Merge tag 'drm/tegra/for-4.16-rc1-fixes' of git://anongit.freedesktop.org/tegra/linux into drm-next

drm/tegra: Changes for v4.16-rc1

The bulk of these changes are preparation work and addition of support
for Tegra186. Currently only HDMI output (the primary output on Jetson
TX2) is supported, but the hardware is also capable of doing DSI and
DisplayPort.

Tegra DRM now also uses the atomic commit helpers instead of the open-
coded variant that was only doing half its job. As a bit of a byproduct
of the Tegra186 support the driver also gained HDMI 2.0 as well as zpos
property support.

Along the way there are also a few patches to clean up a few things and
fix minor issues.

* tag 'drm/tegra/for-4.16-rc1-fixes' of git://anongit.freedesktop.org/tegra/linux: (51 commits)
  drm/tegra: dc: Properly cleanup overlay planes
  drm/tegra: dc: Fix possible_crtcs mask for planes
  drm/tegra: dc: Restore YUV overlay support
  drm/tegra: dc: Implement legacy blending
  drm/tegra: Correct timeout in tegra_syncpt_wait
  drm/tegra: gem: Correct iommu_map_sg() error checking
  drm/tegra: dc: Link DC1 to DC0 on Tegra20
  drm/tegra: Fix non-debugfs builds
  drm/tegra: dpaux: Keep reset defaults for hybrid pad parameters
  drm/tegra: Mark Tegra186 display hub PM functions __maybe_unused
  drm/tegra: Use IOMMU groups
  gpu: host1x: Use IOMMU groups
  drm/tegra: Implement zpos property
  drm/tegra: dc: Remove redundant spinlock
  drm/tegra: dc: Use direct offset to plane registers
  drm/tegra: dc: Support more formats
  drm/tegra: fb: Force alpha formats
  drm/tegra: dpaux: Add Tegra186 support
  drm/tegra: dpaux: Implement runtime PM
  drm/tegra: sor: Support HDMI 2.0 modes
  ...

Documentation/devicetree/bindings/display/tegra/nvidia,tegra20-host1x.txt

@@ -206,21 +206,33 @@ of the following host1x client modules:
     - "nvidia,tegra132-sor": for Tegra132
     - "nvidia,tegra210-sor": for Tegra210
     - "nvidia,tegra210-sor1": for Tegra210
+    - "nvidia,tegra186-sor": for Tegra186
+    - "nvidia,tegra186-sor1": for Tegra186
   - reg: Physical base address and length of the controller's registers.
   - interrupts: The interrupt outputs from the controller.
   - clocks: Must contain an entry for each entry in clock-names.
     See ../clocks/clock-bindings.txt for details.
   - clock-names: Must include the following entries:
     - sor: clock input for the SOR hardware
-    - source: source clock for the SOR clock
+    - out: SOR output clock
     - parent: input for the pixel clock
     - dp: reference clock for the SOR clock
     - safe: safe reference for the SOR clock during power up
+
+    For Tegra186 and later:
+    - pad: SOR pad output clock (on Tegra186 and later)
+
+    Obsolete:
+    - source: source clock for the SOR clock (obsolete, use "out" instead)
+
   - resets: Must contain an entry for each entry in reset-names.
     See ../reset/reset.txt for details.
   - reset-names: Must include the following entries:
     - sor
 
+  Required properties on Tegra186 and later:
+  - nvidia,interface: index of the SOR interface
+
   Optional properties:
   - nvidia,ddc-i2c-bus: phandle of an I2C controller used for DDC EDID probing
   - nvidia,hpd-gpio: specifies a GPIO used for hotplug detection

drivers/gpu/drm/tegra/Makefile

@@ -5,6 +5,8 @@ tegra-drm-y := \
 	drm.o \
 	gem.o \
 	fb.o \
+	hub.o \
+	plane.o \
 	dc.o \
 	output.o \
 	rgb.o \

drivers/gpu/drm/tegra/dpaux.c

@@ -15,6 +15,7 @@
 #include <linux/pinctrl/pinconf-generic.h>
 #include <linux/pinctrl/pinctrl.h>
 #include <linux/pinctrl/pinmux.h>
+#include <linux/pm_runtime.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 #include <linux/regulator/consumer.h>
@@ -321,6 +322,9 @@ static int tegra_dpaux_pad_config(struct tegra_dpaux *dpaux, unsigned function)
 	case DPAUX_PADCTL_FUNC_I2C:
 		value = DPAUX_HYBRID_PADCTL_I2C_SDA_INPUT_RCV |
 			DPAUX_HYBRID_PADCTL_I2C_SCL_INPUT_RCV |
+			DPAUX_HYBRID_PADCTL_AUX_CMH(2) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVZ(4) |
+			DPAUX_HYBRID_PADCTL_AUX_DRVI(0x18) |
 			DPAUX_HYBRID_PADCTL_MODE_I2C;
 		break;
 
@@ -467,52 +471,37 @@ static int tegra_dpaux_probe(struct platform_device *pdev)
 		return PTR_ERR(dpaux->clk);
 	}
 
-	err = clk_prepare_enable(dpaux->clk);
-	if (err < 0) {
-		dev_err(&pdev->dev, "failed to enable module clock: %d\n",
-			err);
-		return err;
-	}
-
-	if (dpaux->rst)
-		reset_control_deassert(dpaux->rst);
-
 	dpaux->clk_parent = devm_clk_get(&pdev->dev, "parent");
 	if (IS_ERR(dpaux->clk_parent)) {
 		dev_err(&pdev->dev, "failed to get parent clock: %ld\n",
 			PTR_ERR(dpaux->clk_parent));
-		err = PTR_ERR(dpaux->clk_parent);
-		goto assert_reset;
-	}
-
-	err = clk_prepare_enable(dpaux->clk_parent);
-	if (err < 0) {
-		dev_err(&pdev->dev, "failed to enable parent clock: %d\n",
-			err);
-		goto assert_reset;
+		return PTR_ERR(dpaux->clk_parent);
 	}
 
 	err = clk_set_rate(dpaux->clk_parent, 270000000);
 	if (err < 0) {
 		dev_err(&pdev->dev, "failed to set clock to 270 MHz: %d\n",
 			err);
-		goto disable_parent_clk;
+		return err;
 	}
 
 	dpaux->vdd = devm_regulator_get(&pdev->dev, "vdd");
 	if (IS_ERR(dpaux->vdd)) {
 		dev_err(&pdev->dev, "failed to get VDD supply: %ld\n",
 			PTR_ERR(dpaux->vdd));
-		err = PTR_ERR(dpaux->vdd);
-		goto disable_parent_clk;
+		return PTR_ERR(dpaux->vdd);
 	}
 
+	platform_set_drvdata(pdev, dpaux);
+	pm_runtime_enable(&pdev->dev);
+	pm_runtime_get_sync(&pdev->dev);
+
 	err = devm_request_irq(dpaux->dev, dpaux->irq, tegra_dpaux_irq, 0,
 			       dev_name(dpaux->dev), dpaux);
 	if (err < 0) {
 		dev_err(dpaux->dev, "failed to request IRQ#%u: %d\n",
 			dpaux->irq, err);
-		goto disable_parent_clk;
+		return err;
 	}
 
 	disable_irq(dpaux->irq);
@@ -522,7 +511,7 @@ static int tegra_dpaux_probe(struct platform_device *pdev)
 
 	err = drm_dp_aux_register(&dpaux->aux);
 	if (err < 0)
-		goto disable_parent_clk;
+		return err;
 
 	/*
 	 * Assume that by default the DPAUX/I2C pads will be used for HDMI,
@@ -560,47 +549,97 @@ static int tegra_dpaux_probe(struct platform_device *pdev)
 	list_add_tail(&dpaux->list, &dpaux_list);
 	mutex_unlock(&dpaux_lock);
 
