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

Daniel needs a few commits from drm-next.

Signed-off-by: Maxime Ripard <maxime@cerno.tech>

MAINTAINERS

@@ -8424,7 +8424,7 @@ M:	Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
 M:	Rodrigo Vivi <rodrigo.vivi@intel.com>
 L:	intel-gfx@lists.freedesktop.org
 W:	https://01.org/linuxgraphics/
-B:	https://01.org/linuxgraphics/documentation/how-report-bugs
+B:	https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs
 C:	irc://chat.freenode.net/intel-gfx
 Q:	http://patchwork.freedesktop.org/project/intel-gfx/
 T:	git git://anongit.freedesktop.org/drm-intel

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

@@ -969,6 +969,7 @@ struct amdgpu_device {
 	int				pstate;
 	/* enable runtime pm on the device */
 	bool                            runpm;
+	bool                            in_runpm;
 
 	bool                            pm_sysfs_en;
 	bool                            ucode_sysfs_en;

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

@@ -126,7 +126,7 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 		/* this is going to have a few of the MSBs set that we need to
 		 * clear
 		 */
-		bitmap_complement(gpu_resources.queue_bitmap,
+		bitmap_complement(gpu_resources.cp_queue_bitmap,
 				  adev->gfx.mec.queue_bitmap,
 				  KGD_MAX_QUEUES);
 
@@ -137,7 +137,7 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 				* adev->gfx.mec.num_pipe_per_mec
 				* adev->gfx.mec.num_queue_per_pipe;
 		for (i = last_valid_bit; i < KGD_MAX_QUEUES; ++i)
-			clear_bit(i, gpu_resources.queue_bitmap);
+			clear_bit(i, gpu_resources.cp_queue_bitmap);
 
 		amdgpu_doorbell_get_kfd_info(adev,
 				&gpu_resources.doorbell_physical_address,
@@ -178,18 +178,18 @@ void amdgpu_amdkfd_interrupt(struct amdgpu_device *adev,
 		kgd2kfd_interrupt(adev->kfd.dev, ih_ring_entry);
 }
 
-void amdgpu_amdkfd_suspend(struct amdgpu_device *adev)
+void amdgpu_amdkfd_suspend(struct amdgpu_device *adev, bool run_pm)
 {
 	if (adev->kfd.dev)
-		kgd2kfd_suspend(adev->kfd.dev);
+		kgd2kfd_suspend(adev->kfd.dev, run_pm);
 }
 
-int amdgpu_amdkfd_resume(struct amdgpu_device *adev)
+int amdgpu_amdkfd_resume(struct amdgpu_device *adev, bool run_pm)
 {
 	int r = 0;
 
 	if (adev->kfd.dev)
-		r = kgd2kfd_resume(adev->kfd.dev);
+		r = kgd2kfd_resume(adev->kfd.dev, run_pm);
 
 	return r;
 }
@@ -402,7 +402,7 @@ void amdgpu_amdkfd_get_local_mem_info(struct kgd_dev *kgd,
 
 	if (amdgpu_sriov_vf(adev))
 		mem_info->mem_clk_max = adev->clock.default_mclk / 100;
-	else if (adev->powerplay.pp_funcs) {
+	else if (adev->pm.dpm_enabled) {
 		if (amdgpu_emu_mode == 1)
 			mem_info->mem_clk_max = 0;
 		else
@@ -427,7 +427,7 @@ uint32_t amdgpu_amdkfd_get_max_engine_clock_in_mhz(struct kgd_dev *kgd)
 	/* the sclk is in quantas of 10kHz */
 	if (amdgpu_sriov_vf(adev))
 		return adev->clock.default_sclk / 100;
-	else if (adev->powerplay.pp_funcs)
+	else if (adev->pm.dpm_enabled)
 		return amdgpu_dpm_get_sclk(adev, false) / 100;
 	else
 		return 100;
@@ -525,6 +525,14 @@ uint64_t amdgpu_amdkfd_get_hive_id(struct kgd_dev *kgd)
 
 	return adev->gmc.xgmi.hive_id;
 }
+
+uint64_t amdgpu_amdkfd_get_unique_id(struct kgd_dev *kgd)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	return adev->unique_id;
+}
+
 uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *src)
 {
 	struct amdgpu_device *peer_adev = (struct amdgpu_device *)src;
@@ -647,13 +655,9 @@ int amdgpu_amdkfd_flush_gpu_tlb_vmid(struct kgd_dev *kgd, uint16_t vmid)
 int amdgpu_amdkfd_flush_gpu_tlb_pasid(struct kgd_dev *kgd, uint16_t pasid)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
-	uint32_t flush_type = 0;
+	const uint32_t flush_type = 0;
 	bool all_hub = false;
 
-	if (adev->gmc.xgmi.num_physical_nodes &&
-		adev->asic_type == CHIP_VEGA20)
-		flush_type = 2;
-
 	if (adev->family == AMDGPU_FAMILY_AI)
 		all_hub = true;
 
@@ -677,6 +681,11 @@ void amdgpu_amdkfd_unreserve_memory_limit(struct amdgpu_bo *bo)
 {
 }
 
+int amdgpu_amdkfd_remove_fence_on_pt_pd_bos(struct amdgpu_bo *bo)
+{
+	return 0;
+}
+
 void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
 					struct amdgpu_vm *vm)
 {
@@ -713,11 +722,11 @@ void kgd2kfd_exit(void)
 {
 }
 
-void kgd2kfd_suspend(struct kfd_dev *kfd)
+void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm)
 {
 }
 
-int kgd2kfd_resume(struct kfd_dev *kfd)
+int kgd2kfd_resume(struct kfd_dev *kfd, bool run_pm)
 {
 	return 0;
 }

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

@@ -96,6 +96,7 @@ struct amdgpu_amdkfd_fence *amdgpu_amdkfd_fence_create(u64 context,
 						       struct mm_struct *mm);
 bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm);
 struct amdgpu_amdkfd_fence *to_amdgpu_amdkfd_fence(struct dma_fence *f);
+int amdgpu_amdkfd_remove_fence_on_pt_pd_bos(struct amdgpu_bo *bo);
 
 struct amdkfd_process_info {
 	/* List head of all VMs that belong to a KFD process */
@@ -122,8 +123,8 @@ struct amdkfd_process_info {
 int amdgpu_amdkfd_init(void);
 void amdgpu_amdkfd_fini(void);
 
-void amdgpu_amdkfd_suspend(struct amdgpu_device *adev);
+void amdgpu_amdkfd_suspend(struct amdgpu_device *adev, bool run_pm);
-int amdgpu_amdkfd_resume(struct amdgpu_device *adev);
+int amdgpu_amdkfd_resume(struct amdgpu_device *adev, bool run_pm);
 void amdgpu_amdkfd_interrupt(struct amdgpu_device *adev,
 			const void *ih_ring_entry);
 void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev);
@@ -171,6 +172,7 @@ int amdgpu_amdkfd_get_dmabuf_info(struct kgd_dev *kgd, int dma_buf_fd,
 				  uint32_t *flags);
 uint64_t amdgpu_amdkfd_get_vram_usage(struct kgd_dev *kgd);
 uint64_t amdgpu_amdkfd_get_hive_id(struct kgd_dev *kgd);
+uint64_t amdgpu_amdkfd_get_unique_id(struct kgd_dev *kgd);
 uint64_t amdgpu_amdkfd_get_mmio_remap_phys_addr(struct kgd_dev *kgd);
 uint32_t amdgpu_amdkfd_get_num_gws(struct kgd_dev *kgd);
 uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *src);
@@ -249,8 +251,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources);
 void kgd2kfd_device_exit(struct kfd_dev *kfd);
-void kgd2kfd_suspend(struct kfd_dev *kfd);
+void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm);
-int kgd2kfd_resume(struct kfd_dev *kfd);
+int kgd2kfd_resume(struct kfd_dev *kfd, bool run_pm);
 int kgd2kfd_pre_reset(struct kfd_dev *kfd);
 int kgd2kfd_post_reset(struct kfd_dev *kfd);
 void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry);

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v10.c

@@ -808,4 +808,5 @@ const struct kfd2kgd_calls gfx_v10_kfd2kgd = {
 	.get_tile_config = amdgpu_amdkfd_get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
 	.get_hive_id = amdgpu_amdkfd_get_hive_id,
+	.get_unique_id = amdgpu_amdkfd_get_unique_id,
 };

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c

@@ -739,4 +739,5 @@ const struct kfd2kgd_calls gfx_v9_kfd2kgd = {
 	.get_tile_config = kgd_gfx_v9_get_tile_config,
 	.set_vm_context_page_table_base = kgd_gfx_v9_set_vm_context_page_table_base,
 	.get_hive_id = amdgpu_amdkfd_get_hive_id,
+	.get_unique_id = amdgpu_amdkfd_get_unique_id,
 };

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

@@ -276,6 +276,42 @@ static int amdgpu_amdkfd_remove_eviction_fence(struct amdgpu_bo *bo,
 	return 0;
 }
 
+int amdgpu_amdkfd_remove_fence_on_pt_pd_bos(struct amdgpu_bo *bo)
+{
+	struct amdgpu_bo *root = bo;
+	struct amdgpu_vm_bo_base *vm_bo;
+	struct amdgpu_vm *vm;
+	struct amdkfd_process_info *info;
+	struct amdgpu_amdkfd_fence *ef;
+	int ret;
+
+	/* we can always get vm_bo from root PD bo.*/
+	while (root->parent)
+		root = root->parent;
+
+	vm_bo = root->vm_bo;
+	if (!vm_bo)
+		return 0;
+
+	vm = vm_bo->vm;
+	if (!vm)
+		return 0;
+
+	info = vm->process_info;
+	if (!info || !info->eviction_fence)
+		return 0;
+
+	ef = container_of(dma_fence_get(&info->eviction_fence->base),
+			struct amdgpu_amdkfd_fence, base);
+
+	BUG_ON(!dma_resv_trylock(bo->tbo.base.resv));
+	ret = amdgpu_amdkfd_remove_eviction_fence(bo, ef);
+	dma_resv_unlock(bo->tbo.base.resv);
+
+	dma_fence_put(&ef->base);
+	return ret;
+}
+
 static int amdgpu_amdkfd_bo_validate(struct amdgpu_bo *bo, uint32_t domain,
 				     bool wait)
 {
@@ -847,9 +883,9 @@ static int process_sync_pds_resv(struct amdkfd_process_info *process_info,
 			    vm_list_node) {
 		struct amdgpu_bo *pd = peer_vm->root.base.bo;
 
-		ret = amdgpu_sync_resv(NULL,
+		ret = amdgpu_sync_resv(NULL, sync, pd->tbo.base.resv,
-					sync, pd->tbo.base.resv,
+				       AMDGPU_SYNC_NE_OWNER,
-					AMDGPU_FENCE_OWNER_KFD, false);
+				       AMDGPU_FENCE_OWNER_KFD);
 		if (ret)
 			return ret;
 	}
@@ -1044,6 +1080,8 @@ void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
 	list_del(&vm->vm_list_node);
 	mutex_unlock(&process_info->lock);
 
+	vm->process_info = NULL;
+
 	/* Release per-process resources when last compute VM is destroyed */
 	if (!process_info->n_vms) {
 		WARN_ON(!list_empty(&process_info->kfd_bo_list));

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

@@ -1461,6 +1461,20 @@ static enum drm_mode_status amdgpu_connector_dp_mode_valid(struct drm_connector
 	return MODE_OK;
 }
 
+static int
+amdgpu_connector_late_register(struct drm_connector *connector)
+{
+	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
+	int r = 0;
+
+	if (amdgpu_connector->ddc_bus->has_aux) {
+		amdgpu_connector->ddc_bus->aux.dev = amdgpu_connector->base.kdev;
+		r = drm_dp_aux_register(&amdgpu_connector->ddc_bus->aux);
+	}
+
+	return r;
+}
+
 static const struct drm_connector_helper_funcs amdgpu_connector_dp_helper_funcs = {
 	.get_modes = amdgpu_connector_dp_get_modes,
 	.mode_valid = amdgpu_connector_dp_mode_valid,
@@ -1475,6 +1489,7 @@ static const struct drm_connector_funcs amdgpu_connector_dp_funcs = {
 	.early_unregister = amdgpu_connector_unregister,
 	.destroy = amdgpu_connector_destroy,
 	.force = amdgpu_connector_dvi_force,
+	.late_register = amdgpu_connector_late_register,
 };
 
 static const struct drm_connector_funcs amdgpu_connector_edp_funcs = {
@@ -1485,6 +1500,7 @@ static const struct drm_connector_funcs amdgpu_connector_edp_funcs = {
 	.early_unregister = amdgpu_connector_unregister,
 	.destroy = amdgpu_connector_destroy,
 	.force = amdgpu_connector_dvi_force,
+	.late_register = amdgpu_connector_late_register,
 };
 
 void
@@ -1931,7 +1947,6 @@ amdgpu_connector_add(struct amdgpu_device *adev,
 		connector->polled = DRM_CONNECTOR_POLL_HPD;
 
 	connector->display_info.subpixel_order = subpixel_order;
-	drm_connector_register(connector);
 
 	if (has_aux)
 		amdgpu_atombios_dp_aux_init(amdgpu_connector);

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

@@ -654,16 +654,19 @@ out:
 
 static int amdgpu_cs_sync_rings(struct amdgpu_cs_parser *p)
 {
+	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
 	struct amdgpu_bo_list_entry *e;
 	int r;
 
 	list_for_each_entry(e, &p->validated, tv.head) {
 		struct amdgpu_bo *bo = ttm_to_amdgpu_bo(e->tv.bo);
 		struct dma_resv *resv = bo->tbo.base.resv;
+		enum amdgpu_sync_mode sync_mode;
 
-		r = amdgpu_sync_resv(p->adev, &p->job->sync, resv, p->filp,
+		sync_mode = amdgpu_bo_explicit_sync(bo) ?
-				     amdgpu_bo_explicit_sync(bo));
+			AMDGPU_SYNC_EXPLICIT : AMDGPU_SYNC_NE_OWNER;
-
+		r = amdgpu_sync_resv(p->adev, &p->job->sync, resv, sync_mode,
+				     &fpriv->vm);
 		if (r)
 			return r;
 	}
@@ -1214,7 +1217,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	job = p->job;
 	p->job = NULL;
 
-	r = drm_sched_job_init(&job->base, entity, p->filp);
+	r = drm_sched_job_init(&job->base, entity, &fpriv->vm);
 	if (r)
 		goto error_unlock;
 

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

@@ -31,6 +31,8 @@
 #include <drm/drm_debugfs.h>
 
 #include "amdgpu.h"
+#include "amdgpu_pm.h"
+#include "amdgpu_dm_debugfs.h"
 
 /**
  * amdgpu_debugfs_add_files - Add simple debugfs entries
@@ -840,6 +842,55 @@ err:
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_gfxoff_write - Enable/disable GFXOFF
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * Write a 32-bit zero to disable or a 32-bit non-zero to enable
+ */
+static ssize_t amdgpu_debugfs_gfxoff_write(struct file *f, const char __user *buf,
+					 size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = file_inode(f)->i_private;
+	ssize_t result = 0;
+	int r;
+
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	r = pm_runtime_get_sync(adev->ddev->dev);
+	if (r < 0)
+		return r;
+
+	while (size) {
+		uint32_t value;
+
+		r = get_user(value, (uint32_t *)buf);
+		if (r) {
+			pm_runtime_mark_last_busy(adev->ddev->dev);
+			pm_runtime_put_autosuspend(adev->ddev->dev);
+			return r;
+		}
+
+		amdgpu_gfx_off_ctrl(adev, value ? true : false);
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	pm_runtime_mark_last_busy(adev->ddev->dev);
+	pm_runtime_put_autosuspend(adev->ddev->dev);
+
+	return result;
+}
+
+
 static const struct file_operations amdgpu_debugfs_regs_fops = {
 	.owner = THIS_MODULE,
 	.read = amdgpu_debugfs_regs_read,
@@ -888,6 +939,11 @@ static const struct file_operations amdgpu_debugfs_gpr_fops = {
 	.llseek = default_llseek
 };
 
+static const struct file_operations amdgpu_debugfs_gfxoff_fops = {
+	.owner = THIS_MODULE,
+	.write = amdgpu_debugfs_gfxoff_write,
+};
+
 static const struct file_operations *debugfs_regs[] = {
 	&amdgpu_debugfs_regs_fops,
 	&amdgpu_debugfs_regs_didt_fops,
@@ -897,6 +953,7 @@ static const struct file_operations *debugfs_regs[] = {
 	&amdgpu_debugfs_sensors_fops,
 	&amdgpu_debugfs_wave_fops,
 	&amdgpu_debugfs_gpr_fops,
+	&amdgpu_debugfs_gfxoff_fops,
 };
 
 static const char *debugfs_regs_names[] = {
@@ -908,6 +965,7 @@ static const char *debugfs_regs_names[] = {
 	"amdgpu_sensors",
 	"amdgpu_wave",
 	"amdgpu_gpr",
+	"amdgpu_gfxoff",
 };
 
 /**
@@ -1216,6 +1274,8 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_ib_preempt, NULL,
 
 int amdgpu_debugfs_init(struct amdgpu_device *adev)
 {
+	int r, i;
+
 	adev->debugfs_preempt =
 		debugfs_create_file("amdgpu_preempt_ib", 0600,
 				    adev->ddev->primary->debugfs_root, adev,
@@ -1225,12 +1285,73 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 		return -EIO;
 	}
 
+	/* Register debugfs entries for amdgpu_ttm */
+	r = amdgpu_ttm_debugfs_init(adev);
+	if (r) {
+		DRM_ERROR("Failed to init debugfs\n");
+		return r;
+	}
+
+	r = amdgpu_debugfs_pm_init(adev);
+	if (r) {
+		DRM_ERROR("Failed to register debugfs file for dpm!\n");
+		return r;
+	}
+
+	if (amdgpu_debugfs_sa_init(adev)) {
+		dev_err(adev->dev, "failed to register debugfs file for SA\n");
+	}
+
+	if (amdgpu_debugfs_fence_init(adev))
+		dev_err(adev->dev, "fence debugfs file creation failed\n");
+
+	r = amdgpu_debugfs_gem_init(adev);
+	if (r)
+		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
+
+	r = amdgpu_debugfs_regs_init(adev);
+	if (r)
+		DRM_ERROR("registering register debugfs failed (%d).\n", r);
+
+	r = amdgpu_debugfs_firmware_init(adev);
+	if (r)
+		DRM_ERROR("registering firmware debugfs failed (%d).\n", r);
+
+#if defined(CONFIG_DRM_AMD_DC)
+	if (amdgpu_device_has_dc_support(adev)) {
+		if (dtn_debugfs_init(adev))
+			DRM_ERROR("amdgpu: failed initialize dtn debugfs support.\n");
+	}
+#endif
+
+	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
+		struct amdgpu_ring *ring = adev->rings[i];
+
+		if (!ring)
+			continue;
+
+		if (amdgpu_debugfs_ring_init(adev, ring)) {
+			DRM_ERROR("Failed to register debugfs file for rings !\n");
+		}
+	}
+
 	return amdgpu_debugfs_add_files(adev, amdgpu_debugfs_list,
 					ARRAY_SIZE(amdgpu_debugfs_list));
 }
 
-void amdgpu_debugfs_preempt_cleanup(struct amdgpu_device *adev)
+void amdgpu_debugfs_fini(struct amdgpu_device *adev)
 {
+	int i;
+
+	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
+		struct amdgpu_ring *ring = adev->rings[i];
+
+		if (!ring)
+			continue;
+
+		amdgpu_debugfs_ring_fini(ring);
+	}
+	amdgpu_ttm_debugfs_fini(adev);
 	debugfs_remove(adev->debugfs_preempt);
 }
 
@@ -1239,7 +1360,7 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 {
 	return 0;
 }
-void amdgpu_debugfs_preempt_cleanup(struct amdgpu_device *adev) { }
+void amdgpu_debugfs_fini(struct amdgpu_device *adev) { }
 int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
 {
 	return 0;

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

@@ -34,7 +34,7 @@ struct amdgpu_debugfs {
 int amdgpu_debugfs_regs_init(struct amdgpu_device *adev);
 void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev);
 int amdgpu_debugfs_init(struct amdgpu_device *adev);
-void amdgpu_debugfs_preempt_cleanup(struct amdgpu_device *adev);
+void amdgpu_debugfs_fini(struct amdgpu_device *adev);
 int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
 			     const struct drm_info_list *files,
 			     unsigned nfiles);

