Merge drm-upstream/drm-next into drm-misc-next

We got a few conflicts in drm_atomic.c after merging the DRM writeback support,
now we need a backmerge to unlock develop development on drm-misc-next.

Signed-off-by: Gustavo Padovan <gustavo.padovan@collabora.com>

12 files changed, 594 insertions, 315 deletions

diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 9f83a65b5ea9..c3afe7b2237e 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -163,28 +163,89 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 			 BLEND_COLOR_KEY_NONE;
 	u32 blendnokey = BLEND_WEIGHT1(255) | BLEND_WEIGHT0(255);
 	struct tegra_plane_state *state;
+	u32 blending[2];
 	unsigned int i;
 
+	/* disable blending for non-overlapping case */
+	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
+	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+
 	state = to_tegra_plane_state(plane->base.state);
 
-	/* alpha contribution is 1 minus sum of overlapping windows */
-	for (i = 0; i < 3; i++) {
-		if (state->dependent[i])
-			background[i] |= BLEND_CONTROL_DEPENDENT;
-	}
+	if (state->opaque) {
+		/*
+		 * Since custom fix-weight blending isn't utilized and weight
+		 * of top window is set to max, we can enforce dependent
+		 * blending which in this case results in transparent bottom
+		 * window if top window is opaque and if top window enables
+		 * alpha blending, then bottom window is getting alpha value
+		 * of 1 minus the sum of alpha components of the overlapping
+		 * plane.
+		 */
+		background[0] |= BLEND_CONTROL_DEPENDENT;
+		background[1] |= BLEND_CONTROL_DEPENDENT;
 
-	/* enable alpha blending if pixel format has an alpha component */
-	if (!state->opaque)
+		/*
+		 * The region where three windows overlap is the intersection
+		 * of the two regions where two windows overlap. It contributes
+		 * to the area if all of the windows on top of it have an alpha
+		 * component.
+		 */
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+
+		case 1:
+			background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+		}
+	} else {
+		/*
+		 * Enable alpha blending if pixel format has an alpha
+		 * component.
+		 */
 		foreground |= BLEND_CONTROL_ALPHA;
 
-	/*
-	 * Disable blending and assume Window A is the bottom-most window,
-	 * Window C is the top-most window and Window B is in the middle.
-	 */
-	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
-	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+		/*
+		 * If any of the windows on top of this window is opaque, it
+		 * will completely conceal this window within that area. If
+		 * top window has an alpha component, it is blended over the
+		 * bottom window.
+		 */
+		for (i = 0; i < 2; i++) {
+			if (state->blending[i].alpha &&
+			    state->blending[i].top)
+				background[i] |= BLEND_CONTROL_DEPENDENT;
+		}
+
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
 
-	switch (plane->index) {
+		case 1:
+			/*
+			 * When both middle and topmost windows have an alpha,
+			 * these windows a mixed together and then the result
+			 * is blended over the bottom window.
+			 */
+			if (state->blending[0].alpha &&
+			    state->blending[0].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+
+			if (state->blending[1].alpha &&
+			    state->blending[1].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+			break;
+		}
+	}
+
+	switch (state->base.normalized_zpos) {
 	case 0:
 		tegra_plane_writel(plane, background[0], DC_WIN_BLEND_2WIN_X);
 		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
@@ -192,8 +253,21 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 		break;
 
 	case 1:
-		tegra_plane_writel(plane, foreground, DC_WIN_BLEND_2WIN_X);
-		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
+		/*
+		 * If window B / C is topmost, then X / Y registers are
+		 * matching the order of blending[...] state indices,
+		 * otherwise a swap is required.
+		 */
+		if (!state->blending[0].top && state->blending[1].top) {
+			blending[0] = foreground;
+			blending[1] = background[1];
+		} else {
+			blending[0] = background[0];
+			blending[1] = foreground;
+		}
+
+		tegra_plane_writel(plane, blending[0], DC_WIN_BLEND_2WIN_X);
+		tegra_plane_writel(plane, blending[1], DC_WIN_BLEND_2WIN_Y);
 		tegra_plane_writel(plane, background[2], DC_WIN_BLEND_3WIN_XY);
 		break;
 
@@ -224,6 +298,39 @@ static void tegra_plane_setup_blending(struct tegra_plane *plane,
 	tegra_plane_writel(plane, value, DC_WIN_BLEND_LAYER_CONTROL);
 }
 
+static bool
+tegra_plane_use_horizontal_filtering(struct tegra_plane *plane,
+				     const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.w == window->dst.w)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	return true;
+}
+
+static bool
+tegra_plane_use_vertical_filtering(struct tegra_plane *plane,
+				   const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.h == window->dst.h)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	if (plane->index == 2 && dc->soc->has_win_c_without_vert_filter)
+		return false;
+
+	return true;
+}
+
 static void tegra_dc_setup_window(struct tegra_plane *plane,
 				  const struct tegra_dc_window *window)
 {
@@ -361,12 +468,50 @@ static void tegra_dc_setup_window(struct tegra_plane *plane,
 	if (window->bottom_up)
 		value |= V_DIRECTION;
 
