Merge branch 'drm-next-4.19' of git://people.freedesktop.org/~agd5f/linux into drm-next

Fixes for 4.19:
- Fix UVD 7.2 instance handling
- Fix UVD 7.2 harvesting
- GPU scheduler fix for when a process is killed
- TTM cleanups
- amdgpu CS bo_list fixes
- Powerplay fixes for polaris12 and CZ/ST
- DC fixes for link training certain HMDs
- DC fix for vega10 blank screen in certain cases

From: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180801222906.1016-1-alexander.deucher@amd.com

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -74,6 +74,7 @@
 #include "amdgpu_gart.h"
 #include "amdgpu_debugfs.h"
 #include "amdgpu_job.h"
+#include "amdgpu_bo_list.h"
 
 /*
  * Modules parameters.
@@ -689,45 +690,6 @@ struct amdgpu_fpriv {
 	struct amdgpu_ctx_mgr	ctx_mgr;
 };
 
-/*
- * residency list
- */
-struct amdgpu_bo_list_entry {
-	struct amdgpu_bo		*robj;
-	struct ttm_validate_buffer	tv;
-	struct amdgpu_bo_va		*bo_va;
-	uint32_t			priority;
-	struct page			**user_pages;
-	int				user_invalidated;
-};
-
-struct amdgpu_bo_list {
-	struct mutex lock;
-	struct rcu_head rhead;
-	struct kref refcount;
-	struct amdgpu_bo *gds_obj;
-	struct amdgpu_bo *gws_obj;
-	struct amdgpu_bo *oa_obj;
-	unsigned first_userptr;
-	unsigned num_entries;
-	struct amdgpu_bo_list_entry *array;
-};
-
-struct amdgpu_bo_list *
-amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id);
-void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
-			     struct list_head *validated);
-void amdgpu_bo_list_put(struct amdgpu_bo_list *list);
-void amdgpu_bo_list_free(struct amdgpu_bo_list *list);
-int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
-				      struct drm_amdgpu_bo_list_entry **info_param);
-
-int amdgpu_bo_list_create(struct amdgpu_device *adev,
-				 struct drm_file *filp,
-				 struct drm_amdgpu_bo_list_entry *info,
-				 unsigned num_entries,
-				 struct amdgpu_bo_list **list);
-
 /*
  * GFX stuff
  */
@@ -1748,6 +1710,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr)))
 #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags)))
 #define amdgpu_ring_parse_cs(r, p, ib) ((r)->funcs->parse_cs((p), (ib)))
+#define amdgpu_ring_patch_cs_in_place(r, p, ib) ((r)->funcs->patch_cs_in_place((p), (ib)))
 #define amdgpu_ring_test_ring(r) (r)->funcs->test_ring((r))
 #define amdgpu_ring_test_ib(r, t) (r)->funcs->test_ib((r), (t))
 #define amdgpu_ring_get_rptr(r) (r)->funcs->get_rptr((r))

drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c

@@ -35,83 +35,53 @@
 #define AMDGPU_BO_LIST_MAX_PRIORITY	32u
 #define AMDGPU_BO_LIST_NUM_BUCKETS	(AMDGPU_BO_LIST_MAX_PRIORITY + 1)
 
-static int amdgpu_bo_list_set(struct amdgpu_device *adev,
+static void amdgpu_bo_list_free_rcu(struct rcu_head *rcu)
-				     struct drm_file *filp,
-				     struct amdgpu_bo_list *list,
-				     struct drm_amdgpu_bo_list_entry *info,
-				     unsigned num_entries);
-
-static void amdgpu_bo_list_release_rcu(struct kref *ref)
 {
-	unsigned i;
+	struct amdgpu_bo_list *list = container_of(rcu, struct amdgpu_bo_list,
-	struct amdgpu_bo_list *list = container_of(ref, struct amdgpu_bo_list,
+						   rhead);
-						   refcount);
 
-	for (i = 0; i < list->num_entries; ++i)
+	kvfree(list);
-		amdgpu_bo_unref(&list->array[i].robj);
-
-	mutex_destroy(&list->lock);
-	kvfree(list->array);
-	kfree_rcu(list, rhead);
 }
 
-int amdgpu_bo_list_create(struct amdgpu_device *adev,
+static void amdgpu_bo_list_free(struct kref *ref)
-				 struct drm_file *filp,
-				 struct drm_amdgpu_bo_list_entry *info,
-				 unsigned num_entries,
-				 struct amdgpu_bo_list **list_out)
 {
+	struct amdgpu_bo_list *list = container_of(ref, struct amdgpu_bo_list,
+						   refcount);
+	struct amdgpu_bo_list_entry *e;
+
+	amdgpu_bo_list_for_each_entry(e, list)
+		amdgpu_bo_unref(&e->robj);
+
+	call_rcu(&list->rhead, amdgpu_bo_list_free_rcu);
+}
+
+int amdgpu_bo_list_create(struct amdgpu_device *adev, struct drm_file *filp,
+			  struct drm_amdgpu_bo_list_entry *info,
+			  unsigned num_entries, struct amdgpu_bo_list **result)
+{
+	unsigned last_entry = 0, first_userptr = num_entries;
+	struct amdgpu_bo_list_entry *array;
 	struct amdgpu_bo_list *list;
+	uint64_t total_size = 0;
+	size_t size;
+	unsigned i;
 	int r;
 
+	if (num_entries > SIZE_MAX / sizeof(struct amdgpu_bo_list_entry))
+		return -EINVAL;
 
-	list = kzalloc(sizeof(struct amdgpu_bo_list), GFP_KERNEL);
+	size = sizeof(struct amdgpu_bo_list);
+	size += num_entries * sizeof(struct amdgpu_bo_list_entry);
+	list = kvmalloc(size, GFP_KERNEL);
 	if (!list)
 		return -ENOMEM;
 
-	/* initialize bo list*/
-	mutex_init(&list->lock);
 	kref_init(&list->refcount);
-	r = amdgpu_bo_list_set(adev, filp, list, info, num_entries);
+	list->gds_obj = adev->gds.gds_gfx_bo;
-	if (r) {
+	list->gws_obj = adev->gds.gws_gfx_bo;
-		kfree(list);
+	list->oa_obj = adev->gds.oa_gfx_bo;
-		return r;
-	}
 
-	*list_out = list;
+	array = amdgpu_bo_list_array_entry(list, 0);
-	return 0;
-}
-
-static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id)
-{
-	struct amdgpu_bo_list *list;
-
-	mutex_lock(&fpriv->bo_list_lock);
-	list = idr_remove(&fpriv->bo_list_handles, id);
-	mutex_unlock(&fpriv->bo_list_lock);
-	if (list)
-		kref_put(&list->refcount, amdgpu_bo_list_release_rcu);
-}
-
-static int amdgpu_bo_list_set(struct amdgpu_device *adev,
-				     struct drm_file *filp,
-				     struct amdgpu_bo_list *list,
-				     struct drm_amdgpu_bo_list_entry *info,
-				     unsigned num_entries)
-{
-	struct amdgpu_bo_list_entry *array;
-	struct amdgpu_bo *gds_obj = adev->gds.gds_gfx_bo;
-	struct amdgpu_bo *gws_obj = adev->gds.gws_gfx_bo;
-	struct amdgpu_bo *oa_obj = adev->gds.oa_gfx_bo;
-
-	unsigned last_entry = 0, first_userptr = num_entries;
-	unsigned i;
-	int r;
-	unsigned long total_size = 0;
-
-	array = kvmalloc_array(num_entries, sizeof(struct amdgpu_bo_list_entry), GFP_KERNEL);
-	if (!array)
-		return -ENOMEM;
 	memset(array, 0, num_entries * sizeof(struct amdgpu_bo_list_entry));
 
 	for (i = 0; i < num_entries; ++i) {
@@ -148,59 +118,56 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev,
 		entry->tv.shared = !entry->robj->prime_shared_count;
 
 		if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GDS)
-			gds_obj = entry->robj;
+			list->gds_obj = entry->robj;
 		if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_GWS)
-			gws_obj = entry->robj;
+			list->gws_obj = entry->robj;
 		if (entry->robj->preferred_domains == AMDGPU_GEM_DOMAIN_OA)
-			oa_obj = entry->robj;
+			list->oa_obj = entry->robj;
 
 		total_size += amdgpu_bo_size(entry->robj);
 		trace_amdgpu_bo_list_set(list, entry->robj);
 	}
 
-	for (i = 0; i < list->num_entries; ++i)
-		amdgpu_bo_unref(&list->array[i].robj);
-
-	kvfree(list->array);
-
-	list->gds_obj = gds_obj;
-	list->gws_obj = gws_obj;
-	list->oa_obj = oa_obj;
 	list->first_userptr = first_userptr;
-	list->array = array;
 	list->num_entries = num_entries;
 
 	trace_amdgpu_cs_bo_status(list->num_entries, total_size);
+
+	*result = list;
 	return 0;
 
 error_free:
 	while (i--)
 		amdgpu_bo_unref(&array[i].robj);
-	kvfree(array);
+	kvfree(list);
 	return r;
+
 }
 
-struct amdgpu_bo_list *
+static void amdgpu_bo_list_destroy(struct amdgpu_fpriv *fpriv, int id)
-amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id)
 {
-	struct amdgpu_bo_list *result;
+	struct amdgpu_bo_list *list;
 
+	mutex_lock(&fpriv->bo_list_lock);
+	list = idr_remove(&fpriv->bo_list_handles, id);
+	mutex_unlock(&fpriv->bo_list_lock);
+	if (list)
+		kref_put(&list->refcount, amdgpu_bo_list_free);
+}
+
+int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id,
+		       struct amdgpu_bo_list **result)
+{
 	rcu_read_lock();
-	result = idr_find(&fpriv->bo_list_handles, id);
+	*result = idr_find(&fpriv->bo_list_handles, id);
 
-	if (result) {
+	if (*result && kref_get_unless_zero(&(*result)->refcount)) {
-		if (kref_get_unless_zero(&result->refcount)) {
-			rcu_read_unlock();
-			mutex_lock(&result->lock);
-		} else {
-			rcu_read_unlock();
-			result = NULL;
-		}
-	} else {
 		rcu_read_unlock();
+		return 0;
 	}
 
-	return result;
+	rcu_read_unlock();
+	return -ENOENT;
 }
 
 void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
@@ -211,6 +178,7 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
 	 * concatenated in descending order.
 	 */
 	struct list_head bucket[AMDGPU_BO_LIST_NUM_BUCKETS];
+	struct amdgpu_bo_list_entry *e;
 	unsigned i;
 
 	for (i = 0; i < AMDGPU_BO_LIST_NUM_BUCKETS; i++)
@@ -221,14 +189,13 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
 	 * in the list, the sort mustn't change the ordering of buffers
 	 * with the same priority, i.e. it must be stable.
 	 */
-	for (i = 0; i < list->num_entries; i++) {
+	amdgpu_bo_list_for_each_entry(e, list) {
-		unsigned priority = list->array[i].priority;
+		unsigned priority = e->priority;
 
-		if (!list->array[i].robj->parent)
+		if (!e->robj->parent)
-			list_add_tail(&list->array[i].tv.head,
+			list_add_tail(&e->tv.head, &bucket[priority]);
-				      &bucket[priority]);
 
-		list->array[i].user_pages = NULL;
+		e->user_pages = NULL;
 	}
 
 	/* Connect the sorted buckets in the output list. */
@@ -238,20 +205,7 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
 
 void amdgpu_bo_list_put(struct amdgpu_bo_list *list)
 {
-	mutex_unlock(&list->lock);
+	kref_put(&list->refcount, amdgpu_bo_list_free);
-	kref_put(&list->refcount, amdgpu_bo_list_release_rcu);
-}
-
-void amdgpu_bo_list_free(struct amdgpu_bo_list *list)
-{
-	unsigned i;
-
-	for (i = 0; i < list->num_entries; ++i)
-		amdgpu_bo_unref(&list->array[i].robj);
-
-	mutex_destroy(&list->lock);
-	kvfree(list->array);
-	kfree(list);
 }
 
 int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
@@ -304,7 +258,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 	union drm_amdgpu_bo_list *args = data;
 	uint32_t handle = args->in.list_handle;
 	struct drm_amdgpu_bo_list_entry *info = NULL;
-	struct amdgpu_bo_list *list;
+	struct amdgpu_bo_list *list, *old;
 	int r;
 
 	r = amdgpu_bo_create_list_entry_array(&args->in, &info);
@@ -322,7 +276,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 		r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL);
 		mutex_unlock(&fpriv->bo_list_lock);
 		if (r < 0) {
-			amdgpu_bo_list_free(list);
+			amdgpu_bo_list_put(list);
 			return r;
 		}
 
@@ -335,17 +289,22 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 		break;
 
 	case AMDGPU_BO_LIST_OP_UPDATE:
-		r = -ENOENT;
+		r = amdgpu_bo_list_create(adev, filp, info, args->in.bo_number,
-		list = amdgpu_bo_list_get(fpriv, handle);
+					  &list);
-		if (!list)
-			goto error_free;
-
-		r = amdgpu_bo_list_set(adev, filp, list, info,
-					      args->in.bo_number);
-		amdgpu_bo_list_put(list);
 		if (r)
 			goto error_free;
 
+		mutex_lock(&fpriv->bo_list_lock);
+		old = idr_replace(&fpriv->bo_list_handles, list, handle);
+		mutex_unlock(&fpriv->bo_list_lock);
+
+		if (IS_ERR(old)) {
+			amdgpu_bo_list_put(list);
+			r = PTR_ERR(old);
+			goto error_free;
+		}
+
+		amdgpu_bo_list_put(old);
 		break;
 
 	default:

drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.h

@@ -0,0 +1,85 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __AMDGPU_BO_LIST_H__
+#define __AMDGPU_BO_LIST_H__
+
+#include <drm/ttm/ttm_execbuf_util.h>
+#include <drm/amdgpu_drm.h>
+
+struct amdgpu_device;
+struct amdgpu_bo;
+struct amdgpu_bo_va;
+struct amdgpu_fpriv;
+
+struct amdgpu_bo_list_entry {
+	struct amdgpu_bo		*robj;
+	struct ttm_validate_buffer	tv;
+	struct amdgpu_bo_va		*bo_va;
+	uint32_t			priority;
+	struct page			**user_pages;
+	int				user_invalidated;
+};
+
+struct amdgpu_bo_list {
+	struct rcu_head rhead;
+	struct kref refcount;
+	struct amdgpu_bo *gds_obj;
+	struct amdgpu_bo *gws_obj;
+	struct amdgpu_bo *oa_obj;
+	unsigned first_userptr;
+	unsigned num_entries;
+};
+
+int amdgpu_bo_list_get(struct amdgpu_fpriv *fpriv, int id,
+		       struct amdgpu_bo_list **result);
+void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
+			     struct list_head *validated);
+void amdgpu_bo_list_put(struct amdgpu_bo_list *list);
+int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
+				      struct drm_amdgpu_bo_list_entry **info_param);
+
+int amdgpu_bo_list_create(struct amdgpu_device *adev,
+				 struct drm_file *filp,
+				 struct drm_amdgpu_bo_list_entry *info,
+				 unsigned num_entries,
+				 struct amdgpu_bo_list **list);
+
+static inline struct amdgpu_bo_list_entry *
+amdgpu_bo_list_array_entry(struct amdgpu_bo_list *list, unsigned index)
+{
+	struct amdgpu_bo_list_entry *array = (void *)&list[1];
+
+	return &array[index];
+}
+
+#define amdgpu_bo_list_for_each_entry(e, list) \
+	for (e = amdgpu_bo_list_array_entry(list, 0); \
+	     e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \
+	     ++e)
+
+#define amdgpu_bo_list_for_each_userptr_entry(e, list) \
+	for (e = amdgpu_bo_list_array_entry(list, (list)->first_userptr); \
+	     e != amdgpu_bo_list_array_entry(list, (list)->num_entries); \
+	     ++e)
+
+#endif

drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c

@@ -561,27 +561,37 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 				union drm_amdgpu_cs *cs)
 {
 	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
+	struct amdgpu_vm *vm = &fpriv->vm;
 	struct amdgpu_bo_list_entry *e;
 	struct list_head duplicates;
-	unsigned i, tries = 10;
 	struct amdgpu_bo *gds;
 	struct amdgpu_bo *gws;
 	struct amdgpu_bo *oa;
+	unsigned tries = 10;
 	int r;
 
 	INIT_LIST_HEAD(&p->validated);
 
 	/* p->bo_list could already be assigned if AMDGPU_CHUNK_ID_BO_HANDLES is present */
-	if (!p->bo_list)
+	if (cs->in.bo_list_handle) {
-		p->bo_list = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle);
+		if (p->bo_list)
-	else
+			return -EINVAL;
-		mutex_lock(&p->bo_list->lock);
+
+		r = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle,
+				       &p->bo_list);
+		if (r)
+			return r;
+	} else if (!p->bo_list) {
+		/* Create a empty bo_list when no handle is provided */
+		r = amdgpu_bo_list_create(p->adev, p->filp, NULL, 0,
+					  &p->bo_list);
+		if (r)
+			return r;
+	}
 
-	if (p->bo_list) {
 	amdgpu_bo_list_get_list(p->bo_list, &p->validated);
 	if (p->bo_list->first_userptr != p->bo_list->num_entries)
 		p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX);
-	}
 
 	INIT_LIST_HEAD(&duplicates);
 	amdgpu_vm_get_pd_bo(&fpriv->vm, &p->validated, &p->vm_pd);
@@ -591,7 +601,6 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 
 	while (1) {
 		struct list_head need_pages;
-		unsigned i;
 
 		r = ttm_eu_reserve_buffers(&p->ticket, &p->validated, true,
 					   &duplicates);
@@ -601,17 +610,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 			goto error_free_pages;
 		}
 
-		/* Without a BO list we don't have userptr BOs */
-		if (!p->bo_list)
-			break;
-
 		INIT_LIST_HEAD(&need_pages);
-		for (i = p->bo_list->first_userptr;
+		amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) {
-		     i < p->bo_list->num_entries; ++i) {
+			struct amdgpu_bo *bo = e->robj;
-			struct amdgpu_bo *bo;
-
-			e = &p->bo_list->array[i];
-			bo = e->robj;
 
 			if (amdgpu_ttm_tt_userptr_invalidated(bo->tbo.ttm,
 				 &e->user_invalidated) && e->user_pages) {
@@ -703,23 +704,12 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 	amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved,
 				     p->bytes_moved_vis);
 
-	if (p->bo_list) {
-		struct amdgpu_vm *vm = &fpriv->vm;
-		unsigned i;
-
 	gds = p->bo_list->gds_obj;
 	gws = p->bo_list->gws_obj;
 	oa = p->bo_list->oa_obj;
-		for (i = 0; i < p->bo_list->num_entries; i++) {
-			struct amdgpu_bo *bo = p->bo_list->array[i].robj;
 
-			p->bo_list->array[i].bo_va = amdgpu_vm_bo_find(vm, bo);
+	amdgpu_bo_list_for_each_entry(e, p->bo_list)
-		}
+		e->bo_va = amdgpu_vm_bo_find(vm, e->robj);
-	} else {
-		gds = p->adev->gds.gds_gfx_bo;
-		gws = p->adev->gds.gws_gfx_bo;
-		oa = p->adev->gds.oa_gfx_bo;
-	}
 
 	if (gds) {
 		p->job->gds_base = amdgpu_bo_gpu_offset(gds);
@@ -747,11 +737,7 @@ error_validate:
 
 error_free_pages:
 
-	if (p->bo_list) {
+	amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) {
-		for (i = p->bo_list->first_userptr;
-		     i < p->bo_list->num_entries; ++i) {
-			e = &p->bo_list->array[i];
-
 		if (!e->user_pages)
 			continue;
 
@@ -759,7 +745,6 @@ error_free_pages:
 			      e->robj->tbo.ttm->num_pages);
 		kvfree(e->user_pages);
 	}
-	}
 
 	return r;
 }
@@ -820,12 +805,13 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error,
 
 static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
 {
-	struct amdgpu_device *adev = p->adev;
 	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
+	struct amdgpu_device *adev = p->adev;
 	struct amdgpu_vm *vm = &fpriv->vm;
+	struct amdgpu_bo_list_entry *e;
 	struct amdgpu_bo_va *bo_va;
 	struct amdgpu_bo *bo;
-	int i, r;
+	int r;
 
 	r = amdgpu_vm_clear_freed(adev, vm, NULL);
 	if (r)
@@ -855,16 +841,15 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
 			return r;
 	}
 
-	if (p->bo_list) {
+	amdgpu_bo_list_for_each_entry(e, p->bo_list) {
-		for (i = 0; i < p->bo_list->num_entries; i++) {
 		struct dma_fence *f;
 
 		/* ignore duplicates */
-			bo = p->bo_list->array[i].robj;
+		bo = e->robj;
 		if (!bo)
 			continue;
 
-			bo_va = p->bo_list->array[i].bo_va;
+		bo_va = e->bo_va;
 		if (bo_va == NULL)
 			continue;
 
@@ -878,8 +863,6 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
 			return r;
 	}
 
-	}
-
 	r = amdgpu_vm_handle_moved(adev, vm);
 	if (r)
 		return r;
@@ -892,15 +875,14 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p)
 	if (r)
 		return r;
 
-	if (amdgpu_vm_debug && p->bo_list) {
+	if (amdgpu_vm_debug) {
 		/* Invalidate all BOs to test for userspace bugs */
-		for (i = 0; i < p->bo_list->num_entries; i++) {
+		amdgpu_bo_list_for_each_entry(e, p->bo_list) {
 			/* ignore duplicates */
-			bo = p->bo_list->array[i].robj;
+			if (!e->robj)
-			if (!bo)
 				continue;
 
-			amdgpu_vm_bo_invalidate(adev, bo, false);
+			amdgpu_vm_bo_invalidate(adev, e->robj, false);
 		}
 	}
 
@@ -916,7 +898,7 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 	int r;
 
 	/* Only for UVD/VCE VM emulation */
-	if (p->ring->funcs->parse_cs) {
+	if (p->ring->funcs->parse_cs || p->ring->funcs->patch_cs_in_place) {
 		unsigned i, j;
 
 		for (i = 0, j = 0; i < p->nchunks && j < p->job->num_ibs; i++) {
@@ -957,12 +939,20 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 			offset = m->start * AMDGPU_GPU_PAGE_SIZE;
 			kptr += va_start - offset;
 
+			if (p->ring->funcs->parse_cs) {
 				memcpy(ib->ptr, kptr, chunk_ib->ib_bytes);
 				amdgpu_bo_kunmap(aobj);
 
 				r = amdgpu_ring_parse_cs(ring, p, j);
 				if (r)
 					return r;
+			} else {
+				ib->ptr = (uint32_t *)kptr;
+				r = amdgpu_ring_patch_cs_in_place(ring, p, j);
+				amdgpu_bo_kunmap(aobj);
+				if (r)
+					return r;
+			}
 
 			j++;
 		}
@@ -1207,27 +1197,25 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p)
 static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 			    union drm_amdgpu_cs *cs)
 {
+	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
 	struct amdgpu_ring *ring = p->ring;
 	struct drm_sched_entity *entity = &p->ctx->rings[ring->idx].entity;
 	enum drm_sched_priority priority;
+	struct amdgpu_bo_list_entry *e;
 	struct amdgpu_job *job;
-	unsigned i;
 	uint64_t seq;
 
 	int r;
 
 	amdgpu_mn_lock(p->mn);
-	if (p->bo_list) {
+	amdgpu_bo_list_for_each_userptr_entry(e, p->bo_list) {
-		for (i = p->bo_list->first_userptr;
+		struct amdgpu_bo *bo = e->robj;
-		     i < p->bo_list->num_entries; ++i) {
-			struct amdgpu_bo *bo = p->bo_list->array[i].robj;
 
 		if (amdgpu_ttm_tt_userptr_needs_pages(bo->tbo.ttm)) {
 			amdgpu_mn_unlock(p->mn);
 			return -ERESTARTSYS;
 		}
 	}
-	}
 
 	job = p->job;
 	p->job = NULL;
@@ -1259,6 +1247,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	amdgpu_job_free_resources(job);
 
 	trace_amdgpu_cs_ioctl(job);
+	amdgpu_vm_bo_trace_cs(&fpriv->vm, &p->ticket);
 	priority = job->base.s_priority;
 	drm_sched_entity_push_job(&job->base, entity);
 

drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c

@@ -286,7 +286,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	struct drm_crtc *crtc;
 	uint32_t ui32 = 0;
 	uint64_t ui64 = 0;
-	int i, found;
+	int i, j, found;
 	int ui32_size = sizeof(ui32);
 
 	if (!info->return_size || !info->return_pointer)
@@ -348,7 +348,11 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_UVD:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			ring_mask |= adev->uvd.inst[0].ring.ready;
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+				if (adev->uvd.harvest_config & (1 << i))
+					continue;
+				ring_mask |= adev->uvd.inst[i].ring.ready;
+			}
 			ib_start_alignment = 64;
 			ib_size_alignment = 64;
 			break;
@@ -361,9 +365,12 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_UVD_ENC:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			for (i = 0; i < adev->uvd.num_enc_rings; i++)
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
-				ring_mask |=
+				if (adev->uvd.harvest_config & (1 << i))
-					adev->uvd.inst[0].ring_enc[i].ready << i;
+					continue;
+				for (j = 0; j < adev->uvd.num_enc_rings; j++)
+					ring_mask |= adev->uvd.inst[i].ring_enc[j].ready << j;
+			}
 			ib_start_alignment = 64;
 			ib_size_alignment = 64;
 			break;
@@ -960,7 +967,7 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 	amdgpu_bo_unref(&pd);
 
 	idr_for_each_entry(&fpriv->bo_list_handles, list, handle)
-		amdgpu_bo_list_free(list);
+		amdgpu_bo_list_put(list);
 
 	idr_destroy(&fpriv->bo_list_handles);
 	mutex_destroy(&fpriv->bo_list_lock);

drivers/gpu/drm/amd/amdgpu/amdgpu_object.c

@@ -825,7 +825,7 @@ struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo)
 	if (bo == NULL)
 		return NULL;
 
-	ttm_bo_reference(&bo->tbo);
+	ttm_bo_get(&bo->tbo);
 	return bo;
 }
 
@@ -843,8 +843,7 @@ void amdgpu_bo_unref(struct amdgpu_bo **bo)
 		return;
 
 	tbo = &((*bo)->tbo);
-	ttm_bo_unref(&tbo);
+	ttm_bo_put(tbo);
-	if (tbo == NULL)
 	*bo = NULL;
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c

@@ -31,7 +31,7 @@
 #include <linux/power_supply.h>
 #include <linux/hwmon.h>
 #include <linux/hwmon-sysfs.h>
-
+#include <linux/nospec.h>
 
 static int amdgpu_debugfs_pm_init(struct amdgpu_device *adev);
 
@@ -403,6 +403,7 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev,
 			count = -EINVAL;
 			goto fail;
 		}
+		idx = array_index_nospec(idx, ARRAY_SIZE(data.states));
 
 		amdgpu_dpm_get_pp_num_states(adev, &data);
 		state = data.states[idx];
@@ -1185,7 +1186,7 @@ static ssize_t amdgpu_hwmon_show_vddnb(struct device *dev,
 	int r, size = sizeof(vddnb);
 
 	/* only APUs have vddnb */
-	if  (adev->flags & AMD_IS_APU)
+	if  (!(adev->flags & AMD_IS_APU))
 		return -EINVAL;
 
 	/* Can't get voltage when the card is off */

drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c

@@ -214,7 +214,7 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 			 u32 hw_ip, u32 instance, u32 ring,
 			 struct amdgpu_ring **out_ring)
 {
-	int r, ip_num_rings;
+	int i, r, ip_num_rings = 0;
 	struct amdgpu_queue_mapper *mapper = &mgr->mapper[hw_ip];
 
 	if (!adev || !mgr || !out_ring)
@@ -243,14 +243,21 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 		ip_num_rings = adev->sdma.num_instances;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		ip_num_rings = adev->uvd.num_uvd_inst;
+		for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+			if (!(adev->uvd.harvest_config & (1 << i)))
+				ip_num_rings++;
+		}
 		break;
 	case AMDGPU_HW_IP_VCE:
 		ip_num_rings = adev->vce.num_rings;
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
+		for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+			if (!(adev->uvd.harvest_config & (1 << i)))
+				ip_num_rings++;
+		}
 		ip_num_rings =
-			adev->uvd.num_enc_rings * adev->uvd.num_uvd_inst;
+			adev->uvd.num_enc_rings * ip_num_rings;
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		ip_num_rings = 1;

drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h

@@ -123,6 +123,7 @@ struct amdgpu_ring_funcs {
 	void (*set_wptr)(struct amdgpu_ring *ring);
 	/* validating and patching of IBs */
 	int (*parse_cs)(struct amdgpu_cs_parser *p, uint32_t ib_idx);
+	int (*patch_cs_in_place)(struct amdgpu_cs_parser *p, uint32_t ib_idx);
 	/* constants to calculate how many DW are needed for an emit */
 	unsigned emit_frame_size;
 	unsigned emit_ib_size;

drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h

@@ -314,6 +314,11 @@ DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_mapping,
 	    TP_ARGS(mapping)
 );
 
+DEFINE_EVENT(amdgpu_vm_mapping, amdgpu_vm_bo_cs,
+	    TP_PROTO(struct amdgpu_bo_va_mapping *mapping),
+	    TP_ARGS(mapping)
+);
+
 TRACE_EVENT(amdgpu_vm_set_ptes,
 	    TP_PROTO(uint64_t pe, uint64_t addr, unsigned count,
 		     uint32_t incr, uint64_t flags),

drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c

@@ -92,11 +92,9 @@ static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref)
 }
 
 /**
- * amdgpu_ttm_global_init - Initialize global TTM memory reference
+ * amdgpu_ttm_global_init - Initialize global TTM memory reference structures.
- * 							structures.
  *
- * @adev:  	AMDGPU device for which the global structures need to be
+ * @adev: AMDGPU device for which the global structures need to be registered.
- *			registered.
  *
  * This is called as part of the AMDGPU ttm init from amdgpu_ttm_init()
  * during bring up.
@@ -162,13 +160,12 @@ static int amdgpu_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 }
 
 /**
- * amdgpu_init_mem_type - 	Initialize a memory manager for a specific
+ * amdgpu_init_mem_type - Initialize a memory manager for a specific type of
- * 							type of memory request.
+ * memory request.
  *
- * @bdev:	The TTM BO device object (contains a reference to
+ * @bdev: The TTM BO device object (contains a reference to amdgpu_device)
- * 			amdgpu_device)
  * @type: The type of memory requested
- * @man:
+ * @man: The memory type manager for each domain
  *
  * This is called by ttm_bo_init_mm() when a buffer object is being
  * initialized.
@@ -321,8 +318,7 @@ static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp)
  * @bo: The bo to assign the memory to
  * @new_mem: The memory to be assigned.
  *
- * Assign the memory from new_mem to the memory of the buffer object
+ * Assign the memory from new_mem to the memory of the buffer object bo.
- * bo.
  */
 static void amdgpu_move_null(struct ttm_buffer_object *bo,
 			     struct ttm_mem_reg *new_mem)
@@ -335,8 +331,12 @@ static void amdgpu_move_null(struct ttm_buffer_object *bo,
 }
 
 /**
- * amdgpu_mm_node_addr -	Compute the GPU relative offset of a GTT
+ * amdgpu_mm_node_addr - Compute the GPU relative offset of a GTT buffer.
- * 							buffer.
+ *
+ * @bo: The bo to assign the memory to.
+ * @mm_node: Memory manager node for drm allocator.
+ * @mem: The region where the bo resides.
+ *
  */
 static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 				    struct drm_mm_node *mm_node,
@@ -352,10 +352,12 @@ static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 }
 
 /**
- * amdgpu_find_mm_node -	Helper function finds the drm_mm_node
+ * amdgpu_find_mm_node - Helper function finds the drm_mm_node corresponding to
- *  						corresponding to @offset. It also modifies
+ * @offset. It also modifies the offset to be within the drm_mm_node returned
- * 							the offset to be within the drm_mm_node
+ *
- * 							returned
+ * @mem: The region where the bo resides.
+ * @offset: The offset that drm_mm_node is used for finding.
+ *
  */
 static struct drm_mm_node *amdgpu_find_mm_node(struct ttm_mem_reg *mem,
 					       unsigned long *offset)
@@ -497,8 +499,8 @@ error:
 /**
  * amdgpu_move_blit - Copy an entire buffer to another buffer
  *
- * This is a helper called by amdgpu_bo_move() and
+ * This is a helper called by amdgpu_bo_move() and amdgpu_move_vram_ram() to
- * amdgpu_move_vram_ram() to help move buffers to and from VRAM.
+ * help move buffers to and from VRAM.
  */
 static int amdgpu_move_blit(struct ttm_buffer_object *bo,
 			    bool evict, bool no_wait_gpu,
@@ -580,7 +582,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 	}
 
 	/* blit VRAM to GTT */
-	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, &tmp_mem, old_mem);
+	r = amdgpu_move_blit(bo, evict, ctx->no_wait_gpu, &tmp_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
@@ -632,7 +634,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	}
 
 	/* copy to VRAM */
-	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, new_mem, old_mem);
+	r = amdgpu_move_blit(bo, evict, ctx->no_wait_gpu, new_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
@@ -794,8 +796,8 @@ struct amdgpu_ttm_tt {
 };
 
 /**
- * amdgpu_ttm_tt_get_user_pages - 	Pin pages of memory pointed to
+ * amdgpu_ttm_tt_get_user_pages - Pin pages of memory pointed to by a USERPTR
- * 									by a USERPTR pointer to memory
+ * pointer to memory
  *
  * Called by amdgpu_gem_userptr_ioctl() and amdgpu_cs_parser_bos().
  * This provides a wrapper around the get_user_pages() call to provide
@@ -818,8 +820,10 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 	down_read(&mm->mmap_sem);
 
 	if (gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) {
-		/* check that we only use anonymous memory
+		/*
-		   to prevent problems with writeback */
+		 * check that we only use anonymous memory to prevent problems
+		 * with writeback
+		 */
 		unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE;
 		struct vm_area_struct *vma;
 
@@ -870,8 +874,7 @@ release_pages:
 }
 
 /**
- * amdgpu_ttm_tt_set_user_pages - 	Copy pages in, putting old pages
+ * amdgpu_ttm_tt_set_user_pages - Copy pages in, putting old pages as necessary.
- * 									as necessary.
  *
  * Called by amdgpu_cs_list_validate(). This creates the page list
  * that backs user memory and will ultimately be mapped into the device
@@ -915,8 +918,7 @@ void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm)
 }
 
 /**
- * amdgpu_ttm_tt_pin_userptr - 	prepare the sg table with the
+ * amdgpu_ttm_tt_pin_userptr - 	prepare the sg table with the user pages
- * 								user pages
  *
  * Called by amdgpu_ttm_backend_bind()
  **/
@@ -1295,8 +1297,8 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 }
 
 /**
- * amdgpu_ttm_tt_set_userptr -	Initialize userptr GTT ttm_tt
+ * amdgpu_ttm_tt_set_userptr - Initialize userptr GTT ttm_tt for the current
- * 								for the current task
+ * task
  *
  * @ttm: The ttm_tt object to bind this userptr object to
  * @addr:  The address in the current tasks VM space to use
@@ -1346,9 +1348,8 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 }
 
 /**
- * amdgpu_ttm_tt_affect_userptr -	Determine if a ttm_tt object lays
+ * amdgpu_ttm_tt_affect_userptr - Determine if a ttm_tt object lays inside an
- * 									inside an address range for the
+ * address range for the current task.
- * 									current task.
  *
  */
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
@@ -1386,8 +1387,7 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 }
 
 /**
- * amdgpu_ttm_tt_userptr_invalidated -	Has the ttm_tt object been
+ * amdgpu_ttm_tt_userptr_invalidated - Has the ttm_tt object been invalidated?
- * 										invalidated?
  */
 bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 				       int *last_invalidated)
@@ -1400,10 +1400,8 @@ bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 }
 
 /**
- * amdgpu_ttm_tt_userptr_needs_pages -	Have the pages backing this
+ * amdgpu_ttm_tt_userptr_needs_pages - Have the pages backing this ttm_tt object
- * 										ttm_tt object been invalidated
+ * been invalidated since the last time they've been set?
- * 										since the last time they've
- * 										been set?
  */
 bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm)
 {
@@ -1459,13 +1457,12 @@ uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 }
 
