drm/i915: turbo & RC6 support for VLV v7

Uses slightly different interfaces than other platforms.

v2: track actual set freq, not requested (Rohit)
    fix debug prints in init code (Jesse)
v3: don't write sleep reg (Jesse)
    re-add RC6 wake limit write (Ben)
    fixup thresholds to match other platforms (Ben)
    clean up mem freq calculation (Ben)
    clean up debug prints (Ben)
v4: move defines from punit patch (Ville)
v5: remove writes to nonexistent regs (Jesse)
    put RP and RC regs together (Jesse)
    fix RC6 enable (Jesse)
v6: use correct fuse reads from NC (Jesse)
    split out min/max funcs for use in sysfs (Jesse)
    add debugfs & sysfs freq controls (Jesse)
v7: update with Ben's hw_max changes (Jesse)

Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net> (v6)
[danvet: Follow checkpatch sugggestion to use min_t to avoid casting
fun.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

drivers/gpu/drm/i915/i915_debugfs.c

@@ -941,7 +941,7 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 			   MEMSTAT_VID_SHIFT);
 		seq_printf(m, "Current P-state: %d\n",
 			   (rgvstat & MEMSTAT_PSTATE_MASK) >> MEMSTAT_PSTATE_SHIFT);
-	} else if (IS_GEN6(dev) || IS_GEN7(dev)) {
+	} else if ((IS_GEN6(dev) || IS_GEN7(dev)) && !IS_VALLEYVIEW(dev)) {
 		u32 gt_perf_status = I915_READ(GEN6_GT_PERF_STATUS);
 		u32 rp_state_limits = I915_READ(GEN6_RP_STATE_LIMITS);
 		u32 rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
@@ -1009,6 +1009,25 @@ static int i915_cur_delayinfo(struct seq_file *m, void *unused)
 
 		seq_printf(m, "Max overclocked frequency: %dMHz\n",
 			   dev_priv->rps.hw_max * GT_FREQUENCY_MULTIPLIER);
+	} else if (IS_VALLEYVIEW(dev)) {
+		u32 freq_sts, val;
+
+		valleyview_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS,
+				      &freq_sts);
+		seq_printf(m, "PUNIT_REG_GPU_FREQ_STS: 0x%08x\n", freq_sts);
+		seq_printf(m, "DDR freq: %d MHz\n", dev_priv->mem_freq);
+
+		valleyview_punit_read(dev_priv, PUNIT_FUSE_BUS1, &val);
+		seq_printf(m, "max GPU freq: %d MHz\n",
+			   vlv_gpu_freq(dev_priv->mem_freq, val));
+
+		valleyview_punit_read(dev_priv, PUNIT_REG_GPU_LFM, &val);
+		seq_printf(m, "min GPU freq: %d MHz\n",
+			   vlv_gpu_freq(dev_priv->mem_freq, val));
+
+		seq_printf(m, "current GPU freq: %d MHz\n",
+			   vlv_gpu_freq(dev_priv->mem_freq,
+					(freq_sts >> 8) & 0xff));
 	} else {
 		seq_printf(m, "no P-state info available\n");
 	}
@@ -1812,7 +1831,11 @@ i915_max_freq_get(void *data, u64 *val)
 	if (ret)
 		return ret;
 
-	*val = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
+	if (IS_VALLEYVIEW(dev))
+		*val = vlv_gpu_freq(dev_priv->mem_freq,
+				    dev_priv->rps.max_delay);
+	else
+		*val = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -1837,9 +1860,16 @@ i915_max_freq_set(void *data, u64 val)
 	/*
 	 * Turbo will still be enabled, but won't go above the set value.
 	 */
-	do_div(val, GT_FREQUENCY_MULTIPLIER);
-	dev_priv->rps.max_delay = val;
-	gen6_set_rps(dev, val);
+	if (IS_VALLEYVIEW(dev)) {
+		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		dev_priv->rps.max_delay = val;
+		gen6_set_rps(dev, val);
+	} else {
+		do_div(val, GT_FREQUENCY_MULTIPLIER);
+		dev_priv->rps.max_delay = val;
+		gen6_set_rps(dev, val);
+	}
+
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -1863,7 +1893,11 @@ i915_min_freq_get(void *data, u64 *val)
 	if (ret)
 		return ret;
 
