1 files changed, 410 insertions, 281 deletions

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 4b70ce664c41..1d18d43292dc 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -1,42 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0-only
 /*
  * Copyright (C) 2012 Avionic Design GmbH
  * Copyright (C) 2012-2016 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 <linux/aperture.h>
 #include <linux/bitops.h>
 #include <linux/host1x.h>
 #include <linux/idr.h>
 #include <linux/iommu.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
 
+#include <drm/clients/drm_client_setup.h>
 #include <drm/drm_atomic.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_debugfs.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_framebuffer.h>
+#include <drm/drm_ioctl.h>
+#include <drm/drm_prime.h>
+#include <drm/drm_print.h>
+#include <drm/drm_vblank.h>
 
 #if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
 #include <asm/dma-iommu.h>
 #endif
 
+#include "dc.h"
 #include "drm.h"
 #include "gem.h"
+#include "uapi.h"
 
 #define DRIVER_NAME "tegra"
 #define DRIVER_DESC "NVIDIA Tegra graphics"
-#define DRIVER_DATE "20120330"
-#define DRIVER_MAJOR 0
+#define DRIVER_MAJOR 1
 #define DRIVER_MINOR 0
 #define DRIVER_PATCHLEVEL 0
 
 #define CARVEOUT_SZ SZ_64M
 #define CDMA_GATHER_FETCHES_MAX_NB 16383
 
-struct tegra_drm_file {
-	struct idr contexts;
-	struct mutex lock;
-};
-
 static int tegra_atomic_check(struct drm_device *drm,
 			      struct drm_atomic_state *state)
 {
@@ -51,29 +57,42 @@ static int tegra_atomic_check(struct drm_device *drm,
 
 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 = tegra_atomic_check,
 	.atomic_commit = drm_atomic_helper_commit,
 };
 
+static void tegra_atomic_post_commit(struct drm_device *drm,
+				     struct drm_atomic_state *old_state)
+{
+	struct drm_crtc_state *old_crtc_state __maybe_unused;
+	struct drm_crtc *crtc;
+	unsigned int i;
+
+	for_each_old_crtc_in_state(old_state, crtc, old_crtc_state, i)
+		tegra_crtc_atomic_post_commit(crtc, old_state);
+}
+
 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) {
+		bool fence_cookie = dma_fence_begin_signalling();
+
 		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);
+		dma_fence_end_signalling(fence_cookie);
 		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);
 	}
+
+	tegra_atomic_post_commit(drm, old_state);
 }
 
 static const struct drm_mode_config_helper_funcs
@@ -81,163 +100,6 @@ tegra_drm_mode_config_helpers = {
 	.atomic_commit_tail = tegra_atomic_commit_tail,
 };
 
-static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
-{
-	struct host1x_device *device = to_host1x_device(drm->dev);
-	struct tegra_drm *tegra;
-	int err;
-
-	tegra = kzalloc(sizeof(*tegra), GFP_KERNEL);
-	if (!tegra)
-		return -ENOMEM;
-
-	if (iommu_present(&platform_bus_type)) {
-		u64 carveout_start, carveout_end, gem_start, gem_end;
-		struct iommu_domain_geometry *geometry;
-		unsigned long order;
-
-		tegra->domain = iommu_domain_alloc(&platform_bus_type);
-		if (!tegra->domain) {
-			err = -ENOMEM;
-			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;
-		carveout_start = gem_end + 1;
-		carveout_end = geometry->aperture_end;
-
-		order = __ffs(tegra->domain->pgsize_bitmap);
-		init_iova_domain(&tegra->carveout.domain, 1UL << order,
-				 carveout_start >> order);
-
-		tegra->carveout.shift = iova_shift(&tegra->carveout.domain);
-		tegra->carveout.limit = carveout_end >> tegra->carveout.shift;
-
-		drm_mm_init(&tegra->mm, gem_start, gem_end - gem_start + 1);
-		mutex_init(&tegra->mm_lock);
-
-		DRM_DEBUG("IOMMU apertures:\n");
-		DRM_DEBUG("  GEM: %#llx-%#llx\n", gem_start, gem_end);
-		DRM_DEBUG("  Carveout: %#llx-%#llx\n", carveout_start,
-			  carveout_end);
-	}
-
-	mutex_init(&tegra->clients_lock);
-	INIT_LIST_HEAD(&tegra->clients);
-
-	drm->dev_private = tegra;
-	tegra->drm = drm;
-
-	drm_mode_config_init(drm);
-
-	drm->mode_config.min_width = 0;
-	drm->mode_config.min_height = 0;
-
-	drm->mode_config.max_width = 4096;
-	drm->mode_config.max_height = 4096;
-
-	drm->mode_config.allow_fb_modifiers = true;
-
-	drm->mode_config.normalize_zpos = true;
-
-	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)
-		goto config;
-
-	drm_kms_helper_poll_init(drm);
-
-	err = host1x_device_init(device);
-	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
-	 * DRM_IOCTL_WAIT_VBLANK to operate correctly.
-	 */
-	drm->irq_enabled = true;
-
-	/* syncpoints are used for full 32-bit hardware VBLANK counters */
-	drm->max_vblank_count = 0xffffffff;
-
-	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
-	if (err < 0)
-		goto hub;
-
-	drm_mode_config_reset(drm);
-
-	err = tegra_drm_fb_init(drm);
-	if (err < 0)
-		goto hub;
-
-	return 0;
-
-hub:
-	if (tegra->hub)
-		tegra_display_hub_cleanup(tegra->hub);
-device:
-	host1x_device_exit(device);
-fbdev:
-	drm_kms_helper_poll_fini(drm);
-	tegra_drm_fb_free(drm);
-config:
-	drm_mode_config_cleanup(drm);
-
-	if (tegra->domain) {
-		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;
-}
-
-static void tegra_drm_unload(struct drm_device *drm)
-{
-	struct host1x_device *device = to_host1x_device(drm->dev);
-	struct tegra_drm *tegra = drm->dev_private;
-	int err;
-
-	drm_kms_helper_poll_fini(drm);
-	tegra_drm_fb_exit(drm);
-	drm_atomic_helper_shutdown(drm);
-	drm_mode_config_cleanup(drm);
-
-	err = host1x_device_exit(device);
-	if (err < 0)
-		return;
-
-	if (tegra->domain) {
-		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);
-}
-
 static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp)
 {
 	struct tegra_drm_file *fpriv;
@@ -246,7 +108,9 @@ static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp)
 	if (!fpriv)
 		return -ENOMEM;
 