+	if (tegra_plane_use_horizontal_filtering(plane, window)) {
+		/*
+		 * Enable horizontal 6-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		tegra_plane_writel(plane, 0x00008000, DC_WIN_H_FILTER_P(0));
+		tegra_plane_writel(plane, 0x3e087ce1, DC_WIN_H_FILTER_P(1));
+		tegra_plane_writel(plane, 0x3b117ac1, DC_WIN_H_FILTER_P(2));
+		tegra_plane_writel(plane, 0x591b73aa, DC_WIN_H_FILTER_P(3));
+		tegra_plane_writel(plane, 0x57256d9a, DC_WIN_H_FILTER_P(4));
+		tegra_plane_writel(plane, 0x552f668b, DC_WIN_H_FILTER_P(5));
+		tegra_plane_writel(plane, 0x73385e8b, DC_WIN_H_FILTER_P(6));
+		tegra_plane_writel(plane, 0x72435583, DC_WIN_H_FILTER_P(7));
+		tegra_plane_writel(plane, 0x714c4c8b, DC_WIN_H_FILTER_P(8));
+		tegra_plane_writel(plane, 0x70554393, DC_WIN_H_FILTER_P(9));
+		tegra_plane_writel(plane, 0x715e389b, DC_WIN_H_FILTER_P(10));
+		tegra_plane_writel(plane, 0x71662faa, DC_WIN_H_FILTER_P(11));
+		tegra_plane_writel(plane, 0x536d25ba, DC_WIN_H_FILTER_P(12));
+		tegra_plane_writel(plane, 0x55731bca, DC_WIN_H_FILTER_P(13));
+		tegra_plane_writel(plane, 0x387a11d9, DC_WIN_H_FILTER_P(14));
+		tegra_plane_writel(plane, 0x3c7c08f1, DC_WIN_H_FILTER_P(15));
+
+		value |= H_FILTER;
+	}
+
+	if (tegra_plane_use_vertical_filtering(plane, window)) {
+		unsigned int i, k;
+
+		/*
+		 * Enable vertical 2-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		for (i = 0, k = 128; i < 16; i++, k -= 8)
+			tegra_plane_writel(plane, k, DC_WIN_V_FILTER_P(i));
+
+		value |= V_FILTER;
+	}
+
 	tegra_plane_writel(plane, value, DC_WIN_WIN_OPTIONS);
 
-	if (dc->soc->supports_blending)
-		tegra_plane_setup_blending(plane, window);
-	else
+	if (dc->soc->has_legacy_blending)
 		tegra_plane_setup_blending_legacy(plane);
+	else
+		tegra_plane_setup_blending(plane, window);
 }
 
 static const u32 tegra20_primary_formats[] = {
@@ -451,17 +596,18 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 				    struct drm_plane_state *state)
 {
 	struct tegra_plane_state *plane_state = to_tegra_plane_state(state);
+	unsigned int rotation = DRM_MODE_ROTATE_0 | DRM_MODE_REFLECT_Y;
 	struct tegra_bo_tiling *tiling = &plane_state->tiling;
 	struct tegra_plane *tegra = to_tegra_plane(plane);
 	struct tegra_dc *dc = to_tegra_dc(state->crtc);
-	unsigned int format;
 	int err;
 
 	/* no need for further checks if the plane is being disabled */
 	if (!state->crtc)
 		return 0;
 
-	err = tegra_plane_format(state->fb->format->format, &format,
+	err = tegra_plane_format(state->fb->format->format,
+				 &plane_state->format,
 				 &plane_state->swap);
 	if (err < 0)
 		return err;
@@ -472,22 +618,12 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 	 * the corresponding opaque formats. However, the opaque formats can
 	 * be emulated by disabling alpha blending for the plane.
 	 */
-	if (!dc->soc->supports_blending) {
-		if (!tegra_plane_format_has_alpha(format)) {
-			err = tegra_plane_format_get_alpha(format, &format);
-			if (err < 0)
-				return err;
-
-			plane_state->opaque = true;
-		} else {
-			plane_state->opaque = false;
-		}
-
-		tegra_plane_check_dependent(tegra, plane_state);
+	if (dc->soc->has_legacy_blending) {
+		err = tegra_plane_setup_legacy_state(tegra, plane_state);
+		if (err < 0)
+			return err;
 	}
 
-	plane_state->format = format;
-
 	err = tegra_fb_get_tiling(state->fb, tiling);
 	if (err < 0)
 		return err;
@@ -498,6 +634,13 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 		return -EINVAL;
 	}
 
+	rotation = drm_rotation_simplify(state->rotation, rotation);
+
+	if (rotation & DRM_MODE_REFLECT_Y)
+		plane_state->bottom_up = true;
+	else
+		plane_state->bottom_up = false;
+
 	/*
 	 * Tegra doesn't support different strides for U and V planes so we
 	 * error out if the user tries to display a framebuffer with such a
@@ -558,7 +701,7 @@ static void tegra_plane_atomic_update(struct drm_plane *plane,
 	window.dst.w = drm_rect_width(&plane->state->dst);
 	window.dst.h = drm_rect_height(&plane->state->dst);
 	window.bits_per_pixel = fb->format->cpp[0] * 8;
-	window.bottom_up = tegra_fb_is_bottom_up(fb);
+	window.bottom_up = tegra_fb_is_bottom_up(fb) || state->bottom_up;
 
 	/* copy from state */
 	window.zpos = plane->state->normalized_zpos;
@@ -639,9 +782,15 @@ static struct drm_plane *tegra_primary_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
 
 	return &plane->base;
 }
@@ -918,9 +1067,15 @@ static struct drm_plane *tegra_dc_overlay_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
 
 	return &plane->base;
 }
@@ -1826,7 +1981,6 @@ static irqreturn_t tegra_dc_irq(int irq, void *data)
 static int tegra_dc_init(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	unsigned long flags = HOST1X_SYNCPT_CLIENT_MANAGED;
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	struct tegra_drm *tegra = drm->dev_private;
@@ -1838,20 +1992,11 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (!dc->syncpt)
 		dev_warn(dc->dev, "failed to allocate syncpoint\n");
 
