On Tue, 2018-02-20 at 10:44 -0800, Rodrigo Vivi wrote:
> On Tue, Feb 20, 2018 at 06:25:38PM +0200, Imre Deak wrote:
> > On Tue, Feb 20, 2018 at 05:35:20PM +0200, Imre Deak wrote:
> > > On Mon, Feb 19, 2018 at 05:49:41PM +0200, Ville Syrjälä wrote:
> > > > On Mon, Feb 19, 2018 at 04:47:41PM +0200, Imre Deak wrote:
> > > > > On Fri, Feb 16, 2018 at 04:27:04PM -0800, Rodrigo Vivi wrote:
> > > > > > On Fri, Feb 16, 2018 at 03:25:55PM -0800, Dhinakaran Pandiyan wrote:
> > > > > > > PSR requires AUX IO well to be kept on and the existing AUX domain
> > > > > > > enables other wells too.
> > > > > > >
> > > > > >
> > > > > > Cc: Imre Deak <imre.d...@intel.com>
> > > > > >
> > > > > > > Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandi...@intel.com>
> > > > > > > ---
> > > > > > > drivers/gpu/drm/i915/intel_display.h | 5 +++++
> > > > > > > drivers/gpu/drm/i915/intel_dp.c | 6 ++++++
> > > > > > > drivers/gpu/drm/i915/intel_drv.h | 1 +
> > > > > > > drivers/gpu/drm/i915/intel_psr.c | 4 ++++
> > > > > > > drivers/gpu/drm/i915/intel_runtime_pm.c | 15 +++++++++++++++
> > > > > > > 5 files changed, 31 insertions(+)
> > > > > > >
> > > > > > > diff --git a/drivers/gpu/drm/i915/intel_display.h
> > > > > > > b/drivers/gpu/drm/i915/intel_display.h
> > > > > > > index c4042e342f50..9857b60a7e0f 100644
> > > > > > > --- a/drivers/gpu/drm/i915/intel_display.h
> > > > > > > +++ b/drivers/gpu/drm/i915/intel_display.h
> > > > > > > @@ -175,6 +175,11 @@ enum intel_display_power_domain {
> > > > > > > POWER_DOMAIN_AUX_C,
> > > > > > > POWER_DOMAIN_AUX_D,
> > > > > > > POWER_DOMAIN_AUX_F,
> > > > > > > + POWER_DOMAIN_AUX_IO_A,
> > > > > > > + POWER_DOMAIN_AUX_IO_B,
> > > > > > > + POWER_DOMAIN_AUX_IO_C,
> > > > > > > + POWER_DOMAIN_AUX_IO_D,
> > > > > > > + POWER_DOMAIN_AUX_IO_F,
> > > > > >
> > > > > > I know that I had suggested something very similar, but looking now
> > > > > > to the code it seems a big duplication of all other aux..
> > > > > >
> > > > > > I mean, on CNL
> > > > > >
> > > > > > POWER_DOMAIN_AUX_{B,C,D,F} are exactly the same as
> > > > > > POWER_DOMAIN_AUX__IO_{B,C,D,F}
> > > > > >
> > > > > > Only the AUX_IO A is different from the AUX_A because AUX_A
> > > > > > disables DMC.
> > > > > > So, question is: What about removing the AUX_A domain from DC_OFF
> > > > > > and just using AUX doamins... Well, on this case calling get/put
> > > > > > only for INTEL_GEN <= 10...
> > > > > >
> > > > > > So, yeap, not beautiful anyways.
> > > > > >
> > > > > > I'd like to hear from Imre what he had in mind for handling
> > > > > > AUX IOs and to know if he has any better idea for this case.
> > > > >
> > > > > Hm, so in general (for AUX transfers) to enable AUX-A we first need to
> > > > > disable DC states _except_ if we enable AUX-A for PSR where we want
> > > > > DC5/6 to stay enabled. Nice. Also not sure why the FW/HW can't have
> > > > > its
> > > > > own request bit..
> > > > >
> > > > > I don't have a better idea, but since on DDI where this matters PSR is
> > > > > tied to port A,
> > > >
> > > > SKL+ (or maybe even BDW?) can do PSR on any port actually. We just don't
> > > > make use of that capability currently.