-	idr_init(&fpriv->contexts);
+	idr_init_base(&fpriv->legacy_contexts, 1);
+	xa_init_flags(&fpriv->contexts, XA_FLAGS_ALLOC1);
+	xa_init(&fpriv->syncpoints);
 	mutex_init(&fpriv->lock);
 	filp->driver_priv = fpriv;
 
@@ -256,23 +120,10 @@ static int tegra_drm_open(struct drm_device *drm, struct drm_file *filp)
 static void tegra_drm_context_free(struct tegra_drm_context *context)
 {
 	context->client->ops->close_channel(context);
+	pm_runtime_put(context->client->base.dev);
 	kfree(context);
 }
 
-static struct host1x_bo *
-host1x_bo_lookup(struct drm_file *file, u32 handle)
-{
-	struct drm_gem_object *gem;
-	struct tegra_bo *bo;
-
-	gem = drm_gem_object_lookup(file, handle);
-	if (!gem)
-		return NULL;
-
-	bo = to_tegra_bo(gem);
-	return &bo->base;
-}
-
 static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 				       struct drm_tegra_reloc __user *src,
 				       struct drm_device *drm,
@@ -301,11 +152,13 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 	if (err < 0)
 		return err;
 
-	dest->cmdbuf.bo = host1x_bo_lookup(file, cmdbuf);
+	dest->flags = HOST1X_RELOC_READ | HOST1X_RELOC_WRITE;
+
+	dest->cmdbuf.bo = tegra_gem_lookup(file, cmdbuf);
 	if (!dest->cmdbuf.bo)
 		return -ENOENT;
 
-	dest->target.bo = host1x_bo_lookup(file, target);
+	dest->target.bo = tegra_gem_lookup(file, target);
 	if (!dest->target.bo)
 		return -ENOENT;
 
@@ -325,7 +178,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 	struct drm_tegra_syncpt syncpt;
 	struct host1x *host1x = dev_get_drvdata(drm->dev->parent);
 	struct drm_gem_object **refs;
-	struct host1x_syncpt *sp;
+	struct host1x_syncpt *sp = NULL;
 	struct host1x_job *job;
 	unsigned int num_refs;
 	int err;
@@ -343,7 +196,7 @@ 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_relocs, false);
 	if (!job)
 		return -ENOMEM;
 
@@ -351,6 +204,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 	job->client = client;
 	job->class = client->class;
 	job->serialize = true;
+	job->syncpt_recovery = true;
 
 	/*
 	 * Track referenced BOs so that they can be unreferenced after the
@@ -387,7 +241,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 			goto fail;
 		}
 
-		bo = host1x_bo_lookup(file, cmdbuf.handle);
+		bo = tegra_gem_lookup(file, cmdbuf.handle);
 		if (!bo) {
 			err = -ENOENT;
 			goto fail;
@@ -452,8 +306,8 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		goto fail;
 	}
 