-	if (group && tegra->domain) {
-		if (group != tegra->group) {
-			err = iommu_attach_group(tegra->domain, group);
-			if (err < 0) {
-				dev_err(dc->dev,
-					"failed to attach to domain: %d\n",
-					err);
-				return err;
-			}
-
-			tegra->group = group;
-		}
-
-		dc->domain = tegra->domain;
+	dc->group = host1x_client_iommu_attach(client, true);
+	if (IS_ERR(dc->group)) {
+		err = PTR_ERR(dc->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		return err;
 	}
 
 	if (dc->soc->wgrps)
@@ -1916,24 +2061,15 @@ cleanup:
 	if (!IS_ERR(primary))
 		drm_plane_cleanup(primary);
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
-			tegra->group = NULL;
-		}
-
-		dc->domain = NULL;
-	}
+	host1x_client_iommu_detach(client, dc->group);
+	host1x_syncpt_free(dc->syncpt);
 
 	return err;
 }
 
 static int tegra_dc_exit(struct host1x_client *client)
 {
-	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
-	struct tegra_drm *tegra = drm->dev_private;
 	int err;
 
 	devm_free_irq(dc->dev, dc->irq, dc);
@@ -1944,15 +2080,7 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
-			tegra->group = NULL;
-		}
-
-		dc->domain = NULL;
-	}
-
+	host1x_client_iommu_detach(client, dc->group);
 	host1x_syncpt_free(dc->syncpt);
 
 	return 0;
@@ -1968,7 +2096,7 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = true,
@@ -1978,6 +2106,8 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = true,
+	.has_win_c_without_vert_filter = true,
 };
 
 static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
@@ -1985,7 +2115,7 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = false,
@@ -1995,6 +2125,8 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
@@ -2002,7 +2134,7 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2012,6 +2144,8 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
@@ -2019,7 +2153,7 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2029,6 +2163,8 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra124_overlay_formats),
 	.overlay_formats = tegra124_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
@@ -2036,7 +2172,7 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2046,6 +2182,8 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_windowgroup_soc tegra186_dc_wgrps[] = {
@@ -2087,7 +2225,7 @@ static const struct tegra_dc_soc_info tegra186_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = false,
 	.coupled_pm = false,
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index d2b50d32de4d..e96f582ca692 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -55,7 +55,7 @@ struct tegra_dc_soc_info {
 	bool supports_interlacing;
 	bool supports_cursor;
 	bool supports_block_linear;
-	bool supports_blending;
+	bool has_legacy_blending;
 	unsigned int pitch_align;
 	bool has_powergate;
 	bool coupled_pm;
@@ -67,6 +67,8 @@ struct tegra_dc_soc_info {
 	const u32 *overlay_formats;
 	unsigned int num_overlay_formats;
 	const u64 *modifiers;
+	bool has_win_a_without_filters;
+	bool has_win_c_without_vert_filter;
 };
 
 struct tegra_dc {
@@ -92,7 +94,7 @@ struct tegra_dc {
 
 	const struct tegra_dc_soc_info *soc;
 
-	struct iommu_domain *domain;
+	struct iommu_group *group;
 };
 
 static inline struct tegra_dc *
@@ -553,6 +555,9 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define  THREAD_NUM(x) (((x) & 0x1f) << 1)
 #define  THREAD_GROUP_ENABLE (1 << 0)
 
+#define DC_WIN_H_FILTER_P(p)			(0x601 + (p))
+#define DC_WIN_V_FILTER_P(p)			(0x619 + (p))
+
 #define DC_WIN_CSC_YOF				0x611
 #define DC_WIN_CSC_KYRGB			0x612
 #define DC_WIN_CSC_KUR				0x613
@@ -566,6 +571,8 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define H_DIRECTION  (1 <<  0)
 #define V_DIRECTION  (1 <<  2)
 #define COLOR_EXPAND (1 <<  6)
+#define H_FILTER     (1 <<  8)
+#define V_FILTER     (1 << 10)
 #define CSC_ENABLE   (1 << 18)
 #define WIN_ENABLE   (1 << 30)
 
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index a0519612ae2c..776c1513e582 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -98,6 +98,10 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 			goto free;
 		}
 
+		err = iova_cache_get();
+		if (err < 0)
+			goto domain;
+
 		geometry = &tegra->domain->geometry;
 		gem_start = geometry->aperture_start;
 		gem_end = geometry->aperture_end - CARVEOUT_SZ;
@@ -191,11 +195,14 @@ config:
 	drm_mode_config_cleanup(drm);
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
 	}
+domain:
+	if (tegra->domain)
+		iommu_domain_free(tegra->domain);
 free:
 	kfree(tegra);
 	return err;
@@ -217,10 +224,11 @@ static void tegra_drm_unload(struct drm_device *drm)
 		return;
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
+		iommu_domain_free(tegra->domain);
 	}
 
 	kfree(tegra);
@@ -300,46 +308,15 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 	return 0;
 }
 
-static int host1x_waitchk_copy_from_user(struct host1x_waitchk *dest,
-					 struct drm_tegra_waitchk __user *src,
-					 struct drm_file *file)
-{
-	u32 cmdbuf;
-	int err;
-
-	err = get_user(cmdbuf, &src->handle);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->offset, &src->offset);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->syncpt_id, &src->syncpt);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->thresh, &src->thresh);
-	if (err < 0)
-		return err;
-
-	dest->bo = host1x_bo_lookup(file, cmdbuf);
-	if (!dest->bo)
-		return -ENOENT;
-
-	return 0;
-}
-
 int tegra_drm_submit(struct tegra_drm_context *context,
 		     struct drm_tegra_submit *args, struct drm_device *drm,
 		     struct drm_file *file)
 {
+	struct host1x_client *client = &context->client->base;
 	unsigned int num_cmdbufs = args->num_cmdbufs;
 	unsigned int num_relocs = args->num_relocs;
-	unsigned int num_waitchks = args->num_waitchks;
 	struct drm_tegra_cmdbuf __user *user_cmdbufs;
 	struct drm_tegra_reloc __user *user_relocs;
-	struct drm_tegra_waitchk __user *user_waitchks;
 	struct drm_tegra_syncpt __user *user_syncpt;
 	struct drm_tegra_syncpt syncpt;
 	struct host1x *host1x = dev_get_drvdata(drm->dev->parent);
@@ -351,7 +328,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 
 	user_cmdbufs = u64_to_user_ptr(args->cmdbufs);
 	user_relocs = u64_to_user_ptr(args->relocs);
-	user_waitchks = u64_to_user_ptr(args->waitchks);
 	user_syncpt = u64_to_user_ptr(args->syncpts);
 