> > >
> > > Ok. But still, one other idea would be to avoid the duplication Rodrigo
> > > also mentioned is to have:
> > >
> > > get_psr_aux_power()
> > > {
> > > if (!HAS_DDI())
> >
> > Err, correctly the above would be if (GEN < 10)
>
> Exactly what I was going to reply here :)
>
> >
> > > return;
> > >
> > > intel_display_power_get(port == PORT_A ? POWER_DOMAIN_AUX_IO_A :
> > > intel_dp->aux_power_domain);
>
> good idea. Although as Ville pointed out we don't support PSR on other ports
> it is better to do the right thing at least here.
>
> > > }
> > >
> > > and a put_psr_aux_power() accordingly. That way we only need to add one
> > > extra power domain for port A.
>
> Also maybe add a comment explaining the difference and purpose of
> POWER_DOMAIN_AUX_IO_A
> from POWER_DOMAIN_AUX_A along with POWER_DOMAIN_AUX_IO_A definition.
I'd sent this as a test patch to get a CI run but now that there's a
consensus on how we should do this, I'll write this up as proper patch.
Thanks for the feedback.
-DK
>
> Thanks,
> Rodrigo.
>
> > >
> > > --Imre
> > >
> > >
> > >
> > >
> > > >
> > > > > we could have only a POWER_DOMAIN_PSR and map it to the
> > > > > AUX-A power well where it's needed.
> > > > >
> > > > > --Imre
> > > > >
> > > > > >
> > > > > > > POWER_DOMAIN_GMBUS,
> > > > > > > POWER_DOMAIN_MODESET,
> > > > > > > POWER_DOMAIN_GT_IRQ,
> > > > > > > diff --git a/drivers/gpu/drm/i915/intel_dp.c
> > > > > > > b/drivers/gpu/drm/i915/intel_dp.c
> > > > > > > index f20b25f98e5a..aa90f6a70689 100644
> > > > > > > --- a/drivers/gpu/drm/i915/intel_dp.c
> > > > > > > +++ b/drivers/gpu/drm/i915/intel_dp.c
> > > > > > > @@ -6280,22 +6280,28 @@ intel_dp_init_connector_port_info(struct
> > > > > > > intel_digital_port *intel_dig_port)
> > > > > > > switch (encoder->port) {
> > > > > > > case PORT_A:
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_A;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_A;
> > > > > > > break;
> > > > > > > case PORT_B:
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_B;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_B;
> > > > > > > break;
> > > > > > > case PORT_C:
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_C;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_C;
> > > > > > > break;
> > > > > > > case PORT_D:
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_D;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_D;
> > > > > > > break;
> > > > > > > case PORT_E:
> > > > > > > /* FIXME: Check VBT for actual wiring of PORT E */
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_D;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_D;
> > > > > > > break;
> > > > > > > case PORT_F:
> > > > > > > intel_dp->aux_power_domain = POWER_DOMAIN_AUX_F;
> > > > > > > + intel_dp->aux_io_power_domain = POWER_DOMAIN_AUX_IO_F;
> > > > > > > break;
> > > > > > > default:
> > > > > > > MISSING_CASE(encoder->port);
> > > > > > > diff --git a/drivers/gpu/drm/i915/intel_drv.h
> > > > > > > b/drivers/gpu/drm/i915/intel_drv.h
> > > > > > > index 898064e8bea7..c7c4180e41d8 100644
> > > > > > > --- a/drivers/gpu/drm/i915/intel_drv.h
> > > > > > > +++ b/drivers/gpu/drm/i915/intel_drv.h
> > > > > > > @@ -1071,6 +1071,7 @@ struct intel_dp {
> > > > > > > struct drm_dp_desc desc;
> > > > > > > struct drm_dp_aux aux;
> > > > > > > enum intel_display_power_domain aux_power_domain;
> > > > > > > + enum intel_display_power_domain aux_io_power_domain;
> > > > > > > uint8_t train_set[4];
> > > > > > > int panel_power_up_delay;
> > > > > > > int panel_power_down_delay;
> > > > > > > diff --git a/drivers/gpu/drm/i915/intel_psr.c
> > > > > > > b/drivers/gpu/drm/i915/intel_psr.c
> > > > > > > index 2ef374f936b9..62de8f0632b0 100644