-	*val = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
+	if (IS_VALLEYVIEW(dev))
+		*val = vlv_gpu_freq(dev_priv->mem_freq,
+				    dev_priv->rps.min_delay);
+	else
+		*val = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;
@@ -1888,9 +1922,15 @@ i915_min_freq_set(void *data, u64 val)
 	/*
 	 * Turbo will still be enabled, but won't go below the set value.
 	 */
-	do_div(val, GT_FREQUENCY_MULTIPLIER);
-	dev_priv->rps.min_delay = val;
-	gen6_set_rps(dev, val);
+	if (IS_VALLEYVIEW(dev)) {
+		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+		dev_priv->rps.min_delay = val;
+		valleyview_set_rps(dev, val);
+	} else {
+		do_div(val, GT_FREQUENCY_MULTIPLIER);
+		dev_priv->rps.min_delay = val;
+		gen6_set_rps(dev, val);
+	}
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return 0;

drivers/gpu/drm/i915/i915_drv.h

@@ -1856,6 +1856,9 @@ extern void intel_disable_fbc(struct drm_device *dev);
 extern bool ironlake_set_drps(struct drm_device *dev, u8 val);
 extern void intel_init_pch_refclk(struct drm_device *dev);
 extern void gen6_set_rps(struct drm_device *dev, u8 val);
+extern void valleyview_set_rps(struct drm_device *dev, u8 val);
+extern int valleyview_rps_max_freq(struct drm_i915_private *dev_priv);
+extern int valleyview_rps_min_freq(struct drm_i915_private *dev_priv);
 extern void intel_detect_pch(struct drm_device *dev);
 extern int intel_trans_dp_port_sel(struct drm_crtc *crtc);
 extern int intel_enable_rc6(const struct drm_device *dev);
@@ -1887,6 +1890,8 @@ int sandybridge_pcode_read(struct drm_i915_private *dev_priv, u8 mbox, u32 *val)
 int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val);
 int valleyview_punit_read(struct drm_i915_private *dev_priv, u8 addr, u32 *val);
 int valleyview_punit_write(struct drm_i915_private *dev_priv, u8 addr, u32 val);
+int valleyview_nc_read(struct drm_i915_private *dev_priv, u8 addr, u32 *val);
+
 int vlv_gpu_freq(int ddr_freq, int val);
 int vlv_freq_opcode(int ddr_freq, int val);
 

drivers/gpu/drm/i915/i915_irq.c

@@ -482,7 +482,10 @@ static void gen6_pm_rps_work(struct work_struct *work)
 	 */
 	if (!(new_delay > dev_priv->rps.max_delay ||
 	      new_delay < dev_priv->rps.min_delay)) {
-		gen6_set_rps(dev_priv->dev, new_delay);
+		if (IS_VALLEYVIEW(dev_priv->dev))
+			valleyview_set_rps(dev_priv->dev, new_delay);
+		else
+			gen6_set_rps(dev_priv->dev, new_delay);
 	}
 
 	mutex_unlock(&dev_priv->rps.hw_lock);

drivers/gpu/drm/i915/i915_reg.h

@@ -4315,6 +4315,7 @@
 #define   GEN6_RC_CTL_RC6_ENABLE		(1<<18)
 #define   GEN6_RC_CTL_RC1e_ENABLE		(1<<20)
 #define   GEN6_RC_CTL_RC7_ENABLE		(1<<22)
+#define   GEN7_RC_CTL_TO_MODE			(1<<28)
 #define   GEN6_RC_CTL_EI_MODE(x)		((x)<<27)
 #define   GEN6_RC_CTL_HW_ENABLE			(1<<31)
 #define GEN6_RP_DOWN_TIMEOUT			0xA010
@@ -4406,12 +4407,32 @@
 #define   IOSF_BAR_SHIFT			1
 #define   IOSF_SB_BUSY				(1<<0)
 #define   IOSF_PORT_PUNIT			0x4
+#define   IOSF_PORT_NC				0x11
 #define VLV_IOSF_DATA				0x182104
 #define VLV_IOSF_ADDR				0x182108
 