-	/* check whether syncpoint ID is valid */
-	sp = host1x_syncpt_get(host1x, syncpt.id);
+	/* Syncpoint ref will be dropped on job release. */
+	sp = host1x_syncpt_get_by_id(host1x, syncpt.id);
 	if (!sp) {
 		err = -ENOENT;
 		goto fail;
@@ -462,7 +316,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 	job->is_addr_reg = context->client->ops->is_addr_reg;
 	job->is_valid_class = context->client->ops->is_valid_class;
 	job->syncpt_incrs = syncpt.incrs;
-	job->syncpt_id = syncpt.id;
+	job->syncpt = sp;
 	job->timeout = 10000;
 
 	if (args->timeout && args->timeout < 10000)
@@ -482,7 +336,7 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 
 fail:
 	while (num_refs--)
-		drm_gem_object_put_unlocked(refs[num_refs]);
+		drm_gem_object_put(refs[num_refs]);
 
 	kfree(refs);
 
@@ -522,7 +376,7 @@ static int tegra_gem_mmap(struct drm_device *drm, void *data,
 
 	args->offset = drm_vma_node_offset_addr(&bo->gem.vma_node);
 
-	drm_gem_object_put_unlocked(gem);
+	drm_gem_object_put(gem);
 
 	return 0;
 }
@@ -534,7 +388,7 @@ static int tegra_syncpt_read(struct drm_device *drm, void *data,
 	struct drm_tegra_syncpt_read *args = data;
 	struct host1x_syncpt *sp;
 
-	sp = host1x_syncpt_get(host, args->id);
+	sp = host1x_syncpt_get_by_id_noref(host, args->id);
 	if (!sp)
 		return -EINVAL;
 
@@ -549,7 +403,7 @@ static int tegra_syncpt_incr(struct drm_device *drm, void *data,
 	struct drm_tegra_syncpt_incr *args = data;
 	struct host1x_syncpt *sp;
 
-	sp = host1x_syncpt_get(host1x, args->id);
+	sp = host1x_syncpt_get_by_id_noref(host1x, args->id);
 	if (!sp)
 		return -EINVAL;
 
@@ -563,7 +417,7 @@ static int tegra_syncpt_wait(struct drm_device *drm, void *data,
 	struct drm_tegra_syncpt_wait *args = data;
 	struct host1x_syncpt *sp;
 
-	sp = host1x_syncpt_get(host1x, args->id);
+	sp = host1x_syncpt_get_by_id_noref(host1x, args->id);
 	if (!sp)
 		return -EINVAL;
 
@@ -578,13 +432,20 @@ static int tegra_client_open(struct tegra_drm_file *fpriv,
 {
 	int err;
 
+	err = pm_runtime_resume_and_get(client->base.dev);
+	if (err)
+		return err;
+
 	err = client->ops->open_channel(client, context);
-	if (err < 0)
+	if (err < 0) {
+		pm_runtime_put(client->base.dev);
 		return err;
+	}
 
-	err = idr_alloc(&fpriv->contexts, context, 1, 0, GFP_KERNEL);
+	err = idr_alloc(&fpriv->legacy_contexts, context, 1, 0, GFP_KERNEL);
 	if (err < 0) {
 		client->ops->close_channel(context);
+		pm_runtime_put(client->base.dev);
 		return err;
 	}
 
@@ -637,13 +498,13 @@ static int tegra_close_channel(struct drm_device *drm, void *data,
 
 	mutex_lock(&fpriv->lock);
 
-	context = idr_find(&fpriv->contexts, args->context);
+	context = idr_find(&fpriv->legacy_contexts, args->context);
 	if (!context) {
 		err = -EINVAL;
 		goto unlock;
 	}
 
-	idr_remove(&fpriv->contexts, context->id);
+	idr_remove(&fpriv->legacy_contexts, context->id);
 	tegra_drm_context_free(context);
 
 unlock:
@@ -662,7 +523,7 @@ static int tegra_get_syncpt(struct drm_device *drm, void *data,
 
 	mutex_lock(&fpriv->lock);
 
-	context = idr_find(&fpriv->contexts, args->context);
+	context = idr_find(&fpriv->legacy_contexts, args->context);
 	if (!context) {
 		err = -ENODEV;
 		goto unlock;
@@ -691,7 +552,7 @@ static int tegra_submit(struct drm_device *drm, void *data,
 
 	mutex_lock(&fpriv->lock);
 
-	context = idr_find(&fpriv->contexts, args->context);
+	context = idr_find(&fpriv->legacy_contexts, args->context);
 	if (!context) {
 		err = -ENODEV;
 		goto unlock;
@@ -716,7 +577,7 @@ static int tegra_get_syncpt_base(struct drm_device *drm, void *data,
 
 	mutex_lock(&fpriv->lock);
 
