diff mbox

[4/9] drm/i915: pass which operation triggered the frontbuffer tracking

Message ID 1423500395-1787-5-git-send-email-przanoni@gmail.com (mailing list archive)
State New, archived
Headers show

Commit Message

Paulo Zanoni Feb. 9, 2015, 4:46 p.m. UTC 
From: Paulo Zanoni <paulo.r.zanoni@intel.com>

We want to port FBC to the frontbuffer tracking infrastructure, but
for that we need to know what caused the object invalidation/flush so
we can react accordingly: CPU mmaps need manual, GTT mmaps and
flips don't need handling and ring rendering needs nukes.

v2: - s/ORIGIN_RENDER/ORIGIN_CS/ (Daniel, Rodrigo)
    - Fix copy/pasted wrong documentation
    - Rebase
v3: - Rebase

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> (v2)
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h            |  7 +++++++
 drivers/gpu/drm/i915/i915_gem.c            | 10 +++++-----
 drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
 drivers/gpu/drm/i915/intel_drv.h           | 11 +++++++----
 drivers/gpu/drm/i915/intel_frontbuffer.c   | 15 ++++++++++-----
 5 files changed, 30 insertions(+), 15 deletions(-)

Comments

Daniel Vetter Feb. 9, 2015, 6:29 p.m. UTC | #1 
On Mon, Feb 09, 2015 at 02:46:30PM -0200, Paulo Zanoni wrote:
> From: Paulo Zanoni <paulo.r.zanoni@intel.com>
> 
> We want to port FBC to the frontbuffer tracking infrastructure, but
> for that we need to know what caused the object invalidation/flush so
> we can react accordingly: CPU mmaps need manual, GTT mmaps and
> flips don't need handling and ring rendering needs nukes.
> 
> v2: - s/ORIGIN_RENDER/ORIGIN_CS/ (Daniel, Rodrigo)
>     - Fix copy/pasted wrong documentation
>     - Rebase
> v3: - Rebase
> 
> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com> (v2)
> Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>

Sorry for being really late with my review here. I fully agree that origin
for invalidate makes a lot of sense, but for flush it kinda doesn't.
There's no origin for a flush, it just means that any previous rendering
(from _any_ origin) has now completed.

I've looked ahead a bit and I'll comment on the patch that uses the flush
origin what should imo be used instead.