 	/* We don't yet support other than one syncpt_incr struct per submit */
@@ -363,21 +339,20 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		return -EINVAL;
 
 	job = host1x_job_alloc(context->channel, args->num_cmdbufs,
-			       args->num_relocs, args->num_waitchks);
+			       args->num_relocs);
 	if (!job)
 		return -ENOMEM;
 
 	job->num_relocs = args->num_relocs;
-	job->num_waitchk = args->num_waitchks;
-	job->client = (u32)args->context;
-	job->class = context->client->base.class;
+	job->client = client;
+	job->class = client->class;
 	job->serialize = true;
 
 	/*
 	 * Track referenced BOs so that they can be unreferenced after the
 	 * submission is complete.
 	 */
-	num_refs = num_cmdbufs + num_relocs * 2 + num_waitchks;
+	num_refs = num_cmdbufs + num_relocs * 2;
 
 	refs = kmalloc_array(num_refs, sizeof(*refs), GFP_KERNEL);
 	if (!refs) {
@@ -438,13 +413,13 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		struct host1x_reloc *reloc;
 		struct tegra_bo *obj;
 
-		err = host1x_reloc_copy_from_user(&job->relocarray[num_relocs],
+		err = host1x_reloc_copy_from_user(&job->relocs[num_relocs],
 						  &user_relocs[num_relocs], drm,
 						  file);
 		if (err < 0)
 			goto fail;
 
-		reloc = &job->relocarray[num_relocs];
+		reloc = &job->relocs[num_relocs];
 		obj = host1x_to_tegra_bo(reloc->cmdbuf.bo);
 		refs[num_refs++] = &obj->gem;
 
@@ -468,30 +443,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		}
 	}
 
-	/* copy and resolve waitchks from submit */
-	while (num_waitchks--) {
-		struct host1x_waitchk *wait = &job->waitchk[num_waitchks];
-		struct tegra_bo *obj;
-
-		err = host1x_waitchk_copy_from_user(
-			wait, &user_waitchks[num_waitchks], file);
-		if (err < 0)
-			goto fail;
-
-		obj = host1x_to_tegra_bo(wait->bo);
-		refs[num_refs++] = &obj->gem;
-
-		/*
-		 * The unaligned offset will cause an unaligned write during
-		 * of the waitchks patching, corrupting the commands stream.
-		 */
-		if (wait->offset & 3 ||
-		    wait->offset >= obj->gem.size) {
-			err = -EINVAL;
-			goto fail;
-		}
-	}
-
 	if (copy_from_user(&syncpt, user_syncpt, sizeof(syncpt))) {
 		err = -EFAULT;
 		goto fail;
@@ -1101,6 +1052,52 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+	struct iommu_group *group = NULL;
+	int err;
+
+	if (tegra->domain) {
+		group = iommu_group_get(client->dev);
+		if (!group) {
+			dev_err(client->dev, "failed to get IOMMU group\n");
+			return ERR_PTR(-ENODEV);
+		}
+
+		if (!shared || (shared && (group != tegra->group))) {
+			err = iommu_attach_group(tegra->domain, group);
+			if (err < 0) {
+				iommu_group_put(group);
+				return ERR_PTR(err);
+			}
+
+			if (shared && !tegra->group)
+				tegra->group = group;
+		}
+	}
+
+	return group;
+}
+
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+
+	if (group) {
+		if (group == tegra->group) {
+			iommu_detach_group(tegra->domain, group);
+			tegra->group = NULL;
+		}
+
+		iommu_group_put(group);
+	}
+}
+
 void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *dma)
 {
 	struct iova *alloc;
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 4f41aaec8530..92d248784396 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -29,16 +29,10 @@
 
 struct reset_control;
 
-struct tegra_fb {
-	struct drm_framebuffer base;
-	struct tegra_bo **planes;
-	unsigned int num_planes;
-};
-
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 struct tegra_fbdev {
 	struct drm_fb_helper base;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 };
 #endif
 
@@ -97,6 +91,7 @@ struct tegra_drm_client {
 	struct host1x_client base;
 	struct list_head list;
 
+	unsigned int version;
 	const struct tegra_drm_client_ops *ops;
 };
 
@@ -110,6 +105,10 @@ int tegra_drm_register_client(struct tegra_drm *tegra,
 			      struct tegra_drm_client *client);
 int tegra_drm_unregister_client(struct tegra_drm *tegra,
 				struct tegra_drm_client *client);
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared);
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group);
 
 int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
 int tegra_drm_exit(struct tegra_drm *tegra);
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index e69434909a42..4c22cdded3c2 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -14,11 +14,7 @@
 
 #include "drm.h"
 #include "gem.h"
-
-static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
-{
-	return container_of(fb, struct tegra_fb, base);
-}
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
@@ -30,19 +26,14 @@ static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	if (index >= framebuffer->format->num_planes)
-		return NULL;
-
-	return fb->planes[index];
+	return to_tegra_bo(drm_gem_fb_get_obj(framebuffer, index));
 }
 
 bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+	struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, 0);
 