 #define PUNIT_OPCODE_REG_READ			6
 #define PUNIT_OPCODE_REG_WRITE			7
 
+#define PUNIT_REG_GPU_LFM			0xd3
+#define PUNIT_REG_GPU_FREQ_REQ			0xd4
+#define PUNIT_REG_GPU_FREQ_STS			0xd8
+#define PUNIT_REG_MEDIA_TURBO_FREQ_REQ		0xdc
+
+#define PUNIT_FUSE_BUS2				0xf6 /* bits 47:40 */
+#define PUNIT_FUSE_BUS1				0xf5 /* bits 55:48 */
+
+#define IOSF_NC_FB_GFX_FREQ_FUSE		0x1c
+#define   FB_GFX_MAX_FREQ_FUSE_SHIFT		3
+#define   FB_GFX_MAX_FREQ_FUSE_MASK		0x000007f8
+#define   FB_GFX_FGUARANTEED_FREQ_FUSE_SHIFT	11
+#define   FB_GFX_FGUARANTEED_FREQ_FUSE_MASK	0x0007f800
+#define IOSF_NC_FB_GFX_FMAX_FUSE_HI		0x34
+#define   FB_FMAX_VMIN_FREQ_HI_MASK		0x00000007
+#define IOSF_NC_FB_GFX_FMAX_FUSE_LO		0x30
+#define   FB_FMAX_VMIN_FREQ_LO_SHIFT		27
+#define   FB_FMAX_VMIN_FREQ_LO_MASK		0xf8000000
+
 #define GEN6_GT_CORE_STATUS		0x138060
 #define   GEN6_CORE_CPD_STATE_MASK	(7<<4)
 #define   GEN6_RCn_MASK			7

drivers/gpu/drm/i915/i915_sysfs.c

@@ -212,7 +212,10 @@ static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 	int ret;
 
 	mutex_lock(&dev_priv->rps.hw_lock);
-	ret = dev_priv->rps.cur_delay * GT_FREQUENCY_MULTIPLIER;
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.cur_delay);
+	else
+		ret = dev_priv->rps.cur_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
@@ -226,7 +229,10 @@ static ssize_t gt_max_freq_mhz_show(struct device *kdev, struct device_attribute
 	int ret;
 
 	mutex_lock(&dev_priv->rps.hw_lock);
-	ret = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.max_delay);
+	else
+		ret = dev_priv->rps.max_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
@@ -246,16 +252,25 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 	if (ret)
 		return ret;
 
-	val /= GT_FREQUENCY_MULTIPLIER;
-
 	mutex_lock(&dev_priv->rps.hw_lock);
 
-	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-	hw_max = dev_priv->rps.hw_max;
-	non_oc_max = (rp_state_cap & 0xff);
-	hw_min = ((rp_state_cap & 0xff0000) >> 16);
+	if (IS_VALLEYVIEW(dev_priv->dev)) {
+		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+
+		hw_max = valleyview_rps_max_freq(dev_priv);
+		hw_min = valleyview_rps_min_freq(dev_priv);
+		non_oc_max = hw_max;
+	} else {
+		val /= GT_FREQUENCY_MULTIPLIER;
 