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

@@ -183,20 +183,51 @@ bool amdgpu_device_supports_baco(struct drm_device *dev)
 void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos,
 			       uint32_t *buf, size_t size, bool write)
 {
-	uint64_t last;
 	unsigned long flags;
+	uint32_t hi = ~0;
+	uint64_t last;
+
+
+#ifdef CONFIG_64BIT
+	last = min(pos + size, adev->gmc.visible_vram_size);
+	if (last > pos) {
+		void __iomem *addr = adev->mman.aper_base_kaddr + pos;
+		size_t count = last - pos;
+
+		if (write) {
+			memcpy_toio(addr, buf, count);
+			mb();
+			amdgpu_asic_flush_hdp(adev, NULL);
+		} else {
+			amdgpu_asic_invalidate_hdp(adev, NULL);
+			mb();
+			memcpy_fromio(buf, addr, count);
+		}
+
+		if (count == size)
+			return;
+
+		pos += count;
+		buf += count / 4;
+		size -= count;
+	}
+#endif
 
-	last = size - 4;
-	for (last += pos; pos <= last; pos += 4) {
 	spin_lock_irqsave(&adev->mmio_idx_lock, flags);
+	for (last = pos + size; pos < last; pos += 4) {
+		uint32_t tmp = pos >> 31;
+
 		WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)pos) | 0x80000000);
-		WREG32_NO_KIQ(mmMM_INDEX_HI, pos >> 31);
+		if (tmp != hi) {
+			WREG32_NO_KIQ(mmMM_INDEX_HI, tmp);
+			hi = tmp;
+		}
 		if (write)
 			WREG32_NO_KIQ(mmMM_DATA, *buf++);
 		else
 			*buf++ = RREG32_NO_KIQ(mmMM_DATA);
-		spin_unlock_irqrestore(&adev->mmio_idx_lock, flags);
 	}
+	spin_unlock_irqrestore(&adev->mmio_idx_lock, flags);
 }
 
 /*
@@ -2344,6 +2375,7 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 		}
 		adev->ip_blocks[i].status.hw = false;
 		/* handle putting the SMC in the appropriate state */
+		if(!amdgpu_sriov_vf(adev)){
 			if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_SMC) {
 				r = amdgpu_dpm_set_mp1_state(adev, adev->mp1_state);
 				if (r) {
@@ -2352,7 +2384,7 @@ static int amdgpu_device_ip_suspend_phase2(struct amdgpu_device *adev)
 					return r;
 				}
 			}
-
+		}
 		adev->ip_blocks[i].status.hw = false;
 	}
 
@@ -2800,7 +2832,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 
 	adev->usec_timeout = AMDGPU_MAX_USEC_TIMEOUT;
 	if (amdgpu_emu_mode == 1)
-		adev->usec_timeout *= 2;
+		adev->usec_timeout *= 10;
 	adev->gmc.gart_size = 512 * 1024 * 1024;
 	adev->accel_working = false;
 	adev->num_rings = 0;
@@ -3088,22 +3120,6 @@ fence_driver_init:
 	} else
 		adev->ucode_sysfs_en = true;
 
-	r = amdgpu_debugfs_gem_init(adev);
-	if (r)
-		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
-
-	r = amdgpu_debugfs_regs_init(adev);
-	if (r)
-		DRM_ERROR("registering register debugfs failed (%d).\n", r);
-
-	r = amdgpu_debugfs_firmware_init(adev);
-	if (r)
-		DRM_ERROR("registering firmware debugfs failed (%d).\n", r);
-
-	r = amdgpu_debugfs_init(adev);
-	if (r)
-		DRM_ERROR("Creating debugfs files failed (%d).\n", r);
-
 	if ((amdgpu_testing & 1)) {
 		if (adev->accel_working)
 			amdgpu_test_moves(adev);
@@ -3225,7 +3241,6 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 		amdgpu_ucode_sysfs_fini(adev);
 	if (IS_ENABLED(CONFIG_PERF_EVENTS))
 		amdgpu_pmu_fini(adev);
-	amdgpu_debugfs_preempt_cleanup(adev);
 	if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10)
 		amdgpu_discovery_fini(adev);
 }
@@ -3309,7 +3324,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon)
 		}
 	}
 
-	amdgpu_amdkfd_suspend(adev);
+	amdgpu_amdkfd_suspend(adev, !fbcon);
 
 	amdgpu_ras_suspend(adev);
 
@@ -3393,7 +3408,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon)
 			}
 		}
 	}
-	r = amdgpu_amdkfd_resume(adev);
+	r = amdgpu_amdkfd_resume(adev, !fbcon);
 	if (r)
 		return r;
 

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

@@ -307,7 +307,7 @@ int amdgpu_discovery_reg_base_init(struct amdgpu_device *adev)
 
 			for (hw_ip = 0; hw_ip < MAX_HWIP; hw_ip++) {
 				if (hw_id_map[hw_ip] == le16_to_cpu(ip->hw_id)) {
-					DRM_INFO("set register base offset for %s\n",
+					DRM_DEBUG("set register base offset for %s\n",
 							hw_id_names[le16_to_cpu(ip->hw_id)]);
 					adev->reg_offset[hw_ip][ip->number_instance] =
 						ip->base_address;

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

@@ -1171,3 +1171,20 @@ int amdgpu_dpm_set_xgmi_pstate(struct amdgpu_device *adev,
 
 	return ret;
 }
+
+int amdgpu_dpm_set_df_cstate(struct amdgpu_device *adev,
+			     uint32_t cstate)
+{
+	int ret = 0;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+	void *pp_handle = adev->powerplay.pp_handle;
+	struct smu_context *smu = &adev->smu;
+
+	if (is_support_sw_smu(adev))
+		ret = smu_set_df_cstate(smu, cstate);
+	else if (pp_funcs &&
+		 pp_funcs->set_df_cstate)
+		ret = pp_funcs->set_df_cstate(pp_handle, cstate);
+
+	return ret;
+}

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

@@ -533,4 +533,7 @@ int amdgpu_dpm_baco_exit(struct amdgpu_device *adev);
 
 int amdgpu_dpm_baco_enter(struct amdgpu_device *adev);
 
+int amdgpu_dpm_set_df_cstate(struct amdgpu_device *adev,
+			     uint32_t cstate);
+
 #endif

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

@@ -1021,6 +1021,7 @@ static int amdgpu_pci_probe(struct pci_dev *pdev,
 			    const struct pci_device_id *ent)
 {
 	struct drm_device *dev;
+	struct amdgpu_device *adev;
 	unsigned long flags = ent->driver_data;
 	int ret, retry = 0;
 	bool supports_atomic = false;
@@ -1090,6 +1091,8 @@ static int amdgpu_pci_probe(struct pci_dev *pdev,
 
 	pci_set_drvdata(pdev, dev);
 
+	amdgpu_driver_load_kms(dev, ent->driver_data);
+
 retry_init:
 	ret = drm_dev_register(dev, ent->driver_data);
 	if (ret == -EAGAIN && ++retry <= 3) {
@@ -1100,6 +1103,11 @@ retry_init:
 	} else if (ret)
 		goto err_pci;
 
+	adev = dev->dev_private;
+	ret = amdgpu_debugfs_init(adev);
+	if (ret)
+		DRM_ERROR("Creating debugfs files failed (%d).\n", ret);
+
 	return 0;
 
 err_pci:
@@ -1113,6 +1121,7 @@ static void
 amdgpu_pci_remove(struct pci_dev *pdev)
 {
 	struct drm_device *dev = pci_get_drvdata(pdev);
+	struct amdgpu_device *adev = dev->dev_private;
 
 #ifdef MODULE
 	if (THIS_MODULE->state != MODULE_STATE_GOING)
@@ -1120,6 +1129,8 @@ amdgpu_pci_remove(struct pci_dev *pdev)
 		DRM_ERROR("Hotplug removal is not supported\n");
 	drm_dev_unplug(dev);
 	drm_dev_put(dev);
+	amdgpu_debugfs_fini(adev);
+	amdgpu_driver_unload_kms(dev);
 	pci_disable_device(pdev);
 	pci_set_drvdata(pdev, NULL);
 }
@@ -1220,11 +1231,15 @@ static int amdgpu_pmops_runtime_suspend(struct device *dev)
 		}
 	}
 
+	adev->in_runpm = true;
 	if (amdgpu_device_supports_boco(drm_dev))
 		drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING;
 	drm_kms_helper_poll_disable(drm_dev);
 
 	ret = amdgpu_device_suspend(drm_dev, false);
+	if (ret)
+		return ret;
+
 	if (amdgpu_device_supports_boco(drm_dev)) {
 		/* Only need to handle PCI state in the driver for ATPX
 		 * PCI core handles it for _PR3.
@@ -1278,6 +1293,7 @@ static int amdgpu_pmops_runtime_resume(struct device *dev)
 	drm_kms_helper_poll_enable(drm_dev);
 	if (amdgpu_device_supports_boco(drm_dev))
 		drm_dev->switch_power_state = DRM_SWITCH_POWER_ON;
+	adev->in_runpm = false;
 	return 0;
 }
 
@@ -1379,15 +1395,13 @@ int amdgpu_file_to_fpriv(struct file *filp, struct amdgpu_fpriv **fpriv)
 
 static struct drm_driver kms_driver = {
 	.driver_features =
-	    DRIVER_USE_AGP | DRIVER_ATOMIC |
+	    DRIVER_ATOMIC |
 	    DRIVER_GEM |
 	    DRIVER_RENDER | DRIVER_MODESET | DRIVER_SYNCOBJ |
 	    DRIVER_SYNCOBJ_TIMELINE,
-	.load = amdgpu_driver_load_kms,
 	.open = amdgpu_driver_open_kms,
 	.postclose = amdgpu_driver_postclose_kms,
 	.lastclose = amdgpu_driver_lastclose_kms,
-	.unload = amdgpu_driver_unload_kms,
 	.irq_handler = amdgpu_irq_handler,
 	.ioctls = amdgpu_ioctls_kms,
 	.gem_free_object_unlocked = amdgpu_gem_object_free,

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

@@ -503,9 +503,6 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
  */
 int amdgpu_fence_driver_init(struct amdgpu_device *adev)
 {
-	if (amdgpu_debugfs_fence_init(adev))
-		dev_err(adev->dev, "fence debugfs file creation failed\n");
-
 	return 0;
 }
 

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

@@ -477,7 +477,7 @@ int amdgpu_gfx_disable_kcq(struct amdgpu_device *adev)
 		kiq->pmf->kiq_unmap_queues(kiq_ring, &adev->gfx.compute_ring[i],
 					   RESET_QUEUES, 0, 0);
 
-	return amdgpu_ring_test_ring(kiq_ring);
+	return amdgpu_ring_test_helper(kiq_ring);
 }
 
 int amdgpu_gfx_enable_kcq(struct amdgpu_device *adev)

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

@@ -195,6 +195,7 @@ struct amdgpu_gmc {
 	uint32_t                srbm_soft_reset;
 	bool			prt_warning;
 	uint64_t		stolen_size;
+	uint32_t		sdpif_register;
 	/* apertures */
 	u64			shared_aperture_start;
 	u64			shared_aperture_end;

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

@@ -48,7 +48,6 @@
  * produce command buffers which are send to the kernel and
  * put in IBs for execution by the requested ring.
  */
-static int amdgpu_debugfs_sa_init(struct amdgpu_device *adev);
 
 /**
  * amdgpu_ib_get - request an IB (Indirect Buffer)
@@ -295,9 +294,7 @@ int amdgpu_ib_pool_init(struct amdgpu_device *adev)
 	}
 
 	adev->ib_pool_ready = true;
-	if (amdgpu_debugfs_sa_init(adev)) {
+
-		dev_err(adev->dev, "failed to register debugfs file for SA\n");
-	}
 	return 0;
 }
 
@@ -421,7 +418,7 @@ static const struct drm_info_list amdgpu_debugfs_sa_list[] = {
 
 #endif
 
-static int amdgpu_debugfs_sa_init(struct amdgpu_device *adev)
+int amdgpu_debugfs_sa_init(struct amdgpu_device *adev)
 {
 #if defined(CONFIG_DEBUG_FS)
 	return amdgpu_debugfs_add_files(adev, amdgpu_debugfs_sa_list, 1);

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

@@ -170,10 +170,16 @@ int amdgpu_driver_load_kms(struct drm_device *dev, unsigned long flags)
 	}
 
 	if (amdgpu_device_supports_boco(dev) &&
-	    (amdgpu_runtime_pm != 0)) /* enable runpm by default */
+	    (amdgpu_runtime_pm != 0)) /* enable runpm by default for boco */
 		adev->runpm = true;
 	else if (amdgpu_device_supports_baco(dev) &&
-		 (amdgpu_runtime_pm > 0)) /* enable runpm if runpm=1 */
+		 (amdgpu_runtime_pm != 0) &&
+		 (adev->asic_type >= CHIP_TOPAZ) &&
+		 (adev->asic_type != CHIP_VEGA20) &&
+		 (adev->asic_type != CHIP_ARCTURUS)) /* enable runpm on VI+ */
+		adev->runpm = true;
+	else if (amdgpu_device_supports_baco(dev) &&
+		 (amdgpu_runtime_pm > 0))  /* enable runpm if runpm=1 on CI */
 		adev->runpm = true;
 
 	/* Call ACPI methods: require modeset init

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

@@ -1318,6 +1318,11 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo)
 	if (abo->kfd_bo)
 		amdgpu_amdkfd_unreserve_memory_limit(abo);
 
+	/* We only remove the fence if the resv has individualized. */
+	WARN_ON_ONCE(bo->base.resv != &bo->base._resv);
+	if (bo->base.resv == &bo->base._resv)
+		amdgpu_amdkfd_remove_fence_on_pt_pd_bos(abo);
+
 	if (bo->mem.mem_type != TTM_PL_VRAM || !bo->mem.mm_node ||
 	    !(abo->flags & AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE))
 		return;
@@ -1414,27 +1419,49 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence,
 }
 
 /**
- * amdgpu_sync_wait_resv - Wait for BO reservation fences
+ * amdgpu_bo_sync_wait_resv - Wait for BO reservation fences
  *
- * @bo: buffer object
+ * @adev: amdgpu device pointer
+ * @resv: reservation object to sync to
+ * @sync_mode: synchronization mode
  * @owner: fence owner
  * @intr: Whether the wait is interruptible
  *
+ * Extract the fences from the reservation object and waits for them to finish.
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
+ */
+int amdgpu_bo_sync_wait_resv(struct amdgpu_device *adev, struct dma_resv *resv,
+			     enum amdgpu_sync_mode sync_mode, void *owner,
+			     bool intr)
+{
+	struct amdgpu_sync sync;
+	int r;
+
+	amdgpu_sync_create(&sync);
+	amdgpu_sync_resv(adev, &sync, resv, sync_mode, owner);
+	r = amdgpu_sync_wait(&sync, intr);
+	amdgpu_sync_free(&sync);
+	return r;
+}
+
+/**
+ * amdgpu_bo_sync_wait - Wrapper for amdgpu_bo_sync_wait_resv
+ * @bo: buffer object to wait for
+ * @owner: fence owner
+ * @intr: Whether the wait is interruptible
+ *
+ * Wrapper to wait for fences in a BO.
  * Returns:
  * 0 on success, errno otherwise.
  */
 int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	struct amdgpu_sync sync;
-	int r;
 
-	amdgpu_sync_create(&sync);
+	return amdgpu_bo_sync_wait_resv(adev, bo->tbo.base.resv,
-	amdgpu_sync_resv(adev, &sync, bo->tbo.base.resv, owner, false);
+					AMDGPU_SYNC_NE_OWNER, owner, intr);
-	r = amdgpu_sync_wait(&sync, intr);
-	amdgpu_sync_free(&sync);
-
-	return r;
 }
 
 /**

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

@@ -277,6 +277,9 @@ void amdgpu_bo_release_notify(struct ttm_buffer_object *bo);
 int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo);
 void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence,
 		     bool shared);
+int amdgpu_bo_sync_wait_resv(struct amdgpu_device *adev, struct dma_resv *resv,
+			     enum amdgpu_sync_mode sync_mode, void *owner,
+			     bool intr);
 int amdgpu_bo_sync_wait(struct amdgpu_bo *bo, void *owner, bool intr);
 u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo);
 int amdgpu_bo_validate(struct amdgpu_bo *bo);
@@ -316,6 +319,7 @@ void amdgpu_sa_bo_free(struct amdgpu_device *adev,
 void amdgpu_sa_bo_dump_debug_info(struct amdgpu_sa_manager *sa_manager,
 					 struct seq_file *m);
 #endif
+int amdgpu_debugfs_sa_init(struct amdgpu_device *adev);
 
 bool amdgpu_bo_support_uswc(u64 bo_flags);
 

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

@@ -41,8 +41,6 @@
 #include "hwmgr.h"
 #define WIDTH_4K 3840
 
-static int amdgpu_debugfs_pm_init(struct amdgpu_device *adev);
-
 static const struct cg_flag_name clocks[] = {
 	{AMD_CG_SUPPORT_GFX_MGCG, "Graphics Medium Grain Clock Gating"},
 	{AMD_CG_SUPPORT_GFX_MGLS, "Graphics Medium Grain memory Light Sleep"},
@@ -3398,11 +3396,6 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 		DRM_ERROR("failed to create device file unique_id\n");
 		return ret;
 	}
-	ret = amdgpu_debugfs_pm_init(adev);
-	if (ret) {
-		DRM_ERROR("Failed to register debugfs file for dpm!\n");
-		return ret;
-	}
 
 	if ((adev->asic_type >= CHIP_VEGA10) &&
 	    !(adev->flags & AMD_IS_APU)) {
@@ -3669,7 +3662,7 @@ static const struct drm_info_list amdgpu_pm_info_list[] = {
 };
 #endif
 
-static int amdgpu_debugfs_pm_init(struct amdgpu_device *adev)
+int amdgpu_debugfs_pm_init(struct amdgpu_device *adev)
 {
 #if defined(CONFIG_DEBUG_FS)
 	return amdgpu_debugfs_add_files(adev, amdgpu_pm_info_list, ARRAY_SIZE(amdgpu_pm_info_list));

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

@@ -43,4 +43,6 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable);
 void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable);
 void amdgpu_dpm_enable_jpeg(struct amdgpu_device *adev, bool enable);
 
+int amdgpu_debugfs_pm_init(struct amdgpu_device *adev);
+
 #endif

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

@@ -38,6 +38,39 @@
 
 static void psp_set_funcs(struct amdgpu_device *adev);
 
+/*
+ * Due to DF Cstate management centralized to PMFW, the firmware
+ * loading sequence will be updated as below:
+ *   - Load KDB
+ *   - Load SYS_DRV
+ *   - Load tOS
+ *   - Load PMFW
+ *   - Setup TMR
+ *   - Load other non-psp fw
+ *   - Load ASD
+ *   - Load XGMI/RAS/HDCP/DTM TA if any
+ *
+ * This new sequence is required for
+ *   - Arcturus
+ *   - Navi12 and onwards
+ */
+static void psp_check_pmfw_centralized_cstate_management(struct psp_context *psp)
+{
+	struct amdgpu_device *adev = psp->adev;
+
+	psp->pmfw_centralized_cstate_management = false;
+
+	if (amdgpu_sriov_vf(adev))
+		return;
+
+	if (adev->flags & AMD_IS_APU)
+		return;
+
+	if ((adev->asic_type == CHIP_ARCTURUS) ||
+	    (adev->asic_type >= CHIP_NAVI12))
+		psp->pmfw_centralized_cstate_management = true;
+}
+
 static int psp_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -75,6 +108,8 @@ static int psp_early_init(void *handle)
 
 	psp->adev = adev;
 
+	psp_check_pmfw_centralized_cstate_management(psp);
+
 	return 0;
 }
 
@@ -558,7 +593,7 @@ int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 	return psp_ta_invoke(psp, ta_cmd_id, psp->xgmi_context.session_id);
 }
 
-static int psp_xgmi_terminate(struct psp_context *psp)
+int psp_xgmi_terminate(struct psp_context *psp)
 {
 	int ret;
 
@@ -579,7 +614,7 @@ static int psp_xgmi_terminate(struct psp_context *psp)
 	return 0;
 }
 
-static int psp_xgmi_initialize(struct psp_context *psp)
+int psp_xgmi_initialize(struct psp_context *psp)
 {
 	struct ta_xgmi_shared_memory *xgmi_cmd;
 	int ret;
@@ -1013,6 +1048,30 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	return 0;
 }
 
+static int psp_dtm_unload(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	/*
+	 * TODO: bypass the unloading in sriov for now
+	 */
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_ta_unload_cmd_buf(cmd, psp->dtm_context.session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
 int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 {
 	/*
@@ -1037,7 +1096,7 @@ static int psp_dtm_terminate(struct psp_context *psp)
 	if (!psp->dtm_context.dtm_initialized)
 		return 0;
 
-	ret = psp_hdcp_unload(psp);
+	ret = psp_dtm_unload(psp);
 	if (ret)
 		return ret;
 
@@ -1057,7 +1116,7 @@ static int psp_hw_start(struct psp_context *psp)
 	struct amdgpu_device *adev = psp->adev;
 	int ret;
 
-	if (!amdgpu_sriov_vf(adev) || !adev->in_gpu_reset) {
+	if (!amdgpu_sriov_vf(adev)) {
 		if (psp->kdb_bin_size &&
 		    (psp->funcs->bootloader_load_kdb != NULL)) {
 			ret = psp_bootloader_load_kdb(psp);
@@ -1092,11 +1151,18 @@ static int psp_hw_start(struct psp_context *psp)
 		return ret;
 	}
 
+	/*
+	 * For those ASICs with DF Cstate management centralized
+	 * to PMFW, TMR setup should be performed after PMFW
+	 * loaded and before other non-psp firmware loaded.
+	 */
+	if (!psp->pmfw_centralized_cstate_management) {
 		ret = psp_tmr_load(psp);
 		if (ret) {
 			DRM_ERROR("PSP load tmr failed!\n");
 			return ret;
 		}
+	}
 
 	return 0;
 }
@@ -1292,9 +1358,10 @@ static int psp_np_fw_load(struct psp_context *psp)
 	struct amdgpu_firmware_info *ucode;
 	struct amdgpu_device* adev = psp->adev;
 
-	if (psp->autoload_supported) {
+	if (psp->autoload_supported ||
+	    psp->pmfw_centralized_cstate_management) {
 		ucode = &adev->firmware.ucode[AMDGPU_UCODE_ID_SMC];
-		if (!ucode->fw)
+		if (!ucode->fw || amdgpu_sriov_vf(adev))
 			goto out;
 
 		ret = psp_execute_np_fw_load(psp, ucode);
@@ -1302,6 +1369,14 @@ static int psp_np_fw_load(struct psp_context *psp)
 			return ret;
 	}
 
+	if (psp->pmfw_centralized_cstate_management) {
+		ret = psp_tmr_load(psp);
+		if (ret) {
+			DRM_ERROR("PSP load tmr failed!\n");
+			return ret;
+		}
+	}
+
 out:
 	for (i = 0; i < adev->firmware.max_ucodes; i++) {
 		ucode = &adev->firmware.ucode[i];
@@ -1309,7 +1384,9 @@ out:
 			continue;
 
 		if (ucode->ucode_id == AMDGPU_UCODE_ID_SMC &&
-		    (psp_smu_reload_quirk(psp) || psp->autoload_supported))
+		    (psp_smu_reload_quirk(psp) ||
+		     psp->autoload_supported ||
+		     psp->pmfw_centralized_cstate_management))
 			continue;
 
 		if (amdgpu_sriov_vf(adev) &&
@@ -1420,16 +1497,6 @@ skip_memalloc:
 		return ret;
 	}
 
-	if (adev->gmc.xgmi.num_physical_nodes > 1) {
-		ret = psp_xgmi_initialize(psp);
-		/* Warning the XGMI seesion initialize failure
-		 * Instead of stop driver initialization
-		 */
-		if (ret)
-			dev_err(psp->adev->dev,
-				"XGMI: Failed to initialize XGMI session\n");
-	}
-
 	if (psp->adev->psp.ta_fw) {
 		ret = psp_ras_initialize(psp);
 		if (ret)
@@ -1494,10 +1561,6 @@ static int psp_hw_fini(void *handle)
 	void *tmr_buf;
 	void **pptr;
 
-	if (adev->gmc.xgmi.num_physical_nodes > 1 &&
-	    psp->xgmi_context.initialized == 1)
-                psp_xgmi_terminate(psp);
-
 	if (psp->adev->psp.ta_fw) {
 		psp_ras_terminate(psp);
 		psp_dtm_terminate(psp);

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

@@ -264,6 +264,8 @@ struct psp_context
 	atomic_t			fence_value;
 	/* flag to mark whether gfx fw autoload is supported or not */
 	bool				autoload_supported;
+	/* flag to mark whether df cstate management centralized to PMFW */
+	bool				pmfw_centralized_cstate_management;
 
 	/* xgmi ta firmware and buffer */
 	const struct firmware		*ta_fw;
@@ -362,6 +364,8 @@ int psp_gpu_reset(struct amdgpu_device *adev);
 int psp_update_vcn_sram(struct amdgpu_device *adev, int inst_idx,
 			uint64_t cmd_gpu_addr, int cmd_size);
 
+int psp_xgmi_initialize(struct psp_context *psp);
+int psp_xgmi_terminate(struct psp_context *psp);
 int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id);
 
 int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id);

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

@@ -31,6 +31,7 @@
 #include "amdgpu.h"
 #include "amdgpu_ras.h"
 #include "amdgpu_atomfirmware.h"
+#include "amdgpu_xgmi.h"
 #include "ivsrcid/nbio/irqsrcs_nbif_7_4.h"
 
 const char *ras_error_string[] = {
@@ -742,20 +743,6 @@ int amdgpu_ras_error_query(struct amdgpu_device *adev,
 	return 0;
 }
 
-uint64_t get_xgmi_relative_phy_addr(struct amdgpu_device *adev, uint64_t addr)
-{
-	uint32_t df_inst_id;
-
-	if ((!adev->df.funcs)                 ||
-	    (!adev->df.funcs->get_df_inst_id) ||
-	    (!adev->df.funcs->get_dram_base_addr))
-		return addr;
-
-	df_inst_id = adev->df.funcs->get_df_inst_id(adev);
-
-	return addr + adev->df.funcs->get_dram_base_addr(adev, df_inst_id);
-}
-
 /* wrapper of psp_ras_trigger_error */
 int amdgpu_ras_error_inject(struct amdgpu_device *adev,
 		struct ras_inject_if *info)
@@ -775,7 +762,8 @@ int amdgpu_ras_error_inject(struct amdgpu_device *adev,
 
 	/* Calculate XGMI relative offset */
 	if (adev->gmc.xgmi.num_physical_nodes > 1) {
-		block_info.address = get_xgmi_relative_phy_addr(adev,
+		block_info.address =
+			amdgpu_xgmi_get_relative_phy_addr(adev,
 							  block_info.address);
 	}
 
@@ -1319,6 +1307,33 @@ static int amdgpu_ras_interrupt_remove_all(struct amdgpu_device *adev)
 }
 /* ih end */
 
+/* traversal all IPs except NBIO to query error counter */
+static void amdgpu_ras_log_on_err_counter(struct amdgpu_device *adev)
+{
+	struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+	struct ras_manager *obj;
+
+	if (!con)
+		return;
+
+	list_for_each_entry(obj, &con->head, node) {
+		struct ras_query_if info = {
+			.head = obj->head,
+		};
+
+		/*
+		 * PCIE_BIF IP has one different isr by ras controller
+		 * interrupt, the specific ras counter query will be
+		 * done in that isr. So skip such block from common
+		 * sync flood interrupt isr calling.
+		 */
+		if (info.head.block == AMDGPU_RAS_BLOCK__PCIE_BIF)
+			continue;
+
+		amdgpu_ras_error_query(adev, &info);
+	}
+}
+
 /* recovery begin */
 
 /* return 0 on success.
@@ -1373,6 +1388,12 @@ static void amdgpu_ras_do_recovery(struct work_struct *work)
 	struct amdgpu_ras *ras =
 		container_of(work, struct amdgpu_ras, recovery_work);
 
+	/*
+	 * Query and print non zero error counter per IP block for
+	 * awareness before recovering GPU.
+	 */
+	amdgpu_ras_log_on_err_counter(ras->adev);
+
 	if (amdgpu_device_should_recover_gpu(ras->adev))
 		amdgpu_device_gpu_recover(ras->adev, 0);
 	atomic_set(&ras->in_recovery, 0);

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

@@ -26,9 +26,11 @@
 #include "amdgpu_ras.h"
 #include <linux/bits.h>
 #include "smu_v11_0_i2c.h"
+#include "atom.h"
 
-#define EEPROM_I2C_TARGET_ADDR_ARCTURUS  0xA8
 #define EEPROM_I2C_TARGET_ADDR_VEGA20    	0xA0
+#define EEPROM_I2C_TARGET_ADDR_ARCTURUS  	0xA8
+#define EEPROM_I2C_TARGET_ADDR_ARCTURUS_D342  	0xA0
 
 /*
  * The 2 macros bellow represent the actual size in bytes that
@@ -55,6 +57,45 @@
 
 #define to_amdgpu_device(x) (container_of(x, struct amdgpu_ras, eeprom_control))->adev
 
+static bool __get_eeprom_i2c_addr_arct(struct amdgpu_device *adev,
+				       uint16_t *i2c_addr)
+{
+	struct atom_context *atom_ctx = adev->mode_info.atom_context;
+
+	if (!i2c_addr || !atom_ctx)
+		return false;
+
+	if (strnstr(atom_ctx->vbios_version,
+	            "D342",
+		    sizeof(atom_ctx->vbios_version)))
+		*i2c_addr = EEPROM_I2C_TARGET_ADDR_ARCTURUS_D342;
+	else
+		*i2c_addr = EEPROM_I2C_TARGET_ADDR_ARCTURUS;
+
+	return true;
+}
+
+static bool __get_eeprom_i2c_addr(struct amdgpu_device *adev,
+				  uint16_t *i2c_addr)
+{
+	if (!i2c_addr)
+		return false;
+
+	switch (adev->asic_type) {
+	case CHIP_VEGA20:
+		*i2c_addr = EEPROM_I2C_TARGET_ADDR_VEGA20;
+		break;
+
+	case CHIP_ARCTURUS:
+		return __get_eeprom_i2c_addr_arct(adev, i2c_addr);
+
+	default:
+		return false;
+	}
+
+	return true;
+}
+
 static void __encode_table_header_to_buff(struct amdgpu_ras_eeprom_table_header *hdr,
 					  unsigned char *buff)
 {
@@ -103,8 +144,6 @@ static int __update_table_header(struct amdgpu_ras_eeprom_control *control,
 	return ret;
 }
 
-
-
 static uint32_t  __calc_hdr_byte_sum(struct amdgpu_ras_eeprom_control *control)
 {
 	int i;
@@ -212,16 +251,17 @@ int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control)
 			.buf	= buff,
 	};
 
+	if (!__get_eeprom_i2c_addr(adev, &control->i2c_address))
+		return -EINVAL;
+
 	mutex_init(&control->tbl_mutex);
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA20:
-		control->i2c_address = EEPROM_I2C_TARGET_ADDR_VEGA20;
 		ret = smu_v11_0_i2c_eeprom_control_init(&control->eeprom_accessor);
 		break;
 
 	case CHIP_ARCTURUS:
-		control->i2c_address = EEPROM_I2C_TARGET_ADDR_ARCTURUS;
 		ret = smu_i2c_eeprom_init(&adev->smu, &control->eeprom_accessor);
 		break;
 

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

@@ -48,9 +48,6 @@
  * wptr.  The GPU then starts fetching commands and executes
  * them until the pointers are equal again.
  */
-static int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
-				    struct amdgpu_ring *ring);
-static void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring);
 
 /**
  * amdgpu_ring_alloc - allocate space on the ring buffer
@@ -334,10 +331,6 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	for (i = 0; i < DRM_SCHED_PRIORITY_MAX; ++i)
 		atomic_set(&ring->num_jobs[i], 0);
 
-	if (amdgpu_debugfs_ring_init(adev, ring)) {
-		DRM_ERROR("Failed to register debugfs file for rings !\n");
-	}
-
 	return 0;
 }
 
@@ -351,12 +344,13 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
  */
 void amdgpu_ring_fini(struct amdgpu_ring *ring)
 {
-	ring->sched.ready = false;
 
 	/* Not to finish a ring which is not initialized */
 	if (!(ring->adev) || !(ring->adev->rings[ring->idx]))
 		return;
 
+	ring->sched.ready = false;
+
 	amdgpu_device_wb_free(ring->adev, ring->rptr_offs);
 	amdgpu_device_wb_free(ring->adev, ring->wptr_offs);
 
@@ -367,8 +361,6 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring)
 			      &ring->gpu_addr,
 			      (void **)&ring->ring);
 
-	amdgpu_debugfs_ring_fini(ring);
-
 	dma_fence_put(ring->vmid_wait);
 	ring->vmid_wait = NULL;
 	ring->me = 0;
@@ -485,7 +477,7 @@ static const struct file_operations amdgpu_debugfs_ring_fops = {
 
 #endif
 
-static int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
+int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
 			     struct amdgpu_ring *ring)
 {
 #if defined(CONFIG_DEBUG_FS)
@@ -507,7 +499,7 @@ static int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
 	return 0;
 }
 
-static void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring)
+void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring)
 {
 #if defined(CONFIG_DEBUG_FS)
 	debugfs_remove(ring->ent);

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

@@ -328,4 +328,8 @@ static inline void amdgpu_ring_write_multiple(struct amdgpu_ring *ring,
 
 int amdgpu_ring_test_helper(struct amdgpu_ring *ring);
 
+int amdgpu_debugfs_ring_init(struct amdgpu_device *adev,
+			     struct amdgpu_ring *ring);
+void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring);
+
 #endif

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

@@ -70,7 +70,8 @@ uint64_t amdgpu_sdma_get_csa_mc_addr(struct amdgpu_ring *ring,
 	uint32_t index = 0;
 	int r;
 
-	if (vmid == 0 || !amdgpu_mcbp)
+	/* don't enable OS preemption on SDMA under SRIOV */
+	if (amdgpu_sriov_vf(adev) || vmid == 0 || !amdgpu_mcbp)
 		return 0;
 
 	r = amdgpu_sdma_get_index_from_ring(ring, &index);

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

@@ -202,18 +202,17 @@ int amdgpu_sync_vm_fence(struct amdgpu_sync *sync, struct dma_fence *fence)
  *
  * @sync: sync object to add fences from reservation object to
  * @resv: reservation object with embedded fence
- * @explicit_sync: true if we should only sync to the exclusive fence
+ * @mode: how owner affects which fences we sync to
+ * @owner: owner of the planned job submission
  *
  * Sync to the fence
  */
-int amdgpu_sync_resv(struct amdgpu_device *adev,
+int amdgpu_sync_resv(struct amdgpu_device *adev, struct amdgpu_sync *sync,
-		     struct amdgpu_sync *sync,
+		     struct dma_resv *resv, enum amdgpu_sync_mode mode,
-		     struct dma_resv *resv,
+		     void *owner)
-		     void *owner, bool explicit_sync)
 {
 	struct dma_resv_list *flist;
 	struct dma_fence *f;
-	void *fence_owner;
 	unsigned i;
 	int r = 0;
 
@@ -229,29 +228,45 @@ int amdgpu_sync_resv(struct amdgpu_device *adev,
 		return r;
 
 	for (i = 0; i < flist->shared_count; ++i) {
+		void *fence_owner;
+
 		f = rcu_dereference_protected(flist->shared[i],
 					      dma_resv_held(resv));
+
+		fence_owner = amdgpu_sync_get_owner(f);
+
+		/* Always sync to moves, no matter what */
+		if (fence_owner == AMDGPU_FENCE_OWNER_UNDEFINED) {
+			r = amdgpu_sync_fence(sync, f, false);
+			if (r)
+				break;
+		}
+
 		/* We only want to trigger KFD eviction fences on
 		 * evict or move jobs. Skip KFD fences otherwise.
 		 */
-		fence_owner = amdgpu_sync_get_owner(f);
 		if (fence_owner == AMDGPU_FENCE_OWNER_KFD &&
 		    owner != AMDGPU_FENCE_OWNER_UNDEFINED)
 			continue;
 
-		if (amdgpu_sync_same_dev(adev, f)) {
+		/* Ignore fences depending on the sync mode */
-			/* VM updates only sync with moves but not with user
+		switch (mode) {
-			 * command submissions or KFD evictions fences
+		case AMDGPU_SYNC_ALWAYS:
-			 */
+			break;
-			if (owner == AMDGPU_FENCE_OWNER_VM &&
-			    fence_owner != AMDGPU_FENCE_OWNER_UNDEFINED)
-				continue;
 
-			/* Ignore fence from the same owner and explicit one as
+		case AMDGPU_SYNC_NE_OWNER:
-			 * long as it isn't undefined.
+			if (amdgpu_sync_same_dev(adev, f) &&
-			 */
+			    fence_owner == owner)
-			if (owner != AMDGPU_FENCE_OWNER_UNDEFINED &&
+				continue;
-			    (fence_owner == owner || explicit_sync))
+			break;
+
+		case AMDGPU_SYNC_EQ_OWNER:
+			if (amdgpu_sync_same_dev(adev, f) &&
+			    fence_owner != owner)
+				continue;
+			break;
+
+		case AMDGPU_SYNC_EXPLICIT:
 			continue;
 		}
 

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

@@ -31,6 +31,13 @@ struct dma_resv;
 struct amdgpu_device;
 struct amdgpu_ring;
 
+enum amdgpu_sync_mode {
+	AMDGPU_SYNC_ALWAYS,
+	AMDGPU_SYNC_NE_OWNER,
+	AMDGPU_SYNC_EQ_OWNER,
+	AMDGPU_SYNC_EXPLICIT
+};
+
 /*
  * Container for fences used to sync command submissions.
  */
@@ -43,11 +50,9 @@ void amdgpu_sync_create(struct amdgpu_sync *sync);
 int amdgpu_sync_fence(struct amdgpu_sync *sync, struct dma_fence *f,
 		      bool explicit);
 int amdgpu_sync_vm_fence(struct amdgpu_sync *sync, struct dma_fence *fence);
-int amdgpu_sync_resv(struct amdgpu_device *adev,
+int amdgpu_sync_resv(struct amdgpu_device *adev, struct amdgpu_sync *sync,
-		     struct amdgpu_sync *sync,
+		     struct dma_resv *resv, enum amdgpu_sync_mode mode,
-		     struct dma_resv *resv,
+		     void *owner);
-		     void *owner,
-		     bool explicit_sync);
 struct dma_fence *amdgpu_sync_peek_fence(struct amdgpu_sync *sync,
 				     struct amdgpu_ring *ring);
 struct dma_fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync,

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

@@ -60,15 +60,14 @@
 #include "amdgpu_ras.h"
 #include "bif/bif_4_1_d.h"
 
+#define AMDGPU_TTM_VRAM_MAX_DW_READ	(size_t)128
+
 static int amdgpu_map_buffer(struct ttm_buffer_object *bo,
 			     struct ttm_mem_reg *mem, unsigned num_pages,
 			     uint64_t offset, unsigned window,
 			     struct amdgpu_ring *ring,
 			     uint64_t *addr);
 
-static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev);
-static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev);
-
 /**
  * amdgpu_init_mem_type - Initialize a memory manager for a specific type of
  * memory request.
@@ -1591,7 +1590,7 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo,
 
 	while (len && pos < adev->gmc.mc_vram_size) {
 		uint64_t aligned_pos = pos & ~(uint64_t)3;
-		uint32_t bytes = 4 - (pos & 3);
+		uint64_t bytes = 4 - (pos & 3);
 		uint32_t shift = (pos & 3) * 8;
 		uint32_t mask = 0xffffffff << shift;
 
@@ -1600,6 +1599,7 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo,
 			bytes = len;
 		}
 
+		if (mask != 0xffffffff) {
 			spin_lock_irqsave(&adev->mmio_idx_lock, flags);
 			WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)aligned_pos) | 0x80000000);
 			WREG32_NO_KIQ(mmMM_INDEX_HI, aligned_pos >> 31);
@@ -1615,6 +1615,13 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo,
 				value = (value & mask) >> shift;
 				memcpy(buf, &value, bytes);
 			}
+		} else {
+			bytes = (nodes->start + nodes->size) << PAGE_SHIFT;
+			bytes = min(bytes - pos, (uint64_t)len & ~0x3ull);
+
+			amdgpu_device_vram_access(adev, pos, (uint32_t *)buf,
+						  bytes, write);
+		}
 
 		ret += bytes;
 		buf = (uint8_t *)buf + bytes;
@@ -1905,12 +1912,6 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	/* Register debugfs entries for amdgpu_ttm */
-	r = amdgpu_ttm_debugfs_init(adev);
-	if (r) {
-		DRM_ERROR("Failed to init debugfs\n");
-		return r;
-	}
 	return 0;
 }
 
@@ -1932,7 +1933,6 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev)
 	if (!adev->mman.initialized)
 		return;
 
-	amdgpu_ttm_debugfs_fini(adev);
 	amdgpu_ttm_training_reserve_vram_fini(adev);
 	/* return the IP Discovery TMR memory back to VRAM */
 	amdgpu_bo_free_kernel(&adev->discovery_memory, NULL, NULL);
@@ -2107,8 +2107,8 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset,
 	}
 	if (resv) {
 		r = amdgpu_sync_resv(adev, &job->sync, resv,
-				     AMDGPU_FENCE_OWNER_UNDEFINED,
+				     AMDGPU_SYNC_ALWAYS,
-				     false);
+				     AMDGPU_FENCE_OWNER_UNDEFINED);
 		if (r) {
 			DRM_ERROR("sync failed (%d).\n", r);
 			goto error_free;
@@ -2192,7 +2192,8 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo,
 
 	if (resv) {
 		r = amdgpu_sync_resv(adev, &job->sync, resv,
-				     AMDGPU_FENCE_OWNER_UNDEFINED, false);
+				     AMDGPU_SYNC_ALWAYS,
+				     AMDGPU_FENCE_OWNER_UNDEFINED);
 		if (r) {
 			DRM_ERROR("sync failed (%d).\n", r);
 			goto error_free;
@@ -2273,7 +2274,6 @@ static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 {
 	struct amdgpu_device *adev = file_inode(f)->i_private;
 	ssize_t result = 0;
-	int r;
 
 	if (size & 0x3 || *pos & 0x3)
 		return -EINVAL;
@@ -2281,27 +2281,19 @@ static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 	if (*pos >= adev->gmc.mc_vram_size)
 		return -ENXIO;
 
+	size = min(size, (size_t)(adev->gmc.mc_vram_size - *pos));
 	while (size) {
-		unsigned long flags;
+		size_t bytes = min(size, AMDGPU_TTM_VRAM_MAX_DW_READ * 4);
-		uint32_t value;
+		uint32_t value[AMDGPU_TTM_VRAM_MAX_DW_READ];
 
-		if (*pos >= adev->gmc.mc_vram_size)
+		amdgpu_device_vram_access(adev, *pos, value, bytes, false);
-			return result;
+		if (copy_to_user(buf, value, bytes))
+			return -EFAULT;
 
-		spin_lock_irqsave(&adev->mmio_idx_lock, flags);
+		result += bytes;
-		WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)*pos) | 0x80000000);
+		buf += bytes;
-		WREG32_NO_KIQ(mmMM_INDEX_HI, *pos >> 31);
+		*pos += bytes;
-		value = RREG32_NO_KIQ(mmMM_DATA);
+		size -= bytes;
-		spin_unlock_irqrestore(&adev->mmio_idx_lock, flags);
-
-		r = put_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
 	}
 
 	return result;
@@ -2538,7 +2530,7 @@ static const struct {
 
 #endif
 
-static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev)
+int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev)
 {
 #if defined(CONFIG_DEBUG_FS)
 	unsigned count;
@@ -2574,7 +2566,7 @@ static int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev)
 #endif
 }
 
-static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev)
+void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev)
 {
 #if defined(CONFIG_DEBUG_FS)
 	unsigned i;

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

@@ -133,4 +133,7 @@ uint64_t amdgpu_ttm_tt_pde_flags(struct ttm_tt *ttm, struct ttm_mem_reg *mem);
 uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 				 struct ttm_mem_reg *mem);
 
+int amdgpu_ttm_debugfs_init(struct amdgpu_device *adev);
+void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev);
+
 #endif

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

@@ -1099,7 +1099,8 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 			goto err_free;
 	} else {
 		r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.base.resv,
-				     AMDGPU_FENCE_OWNER_UNDEFINED, false);
+				     AMDGPU_SYNC_ALWAYS,
+				     AMDGPU_FENCE_OWNER_UNDEFINED);
 		if (r)
 			goto err_free;
 

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

@@ -120,23 +120,17 @@ static inline void amdgpu_vm_eviction_unlock(struct amdgpu_vm *vm)
 static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev,
 				      unsigned level)
 {
-	unsigned shift = 0xff;
-
 	switch (level) {
 	case AMDGPU_VM_PDB2:
 	case AMDGPU_VM_PDB1:
 	case AMDGPU_VM_PDB0:
-		shift = 9 * (AMDGPU_VM_PDB0 - level) +
+		return 9 * (AMDGPU_VM_PDB0 - level) +
 			adev->vm_manager.block_size;
-		break;
 	case AMDGPU_VM_PTB:
-		shift = 0;
+		return 0;
-		break;
 	default:
-		dev_err(adev->dev, "the level%d isn't supported.\n", level);
+		return ~0;
 	}
-
-	return shift;
 }
 
 /**
@@ -235,19 +229,6 @@ static void amdgpu_vm_bo_evicted(struct amdgpu_vm_bo_base *vm_bo)
 	else
 		list_move_tail(&vm_bo->vm_status, &vm->evicted);
 }
-
-/**
- * amdgpu_vm_bo_relocated - vm_bo is reloacted
- *
- * @vm_bo: vm_bo which is relocated
- *
- * State for PDs/PTs which needs to update their parent PD.
- */
-static void amdgpu_vm_bo_relocated(struct amdgpu_vm_bo_base *vm_bo)
-{
-	list_move(&vm_bo->vm_status, &vm_bo->vm->relocated);
-}
-
 /**
  * amdgpu_vm_bo_moved - vm_bo is moved
  *
@@ -290,6 +271,22 @@ static void amdgpu_vm_bo_invalidated(struct amdgpu_vm_bo_base *vm_bo)
 	spin_unlock(&vm_bo->vm->invalidated_lock);
 }
 
+/**
+ * amdgpu_vm_bo_relocated - vm_bo is reloacted
+ *
+ * @vm_bo: vm_bo which is relocated
+ *
+ * State for PDs/PTs which needs to update their parent PD.
+ * For the root PD, just move to idle state.
+ */
+static void amdgpu_vm_bo_relocated(struct amdgpu_vm_bo_base *vm_bo)
+{
+	if (vm_bo->bo->parent)
+		list_move(&vm_bo->vm_status, &vm_bo->vm->relocated);
+	else
+		amdgpu_vm_bo_idle(vm_bo);
+}
+
 /**
  * amdgpu_vm_bo_done - vm_bo is done
  *
@@ -588,8 +585,8 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm,
 {
 	entry->priority = 0;
 	entry->tv.bo = &vm->root.base.bo->tbo;
-	/* One for TTM and one for the CS job */
+	/* Two for VM updates, one for TTM and one for the CS job */
-	entry->tv.num_shared = 2;
+	entry->tv.num_shared = 4;
 	entry->user_pages = NULL;
 	list_add(&entry->tv.head, validated);
 }
@@ -697,10 +694,7 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			amdgpu_vm_bo_moved(bo_base);
 		} else {
 			vm->update_funcs->map_table(bo);
-			if (bo->parent)
 			amdgpu_vm_bo_relocated(bo_base);
-			else
-				amdgpu_vm_bo_idle(bo_base);
 		}
 	}
 
@@ -803,7 +797,7 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 	params.vm = vm;
 	params.direct = direct;
 
-	r = vm->update_funcs->prepare(&params, AMDGPU_FENCE_OWNER_KFD, NULL);
+	r = vm->update_funcs->prepare(&params, NULL, AMDGPU_SYNC_EXPLICIT);
 	if (r)
 		return r;
 
@@ -1299,7 +1293,7 @@ int amdgpu_vm_update_pdes(struct amdgpu_device *adev,
 	params.vm = vm;
 	params.direct = direct;
 
-	r = vm->update_funcs->prepare(&params, AMDGPU_FENCE_OWNER_VM, NULL);
+	r = vm->update_funcs->prepare(&params, NULL, AMDGPU_SYNC_EXPLICIT);
 	if (r)
 		return r;
 
@@ -1448,21 +1442,14 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
 		uint64_t incr, entry_end, pe_start;
 		struct amdgpu_bo *pt;
 
-		/* make sure that the page tables covering the address range are
+		if (flags & AMDGPU_PTE_VALID) {
-		 * actually allocated
+			/* make sure that the page tables covering the
+			 * address range are actually allocated
 			 */
-		r = amdgpu_vm_alloc_pts(params->adev, params->vm, &cursor,
+			r = amdgpu_vm_alloc_pts(params->adev, params->vm,
-					params->direct);
+						&cursor, params->direct);
 			if (r)
 				return r;
-
-		pt = cursor.entry->base.bo;
-
-		/* The root level can't be a huge page */
-		if (cursor.level == adev->vm_manager.root_level) {
-			if (!amdgpu_vm_pt_descendant(adev, &cursor))
-				return -ENOENT;
-			continue;
 		}
 
 		shift = amdgpu_vm_level_shift(adev, cursor.level);
@@ -1480,25 +1467,38 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
 			 * smaller than the address shift. Go to the next
 			 * child entry and try again.
 			 */
-			if (!amdgpu_vm_pt_descendant(adev, &cursor))
+			if (amdgpu_vm_pt_descendant(adev, &cursor))
-				return -ENOENT;
 				continue;
-		} else if (frag >= parent_shift &&
+		} else if (frag >= parent_shift) {
-			   cursor.level - 1 != adev->vm_manager.root_level) {
 			/* If the fragment size is even larger than the parent
-			 * shift we should go up one level and check it again
+			 * shift we should go up one level and check it again.
-			 * unless one level up is the root level.
 			 */
 			if (!amdgpu_vm_pt_ancestor(&cursor))
-				return -ENOENT;
+				return -EINVAL;
 			continue;
 		}
 
+		pt = cursor.entry->base.bo;
+		if (!pt) {
+			/* We need all PDs and PTs for mapping something, */
+			if (flags & AMDGPU_PTE_VALID)
+				return -ENOENT;
+
+			/* but unmapping something can happen at a higher
+			 * level.
+			 */
+			if (!amdgpu_vm_pt_ancestor(&cursor))
+				return -EINVAL;
+
+			pt = cursor.entry->base.bo;
+			shift = parent_shift;
+		}
+
 		/* Looks good so far, calculate parameters for the update */
 		incr = (uint64_t)AMDGPU_GPU_PAGE_SIZE << shift;
 		mask = amdgpu_vm_entries_mask(adev, cursor.level);
 		pe_start = ((cursor.pfn >> shift) & mask) * 8;
-		entry_end = (uint64_t)(mask + 1) << shift;
+		entry_end = ((uint64_t)mask + 1) << shift;
 		entry_end += cursor.pfn & ~(entry_end - 1);
 		entry_end = min(entry_end, end);
 
@@ -1506,6 +1506,10 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
 			uint64_t upd_end = min(entry_end, frag_end);
 			unsigned nptes = (upd_end - frag_start) >> shift;
 
+			/* This can happen when we set higher level PDs to
+			 * silent to stop fault floods.
+			 */
+			nptes = max(nptes, 1u);
 			amdgpu_vm_update_flags(params, pt, cursor.level,
 					       pe_start, dst, nptes, incr,
 					       flags | AMDGPU_PTE_FRAG(frag));
@@ -1550,7 +1554,7 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
  * @adev: amdgpu_device pointer
  * @vm: requested vm
  * @direct: direct submission in a page fault
- * @exclusive: fence we need to sync to
+ * @resv: fences we need to sync to
  * @start: start of mapped range
  * @last: last mapped entry
  * @flags: flags for the entries
@@ -1565,14 +1569,14 @@ static int amdgpu_vm_update_ptes(struct amdgpu_vm_update_params *params,
  */
 static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 				       struct amdgpu_vm *vm, bool direct,
-				       struct dma_fence *exclusive,
+				       struct dma_resv *resv,
 				       uint64_t start, uint64_t last,
 				       uint64_t flags, uint64_t addr,
 				       dma_addr_t *pages_addr,
 				       struct dma_fence **fence)
 {
 	struct amdgpu_vm_update_params params;
-	void *owner = AMDGPU_FENCE_OWNER_VM;
+	enum amdgpu_sync_mode sync_mode;
 	int r;
 
 	memset(&params, 0, sizeof(params));
@@ -1581,9 +1585,13 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 	params.direct = direct;
 	params.pages_addr = pages_addr;
 
-	/* sync to everything except eviction fences on unmapping */
+	/* Implicitly sync to command submissions in the same VM before
+	 * unmapping. Sync to moving fences before mapping.
+	 */
 	if (!(flags & AMDGPU_PTE_VALID))
-		owner = AMDGPU_FENCE_OWNER_KFD;
+		sync_mode = AMDGPU_SYNC_EQ_OWNER;
+	else
+		sync_mode = AMDGPU_SYNC_EXPLICIT;
 
 	amdgpu_vm_eviction_lock(vm);
 	if (vm->evicting) {
@@ -1591,7 +1599,17 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 		goto error_unlock;
 	}
 
-	r = vm->update_funcs->prepare(&params, owner, exclusive);
+	if (flags & AMDGPU_PTE_VALID) {
+		struct amdgpu_bo *root = vm->root.base.bo;
+
+		if (!dma_fence_is_signaled(vm->last_direct))
+			amdgpu_bo_fence(root, vm->last_direct, true);
+
+		if (!dma_fence_is_signaled(vm->last_delayed))
+			amdgpu_bo_fence(root, vm->last_delayed, true);
+	}
+
+	r = vm->update_funcs->prepare(&params, resv, sync_mode);
 	if (r)
 		goto error_unlock;
 
@@ -1610,7 +1628,7 @@ error_unlock:
  * amdgpu_vm_bo_split_mapping - split a mapping into smaller chunks
  *
  * @adev: amdgpu_device pointer
- * @exclusive: fence we need to sync to
+ * @resv: fences we need to sync to
  * @pages_addr: DMA addresses to use for mapping
  * @vm: requested vm
  * @mapping: mapped range and flags to use for the update
@@ -1626,7 +1644,7 @@ error_unlock:
  * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
-				      struct dma_fence *exclusive,
+				      struct dma_resv *resv,
 				      dma_addr_t *pages_addr,
 				      struct amdgpu_vm *vm,
 				      struct amdgpu_bo_va_mapping *mapping,
@@ -1702,7 +1720,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 		}
 
 		last = min((uint64_t)mapping->last, start + max_entries - 1);
-		r = amdgpu_vm_bo_update_mapping(adev, vm, false, exclusive,
+		r = amdgpu_vm_bo_update_mapping(adev, vm, false, resv,
 						start, last, flags, addr,
 						dma_addr, fence);
 		if (r)
@@ -1741,7 +1759,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 	dma_addr_t *pages_addr = NULL;
 	struct ttm_mem_reg *mem;
 	struct drm_mm_node *nodes;
-	struct dma_fence *exclusive, **last_update;
+	struct dma_fence **last_update;
+	struct dma_resv *resv;
 	uint64_t flags;
 	struct amdgpu_device *bo_adev = adev;
 	int r;
@@ -1749,7 +1768,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 	if (clear || !bo) {
 		mem = NULL;
 		nodes = NULL;
-		exclusive = NULL;
+		resv = vm->root.base.bo->tbo.base.resv;
 	} else {
 		struct ttm_dma_tt *ttm;
 
@@ -1759,7 +1778,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 			ttm = container_of(bo->tbo.ttm, struct ttm_dma_tt, ttm);
 			pages_addr = ttm->dma_address;
 		}
-		exclusive = bo->tbo.moving;
+		resv = bo->tbo.base.resv;
 	}
 
 	if (bo) {
@@ -1769,7 +1788,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 		flags = 0x0;
 	}
 
-	if (clear || (bo && bo->tbo.base.resv == vm->root.base.bo->tbo.base.resv))
+	if (clear || (bo && bo->tbo.base.resv ==
+		      vm->root.base.bo->tbo.base.resv))
 		last_update = &vm->last_update;
 	else
 		last_update = &bo_va->last_pt_update;
@@ -1783,7 +1803,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 	}
 
 	list_for_each_entry(mapping, &bo_va->invalids, list) {
-		r = amdgpu_vm_bo_split_mapping(adev, exclusive, pages_addr, vm,
+		r = amdgpu_vm_bo_split_mapping(adev, resv, pages_addr, vm,
 					       mapping, flags, bo_adev, nodes,
 					       last_update);
 		if (r)
@@ -1978,6 +1998,7 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 			  struct amdgpu_vm *vm,
 			  struct dma_fence **fence)
 {
+	struct dma_resv *resv = vm->root.base.bo->tbo.base.resv;
 	struct amdgpu_bo_va_mapping *mapping;
 	uint64_t init_pte_value = 0;
 	struct dma_fence *f = NULL;
@@ -1992,7 +2013,7 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 		    mapping->start < AMDGPU_GMC_HOLE_START)
 			init_pte_value = AMDGPU_PTE_DEFAULT_ATC;
 
-		r = amdgpu_vm_bo_update_mapping(adev, vm, false, NULL,
+		r = amdgpu_vm_bo_update_mapping(adev, vm, false, resv,
 						mapping->start, mapping->last,
 						init_pte_value, 0, NULL, &f);
 		amdgpu_vm_free_mapping(adev, vm, mapping, f);

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

@@ -227,8 +227,8 @@ struct amdgpu_vm_update_params {
 
 struct amdgpu_vm_update_funcs {
 	int (*map_table)(struct amdgpu_bo *bo);
-	int (*prepare)(struct amdgpu_vm_update_params *p, void * owner,
+	int (*prepare)(struct amdgpu_vm_update_params *p, struct dma_resv *resv,
-		       struct dma_fence *exclusive);
+		       enum amdgpu_sync_mode sync_mode);
 	int (*update)(struct amdgpu_vm_update_params *p,
 		      struct amdgpu_bo *bo, uint64_t pe, uint64_t addr,
 		      unsigned count, uint32_t incr, uint64_t flags);

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

@@ -44,26 +44,14 @@ static int amdgpu_vm_cpu_map_table(struct amdgpu_bo *table)
  * Returns:
  * Negativ errno, 0 for success.
  */
-static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p, void *owner,
+static int amdgpu_vm_cpu_prepare(struct amdgpu_vm_update_params *p,
-				 struct dma_fence *exclusive)
+				 struct dma_resv *resv,
+				 enum amdgpu_sync_mode sync_mode)
 {
-	int r;
+	if (!resv)
-
-	/* Wait for any BO move to be completed */
-	if (exclusive) {
-		r = dma_fence_wait(exclusive, true);
-		if (unlikely(r))
-			return r;
-	}
-
-	/* Don't wait for submissions during page fault */
-	if (p->direct)
 		return 0;
 
-	/* Wait for PT BOs to be idle. PTs share the same resv. object
+	return amdgpu_bo_sync_wait_resv(p->adev, resv, sync_mode, p->vm, true);
-	 * as the root PD BO
-	 */
-	return amdgpu_bo_sync_wait(p->vm->root.base.bo, owner, true);
 }
 
 /**
@@ -86,6 +74,13 @@ static int amdgpu_vm_cpu_update(struct amdgpu_vm_update_params *p,
 {
 	unsigned int i;
 	uint64_t value;
+	int r;
+
+	if (bo->tbo.moving) {
+		r = dma_fence_wait(bo->tbo.moving, true);
+		if (r)
+			return r;
+	}
 
 	pe += (unsigned long)amdgpu_bo_kptr(bo);
 

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

@@ -58,9 +58,9 @@ static int amdgpu_vm_sdma_map_table(struct amdgpu_bo *table)
  * Negativ errno, 0 for success.
  */
 static int amdgpu_vm_sdma_prepare(struct amdgpu_vm_update_params *p,
-				  void *owner, struct dma_fence *exclusive)
+				  struct dma_resv *resv,
+				  enum amdgpu_sync_mode sync_mode)
 {
-	struct amdgpu_bo *root = p->vm->root.base.bo;
 	unsigned int ndw = AMDGPU_VM_SDMA_MIN_NUM_DW;
 	int r;
 
@@ -70,17 +70,10 @@ static int amdgpu_vm_sdma_prepare(struct amdgpu_vm_update_params *p,
 
 	p->num_dw_left = ndw;
 
-	/* Wait for moves to be completed */
+	if (!resv)
-	r = amdgpu_sync_fence(&p->job->sync, exclusive, false);
-	if (r)
-		return r;
-
-	/* Don't wait for any submissions during page fault handling */
-	if (p->direct)
 		return 0;
 
-	return amdgpu_sync_resv(p->adev, &p->job->sync, root->tbo.base.resv,
+	return amdgpu_sync_resv(p->adev, &p->job->sync, resv, sync_mode, p->vm);
-				owner, false);
 }
 
 /**
@@ -147,7 +140,7 @@ static void amdgpu_vm_sdma_copy_ptes(struct amdgpu_vm_update_params *p,
 
 	src += p->num_dw_left * 4;
 
-	pe += amdgpu_bo_gpu_offset(bo);
+	pe += amdgpu_gmc_sign_extend(bo->tbo.offset);
 	trace_amdgpu_vm_copy_ptes(pe, src, count, p->direct);
 
 	amdgpu_vm_copy_pte(p->adev, ib, pe, src, count);
@@ -174,7 +167,7 @@ static void amdgpu_vm_sdma_set_ptes(struct amdgpu_vm_update_params *p,
 {
 	struct amdgpu_ib *ib = p->job->ibs;
 
-	pe += amdgpu_bo_gpu_offset(bo);
+	pe += amdgpu_gmc_sign_extend(bo->tbo.offset);
 	trace_amdgpu_vm_set_ptes(pe, addr, count, incr, flags, p->direct);
 	if (count < 3) {
 		amdgpu_vm_write_pte(p->adev, ib, pe, addr | flags,
@@ -208,6 +201,11 @@ static int amdgpu_vm_sdma_update(struct amdgpu_vm_update_params *p,
 	uint64_t *pte;
 	int r;
 
+	/* Wait for PD/PT moves to be completed */
+	r = amdgpu_sync_fence(&p->job->sync, bo->tbo.moving, false);
+	if (r)
+		return r;
+
 	do {
 		ndw = p->num_dw_left;
 		ndw -= p->job->ibs->length_dw;

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

@@ -365,6 +365,13 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 		return 0;
 
 	if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP)) {
+		ret = psp_xgmi_initialize(&adev->psp);
+		if (ret) {
+			dev_err(adev->dev,
+				"XGMI: Failed to initialize xgmi session\n");
+			return ret;
+		}
+
 		ret = psp_xgmi_get_hive_id(&adev->psp, &adev->gmc.xgmi.hive_id);
 		if (ret) {
 			dev_err(adev->dev,
@@ -451,16 +458,16 @@ exit:
 	return ret;
 }
 
-void amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
+int amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
 {
 	struct amdgpu_hive_info *hive;
 
 	if (!adev->gmc.xgmi.supported)
-		return;
+		return -EINVAL;
 
 	hive = amdgpu_get_xgmi_hive(adev, 1);
 	if (!hive)
-		return;
+		return -EINVAL;
 
 	if (!(hive->number_devices--)) {
 		amdgpu_xgmi_sysfs_destroy(adev, hive);
@@ -471,6 +478,8 @@ void amdgpu_xgmi_remove_device(struct amdgpu_device *adev)
 		amdgpu_xgmi_sysfs_rem_dev_info(adev, hive);
 		mutex_unlock(&hive->hive_lock);
 	}
+
+	return psp_xgmi_terminate(&adev->psp);
 }
 
 int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev)
@@ -521,3 +530,33 @@ void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev)
 		kfree(ras_if);
 	}
 }
+
+uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev,
+					   uint64_t addr)
+{
+	uint32_t df_inst_id;
+	uint64_t dram_base_addr = 0;
+	const struct amdgpu_df_funcs *df_funcs = adev->df.funcs;
+
+	if ((!df_funcs)                 ||
+	    (!df_funcs->get_df_inst_id) ||
+	    (!df_funcs->get_dram_base_addr)) {
+		dev_warn(adev->dev,
+			 "XGMI: relative phy_addr algorithm is not supported\n");
+		return addr;
+	}
+
+	if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW)) {
+		dev_warn(adev->dev,
+			 "failed to disable DF-Cstate, DF register may not be accessible\n");
+		return addr;
+	}
+
+	df_inst_id = df_funcs->get_df_inst_id(adev);
+	dram_base_addr = df_funcs->get_dram_base_addr(adev, df_inst_id);
+
+	if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_ALLOW))
+		dev_warn(adev->dev, "failed to enable DF-Cstate\n");
+
+	return addr + dram_base_addr;
+}

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

@@ -40,12 +40,14 @@ struct amdgpu_hive_info {
 struct amdgpu_hive_info *amdgpu_get_xgmi_hive(struct amdgpu_device *adev, int lock);
 int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_device *adev);
 int amdgpu_xgmi_add_device(struct amdgpu_device *adev);
-void amdgpu_xgmi_remove_device(struct amdgpu_device *adev);
+int amdgpu_xgmi_remove_device(struct amdgpu_device *adev);
 int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate);
 int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev,
 		struct amdgpu_device *peer_adev);
 int amdgpu_xgmi_ras_late_init(struct amdgpu_device *adev);
 void amdgpu_xgmi_ras_fini(struct amdgpu_device *adev);
+uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev,
+					   uint64_t addr);
 
 static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev,
 		struct amdgpu_device *bo_adev)

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

@@ -186,16 +186,10 @@ amdgpu_atombios_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *m
 
 void amdgpu_atombios_dp_aux_init(struct amdgpu_connector *amdgpu_connector)
 {
-	int ret;
-
 	amdgpu_connector->ddc_bus->rec.hpd = amdgpu_connector->hpd.hpd;
-	amdgpu_connector->ddc_bus->aux.dev = amdgpu_connector->base.kdev;
 	amdgpu_connector->ddc_bus->aux.transfer = amdgpu_atombios_dp_aux_transfer;
-	ret = drm_dp_aux_register(&amdgpu_connector->ddc_bus->aux);
+	drm_dp_aux_init(&amdgpu_connector->ddc_bus->aux);
-	if (!ret)
 	amdgpu_connector->ddc_bus->has_aux = true;
-
-	WARN(ret, "drm_dp_aux_register_i2c_bus() failed with error %d\n", ret);
 }
 
 /***** general DP utility functions *****/

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

@@ -614,7 +614,6 @@ static int dce_virtual_connector_encoder_init(struct amdgpu_device *adev,
 	connector->display_info.subpixel_order = SubPixelHorizontalRGB;
 	connector->interlace_allowed = false;
 	connector->doublescan_allowed = false;
-	drm_connector_register(connector);
 
 	/* link them */
 	drm_connector_attach_encoder(connector, encoder);

drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c

@@ -3164,12 +3164,7 @@ static int gfx_v10_0_kiq_enable_kgq(struct amdgpu_device *adev)
 	for (i = 0; i < adev->gfx.num_gfx_rings; i++)
 		kiq->pmf->kiq_map_queues(kiq_ring, &adev->gfx.gfx_ring[i]);
 
-	r = amdgpu_ring_test_ring(kiq_ring);
+	return amdgpu_ring_test_helper(kiq_ring);
-	if (r) {
-		DRM_ERROR("kfq enable failed\n");
-		kiq_ring->sched.ready = false;
-	}
-	return r;
 }
 #endif
 
@@ -3785,7 +3780,7 @@ static int gfx_v10_0_kiq_disable_kgq(struct amdgpu_device *adev)
 		kiq->pmf->kiq_unmap_queues(kiq_ring, &adev->gfx.gfx_ring[i],
 					   PREEMPT_QUEUES, 0, 0);
 
-	return amdgpu_ring_test_ring(kiq_ring);
+	return amdgpu_ring_test_helper(kiq_ring);
 }
 #endif
 
@@ -3923,11 +3918,13 @@ static uint64_t gfx_v10_0_get_gpu_clock_counter(struct amdgpu_device *adev)
 {
 	uint64_t clock;
 
+	amdgpu_gfx_off_ctrl(adev, false);
 	mutex_lock(&adev->gfx.gpu_clock_mutex);
 	WREG32_SOC15(GC, 0, mmRLC_CAPTURE_GPU_CLOCK_COUNT, 1);
 	clock = (uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_LSB) |
 		((uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
 	mutex_unlock(&adev->gfx.gpu_clock_mutex);
+	amdgpu_gfx_off_ctrl(adev, true);
 	return clock;
 }
 
@@ -4411,7 +4408,7 @@ static void gfx_v10_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 
 	control |= ib->length_dw | (vmid << 24);
 
-	if (amdgpu_mcbp && (ib->flags & AMDGPU_IB_FLAG_PREEMPT)) {
+	if ((amdgpu_sriov_vf(ring->adev) || amdgpu_mcbp) && (ib->flags & AMDGPU_IB_FLAG_PREEMPT)) {
 		control |= INDIRECT_BUFFER_PRE_ENB(1);
 
 		if (flags & AMDGPU_IB_PREEMPTED)
@@ -4419,7 +4416,7 @@ static void gfx_v10_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 
 		if (!(ib->flags & AMDGPU_IB_FLAG_CE))
 			gfx_v10_0_ring_emit_de_meta(ring,
-				    flags & AMDGPU_IB_PREEMPTED ? true : false);
+				    (!amdgpu_sriov_vf(ring->adev) && flags & AMDGPU_IB_PREEMPTED) ? true : false);
 	}
 
 	amdgpu_ring_write(ring, header);
@@ -4566,9 +4563,9 @@ static void gfx_v10_0_ring_emit_cntxcntl(struct amdgpu_ring *ring, uint32_t flag
 {
 	uint32_t dw2 = 0;
 
-	if (amdgpu_mcbp)
+	if (amdgpu_mcbp || amdgpu_sriov_vf(ring->adev))
 		gfx_v10_0_ring_emit_ce_meta(ring,
-				    flags & AMDGPU_IB_PREEMPTED ? true : false);
+				    (!amdgpu_sriov_vf(ring->adev) && flags & AMDGPU_IB_PREEMPTED) ? true : false);
 
 	gfx_v10_0_ring_emit_tmz(ring, true);
 

drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c

@@ -1106,10 +1106,11 @@ static void gfx_v9_0_check_fw_write_wait(struct amdgpu_device *adev)
 	adev->gfx.me_fw_write_wait = false;
 	adev->gfx.mec_fw_write_wait = false;
 
-	if ((adev->gfx.mec_fw_version < 0x000001a5) ||
+	if ((adev->asic_type != CHIP_ARCTURUS) &&
+	    ((adev->gfx.mec_fw_version < 0x000001a5) ||
 	    (adev->gfx.mec_feature_version < 46) ||
 	    (adev->gfx.pfp_fw_version < 0x000000b7) ||
-	    (adev->gfx.pfp_feature_version < 46))
+	    (adev->gfx.pfp_feature_version < 46)))
 		DRM_WARN_ONCE("CP firmware version too old, please update!");
 
 	switch (adev->asic_type) {
@@ -1193,6 +1194,14 @@ static bool gfx_v9_0_should_disable_gfxoff(struct pci_dev *pdev)
 	return false;
 }
 
+static bool is_raven_kicker(struct amdgpu_device *adev)
+{
+	if (adev->pm.fw_version >= 0x41e2b)
+		return true;
+	else
+		return false;
+}
+
 static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev)
 {
 	if (gfx_v9_0_should_disable_gfxoff(adev->pdev))
@@ -1205,9 +1214,8 @@ static void gfx_v9_0_check_if_need_gfxoff(struct amdgpu_device *adev)
 		break;
 	case CHIP_RAVEN:
 		if (!(adev->rev_id >= 0x8 || adev->pdev->device == 0x15d8) &&
-		    ((adev->gfx.rlc_fw_version != 106 &&
+		    ((!is_raven_kicker(adev) &&
 		      adev->gfx.rlc_fw_version < 531) ||
-		     (adev->gfx.rlc_fw_version == 53815) ||
 		     (adev->gfx.rlc_feature_version < 1) ||
 		     !adev->gfx.rlc.is_rlc_v2_1))
 			adev->pm.pp_feature &= ~PP_GFXOFF_MASK;
@@ -3959,6 +3967,7 @@ static uint64_t gfx_v9_0_get_gpu_clock_counter(struct amdgpu_device *adev)
 {
 	uint64_t clock;
 
+	amdgpu_gfx_off_ctrl(adev, false);
 	mutex_lock(&adev->gfx.gpu_clock_mutex);
 	if (adev->asic_type == CHIP_VEGA10 && amdgpu_sriov_runtime(adev)) {
 		uint32_t tmp, lsb, msb, i = 0;
@@ -3977,6 +3986,7 @@ static uint64_t gfx_v9_0_get_gpu_clock_counter(struct amdgpu_device *adev)
 			((uint64_t)RREG32_SOC15(GC, 0, mmRLC_GPU_CLOCK_COUNT_MSB) << 32ULL);
 	}
 	mutex_unlock(&adev->gfx.gpu_clock_mutex);
+	amdgpu_gfx_off_ctrl(adev, true);
 	return clock;
 }
 

drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c

@@ -135,6 +135,10 @@ static void gfxhub_v2_0_init_cache_regs(struct amdgpu_device *adev)
 {
 	uint32_t tmp;
 
+	/* These regs are not accessible for VF, PF will program these in SRIOV */
+	if (amdgpu_sriov_vf(adev))
+		return;
+
 	/* Setup L2 cache */
 	tmp = RREG32_SOC15(GC, 0, mmGCVM_L2_CNTL);
 	tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL, ENABLE_L2_CACHE, 1);
@@ -298,9 +302,11 @@ void gfxhub_v2_0_gart_disable(struct amdgpu_device *adev)
 			    ENABLE_ADVANCED_DRIVER_MODEL, 0);
 	WREG32_SOC15(GC, 0, mmGCMC_VM_MX_L1_TLB_CNTL, tmp);
 
+	if (!amdgpu_sriov_vf(adev)) {
 		/* Setup L2 cache */
 		WREG32_FIELD15(GC, 0, GCVM_L2_CNTL, ENABLE_L2_CACHE, 0);
 		WREG32_SOC15(GC, 0, mmGCVM_L2_CNTL3, 0);
+	}
 }
 
 /**

drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c

@@ -476,13 +476,26 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 {
 	bool use_semaphore = gmc_v9_0_use_invalidate_semaphore(adev, vmhub);
 	const unsigned eng = 17;
-	u32 j, inv_req, tmp;
+	u32 j, inv_req, inv_req2, tmp;
 	struct amdgpu_vmhub *hub;
 
 	BUG_ON(vmhub >= adev->num_vmhubs);
 
 	hub = &adev->vmhub[vmhub];
+	if (adev->gmc.xgmi.num_physical_nodes &&
+	    adev->asic_type == CHIP_VEGA20) {
+		/* Vega20+XGMI caches PTEs in TC and TLB. Add a
+		 * heavy-weight TLB flush (type 2), which flushes
+		 * both. Due to a race condition with concurrent
+		 * memory accesses using the same TLB cache line, we
+		 * still need a second TLB flush after this.
+		 */
+		inv_req = gmc_v9_0_get_invalidate_req(vmid, 2);
+		inv_req2 = gmc_v9_0_get_invalidate_req(vmid, flush_type);
+	} else {
 		inv_req = gmc_v9_0_get_invalidate_req(vmid, flush_type);
+		inv_req2 = 0;
+	}
 
 	/* This is necessary for a HW workaround under SRIOV as well
 	 * as GFXOFF under bare metal
@@ -521,11 +534,13 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 			DRM_ERROR("Timeout waiting for sem acquire in VM flush!\n");
 	}
 
+	do {
 		WREG32_NO_KIQ(hub->vm_inv_eng0_req + eng, inv_req);
 
 		/*
-	 * Issue a dummy read to wait for the ACK register to be cleared
+		 * Issue a dummy read to wait for the ACK register to
-	 * to avoid a false ACK due to the new fast GRBM interface.
+		 * be cleared to avoid a false ACK due to the new fast
+		 * GRBM interface.
 		 */
 		if (vmhub == AMDGPU_GFXHUB_0)
 			RREG32_NO_KIQ(hub->vm_inv_eng0_req + eng);
@@ -537,6 +552,10 @@ static void gmc_v9_0_flush_gpu_tlb(struct amdgpu_device *adev, uint32_t vmid,
 			udelay(1);
 		}
 
+		inv_req = inv_req2;
+		inv_req2 = 0;
+	} while (inv_req);
+
 	/* TODO: It needs to continue working on debugging with semaphore for GFXHUB as well. */
 	if (use_semaphore)
 		/*
@@ -577,9 +596,26 @@ static int gmc_v9_0_flush_gpu_tlb_pasid(struct amdgpu_device *adev,
 		return -EIO;
 
 	if (ring->sched.ready) {
+		/* Vega20+XGMI caches PTEs in TC and TLB. Add a
+		 * heavy-weight TLB flush (type 2), which flushes
+		 * both. Due to a race condition with concurrent
+		 * memory accesses using the same TLB cache line, we
+		 * still need a second TLB flush after this.
+		 */
+		bool vega20_xgmi_wa = (adev->gmc.xgmi.num_physical_nodes &&
+				       adev->asic_type == CHIP_VEGA20);
+		/* 2 dwords flush + 8 dwords fence */
+		unsigned int ndw = kiq->pmf->invalidate_tlbs_size + 8;
+
+		if (vega20_xgmi_wa)
+			ndw += kiq->pmf->invalidate_tlbs_size;
+
 		spin_lock(&adev->gfx.kiq.ring_lock);
 		/* 2 dwords flush + 8 dwords fence */
-		amdgpu_ring_alloc(ring, kiq->pmf->invalidate_tlbs_size + 8);
+		amdgpu_ring_alloc(ring, ndw);
+		if (vega20_xgmi_wa)
+			kiq->pmf->kiq_invalidate_tlbs(ring,
+						      pasid, 2, all_hub);
 		kiq->pmf->kiq_invalidate_tlbs(ring,
 					pasid, flush_type, all_hub);
 		amdgpu_fence_emit_polling(ring, &seq);
@@ -1271,6 +1307,19 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev)
 	}
 }
 
+/**
+ * gmc_v9_0_restore_registers - restores regs
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * This restores register values, saved at suspend.
+ */
+static void gmc_v9_0_restore_registers(struct amdgpu_device *adev)
+{
+	if (adev->asic_type == CHIP_RAVEN)
+		WREG32(mmDCHUBBUB_SDPIF_MMIO_CNTRL_0, adev->gmc.sdpif_register);
+}
+
 /**
  * gmc_v9_0_gart_enable - gart enable
  *
@@ -1376,6 +1425,20 @@ static int gmc_v9_0_hw_init(void *handle)
 	return r;
 }
 
+/**
+ * gmc_v9_0_save_registers - saves regs
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * This saves potential register values that should be
+ * restored upon resume
+ */
+static void gmc_v9_0_save_registers(struct amdgpu_device *adev)
+{
+	if (adev->asic_type == CHIP_RAVEN)
+		adev->gmc.sdpif_register = RREG32(mmDCHUBBUB_SDPIF_MMIO_CNTRL_0);
+}
+
 /**
  * gmc_v9_0_gart_disable - gart disable
  *
@@ -1412,9 +1475,16 @@ static int gmc_v9_0_hw_fini(void *handle)
 
 static int gmc_v9_0_suspend(void *handle)
 {
+	int r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	return gmc_v9_0_hw_fini(adev);
+	r = gmc_v9_0_hw_fini(adev);
+	if (r)
+		return r;
+
+	gmc_v9_0_save_registers(adev);
+
+	return 0;
 }
 
 static int gmc_v9_0_resume(void *handle)
@@ -1422,6 +1492,7 @@ static int gmc_v9_0_resume(void *handle)
 	int r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	gmc_v9_0_restore_registers(adev);
 	r = gmc_v9_0_hw_init(adev);
 	if (r)
 		return r;

drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c

@@ -318,6 +318,7 @@ static void nbio_v7_4_handle_ras_controller_intr_no_bifring(struct amdgpu_device
 {
 	uint32_t bif_doorbell_intr_cntl;
 	struct ras_manager *obj = amdgpu_ras_find_obj(adev, adev->nbio.ras_if);
+	struct ras_err_data err_data = {0, 0, 0, NULL};
 
 	bif_doorbell_intr_cntl = RREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL);
 	if (REG_GET_FIELD(bif_doorbell_intr_cntl,
@@ -332,7 +333,19 @@ static void nbio_v7_4_handle_ras_controller_intr_no_bifring(struct amdgpu_device
 		 * clear error status after ras_controller_intr according to
 		 * hw team and count ue number for query
 		 */
-		nbio_v7_4_query_ras_error_count(adev, &obj->err_data);
+		nbio_v7_4_query_ras_error_count(adev, &err_data);
+
+		/* logging on error counter and printing for awareness */
+		obj->err_data.ue_count += err_data.ue_count;
+		obj->err_data.ce_count += err_data.ce_count;
+
+		if (err_data.ce_count)
+			DRM_INFO("%ld correctable errors detected in %s block\n",
+				obj->err_data.ce_count, adev->nbio.ras_if->name);
+
+		if (err_data.ue_count)
+			DRM_INFO("%ld uncorrectable errors detected in %s block\n",
+				obj->err_data.ue_count, adev->nbio.ras_if->name);
 