-	platform_set_drvdata(pdev, dpaux);
-
 	return 0;
-
-disable_parent_clk:
-	clk_disable_unprepare(dpaux->clk_parent);
-assert_reset:
-	if (dpaux->rst)
-		reset_control_assert(dpaux->rst);
-
-	clk_disable_unprepare(dpaux->clk);
-
-	return err;
 }
 
 static int tegra_dpaux_remove(struct platform_device *pdev)
 {
 	struct tegra_dpaux *dpaux = platform_get_drvdata(pdev);
 
+	cancel_work_sync(&dpaux->work);
+
 	/* make sure pads are powered down when not in use */
 	tegra_dpaux_pad_power_down(dpaux);
 
+	pm_runtime_put(&pdev->dev);
+	pm_runtime_disable(&pdev->dev);
+
 	drm_dp_aux_unregister(&dpaux->aux);
 
 	mutex_lock(&dpaux_lock);
 	list_del(&dpaux->list);
 	mutex_unlock(&dpaux_lock);
 
-	cancel_work_sync(&dpaux->work);
+	return 0;
+}
 
-	clk_disable_unprepare(dpaux->clk_parent);
+#ifdef CONFIG_PM
+static int tegra_dpaux_suspend(struct device *dev)
+{
+	struct tegra_dpaux *dpaux = dev_get_drvdata(dev);
+	int err = 0;
+
+	if (dpaux->rst) {
+		err = reset_control_assert(dpaux->rst);
+		if (err < 0) {
+			dev_err(dev, "failed to assert reset: %d\n", err);
+			return err;
+		}
+	}
 
-	if (dpaux->rst)
-		reset_control_assert(dpaux->rst);
+	usleep_range(1000, 2000);
 
+	clk_disable_unprepare(dpaux->clk_parent);
 	clk_disable_unprepare(dpaux->clk);
 
+	return err;
+}
+
+static int tegra_dpaux_resume(struct device *dev)
+{
+	struct tegra_dpaux *dpaux = dev_get_drvdata(dev);
+	int err;
+
+	err = clk_prepare_enable(dpaux->clk);
+	if (err < 0) {
+		dev_err(dev, "failed to enable clock: %d\n", err);
+		return err;
+	}
+
+	err = clk_prepare_enable(dpaux->clk_parent);
+	if (err < 0) {
+		dev_err(dev, "failed to enable parent clock: %d\n", err);
+		goto disable_clk;
+	}
+
+	usleep_range(1000, 2000);
+
+	if (dpaux->rst) {
+		err = reset_control_deassert(dpaux->rst);
+		if (err < 0) {
+			dev_err(dev, "failed to deassert reset: %d\n", err);
+			goto disable_parent;
+		}
+
+		usleep_range(1000, 2000);
+	}
+
 	return 0;
+
+disable_parent:
+	clk_disable_unprepare(dpaux->clk_parent);
+disable_clk:
+	clk_disable_unprepare(dpaux->clk);
+	return err;
 }
+#endif
+
+static const struct dev_pm_ops tegra_dpaux_pm_ops = {
+	SET_RUNTIME_PM_OPS(tegra_dpaux_suspend, tegra_dpaux_resume, NULL)
+};
 
 static const struct of_device_id tegra_dpaux_of_match[] = {
+	{ .compatible = "nvidia,tegra186-dpaux", },
 	{ .compatible = "nvidia,tegra210-dpaux", },
 	{ .compatible = "nvidia,tegra124-dpaux", },
 	{ },
@@ -611,6 +650,7 @@ struct platform_driver tegra_dpaux_driver = {
 	.driver = {
 		.name = "tegra-dpaux",
 		.of_match_table = tegra_dpaux_of_match,
+		.pm = &tegra_dpaux_pm_ops,
 	},
 	.probe = tegra_dpaux_probe,
 	.remove = tegra_dpaux_remove,

drivers/gpu/drm/tegra/drm.c

@@ -33,97 +33,91 @@ struct tegra_drm_file {
 	struct mutex lock;
 };
 
-static void tegra_atomic_schedule(struct tegra_drm *tegra,
+static int tegra_atomic_check(struct drm_device *drm,
 			      struct drm_atomic_state *state)
 {
-	tegra->commit.state = state;
-	schedule_work(&tegra->commit.work);
-}
+	int err;
 
-static void tegra_atomic_complete(struct tegra_drm *tegra,
-				  struct drm_atomic_state *state)
-{
-	struct drm_device *drm = tegra->drm;
+	err = drm_atomic_helper_check_modeset(drm, state);
+	if (err < 0)
+		return err;
 
-	/*
-	 * Everything below can be run asynchronously without the need to grab
-	 * any modeset locks at all under one condition: It must be guaranteed
-	 * that the asynchronous work has either been cancelled (if the driver
-	 * supports it, which at least requires that the framebuffers get
-	 * cleaned up with drm_atomic_helper_cleanup_planes()) or completed
-	 * before the new state gets committed on the software side with
-	 * drm_atomic_helper_swap_state().
-	 *
-	 * This scheme allows new atomic state updates to be prepared and
-	 * checked in parallel to the asynchronous completion of the previous
-	 * update. Which is important since compositors need to figure out the
-	 * composition of the next frame right after having submitted the
-	 * current layout.
-	 */
+	err = drm_atomic_normalize_zpos(drm, state);
+	if (err < 0)
+		return err;
 
-	drm_atomic_helper_commit_modeset_disables(drm, state);
-	drm_atomic_helper_commit_modeset_enables(drm, state);
-	drm_atomic_helper_commit_planes(drm, state,
-					DRM_PLANE_COMMIT_ACTIVE_ONLY);
+	err = drm_atomic_helper_check_planes(drm, state);
+	if (err < 0)
+		return err;
 
-	drm_atomic_helper_wait_for_vblanks(drm, state);
+	if (state->legacy_cursor_update)
+		state->async_update = !drm_atomic_helper_async_check(drm, state);
 
-	drm_atomic_helper_cleanup_planes(drm, state);
-	drm_atomic_state_put(state);
+	return 0;
 }
 
-static void tegra_atomic_work(struct work_struct *work)
+static struct drm_atomic_state *
+tegra_atomic_state_alloc(struct drm_device *drm)
 {
-	struct tegra_drm *tegra = container_of(work, struct tegra_drm,
-					       commit.work);
+	struct tegra_atomic_state *state = kzalloc(sizeof(*state), GFP_KERNEL);
+
+	if (!state || drm_atomic_state_init(drm, &state->base) < 0) {
+		kfree(state);
+		return NULL;
+	}
 
-	tegra_atomic_complete(tegra, tegra->commit.state);
+	return &state->base;
 }
 
-static int tegra_atomic_commit(struct drm_device *drm,
-			       struct drm_atomic_state *state, bool nonblock)
+static void tegra_atomic_state_clear(struct drm_atomic_state *state)
 {
-	struct tegra_drm *tegra = drm->dev_private;
-	int err;
+	struct tegra_atomic_state *tegra = to_tegra_atomic_state(state);
 
-	err = drm_atomic_helper_prepare_planes(drm, state);
-	if (err)
-		return err;
-
-	/* serialize outstanding nonblocking commits */
-	mutex_lock(&tegra->commit.lock);
-	flush_work(&tegra->commit.work);
-
-	/*
-	 * This is the point of no return - everything below never fails except
-	 * when the hw goes bonghits. Which means we can commit the new state on
-	 * the software side now.
-	 */
-
-	err = drm_atomic_helper_swap_state(state, true);
-	if (err) {
-		mutex_unlock(&tegra->commit.lock);
-		drm_atomic_helper_cleanup_planes(drm, state);
-		return err;
-	}
-
-	drm_atomic_state_get(state);
-	if (nonblock)
-		tegra_atomic_schedule(tegra, state);
-	else
-		tegra_atomic_complete(tegra, state);
+	drm_atomic_state_default_clear(state);
+	tegra->clk_disp = NULL;
+	tegra->dc = NULL;
+	tegra->rate = 0;
+}
 