Merged the first 3 patches from this series meanwhile, thanks.
-Daniel

> ---
>  drivers/gpu/drm/i915/i915_drv.h            |  7 +++++++
>  drivers/gpu/drm/i915/i915_gem.c            | 10 +++++-----
>  drivers/gpu/drm/i915/i915_gem_execbuffer.c |  2 +-
>  drivers/gpu/drm/i915/intel_drv.h           | 11 +++++++----
>  drivers/gpu/drm/i915/intel_frontbuffer.c   | 15 ++++++++++-----
>  5 files changed, 30 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
> index c0b8644..d578211 100644
> --- a/drivers/gpu/drm/i915/i915_drv.h
> +++ b/drivers/gpu/drm/i915/i915_drv.h
> @@ -771,6 +771,13 @@ struct intel_context {
>  	struct list_head link;
>  };
>  
> +enum fb_op_origin {
> +	ORIGIN_GTT,
> +	ORIGIN_CPU,
> +	ORIGIN_CS,
> +	ORIGIN_FLIP,
> +};
> +
>  struct i915_fbc {
>  	unsigned long uncompressed_size;
>  	unsigned threshold;
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index c26d36c..3d198f8 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -2310,7 +2310,7 @@ i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
>  			list_move_tail(&vma->mm_list, &vma->vm->inactive_list);
>  	}
>  
> -	intel_fb_obj_flush(obj, true);
> +	intel_fb_obj_flush(obj, true, ORIGIN_CS);
>  
>  	list_del_init(&obj->ring_list);
>  
> @@ -3677,7 +3677,7 @@ i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj)
>  	old_write_domain = obj->base.write_domain;
>  	obj->base.write_domain = 0;
>  
> -	intel_fb_obj_flush(obj, false);
> +	intel_fb_obj_flush(obj, false, ORIGIN_GTT);
>  
>  	trace_i915_gem_object_change_domain(obj,
>  					    obj->base.read_domains,
> @@ -3699,7 +3699,7 @@ i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj)
>  	old_write_domain = obj->base.write_domain;
>  	obj->base.write_domain = 0;
>  
> -	intel_fb_obj_flush(obj, false);
> +	intel_fb_obj_flush(obj, false, ORIGIN_CPU);
>  
>  	trace_i915_gem_object_change_domain(obj,
>  					    obj->base.read_domains,
> @@ -3764,7 +3764,7 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
>  	}
>  
>  	if (write)
> -		intel_fb_obj_invalidate(obj, NULL);
> +		intel_fb_obj_invalidate(obj, NULL, ORIGIN_GTT);
>  
>  	trace_i915_gem_object_change_domain(obj,
>  					    old_read_domains,
> @@ -4079,7 +4079,7 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
>  	}
>  
>  	if (write)
> -		intel_fb_obj_invalidate(obj, NULL);
> +		intel_fb_obj_invalidate(obj, NULL, ORIGIN_CPU);
>  
>  	trace_i915_gem_object_change_domain(obj,
>  					    old_read_domains,
> diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> index b773368..f4342f0 100644
> --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
> @@ -971,7 +971,7 @@ i915_gem_execbuffer_move_to_active(struct list_head *vmas,
>  			obj->dirty = 1;
>  			i915_gem_request_assign(&obj->last_write_req, req);
>  
> -			intel_fb_obj_invalidate(obj, ring);
> +			intel_fb_obj_invalidate(obj, ring, ORIGIN_CS);
>  
>  			/* update for the implicit flush after a batch */
>  			obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
> diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
> index 76b3c20..a73d091 100644
> --- a/drivers/gpu/drm/i915/intel_drv.h
> +++ b/drivers/gpu/drm/i915/intel_drv.h
> @@ -853,13 +853,15 @@ void intel_ddi_set_vc_payload_alloc(struct drm_crtc *crtc, bool state);
>  
>  /* intel_frontbuffer.c */
>  void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
> -			     struct intel_engine_cs *ring);
> +			     struct intel_engine_cs *ring,
> +			     enum fb_op_origin origin);
>  void intel_frontbuffer_flip_prepare(struct drm_device *dev,
>  				    unsigned frontbuffer_bits);
>  void intel_frontbuffer_flip_complete(struct drm_device *dev,
>  				     unsigned frontbuffer_bits);
>  void intel_frontbuffer_flush(struct drm_device *dev,
> -			     unsigned frontbuffer_bits);
> +			     unsigned frontbuffer_bits,
> +			     enum fb_op_origin origin);
>  /**
>   * intel_frontbuffer_flip - synchronous frontbuffer flip
>   * @dev: DRM device
> @@ -875,12 +877,13 @@ static inline
>  void intel_frontbuffer_flip(struct drm_device *dev,
>  			    unsigned frontbuffer_bits)
>  {
> -	intel_frontbuffer_flush(dev, frontbuffer_bits);
> +	intel_frontbuffer_flush(dev, frontbuffer_bits, ORIGIN_FLIP);
>  }
>  
>  int intel_fb_align_height(struct drm_device *dev, int height,
>  			  unsigned int tiling);
> -void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire);
> +void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire,
> +			enum fb_op_origin origin);
>  
>  
>  /* intel_audio.c */
> diff --git a/drivers/gpu/drm/i915/intel_frontbuffer.c b/drivers/gpu/drm/i915/intel_frontbuffer.c
> index 73cb6e0..0a773981 100644
> --- a/drivers/gpu/drm/i915/intel_frontbuffer.c
> +++ b/drivers/gpu/drm/i915/intel_frontbuffer.c
> @@ -127,6 +127,7 @@ static void intel_mark_fb_busy(struct drm_device *dev,
>   * intel_fb_obj_invalidate - invalidate frontbuffer object
>   * @obj: GEM object to invalidate
>   * @ring: set for asynchronous rendering
> + * @origin: which operation caused the invalidation
>   *
>   * This function gets called every time rendering on the given object starts and
>   * frontbuffer caching (fbc, low refresh rate for DRRS, panel self refresh) must
> @@ -135,7 +136,8 @@ static void intel_mark_fb_busy(struct drm_device *dev,
>   * scheduled.
>   */
>  void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
> -			     struct intel_engine_cs *ring)
> +			     struct intel_engine_cs *ring,
> +			     enum fb_op_origin origin)
>  {
>  	struct drm_device *dev = obj->base.dev;
>  	struct drm_i915_private *dev_priv = dev->dev_private;
> @@ -164,6 +166,7 @@ void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
>   * intel_frontbuffer_flush - flush frontbuffer
>   * @dev: DRM device
>   * @frontbuffer_bits: frontbuffer plane tracking bits
> + * @origin: which operation caused the flush
>   *
>   * This function gets called every time rendering on the given planes has
>   * completed and frontbuffer caching can be started again. Flushes will get
> @@ -172,7 +175,8 @@ void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
>   * Can be called without any locks held.
>   */
>  void intel_frontbuffer_flush(struct drm_device *dev,
> -			     unsigned frontbuffer_bits)
> +			     unsigned frontbuffer_bits,
> +			     enum fb_op_origin origin)
>  {
>  	struct drm_i915_private *dev_priv = dev->dev_private;
>  
> @@ -201,13 +205,14 @@ void intel_frontbuffer_flush(struct drm_device *dev,
>   * intel_fb_obj_flush - flush frontbuffer object
>   * @obj: GEM object to flush
>   * @retire: set when retiring asynchronous rendering
> + * @origin: which operation caused the flush
>   *
>   * This function gets called every time rendering on the given object has
>   * completed and frontbuffer caching can be started again. If @retire is true
>   * then any delayed flushes will be unblocked.
>   */
>  void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
> -			bool retire)
> +			bool retire, enum fb_op_origin origin)
>  {
>  	struct drm_device *dev = obj->base.dev;
>  	struct drm_i915_private *dev_priv = dev->dev_private;
> @@ -229,7 +234,7 @@ void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
>  		mutex_unlock(&dev_priv->fb_tracking.lock);
>  	}
>  
> -	intel_frontbuffer_flush(dev, frontbuffer_bits);
> +	intel_frontbuffer_flush(dev, frontbuffer_bits, origin);
>  }
>  
>  /**
> @@ -277,5 +282,5 @@ void intel_frontbuffer_flip_complete(struct drm_device *dev,
>  	dev_priv->fb_tracking.flip_bits &= ~frontbuffer_bits;
>  	mutex_unlock(&dev_priv->fb_tracking.lock);
>  
> -	intel_frontbuffer_flush(dev, frontbuffer_bits);
> +	intel_frontbuffer_flush(dev, frontbuffer_bits, ORIGIN_FLIP);
>  }
> -- 
> 2.1.4
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx
diff mbox