-	if (val < hw_min || val > hw_max || val < dev_priv->rps.min_delay) {
+		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+		hw_max = dev_priv->rps.hw_max;
+		non_oc_max = (rp_state_cap & 0xff);
+		hw_min = ((rp_state_cap & 0xff0000) >> 16);
+	}
+
+	if (val < hw_min || val > hw_max ||
+	    val < dev_priv->rps.min_delay) {
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		return -EINVAL;
 	}
@@ -264,8 +279,12 @@ static ssize_t gt_max_freq_mhz_store(struct device *kdev,
 		DRM_DEBUG("User requested overclocking to %d\n",
 			  val * GT_FREQUENCY_MULTIPLIER);
 
-	if (dev_priv->rps.cur_delay > val)
-		gen6_set_rps(dev_priv->dev, val);
+	if (dev_priv->rps.cur_delay > val) {
+		if (IS_VALLEYVIEW(dev_priv->dev))
+			valleyview_set_rps(dev_priv->dev, val);
+		else
+			gen6_set_rps(dev_priv->dev, val);
+	}
 
 	dev_priv->rps.max_delay = val;
 
@@ -282,7 +301,10 @@ static ssize_t gt_min_freq_mhz_show(struct device *kdev, struct device_attribute
 	int ret;
 
 	mutex_lock(&dev_priv->rps.hw_lock);
-	ret = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
+	if (IS_VALLEYVIEW(dev_priv->dev))
+		ret = vlv_gpu_freq(dev_priv->mem_freq, dev_priv->rps.min_delay);
+	else
+		ret = dev_priv->rps.min_delay * GT_FREQUENCY_MULTIPLIER;
 	mutex_unlock(&dev_priv->rps.hw_lock);
 
 	return snprintf(buf, PAGE_SIZE, "%d\n", ret);
@@ -302,21 +324,32 @@ static ssize_t gt_min_freq_mhz_store(struct device *kdev,
 	if (ret)
 		return ret;
 
-	val /= GT_FREQUENCY_MULTIPLIER;
-
 	mutex_lock(&dev_priv->rps.hw_lock);
 
-	rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
-	hw_max = dev_priv->rps.hw_max;
-	hw_min = ((rp_state_cap & 0xff0000) >> 16);
+	if (IS_VALLEYVIEW(dev)) {
+		val = vlv_freq_opcode(dev_priv->mem_freq, val);
+
+		hw_max = valleyview_rps_max_freq(dev_priv);
+		hw_min = valleyview_rps_min_freq(dev_priv);
+	} else {
+		val /= GT_FREQUENCY_MULTIPLIER;
+
+		rp_state_cap = I915_READ(GEN6_RP_STATE_CAP);
+		hw_max = dev_priv->rps.hw_max;
+		hw_min = ((rp_state_cap & 0xff0000) >> 16);
+	}
 
 	if (val < hw_min || val > hw_max || val > dev_priv->rps.max_delay) {
 		mutex_unlock(&dev_priv->rps.hw_lock);
 		return -EINVAL;
 	}
 
-	if (dev_priv->rps.cur_delay < val)
-		gen6_set_rps(dev_priv->dev, val);
+	if (dev_priv->rps.cur_delay < val) {
+		if (IS_VALLEYVIEW(dev))
+			valleyview_set_rps(dev, val);
+		else
+			gen6_set_rps(dev_priv->dev, val);
+	}
 
 	dev_priv->rps.min_delay = val;
 

drivers/gpu/drm/i915/intel_pm.c

@@ -2481,6 +2481,52 @@ void gen6_set_rps(struct drm_device *dev, u8 val)
 	trace_intel_gpu_freq_change(val * 50);
 }
 