-	mutex_unlock(&tegra->commit.lock);
-	return 0;
+static void tegra_atomic_state_free(struct drm_atomic_state *state)
+{
+	drm_atomic_state_default_release(state);
+	kfree(state);
 }
 
-static const struct drm_mode_config_funcs tegra_drm_mode_funcs = {
+static const struct drm_mode_config_funcs tegra_drm_mode_config_funcs = {
 	.fb_create = tegra_fb_create,
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 	.output_poll_changed = drm_fb_helper_output_poll_changed,
 #endif
-	.atomic_check = drm_atomic_helper_check,
-	.atomic_commit = tegra_atomic_commit,
+	.atomic_check = tegra_atomic_check,
+	.atomic_commit = drm_atomic_helper_commit,
+	.atomic_state_alloc = tegra_atomic_state_alloc,
+	.atomic_state_clear = tegra_atomic_state_clear,
+	.atomic_state_free = tegra_atomic_state_free,
+};
+
+static void tegra_atomic_commit_tail(struct drm_atomic_state *old_state)
+{
+	struct drm_device *drm = old_state->dev;
+	struct tegra_drm *tegra = drm->dev_private;
+
+	if (tegra->hub) {
+		drm_atomic_helper_commit_modeset_disables(drm, old_state);
+		tegra_display_hub_atomic_commit(drm, old_state);
+		drm_atomic_helper_commit_planes(drm, old_state, 0);
+		drm_atomic_helper_commit_modeset_enables(drm, old_state);
+		drm_atomic_helper_commit_hw_done(old_state);
+		drm_atomic_helper_wait_for_vblanks(drm, old_state);
+		drm_atomic_helper_cleanup_planes(drm, old_state);
+	} else {
+		drm_atomic_helper_commit_tail_rpm(old_state);
+	}
+}
+
+static const struct drm_mode_config_helper_funcs
+tegra_drm_mode_config_helpers = {
+	.atomic_commit_tail = tegra_atomic_commit_tail,
 };
 
 static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
@@ -172,9 +166,6 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 	mutex_init(&tegra->clients_lock);
 	INIT_LIST_HEAD(&tegra->clients);
 
-	mutex_init(&tegra->commit.lock);
-	INIT_WORK(&tegra->commit.work, tegra_atomic_work);
-
 	drm->dev_private = tegra;
 	tegra->drm = drm;
 
@@ -188,7 +179,8 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 
 	drm->mode_config.allow_fb_modifiers = true;
 
-	drm->mode_config.funcs = &tegra_drm_mode_funcs;
+	drm->mode_config.funcs = &tegra_drm_mode_config_funcs;
+	drm->mode_config.helper_private = &tegra_drm_mode_config_helpers;
 
 	err = tegra_drm_fb_prepare(drm);
 	if (err < 0)
@@ -200,6 +192,12 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 	if (err < 0)
 		goto fbdev;
 
+	if (tegra->hub) {
+		err = tegra_display_hub_prepare(tegra->hub);
+		if (err < 0)
+			goto device;
+	}
+
 	/*
 	 * We don't use the drm_irq_install() helpers provided by the DRM
 	 * core, so we need to set this manually in order to allow the
@@ -212,16 +210,19 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 
 	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
 	if (err < 0)
-		goto device;
+		goto hub;
 
 	drm_mode_config_reset(drm);
 
 	err = tegra_drm_fb_init(drm);
 	if (err < 0)
-		goto device;
+		goto hub;
 
 	return 0;
 
+hub:
+	if (tegra->hub)
+		tegra_display_hub_cleanup(tegra->hub);
 device:
 	host1x_device_exit(device);
 fbdev:
@@ -651,7 +652,8 @@ static int tegra_syncpt_wait(struct drm_device *drm, void *data,
 	if (!sp)
 		return -EINVAL;
 
-	return host1x_syncpt_wait(sp, args->thresh, args->timeout,
+	return host1x_syncpt_wait(sp, args->thresh,
+				  msecs_to_jiffies(args->timeout),
 				  &args->value);
 }
 
@@ -1139,8 +1141,7 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
-void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size,
-			      dma_addr_t *dma)
+void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *dma)
 {
 	struct iova *alloc;
 	void *virt;
@@ -1308,6 +1309,10 @@ static const struct of_device_id host1x_drm_subdevs[] = {
 	{ .compatible = "nvidia,tegra210-sor", },
 	{ .compatible = "nvidia,tegra210-sor1", },
 	{ .compatible = "nvidia,tegra210-vic", },
+	{ .compatible = "nvidia,tegra186-display", },
+	{ .compatible = "nvidia,tegra186-dc", },
+	{ .compatible = "nvidia,tegra186-sor", },
+	{ .compatible = "nvidia,tegra186-sor1", },
 	{ .compatible = "nvidia,tegra186-vic", },
 	{ /* sentinel */ }
 };
@@ -1323,6 +1328,7 @@ static struct host1x_driver host1x_drm_driver = {
 };
 
 static struct platform_driver * const drivers[] = {
+	&tegra_display_hub_driver,
 	&tegra_dc_driver,
 	&tegra_hdmi_driver,
 	&tegra_dsi_driver,

drivers/gpu/drm/tegra/drm.h

@@ -16,6 +16,7 @@
 #include <linux/of_gpio.h>
 
 #include <drm/drmP.h>
+#include <drm/drm_atomic.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_encoder.h>
@@ -23,6 +24,7 @@
 #include <drm/drm_fixed.h>
 
 #include "gem.h"
+#include "hub.h"
 #include "trace.h"
 
 struct reset_control;
@@ -40,10 +42,25 @@ struct tegra_fbdev {
 };
 #endif
 
+struct tegra_atomic_state {
+	struct drm_atomic_state base;
+
+	struct clk *clk_disp;
+	struct tegra_dc *dc;
+	unsigned long rate;
+};
+
+static inline struct tegra_atomic_state *
+to_tegra_atomic_state(struct drm_atomic_state *state)
+{
+	return container_of(state, struct tegra_atomic_state, base);
+}
+
 struct tegra_drm {
 	struct drm_device *drm;
 
 	struct iommu_domain *domain;
+	struct iommu_group *group;
 	struct mutex mm_lock;
 	struct drm_mm mm;
 
@@ -62,11 +79,7 @@ struct tegra_drm {
 
 	unsigned int pitch_align;
 
-	struct {
-		struct drm_atomic_state *state;
-		struct work_struct work;
-		struct mutex lock;
-	} commit;
+	struct tegra_display_hub *hub;
 
 	struct drm_atomic_state *state;
 };
@@ -152,6 +165,8 @@ int tegra_output_probe(struct tegra_output *output);
 void tegra_output_remove(struct tegra_output *output);
 int tegra_output_init(struct drm_device *drm, struct tegra_output *output);
 void tegra_output_exit(struct tegra_output *output);
+void tegra_output_find_possible_crtcs(struct tegra_output *output,
+				      struct drm_device *drm);
 
 int tegra_output_connector_get_modes(struct drm_connector *connector);
 enum drm_connector_status
@@ -189,6 +204,7 @@ void tegra_drm_fb_exit(struct drm_device *drm);
 void tegra_drm_fb_suspend(struct drm_device *drm);
 void tegra_drm_fb_resume(struct drm_device *drm);
 
+extern struct platform_driver tegra_display_hub_driver;
 extern struct platform_driver tegra_dc_driver;
 extern struct platform_driver tegra_hdmi_driver;
 extern struct platform_driver tegra_dsi_driver;

drivers/gpu/drm/tegra/dsi.c

@@ -65,8 +65,6 @@ struct tegra_dsi {
 	struct clk *clk;
 
 	struct drm_info_list *debugfs_files;
-	struct drm_minor *minor;
-	struct dentry *debugfs;
 
 	unsigned long flags;
 	enum mipi_dsi_pixel_format format;
@@ -122,12 +120,89 @@ static inline void tegra_dsi_writel(struct tegra_dsi *dsi, u32 value,
 	writel(value, dsi->regs + (offset << 2));
 }
 
