6 files changed, 82 insertions, 22 deletions

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index f990ab4c3efb..fc8cfddbf232 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -811,6 +811,8 @@ int i915_reset(struct drm_device *dev)
         if (!i915.reset)
                 return 0;
 
+        intel_reset_gt_powersave(dev);
+
         mutex_lock(&dev->struct_mutex);
 
         i915_gem_reset(dev);
@@ -880,7 +882,7 @@ int i915_reset(struct drm_device *dev)
                  * of re-init after reset.
                  */
                 if (INTEL_INFO(dev)->gen > 5)
-                        intel_reset_gt_powersave(dev);
+                        intel_enable_gt_powersave(dev);
         } else {
                 mutex_unlock(&dev->struct_mutex);
         }
diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c
index d17ff435f276..d011ec82ef1e 100644
--- a/drivers/gpu/drm/i915/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/i915_gem_context.c
@@ -473,7 +473,12 @@ mi_set_context(struct intel_engine_cs *ring,
                u32 hw_flags)
 {
         u32 flags = hw_flags | MI_MM_SPACE_GTT;
-        int ret;
+        const int num_rings =
+                /* Use an extended w/a on ivb+ if signalling from other rings */
+                i915_semaphore_is_enabled(ring->dev) ?
+                hweight32(INTEL_INFO(ring->dev)->ring_mask) - 1 :
+                0;
+        int len, i, ret;
 
         /* w/a: If Flush TLB Invalidation Mode is enabled, driver must do a TLB
          * invalidation prior to MI_SET_CONTEXT. On GEN6 we don't set the value
@@ -490,15 +495,31 @@ mi_set_context(struct intel_engine_cs *ring,
         if (!IS_HASWELL(ring->dev) && INTEL_INFO(ring->dev)->gen < 8)
                 flags |= (MI_SAVE_EXT_STATE_EN | MI_RESTORE_EXT_STATE_EN);
 
-        ret = intel_ring_begin(ring, 6);
+
+        len = 4;
+        if (INTEL_INFO(ring->dev)->gen >= 7)
+                len += 2 + (num_rings ? 4*num_rings + 2 : 0);
+
+        ret = intel_ring_begin(ring, len);
         if (ret)
                 return ret;
 
         /* WaProgramMiArbOnOffAroundMiSetContext:ivb,vlv,hsw,bdw,chv */
-        if (INTEL_INFO(ring->dev)->gen >= 7)
+        if (INTEL_INFO(ring->dev)->gen >= 7) {
                 intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_DISABLE);
-        else
-                intel_ring_emit(ring, MI_NOOP);
+                if (num_rings) {
+                        struct intel_engine_cs *signaller;
+
+                        intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
+                        for_each_ring(signaller, to_i915(ring->dev), i) {
+                                if (signaller == ring)
+                                        continue;
+
+                                intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
+                                intel_ring_emit(ring, _MASKED_BIT_ENABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
+                        }
+                }
+        }
 
         intel_ring_emit(ring, MI_NOOP);
         intel_ring_emit(ring, MI_SET_CONTEXT);
@@ -510,10 +531,21 @@ mi_set_context(struct intel_engine_cs *ring,
          */
         intel_ring_emit(ring, MI_NOOP);
 
-        if (INTEL_INFO(ring->dev)->gen >= 7)
+        if (INTEL_INFO(ring->dev)->gen >= 7) {
+                if (num_rings) {
+                        struct intel_engine_cs *signaller;
+
+                        intel_ring_emit(ring, MI_LOAD_REGISTER_IMM(num_rings));
+                        for_each_ring(signaller, to_i915(ring->dev), i) {
+                                if (signaller == ring)
+                                        continue;
+
+                                intel_ring_emit(ring, RING_PSMI_CTL(signaller->mmio_base));
+                                intel_ring_emit(ring, _MASKED_BIT_DISABLE(GEN6_PSMI_SLEEP_MSG_DISABLE));
+                        }
+                }
                 intel_ring_emit(ring, MI_ARB_ON_OFF | MI_ARB_ENABLE);
-        else
-                intel_ring_emit(ring, MI_NOOP);
+        }
 
         intel_ring_advance(ring);
 
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index 981834b0f9b6..996c2931c499 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -281,10 +281,14 @@ void gen6_enable_rps_interrupts(struct drm_device *dev)
         struct drm_i915_private *dev_priv = dev->dev_private;
 
         spin_lock_irq(&dev_priv->irq_lock);
+
         WARN_ON(dev_priv->rps.pm_iir);
         WARN_ON(I915_READ(gen6_pm_iir(dev_priv)) & dev_priv->pm_rps_events);
         dev_priv->rps.interrupts_enabled = true;
+        I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) |
+                                dev_priv->pm_rps_events);
         gen6_enable_pm_irq(dev_priv, dev_priv->pm_rps_events);
+
         spin_unlock_irq(&dev_priv->irq_lock);
 }
 
@@ -3307,8 +3311,10 @@ static void gen5_gt_irq_postinstall(struct drm_device *dev)
         GEN5_IRQ_INIT(GT, dev_priv->gt_irq_mask, gt_irqs);
 
         if (INTEL_INFO(dev)->gen >= 6) {
-                pm_irqs |= dev_priv->pm_rps_events;
-
+                /*
+                 * RPS interrupts will get enabled/disabled on demand when RPS
+                 * itself is enabled/disabled.
+                 */
                 if (HAS_VEBOX(dev))
                         pm_irqs |= PM_VEBOX_USER_INTERRUPT;
 