-	context = idr_find(&fpriv->contexts, args->context);
+	context = idr_find(&fpriv->legacy_contexts, args->context);
 	if (!context) {
 		err = -ENODEV;
 		goto unlock;
@@ -790,7 +651,7 @@ static int tegra_gem_set_tiling(struct drm_device *drm, void *data,
 	bo->tiling.mode = mode;
 	bo->tiling.value = value;
 
-	drm_gem_object_put_unlocked(gem);
+	drm_gem_object_put(gem);
 
 	return 0;
 }
@@ -830,7 +691,7 @@ static int tegra_gem_get_tiling(struct drm_device *drm, void *data,
 		break;
 	}
 
-	drm_gem_object_put_unlocked(gem);
+	drm_gem_object_put(gem);
 
 	return err;
 }
@@ -855,7 +716,7 @@ static int tegra_gem_set_flags(struct drm_device *drm, void *data,
 	if (args->flags & DRM_TEGRA_GEM_BOTTOM_UP)
 		bo->flags |= TEGRA_BO_BOTTOM_UP;
 
-	drm_gem_object_put_unlocked(gem);
+	drm_gem_object_put(gem);
 
 	return 0;
 }
@@ -877,7 +738,7 @@ static int tegra_gem_get_flags(struct drm_device *drm, void *data,
 	if (bo->flags & TEGRA_BO_BOTTOM_UP)
 		args->flags |= DRM_TEGRA_GEM_BOTTOM_UP;
 
-	drm_gem_object_put_unlocked(gem);
+	drm_gem_object_put(gem);
 
 	return 0;
 }
@@ -885,34 +746,49 @@ static int tegra_gem_get_flags(struct drm_device *drm, void *data,
 
 static const struct drm_ioctl_desc tegra_drm_ioctls[] = {
 #ifdef CONFIG_DRM_TEGRA_STAGING
-	DRM_IOCTL_DEF_DRV(TEGRA_GEM_CREATE, tegra_gem_create,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF_DRV(TEGRA_GEM_MMAP, tegra_gem_mmap,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_CHANNEL_OPEN, tegra_drm_ioctl_channel_open,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_CHANNEL_CLOSE, tegra_drm_ioctl_channel_close,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_CHANNEL_MAP, tegra_drm_ioctl_channel_map,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_CHANNEL_UNMAP, tegra_drm_ioctl_channel_unmap,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_CHANNEL_SUBMIT, tegra_drm_ioctl_channel_submit,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPOINT_ALLOCATE, tegra_drm_ioctl_syncpoint_allocate,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPOINT_FREE, tegra_drm_ioctl_syncpoint_free,
+			  DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPOINT_WAIT, tegra_drm_ioctl_syncpoint_wait,
+			  DRM_RENDER_ALLOW),
+
+	DRM_IOCTL_DEF_DRV(TEGRA_GEM_CREATE, tegra_gem_create, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(TEGRA_GEM_MMAP, tegra_gem_mmap, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_READ, tegra_syncpt_read,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_INCR, tegra_syncpt_incr,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_SYNCPT_WAIT, tegra_syncpt_wait,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_OPEN_CHANNEL, tegra_open_channel,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_CLOSE_CHANNEL, tegra_close_channel,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GET_SYNCPT, tegra_get_syncpt,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_SUBMIT, tegra_submit,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GET_SYNCPT_BASE, tegra_get_syncpt_base,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GEM_SET_TILING, tegra_gem_set_tiling,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GEM_GET_TILING, tegra_gem_get_tiling,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GEM_SET_FLAGS, tegra_gem_set_flags,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(TEGRA_GEM_GET_FLAGS, tegra_gem_get_flags,
-			  DRM_UNLOCKED | DRM_RENDER_ALLOW),
+			  DRM_RENDER_ALLOW),
 #endif
 };
 
@@ -926,6 +802,7 @@ static const struct file_operations tegra_drm_fops = {
 	.read = drm_read,
 	.compat_ioctl = drm_compat_ioctl,
 	.llseek = noop_llseek,
+	.fop_flags = FOP_UNSIGNED_OFFSET,
 };
 
 static int tegra_drm_context_cleanup(int id, void *p, void *data)
@@ -942,10 +819,11 @@ static void tegra_drm_postclose(struct drm_device *drm, struct drm_file *file)
 	struct tegra_drm_file *fpriv = file->driver_priv;
 
 	mutex_lock(&fpriv->lock);
-	idr_for_each(&fpriv->contexts, tegra_drm_context_cleanup, NULL);
+	idr_for_each(&fpriv->legacy_contexts, tegra_drm_context_cleanup, NULL);
+	tegra_drm_uapi_close_file(fpriv);
 	mutex_unlock(&fpriv->lock);
 