+#define DEBUGFS_REG32(_name) { .name = #_name, .offset = _name }
+
+static const struct debugfs_reg32 tegra_dsi_regs[] = {
+	DEBUGFS_REG32(DSI_INCR_SYNCPT),
+	DEBUGFS_REG32(DSI_INCR_SYNCPT_CONTROL),
+	DEBUGFS_REG32(DSI_INCR_SYNCPT_ERROR),
+	DEBUGFS_REG32(DSI_CTXSW),
+	DEBUGFS_REG32(DSI_RD_DATA),
+	DEBUGFS_REG32(DSI_WR_DATA),
+	DEBUGFS_REG32(DSI_POWER_CONTROL),
+	DEBUGFS_REG32(DSI_INT_ENABLE),
+	DEBUGFS_REG32(DSI_INT_STATUS),
+	DEBUGFS_REG32(DSI_INT_MASK),
+	DEBUGFS_REG32(DSI_HOST_CONTROL),
+	DEBUGFS_REG32(DSI_CONTROL),
+	DEBUGFS_REG32(DSI_SOL_DELAY),
+	DEBUGFS_REG32(DSI_MAX_THRESHOLD),
+	DEBUGFS_REG32(DSI_TRIGGER),
+	DEBUGFS_REG32(DSI_TX_CRC),
+	DEBUGFS_REG32(DSI_STATUS),
+	DEBUGFS_REG32(DSI_INIT_SEQ_CONTROL),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_0),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_1),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_2),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_3),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_4),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_5),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_6),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_7),
+	DEBUGFS_REG32(DSI_PKT_SEQ_0_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_0_HI),
+	DEBUGFS_REG32(DSI_PKT_SEQ_1_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_1_HI),
+	DEBUGFS_REG32(DSI_PKT_SEQ_2_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_2_HI),
+	DEBUGFS_REG32(DSI_PKT_SEQ_3_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_3_HI),
+	DEBUGFS_REG32(DSI_PKT_SEQ_4_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_4_HI),
+	DEBUGFS_REG32(DSI_PKT_SEQ_5_LO),
+	DEBUGFS_REG32(DSI_PKT_SEQ_5_HI),
+	DEBUGFS_REG32(DSI_DCS_CMDS),
+	DEBUGFS_REG32(DSI_PKT_LEN_0_1),
+	DEBUGFS_REG32(DSI_PKT_LEN_2_3),
+	DEBUGFS_REG32(DSI_PKT_LEN_4_5),
+	DEBUGFS_REG32(DSI_PKT_LEN_6_7),
+	DEBUGFS_REG32(DSI_PHY_TIMING_0),
+	DEBUGFS_REG32(DSI_PHY_TIMING_1),
+	DEBUGFS_REG32(DSI_PHY_TIMING_2),
+	DEBUGFS_REG32(DSI_BTA_TIMING),
+	DEBUGFS_REG32(DSI_TIMEOUT_0),
+	DEBUGFS_REG32(DSI_TIMEOUT_1),
+	DEBUGFS_REG32(DSI_TO_TALLY),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_0),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_CD),
+	DEBUGFS_REG32(DSI_PAD_CD_STATUS),
+	DEBUGFS_REG32(DSI_VIDEO_MODE_CONTROL),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_1),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_2),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_3),
+	DEBUGFS_REG32(DSI_PAD_CONTROL_4),
+	DEBUGFS_REG32(DSI_GANGED_MODE_CONTROL),
+	DEBUGFS_REG32(DSI_GANGED_MODE_START),
+	DEBUGFS_REG32(DSI_GANGED_MODE_SIZE),
+	DEBUGFS_REG32(DSI_RAW_DATA_BYTE_COUNT),
+	DEBUGFS_REG32(DSI_ULTRA_LOW_POWER_CONTROL),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_8),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_9),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_10),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_11),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_12),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_13),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_14),
+	DEBUGFS_REG32(DSI_INIT_SEQ_DATA_15),
+};
+
 static int tegra_dsi_show_regs(struct seq_file *s, void *data)
 {
 	struct drm_info_node *node = s->private;
 	struct tegra_dsi *dsi = node->info_ent->data;
 	struct drm_crtc *crtc = dsi->output.encoder.crtc;
 	struct drm_device *drm = node->minor->dev;
+	unsigned int i;
 	int err = 0;
 
 	drm_modeset_lock_all(drm);
@@ -137,93 +212,12 @@ static int tegra_dsi_show_regs(struct seq_file *s, void *data)
 		goto unlock;
 	}
 
-#define DUMP_REG(name)						\
-	seq_printf(s, "%-32s %#05x %08x\n", #name, name,	\
-		   tegra_dsi_readl(dsi, name))
-
-	DUMP_REG(DSI_INCR_SYNCPT);
-	DUMP_REG(DSI_INCR_SYNCPT_CONTROL);
-	DUMP_REG(DSI_INCR_SYNCPT_ERROR);
-	DUMP_REG(DSI_CTXSW);
-	DUMP_REG(DSI_RD_DATA);
-	DUMP_REG(DSI_WR_DATA);
-	DUMP_REG(DSI_POWER_CONTROL);
-	DUMP_REG(DSI_INT_ENABLE);
-	DUMP_REG(DSI_INT_STATUS);
-	DUMP_REG(DSI_INT_MASK);
-	DUMP_REG(DSI_HOST_CONTROL);
-	DUMP_REG(DSI_CONTROL);
-	DUMP_REG(DSI_SOL_DELAY);
-	DUMP_REG(DSI_MAX_THRESHOLD);
-	DUMP_REG(DSI_TRIGGER);
-	DUMP_REG(DSI_TX_CRC);
-	DUMP_REG(DSI_STATUS);
-
-	DUMP_REG(DSI_INIT_SEQ_CONTROL);
-	DUMP_REG(DSI_INIT_SEQ_DATA_0);
-	DUMP_REG(DSI_INIT_SEQ_DATA_1);
-	DUMP_REG(DSI_INIT_SEQ_DATA_2);
-	DUMP_REG(DSI_INIT_SEQ_DATA_3);
-	DUMP_REG(DSI_INIT_SEQ_DATA_4);
-	DUMP_REG(DSI_INIT_SEQ_DATA_5);
-	DUMP_REG(DSI_INIT_SEQ_DATA_6);
-	DUMP_REG(DSI_INIT_SEQ_DATA_7);
-
-	DUMP_REG(DSI_PKT_SEQ_0_LO);
-	DUMP_REG(DSI_PKT_SEQ_0_HI);
-	DUMP_REG(DSI_PKT_SEQ_1_LO);
-	DUMP_REG(DSI_PKT_SEQ_1_HI);
-	DUMP_REG(DSI_PKT_SEQ_2_LO);
-	DUMP_REG(DSI_PKT_SEQ_2_HI);
-	DUMP_REG(DSI_PKT_SEQ_3_LO);
-	DUMP_REG(DSI_PKT_SEQ_3_HI);
-	DUMP_REG(DSI_PKT_SEQ_4_LO);
-	DUMP_REG(DSI_PKT_SEQ_4_HI);
-	DUMP_REG(DSI_PKT_SEQ_5_LO);
-	DUMP_REG(DSI_PKT_SEQ_5_HI);
-
-	DUMP_REG(DSI_DCS_CMDS);
-
-	DUMP_REG(DSI_PKT_LEN_0_1);
-	DUMP_REG(DSI_PKT_LEN_2_3);
-	DUMP_REG(DSI_PKT_LEN_4_5);
-	DUMP_REG(DSI_PKT_LEN_6_7);
-
-	DUMP_REG(DSI_PHY_TIMING_0);
-	DUMP_REG(DSI_PHY_TIMING_1);
-	DUMP_REG(DSI_PHY_TIMING_2);
-	DUMP_REG(DSI_BTA_TIMING);
-
-	DUMP_REG(DSI_TIMEOUT_0);
-	DUMP_REG(DSI_TIMEOUT_1);
-	DUMP_REG(DSI_TO_TALLY);
-
-	DUMP_REG(DSI_PAD_CONTROL_0);
-	DUMP_REG(DSI_PAD_CONTROL_CD);
-	DUMP_REG(DSI_PAD_CD_STATUS);
-	DUMP_REG(DSI_VIDEO_MODE_CONTROL);
-	DUMP_REG(DSI_PAD_CONTROL_1);
-	DUMP_REG(DSI_PAD_CONTROL_2);
-	DUMP_REG(DSI_PAD_CONTROL_3);
-	DUMP_REG(DSI_PAD_CONTROL_4);
-
-	DUMP_REG(DSI_GANGED_MODE_CONTROL);
-	DUMP_REG(DSI_GANGED_MODE_START);
-	DUMP_REG(DSI_GANGED_MODE_SIZE);
-
-	DUMP_REG(DSI_RAW_DATA_BYTE_COUNT);
-	DUMP_REG(DSI_ULTRA_LOW_POWER_CONTROL);
-
-	DUMP_REG(DSI_INIT_SEQ_DATA_8);
-	DUMP_REG(DSI_INIT_SEQ_DATA_9);
-	DUMP_REG(DSI_INIT_SEQ_DATA_10);
-	DUMP_REG(DSI_INIT_SEQ_DATA_11);
-	DUMP_REG(DSI_INIT_SEQ_DATA_12);
-	DUMP_REG(DSI_INIT_SEQ_DATA_13);
-	DUMP_REG(DSI_INIT_SEQ_DATA_14);
-	DUMP_REG(DSI_INIT_SEQ_DATA_15);
-
-#undef DUMP_REG
+	for (i = 0; i < ARRAY_SIZE(tegra_dsi_regs); i++) {
+		unsigned int offset = tegra_dsi_regs[i].offset;
+
+		seq_printf(s, "%-32s %#05x %08x\n", tegra_dsi_regs[i].name,
+			   offset, tegra_dsi_readl(dsi, offset));
+	}
 