+void valleyview_set_rps(struct drm_device *dev, u8 val)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	unsigned long timeout = jiffies + msecs_to_jiffies(10);
+	u32 limits = gen6_rps_limits(dev_priv, &val);
+	u32 pval;
+
+	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
+	WARN_ON(val > dev_priv->rps.max_delay);
+	WARN_ON(val < dev_priv->rps.min_delay);
+
+	DRM_DEBUG_DRIVER("gpu freq request from %d to %d\n",
+			 vlv_gpu_freq(dev_priv->mem_freq,
+				      dev_priv->rps.cur_delay),
+			 vlv_gpu_freq(dev_priv->mem_freq, val));
+
+	if (val == dev_priv->rps.cur_delay)
+		return;
+
+	valleyview_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val);
+
+	do {
+		valleyview_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS, &pval);
+		if (time_after(jiffies, timeout)) {
+			DRM_DEBUG_DRIVER("timed out waiting for Punit\n");
+			break;
+		}
+		udelay(10);
+	} while (pval & 1);
+
+	valleyview_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS, &pval);
+	if ((pval >> 8) != val)
+		DRM_DEBUG_DRIVER("punit overrode freq: %d requested, but got %d\n",
+			  val, pval >> 8);
+
+	/* Make sure we continue to get interrupts
+	 * until we hit the minimum or maximum frequencies.
+	 */
+	I915_WRITE(GEN6_RP_INTERRUPT_LIMITS, limits);
+
+	dev_priv->rps.cur_delay = pval >> 8;
+
+	trace_intel_gpu_freq_change(vlv_gpu_freq(dev_priv->mem_freq, val));
+}
+
+
 static void gen6_disable_rps(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -2742,6 +2788,127 @@ static void gen6_update_ring_freq(struct drm_device *dev)
 	}
 }
 