-	idr_destroy(&fpriv->contexts);
+	idr_destroy(&fpriv->legacy_contexts);
 	mutex_destroy(&fpriv->lock);
 	kfree(fpriv);
 }
@@ -993,44 +871,36 @@ static struct drm_info_list tegra_debugfs_list[] = {
 	{ "iova", tegra_debugfs_iova, 0 },
 };
 
-static int tegra_debugfs_init(struct drm_minor *minor)
+static void tegra_debugfs_init(struct drm_minor *minor)
 {
-	return drm_debugfs_create_files(tegra_debugfs_list,
-					ARRAY_SIZE(tegra_debugfs_list),
-					minor->debugfs_root, minor);
+	drm_debugfs_create_files(tegra_debugfs_list,
+				 ARRAY_SIZE(tegra_debugfs_list),
+				 minor->debugfs_root, minor);
 }
 #endif
 
-static struct drm_driver tegra_drm_driver = {
-	.driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME |
-			   DRIVER_ATOMIC | DRIVER_RENDER,
-	.load = tegra_drm_load,
-	.unload = tegra_drm_unload,
+static const struct drm_driver tegra_drm_driver = {
+	.driver_features = DRIVER_MODESET | DRIVER_GEM |
+			   DRIVER_ATOMIC | DRIVER_RENDER | DRIVER_SYNCOBJ,
 	.open = tegra_drm_open,
 	.postclose = tegra_drm_postclose,
-	.lastclose = drm_fb_helper_lastclose,
 
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = tegra_debugfs_init,
 #endif
 
-	.gem_free_object_unlocked = tegra_bo_free_object,
-	.gem_vm_ops = &tegra_bo_vm_ops,
-
-	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
-	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
-	.gem_prime_export = tegra_gem_prime_export,
 	.gem_prime_import = tegra_gem_prime_import,
 
 	.dumb_create = tegra_bo_dumb_create,
 
+	TEGRA_FBDEV_DRIVER_OPS,
+
 	.ioctls = tegra_drm_ioctls,
 	.num_ioctls = ARRAY_SIZE(tegra_drm_ioctls),
 	.fops = &tegra_drm_fops,
 
 	.name = DRIVER_NAME,
 	.desc = DRIVER_DESC,
-	.date = DRIVER_DATE,
 	.major = DRIVER_MAJOR,
 	.minor = DRIVER_MINOR,
 	.patchlevel = DRIVER_PATCHLEVEL,
@@ -1039,8 +909,17 @@ static struct drm_driver tegra_drm_driver = {
 int tegra_drm_register_client(struct tegra_drm *tegra,
 			      struct tegra_drm_client *client)
 {
+	/*
+	 * When MLOCKs are implemented, change to allocate a shared channel
+	 * only when MLOCKs are disabled.
+	 */
+	client->shared_channel = host1x_channel_request(&client->base);
+	if (!client->shared_channel)
+		return -EBUSY;
+
 	mutex_lock(&tegra->clients_lock);
 	list_add_tail(&client->list, &tegra->clients);
+	client->drm = tegra;
 	mutex_unlock(&tegra->clients_lock);
 
 	return 0;
@@ -1051,62 +930,82 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 {
 	mutex_lock(&tegra->clients_lock);
 	list_del_init(&client->list);
+	client->drm = NULL;
 	mutex_unlock(&tegra->clients_lock);
 
+	if (client->shared_channel)
+		host1x_channel_put(client->shared_channel);
+
 	return 0;
 }
 
-struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
-					       bool shared)
+int host1x_client_iommu_attach(struct host1x_client *client)
 {
-	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct iommu_domain *domain = iommu_get_domain_for_dev(client->dev);
+	struct drm_device *drm = dev_get_drvdata(client->host);
 	struct tegra_drm *tegra = drm->dev_private;
 	struct iommu_group *group = NULL;
 	int err;
 
+#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
+	if (client->dev->archdata.mapping) {
+		struct dma_iommu_mapping *mapping =
+				to_dma_iommu_mapping(client->dev);
+		arm_iommu_detach_device(client->dev);
+		arm_iommu_release_mapping(mapping);
+
+		domain = iommu_get_domain_for_dev(client->dev);
+	}
+#endif
+
+	/*
+	 * If the host1x client is already attached to an IOMMU domain that is
+	 * not the shared IOMMU domain, don't try to attach it to a different
+	 * domain. This allows using the IOMMU-backed DMA API.
+	 */
+	if (domain && domain->type != IOMMU_DOMAIN_IDENTITY &&
+	    domain != tegra->domain)
+		return 0;
+
 	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 (!group)
+			return -ENODEV;
 