 unlock:
 	drm_modeset_unlock_all(drm);
@@ -234,58 +228,46 @@ static struct drm_info_list debugfs_files[] = {
 	{ "regs", tegra_dsi_show_regs, 0, NULL },
 };
 
-static int tegra_dsi_debugfs_init(struct tegra_dsi *dsi,
-				  struct drm_minor *minor)
+static int tegra_dsi_late_register(struct drm_connector *connector)
 {
-	const char *name = dev_name(dsi->dev);
-	unsigned int i;
+	struct tegra_output *output = connector_to_output(connector);
+	unsigned int i, count = ARRAY_SIZE(debugfs_files);
+	struct drm_minor *minor = connector->dev->primary;
+	struct dentry *root = connector->debugfs_entry;
+	struct tegra_dsi *dsi = to_dsi(output);
 	int err;
 
-	dsi->debugfs = debugfs_create_dir(name, minor->debugfs_root);
-	if (!dsi->debugfs)
-		return -ENOMEM;
-
 	dsi->debugfs_files = kmemdup(debugfs_files, sizeof(debugfs_files),
 				     GFP_KERNEL);
-	if (!dsi->debugfs_files) {
-		err = -ENOMEM;
-		goto remove;
-	}
+	if (!dsi->debugfs_files)
+		return -ENOMEM;
 
-	for (i = 0; i < ARRAY_SIZE(debugfs_files); i++)
+	for (i = 0; i < count; i++)
 		dsi->debugfs_files[i].data = dsi;
 
-	err = drm_debugfs_create_files(dsi->debugfs_files,
-				       ARRAY_SIZE(debugfs_files),
-				       dsi->debugfs, minor);
+	err = drm_debugfs_create_files(dsi->debugfs_files, count, root, minor);
 	if (err < 0)
 		goto free;
 
-	dsi->minor = minor;
-
 	return 0;
 
 free:
 	kfree(dsi->debugfs_files);
 	dsi->debugfs_files = NULL;
-remove:
-	debugfs_remove(dsi->debugfs);
-	dsi->debugfs = NULL;
 
 	return err;
 }
 
-static void tegra_dsi_debugfs_exit(struct tegra_dsi *dsi)
+static void tegra_dsi_early_unregister(struct drm_connector *connector)
 {
-	drm_debugfs_remove_files(dsi->debugfs_files, ARRAY_SIZE(debugfs_files),
-				 dsi->minor);
-	dsi->minor = NULL;
+	struct tegra_output *output = connector_to_output(connector);
+	unsigned int count = ARRAY_SIZE(debugfs_files);
+	struct tegra_dsi *dsi = to_dsi(output);
 
+	drm_debugfs_remove_files(dsi->debugfs_files, count,
+				 connector->dev->primary);
 	kfree(dsi->debugfs_files);
 	dsi->debugfs_files = NULL;
-
-	debugfs_remove(dsi->debugfs);
-	dsi->debugfs = NULL;
 }
 
 #define PKT_ID0(id)	((((id) & 0x3f) <<  3) | (1 <<  9))
@@ -827,6 +809,8 @@ static const struct drm_connector_funcs tegra_dsi_connector_funcs = {
 	.destroy = tegra_output_connector_destroy,
 	.atomic_duplicate_state = tegra_dsi_connector_duplicate_state,
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+	.late_register = tegra_dsi_late_register,
+	.early_unregister = tegra_dsi_early_unregister,
 };
 
 static enum drm_mode_status
@@ -1080,12 +1064,6 @@ static int tegra_dsi_init(struct host1x_client *client)
 		dsi->output.encoder.possible_crtcs = 0x3;
 	}
 
-	if (IS_ENABLED(CONFIG_DEBUG_FS)) {
-		err = tegra_dsi_debugfs_init(dsi, drm->primary);
-		if (err < 0)
-			dev_err(dsi->dev, "debugfs setup failed: %d\n", err);
-	}
-
 	return 0;
 }
 
@@ -1094,10 +1072,6 @@ static int tegra_dsi_exit(struct host1x_client *client)
 	struct tegra_dsi *dsi = host1x_client_to_dsi(client);
 
 	tegra_output_exit(&dsi->output);
-
-	if (IS_ENABLED(CONFIG_DEBUG_FS))
-		tegra_dsi_debugfs_exit(dsi);
-
 	regulator_disable(dsi->vdd);
 
 	return 0;

drivers/gpu/drm/tegra/fb.c

@@ -54,17 +54,40 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 	struct tegra_fb *fb = to_tegra_fb(framebuffer);
 	uint64_t modifier = fb->base.modifier;
 
-	switch (fourcc_mod_tegra_mod(modifier)) {
-	case NV_FORMAT_MOD_TEGRA_TILED:
+	switch (modifier) {
+	case DRM_FORMAT_MOD_NVIDIA_TEGRA_TILED:
 		tiling->mode = TEGRA_BO_TILING_MODE_TILED;
 		tiling->value = 0;
 		break;
 