-	if (fb->planes[0]->flags & TEGRA_BO_BOTTOM_UP)
+	if (bo->flags & TEGRA_BO_BOTTOM_UP)
 		return true;
 
 	return false;
@@ -51,8 +42,7 @@ bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 			struct tegra_bo_tiling *tiling)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-	uint64_t modifier = fb->base.modifier;
+	uint64_t modifier = framebuffer->modifier;
 
 	switch (modifier) {
 	case DRM_FORMAT_MOD_LINEAR:
@@ -102,46 +92,17 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 	return 0;
 }
 
-static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
-{
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-	unsigned int i;
-
-	for (i = 0; i < fb->num_planes; i++) {
-		struct tegra_bo *bo = fb->planes[i];
-
-		if (bo) {
-			if (bo->pages)
-				vunmap(bo->vaddr);
-
-			drm_gem_object_put_unlocked(&bo->gem);
-		}
-	}
-
-	drm_framebuffer_cleanup(framebuffer);
-	kfree(fb->planes);
-	kfree(fb);
-}
-
-static int tegra_fb_create_handle(struct drm_framebuffer *framebuffer,
-				  struct drm_file *file, unsigned int *handle)
-{
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	return drm_gem_handle_create(file, &fb->planes[0]->gem, handle);
-}
-
 static const struct drm_framebuffer_funcs tegra_fb_funcs = {
-	.destroy = tegra_fb_destroy,
-	.create_handle = tegra_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
-static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
-				       const struct drm_mode_fb_cmd2 *mode_cmd,
-				       struct tegra_bo **planes,
-				       unsigned int num_planes)
+static struct drm_framebuffer *tegra_fb_alloc(struct drm_device *drm,
+					      const struct drm_mode_fb_cmd2 *mode_cmd,
+					      struct tegra_bo **planes,
+					      unsigned int num_planes)
 {
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	unsigned int i;
 	int err;
 
@@ -149,24 +110,15 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
-	if (!fb->planes) {
-		kfree(fb);
-		return ERR_PTR(-ENOMEM);
-	}
-
-	fb->num_planes = num_planes;
+	drm_helper_mode_fill_fb_struct(drm, fb, mode_cmd);
 
-	drm_helper_mode_fill_fb_struct(drm, &fb->base, mode_cmd);
+	for (i = 0; i < fb->format->num_planes; i++)
+		fb->obj[i] = &planes[i]->gem;
 
-	for (i = 0; i < fb->num_planes; i++)
-		fb->planes[i] = planes[i];
-
-	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
+	err = drm_framebuffer_init(drm, fb, &tegra_fb_funcs);
 	if (err < 0) {
 		dev_err(drm->dev, "failed to initialize framebuffer: %d\n",
 			err);
-		kfree(fb->planes);
 		kfree(fb);
 		return ERR_PTR(err);
 	}
@@ -181,7 +133,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 	unsigned int hsub, vsub, i;
 	struct tegra_bo *planes[4];
 	struct drm_gem_object *gem;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	int err;
 
 	hsub = drm_format_horz_chroma_subsampling(cmd->pixel_format);
@@ -217,7 +169,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 		goto unreference;
 	}
 
-	return &fb->base;
+	return fb;
 
 unreference:
 	while (i--)
@@ -298,7 +250,7 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 		return PTR_ERR(fbdev->fb);
 	}
 
-	fb = &fbdev->fb->base;
+	fb = fbdev->fb;
 	helper->fb = fb;
 	helper->fbdev = info;
 
@@ -398,8 +350,17 @@ static void tegra_fbdev_exit(struct tegra_fbdev *fbdev)
 {
 	drm_fb_helper_unregister_fbi(&fbdev->base);
 
-	if (fbdev->fb)
-		drm_framebuffer_remove(&fbdev->fb->base);
+	if (fbdev->fb) {
+		struct tegra_bo *bo = tegra_fb_get_plane(fbdev->fb, 0);
+
+		/* Undo the special mapping we made in fbdev probe. */
+		if (bo && bo->pages) {
+			vunmap(bo->vaddr);
+			bo->vaddr = 0;
+		}
+
+		drm_framebuffer_remove(fbdev->fb);
+	}
 
 	drm_fb_helper_fini(&fbdev->base);
 	tegra_fbdev_free(fbdev);
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index d7661702c11c..4f80100ff5f3 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -422,14 +422,13 @@ int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 	return 0;
 }
 
-static int tegra_bo_fault(struct vm_fault *vmf)
+static vm_fault_t tegra_bo_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *gem = vma->vm_private_data;
 	struct tegra_bo *bo = to_tegra_bo(gem);
 	struct page *page;
 	pgoff_t offset;
-	int err;
 
 	if (!bo->pages)
 		return VM_FAULT_SIGBUS;
@@ -437,20 +436,7 @@ static int tegra_bo_fault(struct vm_fault *vmf)
 	offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 	page = bo->pages[offset];
 
-	err = vm_insert_page(vma, vmf->address, page);
-	switch (err) {
-	case -EAGAIN:
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-	case -EBUSY:
-		return VM_FAULT_NOPAGE;
-
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	}
-
-	return VM_FAULT_SIGBUS;
+	return vmf_insert_page(vma, vmf->address, page);
 }
 
 const struct vm_operations_struct tegra_bo_vm_ops = {
@@ -649,6 +635,8 @@ struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
 {
 	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
 
+	exp_info.exp_name = KBUILD_MODNAME;
+	exp_info.owner = drm->driver->fops->owner;
 	exp_info.ops = &tegra_gem_prime_dmabuf_ops;
 	exp_info.size = gem->size;
 	exp_info.flags = flags;
diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 9a8ea93016a9..673059fd2fcb 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -7,16 +7,25 @@
  */
 