+int valleyview_rps_max_freq(struct drm_i915_private *dev_priv)
+{
+	u32 val, rp0;
+
+	valleyview_nc_read(dev_priv, IOSF_NC_FB_GFX_FREQ_FUSE, &val);
+
+	rp0 = (val & FB_GFX_MAX_FREQ_FUSE_MASK) >> FB_GFX_MAX_FREQ_FUSE_SHIFT;
+	/* Clamp to max */
+	rp0 = min_t(u32, rp0, 0xea);
+
+	return rp0;
+}
+
+static int valleyview_rps_rpe_freq(struct drm_i915_private *dev_priv)
+{
+	u32 val, rpe;
+
+	valleyview_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_LO, &val);
+	rpe = (val & FB_FMAX_VMIN_FREQ_LO_MASK) >> FB_FMAX_VMIN_FREQ_LO_SHIFT;
+	valleyview_nc_read(dev_priv, IOSF_NC_FB_GFX_FMAX_FUSE_HI, &val);
+	rpe |= (val & FB_FMAX_VMIN_FREQ_HI_MASK) << 5;
+
+	return rpe;
+}
+
+int valleyview_rps_min_freq(struct drm_i915_private *dev_priv)
+{
+	u32 val;
+
+	valleyview_punit_read(dev_priv, PUNIT_REG_GPU_LFM, &val);
+
+	return val & 0xff;
+}
+
+static void valleyview_enable_rps(struct drm_device *dev)
+{
+	struct drm_i915_private *dev_priv = dev->dev_private;
+	struct intel_ring_buffer *ring;
+	u32 gtfifodbg, val, rpe;
+	int i;
+
+	WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock));
+
+	if ((gtfifodbg = I915_READ(GTFIFODBG))) {
+		DRM_ERROR("GT fifo had a previous error %x\n", gtfifodbg);
+		I915_WRITE(GTFIFODBG, gtfifodbg);
+	}
+
+	gen6_gt_force_wake_get(dev_priv);
+
+	I915_WRITE(GEN6_RP_UP_THRESHOLD, 59400);
+	I915_WRITE(GEN6_RP_DOWN_THRESHOLD, 245000);
+	I915_WRITE(GEN6_RP_UP_EI, 66000);
+	I915_WRITE(GEN6_RP_DOWN_EI, 350000);
+
+	I915_WRITE(GEN6_RP_IDLE_HYSTERSIS, 10);
+
+	I915_WRITE(GEN6_RP_CONTROL,
+		   GEN6_RP_MEDIA_TURBO |
+		   GEN6_RP_MEDIA_HW_NORMAL_MODE |
+		   GEN6_RP_MEDIA_IS_GFX |
+		   GEN6_RP_ENABLE |
+		   GEN6_RP_UP_BUSY_AVG |
+		   GEN6_RP_DOWN_IDLE_CONT);
+
+	I915_WRITE(GEN6_RC6_WAKE_RATE_LIMIT, 0x00280000);
+	I915_WRITE(GEN6_RC_EVALUATION_INTERVAL, 125000);
+	I915_WRITE(GEN6_RC_IDLE_HYSTERSIS, 25);
+
+	for_each_ring(ring, dev_priv, i)
+		I915_WRITE(RING_MAX_IDLE(ring->mmio_base), 10);
+
+	I915_WRITE(GEN6_RC6_THRESHOLD, 0xc350);
+
+	/* allows RC6 residency counter to work */
+	I915_WRITE(0x138104, _MASKED_BIT_ENABLE(0x3));
+	I915_WRITE(GEN6_RC_CONTROL,
+		   GEN7_RC_CTL_TO_MODE);
+
+	valleyview_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS, &val);
+	dev_priv->mem_freq = 800 + (266 * (val >> 6) & 3);
+	DRM_DEBUG_DRIVER("DDR speed: %d MHz", dev_priv->mem_freq);
+
+	DRM_DEBUG_DRIVER("GPLL enabled? %s\n", val & 0x10 ? "yes" : "no");
+	DRM_DEBUG_DRIVER("GPU status: 0x%08x\n", val);
+
+	DRM_DEBUG_DRIVER("current GPU freq: %d\n",
+			 vlv_gpu_freq(dev_priv->mem_freq, (val >> 8) & 0xff));
+	dev_priv->rps.cur_delay = (val >> 8) & 0xff;
+
+	dev_priv->rps.max_delay = valleyview_rps_max_freq(dev_priv);
+	dev_priv->rps.hw_max = dev_priv->rps.max_delay;
+	DRM_DEBUG_DRIVER("max GPU freq: %d\n", vlv_gpu_freq(dev_priv->mem_freq,
+						     dev_priv->rps.max_delay));
+
+	rpe = valleyview_rps_rpe_freq(dev_priv);
+	DRM_DEBUG_DRIVER("RPe GPU freq: %d\n",
+			 vlv_gpu_freq(dev_priv->mem_freq, rpe));
+
+	val = valleyview_rps_min_freq(dev_priv);
+	DRM_DEBUG_DRIVER("min GPU freq: %d\n", vlv_gpu_freq(dev_priv->mem_freq,
+							    val));
+	dev_priv->rps.min_delay = val;
+
+	DRM_DEBUG_DRIVER("setting GPU freq to %d\n",
+			 vlv_gpu_freq(dev_priv->mem_freq, rpe));
+
+	valleyview_set_rps(dev_priv->dev, rpe);
+
+	/* requires MSI enabled */
+	I915_WRITE(GEN6_PMIER, GEN6_PM_DEFERRED_EVENTS);
+	spin_lock_irq(&dev_priv->rps.lock);
+	WARN_ON(dev_priv->rps.pm_iir != 0);
+	I915_WRITE(GEN6_PMIMR, 0);
+	spin_unlock_irq(&dev_priv->rps.lock);
+	/* enable all PM interrupts */
+	I915_WRITE(GEN6_PMINTRMSK, 0);
+
+	gen6_gt_force_wake_put(dev_priv);
+}
+
 void ironlake_teardown_rc6(struct drm_device *dev)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -3468,7 +3635,7 @@ void intel_disable_gt_powersave(struct drm_device *dev)
 	if (IS_IRONLAKE_M(dev)) {
 		ironlake_disable_drps(dev);
 		ironlake_disable_rc6(dev);
-	} else if (INTEL_INFO(dev)->gen >= 6 && !IS_VALLEYVIEW(dev)) {
+	} else if (INTEL_INFO(dev)->gen >= 6) {
 		cancel_delayed_work_sync(&dev_priv->rps.delayed_resume_work);
 		mutex_lock(&dev_priv->rps.hw_lock);
 		gen6_disable_rps(dev);
@@ -3484,8 +3651,13 @@ static void intel_gen6_powersave_work(struct work_struct *work)
 	struct drm_device *dev = dev_priv->dev;
 