 /**
- * amdgpu_ttm_bo_eviction_valuable -	Check to see if we can evict
+ * amdgpu_ttm_bo_eviction_valuable - Check to see if we can evict a buffer
- * 										a buffer object.
+ * object.
  *
- * Return true if eviction is sensible.  Called by
+ * Return true if eviction is sensible. Called by ttm_mem_evict_first() on
- * ttm_mem_evict_first() on behalf of ttm_bo_mem_force_space()
+ * behalf of ttm_bo_mem_force_space() which tries to evict buffer objects until
- * which tries to evict buffer objects until it can find space
+ * it can find space for a new object and by ttm_bo_force_list_clean() which is
- * for a new object and by ttm_bo_force_list_clean() which is
  * used to clean out a memory space.
  */
 static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
@@ -1515,8 +1512,7 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 }
 
 /**
- * amdgpu_ttm_access_memory -	Read or Write memory that backs a
+ * amdgpu_ttm_access_memory - Read or Write memory that backs a buffer object.
- * 								buffer object.
  *
  * @bo:  The buffer object to read/write
  * @offset:  Offset into buffer object
@@ -1704,8 +1700,8 @@ error_create:
 	return r;
 }
 /**
- * amdgpu_ttm_init -	Init the memory management (ttm) as well as
+ * amdgpu_ttm_init - Init the memory management (ttm) as well as various
- * 						various gtt/vram related fields.
+ * gtt/vram related fields.
  *
  * This initializes all of the memory space pools that the TTM layer
  * will need such as the GTT space (system memory mapped to the device),
@@ -1856,8 +1852,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_ttm_late_init -	Handle any late initialization for
+ * amdgpu_ttm_late_init - Handle any late initialization for amdgpu_ttm
- * 							amdgpu_ttm
  */
 void amdgpu_ttm_late_init(struct amdgpu_device *adev)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c

@@ -255,7 +255,8 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 		bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
-
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
 					    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst[j].vcpu_bo,
 					    &adev->uvd.inst[j].gpu_addr, &adev->uvd.inst[j].cpu_addr);
@@ -308,6 +309,8 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 	drm_sched_entity_destroy(&adev->uvd.entity);
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		kfree(adev->uvd.inst[j].saved_bo);
 
 		amdgpu_bo_free_kernel(&adev->uvd.inst[j].vcpu_bo,
@@ -343,6 +346,8 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 	}
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		if (adev->uvd.inst[j].vcpu_bo == NULL)
 			continue;
 
@@ -365,6 +370,8 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		if (adev->uvd.inst[i].vcpu_bo == NULL)
 			return -EINVAL;
 
@@ -1159,6 +1166,8 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 	unsigned fences = 0, i, j;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring);
 		for (j = 0; j < adev->uvd.num_enc_rings; ++j) {
 			fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring_enc[j]);

drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h

@@ -48,6 +48,9 @@ struct amdgpu_uvd_inst {
 	uint32_t                srbm_soft_reset;
 };
 
+#define AMDGPU_UVD_HARVEST_UVD0 (1 << 0)
+#define AMDGPU_UVD_HARVEST_UVD1 (1 << 1)
+
 struct amdgpu_uvd {
 	const struct firmware	*fw;	/* UVD firmware */
 	unsigned		fw_version;
@@ -61,6 +64,7 @@ struct amdgpu_uvd {
 	atomic_t		handles[AMDGPU_MAX_UVD_HANDLES];
 	struct drm_sched_entity entity;
 	struct delayed_work	idle_work;
+	unsigned		harvest_config;
 };
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev);

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

@@ -2344,6 +2344,35 @@ struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm,
 	return amdgpu_vm_it_iter_first(&vm->va, addr, addr);
 }
 
+/**
+ * amdgpu_vm_bo_trace_cs - trace all reserved mappings
+ *
+ * @vm: the requested vm
+ * @ticket: CS ticket
+ *
+ * Trace all mappings of BOs reserved during a command submission.
+ */
+void amdgpu_vm_bo_trace_cs(struct amdgpu_vm *vm, struct ww_acquire_ctx *ticket)
+{
+	struct amdgpu_bo_va_mapping *mapping;
+
+	if (!trace_amdgpu_vm_bo_cs_enabled())
+		return;
+
+	for (mapping = amdgpu_vm_it_iter_first(&vm->va, 0, U64_MAX); mapping;
+	     mapping = amdgpu_vm_it_iter_next(mapping, 0, U64_MAX)) {
+		if (mapping->bo_va && mapping->bo_va->base.bo) {
+			struct amdgpu_bo *bo;
+
+			bo = mapping->bo_va->base.bo;
+			if (READ_ONCE(bo->tbo.resv->lock.ctx) != ticket)
+				continue;
+		}
+
+		trace_amdgpu_vm_bo_cs(mapping);
+	}
+}
+
 /**
  * amdgpu_vm_bo_rmv - remove a bo to a specific vm
  *

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h

@@ -318,6 +318,7 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 				uint64_t saddr, uint64_t size);
 struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm,
 							 uint64_t addr);
+void amdgpu_vm_bo_trace_cs(struct amdgpu_vm *vm, struct ww_acquire_ctx *ticket);
 void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
 		      struct amdgpu_bo_va *bo_va);
 void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size,

drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c

@@ -41,6 +41,12 @@
 #include "mmhub/mmhub_1_0_sh_mask.h"
 #include "ivsrcid/uvd/irqsrcs_uvd_7_0.h"
 
+#define mmUVD_PG0_CC_UVD_HARVESTING                                                                    0x00c7
+#define mmUVD_PG0_CC_UVD_HARVESTING_BASE_IDX                                                           1
+//UVD_PG0_CC_UVD_HARVESTING
+#define UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE__SHIFT                                                         0x1
+#define UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE_MASK                                                           0x00000002L
+
 #define UVD7_MAX_HW_INSTANCES_VEGA20			2
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev);
@@ -370,10 +376,25 @@ error:
 static int uvd_v7_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	if (adev->asic_type == CHIP_VEGA20)
+
+	if (adev->asic_type == CHIP_VEGA20) {
+		u32 harvest;
+		int i;
+
 		adev->uvd.num_uvd_inst = UVD7_MAX_HW_INSTANCES_VEGA20;
-	else
+		for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+			harvest = RREG32_SOC15(UVD, i, mmUVD_PG0_CC_UVD_HARVESTING);
+			if (harvest & UVD_PG0_CC_UVD_HARVESTING__UVD_DISABLE_MASK) {
+				adev->uvd.harvest_config |= 1 << i;
+			}
+		}
+		if (adev->uvd.harvest_config == (AMDGPU_UVD_HARVEST_UVD0 |
+						 AMDGPU_UVD_HARVEST_UVD1))
+			/* both instances are harvested, disable the block */
+			return -ENOENT;
+	} else {
 		adev->uvd.num_uvd_inst = 1;
+	}
 
 	if (amdgpu_sriov_vf(adev))
 		adev->uvd.num_enc_rings = 1;
@@ -393,6 +414,8 @@ static int uvd_v7_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		/* UVD TRAP */
 		r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], UVD_7_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT, &adev->uvd.inst[j].irq);
 		if (r)
@@ -425,6 +448,8 @@ static int uvd_v7_0_sw_init(void *handle)
 		return r;
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		if (!amdgpu_sriov_vf(adev)) {
 			ring = &adev->uvd.inst[j].ring;
 			sprintf(ring->name, "uvd<%d>", j);
@@ -472,6 +497,8 @@ static int uvd_v7_0_sw_fini(void *handle)
 		return r;
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
 			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
 	}
@@ -500,6 +527,8 @@ static int uvd_v7_0_hw_init(void *handle)
 		goto done;
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		ring = &adev->uvd.inst[j].ring;
 
 		if (!amdgpu_sriov_vf(adev)) {
@@ -579,8 +608,11 @@ static int uvd_v7_0_hw_fini(void *handle)
 		DRM_DEBUG("For SRIOV client, shouldn't do anything.\n");
 	}
 
-	for (i = 0; i < adev->uvd.num_uvd_inst; ++i)
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		adev->uvd.inst[i].ring.ready = false;
+	}
 
 	return 0;
 }
@@ -623,6 +655,8 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
 				lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
@@ -695,6 +729,8 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP, 0);
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		WDOORBELL32(adev->uvd.inst[i].ring_enc[0].doorbell_index, 0);
 		adev->wb.wb[adev->uvd.inst[i].ring_enc[0].wptr_offs] = 0;
 		adev->uvd.inst[i].ring_enc[0].wptr = 0;
@@ -751,6 +787,8 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 		init_table += header->uvd_table_offset;
 
 		for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+			if (adev->uvd.harvest_config & (1 << i))
+				continue;
 			ring = &adev->uvd.inst[i].ring;
 			ring->wptr = 0;
 			size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
@@ -890,6 +928,8 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
 	int i, j, k, r;
 
 	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		if (adev->uvd.harvest_config & (1 << k))
+			continue;
 		/* disable DPG */
 		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_POWER_STATUS), 0,
 				~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
@@ -902,6 +942,8 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
 	uvd_v7_0_mc_resume(adev);
 
 	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		if (adev->uvd.harvest_config & (1 << k))
+			continue;
 		ring = &adev->uvd.inst[k].ring;
 		/* disable clock gating */
 		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_CGC_CTRL), 0,
@@ -1069,6 +1111,8 @@ static void uvd_v7_0_stop(struct amdgpu_device *adev)
 	uint8_t i = 0;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		/* force RBC into idle state */
 		WREG32_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, 0x11010101);
 
@@ -1205,6 +1249,34 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 	return r;
 }
 
+/**
+ * uvd_v7_0_ring_patch_cs_in_place - Patch the IB for command submission.
+ *
+ * @p: the CS parser with the IBs
+ * @ib_idx: which IB to patch
+ *
+ */
+static int uvd_v7_0_ring_patch_cs_in_place(struct amdgpu_cs_parser *p,
+					   uint32_t ib_idx)
+{
+	struct amdgpu_ib *ib = &p->job->ibs[ib_idx];
+	unsigned i;
+
+	/* No patching necessary for the first instance */
+	if (!p->ring->me)
+		return 0;
+
+	for (i = 0; i < ib->length_dw; i += 2) {
+		uint32_t reg = amdgpu_get_ib_value(p, ib_idx, i);
+
+		reg -= p->adev->reg_offset[UVD_HWIP][0][1];
+		reg += p->adev->reg_offset[UVD_HWIP][1][1];
+
+		amdgpu_set_ib_value(p, ib_idx, i, reg);
+	}
+	return 0;
+}
+
 /**
  * uvd_v7_0_ring_emit_ib - execute indirect buffer
  *
@@ -1697,6 +1769,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.get_rptr = uvd_v7_0_ring_get_rptr,
 	.get_wptr = uvd_v7_0_ring_get_wptr,
 	.set_wptr = uvd_v7_0_ring_set_wptr,
+	.patch_cs_in_place = uvd_v7_0_ring_patch_cs_in_place,
 	.emit_frame_size =
 		6 + /* hdp invalidate */
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
@@ -1756,6 +1829,8 @@ static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		adev->uvd.inst[i].ring.funcs = &uvd_v7_0_ring_vm_funcs;
 		adev->uvd.inst[i].ring.me = i;
 		DRM_INFO("UVD(%d) is enabled in VM mode\n", i);
@@ -1767,6 +1842,8 @@ static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 	int i, j;
 
 	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		if (adev->uvd.harvest_config & (1 << j))
+			continue;
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
 			adev->uvd.inst[j].ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
 			adev->uvd.inst[j].ring_enc[i].me = j;
@@ -1786,6 +1863,8 @@ static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		if (adev->uvd.harvest_config & (1 << i))
+			continue;
 		adev->uvd.inst[i].irq.num_types = adev->uvd.num_enc_rings + 1;
 		adev->uvd.inst[i].irq.funcs = &uvd_v7_0_irq_funcs;
 	}

drivers/gpu/drm/amd/display/dc/core/dc_link.c

@@ -60,7 +60,14 @@
 
 enum {
 	LINK_RATE_REF_FREQ_IN_MHZ = 27,
-	PEAK_FACTOR_X1000 = 1006
+	PEAK_FACTOR_X1000 = 1006,
+	/*
+	* Some receivers fail to train on first try and are good
+	* on subsequent tries. 2 retries should be plenty. If we
+	* don't have a successful training then we don't expect to
+	* ever get one.
+	*/
+	LINK_TRAINING_MAX_VERIFY_RETRY = 2
 };
 
 /*******************************************************************************
@@ -760,7 +767,16 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				 */
 
 				/* deal with non-mst cases */
-				dp_verify_link_cap(link, &link->reported_link_cap);
+				for (i = 0; i < LINK_TRAINING_MAX_VERIFY_RETRY; i++) {
+					int fail_count = 0;
+
+					dp_verify_link_cap(link,
+							  &link->reported_link_cap,
+							  &fail_count);
+
+					if (fail_count == 0)
+						break;
+				}
 			}
 
 			/* HDMI-DVI Dongle */

drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c

@@ -33,7 +33,6 @@
 #include "include/vector.h"
 #include "core_types.h"
 #include "dc_link_ddc.h"
-#include "engine.h"
 #include "aux_engine.h"
 
 #define AUX_POWER_UP_WA_DELAY 500
@@ -640,7 +639,6 @@ int dc_link_aux_transfer(struct ddc_service *ddc,
 			     enum i2caux_transaction_action action)
 {
 	struct ddc *ddc_pin = ddc->ddc_pin;
-	struct engine *engine;
 	struct aux_engine *aux_engine;
 	enum aux_channel_operation_result operation_result;
 	struct aux_request_transaction_data aux_req;
@@ -652,8 +650,8 @@ int dc_link_aux_transfer(struct ddc_service *ddc,
 	memset(&aux_req, 0, sizeof(aux_req));
 	memset(&aux_rep, 0, sizeof(aux_rep));
 
-	engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en];
+	aux_engine = ddc->ctx->dc->res_pool->engines[ddc_pin->pin_data->en];
-	aux_engine = engine->funcs->acquire(engine, ddc_pin);
+	aux_engine->funcs->acquire(aux_engine, ddc_pin);
 
 	aux_req.type = type;
 	aux_req.action = action;
@@ -685,7 +683,7 @@ int dc_link_aux_transfer(struct ddc_service *ddc,
 		res = -1;
 		break;
 	}
-	aux_engine->base.funcs->release_engine(&aux_engine->base);
+	aux_engine->funcs->release_engine(aux_engine);
 	return res;
 }
 

drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c

@@ -1088,7 +1088,8 @@ static struct dc_link_settings get_max_link_cap(struct dc_link *link)
 
 bool dp_verify_link_cap(
 	struct dc_link *link,
-	struct dc_link_settings *known_limit_link_setting)
+	struct dc_link_settings *known_limit_link_setting,
+	int *fail_count)
 {
 	struct dc_link_settings max_link_cap = {0};
 	struct dc_link_settings cur_link_setting = {0};
@@ -1160,6 +1161,8 @@ bool dp_verify_link_cap(
 							skip_video_pattern);
 			if (status == LINK_TRAINING_SUCCESS)
 				success = true;
+			else
+				(*fail_count)++;
 		}
 
 		if (success)

drivers/gpu/drm/amd/display/dc/core/dc_surface.c

@@ -192,7 +192,7 @@ void dc_transfer_func_release(struct dc_transfer_func *tf)
 	kref_put(&tf->refcount, dc_transfer_func_free);
 }
 
-struct dc_transfer_func *dc_create_transfer_func()
+struct dc_transfer_func *dc_create_transfer_func(void)
 {
 	struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL);
 

drivers/gpu/drm/amd/display/dc/dc.h

@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.58"
+#define DC_VER "3.1.59"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6

drivers/gpu/drm/amd/display/dc/dce/dce_aux.c

@@ -28,12 +28,12 @@
 #include "dce/dce_11_0_sh_mask.h"
 
 #define CTX \
-	aux110->base.base.ctx
+	aux110->base.ctx
 #define REG(reg_name)\
 	(aux110->regs->reg_name)
 
 #define DC_LOGGER \
-	engine->base.ctx->logger
+	engine->ctx->logger
 
 #include "reg_helper.h"
 
@@ -51,9 +51,9 @@ enum {
 	AUX_DEFER_RETRY_COUNTER = 6
 };
 static void release_engine(
-	struct engine *engine)
+	struct aux_engine *engine)
 {
-	struct aux_engine_dce110 *aux110 = FROM_ENGINE(engine);
+	struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine);
 
 	dal_ddc_close(engine->ddc);
 
@@ -827,22 +827,21 @@ static bool end_of_transaction_command(
 
 	/* according Syed, it does not need now DoDummyMOT */
 }
-bool submit_request(
+static bool submit_request(
-	struct engine *engine,
+	struct aux_engine *engine,
 	struct i2caux_transaction_request *request,
 	bool middle_of_transaction)
 {
-	struct aux_engine *aux_engine = FROM_AUX_ENGINE_ENGINE(engine);
 
 	bool result;
 	bool mot_used = true;
 
 	switch (request->operation) {
 	case I2CAUX_TRANSACTION_READ:
-		result = read_command(aux_engine, request, mot_used);
+		result = read_command(engine, request, mot_used);
 	break;
 	case I2CAUX_TRANSACTION_WRITE:
-		result = write_command(aux_engine, request, mot_used);
+		result = write_command(engine, request, mot_used);
 	break;
 	default:
 		result = false;
@@ -854,45 +853,45 @@ bool submit_request(
 	 */
 
 	if (!middle_of_transaction || !result)
-		end_of_transaction_command(aux_engine, request);
+		end_of_transaction_command(engine, request);
 
 	/* mask AUX interrupt */
 
 	return result;
 }
 enum i2caux_engine_type get_engine_type(
-		const struct engine *engine)
+		const struct aux_engine *engine)
 {
 	return I2CAUX_ENGINE_TYPE_AUX;
 }
 
-static struct aux_engine *acquire(
+static bool acquire(
-	struct engine *engine,
+	struct aux_engine *engine,
 	struct ddc *ddc)
 {
-	struct aux_engine *aux_engine = FROM_AUX_ENGINE_ENGINE(engine);
+
 	enum gpio_result result;
 
-	if (aux_engine->funcs->is_engine_available) {
+	if (engine->funcs->is_engine_available) {
 		/*check whether SW could use the engine*/
-		if (!aux_engine->funcs->is_engine_available(aux_engine))
+		if (!engine->funcs->is_engine_available(engine))
-			return NULL;
+			return false;
 	}
 
 	result = dal_ddc_open(ddc, GPIO_MODE_HARDWARE,
 		GPIO_DDC_CONFIG_TYPE_MODE_AUX);
 
 	if (result != GPIO_RESULT_OK)
-		return NULL;
+		return false;
 
-	if (!aux_engine->funcs->acquire_engine(aux_engine)) {
+	if (!engine->funcs->acquire_engine(engine)) {
 		dal_ddc_close(ddc);
-		return NULL;
+		return false;
 	}
 
 	engine->ddc = ddc;
 
-	return aux_engine;
+	return true;
 }
 
 static const struct aux_engine_funcs aux_engine_funcs = {
@@ -902,9 +901,6 @@ static const struct aux_engine_funcs aux_engine_funcs = {
 	.read_channel_reply = read_channel_reply,
 	.get_channel_status = get_channel_status,
 	.is_engine_available = is_engine_available,
-};
-
-static const struct engine_funcs engine_funcs = {
 	.release_engine = release_engine,
 	.destroy_engine = dce110_engine_destroy,
 	.submit_request = submit_request,
@@ -912,10 +908,10 @@ static const struct engine_funcs engine_funcs = {
 	.acquire = acquire,
 };
 
-void dce110_engine_destroy(struct engine **engine)
+void dce110_engine_destroy(struct aux_engine **engine)
 {
 
-	struct aux_engine_dce110 *engine110 = FROM_ENGINE(*engine);
+	struct aux_engine_dce110 *engine110 = FROM_AUX_ENGINE(*engine);
 
 	kfree(engine110);
 	*engine = NULL;
@@ -927,13 +923,12 @@ struct aux_engine *dce110_aux_engine_construct(struct aux_engine_dce110 *aux_eng
 		uint32_t timeout_period,
 		const struct dce110_aux_registers *regs)
 {
-	aux_engine110->base.base.ddc = NULL;
+	aux_engine110->base.ddc = NULL;
-	aux_engine110->base.base.ctx = ctx;
+	aux_engine110->base.ctx = ctx;
 	aux_engine110->base.delay = 0;
 	aux_engine110->base.max_defer_write_retry = 0;
-	aux_engine110->base.base.funcs = &engine_funcs;
 	aux_engine110->base.funcs = &aux_engine_funcs;
-	aux_engine110->base.base.inst = inst;
+	aux_engine110->base.inst = inst;
 	aux_engine110->timeout_period = timeout_period;
 	aux_engine110->regs = regs;
 

drivers/gpu/drm/amd/display/dc/dce/dce_aux.h

@@ -103,9 +103,9 @@ struct aux_engine *dce110_aux_engine_construct(
 		uint32_t timeout_period,
 		const struct dce110_aux_registers *regs);
 
-void dce110_engine_destroy(struct engine **engine);
+void dce110_engine_destroy(struct aux_engine **engine);
 
 bool dce110_aux_engine_acquire(
-	struct engine *aux_engine,
+	struct aux_engine *aux_engine,
 	struct ddc *ddc);
 #endif

drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c

@@ -150,7 +150,7 @@ static void dce_dmcu_set_psr_enable(struct dmcu *dmcu, bool enable, bool wait)
 	}
 }
 
-static void dce_dmcu_setup_psr(struct dmcu *dmcu,
+static bool dce_dmcu_setup_psr(struct dmcu *dmcu,
 		struct dc_link *link,
 		struct psr_context *psr_context)
 {
@@ -261,6 +261,8 @@ static void dce_dmcu_setup_psr(struct dmcu *dmcu,
 
 	/* notifyDMCUMsg */
 	REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1);
+
+	return true;
 }
 
 static bool dce_is_dmcu_initialized(struct dmcu *dmcu)
@@ -558,11 +560,12 @@ static void dcn10_dmcu_set_psr_enable(struct dmcu *dmcu, bool enable, bool wait)
 		}
 
 		/* assert if max retry hit */
-	ASSERT(retryCount <= 1000);
+		if (retryCount >= 1000)
+			ASSERT(0);
 	}
 }
 
-static void dcn10_dmcu_setup_psr(struct dmcu *dmcu,
+static bool dcn10_dmcu_setup_psr(struct dmcu *dmcu,
 		struct dc_link *link,
 		struct psr_context *psr_context)
 {
@@ -577,7 +580,7 @@ static void dcn10_dmcu_setup_psr(struct dmcu *dmcu,
 
 	/* If microcontroller is not running, do nothing */
 	if (dmcu->dmcu_state != DMCU_RUNNING)
-		return;
+		return false;
 
 	link->link_enc->funcs->psr_program_dp_dphy_fast_training(link->link_enc,
 			psr_context->psrExitLinkTrainingRequired);
@@ -677,6 +680,11 @@ static void dcn10_dmcu_setup_psr(struct dmcu *dmcu,
 
 	/* notifyDMCUMsg */
 	REG_UPDATE(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 1);
+
+	/* waitDMCUReadyForCmd */
+	REG_WAIT(MASTER_COMM_CNTL_REG, MASTER_COMM_INTERRUPT, 0, 1, 10000);
+
+	return true;
 }
 
 static void dcn10_psr_wait_loop(

drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c

@@ -586,7 +586,7 @@ struct output_pixel_processor *dce100_opp_create(
 	return &opp->base;
 }
 
-struct engine *dce100_aux_engine_create(
+struct aux_engine *dce100_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -600,7 +600,7 @@ struct engine *dce100_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 struct clock_source *dce100_clock_source_create(

drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c

@@ -2552,7 +2552,7 @@ static void pplib_apply_display_requirements(
 	dc->prev_display_config = *pp_display_cfg;
 }
 
-static void dce110_set_bandwidth(
+void dce110_set_bandwidth(
 		struct dc *dc,
 		struct dc_state *context,
 		bool decrease_allowed)

drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h

@@ -68,6 +68,11 @@ void dce110_fill_display_configs(
 	const struct dc_state *context,
 	struct dm_pp_display_configuration *pp_display_cfg);
 
+void dce110_set_bandwidth(
+		struct dc *dc,
+		struct dc_state *context,
+		bool decrease_allowed);
+
 uint32_t dce110_get_min_vblank_time_us(const struct dc_state *context);
 
 void dp_receiver_power_ctrl(struct dc_link *link, bool on);

drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c

@@ -604,7 +604,7 @@ static struct output_pixel_processor *dce110_opp_create(
 	return &opp->base;
 }
 
-struct engine *dce110_aux_engine_create(
+struct aux_engine *dce110_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -618,7 +618,7 @@ struct engine *dce110_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 struct clock_source *dce110_clock_source_create(

drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c

@@ -604,7 +604,7 @@ struct output_pixel_processor *dce112_opp_create(
 	return &opp->base;
 }
 
-struct engine *dce112_aux_engine_create(
+struct aux_engine *dce112_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -618,7 +618,7 @@ struct engine *dce112_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 struct clock_source *dce112_clock_source_create(

drivers/gpu/drm/amd/display/dc/dce120/dce120_hw_sequencer.c

@@ -244,7 +244,16 @@ static void dce120_update_dchub(
 	dh_data->dchub_info_valid = false;
 }
 
+static void dce120_set_bandwidth(
+		struct dc *dc,
+		struct dc_state *context,
+		bool decrease_allowed)
+{
+	if (context->stream_count <= 0)
+		return;
 
+	dce110_set_bandwidth(dc, context, decrease_allowed);
+}
 
 void dce120_hw_sequencer_construct(struct dc *dc)
 {
@@ -254,5 +263,6 @@ void dce120_hw_sequencer_construct(struct dc *dc)
 	dce110_hw_sequencer_construct(dc);
 	dc->hwss.enable_display_power_gating = dce120_enable_display_power_gating;
 	dc->hwss.update_dchub = dce120_update_dchub;
+	dc->hwss.set_bandwidth = dce120_set_bandwidth;
 }
 

drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c

@@ -376,7 +376,7 @@ struct output_pixel_processor *dce120_opp_create(
 			     ctx, inst, &opp_regs[inst], &opp_shift, &opp_mask);
 	return &opp->base;
 }
-struct engine *dce120_aux_engine_create(
+struct aux_engine *dce120_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -390,7 +390,7 @@ struct engine *dce120_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 static const struct bios_registers bios_regs = {

drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c

@@ -464,7 +464,7 @@ static struct output_pixel_processor *dce80_opp_create(
 	return &opp->base;
 }
 
-struct engine *dce80_aux_engine_create(
+struct aux_engine *dce80_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -478,7 +478,7 @@ struct engine *dce80_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 static struct stream_encoder *dce80_stream_encoder_create(

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c

@@ -594,7 +594,7 @@ static struct output_pixel_processor *dcn10_opp_create(
 	return &opp->base;
 }
 
-struct engine *dcn10_aux_engine_create(
+struct aux_engine *dcn10_aux_engine_create(
 	struct dc_context *ctx,
 	uint32_t inst)
 {
@@ -608,7 +608,7 @@ struct engine *dcn10_aux_engine_create(
 				    SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
 				    &aux_engine_regs[inst]);
 
-	return &aux_engine->base.base;
+	return &aux_engine->base;
 }
 
 static struct mpc *dcn10_mpc_create(struct dc_context *ctx)

drivers/gpu/drm/amd/display/dc/inc/core_types.h

@@ -138,7 +138,7 @@ struct resource_pool {
 	struct output_pixel_processor *opps[MAX_PIPES];
 	struct timing_generator *timing_generators[MAX_PIPES];
 	struct stream_encoder *stream_enc[MAX_PIPES * 2];
-	struct engine *engines[MAX_PIPES];
+	struct aux_engine *engines[MAX_PIPES];
 	struct hubbub *hubbub;
 	struct mpc *mpc;
 	struct pp_smu_funcs_rv *pp_smu;

drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h

@@ -35,7 +35,8 @@ struct dc_link_settings;
 
 bool dp_verify_link_cap(
 	struct dc_link *link,
-	struct dc_link_settings *known_limit_link_setting);
+	struct dc_link_settings *known_limit_link_setting,
+	int *fail_count);
 
 bool dp_validate_mode_timing(
 	struct dc_link *link,

drivers/gpu/drm/amd/display/dc/inc/hw/aux_engine.h

@@ -26,46 +26,72 @@
 #ifndef __DAL_AUX_ENGINE_H__
 #define __DAL_AUX_ENGINE_H__
 
-#include "engine.h"
+#include "dc_ddc_types.h"
 #include "include/i2caux_interface.h"
 
-struct aux_engine;
+enum i2caux_transaction_operation {
-union aux_config;
+	I2CAUX_TRANSACTION_READ,
-struct aux_engine_funcs {
+	I2CAUX_TRANSACTION_WRITE
-	void (*destroy)(
-		struct aux_engine **ptr);
-	bool (*acquire_engine)(
-		struct aux_engine *engine);
-	void (*configure)(
-		struct aux_engine *engine,
-		union aux_config cfg);
-	void (*submit_channel_request)(
-		struct aux_engine *engine,
-		struct aux_request_transaction_data *request);
-	void (*process_channel_reply)(
-		struct aux_engine *engine,
-		struct aux_reply_transaction_data *reply);
-	int (*read_channel_reply)(
-		struct aux_engine *engine,
-		uint32_t size,
-		uint8_t *buffer,
-		uint8_t *reply_result,
-		uint32_t *sw_status);
-	enum aux_channel_operation_result (*get_channel_status)(
-		struct aux_engine *engine,
-		uint8_t *returned_bytes);
-	bool (*is_engine_available)(struct aux_engine *engine);
 };
-struct engine;
+
+enum i2caux_transaction_address_space {
+	I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C = 1,
+	I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD
+};
+
+struct i2caux_transaction_payload {
+	enum i2caux_transaction_address_space address_space;
+	uint32_t address;
+	uint32_t length;
+	uint8_t *data;
+};
+
+enum i2caux_transaction_status {
+	I2CAUX_TRANSACTION_STATUS_UNKNOWN = (-1L),
+	I2CAUX_TRANSACTION_STATUS_SUCCEEDED,
+	I2CAUX_TRANSACTION_STATUS_FAILED_CHANNEL_BUSY,
+	I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT,
+	I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR,
+	I2CAUX_TRANSACTION_STATUS_FAILED_NACK,
+	I2CAUX_TRANSACTION_STATUS_FAILED_INCOMPLETE,
+	I2CAUX_TRANSACTION_STATUS_FAILED_OPERATION,
+	I2CAUX_TRANSACTION_STATUS_FAILED_INVALID_OPERATION,
+	I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW,
+	I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON
+};
+
+struct i2caux_transaction_request {
+	enum i2caux_transaction_operation operation;
+	struct i2caux_transaction_payload payload;
+	enum i2caux_transaction_status status;
+};
+
+enum i2caux_engine_type {
+	I2CAUX_ENGINE_TYPE_UNKNOWN = (-1L),
+	I2CAUX_ENGINE_TYPE_AUX,
+	I2CAUX_ENGINE_TYPE_I2C_DDC_HW,
+	I2CAUX_ENGINE_TYPE_I2C_GENERIC_HW,
+	I2CAUX_ENGINE_TYPE_I2C_SW
+};
+
+enum i2c_default_speed {
+	I2CAUX_DEFAULT_I2C_HW_SPEED = 50,
+	I2CAUX_DEFAULT_I2C_SW_SPEED = 50
+};
+
+union aux_config;
+
 struct aux_engine {
-	struct engine base;
+	uint32_t inst;
+	struct ddc *ddc;
+	struct dc_context *ctx;
 	const struct aux_engine_funcs *funcs;
 	/* following values are expressed in milliseconds */
 	uint32_t delay;
 	uint32_t max_defer_write_retry;
-
 	bool acquire_reset;
 };
+
 struct read_command_context {
 	uint8_t *buffer;
 	uint32_t current_read_length;
@@ -86,6 +112,7 @@ struct read_command_context {
 	bool transaction_complete;
 	bool operation_succeeded;
 };
+
 struct write_command_context {
 	bool mot;
 
@@ -110,4 +137,44 @@ struct write_command_context {
 	bool transaction_complete;
 	bool operation_succeeded;
 };
+
+
+struct aux_engine_funcs {
+	void (*destroy)(
+		struct aux_engine **ptr);
+	bool (*acquire_engine)(
+		struct aux_engine *engine);
+	void (*configure)(
+		struct aux_engine *engine,
+		union aux_config cfg);
+	void (*submit_channel_request)(
+		struct aux_engine *engine,
+		struct aux_request_transaction_data *request);
+	void (*process_channel_reply)(
+		struct aux_engine *engine,
+		struct aux_reply_transaction_data *reply);
+	int (*read_channel_reply)(
+		struct aux_engine *engine,
+		uint32_t size,
+		uint8_t *buffer,
+		uint8_t *reply_result,
+		uint32_t *sw_status);
+	enum aux_channel_operation_result (*get_channel_status)(
+		struct aux_engine *engine,
+		uint8_t *returned_bytes);
+	bool (*is_engine_available)(struct aux_engine *engine);
+	enum i2caux_engine_type (*get_engine_type)(
+		const struct aux_engine *engine);
+	bool (*acquire)(
+		struct aux_engine *engine,
+		struct ddc *ddc);
+	bool (*submit_request)(
+		struct aux_engine *engine,
+		struct i2caux_transaction_request *request,
+		bool middle_of_transaction);
+	void (*release_engine)(
+		struct aux_engine *engine);
+	void (*destroy_engine)(
+		struct aux_engine **engine);
+};
 #endif

drivers/gpu/drm/amd/display/dc/inc/hw/dmcu.h

@@ -48,7 +48,7 @@ struct dmcu_funcs {
 			const char *src,
 			unsigned int bytes);
 	void (*set_psr_enable)(struct dmcu *dmcu, bool enable, bool wait);
-	void (*setup_psr)(struct dmcu *dmcu,
+	bool (*setup_psr)(struct dmcu *dmcu,
 			struct dc_link *link,
 			struct psr_context *psr_context);
 	void (*get_psr_state)(struct dmcu *dmcu, uint32_t *psr_state);

drivers/gpu/drm/amd/display/dc/inc/hw/engine.h

@@ -1,106 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-#ifndef __DAL_ENGINE_H__
-#define __DAL_ENGINE_H__
-
-#include "dc_ddc_types.h"
-
-enum i2caux_transaction_operation {
-	I2CAUX_TRANSACTION_READ,
-	I2CAUX_TRANSACTION_WRITE
-};
-
-enum i2caux_transaction_address_space {
-	I2CAUX_TRANSACTION_ADDRESS_SPACE_I2C = 1,
-	I2CAUX_TRANSACTION_ADDRESS_SPACE_DPCD
-};
-
-struct i2caux_transaction_payload {
-	enum i2caux_transaction_address_space address_space;
-	uint32_t address;
-	uint32_t length;
-	uint8_t *data;
-};
-
-enum i2caux_transaction_status {
-	I2CAUX_TRANSACTION_STATUS_UNKNOWN = (-1L),
-	I2CAUX_TRANSACTION_STATUS_SUCCEEDED,
-	I2CAUX_TRANSACTION_STATUS_FAILED_CHANNEL_BUSY,
-	I2CAUX_TRANSACTION_STATUS_FAILED_TIMEOUT,
-	I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR,
-	I2CAUX_TRANSACTION_STATUS_FAILED_NACK,
-	I2CAUX_TRANSACTION_STATUS_FAILED_INCOMPLETE,
-	I2CAUX_TRANSACTION_STATUS_FAILED_OPERATION,
-	I2CAUX_TRANSACTION_STATUS_FAILED_INVALID_OPERATION,
-	I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW,
-	I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON
-};
-
-struct i2caux_transaction_request {
-	enum i2caux_transaction_operation operation;
-	struct i2caux_transaction_payload payload;
-	enum i2caux_transaction_status status;
-};
-
-enum i2caux_engine_type {
-	I2CAUX_ENGINE_TYPE_UNKNOWN = (-1L),
-	I2CAUX_ENGINE_TYPE_AUX,
-	I2CAUX_ENGINE_TYPE_I2C_DDC_HW,
-	I2CAUX_ENGINE_TYPE_I2C_GENERIC_HW,
-	I2CAUX_ENGINE_TYPE_I2C_SW
-};
-
-enum i2c_default_speed {
-	I2CAUX_DEFAULT_I2C_HW_SPEED = 50,
-	I2CAUX_DEFAULT_I2C_SW_SPEED = 50
-};
-
-struct engine;
-
-struct engine_funcs {
-	enum i2caux_engine_type (*get_engine_type)(
-		const struct engine *engine);
-	struct aux_engine* (*acquire)(
-		struct engine *engine,
-		struct ddc *ddc);
-	bool (*submit_request)(
-		struct engine *engine,
-		struct i2caux_transaction_request *request,
-		bool middle_of_transaction);
-	void (*release_engine)(
-		struct engine *engine);
-	void (*destroy_engine)(
-		struct engine **engine);
-};
-
-struct engine {
-	const struct engine_funcs *funcs;
-	uint32_t inst;
-	struct ddc *ddc;
-	struct dc_context *ctx;
-};
-
-#endif

drivers/gpu/drm/amd/include/atomfirmware.h

@@ -1074,7 +1074,7 @@ struct atom_integrated_system_info_v1_11
   uint16_t  dpphy_override;                   // bit vector, enum of atom_sysinfo_dpphy_override_def
   uint16_t  lvds_misc;                        // enum of atom_sys_info_lvds_misc_def
   uint16_t  backlight_pwm_hz;                 // pwm frequency in hz
-  uint8_t   memorytype;                       // enum of atom_sys_mem_type
+  uint8_t   memorytype;                       // enum of atom_dmi_t17_mem_type_def, APU memory type indication.
   uint8_t   umachannelnumber;                 // number of memory channels
   uint8_t   pwr_on_digon_to_de;               /* all pwr sequence numbers below are in uint of 4ms */
   uint8_t   pwr_on_de_to_vary_bl;
@@ -1084,18 +1084,25 @@ struct atom_integrated_system_info_v1_11
   uint8_t   pwr_on_vary_bl_to_blon;
   uint8_t   pwr_down_bloff_to_vary_bloff;
   uint8_t   min_allowed_bl_level;
+  uint8_t   htc_hyst_limit;
+  uint8_t   htc_tmp_limit;
+  uint8_t   reserved1;
+  uint8_t   reserved2;
   struct atom_external_display_connection_info extdispconninfo;
   struct atom_14nm_dpphy_dvihdmi_tuningset dvi_tuningset;
   struct atom_14nm_dpphy_dvihdmi_tuningset hdmi_tuningset;
   struct atom_14nm_dpphy_dvihdmi_tuningset hdmi6g_tuningset;
-  struct atom_14nm_dpphy_dp_tuningset dp_tuningset;
+  struct atom_14nm_dpphy_dp_tuningset dp_tuningset;        // rbr 1.62G dp tuning set
-  struct atom_14nm_dpphy_dp_tuningset dp_hbr3_tuningset;
+  struct atom_14nm_dpphy_dp_tuningset dp_hbr3_tuningset;   // HBR3 dp tuning set
   struct atom_camera_data  camera_info;
   struct atom_hdmi_retimer_redriver_set dp0_retimer_set;   //for DP0
   struct atom_hdmi_retimer_redriver_set dp1_retimer_set;   //for DP1
   struct atom_hdmi_retimer_redriver_set dp2_retimer_set;   //for DP2
   struct atom_hdmi_retimer_redriver_set dp3_retimer_set;   //for DP3
-  uint32_t  reserved[108];
+  struct atom_14nm_dpphy_dp_tuningset dp_hbr_tuningset;    //hbr 2.7G dp tuning set
+  struct atom_14nm_dpphy_dp_tuningset dp_hbr2_tuningset;   //hbr2 5.4G dp turnig set
+  struct atom_14nm_dpphy_dp_tuningset edp_tuningset;       //edp tuning set
+  uint32_t  reserved[66];
 };
 
 

drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c

@@ -403,6 +403,49 @@ static const struct gpu_pt_config_reg DIDTConfig_Polaris12[] = {
 	{   ixDIDT_SQ_CTRL1,                   DIDT_SQ_CTRL1__MAX_POWER_MASK,                      DIDT_SQ_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
 
 	{   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__UNUSED_0_MASK,                    DIDT_SQ_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_SQ_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3853,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_0_MASK,                       DIDT_SQ_CTRL2__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,       DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,     0x005a,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_1_MASK,                       DIDT_SQ_CTRL2__UNUSED_1__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,       DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_2_MASK,                       DIDT_SQ_CTRL2__UNUSED_2__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,    DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT,  0x0001,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,   DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x0ebb,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__UNUSED_0_MASK,                  DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,       DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,     0x3853,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,     0x3153,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK,                 DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_SQ_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__PHASE_OFFSET_MASK,                   DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__UNUSED_0_MASK,                       DIDT_SQ_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT0_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT0__SHIFT,                  0x000a,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT1_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT1__SHIFT,                  0x0010,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT2_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT2__SHIFT,                  0x0017,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT3_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT3__SHIFT,                  0x002f,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT4_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT4__SHIFT,                  0x0046,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT5_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT5__SHIFT,                  0x005d,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT6_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT6__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT7_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT7__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MIN_POWER_MASK,                      DIDT_TD_CTRL1__MIN_POWER__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+	{   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MAX_POWER_MASK,                      DIDT_TD_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+	{   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__UNUSED_0_MASK,                    DIDT_TD_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
 	{   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_TD_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0x00ff,     GPU_CONFIGREG_DIDT_IND },
 
 	{   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3fff,     GPU_CONFIGREG_DIDT_IND },

drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c

@@ -244,6 +244,7 @@ static int smu8_initialize_dpm_defaults(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+/* convert form 8bit vid to real voltage in mV*4 */
 static uint32_t smu8_convert_8Bit_index_to_voltage(
 			struct pp_hwmgr *hwmgr, uint16_t voltage)
 {
@@ -1702,13 +1703,13 @@ static int smu8_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 	case AMDGPU_PP_SENSOR_VDDNB:
 		tmp = (cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMUSVI_NB_CURRENTVID) &
 			CURRENT_NB_VID_MASK) >> CURRENT_NB_VID__SHIFT;
-		vddnb = smu8_convert_8Bit_index_to_voltage(hwmgr, tmp);
+		vddnb = smu8_convert_8Bit_index_to_voltage(hwmgr, tmp) / 4;
 		*((uint32_t *)value) = vddnb;
 		return 0;
 	case AMDGPU_PP_SENSOR_VDDGFX:
 		tmp = (cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMUSVI_GFX_CURRENTVID) &
 			CURRENT_GFX_VID_MASK) >> CURRENT_GFX_VID__SHIFT;
-		vddgfx = smu8_convert_8Bit_index_to_voltage(hwmgr, (u16)tmp);
+		vddgfx = smu8_convert_8Bit_index_to_voltage(hwmgr, (u16)tmp) / 4;
 		*((uint32_t *)value) = vddgfx;
 		return 0;
 	case AMDGPU_PP_SENSOR_UVD_VCLK:

drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c

@@ -1204,7 +1204,6 @@ static int polaris10_populate_smc_acpi_level(struct pp_hwmgr *hwmgr,
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
 	SMIO_Pattern vol_level;
 	uint32_t mvdd;
-	uint16_t us_mvdd;
 
 	table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC;
 
@@ -1255,16 +1254,11 @@ static int polaris10_populate_smc_acpi_level(struct pp_hwmgr *hwmgr,
 			"in Clock Dependency Table",
 			);
 
-	us_mvdd = 0;
+	if (!((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) ||
-	if ((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) ||
+			(data->mclk_dpm_key_disabled)))
-			(data->mclk_dpm_key_disabled))
+		polaris10_populate_mvdd_value(hwmgr,
-		us_mvdd = data->vbios_boot_state.mvdd_bootup_value;
-	else {
-		if (!polaris10_populate_mvdd_value(hwmgr,
 				data->dpm_table.mclk_table.dpm_levels[0].value,
-				&vol_level))
+				&vol_level);
-			us_mvdd = vol_level.Voltage;
-	}
 
 	if (0 == polaris10_populate_mvdd_value(hwmgr, 0, &vol_level))
 		table->MemoryACPILevel.MinMvdd = PP_HOST_TO_SMC_UL(vol_level.Voltage);
@@ -1517,7 +1511,7 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
 	uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min;
 	struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend);
 
-	uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0;
+	uint8_t i, stretch_amount, volt_offset = 0;
 	struct phm_ppt_v1_information *table_info =
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
 	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
@@ -1568,11 +1562,7 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
 
 	smu_data->smc_state_table.LdoRefSel = (table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ? table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 6;
 	/* Populate CKS Lookup Table */
-	if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5)
+	if (stretch_amount == 0 || stretch_amount > 5) {
-		stretch_amount2 = 0;
-	else if (stretch_amount == 3 || stretch_amount == 4)
-		stretch_amount2 = 1;
-	else {
 		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_ClockStretcher);
 		PP_ASSERT_WITH_CODE(false,

drivers/gpu/drm/radeon/radeon_object.c

@@ -307,7 +307,7 @@ struct radeon_bo *radeon_bo_ref(struct radeon_bo *bo)
 	if (bo == NULL)
 		return NULL;
 
-	ttm_bo_reference(&bo->tbo);
+	ttm_bo_get(&bo->tbo);
 	return bo;
 }
 
@@ -320,8 +320,7 @@ void radeon_bo_unref(struct radeon_bo **bo)
 		return;
 	rdev = (*bo)->rdev;
 	tbo = &((*bo)->tbo);
-	ttm_bo_unref(&tbo);
+	ttm_bo_put(tbo);
-	if (tbo == NULL)
 	*bo = NULL;
 }
 

drivers/gpu/drm/scheduler/Makefile

@@ -20,7 +20,6 @@
 # OTHER DEALINGS IN THE SOFTWARE.
 #
 #
-ccflags-y := -Iinclude/drm
 gpu-sched-y := gpu_scheduler.o sched_fence.o
 
 obj-$(CONFIG_DRM_SCHED) += gpu-sched.o

drivers/gpu/drm/scheduler/gpu_scheduler.c

@@ -198,21 +198,6 @@ int drm_sched_entity_init(struct drm_sched_entity *entity,
 }
 EXPORT_SYMBOL(drm_sched_entity_init);
 
-/**
- * drm_sched_entity_is_initialized - Query if entity is initialized
- *
- * @sched: Pointer to scheduler instance
- * @entity: The pointer to a valid scheduler entity
- *
- * return true if entity is initialized, false otherwise
-*/
-static bool drm_sched_entity_is_initialized(struct drm_gpu_scheduler *sched,
-					    struct drm_sched_entity *entity)
-{
-	return entity->rq != NULL &&
-		entity->rq->sched == sched;
-}
-
 /**
  * drm_sched_entity_is_idle - Check if entity is idle
  *
@@ -224,7 +209,8 @@ static bool drm_sched_entity_is_idle(struct drm_sched_entity *entity)
 {
 	rmb();
 
-	if (!entity->rq || spsc_queue_peek(&entity->job_queue) == NULL)
+	if (list_empty(&entity->list) ||
+	    spsc_queue_peek(&entity->job_queue) == NULL)
 		return true;
 
 	return false;
@@ -275,11 +261,10 @@ static void drm_sched_entity_kill_jobs_cb(struct dma_fence *f,
 long drm_sched_entity_flush(struct drm_sched_entity *entity, long timeout)
 {
 	struct drm_gpu_scheduler *sched;
+	struct task_struct *last_user;
 	long ret = timeout;
 
 	sched = entity->rq->sched;
-	if (!drm_sched_entity_is_initialized(sched, entity))
-		return ret;
 	/**
 	 * The client will not queue more IBs during this fini, consume existing
 	 * queued IBs or discard them on SIGKILL
@@ -295,8 +280,10 @@ long drm_sched_entity_flush(struct drm_sched_entity *entity, long timeout)
 
 
 	/* For killed process disable any more IBs enqueue right now */
-	if ((current->flags & PF_EXITING) && (current->exit_code == SIGKILL))
+	last_user = cmpxchg(&entity->last_user, current->group_leader, NULL);
-		drm_sched_entity_set_rq(entity, NULL);
+	if ((!last_user || last_user == current->group_leader) &&
+	    (current->flags & PF_EXITING) && (current->exit_code == SIGKILL))
+		drm_sched_rq_remove_entity(entity->rq, entity);
 
 	return ret;
 }
@@ -317,7 +304,7 @@ void drm_sched_entity_fini(struct drm_sched_entity *entity)
 	struct drm_gpu_scheduler *sched;
 
 	sched = entity->rq->sched;
-	drm_sched_entity_set_rq(entity, NULL);
+	drm_sched_rq_remove_entity(entity->rq, entity);
 
 	/* Consumption of existing IBs wasn't completed. Forcefully
 	 * remove them here.
@@ -413,15 +400,12 @@ void drm_sched_entity_set_rq(struct drm_sched_entity *entity,
 	if (entity->rq == rq)
 		return;
 
+	BUG_ON(!rq);
+
 	spin_lock(&entity->rq_lock);
-
-	if (entity->rq)
 	drm_sched_rq_remove_entity(entity->rq, entity);
-
 	entity->rq = rq;
-	if (rq)
 	drm_sched_rq_add_entity(rq, entity);
-
 	spin_unlock(&entity->rq_lock);
 }
 EXPORT_SYMBOL(drm_sched_entity_set_rq);
@@ -541,6 +525,7 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job,
 
 	trace_drm_sched_job(sched_job, entity);
 