> > > > > > > --- a/drivers/gpu/drm/i915/intel_psr.c
> > > > > > > +++ b/drivers/gpu/drm/i915/intel_psr.c
> > > > > > > @@ -484,6 +484,8 @@ static void hsw_psr_enable_source(struct
> > > > > > > intel_dp *intel_dp,
> > > > > > > EDP_PSR_DEBUG_MASK_HPD |
> > > > > > > EDP_PSR_DEBUG_MASK_LPSP);
> > > > > > > }
> > > > > > > +
> > > > > > > + intel_display_power_get(dev_priv,
> > > > > > > intel_dp->aux_io_power_domain);
> > > > > > > }
> > > > > > >
> > > > > > > /**
> > > > > > > @@ -617,6 +619,8 @@ static void hsw_psr_disable(struct intel_dp
> > > > > > > *intel_dp,
> > > > > > > else
> > > > > > > WARN_ON(I915_READ(EDP_PSR_CTL) &
> > > > > > > EDP_PSR_ENABLE);
> > > > > > > }
> > > > > > > +
> > > > > > > + intel_display_power_put(dev_priv,
> > > > > > > intel_dp->aux_io_power_domain);
> > > > > > > }
> > > > > > >
> > > > > > > /**
> > > > > > > diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c
> > > > > > > b/drivers/gpu/drm/i915/intel_runtime_pm.c
> > > > > > > index 16790f2576ec..323c1babb0a0 100644
> > > > > > > --- a/drivers/gpu/drm/i915/intel_runtime_pm.c
> > > > > > > +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c
> > > > > > > @@ -130,6 +130,16 @@ intel_display_power_domain_str(enum
> > > > > > > intel_display_power_domain domain)
> > > > > > > return "AUX_D";
> > > > > > > case POWER_DOMAIN_AUX_F:
> > > > > > > return "AUX_F";
> > > > > > > + case POWER_DOMAIN_AUX_IO_A:
> > > > > > > + return "AUX_IO_A";
> > > > > > > + case POWER_DOMAIN_AUX_IO_B:
> > > > > > > + return "AUX_IO_B";
> > > > > > > + case POWER_DOMAIN_AUX_IO_C:
> > > > > > > + return "AUX_IO_C";
> > > > > > > + case POWER_DOMAIN_AUX_IO_D:
> > > > > > > + return "AUX_IO_D";
> > > > > > > + case POWER_DOMAIN_AUX_IO_F:
> > > > > > > + return "AUX_IO_F";
> > > > > > > case POWER_DOMAIN_GMBUS:
> > > > > > > return "GMBUS";
> > > > > > > case POWER_DOMAIN_INIT:
> > > > > > > @@ -1853,18 +1863,23 @@ void intel_display_power_put(struct
> > > > > > > drm_i915_private *dev_priv,
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_AUX_A_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_AUX_A) | \
> > > > > > > + BIT_ULL(POWER_DOMAIN_AUX_IO_A) | \
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_AUX_B_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_AUX_B) | \
> > > > > > > + BIT_ULL(POWER_DOMAIN_AUX_IO_B) | \
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_AUX_C_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_AUX_C) | \
> > > > > > > + BIT_ULL(POWER_DOMAIN_AUX_IO_C) | \
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_AUX_D_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_AUX_D) | \
> > > > > > > + BIT_ULL(POWER_DOMAIN_AUX_IO_D) | \
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_AUX_F_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_AUX_F) | \
> > > > > > > + BIT_ULL(POWER_DOMAIN_AUX_IO_F) | \
> > > > > > > BIT_ULL(POWER_DOMAIN_INIT))
> > > > > > > #define CNL_DISPLAY_DDI_F_IO_POWER_DOMAINS ( \
> > > > > > > BIT_ULL(POWER_DOMAIN_PORT_DDI_F_IO) | \
> > > > > > > --
> > > > > > > 2.14.1
> > > > > > >
> > > > > > > _______________________________________________
> > > > > > > Intel-gfx mailing list
> > > > > > > Intel-gfx@lists.freedesktop.org
> > > > > > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > > > > _______________________________________________
> > > > > Intel-gfx mailing list
> > > > > Intel-gfx@lists.freedesktop.org
> > > > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> > > >
> > > > --
> > > > Ville Syrjälä
> > > > Intel OTC
> > > _______________________________________________
> > > Intel-gfx mailing list
> > > Intel-gfx@lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/intel-gfx
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