-		if (!shared || (shared && (group != tegra->group))) {
-#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
-			if (client->dev->archdata.mapping) {
-				struct dma_iommu_mapping *mapping =
-					to_dma_iommu_mapping(client->dev);
-				arm_iommu_detach_device(client->dev);
-				arm_iommu_release_mapping(mapping);
-			}
-#endif
+		if (domain != tegra->domain) {
 			err = iommu_attach_group(tegra->domain, group);
 			if (err < 0) {
 				iommu_group_put(group);
-				return ERR_PTR(err);
+				return err;
 			}
-
-			if (shared && !tegra->group)
-				tegra->group = group;
 		}
+
+		tegra->use_explicit_iommu = true;
 	}
 
-	return group;
+	client->group = group;
+
+	return 0;
 }
 
-void host1x_client_iommu_detach(struct host1x_client *client,
-				struct iommu_group *group)
+void host1x_client_iommu_detach(struct host1x_client *client)
 {
-	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct drm_device *drm = dev_get_drvdata(client->host);
 	struct tegra_drm *tegra = drm->dev_private;
+	struct iommu_domain *domain;
 
-	if (group) {
-		if (group == tegra->group) {
-			iommu_detach_group(tegra->domain, group);
-			tegra->group = NULL;
-		}
+	if (client->group) {
+		/*
+		 * Devices that are part of the same group may no longer be
+		 * attached to a domain at this point because their group may
+		 * have been detached by an earlier client.
+		 */
+		domain = iommu_get_domain_for_dev(client->dev);
+		if (domain)
+			iommu_detach_group(tegra->domain, client->group);
 
-		iommu_group_put(group);
+		iommu_group_put(client->group);
+		client->group = NULL;
 	}
 }
 
@@ -1156,7 +1055,7 @@ void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *dma)
 
 	*dma = iova_dma_addr(&tegra->carveout.domain, alloc);
 	err = iommu_map(tegra->domain, *dma, virt_to_phys(virt),
-			size, IOMMU_READ | IOMMU_WRITE);
+			size, IOMMU_READ | IOMMU_WRITE, GFP_KERNEL);
 	if (err < 0)
 		goto free_iova;
 
@@ -1187,28 +1086,216 @@ void tegra_drm_free(struct tegra_drm *tegra, size_t size, void *virt,
 	free_pages((unsigned long)virt, get_order(size));
 }
 
+static bool host1x_drm_wants_iommu(struct host1x_device *dev)
+{
+	struct host1x *host1x = dev_get_drvdata(dev->dev.parent);
+	struct iommu_domain *domain;
+
+	/* Our IOMMU usage policy doesn't currently play well with GART */
+	if (of_machine_is_compatible("nvidia,tegra20"))
+		return false;
+
+	/*
+	 * If the Tegra DRM clients are backed by an IOMMU, push buffers are
+	 * likely to be allocated beyond the 32-bit boundary if sufficient
+	 * system memory is available. This is problematic on earlier Tegra
+	 * generations where host1x supports a maximum of 32 address bits in
+	 * the GATHER opcode. In this case, unless host1x is behind an IOMMU
+	 * as well it won't be able to process buffers allocated beyond the
+	 * 32-bit boundary.
+	 *
+	 * The DMA API will use bounce buffers in this case, so that could
+	 * perhaps still be made to work, even if less efficient, but there
+	 * is another catch: in order to perform cache maintenance on pages
+	 * allocated for discontiguous buffers we need to map and unmap the
+	 * SG table representing these buffers. This is fine for something
+	 * small like a push buffer, but it exhausts the bounce buffer pool
+	 * (typically on the order of a few MiB) for framebuffers (many MiB
+	 * for any modern resolution).
+	 *
+	 * Work around this by making sure that Tegra DRM clients only use
+	 * an IOMMU if the parent host1x also uses an IOMMU.
+	 *
+	 * Note that there's still a small gap here that we don't cover: if
+	 * the DMA API is backed by an IOMMU there's no way to control which
+	 * device is attached to an IOMMU and which isn't, except via wiring
+	 * up the device tree appropriately. This is considered an problem
+	 * of integration, so care must be taken for the DT to be consistent.
+	 */
+	domain = iommu_get_domain_for_dev(dev->dev.parent);
+
+	/*
+	 * Tegra20 and Tegra30 don't support addressing memory beyond the
+	 * 32-bit boundary, so the regular GATHER opcodes will always be
+	 * sufficient and whether or not the host1x is attached to an IOMMU
+	 * doesn't matter.
+	 */
+	if (!domain && host1x_get_dma_mask(host1x) <= DMA_BIT_MASK(32))
+		return true;
+
+	return domain != NULL;
+}
+
 static int host1x_drm_probe(struct host1x_device *dev)
 {
-	struct drm_driver *driver = &tegra_drm_driver;
+	struct device *dma_dev = dev->dev.parent;
+	struct tegra_drm *tegra;
 	struct drm_device *drm;
 	int err;
 
-	drm = drm_dev_alloc(driver, &dev->dev);
+	drm = drm_dev_alloc(&tegra_drm_driver, &dev->dev);
 	if (IS_ERR(drm))
 		return PTR_ERR(drm);
 
