[Intel-gfx] [PATCH 2/7] drm/i915/psr: Try to program link training times correctly

[Daniel Vetter]

On Thu, May 19, 2016 at 04:20:02PM +0530, Jindal, Sonika wrote:
> 
> 
> On 5/18/2016 10:17 PM, Daniel Vetter wrote:
> >Oops. Hw default for programming these fields to 0 is "skip link
> >training". Display won't take that too well usually.
> But we were defaulting it to value 0, which means 500us for both TP1 and TP2
> or TP3 time.
> I dont think it means skip link training. This is just to set the time for
> the patterns.
> Skip aux handshake can happen if bit 12 of SRD_CTL is set.
> 
> Does this solution help in fixing the bug mentioned here?

See the other thread, I misread and yes it does help.
-Daniel

> 
> >
> >v2: Unbotch the math a bit.
> >
> >v3: Drop debug hunk.
> >
> >Tested-by: Lyude <cpaul at redhat.com>
> >Cc: Lyude <cpaul at redhat.com>
> >Cc: stable at vger.kernel.org
> >Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95176
> >Cc: Rodrigo Vivi <rodrigo.vivi at intel.com>
> >Cc: Sonika Jindal <sonika.jindal at intel.com>
> >Cc: Durgadoss R <durgadoss.r at intel.com>
> >Cc: "Pandiyan, Dhinakaran" <dhinakaran.pandiyan at intel.com>
> >Signed-off-by: Daniel Vetter <daniel.vetter at intel.com>
> >---
> >  drivers/gpu/drm/i915/intel_psr.c | 55 ++++++++++++++++++++++++++++++++++------
> >  1 file changed, 47 insertions(+), 8 deletions(-)
> >
> >diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c
> >index c3abae4bc596..a788d1e9589b 100644
> >--- a/drivers/gpu/drm/i915/intel_psr.c
> >+++ b/drivers/gpu/drm/i915/intel_psr.c
> >@@ -280,7 +280,10 @@ static void hsw_psr_enable_source(struct intel_dp *intel_dp)
> >  	 * with the 5 or 6 idle patterns.
> >  	 */
> >  	uint32_t idle_frames = max(6, dev_priv->vbt.psr.idle_frames);
> >-	uint32_t val = 0x0;
> >+	uint32_t val = EDP_PSR_ENABLE;
> >+
> >+	val |= max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT;
> >+	val |= idle_frames << EDP_PSR_IDLE_FRAME_SHIFT;
> >  	if (IS_HASWELL(dev))
> >  		val |= EDP_PSR_MIN_LINK_ENTRY_TIME_8_LINES;
> >@@ -288,14 +291,50 @@ static void hsw_psr_enable_source(struct intel_dp *intel_dp)
> >  	if (dev_priv->psr.link_standby)
> >  		val |= EDP_PSR_LINK_STANDBY;
> >-	I915_WRITE(EDP_PSR_CTL, val |
> >-		   max_sleep_time << EDP_PSR_MAX_SLEEP_TIME_SHIFT |
> >-		   idle_frames << EDP_PSR_IDLE_FRAME_SHIFT |
> >-		   EDP_PSR_ENABLE);
> >+	if (dev_priv->vbt.psr.tp1_wakeup_time > 5)
> >+		val |= EDP_PSR_TP1_TIME_2500us;
> >+	else if (dev_priv->vbt.psr.tp1_wakeup_time > 1)
> >+		val |= EDP_PSR_TP1_TIME_500us;
> >+	else if (dev_priv->vbt.psr.tp1_wakeup_time > 0)
> >+		val |= EDP_PSR_TP1_TIME_100us;
> >+	else
> >+		val |= EDP_PSR_TP1_TIME_0us;
> >+
> >+	if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 5)
> >+		val |= EDP_PSR_TP2_TP3_TIME_2500us;
> >+	else if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 1)
> >+		val |= EDP_PSR_TP2_TP3_TIME_500us;
> >+	else if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 0)
> >+		val |= EDP_PSR_TP2_TP3_TIME_100us;
> >+	else
> >+		val |= EDP_PSR_TP2_TP3_TIME_0us;
> >+
> >+	if (intel_dp_source_supports_hbr2(intel_dp) &&
> >+	    drm_dp_tps3_supported(intel_dp->dpcd))
> >+		val |= EDP_PSR_TP1_TP3_SEL;
> >+	else
> >+		val |= EDP_PSR_TP1_TP2_SEL;
> >+
> >+	I915_WRITE(EDP_PSR_CTL, val);
> >+
> >+	if (!dev_priv->psr.psr2_support)
> >+		return;
> >+
> >+	/* FIXME: selective update is probably totally broken because it doesn't
> >+	 * mesh at all with our frontbuffer tracking. And the hw alone isn't
> >+	 * good enough. */
> >+	val = EDP_PSR2_ENABLE | EDP_SU_TRACK_ENABLE;
> >+
> >+	if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 5)
> >+		val |= EDP_PSR2_TP2_TIME_2500;
> >+	else if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 1)
> >+		val |= EDP_PSR2_TP2_TIME_500;
> >+	else if (dev_priv->vbt.psr.tp2_tp3_wakeup_time > 0)
> >+		val |= EDP_PSR2_TP2_TIME_100;
> >+	else
> >+		val |= EDP_PSR2_TP2_TIME_50;
> >-	if (dev_priv->psr.psr2_support)
> >-		I915_WRITE(EDP_PSR2_CTL, EDP_PSR2_ENABLE |
> >-				EDP_SU_TRACK_ENABLE | EDP_PSR2_TP2_TIME_100);
> >+	I915_WRITE(EDP_PSR2_CTL, val);
> >  }
> >  static bool intel_psr_match_conditions(struct intel_dp *intel_dp)
> 

-- 
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch