[PATCH 15/20] drm/i915/dp: Read/ack sink count and sink IRQs for SST as it's done for MST
Imre Deak
imre.deak at intel.com
Thu Jul 3 13:14:08 UTC 2025
On Thu, Jul 03, 2025 at 04:02:18PM +0300, Luca Coelho wrote:
> On Thu, 2025-06-26 at 11:20 +0300, Imre Deak wrote:
> > From: Imre Deak <imre.deak at gmail.com>
> >
> > Read and ack the sink count, sink device and link service IRQs for SST
> > the same way this is done for MST, the read/ack happening in separate
> > steps via an ESI (Event Status Indicator) vector.
> >
> > The above way is more efficient, since on newer (DPCD_REV >= 1.2) sinks
> > the DP_SINK_COUNT_ESI..DP_LINK_SERVICE_IRQ_VECTOR_ESI0 registers can be
> > read out in one AUX transaction - and the 3 last one written in one
> > transaction. Also this allows sharing more of the SST and MST IRQ
> > handling code (done as a follow-up).
> >
> > For now keep the current behavior of always reading the legacy
> > DP_SINK_COUNT, DP_DEVICE_SERVICE_IRQ_VECTOR registers and not reading
> > the DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1 register.
> >
> > Signed-off-by: Imre Deak <imre.deak at gmail.com>
> > ---
> > drivers/gpu/drm/i915/display/intel_dp.c | 132 +++++++++++++-----------
> > 1 file changed, 73 insertions(+), 59 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
> > index 2ba4a810f22c2..2e6ed7d2a64a6 100644
> > --- a/drivers/gpu/drm/i915/display/intel_dp.c
> > +++ b/drivers/gpu/drm/i915/display/intel_dp.c
> > @@ -4573,6 +4573,70 @@ static bool intel_dp_ack_sink_irq_esi(struct intel_dp *intel_dp, u8 esi[4])
> > return false;
> > }
> >
> > +static bool intel_dp_get_sink_irq_esi_sst(struct intel_dp *intel_dp, u8 esi[4])
> > +{
> > + memset(esi, 0, 4);
> > +
> > + /*
> > + * TODO: For DP_DPCD_REV >= 0x12 read
> > + * DP_SINK_COUNT_ESI and DP_DEVICE_SERVICE_IRQ_VECTOR_ESI0.
> > + */
> > + if (drm_dp_dpcd_read_data(&intel_dp->aux, DP_SINK_COUNT, esi, 2) != 0)
> > + return false;
> > +
> > + if (intel_dp->dpcd[DP_DPCD_REV] < DP_DPCD_REV_12)
> > + return true;
> > +
> > + /* TODO: Read DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1 as well */
> > + if (drm_dp_dpcd_read_byte(&intel_dp->aux, DP_LINK_SERVICE_IRQ_VECTOR_ESI0, &esi[3]) != 0)
> > + return false;
> > +
> > + return true;
> > +}
> > +
> > +static bool intel_dp_ack_sink_irq_esi_sst(struct intel_dp *intel_dp, u8 esi[4])
> > +{
> > + /*
> > + * TODO: For DP_DPCD_REV >= 0x12 write
> > + * DP_DEVICE_SERVICE_IRQ_VECTOR_ESI0
> > + */
> > + if (drm_dp_dpcd_write_byte(&intel_dp->aux, DP_DEVICE_SERVICE_IRQ_VECTOR, esi[1]) != 0)
> > + return false;
> > +
> > + if (intel_dp->dpcd[DP_DPCD_REV] < DP_DPCD_REV_12)
> > + return true;
> > +
> > + /* TODO: Read DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1 as well */
> > + if (drm_dp_dpcd_write_byte(&intel_dp->aux, DP_LINK_SERVICE_IRQ_VECTOR_ESI0, esi[3]) != 0)
> > + return false;
> > +
> > + return true;
> > +}
> > +
> > +static bool intel_dp_get_and_ack_sink_irq_esi_sst(struct intel_dp *intel_dp, u8 esi[4])
> > +{
> > + struct intel_display *display = to_intel_display(intel_dp);
> > + struct intel_connector *connector = intel_dp->attached_connector;
> > + struct intel_encoder *encoder = &dp_to_dig_port(intel_dp)->base;
> > +
> > + if (!intel_dp_get_sink_irq_esi_sst(intel_dp, esi))
> > + return false;
> > +
> > + drm_dbg_kms(display->drm,
> > + "[CONNECTOR:%d:%s][ENCODER:%d:%s] DPRX ESI: %4ph\n",
> > + connector->base.base.id, connector->base.name,
> > + encoder->base.base.id, encoder->base.name,
> > + esi);
> > +
> > + if (mem_is_zero(&esi[1], 3))
> > + return true;
> > +
> > + if (!intel_dp_ack_sink_irq_esi_sst(intel_dp, esi))
> > + return false;
> > +
> > + return true;
> > +}
> > +
>
> Again, I think it's better to propagate the error than to swallow it
> and return a bool.