+	tegra = kzalloc(sizeof(*tegra), GFP_KERNEL);
+	if (!tegra) {
+		err = -ENOMEM;
+		goto put;
+	}
+
+	if (host1x_drm_wants_iommu(dev) && device_iommu_mapped(dma_dev)) {
+		tegra->domain = iommu_paging_domain_alloc(dma_dev);
+		if (IS_ERR(tegra->domain)) {
+			err = PTR_ERR(tegra->domain);
+			goto free;
+		}
+
+		err = iova_cache_get();
+		if (err < 0)
+			goto domain;
+	}
+
+	mutex_init(&tegra->clients_lock);
+	INIT_LIST_HEAD(&tegra->clients);
+
 	dev_set_drvdata(&dev->dev, drm);
+	drm->dev_private = tegra;
+	tegra->drm = drm;
+
+	drm_mode_config_init(drm);
+
+	drm->mode_config.min_width = 0;
+	drm->mode_config.min_height = 0;
+	drm->mode_config.max_width = 0;
+	drm->mode_config.max_height = 0;
+
+	drm->mode_config.normalize_zpos = true;
+
+	drm->mode_config.funcs = &tegra_drm_mode_config_funcs;
+	drm->mode_config.helper_private = &tegra_drm_mode_config_helpers;
 
-	err = drm_fb_helper_remove_conflicting_framebuffers(NULL, "tegradrmfb", false);
+	drm_kms_helper_poll_init(drm);
+
+	err = host1x_device_init(dev);
 	if (err < 0)
-		goto put;
+		goto poll;
+
+	/*
+	 * Now that all display controller have been initialized, the maximum
+	 * supported resolution is known and the bitmask for horizontal and
+	 * vertical bitfields can be computed.
+	 */
+	tegra->hmask = drm->mode_config.max_width - 1;
+	tegra->vmask = drm->mode_config.max_height - 1;
+
+	if (tegra->use_explicit_iommu) {
+		u64 carveout_start, carveout_end, gem_start, gem_end;
+		u64 dma_mask = dma_get_mask(&dev->dev);
+		dma_addr_t start, end;
+		unsigned long order;
+
+		start = tegra->domain->geometry.aperture_start & dma_mask;
+		end = tegra->domain->geometry.aperture_end & dma_mask;
+
+		gem_start = start;
+		gem_end = end - CARVEOUT_SZ;
+		carveout_start = gem_end + 1;
+		carveout_end = end;
+
+		order = __ffs(tegra->domain->pgsize_bitmap);
+		init_iova_domain(&tegra->carveout.domain, 1UL << order,
+				 carveout_start >> order);
+
+		tegra->carveout.shift = iova_shift(&tegra->carveout.domain);
+		tegra->carveout.limit = carveout_end >> tegra->carveout.shift;
+
+		drm_mm_init(&tegra->mm, gem_start, gem_end - gem_start + 1);
+		mutex_init(&tegra->mm_lock);
+
+		DRM_DEBUG_DRIVER("IOMMU apertures:\n");
+		DRM_DEBUG_DRIVER("  GEM: %#llx-%#llx\n", gem_start, gem_end);
+		DRM_DEBUG_DRIVER("  Carveout: %#llx-%#llx\n", carveout_start,
+				 carveout_end);
+	} else if (tegra->domain) {
+		iommu_domain_free(tegra->domain);
+		tegra->domain = NULL;
+		iova_cache_put();
+	}
+
+	if (tegra->hub) {
+		err = tegra_display_hub_prepare(tegra->hub);
+		if (err < 0)
+			goto device;
+	}
+
+	/* syncpoints are used for full 32-bit hardware VBLANK counters */
+	drm->max_vblank_count = 0xffffffff;
+
+	err = drm_vblank_init(drm, drm->mode_config.num_crtc);
+	if (err < 0)
+		goto hub;
+
+	drm_mode_config_reset(drm);
+
+	/*
+	 * Only take over from a potential firmware framebuffer if any CRTCs
+	 * have been registered. This must not be a fatal error because there
+	 * are other accelerators that are exposed via this driver.
+	 *
+	 * Another case where this happens is on Tegra234 where the display
+	 * hardware is no longer part of the host1x complex, so this driver
+	 * will not expose any modesetting features.
+	 */
+	if (drm->mode_config.num_crtc > 0) {
+		err = aperture_remove_all_conflicting_devices(tegra_drm_driver.name);
+		if (err < 0)
+			goto hub;
+	} else {
+		/*
+		 * Indicate to userspace that this doesn't expose any display
+		 * capabilities.
+		 */
+		drm->driver_features &= ~(DRIVER_MODESET | DRIVER_ATOMIC);
+	}
 
 	err = drm_dev_register(drm, 0);
 	if (err < 0)
-		goto put;
+		goto hub;
+
+	drm_client_setup(drm, NULL);
 
 	return 0;
 