-	case NV_FORMAT_MOD_TEGRA_16BX2_BLOCK(0):
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(0):
 		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
-		tiling->value = fourcc_mod_tegra_param(modifier);
-		if (tiling->value > 5)
-			return -EINVAL;
+		tiling->value = 0;
+		break;
+
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(1):
+		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
+		tiling->value = 1;
+		break;
+
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(2):
+		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
+		tiling->value = 2;
+		break;
+
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(3):
+		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
+		tiling->value = 3;
+		break;
+
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(4):
+		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
+		tiling->value = 4;
+		break;
+
+	case DRM_FORMAT_MOD_NVIDIA_16BX2_BLOCK(5):
+		tiling->mode = TEGRA_BO_TILING_MODE_BLOCK;
+		tiling->value = 5;
 		break;
 
 	default:
@@ -230,6 +253,7 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 	cmd.height = sizes->surface_height;
 	cmd.pitches[0] = round_up(sizes->surface_width * bytes_per_pixel,
 				  tegra->pitch_align);
+
 	cmd.pixel_format = drm_mode_legacy_fb_format(sizes->surface_bpp,
 						     sizes->surface_depth);
 

drivers/gpu/drm/tegra/gem.c

@@ -114,7 +114,7 @@ static const struct host1x_bo_ops tegra_bo_ops = {
 static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo)
 {
 	int prot = IOMMU_READ | IOMMU_WRITE;
-	ssize_t err;
+	int err;
 
 	if (bo->mm)
 		return -EBUSY;
@@ -128,22 +128,21 @@ static int tegra_bo_iommu_map(struct tegra_drm *tegra, struct tegra_bo *bo)
 	err = drm_mm_insert_node_generic(&tegra->mm,
 					 bo->mm, bo->gem.size, PAGE_SIZE, 0, 0);
 	if (err < 0) {
-		dev_err(tegra->drm->dev, "out of I/O virtual memory: %zd\n",
+		dev_err(tegra->drm->dev, "out of I/O virtual memory: %d\n",
 			err);
 		goto unlock;
 	}
 
 	bo->paddr = bo->mm->start;
 
-	err = iommu_map_sg(tegra->domain, bo->paddr, bo->sgt->sgl,
+	bo->size = iommu_map_sg(tegra->domain, bo->paddr, bo->sgt->sgl,
 				bo->sgt->nents, prot);
-	if (err < 0) {
-		dev_err(tegra->drm->dev, "failed to map buffer: %zd\n", err);
+	if (!bo->size) {
+		dev_err(tegra->drm->dev, "failed to map buffer\n");
+		err = -ENOMEM;
 		goto remove;
 	}
 
-	bo->size = err;
-
 	mutex_unlock(&tegra->mm_lock);
 
 	return 0;

drivers/gpu/drm/tegra/hub.h

+/*
+ * Copyright (C) 2017 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef TEGRA_HUB_H
+#define TEGRA_HUB_H 1
+
+#include <drm/drmP.h>
+#include <drm/drm_plane.h>
+
+#include "plane.h"
+
+struct tegra_dc;
+
+struct tegra_windowgroup {
+	unsigned int usecount;
+	struct mutex lock;
+
+	unsigned int index;
+	struct device *parent;
+	struct reset_control *rst;
+};
+
+struct tegra_shared_plane {
+	struct tegra_plane base;
+	struct tegra_windowgroup *wgrp;
+};
+
+static inline struct tegra_shared_plane *
+to_tegra_shared_plane(struct drm_plane *plane)
+{
+	return container_of(plane, struct tegra_shared_plane, base.base);
+}
+
+struct tegra_display_hub_soc {
+	unsigned int num_wgrps;
+};
+
+struct tegra_display_hub {
+	struct host1x_client client;
+	struct clk *clk_disp;
+	struct clk *clk_dsc;
+	struct clk *clk_hub;
+	struct reset_control *rst;
+
+	const struct tegra_display_hub_soc *soc;
+	struct tegra_windowgroup *wgrps;
+};
+
+static inline struct tegra_display_hub *
+to_tegra_display_hub(struct host1x_client *client)
+{
+	return container_of(client, struct tegra_display_hub, client);
+}
+
+struct tegra_dc;
+struct tegra_plane;
+
+int tegra_display_hub_prepare(struct tegra_display_hub *hub);
+void tegra_display_hub_cleanup(struct tegra_display_hub *hub);
+
+struct drm_plane *tegra_shared_plane_create(struct drm_device *drm,
+					    struct tegra_dc *dc,
+					    unsigned int wgrp,
+					    unsigned int index);
+
+void tegra_display_hub_atomic_commit(struct drm_device *drm,
+				     struct drm_atomic_state *state);
+
+#define DC_CMD_IHUB_COMMON_MISC_CTL 0x068
+#define  LATENCY_EVENT (1 << 3)
+
+#define DC_DISP_IHUB_COMMON_DISPLAY_FETCH_METER 0x451
+#define  CURS_SLOTS(x) (((x) & 0xff) << 8)
+#define  WGRP_SLOTS(x) (((x) & 0xff) << 0)
+
+#endif /* TEGRA_HUB_H */

drivers/gpu/drm/tegra/output.c

@@ -9,7 +9,9 @@
 
 #include <drm/drm_atomic_helper.h>
 #include <drm/drm_panel.h>
+
 #include "drm.h"
+#include "dc.h"
 
 #include <media/cec-notifier.h>
 
@@ -218,3 +220,25 @@ void tegra_output_exit(struct tegra_output *output)
 	if (output->panel)
 		drm_panel_detach(output->panel);
 }
+
+void tegra_output_find_possible_crtcs(struct tegra_output *output,
+				      struct drm_device *drm)
+{
+	struct device *dev = output->dev;
+	struct drm_crtc *crtc;
+	unsigned int mask = 0;
+
+	drm_for_each_crtc(crtc, drm) {
+		struct tegra_dc *dc = to_tegra_dc(crtc);
+
+		if (tegra_dc_has_output(dc, dev))
+			mask |= drm_crtc_mask(crtc);
+	}
+
+	if (mask == 0) {
+		dev_warn(dev, "missing output definition for heads in DT\n");
+		mask = 0x3;
+	}
+
+	output->encoder.possible_crtcs = mask;
+}

