[Intel-gfx] [PATCH 3/6] drm/i915: Extend vlv/chv residency resolution

Mika Kuoppala mika.kuoppala at linux.intel.com
Wed Mar 15 15:43:01 UTC 2017


Vlv and chv residency counters are 40 bits in width.
With a control bit, we can choose between upper or lower
32 bit window into this counter.

Lets toggle this bit on and off on and read both parts.
As a result we can push the wrap from 13 seconds to 54
minutes.

v2: commit msg, loop readability, goto elimination (Chris)

Reported-by: Len Brown <len.brown at intel.com>
Cc: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala at linux.intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala at intel.com>
---
 drivers/gpu/drm/i915/intel_pm.c | 60 +++++++++++++++++++++++++++++++++++------
 1 file changed, 52 insertions(+), 8 deletions(-)

diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index da742a9..19fd11b 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -8350,12 +8350,51 @@ void intel_pm_setup(struct drm_i915_private *dev_priv)
 	atomic_set(&dev_priv->pm.wakeref_count, 0);
 }
 
+static u64 vlv_residency_raw(struct drm_i915_private *dev_priv,
+			     const i915_reg_t reg)
+{
+	u32 lower, upper, tmp, saved_ctl;
+
+	/* The register accessed do not need forcewake. We borrow
+	 * uncore lock to prevent concurrent access to range reg.
+	 */
+	spin_lock_irq(&dev_priv->uncore.lock);
+	saved_ctl = I915_READ_FW(VLV_COUNTER_CONTROL);
+
+	if (!(saved_ctl & VLV_COUNT_RANGE_HIGH))
+		I915_WRITE_FW(VLV_COUNTER_CONTROL,
+			      _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
+
+	/* vlv and chv residency counters are 40 bits in width.
+	 * With a control bit, we can choose between upper or lower
+	 * 32bit window into this counter.
+	 */
+	upper = I915_READ_FW(reg);
+	do {
+		tmp = upper;
+
+		I915_WRITE_FW(VLV_COUNTER_CONTROL,
+			      _MASKED_BIT_DISABLE(VLV_COUNT_RANGE_HIGH));
+		lower = I915_READ_FW(reg);
+
+		I915_WRITE_FW(VLV_COUNTER_CONTROL,
+			      _MASKED_BIT_ENABLE(VLV_COUNT_RANGE_HIGH));
+		upper = I915_READ_FW(reg);
+	} while (upper != tmp);
+
+	if (!(saved_ctl & VLV_COUNT_RANGE_HIGH))
+		I915_WRITE_FW(VLV_COUNTER_CONTROL,
+			      _MASKED_BIT_DISABLE(VLV_COUNT_RANGE_HIGH));
+
+	spin_unlock_irq(&dev_priv->uncore.lock);
+
+	return lower | (u64)upper << 8;
+}
+
 u64 intel_rc6_residency_us(struct drm_i915_private *dev_priv,
 			   const i915_reg_t reg)
 {
-	u64 raw_time; /* 32b value may overflow during fixed point math */
-	u64 units = 128000ULL, div = 100000ULL;
-	u64 ret;
+	u64 time_hw, units, div, residency_us;
 
 	if (!intel_enable_rc6())
 		return 0;
@@ -8367,16 +8406,21 @@ u64 intel_rc6_residency_us(struct drm_i915_private *dev_priv,
 		units = 1000;
 		div = dev_priv->czclk_freq;
 
-		if (I915_READ(VLV_COUNTER_CONTROL) & VLV_COUNT_RANGE_HIGH)
-			units <<= 8;
+		time_hw = vlv_residency_raw(dev_priv, reg);
 	} else if (IS_GEN9_LP(dev_priv)) {
 		units = 1000;
 		div = 1200;		/* 833.33ns */
+
+		time_hw = I915_READ(reg);
+	} else {
+		units = 128000; /* 1.28us */
+		div = 100000;
+
+		time_hw = I915_READ(reg);
 	}
 
-	raw_time = I915_READ(reg) * units;
-	ret = DIV_ROUND_UP_ULL(raw_time, div);
+	residency_us = DIV_ROUND_UP_ULL(time_hw * units, div);
 
 	intel_runtime_pm_put(dev_priv);
-	return ret;
+	return residency_us;
 }
-- 
2.7.4



More information about the Intel-gfx mailing list