+hub:
+	if (tegra->hub)
+		tegra_display_hub_cleanup(tegra->hub);
+device:
+	if (tegra->domain) {
+		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
+		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
+	}
+
+	host1x_device_exit(dev);
+poll:
+	drm_kms_helper_poll_fini(drm);
+	drm_mode_config_cleanup(drm);
+domain:
+	if (tegra->domain)
+		iommu_domain_free(tegra->domain);
+free:
+	kfree(tegra);
 put:
 	drm_dev_put(drm);
 	return err;
@@ -1217,13 +1304,41 @@ put:
 static int host1x_drm_remove(struct host1x_device *dev)
 {
 	struct drm_device *drm = dev_get_drvdata(&dev->dev);
+	struct tegra_drm *tegra = drm->dev_private;
+	int err;
 
 	drm_dev_unregister(drm);
+
+	drm_kms_helper_poll_fini(drm);
+	drm_atomic_helper_shutdown(drm);
+	drm_mode_config_cleanup(drm);
+
+	if (tegra->hub)
+		tegra_display_hub_cleanup(tegra->hub);
+
+	err = host1x_device_exit(dev);
+	if (err < 0)
+		dev_err(&dev->dev, "host1x device cleanup failed: %d\n", err);
+
+	if (tegra->domain) {
+		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);
 	drm_dev_put(drm);
 
 	return 0;
 }
 
+static void host1x_drm_shutdown(struct host1x_device *dev)
+{
+	drm_atomic_helper_shutdown(dev_get_drvdata(&dev->dev));
+}
+
 #ifdef CONFIG_PM_SLEEP
 static int host1x_drm_suspend(struct device *dev)
 {
@@ -1252,8 +1367,10 @@ static const struct of_device_id host1x_drm_subdevs[] = {
 	{ .compatible = "nvidia,tegra30-hdmi", },
 	{ .compatible = "nvidia,tegra30-gr2d", },
 	{ .compatible = "nvidia,tegra30-gr3d", },
+	{ .compatible = "nvidia,tegra114-dc", },
 	{ .compatible = "nvidia,tegra114-dsi", },
 	{ .compatible = "nvidia,tegra114-hdmi", },
+	{ .compatible = "nvidia,tegra114-gr2d", },
 	{ .compatible = "nvidia,tegra114-gr3d", },
 	{ .compatible = "nvidia,tegra124-dc", },
 	{ .compatible = "nvidia,tegra124-sor", },
@@ -1266,15 +1383,21 @@ static const struct of_device_id host1x_drm_subdevs[] = {
 	{ .compatible = "nvidia,tegra210-sor", },
 	{ .compatible = "nvidia,tegra210-sor1", },
 	{ .compatible = "nvidia,tegra210-vic", },
+	{ .compatible = "nvidia,tegra210-nvdec", },
+	{ .compatible = "nvidia,tegra210-nvjpg", },
 	{ .compatible = "nvidia,tegra186-display", },
 	{ .compatible = "nvidia,tegra186-dc", },
 	{ .compatible = "nvidia,tegra186-sor", },
 	{ .compatible = "nvidia,tegra186-sor1", },
 	{ .compatible = "nvidia,tegra186-vic", },
+	{ .compatible = "nvidia,tegra186-nvdec", },
 	{ .compatible = "nvidia,tegra194-display", },
 	{ .compatible = "nvidia,tegra194-dc", },
 	{ .compatible = "nvidia,tegra194-sor", },
 	{ .compatible = "nvidia,tegra194-vic", },
+	{ .compatible = "nvidia,tegra194-nvdec", },
+	{ .compatible = "nvidia,tegra234-vic", },
+	{ .compatible = "nvidia,tegra234-nvdec", },
 	{ /* sentinel */ }
 };
 
@@ -1285,6 +1408,7 @@ static struct host1x_driver host1x_drm_driver = {
 	},
 	.probe = host1x_drm_probe,
 	.remove = host1x_drm_remove,
+	.shutdown = host1x_drm_shutdown,
 	.subdevs = host1x_drm_subdevs,
 };
 
@@ -1298,12 +1422,17 @@ static struct platform_driver * const drivers[] = {
 	&tegra_gr2d_driver,
 	&tegra_gr3d_driver,
 	&tegra_vic_driver,
+	&tegra_nvdec_driver,
+	&tegra_nvjpg_driver,
 };
 
 static int __init host1x_drm_init(void)
 {
 	int err;
 
+	if (drm_firmware_drivers_only())
+		return -ENODEV;
+
 	err = host1x_driver_register(&host1x_drm_driver);
 	if (err < 0)
 		return err;