drivers/gpu/drm/tegra/plane.c

+/*
+ * Copyright (C) 2017 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+
+#include "dc.h"
+#include "plane.h"
+
+static void tegra_plane_destroy(struct drm_plane *plane)
+{
+	struct tegra_plane *p = to_tegra_plane(plane);
+
+	drm_plane_cleanup(plane);
+	kfree(p);
+}
+
+static void tegra_plane_reset(struct drm_plane *plane)
+{
+	struct tegra_plane_state *state;
+
+	if (plane->state)
+		__drm_atomic_helper_plane_destroy_state(plane->state);
+
+	kfree(plane->state);
+	plane->state = NULL;
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (state) {
+		plane->state = &state->base;
+		plane->state->plane = plane;
+	}
+}
+
+static struct drm_plane_state *
+tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
+{
+	struct tegra_plane_state *state = to_tegra_plane_state(plane->state);
+	struct tegra_plane_state *copy;
+	unsigned int i;
+
+	copy = kmalloc(sizeof(*copy), GFP_KERNEL);
+	if (!copy)
+		return NULL;
+
+	__drm_atomic_helper_plane_duplicate_state(plane, &copy->base);
+	copy->tiling = state->tiling;
+	copy->format = state->format;
+	copy->swap = state->swap;
+	copy->opaque = state->opaque;
+
+	for (i = 0; i < 3; i++)
+		copy->dependent[i] = state->dependent[i];
+
+	return &copy->base;
+}
+
+static void tegra_plane_atomic_destroy_state(struct drm_plane *plane,
+					     struct drm_plane_state *state)
+{
+	__drm_atomic_helper_plane_destroy_state(state);
+	kfree(state);
+}
+
+const struct drm_plane_funcs tegra_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = tegra_plane_destroy,
+	.reset = tegra_plane_reset,
+	.atomic_duplicate_state = tegra_plane_atomic_duplicate_state,
+	.atomic_destroy_state = tegra_plane_atomic_destroy_state,
+};
+
+int tegra_plane_state_add(struct tegra_plane *plane,
+			  struct drm_plane_state *state)
+{
+	struct drm_crtc_state *crtc_state;
+	struct tegra_dc_state *tegra;
+	struct drm_rect clip;
+	int err;
+
+	/* Propagate errors from allocation or locking failures. */
+	crtc_state = drm_atomic_get_crtc_state(state->state, state->crtc);
+	if (IS_ERR(crtc_state))
+		return PTR_ERR(crtc_state);
+
+	clip.x1 = 0;
+	clip.y1 = 0;
+	clip.x2 = crtc_state->mode.hdisplay;
+	clip.y2 = crtc_state->mode.vdisplay;
+
+	/* Check plane state for visibility and calculate clipping bounds */
+	err = drm_atomic_helper_check_plane_state(state, crtc_state, &clip,
+						  0, INT_MAX, true, true);
+	if (err < 0)
+		return err;
+
+	tegra = to_dc_state(crtc_state);
+
+	tegra->planes |= WIN_A_ACT_REQ << plane->index;
+
+	return 0;
+}
+
+int tegra_plane_format(u32 fourcc, u32 *format, u32 *swap)
+{
+	/* assume no swapping of fetched data */
+	if (swap)
+		*swap = BYTE_SWAP_NOSWAP;
+
+	switch (fourcc) {
+	case DRM_FORMAT_ARGB4444:
+		*format = WIN_COLOR_DEPTH_B4G4R4A4;
+		break;
+
+	case DRM_FORMAT_ARGB1555:
+		*format = WIN_COLOR_DEPTH_B5G5R5A1;
+		break;
+
+	case DRM_FORMAT_RGB565:
+		*format = WIN_COLOR_DEPTH_B5G6R5;
+		break;
+
+	case DRM_FORMAT_RGBA5551:
+		*format = WIN_COLOR_DEPTH_A1B5G5R5;
+		break;
+
+	case DRM_FORMAT_ARGB8888:
+		*format = WIN_COLOR_DEPTH_B8G8R8A8;
+		break;
+
+	case DRM_FORMAT_ABGR8888:
+		*format = WIN_COLOR_DEPTH_R8G8B8A8;
+		break;
+
+	case DRM_FORMAT_ABGR4444:
+		*format = WIN_COLOR_DEPTH_R4G4B4A4;
+		break;
+
+	case DRM_FORMAT_ABGR1555:
+		*format = WIN_COLOR_DEPTH_R5G5B5A;
+		break;
+
+	case DRM_FORMAT_BGRA5551:
+		*format = WIN_COLOR_DEPTH_AR5G5B5;
+		break;
+
+	case DRM_FORMAT_XRGB1555:
+		*format = WIN_COLOR_DEPTH_B5G5R5X1;
+		break;
+
+	case DRM_FORMAT_RGBX5551:
+		*format = WIN_COLOR_DEPTH_X1B5G5R5;
+		break;
+
+	case DRM_FORMAT_XBGR1555:
+		*format = WIN_COLOR_DEPTH_R5G5B5X1;
+		break;
+
+	case DRM_FORMAT_BGRX5551:
+		*format = WIN_COLOR_DEPTH_X1R5G5B5;
+		break;
+
+	case DRM_FORMAT_BGR565:
+		*format = WIN_COLOR_DEPTH_R5G6B5;
+		break;
+
+	case DRM_FORMAT_BGRA8888:
+		*format = WIN_COLOR_DEPTH_A8R8G8B8;
+		break;
+
+	case DRM_FORMAT_RGBA8888:
+		*format = WIN_COLOR_DEPTH_A8B8G8R8;
+		break;
+
+	case DRM_FORMAT_XRGB8888:
+		*format = WIN_COLOR_DEPTH_B8G8R8X8;
+		break;
+
+	case DRM_FORMAT_XBGR8888:
+		*format = WIN_COLOR_DEPTH_R8G8B8X8;
+		break;
+
+	case DRM_FORMAT_UYVY:
+		*format = WIN_COLOR_DEPTH_YCbCr422;
+		break;
+
+	case DRM_FORMAT_YUYV:
+		if (!swap)
+			return -EINVAL;
+
+		*format = WIN_COLOR_DEPTH_YCbCr422;
+		*swap = BYTE_SWAP_SWAP2;
+		break;
+
+	case DRM_FORMAT_YUV420:
+		*format = WIN_COLOR_DEPTH_YCbCr420P;
+		break;
+
+	case DRM_FORMAT_YUV422:
+		*format = WIN_COLOR_DEPTH_YCbCr422P;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+bool tegra_plane_format_is_yuv(unsigned int format, bool *planar)
+{
+	switch (format) {
+	case WIN_COLOR_DEPTH_YCbCr422:
+	case WIN_COLOR_DEPTH_YUV422:
+		if (planar)
+			*planar = false;
+
+		return true;
+
+	case WIN_COLOR_DEPTH_YCbCr420P:
+	case WIN_COLOR_DEPTH_YUV420P:
+	case WIN_COLOR_DEPTH_YCbCr422P:
+	case WIN_COLOR_DEPTH_YUV422P:
+	case WIN_COLOR_DEPTH_YCbCr422R:
+	case WIN_COLOR_DEPTH_YUV422R:
+	case WIN_COLOR_DEPTH_YCbCr422RA:
+	case WIN_COLOR_DEPTH_YUV422RA:
+		if (planar)
+			*planar = true;
+
+		return true;
+	}
+
+	if (planar)
+		*planar = false;
+
+	return false;
+}
+
+static bool __drm_format_has_alpha(u32 format)
+{
+	switch (format) {
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_RGBA5551:
+	case DRM_FORMAT_ABGR8888:
+	case DRM_FORMAT_ARGB8888:
+		return true;
+	}
+
+	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)
+{
+	if (tegra_plane_format_is_yuv(opaque, NULL)) {
+		*alpha = opaque;
+		return 0;
+	}
+
+	switch (opaque) {
+	case WIN_COLOR_DEPTH_B5G5R5X1:
+		*alpha = WIN_COLOR_DEPTH_B5G5R5A1;
+		return 0;
+
+	case WIN_COLOR_DEPTH_X1B5G5R5:
+		*alpha = WIN_COLOR_DEPTH_A1B5G5R5;
+		return 0;
+
+	case WIN_COLOR_DEPTH_R8G8B8X8:
+		*alpha = WIN_COLOR_DEPTH_R8G8B8A8;
+		return 0;
+
+	case WIN_COLOR_DEPTH_B8G8R8X8:
+		*alpha = WIN_COLOR_DEPTH_B8G8R8A8;
+		return 0;
+	}
+
+	return -EINVAL;
+}
+
+unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
+					   struct tegra_plane *other)
+{
+	unsigned int index = 0, i;
+
+	WARN_ON(plane == other);
+
+	for (i = 0; i < 3; i++) {
+		if (i == plane->index)
+			continue;
+
+		if (i == other->index)
+			break;
+
+		index++;
+	}
+
+	return index;
+}
+
+void tegra_plane_check_dependent(struct tegra_plane *tegra,
+				 struct tegra_plane_state *state)
+{
+	struct drm_plane_state *old, *new;
+	struct drm_plane *plane;
+	unsigned int zpos[2];
+	unsigned int i;
+
+	for (i = 0; i < 3; i++)
+		state->dependent[i] = false;
+
+	for (i = 0; i < 2; i++)
+		zpos[i] = 0;
+
+	for_each_oldnew_plane_in_state(state->base.state, plane, old, 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)
+			continue;
+
+		index = tegra_plane_get_overlap_index(tegra, p);
+
+		/*
+		 * 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.
+		 */
+		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;
+		}
+	}
+
+	/*
+	 * 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.
+	 */
+	for (i = 0; i < 2; i++)
+		state->dependent[2] = state->dependent[2] ||
+				      state->dependent[i];
+
+	/*
+	 * 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.
+	 */
+	for (i = 0; i < 2; i++) {
+		if (zpos[i] > tegra->index)
+			state->dependent[2] = state->dependent[2] &&
+					      state->dependent[i];
+	}
+}