Patch

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index c0b8644..d578211 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -771,6 +771,13 @@  struct intel_context {
 	struct list_head link;
 };
 
+enum fb_op_origin {
+	ORIGIN_GTT,
+	ORIGIN_CPU,
+	ORIGIN_CS,
+	ORIGIN_FLIP,
+};
+
 struct i915_fbc {
 	unsigned long uncompressed_size;
 	unsigned threshold;
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index c26d36c..3d198f8 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2310,7 +2310,7 @@  i915_gem_object_move_to_inactive(struct drm_i915_gem_object *obj)
 			list_move_tail(&vma->mm_list, &vma->vm->inactive_list);
 	}
 
-	intel_fb_obj_flush(obj, true);
+	intel_fb_obj_flush(obj, true, ORIGIN_CS);
 
 	list_del_init(&obj->ring_list);
 
@@ -3677,7 +3677,7 @@  i915_gem_object_flush_gtt_write_domain(struct drm_i915_gem_object *obj)
 	old_write_domain = obj->base.write_domain;
 	obj->base.write_domain = 0;
 
-	intel_fb_obj_flush(obj, false);
+	intel_fb_obj_flush(obj, false, ORIGIN_GTT);
 
 	trace_i915_gem_object_change_domain(obj,
 					    obj->base.read_domains,
@@ -3699,7 +3699,7 @@  i915_gem_object_flush_cpu_write_domain(struct drm_i915_gem_object *obj)
 	old_write_domain = obj->base.write_domain;
 	obj->base.write_domain = 0;
 
-	intel_fb_obj_flush(obj, false);
+	intel_fb_obj_flush(obj, false, ORIGIN_CPU);
 
 	trace_i915_gem_object_change_domain(obj,
 					    obj->base.read_domains,
@@ -3764,7 +3764,7 @@  i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	}
 
 	if (write)
-		intel_fb_obj_invalidate(obj, NULL);
+		intel_fb_obj_invalidate(obj, NULL, ORIGIN_GTT);
 
 	trace_i915_gem_object_change_domain(obj,
 					    old_read_domains,
@@ -4079,7 +4079,7 @@  i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write)
 	}
 
 	if (write)
-		intel_fb_obj_invalidate(obj, NULL);
+		intel_fb_obj_invalidate(obj, NULL, ORIGIN_CPU);
 
 	trace_i915_gem_object_change_domain(obj,
 					    old_read_domains,
diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
index b773368..f4342f0 100644
--- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c
@@ -971,7 +971,7 @@  i915_gem_execbuffer_move_to_active(struct list_head *vmas,
 			obj->dirty = 1;
 			i915_gem_request_assign(&obj->last_write_req, req);
 
-			intel_fb_obj_invalidate(obj, ring);
+			intel_fb_obj_invalidate(obj, ring, ORIGIN_CS);
 
 			/* update for the implicit flush after a batch */
 			obj->base.write_domain &= ~I915_GEM_GPU_DOMAINS;
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index 76b3c20..a73d091 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -853,13 +853,15 @@  void intel_ddi_set_vc_payload_alloc(struct drm_crtc *crtc, bool state);
 
