The trick is not strictly necessary on SKL because the offset
registers allow more bits. But for FBC, doing this changes how the
hardware tracking works - it starts at the surface address we provide
- so there's a higher chance that the CRTC will be pointing to an area
of the frontbuffer that is actually being covered by the hardware
tracking mechanism. This fixes fbc-farfromfence on SKL.
Testcase: igt/kms_frontbuffer_tracking/fbc-farfromfence
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@xxxxxxxxx>
---
drivers/gpu/drm/i915/intel_display.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 24b8a72..d40ae71 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -3031,6 +3031,7 @@ static void skylake_update_primary_plane(struct drm_crtc *crtc,
int src_x = 0, src_y = 0, src_w = 0, src_h = 0;
int dst_x = 0, dst_y = 0, dst_w = 0, dst_h = 0;
int scaler_id = -1;
+ int pixel_size;
plane_state = to_intel_plane_state(plane->state);
@@ -3079,6 +3080,12 @@ static void skylake_update_primary_plane(struct drm_crtc *crtc,
src_h = intel_crtc->config->pipe_src_h;
}
+ pixel_size = drm_format_plane_cpp(fb->pixel_format, 0);
+ intel_crtc->dspaddr_offset = intel_gen4_compute_page_offset(dev_priv,
+ &x, &y, obj->tiling_mode,
+ pixel_size, fb->pitches[0]);
+ surf_addr += intel_crtc->dspaddr_offset;
+
if (intel_rotation_90_or_270(rotation)) {
/* stride = Surface height in tiles */
tile_height = intel_tile_height(dev, fb->pixel_format,
--
2.5.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