 #include <linux/clk.h>
+#include <linux/iommu.h>
+#include <linux/of_device.h>
 
 #include "drm.h"
 #include "gem.h"
 #include "gr2d.h"
 
+struct gr2d_soc {
+	unsigned int version;
+};
+
 struct gr2d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk;
 
+	const struct gr2d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR2D_NUM_REGS);
 };
 
@@ -31,6 +40,7 @@ static int gr2d_init(struct host1x_client *client)
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr2d *gr2d = to_gr2d(drm);
+	int err;
 
 	gr2d->channel = host1x_channel_request(client->dev);
 	if (!gr2d->channel)
@@ -38,24 +48,48 @@ static int gr2d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr2d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
+	}
+
+	gr2d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr2d->group)) {
+		err = PTR_ERR(gr2d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
+	}
+
+	err = tegra_drm_register_client(dev->dev_private, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	return 0;
+
+detach:
+	host1x_client_iommu_detach(client, gr2d->group);
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr2d->channel);
+	return err;
 }
 
 static int gr2d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
-	err = tegra_drm_unregister_client(dev->dev_private, drm);
+	err = tegra_drm_unregister_client(tegra, drm);
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr2d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr2d->channel);
 
@@ -123,9 +157,17 @@ static const struct tegra_drm_client_ops gr2d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr2d_soc tegra20_gr2d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr2d_soc tegra30_gr2d_soc = {
+	.version = 0x30,
+};
+
 static const struct of_device_id gr2d_match[] = {
-	{ .compatible = "nvidia,tegra30-gr2d" },
-	{ .compatible = "nvidia,tegra20-gr2d" },
+	{ .compatible = "nvidia,tegra30-gr2d", .data = &tegra20_gr2d_soc },
+	{ .compatible = "nvidia,tegra20-gr2d", .data = &tegra30_gr2d_soc },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, gr2d_match);
@@ -158,6 +200,8 @@ static int gr2d_probe(struct platform_device *pdev)
 	if (!gr2d)
 		return -ENOMEM;
 
+	gr2d->soc = of_device_get_match_data(dev);
+
 	syncpts = devm_kzalloc(dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -182,6 +226,7 @@ static int gr2d_probe(struct platform_device *pdev)
 	gr2d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr2d->client.list);
+	gr2d->client.version = gr2d->soc->version;
 	gr2d->client.ops = &gr2d_ops;
 
 	err = host1x_client_register(&gr2d->client.base);
diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 28c4ef63065b..4778ae999668 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -9,7 +9,9 @@
 
 #include <linux/clk.h>
 #include <linux/host1x.h>
+#include <linux/iommu.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 
@@ -19,7 +21,12 @@
 #include "gem.h"
 #include "gr3d.h"
 
+struct gr3d_soc {
+	unsigned int version;
+};
+
 struct gr3d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk_secondary;
@@ -27,6 +34,8 @@ struct gr3d {
 	struct reset_control *rst_secondary;
 	struct reset_control *rst;
 
+	const struct gr3d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR3D_NUM_REGS);
 };
 
@@ -41,6 +50,7 @@ static int gr3d_init(struct host1x_client *client)
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr3d *gr3d = to_gr3d(drm);
+	int err;
 
 	gr3d->channel = host1x_channel_request(client->dev);
 	if (!gr3d->channel)
@@ -48,11 +58,33 @@ static int gr3d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr3d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	gr3d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr3d->group)) {
+		err = PTR_ERR(gr3d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
+	}
+
+	err = tegra_drm_register_client(dev->dev_private, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
+	}
+
+	return 0;
+
+detach:
+	host1x_client_iommu_detach(client, gr3d->group);
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr3d->channel);
+	return err;
 }
 
 static int gr3d_exit(struct host1x_client *client)
@@ -66,6 +98,7 @@ static int gr3d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr3d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr3d->channel);
 
@@ -125,10 +158,22 @@ static const struct tegra_drm_client_ops gr3d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr3d_soc tegra20_gr3d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr3d_soc tegra30_gr3d_soc = {
+	.version = 0x30,
+};
+
+static const struct gr3d_soc tegra114_gr3d_soc = {
+	.version = 0x35,
+};
+
 static const struct of_device_id tegra_gr3d_match[] = {
-	{ .compatible = "nvidia,tegra114-gr3d" },
-	{ .compatible = "nvidia,tegra30-gr3d" },
-	{ .compatible = "nvidia,tegra20-gr3d" },
+	{ .compatible = "nvidia,tegra114-gr3d", .data = &tegra114_gr3d_soc },
+	{ .compatible = "nvidia,tegra30-gr3d", .data = &tegra30_gr3d_soc },
+	{ .compatible = "nvidia,tegra20-gr3d", .data = &tegra20_gr3d_soc },
 	{ }
 };
 MODULE_DEVICE_TABLE(of, tegra_gr3d_match);
@@ -250,6 +295,8 @@ static int gr3d_probe(struct platform_device *pdev)
 	if (!gr3d)
 		return -ENOMEM;
 
+	gr3d->soc = of_device_get_match_data(&pdev->dev);
+
 	syncpts = devm_kzalloc(&pdev->dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -307,6 +354,7 @@ static int gr3d_probe(struct platform_device *pdev)
 	gr3d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr3d->client.list);
+	gr3d->client.version = gr3d->soc->version;
 	gr3d->client.ops = &gr3d_ops;
 
 	err = host1x_client_register(&gr3d->client.base);
diff --git a/drivers/gpu/drm/tegra/hub.c b/drivers/gpu/drm/tegra/hub.c
index 9a3f23d4780f..8f4fcbb515fb 100644
--- a/drivers/gpu/drm/tegra/hub.c
+++ b/drivers/gpu/drm/tegra/hub.c
@@ -687,7 +687,7 @@ void tegra_display_hub_atomic_commit(struct drm_device *drm,
 	struct device *dev = hub->client.dev;
 	int err;
 