 	mutex_lock(&dev_priv->rps.hw_lock);
-	gen6_enable_rps(dev);
-	gen6_update_ring_freq(dev);
+
+	if (IS_VALLEYVIEW(dev)) {
+		valleyview_enable_rps(dev);
+	} else {
+		gen6_enable_rps(dev);
+		gen6_update_ring_freq(dev);
+	}
 	mutex_unlock(&dev_priv->rps.hw_lock);
 }
 
@@ -3497,7 +3669,7 @@ void intel_enable_gt_powersave(struct drm_device *dev)
 		ironlake_enable_drps(dev);
 		ironlake_enable_rc6(dev);
 		intel_init_emon(dev);
-	} else if ((IS_GEN6(dev) || IS_GEN7(dev)) && !IS_VALLEYVIEW(dev)) {
+	} else if (IS_GEN6(dev) || IS_GEN7(dev)) {
 		/*
 		 * PCU communication is slow and this doesn't need to be
 		 * done at any specific time, so do this out of our fast path
@@ -4568,14 +4740,13 @@ int sandybridge_pcode_write(struct drm_i915_private *dev_priv, u8 mbox, u32 val)
 	return 0;
 }
 
-static int vlv_punit_rw(struct drm_i915_private *dev_priv, u8 opcode,
+static int vlv_punit_rw(struct drm_i915_private *dev_priv, u32 port, u8 opcode,
 			u8 addr, u32 *val)
 {
-	u32 cmd, devfn, port, be, bar;
+	u32 cmd, devfn, be, bar;
 
 	bar = 0;
 	be = 0xf;
-	port = IOSF_PORT_PUNIT;
 	devfn = PCI_DEVFN(2, 0);
 
 	cmd = (devfn << IOSF_DEVFN_SHIFT) | (opcode << IOSF_OPCODE_SHIFT) |
@@ -4597,7 +4768,7 @@ static int vlv_punit_rw(struct drm_i915_private *dev_priv, u8 opcode,
 	I915_WRITE(VLV_IOSF_DOORBELL_REQ, cmd);
 
 	if (wait_for((I915_READ(VLV_IOSF_DOORBELL_REQ) & IOSF_SB_BUSY) == 0,
-		     500)) {
+		     5)) {
 		DRM_ERROR("timeout waiting for pcode %s (%d) to finish\n",
 			  opcode == PUNIT_OPCODE_REG_READ ? "read" : "write",
 			  addr);
@@ -4613,12 +4784,20 @@ static int vlv_punit_rw(struct drm_i915_private *dev_priv, u8 opcode,
 
 int valleyview_punit_read(struct drm_i915_private *dev_priv, u8 addr, u32 *val)
 {
-	return vlv_punit_rw(dev_priv, PUNIT_OPCODE_REG_READ, addr, val);
+	return vlv_punit_rw(dev_priv, IOSF_PORT_PUNIT, PUNIT_OPCODE_REG_READ,
+			    addr, val);
 }
 
 int valleyview_punit_write(struct drm_i915_private *dev_priv, u8 addr, u32 val)
 {
-	return vlv_punit_rw(dev_priv, PUNIT_OPCODE_REG_WRITE, addr, &val);
+	return vlv_punit_rw(dev_priv, IOSF_PORT_PUNIT, PUNIT_OPCODE_REG_WRITE,
+			    addr, &val);
+}
+
+int valleyview_nc_read(struct drm_i915_private *dev_priv, u8 addr, u32 *val)
+{
+	return vlv_punit_rw(dev_priv, IOSF_PORT_NC, PUNIT_OPCODE_REG_READ,
+			    addr, val);
 }
 
 int vlv_gpu_freq(int ddr_freq, int val)