 		DRM_WARN("RAS controller interrupt triggered by NBIF error\n");
 

drivers/gpu/drm/amd/amdgpu/psp_v11_0.c

@@ -420,6 +420,7 @@ static int psp_v11_0_ring_init(struct psp_context *psp,
 	struct psp_ring *ring;
 	struct amdgpu_device *adev = psp->adev;
 
+	if (!amdgpu_sriov_vf(adev))
 		psp_v11_0_reroute_ih(psp);
 
 	ring = &psp->km_ring;

drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c

@@ -746,11 +746,9 @@ static int sdma_v5_0_gfx_resume(struct amdgpu_device *adev)
 			sdma_v5_0_enable(adev, true);
 		}
 
-		r = amdgpu_ring_test_ring(ring);
+		r = amdgpu_ring_test_helper(ring);
-		if (r) {
+		if (r)
-			ring->sched.ready = false;
 			return r;
-		}
 
 		if (adev->mman.buffer_funcs_ring == ring)
 			amdgpu_ttm_set_buffer_funcs_status(adev, true);

drivers/gpu/drm/amd/amdgpu/soc15.c

@@ -272,7 +272,12 @@ static u32 soc15_get_config_memsize(struct amdgpu_device *adev)
 
 static u32 soc15_get_xclk(struct amdgpu_device *adev)
 {
-	return adev->clock.spll.reference_freq;
+	u32 reference_clock = adev->clock.spll.reference_freq;
+
+	if (adev->asic_type == CHIP_RAVEN)
+		return reference_clock / 4;
+
+	return reference_clock;
 }
 
 

drivers/gpu/drm/amd/amdkfd/kfd_chardev.c

@@ -127,6 +127,8 @@ static int kfd_open(struct inode *inode, struct file *filep)
 		return PTR_ERR(process);
 
 	if (kfd_is_locked()) {
+		dev_dbg(kfd_device, "kfd is locked!\n"
+				"process %d unreferenced", process->pasid);
 		kfd_unref_process(process);
 		return -EAGAIN;
 	}

drivers/gpu/drm/amd/amdkfd/kfd_device.c

@@ -648,6 +648,9 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 	if (kfd->kfd2kgd->get_hive_id)
 		kfd->hive_id = kfd->kfd2kgd->get_hive_id(kfd->kgd);
 
+	if (kfd->kfd2kgd->get_unique_id)
+		kfd->unique_id = kfd->kfd2kgd->get_unique_id(kfd->kgd);
+
 	if (kfd_interrupt_init(kfd)) {
 		dev_err(kfd_device, "Error initializing interrupts\n");
 		goto kfd_interrupt_error;
@@ -710,7 +713,7 @@ out:
 void kgd2kfd_device_exit(struct kfd_dev *kfd)
 {
 	if (kfd->init_complete) {
-		kgd2kfd_suspend(kfd);
+		kgd2kfd_suspend(kfd, false);
 		device_queue_manager_uninit(kfd->dqm);
 		kfd_interrupt_exit(kfd);
 		kfd_topology_remove_device(kfd);
@@ -731,7 +734,7 @@ int kgd2kfd_pre_reset(struct kfd_dev *kfd)
 
 	kfd->dqm->ops.pre_reset(kfd->dqm);
 
-	kgd2kfd_suspend(kfd);
+	kgd2kfd_suspend(kfd, false);
 
 	kfd_signal_reset_event(kfd);
 	return 0;
@@ -765,21 +768,23 @@ bool kfd_is_locked(void)
 	return  (atomic_read(&kfd_locked) > 0);
 }
 
-void kgd2kfd_suspend(struct kfd_dev *kfd)
+void kgd2kfd_suspend(struct kfd_dev *kfd, bool run_pm)
 {
 	if (!kfd->init_complete)
 		return;
 
+	/* for runtime suspend, skip locking kfd */
+	if (!run_pm) {
 		/* For first KFD device suspend all the KFD processes */
 		if (atomic_inc_return(&kfd_locked) == 1)
 			kfd_suspend_all_processes();
+	}
 
 	kfd->dqm->ops.stop(kfd->dqm);
-
 	kfd_iommu_suspend(kfd);
 }
 
-int kgd2kfd_resume(struct kfd_dev *kfd)
+int kgd2kfd_resume(struct kfd_dev *kfd, bool run_pm)
 {
 	int ret, count;
 
@@ -790,10 +795,13 @@ int kgd2kfd_resume(struct kfd_dev *kfd)
 	if (ret)
 		return ret;
 
+	/* for runtime resume, skip unlocking kfd */
+	if (!run_pm) {
 		count = atomic_dec_return(&kfd_locked);
 		WARN_ONCE(count < 0, "KFD suspend / resume ref. error");
 		if (count == 0)
 			ret = kfd_resume_all_processes();
+	}
 
 	return ret;
 }

drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c

@@ -78,14 +78,14 @@ static bool is_pipe_enabled(struct device_queue_manager *dqm, int mec, int pipe)
 	/* queue is available for KFD usage if bit is 1 */
 	for (i = 0; i <  dqm->dev->shared_resources.num_queue_per_pipe; ++i)
 		if (test_bit(pipe_offset + i,
-			      dqm->dev->shared_resources.queue_bitmap))
+			      dqm->dev->shared_resources.cp_queue_bitmap))
 			return true;
 	return false;
 }
 
-unsigned int get_queues_num(struct device_queue_manager *dqm)
+unsigned int get_cp_queues_num(struct device_queue_manager *dqm)
 {
-	return bitmap_weight(dqm->dev->shared_resources.queue_bitmap,
+	return bitmap_weight(dqm->dev->shared_resources.cp_queue_bitmap,
 				KGD_MAX_QUEUES);
 }
 
@@ -109,6 +109,11 @@ static unsigned int get_num_xgmi_sdma_engines(struct device_queue_manager *dqm)
 	return dqm->dev->device_info->num_xgmi_sdma_engines;
 }
 
+static unsigned int get_num_all_sdma_engines(struct device_queue_manager *dqm)
+{
+	return get_num_sdma_engines(dqm) + get_num_xgmi_sdma_engines(dqm);
+}
+
 unsigned int get_num_sdma_queues(struct device_queue_manager *dqm)
 {
 	return dqm->dev->device_info->num_sdma_engines
@@ -132,6 +137,22 @@ void program_sh_mem_settings(struct device_queue_manager *dqm,
 						qpd->sh_mem_bases);
 }
 
+void increment_queue_count(struct device_queue_manager *dqm,
+			enum kfd_queue_type type)
+{
+	dqm->active_queue_count++;
+	if (type == KFD_QUEUE_TYPE_COMPUTE || type == KFD_QUEUE_TYPE_DIQ)
+		dqm->active_cp_queue_count++;
+}
+
+void decrement_queue_count(struct device_queue_manager *dqm,
+			enum kfd_queue_type type)
+{
+	dqm->active_queue_count--;
+	if (type == KFD_QUEUE_TYPE_COMPUTE || type == KFD_QUEUE_TYPE_DIQ)
+		dqm->active_cp_queue_count--;
+}
+
 static int allocate_doorbell(struct qcm_process_device *qpd, struct queue *q)
 {
 	struct kfd_dev *dev = qpd->dqm->dev;
@@ -281,8 +302,6 @@ static int create_queue_nocpsch(struct device_queue_manager *dqm,
 	struct mqd_manager *mqd_mgr;
 	int retval;
 
-	print_queue(q);
-
 	dqm_lock(dqm);
 
 	if (dqm->total_queue_count >= max_num_of_queues_per_device) {
@@ -359,12 +378,7 @@ add_queue_to_list:
 	list_add(&q->list, &qpd->queues_list);
 	qpd->queue_count++;
 	if (q->properties.is_active)
-		dqm->queue_count++;
+		increment_queue_count(dqm, q->properties.type);
-
-	if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
-		dqm->sdma_queue_count++;
-	else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)
-		dqm->xgmi_sdma_queue_count++;
 
 	/*
 	 * Unconditionally increment this counter, regardless of the queue's
@@ -446,15 +460,13 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm,
 	mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type(
 			q->properties.type)];
 
-	if (q->properties.type == KFD_QUEUE_TYPE_COMPUTE) {
+	if (q->properties.type == KFD_QUEUE_TYPE_COMPUTE)
 		deallocate_hqd(dqm, q);
-	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
+	else if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
-		dqm->sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q);
-	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI) {
+	else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)
-		dqm->xgmi_sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q);
-	} else {
+	else {
 		pr_debug("q->properties.type %d is invalid\n",
 				q->properties.type);
 		return -EINVAL;
@@ -494,7 +506,7 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm,
 	}
 	qpd->queue_count--;
 	if (q->properties.is_active)
-		dqm->queue_count--;
+		decrement_queue_count(dqm, q->properties.type);
 
 	return retval;
 }
@@ -563,13 +575,13 @@ static int update_queue(struct device_queue_manager *dqm, struct queue *q)
 	/*
 	 * check active state vs. the previous state and modify
 	 * counter accordingly. map_queues_cpsch uses the
-	 * dqm->queue_count to determine whether a new runlist must be
+	 * dqm->active_queue_count to determine whether a new runlist must be
 	 * uploaded.
 	 */
 	if (q->properties.is_active && !prev_active)
-		dqm->queue_count++;
+		increment_queue_count(dqm, q->properties.type);
 	else if (!q->properties.is_active && prev_active)
-		dqm->queue_count--;
+		decrement_queue_count(dqm, q->properties.type);
 
 	if (dqm->sched_policy != KFD_SCHED_POLICY_NO_HWS)
 		retval = map_queues_cpsch(dqm);
@@ -618,7 +630,7 @@ static int evict_process_queues_nocpsch(struct device_queue_manager *dqm,
 		mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type(
 				q->properties.type)];
 		q->properties.is_active = false;
-		dqm->queue_count--;
+		decrement_queue_count(dqm, q->properties.type);
 
 		if (WARN_ONCE(!dqm->sched_running, "Evict when stopped\n"))
 			continue;
@@ -662,7 +674,7 @@ static int evict_process_queues_cpsch(struct device_queue_manager *dqm,
 			continue;
 
 		q->properties.is_active = false;
-		dqm->queue_count--;
+		decrement_queue_count(dqm, q->properties.type);
 	}
 	retval = execute_queues_cpsch(dqm,
 				qpd->is_debug ?
@@ -731,7 +743,7 @@ static int restore_process_queues_nocpsch(struct device_queue_manager *dqm,
 		mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type(
 				q->properties.type)];
 		q->properties.is_active = true;
-		dqm->queue_count++;
+		increment_queue_count(dqm, q->properties.type);
 
 		if (WARN_ONCE(!dqm->sched_running, "Restore when stopped\n"))
 			continue;
@@ -786,7 +798,7 @@ static int restore_process_queues_cpsch(struct device_queue_manager *dqm,
 			continue;
 
 		q->properties.is_active = true;
-		dqm->queue_count++;
+		increment_queue_count(dqm, q->properties.type);
 	}
 	retval = execute_queues_cpsch(dqm,
 				KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0);
@@ -899,16 +911,15 @@ static int initialize_nocpsch(struct device_queue_manager *dqm)
 
 	mutex_init(&dqm->lock_hidden);
 	INIT_LIST_HEAD(&dqm->queues);
-	dqm->queue_count = dqm->next_pipe_to_allocate = 0;
+	dqm->active_queue_count = dqm->next_pipe_to_allocate = 0;
-	dqm->sdma_queue_count = 0;
+	dqm->active_cp_queue_count = 0;
-	dqm->xgmi_sdma_queue_count = 0;
 
 	for (pipe = 0; pipe < get_pipes_per_mec(dqm); pipe++) {
 		int pipe_offset = pipe * get_queues_per_pipe(dqm);
 
 		for (queue = 0; queue < get_queues_per_pipe(dqm); queue++)
 			if (test_bit(pipe_offset + queue,
-				     dqm->dev->shared_resources.queue_bitmap))
+				     dqm->dev->shared_resources.cp_queue_bitmap))
 				dqm->allocated_queues[pipe] |= 1 << queue;
 	}
 
@@ -924,7 +935,7 @@ static void uninitialize(struct device_queue_manager *dqm)
 {
 	int i;
 
-	WARN_ON(dqm->queue_count > 0 || dqm->processes_count > 0);
+	WARN_ON(dqm->active_queue_count > 0 || dqm->processes_count > 0);
 
 	kfree(dqm->allocated_queues);
 	for (i = 0 ; i < KFD_MQD_TYPE_MAX ; i++)
@@ -966,8 +977,11 @@ static int allocate_sdma_queue(struct device_queue_manager *dqm,
 	int bit;
 
 	if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
-		if (dqm->sdma_bitmap == 0)
+		if (dqm->sdma_bitmap == 0) {
+			pr_err("No more SDMA queue to allocate\n");
 			return -ENOMEM;
+		}
+
 		bit = __ffs64(dqm->sdma_bitmap);
 		dqm->sdma_bitmap &= ~(1ULL << bit);
 		q->sdma_id = bit;
@@ -976,8 +990,10 @@ static int allocate_sdma_queue(struct device_queue_manager *dqm,
 		q->properties.sdma_queue_id = q->sdma_id /
 				get_num_sdma_engines(dqm);
 	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI) {
-		if (dqm->xgmi_sdma_bitmap == 0)
+		if (dqm->xgmi_sdma_bitmap == 0) {
+			pr_err("No more XGMI SDMA queue to allocate\n");
 			return -ENOMEM;
+		}
 		bit = __ffs64(dqm->xgmi_sdma_bitmap);
 		dqm->xgmi_sdma_bitmap &= ~(1ULL << bit);
 		q->sdma_id = bit;
@@ -1029,7 +1045,7 @@ static int set_sched_resources(struct device_queue_manager *dqm)
 		mec = (i / dqm->dev->shared_resources.num_queue_per_pipe)
 			/ dqm->dev->shared_resources.num_pipe_per_mec;
 
-		if (!test_bit(i, dqm->dev->shared_resources.queue_bitmap))
+		if (!test_bit(i, dqm->dev->shared_resources.cp_queue_bitmap))
 			continue;
 
 		/* only acquire queues from the first MEC */
@@ -1064,9 +1080,9 @@ static int initialize_cpsch(struct device_queue_manager *dqm)
 
 	mutex_init(&dqm->lock_hidden);
 	INIT_LIST_HEAD(&dqm->queues);
-	dqm->queue_count = dqm->processes_count = 0;
+	dqm->active_queue_count = dqm->processes_count = 0;
-	dqm->sdma_queue_count = 0;
+	dqm->active_cp_queue_count = 0;
-	dqm->xgmi_sdma_queue_count = 0;
+
 	dqm->active_runlist = false;
 	dqm->sdma_bitmap = ~0ULL >> (64 - get_num_sdma_queues(dqm));
 	dqm->xgmi_sdma_bitmap = ~0ULL >> (64 - get_num_xgmi_sdma_queues(dqm));
@@ -1158,7 +1174,7 @@ static int create_kernel_queue_cpsch(struct device_queue_manager *dqm,
 			dqm->total_queue_count);
 
 	list_add(&kq->list, &qpd->priv_queue_list);
-	dqm->queue_count++;
+	increment_queue_count(dqm, kq->queue->properties.type);
 	qpd->is_debug = true;
 	execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0);
 	dqm_unlock(dqm);
@@ -1172,7 +1188,7 @@ static void destroy_kernel_queue_cpsch(struct device_queue_manager *dqm,
 {
 	dqm_lock(dqm);
 	list_del(&kq->list);
-	dqm->queue_count--;
+	decrement_queue_count(dqm, kq->queue->properties.type);
 	qpd->is_debug = false;
 	execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0);
 	/*
@@ -1238,13 +1254,9 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 	list_add(&q->list, &qpd->queues_list);
 	qpd->queue_count++;
 
-	if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
-		dqm->sdma_queue_count++;
-	else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)
-		dqm->xgmi_sdma_queue_count++;
-
 	if (q->properties.is_active) {
-		dqm->queue_count++;
+		increment_queue_count(dqm, q->properties.type);
+
 		retval = execute_queues_cpsch(dqm,
 				KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0);
 	}
@@ -1298,20 +1310,6 @@ int amdkfd_fence_wait_timeout(unsigned int *fence_addr,
 	return 0;
 }
 
-static int unmap_sdma_queues(struct device_queue_manager *dqm)
-{
-	int i, retval = 0;
-
-	for (i = 0; i < dqm->dev->device_info->num_sdma_engines +
-		dqm->dev->device_info->num_xgmi_sdma_engines; i++) {
-		retval = pm_send_unmap_queue(&dqm->packets, KFD_QUEUE_TYPE_SDMA,
-			KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0, false, i);
-		if (retval)
-			return retval;
-	}
-	return retval;
-}
-
 /* dqm->lock mutex has to be locked before calling this function */
 static int map_queues_cpsch(struct device_queue_manager *dqm)
 {
@@ -1319,7 +1317,7 @@ static int map_queues_cpsch(struct device_queue_manager *dqm)
 
 	if (!dqm->sched_running)
 		return 0;
-	if (dqm->queue_count <= 0 || dqm->processes_count <= 0)
+	if (dqm->active_queue_count <= 0 || dqm->processes_count <= 0)
 		return 0;
 	if (dqm->active_runlist)
 		return 0;
@@ -1349,12 +1347,6 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm,
 	if (!dqm->active_runlist)
 		return retval;
 
-	pr_debug("Before destroying queues, sdma queue count is : %u, xgmi sdma queue count is : %u\n",
-		dqm->sdma_queue_count, dqm->xgmi_sdma_queue_count);
-
-	if (dqm->sdma_queue_count > 0 || dqm->xgmi_sdma_queue_count)
-		unmap_sdma_queues(dqm);
-
 	retval = pm_send_unmap_queue(&dqm->packets, KFD_QUEUE_TYPE_COMPUTE,
 			filter, filter_param, false, 0);
 	if (retval)
@@ -1427,18 +1419,15 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm,
 
 	deallocate_doorbell(qpd, q);
 
-	if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
+	if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
-		dqm->sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q);
-	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI) {
+	else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)
-		dqm->xgmi_sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q);
-	}
 
 	list_del(&q->list);
 	qpd->queue_count--;
 	if (q->properties.is_active) {
-		dqm->queue_count--;
+		decrement_queue_count(dqm, q->properties.type);
 		retval = execute_queues_cpsch(dqm,
 				KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES, 0);
 		if (retval == -ETIME)
@@ -1648,7 +1637,7 @@ static int process_termination_cpsch(struct device_queue_manager *dqm,
 	/* Clean all kernel queues */
 	list_for_each_entry_safe(kq, kq_next, &qpd->priv_queue_list, list) {
 		list_del(&kq->list);
-		dqm->queue_count--;
+		decrement_queue_count(dqm, kq->queue->properties.type);
 		qpd->is_debug = false;
 		dqm->total_queue_count--;
 		filter = KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES;
@@ -1656,16 +1645,13 @@ static int process_termination_cpsch(struct device_queue_manager *dqm,
 