@@ -3520,7 +3526,11 @@ static void gen8_gt_irq_postinstall(struct drm_i915_private *dev_priv)
         dev_priv->pm_irq_mask = 0xffffffff;
         GEN8_IRQ_INIT_NDX(GT, 0, ~gt_interrupts[0], gt_interrupts[0]);
         GEN8_IRQ_INIT_NDX(GT, 1, ~gt_interrupts[1], gt_interrupts[1]);
-        GEN8_IRQ_INIT_NDX(GT, 2, dev_priv->pm_irq_mask, dev_priv->pm_rps_events);
+        /*
+         * RPS interrupts will get enabled/disabled on demand when RPS itself
+         * is enabled/disabled.
+         */
+        GEN8_IRQ_INIT_NDX(GT, 2, dev_priv->pm_irq_mask, 0);
         GEN8_IRQ_INIT_NDX(GT, 3, ~gt_interrupts[3], gt_interrupts[3]);
 }
 
@@ -3609,7 +3619,7 @@ static void vlv_display_irq_uninstall(struct drm_i915_private *dev_priv)
 
         vlv_display_irq_reset(dev_priv);
 
-        dev_priv->irq_mask = 0;
+        dev_priv->irq_mask = ~0;
 }
 
 static void valleyview_irq_uninstall(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index eefdc238f70b..172de3b3433b 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -395,6 +395,7 @@
 #define   PIPE_CONTROL_STORE_DATA_INDEX                 (1<<21)
 #define   PIPE_CONTROL_CS_STALL                         (1<<20)
 #define   PIPE_CONTROL_TLB_INVALIDATE                   (1<<18)
+#define   PIPE_CONTROL_MEDIA_STATE_CLEAR                (1<<16)
 #define   PIPE_CONTROL_QW_WRITE                         (1<<14)
 #define   PIPE_CONTROL_POST_SYNC_OP_MASK                (3<<14)
 #define   PIPE_CONTROL_DEPTH_STALL                      (1<<13)
@@ -1128,6 +1129,7 @@ enum punit_power_well {
 #define GEN6_VERSYNC    (RING_SYNC_1(VEBOX_RING_BASE))
 #define GEN6_VEVSYNC    (RING_SYNC_2(VEBOX_RING_BASE))
 #define GEN6_NOSYNC 0
+#define RING_PSMI_CTL(base)     ((base)+0x50)
 #define RING_MAX_IDLE(base)     ((base)+0x54)
 #define RING_HWS_PGA(base)      ((base)+0x80)
 #define RING_HWS_PGA_GEN6(base) ((base)+0x2080)
@@ -1458,6 +1460,7 @@ enum punit_power_well {
 #define   GEN6_BLITTER_FBC_NOTIFY                       (1<<3)
 
 #define GEN6_RC_SLEEP_PSMI_CONTROL      0x2050
+#define   GEN6_PSMI_SLEEP_MSG_DISABLE   (1 << 0)
 #define   GEN8_RC_SEMA_IDLE_MSG_DISABLE (1 << 12)
 #define   GEN8_FF_DOP_CLOCK_GATE_DISABLE        (1<<10)
 
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 1f4b56e273c8..964b28e3c630 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -6191,6 +6191,20 @@ void intel_cleanup_gt_powersave(struct drm_device *dev)
                 valleyview_cleanup_gt_powersave(dev);
 }
 
+static void gen6_suspend_rps(struct drm_device *dev)
+{
+        struct drm_i915_private *dev_priv = dev->dev_private;
+
+        flush_delayed_work(&dev_priv->rps.delayed_resume_work);
+
+        /*
+         * TODO: disable RPS interrupts on GEN9+ too once RPS support
+         * is added for it.
+         */
+        if (INTEL_INFO(dev)->gen < 9)
+                gen6_disable_rps_interrupts(dev);
+}
+
 /**
  * intel_suspend_gt_powersave - suspend PM work and helper threads
  * @dev: drm device
@@ -6206,14 +6220,7 @@ void intel_suspend_gt_powersave(struct drm_device *dev)
         if (INTEL_INFO(dev)->gen < 6)
                 return;
 
-        flush_delayed_work(&dev_priv->rps.delayed_resume_work);
-
-        /*
-         * TODO: disable RPS interrupts on GEN9+ too once RPS support
-         * is added for it.
-         */
-        if (INTEL_INFO(dev)->gen < 9)
-                gen6_disable_rps_interrupts(dev);
+        gen6_suspend_rps(dev);
 
         /* Force GPU to min freq during suspend */
         gen6_rps_idle(dev_priv);
@@ -6316,8 +6323,11 @@ void intel_reset_gt_powersave(struct drm_device *dev)
 {
         struct drm_i915_private *dev_priv = dev->dev_private;
 
+        if (INTEL_INFO(dev)->gen < 6)
+                return;
+
+        gen6_suspend_rps(dev);
         dev_priv->rps.enabled = false;
-        intel_enable_gt_powersave(dev);
 }
 
 static void ibx_init_clock_gating(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 9f445e9a75d1..c7bc93d28d84 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -362,12 +362,15 @@ gen7_render_ring_flush(struct intel_engine_cs *ring,
                 flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
                 flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
                 flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
+                flags |= PIPE_CONTROL_MEDIA_STATE_CLEAR;
                 /*
                  * TLB invalidate requires a post-sync write.
                  */
                 flags |= PIPE_CONTROL_QW_WRITE;
                 flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
 
+                flags |= PIPE_CONTROL_STALL_AT_SCOREBOARD;
+
                 /* Workaround: we must issue a pipe_control with CS-stall bit
                  * set before a pipe_control command that has the state cache
                  * invalidate bit set. */