[Intel-gfx] [PATCH] drm/i915: Install fence register for tiled scanout on i915

[Chris Wilson]

With the work by Jesse Barnes to eliminate allocation of fences during
execbuffer, it becomes possible to write to the scan-out buffer with it
never acquiring a fence (simply by only ever writing to the object using
tiled GPU commands and never writing to it via the GTT). So for pre-i965
chipsets which require fenced access for tiled scan-out buffers, we need
to obtain a fence register.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/i915_drv.h      |    1 +
 drivers/gpu/drm/i915/i915_gem.c      |   10 ++++------
 drivers/gpu/drm/i915/intel_display.c |   20 ++++++++++++++++----
 3 files changed, 21 insertions(+), 10 deletions(-)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 8ef6bce..451b547 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -646,6 +646,7 @@ void i915_gem_object_unpin(struct drm_gem_object *obj);
 int i915_gem_object_unbind(struct drm_gem_object *obj);
 void i915_gem_lastclose(struct drm_device *dev);
 uint32_t i915_get_gem_seqno(struct drm_device *dev);
+int i915_gem_object_get_fence_reg(struct drm_gem_object *obj);
 void i915_gem_retire_requests(struct drm_device *dev);
 void i915_gem_retire_work_handler(struct work_struct *work);
 void i915_gem_clflush_object(struct drm_gem_object *obj);
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index c0ae6bb..c88e3ab 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -46,7 +46,6 @@ static void i915_gem_object_set_to_full_cpu_read_domain(struct drm_gem_object *o
 static int i915_gem_object_wait_rendering(struct drm_gem_object *obj);
 static int i915_gem_object_bind_to_gtt(struct drm_gem_object *obj,
 					   unsigned alignment);
-static int i915_gem_object_get_fence_reg(struct drm_gem_object *obj, bool write);
 static void i915_gem_clear_fence_reg(struct drm_gem_object *obj);
 static int i915_gem_evict_something(struct drm_device *dev);
 static int i915_gem_phys_pwrite(struct drm_device *dev, struct drm_gem_object *obj,
@@ -1158,7 +1157,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 	/* Need a new fence register? */
 	if (obj_priv->fence_reg == I915_FENCE_REG_NONE &&
 	    obj_priv->tiling_mode != I915_TILING_NONE) {
-		ret = i915_gem_object_get_fence_reg(obj, write);
+		ret = i915_gem_object_get_fence_reg(obj);
 		if (ret) {
 			mutex_unlock(&dev->struct_mutex);
 			return VM_FAULT_SIGBUS;
@@ -2169,7 +2168,6 @@ static void i830_write_fence_reg(struct drm_i915_fence_reg *reg)
 /**
  * i915_gem_object_get_fence_reg - set up a fence reg for an object
  * @obj: object to map through a fence reg
- * @write: object is about to be written
  *
  * When mapping objects through the GTT, userspace wants to be able to write
  * to them without having to worry about swizzling if the object is tiled.
@@ -2180,8 +2178,8 @@ static void i830_write_fence_reg(struct drm_i915_fence_reg *reg)
  * It then sets up the reg based on the object's properties: address, pitch
  * and tiling format.
  */
-static int
-i915_gem_object_get_fence_reg(struct drm_gem_object *obj, bool write)
+int
+i915_gem_object_get_fence_reg(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -3550,7 +3548,7 @@ i915_gem_object_pin(struct drm_gem_object *obj, uint32_t alignment)
 	if (!IS_I965G(dev) &&
 	    obj_priv->fence_reg == I915_FENCE_REG_NONE &&
 	    obj_priv->tiling_mode != I915_TILING_NONE) {
-		ret = i915_gem_object_get_fence_reg(obj, true);
+		ret = i915_gem_object_get_fence_reg(obj);
 		if (ret != 0) {
 			if (ret != -EBUSY && ret != -ERESTARTSYS)
 				DRM_ERROR("Failure to install fence: %d\n",
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index b32a51f..83c1718 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -828,19 +828,31 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	}
 
 	mutex_lock(&dev->struct_mutex);
-	ret = i915_gem_object_pin(intel_fb->obj, alignment);
+	ret = i915_gem_object_pin(obj, alignment);
 	if (ret != 0) {
 		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
 
-	ret = i915_gem_object_set_to_gtt_domain(intel_fb->obj, 1);
+	ret = i915_gem_object_set_to_gtt_domain(obj, 1);
 	if (ret != 0) {
-		i915_gem_object_unpin(intel_fb->obj);
+		i915_gem_object_unpin(obj);
 		mutex_unlock(&dev->struct_mutex);
 		return ret;
 	}
 
+	/* Pre-i965 needs to install a fence for tiled scan-out */
+	if (!IS_I965G(dev) &&
+	    obj_priv->fence_reg == I915_FENCE_REG_NONE &&
+	    obj_priv->tiling_mode != I915_TILING_NONE) {
+		ret = i915_gem_object_get_fence_reg(obj);
+		if (ret != 0) {
+			i915_gem_object_unpin(obj);
+			mutex_unlock(&dev->struct_mutex);
+			return ret;
+		}
+	}
+
 	dspcntr = I915_READ(dspcntr_reg);
 	/* Mask out pixel format bits in case we change it */
 	dspcntr &= ~DISPPLANE_PIXFORMAT_MASK;
@@ -860,7 +872,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		break;
 	default:
 		DRM_ERROR("Unknown color depth\n");
-		i915_gem_object_unpin(intel_fb->obj);
+		i915_gem_object_unpin(obj);
 		mutex_unlock(&dev->struct_mutex);
 		return -EINVAL;
 	}
-- 
1.6.3.1