 	/* Clear all user mode queues */
 	list_for_each_entry(q, &qpd->queues_list, list) {
-		if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
+		if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
-			dqm->sdma_queue_count--;
 			deallocate_sdma_queue(dqm, q);
-		} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI) {
+		else if (q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI)
-			dqm->xgmi_sdma_queue_count--;
 			deallocate_sdma_queue(dqm, q);
-		}
 
 		if (q->properties.is_active)
-			dqm->queue_count--;
+			decrement_queue_count(dqm, q->properties.type);
 
 		dqm->total_queue_count--;
 	}
@@ -1742,8 +1728,7 @@ static int allocate_hiq_sdma_mqd(struct device_queue_manager *dqm)
 	struct kfd_dev *dev = dqm->dev;
 	struct kfd_mem_obj *mem_obj = &dqm->hiq_sdma_mqd;
 	uint32_t size = dqm->mqd_mgrs[KFD_MQD_TYPE_SDMA]->mqd_size *
-		(dev->device_info->num_sdma_engines +
+		get_num_all_sdma_engines(dqm) *
-		dev->device_info->num_xgmi_sdma_engines) *
 		dev->device_info->num_sdma_queues_per_engine +
 		dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ]->mqd_size;
 
@@ -1979,7 +1964,7 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 
 		for (queue = 0; queue < get_queues_per_pipe(dqm); queue++) {
 			if (!test_bit(pipe_offset + queue,
-				      dqm->dev->shared_resources.queue_bitmap))
+				      dqm->dev->shared_resources.cp_queue_bitmap))
 				continue;
 
 			r = dqm->dev->kfd2kgd->hqd_dump(
@@ -1995,8 +1980,7 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 		}
 	}
 
-	for (pipe = 0; pipe < get_num_sdma_engines(dqm) +
+	for (pipe = 0; pipe < get_num_all_sdma_engines(dqm); pipe++) {
-			get_num_xgmi_sdma_engines(dqm); pipe++) {
 		for (queue = 0;
 		     queue < dqm->dev->device_info->num_sdma_queues_per_engine;
 		     queue++) {

drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h

@@ -180,9 +180,8 @@ struct device_queue_manager {
 	struct list_head	queues;
 	unsigned int		saved_flags;
 	unsigned int		processes_count;
-	unsigned int		queue_count;
+	unsigned int		active_queue_count;
-	unsigned int		sdma_queue_count;
+	unsigned int		active_cp_queue_count;
-	unsigned int		xgmi_sdma_queue_count;
 	unsigned int		total_queue_count;
 	unsigned int		next_pipe_to_allocate;
 	unsigned int		*allocated_queues;
@@ -219,7 +218,7 @@ void device_queue_manager_init_v10_navi10(
 		struct device_queue_manager_asic_ops *asic_ops);
 void program_sh_mem_settings(struct device_queue_manager *dqm,
 					struct qcm_process_device *qpd);
-unsigned int get_queues_num(struct device_queue_manager *dqm);
+unsigned int get_cp_queues_num(struct device_queue_manager *dqm);
 unsigned int get_queues_per_pipe(struct device_queue_manager *dqm);
 unsigned int get_pipes_per_mec(struct device_queue_manager *dqm);
 unsigned int get_num_sdma_queues(struct device_queue_manager *dqm);

drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c

@@ -47,9 +47,8 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 	struct kfd_dev *dev = pm->dqm->dev;
 
 	process_count = pm->dqm->processes_count;
-	queue_count = pm->dqm->queue_count;
+	queue_count = pm->dqm->active_queue_count;
-	compute_queue_count = queue_count - pm->dqm->sdma_queue_count -
+	compute_queue_count = pm->dqm->active_cp_queue_count;
-				pm->dqm->xgmi_sdma_queue_count;
 
 	/* check if there is over subscription
 	 * Note: the arbitration between the number of VMIDs and
@@ -62,7 +61,7 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 		max_proc_per_quantum = dev->max_proc_per_quantum;
 
 	if ((process_count > max_proc_per_quantum) ||
-	    compute_queue_count > get_queues_num(pm->dqm)) {
+	    compute_queue_count > get_cp_queues_num(pm->dqm)) {
 		*over_subscription = true;
 		pr_debug("Over subscribed runlist\n");
 	}
@@ -141,7 +140,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 	pm->ib_size_bytes = alloc_size_bytes;
 
 	pr_debug("Building runlist ib process count: %d queues count %d\n",
-		pm->dqm->processes_count, pm->dqm->queue_count);
+		pm->dqm->processes_count, pm->dqm->active_queue_count);
 
 	/* build the run list ib packet */
 	list_for_each_entry(cur, queues, list) {

drivers/gpu/drm/amd/amdkfd/kfd_priv.h

@@ -41,6 +41,7 @@
 #include <drm/drm_drv.h>
 #include <drm/drm_device.h>
 #include <kgd_kfd_interface.h>
+#include <linux/swap.h>
 
 #include "amd_shared.h"
 
@@ -294,6 +295,9 @@ struct kfd_dev {
 	/* xGMI */
 	uint64_t hive_id;
     
+	/* UUID */
+	uint64_t unique_id;
+
 	bool pci_atomic_requested;
 
 	/* SRAM ECC flag */
@@ -502,6 +506,9 @@ struct queue {
 	struct kfd_process	*process;
 	struct kfd_dev		*device;
 	void *gws;
+
+	/* procfs */
+	struct kobject kobj;
 };
 
 /*
@@ -646,6 +653,7 @@ struct kfd_process_device {
 	 * function.
 	 */
 	bool already_dequeued;
+	bool runtime_inuse;
 
 	/* Is this process/pasid bound to this device? (amd_iommu_bind_pasid) */
 	enum kfd_pdd_bound bound;
@@ -729,6 +737,7 @@ struct kfd_process {
 
 	/* Kobj for our procfs */
 	struct kobject *kobj;
+	struct kobject *kobj_queues;
 	struct attribute attr_pasid;
 };
 
@@ -835,6 +844,8 @@ extern struct device *kfd_device;
 /* KFD's procfs */
 void kfd_procfs_init(void);
 void kfd_procfs_shutdown(void);
+int kfd_procfs_add_queue(struct queue *q);
+void kfd_procfs_del_queue(struct queue *q);
 
 /* Topology */
 int kfd_topology_init(void);

drivers/gpu/drm/amd/amdkfd/kfd_process.c

@@ -31,6 +31,7 @@
 #include <linux/compat.h>
 #include <linux/mman.h>
 #include <linux/file.h>
+#include <linux/pm_runtime.h>
 #include "amdgpu_amdkfd.h"
 #include "amdgpu.h"
 
@@ -132,6 +133,88 @@ void kfd_procfs_shutdown(void)
 	}
 }
 
+static ssize_t kfd_procfs_queue_show(struct kobject *kobj,
+				     struct attribute *attr, char *buffer)
+{
+	struct queue *q = container_of(kobj, struct queue, kobj);
+
+	if (!strcmp(attr->name, "size"))
+		return snprintf(buffer, PAGE_SIZE, "%llu",
+				q->properties.queue_size);
+	else if (!strcmp(attr->name, "type"))
+		return snprintf(buffer, PAGE_SIZE, "%d", q->properties.type);
+	else if (!strcmp(attr->name, "gpuid"))
+		return snprintf(buffer, PAGE_SIZE, "%u", q->device->id);
+	else
+		pr_err("Invalid attribute");
+
+	return 0;
+}
+
+static struct attribute attr_queue_size = {
+	.name = "size",
+	.mode = KFD_SYSFS_FILE_MODE
+};
+
+static struct attribute attr_queue_type = {
+	.name = "type",
+	.mode = KFD_SYSFS_FILE_MODE
+};
+
+static struct attribute attr_queue_gpuid = {
+	.name = "gpuid",
+	.mode = KFD_SYSFS_FILE_MODE
+};
+
+static struct attribute *procfs_queue_attrs[] = {
+	&attr_queue_size,
+	&attr_queue_type,
+	&attr_queue_gpuid,
+	NULL
+};
+
+static const struct sysfs_ops procfs_queue_ops = {
+	.show = kfd_procfs_queue_show,
+};
+
+static struct kobj_type procfs_queue_type = {
+	.sysfs_ops = &procfs_queue_ops,
+	.default_attrs = procfs_queue_attrs,
+};
+
+int kfd_procfs_add_queue(struct queue *q)
+{
+	struct kfd_process *proc;
+	int ret;
+
+	if (!q || !q->process)
+		return -EINVAL;
+	proc = q->process;
+
+	/* Create proc/<pid>/queues/<queue id> folder */
+	if (!proc->kobj_queues)
+		return -EFAULT;
+	ret = kobject_init_and_add(&q->kobj, &procfs_queue_type,
+			proc->kobj_queues, "%u", q->properties.queue_id);
+	if (ret < 0) {
+		pr_warn("Creating proc/<pid>/queues/%u failed",
+			q->properties.queue_id);
+		kobject_put(&q->kobj);
+		return ret;
+	}
+
+	return 0;
+}
+
+void kfd_procfs_del_queue(struct queue *q)
+{
+	if (!q)
+		return;
+
+	kobject_del(&q->kobj);
+	kobject_put(&q->kobj);
+}
+
 int kfd_process_create_wq(void)
 {
 	if (!kfd_process_wq)
@@ -323,6 +406,11 @@ struct kfd_process *kfd_create_process(struct file *filep)
 		if (ret)
 			pr_warn("Creating pasid for pid %d failed",
 					(int)process->lead_thread->pid);
+
+		process->kobj_queues = kobject_create_and_add("queues",
+							process->kobj);
+		if (!process->kobj_queues)
+			pr_warn("Creating KFD proc/queues folder failed");
 	}
 out:
 	if (!IS_ERR(process))
@@ -440,6 +528,16 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 		kfree(pdd->qpd.doorbell_bitmap);
 		idr_destroy(&pdd->alloc_idr);
 
+		/*
+		 * before destroying pdd, make sure to report availability
+		 * for auto suspend
+		 */
+		if (pdd->runtime_inuse) {
+			pm_runtime_mark_last_busy(pdd->dev->ddev->dev);
+			pm_runtime_put_autosuspend(pdd->dev->ddev->dev);
+			pdd->runtime_inuse = false;
+		}
+
 		kfree(pdd);
 	}
 }
@@ -457,6 +555,9 @@ static void kfd_process_wq_release(struct work_struct *work)
 	/* Remove the procfs files */
 	if (p->kobj) {
 		sysfs_remove_file(p->kobj, &p->attr_pasid);
+		kobject_del(p->kobj_queues);
+		kobject_put(p->kobj_queues);
+		p->kobj_queues = NULL;
 		kobject_del(p->kobj);
 		kobject_put(p->kobj);
 		p->kobj = NULL;
@@ -754,6 +855,7 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev,
 	pdd->process = p;
 	pdd->bound = PDD_UNBOUND;
 	pdd->already_dequeued = false;
+	pdd->runtime_inuse = false;
 	list_add(&pdd->per_device_list, &p->per_device_data);
 
 	/* Init idr used for memory handle translation */
@@ -843,15 +945,41 @@ struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev,
 		return ERR_PTR(-ENOMEM);
 	}
 
+	/*
+	 * signal runtime-pm system to auto resume and prevent
+	 * further runtime suspend once device pdd is created until
+	 * pdd is destroyed.
+	 */
+	if (!pdd->runtime_inuse) {
+		err = pm_runtime_get_sync(dev->ddev->dev);
+		if (err < 0)
+			return ERR_PTR(err);
+	}
+
 	err = kfd_iommu_bind_process_to_device(pdd);
 	if (err)
-		return ERR_PTR(err);
+		goto out;
 
 	err = kfd_process_device_init_vm(pdd, NULL);
 	if (err)
-		return ERR_PTR(err);
+		goto out;
+
+	/*
+	 * make sure that runtime_usage counter is incremented just once
+	 * per pdd
+	 */
+	pdd->runtime_inuse = true;
 
 	return pdd;
+
+out:
+	/* balance runpm reference count and exit with error */
+	if (!pdd->runtime_inuse) {
+		pm_runtime_mark_last_busy(dev->ddev->dev);
+		pm_runtime_put_autosuspend(dev->ddev->dev);
+	}
+
+	return ERR_PTR(err);
 }
 
 struct kfd_process_device *kfd_get_first_process_device_data(

drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c

@@ -241,23 +241,18 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	switch (type) {
 	case KFD_QUEUE_TYPE_SDMA:
 	case KFD_QUEUE_TYPE_SDMA_XGMI:
-		if ((type == KFD_QUEUE_TYPE_SDMA && dev->dqm->sdma_queue_count
+		/* SDMA queues are always allocated statically no matter
-			>= get_num_sdma_queues(dev->dqm)) ||
+		 * which scheduler mode is used. We also do not need to
-			(type == KFD_QUEUE_TYPE_SDMA_XGMI &&
+		 * check whether a SDMA queue can be allocated here, because
-			dev->dqm->xgmi_sdma_queue_count
+		 * allocate_sdma_queue() in create_queue() has the
-			>= get_num_xgmi_sdma_queues(dev->dqm))) {
+		 * corresponding check logic.
-			pr_debug("Over-subscription is not allowed for SDMA.\n");
+		 */
-			retval = -EPERM;
-			goto err_create_queue;
-		}
-
 		retval = init_user_queue(pqm, dev, &q, properties, f, *qid);
 		if (retval != 0)
 			goto err_create_queue;
 		pqn->q = q;
 		pqn->kq = NULL;
 		retval = dev->dqm->ops.create_queue(dev->dqm, q, &pdd->qpd);
-		pr_debug("DQM returned %d for create_queue\n", retval);
 		print_queue(q);
 		break;
 
@@ -266,7 +261,7 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 		if ((dev->dqm->sched_policy ==
 		     KFD_SCHED_POLICY_HWS_NO_OVERSUBSCRIPTION) &&
 		((dev->dqm->processes_count >= dev->vm_info.vmid_num_kfd) ||
-		(dev->dqm->queue_count >= get_queues_num(dev->dqm)))) {
+		(dev->dqm->active_queue_count >= get_cp_queues_num(dev->dqm)))) {
 			pr_debug("Over-subscription is not allowed when amdkfd.sched_policy == 1\n");
 			retval = -EPERM;
 			goto err_create_queue;
@@ -278,7 +273,6 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 		pqn->q = q;
 		pqn->kq = NULL;
 		retval = dev->dqm->ops.create_queue(dev->dqm, q, &pdd->qpd);
-		pr_debug("DQM returned %d for create_queue\n", retval);
 		print_queue(q);
 		break;
 	case KFD_QUEUE_TYPE_DIQ:
@@ -299,7 +293,7 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	}
 
 	if (retval != 0) {
-		pr_err("Pasid 0x%x DQM create queue %d failed. ret %d\n",
+		pr_err("Pasid 0x%x DQM create queue type %d failed. ret %d\n",
 			pqm->process->pasid, type, retval);
 		goto err_create_queue;
 	}
@@ -322,12 +316,16 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 
 	if (q) {
 		pr_debug("PQM done creating queue\n");
+		kfd_procfs_add_queue(q);
 		print_queue_properties(&q->properties);
 	}
 
 	return retval;
 
 err_create_queue:
+	uninit_queue(q);
+	if (kq)
+		kernel_queue_uninit(kq, false);
 	kfree(pqn);
 err_allocate_pqn:
 	/* check if queues list is empty unregister process from device */
@@ -378,6 +376,7 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
 	}
 
 	if (pqn->q) {
+		kfd_procfs_del_queue(pqn->q);
 		dqm = pqn->q->device->dqm;
 		retval = dqm->ops.destroy_queue(dqm, &pdd->qpd, pqn->q);
 		if (retval) {

drivers/gpu/drm/amd/amdkfd/kfd_topology.c

@@ -490,6 +490,8 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr,
 			dev->node_props.num_sdma_queues_per_engine);
 	sysfs_show_32bit_prop(buffer, "num_cp_queues",
 			dev->node_props.num_cp_queues);
+    sysfs_show_64bit_prop(buffer, "unique_id",
+			dev->node_props.unique_id);
 
 	if (dev->gpu) {
 		log_max_watch_addr =
@@ -1318,7 +1320,8 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 	dev->node_props.num_gws = (hws_gws_support &&
 		dev->gpu->dqm->sched_policy != KFD_SCHED_POLICY_NO_HWS) ?
 		amdgpu_amdkfd_get_num_gws(dev->gpu->kgd) : 0;
-	dev->node_props.num_cp_queues = get_queues_num(dev->gpu->dqm);
+	dev->node_props.num_cp_queues = get_cp_queues_num(dev->gpu->dqm);
+	dev->node_props.unique_id = gpu->unique_id;
 
 	kfd_fill_mem_clk_max_info(dev);
 	kfd_fill_iolink_non_crat_info(dev);

drivers/gpu/drm/amd/amdkfd/kfd_topology.h

@@ -54,6 +54,7 @@
 
 struct kfd_node_properties {
 	uint64_t hive_id;
+	uint64_t unique_id;
 	uint32_t cpu_cores_count;
 	uint32_t simd_count;
 	uint32_t mem_banks_count;

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -98,6 +98,9 @@ MODULE_FIRMWARE(FIRMWARE_RENOIR_DMUB);
 #define FIRMWARE_RAVEN_DMCU		"amdgpu/raven_dmcu.bin"
 MODULE_FIRMWARE(FIRMWARE_RAVEN_DMCU);
 
+#define FIRMWARE_NAVI12_DMCU            "amdgpu/navi12_dmcu.bin"
+MODULE_FIRMWARE(FIRMWARE_NAVI12_DMCU);
+
 /* Number of bytes in PSP header for firmware. */
 #define PSP_HEADER_BYTES 0x100
 
@@ -801,10 +804,20 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
 
 	fw_bss_data_size = le32_to_cpu(hdr->bss_data_bytes);
 
+	/* if adev->firmware.load_type == AMDGPU_FW_LOAD_PSP,
+	 * amdgpu_ucode_init_single_fw will load dmub firmware
+	 * fw_inst_const part to cw0; otherwise, the firmware back door load
+	 * will be done by dm_dmub_hw_init
+	 */
+	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
 		memcpy(fb_info->fb[DMUB_WINDOW_0_INST_CONST].cpu_addr, fw_inst_const,
 				fw_inst_const_size);
+	}
+
 	memcpy(fb_info->fb[DMUB_WINDOW_2_BSS_DATA].cpu_addr, fw_bss_data,
 	       fw_bss_data_size);
+
+	/* Copy firmware bios info into FB memory. */
 	memcpy(fb_info->fb[DMUB_WINDOW_3_VBIOS].cpu_addr, adev->bios,
 	       adev->bios_size);
 
@@ -823,6 +836,10 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
 	hw_params.fb_base = adev->gmc.fb_start;
 	hw_params.fb_offset = adev->gmc.aper_base;
 
+	/* backdoor load firmware and trigger dmub running */
+	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
+		hw_params.load_inst_const = true;
+
 	if (dmcu)
 		hw_params.psp_version = dmcu->psp_version;
 
@@ -960,7 +977,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 
 #ifdef CONFIG_DRM_AMD_DC_HDCP
 	if (adev->asic_type >= CHIP_RAVEN) {
-		adev->dm.hdcp_workqueue = hdcp_create_workqueue(&adev->psp, &init_params.cp_psp, adev->dm.dc);
+		adev->dm.hdcp_workqueue = hdcp_create_workqueue(adev, &init_params.cp_psp, adev->dm.dc);
 
 		if (!adev->dm.hdcp_workqueue)
 			DRM_ERROR("amdgpu: failed to initialize hdcp_workqueue.\n");
@@ -991,11 +1008,6 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 		goto error;
 	}
 
-#if defined(CONFIG_DEBUG_FS)
-	if (dtn_debugfs_init(adev))
-		DRM_ERROR("amdgpu: failed initialize dtn debugfs support.\n");
-#endif
-
 	DRM_DEBUG_DRIVER("KMS initialized.\n");
 
 	return 0;
@@ -1079,9 +1091,11 @@ static int load_dmcu_fw(struct amdgpu_device *adev)
 	case CHIP_VEGA20:
 	case CHIP_NAVI10:
 	case CHIP_NAVI14:
-	case CHIP_NAVI12:
 	case CHIP_RENOIR:
 		return 0;
+	case CHIP_NAVI12:
+		fw_name_dmcu = FIRMWARE_NAVI12_DMCU;
+		break;
 	case CHIP_RAVEN:
 		if (ASICREV_IS_PICASSO(adev->external_rev_id))
 			fw_name_dmcu = FIRMWARE_RAVEN_DMCU;
@@ -1192,22 +1206,21 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
 		return 0;
 	}
 