drivers/gpu/drm/tegra/plane.h

+/*
+ * Copyright (C) 2017 NVIDIA CORPORATION.  All rights reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef TEGRA_PLANE_H
+#define TEGRA_PLANE_H 1
+
+#include <drm/drm_plane.h>
+
+struct tegra_bo;
+struct tegra_dc;
+
+struct tegra_plane {
+	struct drm_plane base;
+	struct tegra_dc *dc;
+	unsigned int offset;
+	unsigned int index;
+};
+
+struct tegra_cursor {
+	struct tegra_plane base;
+
+	struct tegra_bo *bo;
+	unsigned int width;
+	unsigned int height;
+};
+
+static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
+{
+	return container_of(plane, struct tegra_plane, base);
+}
+
+struct tegra_plane_state {
+	struct drm_plane_state base;
+
+	struct tegra_bo_tiling tiling;
+	u32 format;
+	u32 swap;
+
+	/* used for legacy blending support only */
+	bool opaque;
+	bool dependent[3];
+};
+
+static inline struct tegra_plane_state *
+to_tegra_plane_state(struct drm_plane_state *state)
+{
+	if (state)
+		return container_of(state, struct tegra_plane_state, base);
+
+	return NULL;
+}
+
+extern const struct drm_plane_funcs tegra_plane_funcs;
+
+int tegra_plane_state_add(struct tegra_plane *plane,
+			  struct drm_plane_state *state);
+
+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);
+
+#endif /* TEGRA_PLANE_H */

drivers/gpu/drm/tegra/sor.h

@@ -89,6 +89,8 @@
 #define SOR_PLL0 0x17
 #define  SOR_PLL0_ICHPMP_MASK			(0xf << 24)
 #define  SOR_PLL0_ICHPMP(x)			(((x) & 0xf) << 24)
+#define  SOR_PLL0_FILTER_MASK			(0xf << 16)
+#define  SOR_PLL0_FILTER(x)			(((x) & 0xf) << 16)
 #define  SOR_PLL0_VCOCAP_MASK			(0xf << 8)
 #define  SOR_PLL0_VCOCAP(x)			(((x) & 0xf) << 8)
 #define  SOR_PLL0_VCOCAP_RST			SOR_PLL0_VCOCAP(3)
@@ -122,10 +124,16 @@
 #define  SOR_PLL2_SEQ_PLL_PULLDOWN		(1 << 16)
 
 #define SOR_PLL3 0x1a
+#define  SOR_PLL3_BG_TEMP_COEF_MASK		(0xf << 28)
+#define  SOR_PLL3_BG_TEMP_COEF(x)		(((x) & 0xf) << 28)
 #define  SOR_PLL3_BG_VREF_LEVEL_MASK		(0xf << 24)
 #define  SOR_PLL3_BG_VREF_LEVEL(x)		(((x) & 0xf) << 24)
 #define  SOR_PLL3_PLL_VDD_MODE_1V8		(0 << 13)
 #define  SOR_PLL3_PLL_VDD_MODE_3V3		(1 << 13)
+#define  SOR_PLL3_AVDD10_LEVEL_MASK		(0xf << 8)
+#define  SOR_PLL3_AVDD10_LEVEL(x)		(((x) & 0xf) << 8)
+#define  SOR_PLL3_AVDD14_LEVEL_MASK		(0xf << 4)
+#define  SOR_PLL3_AVDD14_LEVEL(x)		(((x) & 0xf) << 4)
 
 #define SOR_CSTM 0x1b
 #define  SOR_CSTM_ROTCLK_MASK			(0xf << 24)
@@ -334,6 +342,10 @@
 #define SOR_DP_LQ_CSTM1 0x70
 #define SOR_DP_LQ_CSTM2 0x71
 
+#define SOR_DP_PADCTL2 0x73
+#define  SOR_DP_PADCTL_SPAREPLL_MASK (0xff << 24)
+#define  SOR_DP_PADCTL_SPAREPLL(x) (((x) & 0xff) << 24)
+
 #define SOR_HDMI_AUDIO_INFOFRAME_CTRL 0x9a
 #define SOR_HDMI_AUDIO_INFOFRAME_STATUS 0x9b
 #define SOR_HDMI_AUDIO_INFOFRAME_HEADER 0x9c
@@ -370,4 +382,8 @@
 #define SOR_HDMI_VSI_INFOFRAME_STATUS 0x124
 #define SOR_HDMI_VSI_INFOFRAME_HEADER 0x125
 
+#define SOR_HDMI2_CTRL 0x13e
+#define  SOR_HDMI2_CTRL_CLOCK_MODE_DIV_BY_4 (1 << 1)
+#define  SOR_HDMI2_CTRL_SCRAMBLE (1 << 0)
+
 #endif

drivers/gpu/drm/tegra/vic.c

@@ -138,13 +138,14 @@ static const struct falcon_ops vic_falcon_ops = {
 static int vic_init(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
+	struct iommu_group *group = iommu_group_get(client->dev);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = dev->dev_private;
 	struct vic *vic = to_vic(drm);
 	int err;
 
-	if (tegra->domain) {
-		err = iommu_attach_device(tegra->domain, vic->dev);
+	if (group && tegra->domain) {
+		err = iommu_attach_group(tegra->domain, group);
 		if (err < 0) {
 			dev_err(vic->dev, "failed to attach to domain: %d\n",
 				err);
@@ -158,13 +159,13 @@ static int vic_init(struct host1x_client *client)
 		vic->falcon.data = tegra;
 		err = falcon_load_firmware(&vic->falcon);
 		if (err < 0)
-			goto detach_device;
+			goto detach;
 	}
 
 	vic->channel = host1x_channel_request(client->dev);
 	if (!vic->channel) {
 		err = -ENOMEM;
-		goto detach_device;
+		goto detach;
 	}
 
 	client->syncpts[0] = host1x_syncpt_request(client, 0);
@@ -183,9 +184,9 @@ static int vic_init(struct host1x_client *client)
 	host1x_syncpt_free(client->syncpts[0]);
 free_channel:
 	host1x_channel_put(vic->channel);
-detach_device:
-	if (tegra->domain)
-		iommu_detach_device(tegra->domain, vic->dev);
+detach:
+	if (group && tegra->domain)
+		iommu_detach_group(tegra->domain, group);
 
 	return err;
 }
@@ -193,6 +194,7 @@ static int vic_init(struct host1x_client *client)
 static int vic_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
+	struct iommu_group *group = iommu_group_get(client->dev);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	struct tegra_drm *tegra = dev->dev_private;
 	struct vic *vic = to_vic(drm);
@@ -206,7 +208,7 @@ static int vic_exit(struct host1x_client *client)
 	host1x_channel_put(vic->channel);
 
 	if (vic->domain) {
-		iommu_detach_device(vic->domain, vic->dev);
+		iommu_detach_group(vic->domain, group);
 		vic->domain = NULL;
 	}
 

drivers/gpu/host1x/bus.c

@@ -211,8 +211,7 @@ int host1x_device_init(struct host1x_device *device)
 				dev_err(&device->dev,
 					"failed to initialize %s: %d\n",
 					dev_name(client->dev), err);
-				mutex_unlock(&device->clients_lock);
-				return err;
+				goto teardown;
 			}
 		}
 	}
@@ -220,6 +219,14 @@ int host1x_device_init(struct host1x_device *device)
 	mutex_unlock(&device->clients_lock);
 
 	return 0;
+
+teardown:
+	list_for_each_entry_continue_reverse(client, &device->clients, list)
+		if (client->ops->exit)
+			client->ops->exit(client);
+
+	mutex_unlock(&device->clients_lock);
+	return err;
 }
 EXPORT_SYMBOL(host1x_device_init);
 

drivers/gpu/host1x/dev.h

@@ -117,6 +117,7 @@ struct host1x {
 	struct clk *clk;
 	struct reset_control *rst;
 
+	struct iommu_group *group;
 	struct iommu_domain *domain;
 	struct iova_domain iova;
 	dma_addr_t iova_end;