-	hub_state = tegra_display_hub_get_state(hub, state);
+	hub_state = to_tegra_display_hub_state(hub->base.state);
 
 	if (hub_state->clk) {
 		err = clk_set_rate(hub_state->clk, hub_state->rate);
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 176ef46c615c..d068e8aa3553 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -23,6 +23,7 @@ static void tegra_plane_destroy(struct drm_plane *plane)
 
 static void tegra_plane_reset(struct drm_plane *plane)
 {
+	struct tegra_plane *p = to_tegra_plane(plane);
 	struct tegra_plane_state *state;
 
 	if (plane->state)
@@ -35,6 +36,8 @@ static void tegra_plane_reset(struct drm_plane *plane)
 	if (state) {
 		plane->state = &state->base;
 		plane->state->plane = plane;
+		plane->state->zpos = p->index;
+		plane->state->normalized_zpos = p->index;
 	}
 }
 
@@ -53,10 +56,11 @@ tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
 	copy->tiling = state->tiling;
 	copy->format = state->format;
 	copy->swap = state->swap;
+	copy->bottom_up = state->bottom_up;
 	copy->opaque = state->opaque;
 
-	for (i = 0; i < 3; i++)
-		copy->dependent[i] = state->dependent[i];
+	for (i = 0; i < 2; i++)
+		copy->blending[i] = state->blending[i];
 
 	return &copy->base;
 }
@@ -267,24 +271,8 @@ static bool __drm_format_has_alpha(u32 format)
 	return false;
 }
 
-/*
- * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
- * be emulated using the alpha formats and alpha blending disabled.
- */
-bool tegra_plane_format_has_alpha(unsigned int format)
-{
-	switch (format) {
-	case WIN_COLOR_DEPTH_B5G5R5A1:
-	case WIN_COLOR_DEPTH_A1B5G5R5:
-	case WIN_COLOR_DEPTH_R8G8B8A8:
-	case WIN_COLOR_DEPTH_B8G8R8A8:
-		return true;
-	}
-
-	return false;
-}
-
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
+static int tegra_plane_format_get_alpha(unsigned int opaque,
+					unsigned int *alpha)
 {
 	if (tegra_plane_format_is_yuv(opaque, NULL)) {
 		*alpha = opaque;
@@ -316,6 +304,67 @@ int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
 	return -EINVAL;
 }
 
+/*
+ * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
+ * be emulated using the alpha formats and alpha blending disabled.
+ */
+static int tegra_plane_setup_opacity(struct tegra_plane *tegra,
+				     struct tegra_plane_state *state)
+{
+	unsigned int format;
+	int err;
+
+	switch (state->format) {
+	case WIN_COLOR_DEPTH_B5G5R5A1:
+	case WIN_COLOR_DEPTH_A1B5G5R5:
+	case WIN_COLOR_DEPTH_R8G8B8A8:
+	case WIN_COLOR_DEPTH_B8G8R8A8:
+		state->opaque = false;
+		break;
+
+	default:
+		err = tegra_plane_format_get_alpha(state->format, &format);
+		if (err < 0)
+			return err;
+
+		state->format = format;
+		state->opaque = true;
+		break;
+	}
+
+	return 0;
+}
+
+static int tegra_plane_check_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct drm_plane_state *old, *plane_state;
+	struct drm_plane *plane;
+
+	old = drm_atomic_get_old_plane_state(state->base.state, &tegra->base);
+
+	/* check if zpos / transparency changed */
+	if (old->normalized_zpos == state->base.normalized_zpos &&
+	    to_tegra_plane_state(old)->opaque == state->opaque)
+		return 0;
+
+	/* include all sibling planes into this commit */
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip this plane and planes on different CRTCs */
+		if (p == tegra || p->dc != tegra->dc)
+			continue;
+
+		plane_state = drm_atomic_get_plane_state(state->base.state,
+							 plane);
+		if (IS_ERR(plane_state))
+			return PTR_ERR(plane_state);
+	}
+
+	return 1;
+}
+
 static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 						  struct tegra_plane *other)
 {
@@ -336,61 +385,98 @@ static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 	return index;
 }
 
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state)
+static void tegra_plane_update_transparency(struct tegra_plane *tegra,
+					    struct tegra_plane_state *state)
 {
-	struct drm_plane_state *old, *new;
+	struct drm_plane_state *new;
 	struct drm_plane *plane;
-	unsigned int zpos[2];
 	unsigned int i;
 
-	for (i = 0; i < 2; i++)
-		zpos[i] = 0;
-
-	for_each_oldnew_plane_in_state(state->base.state, plane, old, new, i) {
+	for_each_new_plane_in_state(state->base.state, plane, new, i) {
 		struct tegra_plane *p = to_tegra_plane(plane);
 		unsigned index;
 
 		/* skip this plane and planes on different CRTCs */
-		if (p == tegra || new->crtc != state->base.crtc)
+		if (p == tegra || p->dc != tegra->dc)
 			continue;
 
 		index = tegra_plane_get_overlap_index(tegra, p);
 
-		state->dependent[index] = false;
+		if (new->fb && __drm_format_has_alpha(new->fb->format->format))
+			state->blending[index].alpha = true;
+		else
+			state->blending[index].alpha = false;
+
+		if (new->normalized_zpos > state->base.normalized_zpos)
+			state->blending[index].top = true;
+		else
+			state->blending[index].top = false;
 