 /* intel_frontbuffer.c */
 void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
-			     struct intel_engine_cs *ring);
+			     struct intel_engine_cs *ring,
+			     enum fb_op_origin origin);
 void intel_frontbuffer_flip_prepare(struct drm_device *dev,
 				    unsigned frontbuffer_bits);
 void intel_frontbuffer_flip_complete(struct drm_device *dev,
 				     unsigned frontbuffer_bits);
 void intel_frontbuffer_flush(struct drm_device *dev,
-			     unsigned frontbuffer_bits);
+			     unsigned frontbuffer_bits,
+			     enum fb_op_origin origin);
 /**
  * intel_frontbuffer_flip - synchronous frontbuffer flip
  * @dev: DRM device
@@ -875,12 +877,13 @@  static inline
 void intel_frontbuffer_flip(struct drm_device *dev,
 			    unsigned frontbuffer_bits)
 {
-	intel_frontbuffer_flush(dev, frontbuffer_bits);
+	intel_frontbuffer_flush(dev, frontbuffer_bits, ORIGIN_FLIP);
 }
 
 int intel_fb_align_height(struct drm_device *dev, int height,
 			  unsigned int tiling);
-void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire);
+void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire,
+			enum fb_op_origin origin);
 
 
 /* intel_audio.c */
diff --git a/drivers/gpu/drm/i915/intel_frontbuffer.c b/drivers/gpu/drm/i915/intel_frontbuffer.c
index 73cb6e0..0a773981 100644
--- a/drivers/gpu/drm/i915/intel_frontbuffer.c
+++ b/drivers/gpu/drm/i915/intel_frontbuffer.c
@@ -127,6 +127,7 @@  static void intel_mark_fb_busy(struct drm_device *dev,
  * intel_fb_obj_invalidate - invalidate frontbuffer object
  * @obj: GEM object to invalidate
  * @ring: set for asynchronous rendering
+ * @origin: which operation caused the invalidation
  *
  * This function gets called every time rendering on the given object starts and
  * frontbuffer caching (fbc, low refresh rate for DRRS, panel self refresh) must
@@ -135,7 +136,8 @@  static void intel_mark_fb_busy(struct drm_device *dev,
  * scheduled.
  */
 void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
-			     struct intel_engine_cs *ring)
+			     struct intel_engine_cs *ring,
+			     enum fb_op_origin origin)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -164,6 +166,7 @@  void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
  * intel_frontbuffer_flush - flush frontbuffer
  * @dev: DRM device
  * @frontbuffer_bits: frontbuffer plane tracking bits
+ * @origin: which operation caused the flush
  *
  * This function gets called every time rendering on the given planes has
  * completed and frontbuffer caching can be started again. Flushes will get
@@ -172,7 +175,8 @@  void intel_fb_obj_invalidate(struct drm_i915_gem_object *obj,
  * Can be called without any locks held.
  */
 void intel_frontbuffer_flush(struct drm_device *dev,
-			     unsigned frontbuffer_bits)
+			     unsigned frontbuffer_bits,
+			     enum fb_op_origin origin)
 {
 	struct drm_i915_private *dev_priv = dev->dev_private;
 
@@ -201,13 +205,14 @@  void intel_frontbuffer_flush(struct drm_device *dev,
  * intel_fb_obj_flush - flush frontbuffer object
  * @obj: GEM object to flush
  * @retire: set when retiring asynchronous rendering
+ * @origin: which operation caused the flush
  *
  * This function gets called every time rendering on the given object has
  * completed and frontbuffer caching can be started again. If @retire is true
  * then any delayed flushes will be unblocked.
  */
 void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
-			bool retire)
+			bool retire, enum fb_op_origin origin)
 {
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *dev_priv = dev->dev_private;
@@ -229,7 +234,7 @@  void intel_fb_obj_flush(struct drm_i915_gem_object *obj,
 		mutex_unlock(&dev_priv->fb_tracking.lock);
 	}
 
-	intel_frontbuffer_flush(dev, frontbuffer_bits);
+	intel_frontbuffer_flush(dev, frontbuffer_bits, origin);
 }
 
 /**
@@ -277,5 +282,5 @@  void intel_frontbuffer_flip_complete(struct drm_device *dev,
 	dev_priv->fb_tracking.flip_bits &= ~frontbuffer_bits;
 	mutex_unlock(&dev_priv->fb_tracking.lock);
 
-	intel_frontbuffer_flush(dev, frontbuffer_bits);
+	intel_frontbuffer_flush(dev, frontbuffer_bits, ORIGIN_FLIP);
 }