEP_A0, STEP_C0)

I think we should only check for the IP version if we eliminate IS_MTL_
from the name. Or... when the macro is renamed/reimplemented, it may be a good time to
change.


Lucas De Marchi

>because the expectation is that none of this is actually tied to the
>platform anymore, just to the IP versions of the various chiplets (which
>can be mixed-and-matched and in theory could be re-used on other
>platforms).  But today MTL is the first/only hardware we have using this
>design, so it doesn't matter too much; we don't have to clean this all
>up immediately.
>
>
>Matt
>
>>
>> Lucas De Marchi
>
>-- 
>Matt Roper
>Graphics Software Engineer
>VTT-OSGC Platform Enablement
>Intel Corporation


More information about the Intel-gfx mailing list