+	WRITE_ONCE(entity->last_user, current->group_leader);
 	first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node);
 
 	/* first job wakes up scheduler */

drivers/gpu/drm/ttm/ttm_page_alloc.c

@@ -47,13 +47,7 @@
 
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_page_alloc.h>
-
+#include <drm/ttm/ttm_set_memory.h>
-#if IS_ENABLED(CONFIG_AGP)
-#include <asm/agp.h>
-#endif
-#ifdef CONFIG_X86
-#include <asm/set_memory.h>
-#endif
 
 #define NUM_PAGES_TO_ALLOC		(PAGE_SIZE/sizeof(struct page *))
 #define SMALL_ALLOCATION		16
@@ -222,52 +216,6 @@ static struct kobj_type ttm_pool_kobj_type = {
 
 static struct ttm_pool_manager *_manager;
 
-#ifndef CONFIG_X86
-static int set_pages_wb(struct page *page, int numpages)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < numpages; i++)
-		unmap_page_from_agp(page++);
-#endif
-	return 0;
-}
-
-static int set_pages_array_wb(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		unmap_page_from_agp(pages[i]);
-#endif
-	return 0;
-}
-
-static int set_pages_array_wc(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		map_page_into_agp(pages[i]);
-#endif
-	return 0;
-}
-
-static int set_pages_array_uc(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		map_page_into_agp(pages[i]);
-#endif
-	return 0;
-}
-#endif
-
 /**
  * Select the right pool or requested caching state and ttm flags. */
 static struct ttm_page_pool *ttm_get_pool(int flags, bool huge,
@@ -302,13 +250,13 @@ static void ttm_pages_put(struct page *pages[], unsigned npages,
 	unsigned int i, pages_nr = (1 << order);
 
 	if (order == 0) {
-		if (set_pages_array_wb(pages, npages))
+		if (ttm_set_pages_array_wb(pages, npages))
 			pr_err("Failed to set %d pages to wb!\n", npages);
 	}
 
 	for (i = 0; i < npages; ++i) {
 		if (order > 0) {
-			if (set_pages_wb(pages[i], pages_nr))
+			if (ttm_set_pages_wb(pages[i], pages_nr))
 				pr_err("Failed to set %d pages to wb!\n", pages_nr);
 		}
 		__free_pages(pages[i], order);
@@ -498,12 +446,12 @@ static int ttm_set_pages_caching(struct page **pages,
 	/* Set page caching */
 	switch (cstate) {
 	case tt_uncached:
-		r = set_pages_array_uc(pages, cpages);
+		r = ttm_set_pages_array_uc(pages, cpages);
 		if (r)
 			pr_err("Failed to set %d pages to uc!\n", cpages);
 		break;
 	case tt_wc:
-		r = set_pages_array_wc(pages, cpages);
+		r = ttm_set_pages_array_wc(pages, cpages);
 		if (r)
 			pr_err("Failed to set %d pages to wc!\n", cpages);
 		break;

drivers/gpu/drm/ttm/ttm_page_alloc_dma.c

@@ -50,12 +50,7 @@
 #include <linux/kthread.h>
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_page_alloc.h>
-#if IS_ENABLED(CONFIG_AGP)
+#include <drm/ttm/ttm_set_memory.h>
-#include <asm/agp.h>
-#endif
-#ifdef CONFIG_X86
-#include <asm/set_memory.h>
-#endif
 
 #define NUM_PAGES_TO_ALLOC		(PAGE_SIZE/sizeof(struct page *))
 #define SMALL_ALLOCATION		4
@@ -268,54 +263,19 @@ static struct kobj_type ttm_pool_kobj_type = {
 	.default_attrs = ttm_pool_attrs,
 };
 
-#ifndef CONFIG_X86
-static int set_pages_array_wb(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		unmap_page_from_agp(pages[i]);
-#endif
-	return 0;
-}
-
-static int set_pages_array_wc(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		map_page_into_agp(pages[i]);
-#endif
-	return 0;
-}
-
-static int set_pages_array_uc(struct page **pages, int addrinarray)
-{
-#if IS_ENABLED(CONFIG_AGP)
-	int i;
-
-	for (i = 0; i < addrinarray; i++)
-		map_page_into_agp(pages[i]);
-#endif
-	return 0;
-}
-#endif /* for !CONFIG_X86 */
-
 static int ttm_set_pages_caching(struct dma_pool *pool,
 				 struct page **pages, unsigned cpages)
 {
 	int r = 0;
 	/* Set page caching */
 	if (pool->type & IS_UC) {
-		r = set_pages_array_uc(pages, cpages);
+		r = ttm_set_pages_array_uc(pages, cpages);
 		if (r)
 			pr_err("%s: Failed to set %d pages to uc!\n",
 			       pool->dev_name, cpages);
 	}
 	if (pool->type & IS_WC) {
-		r = set_pages_array_wc(pages, cpages);
+		r = ttm_set_pages_array_wc(pages, cpages);
 		if (r)
 			pr_err("%s: Failed to set %d pages to wc!\n",
 			       pool->dev_name, cpages);
@@ -389,17 +349,14 @@ static void ttm_pool_update_free_locked(struct dma_pool *pool,
 static void ttm_dma_page_put(struct dma_pool *pool, struct dma_page *d_page)
 {
 	struct page *page = d_page->p;
-	unsigned i, num_pages;
+	unsigned num_pages;
 
 	/* Don't set WB on WB page pool. */
 	if (!(pool->type & IS_CACHED)) {
 		num_pages = pool->size / PAGE_SIZE;
-		for (i = 0; i < num_pages; ++i, ++page) {
+		if (ttm_set_pages_wb(page, num_pages))
-			if (set_pages_array_wb(&page, 1)) {
 			pr_err("%s: Failed to set %d pages to wb!\n",
-				       pool->dev_name, 1);
+			       pool->dev_name, num_pages);
-			}
-		}
 	}
 
 	list_del(&d_page->page_list);
@@ -420,7 +377,7 @@ static void ttm_dma_pages_put(struct dma_pool *pool, struct list_head *d_pages,
 
 	/* Don't set WB on WB page pool. */
 	if (npages && !(pool->type & IS_CACHED) &&
-	    set_pages_array_wb(pages, npages))
+	    ttm_set_pages_array_wb(pages, npages))
 		pr_err("%s: Failed to set %d pages to wb!\n",
 		       pool->dev_name, npages);
 

drivers/gpu/drm/ttm/ttm_tt.c

@@ -38,9 +38,7 @@
 #include <drm/drm_cache.h>
 #include <drm/ttm/ttm_bo_driver.h>
 #include <drm/ttm/ttm_page_alloc.h>
-#ifdef CONFIG_X86
+#include <drm/ttm/ttm_set_memory.h>
-#include <asm/set_memory.h>
-#endif
 
 /**
  * Allocates a ttm structure for the given BO.
@@ -115,8 +113,7 @@ static int ttm_sg_tt_alloc_page_directory(struct ttm_dma_tt *ttm)
 	return 0;
 }
 
-#ifdef CONFIG_X86
+static int ttm_tt_set_page_caching(struct page *p,
-static inline int ttm_tt_set_page_caching(struct page *p,
 				   enum ttm_caching_state c_old,
 				   enum ttm_caching_state c_new)
 {
@@ -129,26 +126,18 @@ static inline int ttm_tt_set_page_caching(struct page *p,
 		/* p isn't in the default caching state, set it to
 		 * writeback first to free its current memtype. */
 
-		ret = set_pages_wb(p, 1);
+		ret = ttm_set_pages_wb(p, 1);
 		if (ret)
 			return ret;
 	}
 
 	if (c_new == tt_wc)
-		ret = set_memory_wc((unsigned long) page_address(p), 1);
+		ret = ttm_set_pages_wc(p, 1);
 	else if (c_new == tt_uncached)
-		ret = set_pages_uc(p, 1);
+		ret = ttm_set_pages_uc(p, 1);
 
 	return ret;
 }
-#else /* CONFIG_X86 */
-static inline int ttm_tt_set_page_caching(struct page *p,
-					  enum ttm_caching_state c_old,
-					  enum ttm_caching_state c_new)
-{
-	return 0;
-}
-#endif /* CONFIG_X86 */
 
 /*
  * Change caching policy for the linear kernel map

include/drm/gpu_scheduler.h

@@ -66,6 +66,7 @@ enum drm_sched_priority {
  * @guilty: points to ctx's guilty.
  * @fini_status: contains the exit status in case the process was signalled.
  * @last_scheduled: points to the finished fence of the last scheduled job.
+ * @last_user: last group leader pushing a job into the entity.
  *
  * Entities will emit jobs in order to their corresponding hardware
  * ring, and the scheduler will alternate between entities based on
@@ -85,6 +86,7 @@ struct drm_sched_entity {
 	struct dma_fence_cb		cb;
 	atomic_t			*guilty;
 	struct dma_fence                *last_scheduled;
+	struct task_struct		*last_user;
 };
 
 /**

include/drm/ttm/ttm_set_memory.h

@@ -0,0 +1,150 @@
+/**************************************************************************
+ *
+ * Copyright (c) 2018 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+/*
+ * Authors: Huang Rui <ray.huang@amd.com>
+ */
+
+#ifndef TTM_SET_MEMORY
+#define TTM_SET_MEMORY
+
+#include <linux/mm.h>
+
+#ifdef CONFIG_X86
+
+#include <asm/set_memory.h>
+
+static inline int ttm_set_pages_array_wb(struct page **pages, int addrinarray)
+{
+	return set_pages_array_wb(pages, addrinarray);
+}
+
+static inline int ttm_set_pages_array_wc(struct page **pages, int addrinarray)
+{
+	return set_pages_array_wc(pages, addrinarray);
+}
+
+static inline int ttm_set_pages_array_uc(struct page **pages, int addrinarray)
+{
+	return set_pages_array_uc(pages, addrinarray);
+}
+
+static inline int ttm_set_pages_wb(struct page *page, int numpages)
+{
+	return set_pages_wb(page, numpages);
+}
+
+static inline int ttm_set_pages_wc(struct page *page, int numpages)
+{
+	unsigned long addr = (unsigned long)page_address(page);
+
+	return set_memory_wc(addr, numpages);
+}
+
+static inline int ttm_set_pages_uc(struct page *page, int numpages)
+{
+	return set_pages_uc(page, numpages);
+}
+
+#else /* for CONFIG_X86 */
+
+#if IS_ENABLED(CONFIG_AGP)
+
+#include <asm/agp.h>
+
+static inline int ttm_set_pages_array_wb(struct page **pages, int addrinarray)
+{
+	int i;
+
+	for (i = 0; i < addrinarray; i++)
+		unmap_page_from_agp(pages[i]);
+	return 0;
+}
+
+static inline int ttm_set_pages_array_wc(struct page **pages, int addrinarray)
+{
+	int i;
+
+	for (i = 0; i < addrinarray; i++)
+		map_page_into_agp(pages[i]);
+	return 0;
+}
+
+static inline int ttm_set_pages_array_uc(struct page **pages, int addrinarray)
+{
+	int i;
+
+	for (i = 0; i < addrinarray; i++)
+		map_page_into_agp(pages[i]);
+	return 0;
+}
+
+static inline int ttm_set_pages_wb(struct page *page, int numpages)
+{
+	int i;
+
+	for (i = 0; i < numpages; i++)
+		unmap_page_from_agp(page++);
+	return 0;
+}
+
+#else /* for CONFIG_AGP */
+
+static inline int ttm_set_pages_array_wb(struct page **pages, int addrinarray)
+{
+	return 0;
+}
+
+static inline int ttm_set_pages_array_wc(struct page **pages, int addrinarray)
+{
+	return 0;
+}
+
+static inline int ttm_set_pages_array_uc(struct page **pages, int addrinarray)
+{
+	return 0;
+}
+
+static inline int ttm_set_pages_wb(struct page *page, int numpages)
+{
+	return 0;
+}
+
+#endif /* for CONFIG_AGP */
+
+static inline int ttm_set_pages_wc(struct page *page, int numpages)
+{
+	return 0;
+}
+
+static inline int ttm_set_pages_uc(struct page *page, int numpages)
+{
+	return 0;
+}
+
+#endif /* for CONFIG_X86 */
+
+#endif