I agree. But doing that would make these functions return error in
different ways than the MST
intel_dp_get_sink_irq_esi(), intel_dp_ack_sink_irq_esi()
functions, which return a pass/fail bool. Imo the error return should be
the same for both the SST and MST variety of functions and converting
to propagate an error instead of a pass/fail bool should be done for
both (SST and MST), which is best done as a follow-up. Are you ok with
that?
> Other than that, it looks good to me. So if you agree with this
> change:
>
> Reviewed-by: Luca Coelho <luciano.coelho at tintel.com>
>
> --
> Cheers,
> Luca.
>
> > @@ -5393,27 +5457,6 @@ void intel_dp_check_link_state(struct intel_dp *intel_dp)
> > intel_encoder_link_check_queue_work(encoder, 0);
> > }
> >
> > -static bool intel_dp_get_and_ack_device_service_irq(struct intel_dp *intel_dp, u8 *irq_mask)
> > -{
> > - u8 val;
> > -
> > - *irq_mask = 0;
> > -
> > - if (drm_dp_dpcd_readb(&intel_dp->aux,
> > - DP_DEVICE_SERVICE_IRQ_VECTOR, &val) != 1)
> > - return false;
> > -
> > - if (!val)
> > - return true;
> > -
> > - if (drm_dp_dpcd_writeb(&intel_dp->aux, DP_DEVICE_SERVICE_IRQ_VECTOR, val) != 1)
> > - return false;
> > -
> > - *irq_mask = val;
> > -
> > - return true;
> > -}
> > -
> > static void intel_dp_handle_device_service_irq(struct intel_dp *intel_dp, u8 irq_mask)
> > {
> > struct intel_display *display = to_intel_display(intel_dp);
> > @@ -5428,31 +5471,6 @@ static void intel_dp_handle_device_service_irq(struct intel_dp *intel_dp, u8 irq
> > drm_dbg_kms(display->drm, "Sink specific irq unhandled\n");
> > }
> >
> > -static bool intel_dp_get_and_ack_link_service_irq(struct intel_dp *intel_dp, u8 *irq_mask)
> > -{
> > - u8 val;
> > -
> > - *irq_mask = 0;
> > -
> > - if (intel_dp->dpcd[DP_DPCD_REV] < DP_DPCD_REV_12)
> > - return true;
> > -
> > - if (drm_dp_dpcd_readb(&intel_dp->aux,
> > - DP_LINK_SERVICE_IRQ_VECTOR_ESI0, &val) != 1)
> > - return false;
> > -
> > - if (!val)
> > - return true;
> > -
> > - if (drm_dp_dpcd_writeb(&intel_dp->aux,
> > - DP_LINK_SERVICE_IRQ_VECTOR_ESI0, val) != 1)
> > - return false;
> > -
> > - *irq_mask = val;
> > -
> > - return true;
> > -}
> > -
> > static bool intel_dp_handle_link_service_irq(struct intel_dp *intel_dp, u8 irq_mask)
> > {
> > struct intel_display *display = to_intel_display(intel_dp);
> > @@ -5489,30 +5507,26 @@ static bool
> > intel_dp_short_pulse(struct intel_dp *intel_dp)
> > {
> > bool reprobe_needed = false;
> > - u8 irq_mask;
> > + u8 esi[4] = {};
> >
> > intel_dp_test_reset(intel_dp);
> >
> > + if (!intel_dp_get_and_ack_sink_irq_esi_sst(intel_dp, esi))
> > + return false;
> > +
> > /*
> > - * Now read the DPCD to see if it's actually running
> > * If the current value of sink count doesn't match with
> > - * the value that was stored earlier or dpcd read failed
> > - * we need to do full detection
> > + * the value that was stored earlier we need to do full
> > + * detection.
> > */
> > if (intel_dp_has_sink_count(intel_dp) &&
> > - drm_dp_read_sink_count(&intel_dp->aux) != intel_dp->sink_count)
> > + DP_GET_SINK_COUNT(esi[0]) != intel_dp->sink_count)
> > /* No need to proceed if we are going to do full detect */
> > return false;
> >
> > - if (!intel_dp_get_and_ack_device_service_irq(intel_dp, &irq_mask))
> > - return false;
> > + intel_dp_handle_device_service_irq(intel_dp, esi[1]);
> >
> > - intel_dp_handle_device_service_irq(intel_dp, irq_mask);
> > -
> > - if (!intel_dp_get_and_ack_link_service_irq(intel_dp, &irq_mask))
> > - return false;
> > -
> > - if (intel_dp_handle_link_service_irq(intel_dp, irq_mask))
> > + if (intel_dp_handle_link_service_irq(intel_dp, esi[3]))
> > reprobe_needed = true;
> >
> > /* Handle CEC interrupts, if any */
More information about the Intel-xe
mailing list