 		/*
-		 * If any of the other planes is on top of this plane and uses
-		 * a format with an alpha component, mark this plane as being
-		 * dependent, meaning it's alpha value will be 1 minus the sum
-		 * of alpha components of the overlapping planes.
+		 * Missing framebuffer means that plane is disabled, in this
+		 * case mark B / C window as top to be able to differentiate
+		 * windows indices order in regards to zPos for the middle
+		 * window X / Y registers programming.
 		 */
-		if (p->index > tegra->index) {
-			if (__drm_format_has_alpha(new->fb->format->format))
-				state->dependent[index] = true;
-
-			/* keep track of the Z position */
-			zpos[index] = p->index;
-		}
+		if (!new->fb)
+			state->blending[index].top = (index == 1);
 	}
+}
+
+static int tegra_plane_setup_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct tegra_plane_state *tegra_state;
+	struct drm_plane_state *new;
+	struct drm_plane *plane;
+	int err;
 
 	/*
-	 * The region where three windows overlap is the intersection of the
-	 * two regions where two windows overlap. It contributes to the area
-	 * if any of the windows on top of it have an alpha component.
+	 * If planes zpos / transparency changed, sibling planes blending
+	 * state may require adjustment and in this case they will be included
+	 * into this atom commit, otherwise blending state is unchanged.
 	 */
-	for (i = 0; i < 2; i++)
-		state->dependent[2] = state->dependent[2] ||
-				      state->dependent[i];
+	err = tegra_plane_check_transparency(tegra, state);
+	if (err <= 0)
+		return err;
 
 	/*
-	 * However, if any of the windows on top of this window is opaque, it
-	 * will completely conceal this window within that area, so avoid the
-	 * window from contributing to the area.
+	 * All planes are now in the atomic state, walk them up and update
+	 * transparency state for each plane.
 	 */
-	for (i = 0; i < 2; i++) {
-		if (zpos[i] > tegra->index)
-			state->dependent[2] = state->dependent[2] &&
-					      state->dependent[i];
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip planes on different CRTCs */
+		if (p->dc != tegra->dc)
+			continue;
+
+		new = drm_atomic_get_new_plane_state(state->base.state, plane);
+		tegra_state = to_tegra_plane_state(new);
+
+		/*
+		 * There is no need to update blending state for the disabled
+		 * plane.
+		 */
+		if (new->fb)
+			tegra_plane_update_transparency(p, tegra_state);
 	}
+
+	return 0;
+}
+
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state)
+{
+	int err;
+
+	err = tegra_plane_setup_opacity(tegra, state);
+	if (err < 0)
+		return err;
+
+	err = tegra_plane_setup_transparency(tegra, state);
+	if (err < 0)
+		return err;
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/plane.h b/drivers/gpu/drm/tegra/plane.h
index 6938719e7e5d..e79e6b4a8e0a 100644
--- a/drivers/gpu/drm/tegra/plane.h
+++ b/drivers/gpu/drm/tegra/plane.h
@@ -34,6 +34,11 @@ static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
 	return container_of(plane, struct tegra_plane, base);
 }
 
+struct tegra_plane_legacy_blending_state {
+	bool alpha;
+	bool top;
+};
+
 struct tegra_plane_state {
 	struct drm_plane_state base;
 
@@ -41,9 +46,11 @@ struct tegra_plane_state {
 	u32 format;
 	u32 swap;
 
+	bool bottom_up;
+
 	/* used for legacy blending support only */
+	struct tegra_plane_legacy_blending_state blending[2];
 	bool opaque;
-	bool dependent[3];
 };
 
 static inline struct tegra_plane_state *
@@ -62,9 +69,7 @@ int tegra_plane_state_add(struct tegra_plane *plane,
 
 int tegra_plane_format(u32 fourcc, u32 *format, u32 *swap);
 bool tegra_plane_format_is_yuv(unsigned int format, bool *planar);
-bool tegra_plane_format_has_alpha(unsigned int format);
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha);
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state);
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state);
 
 #endif /* TEGRA_PLANE_H */
diff --git a/drivers/gpu/drm/tegra/vic.c b/drivers/gpu/drm/tegra/vic.c
index f5794dd49f3b..9f657a63b0bb 100644
--- a/drivers/gpu/drm/tegra/vic.c
+++ b/drivers/gpu/drm/tegra/vic.c
@@ -25,6 +25,7 @@
 
 struct vic_config {
 	const char *firmware;
+	unsigned int version;
 };
 
 struct vic {
@@ -264,18 +265,21 @@ static const struct tegra_drm_client_ops vic_ops = {
 
 static const struct vic_config vic_t124_config = {
 	.firmware = NVIDIA_TEGRA_124_VIC_FIRMWARE,
+	.version = 0x40,
 };
 
 #define NVIDIA_TEGRA_210_VIC_FIRMWARE "nvidia/tegra210/vic04_ucode.bin"
 
 static const struct vic_config vic_t210_config = {
 	.firmware = NVIDIA_TEGRA_210_VIC_FIRMWARE,
+	.version = 0x21,
 };
 
 #define NVIDIA_TEGRA_186_VIC_FIRMWARE "nvidia/tegra186/vic04_ucode.bin"
 
 static const struct vic_config vic_t186_config = {
 	.firmware = NVIDIA_TEGRA_186_VIC_FIRMWARE,
+	.version = 0x18,
 };
 
 static const struct of_device_id vic_match[] = {
@@ -342,6 +346,7 @@ static int vic_probe(struct platform_device *pdev)
 	vic->dev = dev;
 
 	INIT_LIST_HEAD(&vic->client.list);
+	vic->client.version = vic->config->version;
 	vic->client.ops = &vic_ops;
 
 	err = host1x_client_register(&vic->client.base);