-	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
-		DRM_WARN("Only PSP firmware loading is supported for DMUB\n");
-		return 0;
-	}
-
 	hdr = (const struct dmcub_firmware_header_v1_0 *)adev->dm.dmub_fw->data;
+
+	if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
 		adev->firmware.ucode[AMDGPU_UCODE_ID_DMCUB].ucode_id =
 			AMDGPU_UCODE_ID_DMCUB;
-	adev->firmware.ucode[AMDGPU_UCODE_ID_DMCUB].fw = adev->dm.dmub_fw;
+		adev->firmware.ucode[AMDGPU_UCODE_ID_DMCUB].fw =
+			adev->dm.dmub_fw;
 		adev->firmware.fw_size +=
 			ALIGN(le32_to_cpu(hdr->inst_const_bytes), PAGE_SIZE);
 
-	adev->dm.dmcub_fw_version = le32_to_cpu(hdr->header.ucode_version);
-
 		DRM_INFO("Loading DMUB firmware via PSP: version=0x%08X\n",
 			 adev->dm.dmcub_fw_version);
+	}
+
+	adev->dm.dmcub_fw_version = le32_to_cpu(hdr->header.ucode_version);
 
 	adev->dm.dmub_srv = kzalloc(sizeof(*adev->dm.dmub_srv), GFP_KERNEL);
 	dmub_srv = adev->dm.dmub_srv;
@@ -1758,6 +1771,61 @@ static struct drm_mode_config_helper_funcs amdgpu_dm_mode_config_helperfuncs = {
 	.atomic_commit_tail = amdgpu_dm_atomic_commit_tail
 };
 
+static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector)
+{
+	u32 max_cll, min_cll, max, min, q, r;
+	struct amdgpu_dm_backlight_caps *caps;
+	struct amdgpu_display_manager *dm;
+	struct drm_connector *conn_base;
+	struct amdgpu_device *adev;
+	static const u8 pre_computed_values[] = {
+		50, 51, 52, 53, 55, 56, 57, 58, 59, 61, 62, 63, 65, 66, 68, 69,
+		71, 72, 74, 75, 77, 79, 81, 82, 84, 86, 88, 90, 92, 94, 96, 98};
+
+	if (!aconnector || !aconnector->dc_link)
+		return;
+
+	conn_base = &aconnector->base;
+	adev = conn_base->dev->dev_private;
+	dm = &adev->dm;
+	caps = &dm->backlight_caps;
+	caps->ext_caps = &aconnector->dc_link->dpcd_sink_ext_caps;
+	caps->aux_support = false;
+	max_cll = conn_base->hdr_sink_metadata.hdmi_type1.max_cll;
+	min_cll = conn_base->hdr_sink_metadata.hdmi_type1.min_cll;
+
+	if (caps->ext_caps->bits.oled == 1 ||
+	    caps->ext_caps->bits.sdr_aux_backlight_control == 1 ||
+	    caps->ext_caps->bits.hdr_aux_backlight_control == 1)
+		caps->aux_support = true;
+
+	/* From the specification (CTA-861-G), for calculating the maximum
+	 * luminance we need to use:
+	 *	Luminance = 50*2**(CV/32)
+	 * Where CV is a one-byte value.
+	 * For calculating this expression we may need float point precision;
+	 * to avoid this complexity level, we take advantage that CV is divided
+	 * by a constant. From the Euclids division algorithm, we know that CV
+	 * can be written as: CV = 32*q + r. Next, we replace CV in the
+	 * Luminance expression and get 50*(2**q)*(2**(r/32)), hence we just
+	 * need to pre-compute the value of r/32. For pre-computing the values
+	 * We just used the following Ruby line:
+	 *	(0...32).each {|cv| puts (50*2**(cv/32.0)).round}
+	 * The results of the above expressions can be verified at
+	 * pre_computed_values.
+	 */
+	q = max_cll >> 5;
+	r = max_cll % 32;
+	max = (1 << q) * pre_computed_values[r];
+
+	// min luminance: maxLum * (CV/255)^2 / 100
+	q = DIV_ROUND_CLOSEST(min_cll, 255);
+	min = max * DIV_ROUND_CLOSEST((q * q), 100);
+
+	caps->aux_max_input_signal = max;
+	caps->aux_min_input_signal = min;
+}
+
 static void
 amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector)
 {
@@ -1872,7 +1940,7 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector)
 					    aconnector->edid);
 		}
 		amdgpu_dm_update_freesync_caps(connector, aconnector->edid);
-
+		update_connector_ext_caps(aconnector);
 	} else {
 		drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux);
 		amdgpu_dm_update_freesync_caps(connector, NULL);
@@ -1911,7 +1979,7 @@ static void handle_hpd_irq(void *param)
 	mutex_lock(&aconnector->hpd_lock);
 
 #ifdef CONFIG_DRM_AMD_DC_HDCP
-	if (adev->asic_type >= CHIP_RAVEN)
+	if (adev->dm.hdcp_workqueue)
 		hdcp_reset_display(adev->dm.hdcp_workqueue, aconnector->dc_link->link_index);
 #endif
 	if (aconnector->fake_enable)
@@ -2088,8 +2156,10 @@ static void handle_hpd_rx_irq(void *param)
 		}
 	}
 #ifdef CONFIG_DRM_AMD_DC_HDCP
-	if (hpd_irq_data.bytes.device_service_irq.bits.CP_IRQ)
+	    if (hpd_irq_data.bytes.device_service_irq.bits.CP_IRQ) {
+		    if (adev->dm.hdcp_workqueue)
 			    hdcp_handle_cpirq(adev->dm.hdcp_workqueue,  aconnector->base.index);
+	    }
 #endif
 	if ((dc_link->cur_link_settings.lane_count != LANE_COUNT_UNKNOWN) ||
 	    (dc_link->type == dc_connection_mst_branch))
@@ -2484,6 +2554,7 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev)
 
 #define AMDGPU_DM_DEFAULT_MIN_BACKLIGHT 12
 #define AMDGPU_DM_DEFAULT_MAX_BACKLIGHT 255
+#define AUX_BL_DEFAULT_TRANSITION_TIME_MS 50
 
 #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
 	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
@@ -2498,9 +2569,11 @@ static void amdgpu_dm_update_backlight_caps(struct amdgpu_display_manager *dm)
 
 	amdgpu_acpi_get_backlight_caps(dm->adev, &caps);
 	if (caps.caps_valid) {
+		dm->backlight_caps.caps_valid = true;
+		if (caps.aux_support)
+			return;
 		dm->backlight_caps.min_input_signal = caps.min_input_signal;
 		dm->backlight_caps.max_input_signal = caps.max_input_signal;
-		dm->backlight_caps.caps_valid = true;
 	} else {
 		dm->backlight_caps.min_input_signal =
 				AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
@@ -2508,40 +2581,95 @@ static void amdgpu_dm_update_backlight_caps(struct amdgpu_display_manager *dm)
 				AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
 	}
 #else
+	if (dm->backlight_caps.aux_support)
+		return;
+
 	dm->backlight_caps.min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
 	dm->backlight_caps.max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
 #endif
 }
 
+static int set_backlight_via_aux(struct dc_link *link, uint32_t brightness)
+{
+	bool rc;
+
+	if (!link)
+		return 1;
+
+	rc = dc_link_set_backlight_level_nits(link, true, brightness,
+					      AUX_BL_DEFAULT_TRANSITION_TIME_MS);
+
+	return rc ? 0 : 1;
+}
+
+static u32 convert_brightness(const struct amdgpu_dm_backlight_caps *caps,
+			      const uint32_t user_brightness)
+{
+	u32 min, max, conversion_pace;
+	u32 brightness = user_brightness;
+
+	if (!caps)
+		goto out;
+
+	if (!caps->aux_support) {
+		max = caps->max_input_signal;
+		min = caps->min_input_signal;
+		/*
+		 * The brightness input is in the range 0-255
+		 * It needs to be rescaled to be between the
+		 * requested min and max input signal
+		 * It also needs to be scaled up by 0x101 to
+		 * match the DC interface which has a range of
+		 * 0 to 0xffff
+		 */
+		conversion_pace = 0x101;
+		brightness =
+			user_brightness
+			* conversion_pace
+			* (max - min)
+			/ AMDGPU_MAX_BL_LEVEL
+			+ min * conversion_pace;
+	} else {
+		/* TODO
+		 * We are doing a linear interpolation here, which is OK but
+		 * does not provide the optimal result. We probably want
+		 * something close to the Perceptual Quantizer (PQ) curve.
+		 */
+		max = caps->aux_max_input_signal;
+		min = caps->aux_min_input_signal;
+
+		brightness = (AMDGPU_MAX_BL_LEVEL - user_brightness) * min
+			       + user_brightness * max;
+		// Multiple the value by 1000 since we use millinits
+		brightness *= 1000;
+		brightness = DIV_ROUND_CLOSEST(brightness, AMDGPU_MAX_BL_LEVEL);
+	}
+
+out:
+	return brightness;
+}
+
 static int amdgpu_dm_backlight_update_status(struct backlight_device *bd)
 {
 	struct amdgpu_display_manager *dm = bl_get_data(bd);
 	struct amdgpu_dm_backlight_caps caps;
-	uint32_t brightness = bd->props.brightness;
+	struct dc_link *link = NULL;
+	u32 brightness;
+	bool rc;
 
 	amdgpu_dm_update_backlight_caps(dm);
 	caps = dm->backlight_caps;
-	/*
-	 * The brightness input is in the range 0-255
-	 * It needs to be rescaled to be between the
-	 * requested min and max input signal
-	 *
-	 * It also needs to be scaled up by 0x101 to
-	 * match the DC interface which has a range of
-	 * 0 to 0xffff
-	 */
-	brightness =
-		brightness
-		* 0x101
-		* (caps.max_input_signal - caps.min_input_signal)
-		/ AMDGPU_MAX_BL_LEVEL
-		+ caps.min_input_signal * 0x101;
 
-	if (dc_link_set_backlight_level(dm->backlight_link,
+	link = (struct dc_link *)dm->backlight_link;
-			brightness, 0))
+
-		return 0;
+	brightness = convert_brightness(&caps, bd->props.brightness);
-	else
+	// Change brightness based on AUX property
-		return 1;
+	if (caps.aux_support)
+		return set_backlight_via_aux(link, brightness);
+
+	rc = dc_link_set_backlight_level(dm->backlight_link, brightness, 0);
+
+	return rc ? 0 : 1;
 }
 
 static int amdgpu_dm_backlight_get_brightness(struct backlight_device *bd)
@@ -4493,6 +4621,19 @@ amdgpu_dm_connector_atomic_duplicate_state(struct drm_connector *connector)
 	return &new_state->base;
 }
 
+static int
+amdgpu_dm_connector_late_register(struct drm_connector *connector)
+{
+	struct amdgpu_dm_connector *amdgpu_dm_connector =
+		to_amdgpu_dm_connector(connector);
+
+#if defined(CONFIG_DEBUG_FS)
+	connector_debugfs_init(amdgpu_dm_connector);
+#endif
+
+	return 0;
+}
+
 static const struct drm_connector_funcs amdgpu_dm_connector_funcs = {
 	.reset = amdgpu_dm_connector_funcs_reset,
 	.detect = amdgpu_dm_connector_detect,
@@ -4502,6 +4643,7 @@ static const struct drm_connector_funcs amdgpu_dm_connector_funcs = {
 	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
 	.atomic_set_property = amdgpu_dm_connector_atomic_set_property,
 	.atomic_get_property = amdgpu_dm_connector_atomic_get_property,
+	.late_register = amdgpu_dm_connector_late_register,
 	.early_unregister = amdgpu_dm_connector_unregister
 };
 
@@ -5705,7 +5847,7 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 		drm_connector_attach_vrr_capable_property(
 			&aconnector->base);
 #ifdef CONFIG_DRM_AMD_DC_HDCP
-		if (adev->asic_type >= CHIP_RAVEN)
+		if (adev->dm.hdcp_workqueue)
 			drm_connector_attach_content_protection_property(&aconnector->base, true);
 #endif
 	}
@@ -5842,13 +5984,6 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
 	drm_connector_attach_encoder(
 		&aconnector->base, &aencoder->base);
 
-	drm_connector_register(&aconnector->base);
-#if defined(CONFIG_DEBUG_FS)
-	connector_debugfs_init(aconnector);
-	aconnector->debugfs_dpcd_address = 0;
-	aconnector->debugfs_dpcd_size = 0;
-#endif
-
 	if (connector_type == DRM_MODE_CONNECTOR_DisplayPort
 		|| connector_type == DRM_MODE_CONNECTOR_eDP)
 		amdgpu_dm_initialize_dp_connector(dm, aconnector);

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

@@ -90,15 +90,41 @@ struct dm_comressor_info {
 };
 
 /**
- * struct amdgpu_dm_backlight_caps - Usable range of backlight values from ACPI
+ * struct amdgpu_dm_backlight_caps - Information about backlight
- * @min_input_signal: minimum possible input in range 0-255
+ *
- * @max_input_signal: maximum possible input in range 0-255
+ * Describe the backlight support for ACPI or eDP AUX.
- * @caps_valid: true if these values are from the ACPI interface
  */
 struct amdgpu_dm_backlight_caps {
+	/**
+	 * @ext_caps: Keep the data struct with all the information about the
+	 * display support for HDR.
+	 */
+	union dpcd_sink_ext_caps *ext_caps;
+	/**
+	 * @aux_min_input_signal: Min brightness value supported by the display
+	 */
+	u32 aux_min_input_signal;
+	/**
+	 * @aux_max_input_signal: Max brightness value supported by the display
+	 * in nits.
+	 */
+	u32 aux_max_input_signal;
+	/**
+	 * @min_input_signal: minimum possible input in range 0-255.
+	 */
 	int min_input_signal;
+	/**
+	 * @max_input_signal: maximum possible input in range 0-255.
+	 */
 	int max_input_signal;
+	/**
+	 * @caps_valid: true if these values are from the ACPI interface.
+	 */
 	bool caps_valid;
+	/**
+	 * @aux_support: Describes if the display supports AUX backlight.
+	 */
+	bool aux_support;
 };
 
 /**

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c

@@ -32,6 +32,19 @@
 #include "amdgpu_dm.h"
 #include "amdgpu_dm_debugfs.h"
 #include "dm_helpers.h"
+#include "dmub/inc/dmub_srv.h"
+
+struct dmub_debugfs_trace_header {
+	uint32_t entry_count;
+	uint32_t reserved[3];
+};
+
+struct dmub_debugfs_trace_entry {
+	uint32_t trace_code;
+	uint32_t tick_count;
+	uint32_t param0;
+	uint32_t param1;
+};
 
 /* function description
  * get/ set DP configuration: lane_count, link_rate, spread_spectrum
@@ -675,6 +688,73 @@ static ssize_t dp_phy_test_pattern_debugfs_write(struct file *f, const char __us
 	return bytes_from_user;
 }
 
+/**
+ * Returns the DMCUB tracebuffer contents.
+ * Example usage: cat /sys/kernel/debug/dri/0/amdgpu_dm_dmub_tracebuffer
+ */
+static int dmub_tracebuffer_show(struct seq_file *m, void *data)
+{
+	struct amdgpu_device *adev = m->private;
+	struct dmub_srv_fb_info *fb_info = adev->dm.dmub_fb_info;
+	struct dmub_debugfs_trace_entry *entries;
+	uint8_t *tbuf_base;
+	uint32_t tbuf_size, max_entries, num_entries, i;
+
+	if (!fb_info)
+		return 0;
+
+	tbuf_base = (uint8_t *)fb_info->fb[DMUB_WINDOW_5_TRACEBUFF].cpu_addr;
+	if (!tbuf_base)
+		return 0;
+
+	tbuf_size = fb_info->fb[DMUB_WINDOW_5_TRACEBUFF].size;
+	max_entries = (tbuf_size - sizeof(struct dmub_debugfs_trace_header)) /
+		      sizeof(struct dmub_debugfs_trace_entry);
+
+	num_entries =
+		((struct dmub_debugfs_trace_header *)tbuf_base)->entry_count;
+
+	num_entries = min(num_entries, max_entries);
+
+	entries = (struct dmub_debugfs_trace_entry
+			   *)(tbuf_base +
+			      sizeof(struct dmub_debugfs_trace_header));
+
+	for (i = 0; i < num_entries; ++i) {
+		struct dmub_debugfs_trace_entry *entry = &entries[i];
+
+		seq_printf(m,
+			   "trace_code=%u tick_count=%u param0=%u param1=%u\n",
+			   entry->trace_code, entry->tick_count, entry->param0,
+			   entry->param1);
+	}
+
+	return 0;
+}
+
+/**
+ * Returns the DMCUB firmware state contents.
+ * Example usage: cat /sys/kernel/debug/dri/0/amdgpu_dm_dmub_fw_state
+ */
+static int dmub_fw_state_show(struct seq_file *m, void *data)
+{
+	struct amdgpu_device *adev = m->private;
+	struct dmub_srv_fb_info *fb_info = adev->dm.dmub_fb_info;
+	uint8_t *state_base;
+	uint32_t state_size;
+
+	if (!fb_info)
+		return 0;
+
+	state_base = (uint8_t *)fb_info->fb[DMUB_WINDOW_6_FW_STATE].cpu_addr;
+	if (!state_base)
+		return 0;
+
+	state_size = fb_info->fb[DMUB_WINDOW_6_FW_STATE].size;
+
+	return seq_write(m, state_base, state_size);
+}
+
 /*
  * Returns the current and maximum output bpc for the connector.
  * Example usage: cat /sys/kernel/debug/dri/0/DP-1/output_bpc
@@ -880,6 +960,8 @@ static ssize_t dp_dpcd_data_read(struct file *f, char __user *buf,
 	return read_size - r;
 }
 
+DEFINE_SHOW_ATTRIBUTE(dmub_fw_state);
+DEFINE_SHOW_ATTRIBUTE(dmub_tracebuffer);
 DEFINE_SHOW_ATTRIBUTE(output_bpc);
 DEFINE_SHOW_ATTRIBUTE(vrr_range);
 
@@ -1008,6 +1090,9 @@ void connector_debugfs_init(struct amdgpu_dm_connector *connector)
 	debugfs_create_file_unsafe("force_yuv420_output", 0644, dir, connector,
 				   &force_yuv420_output_fops);
 
+	connector->debugfs_dpcd_address = 0;
+	connector->debugfs_dpcd_size = 0;
+
 }
 
 /*
@@ -1188,5 +1273,11 @@ int dtn_debugfs_init(struct amdgpu_device *adev)
 	debugfs_create_file_unsafe("amdgpu_dm_visual_confirm", 0644, root, adev,
 				   &visual_confirm_fops);
 
+	debugfs_create_file_unsafe("amdgpu_dm_dmub_tracebuffer", 0644, root,
+				   adev, &dmub_tracebuffer_fops);
+
+	debugfs_create_file_unsafe("amdgpu_dm_dmub_fw_state", 0644, root,
+				   adev, &dmub_fw_state_fops);
+
 	return 0;
 }

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c

@@ -28,6 +28,13 @@
 #include "amdgpu_dm.h"
 #include "dm_helpers.h"
 #include <drm/drm_hdcp.h>
+#include "hdcp_psp.h"
+
+/*
+ * If the SRM version being loaded is less than or equal to the
+ * currently loaded SRM, psp will return 0xFFFF as the version
+ */
+#define PSP_SRM_VERSION_MAX 0xFFFF
 
 static bool
 lp_write_i2c(void *handle, uint32_t address, const uint8_t *data, uint32_t size)
@@ -67,6 +74,59 @@ lp_read_dpcd(void *handle, uint32_t address, uint8_t *data, uint32_t size)
 	return dm_helpers_dp_read_dpcd(link->ctx, link, address, data, size);
 }
 
+static uint8_t *psp_get_srm(struct psp_context *psp, uint32_t *srm_version, uint32_t *srm_size)
+{
+
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	if (!psp->hdcp_context.hdcp_initialized) {
+		DRM_WARN("Failed to get hdcp srm. HDCP TA is not initialized.");
+		return NULL;
+	}
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP_GET_SRM;
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS)
+		return NULL;
+
+	*srm_version = hdcp_cmd->out_msg.hdcp_get_srm.srm_version;
+	*srm_size = hdcp_cmd->out_msg.hdcp_get_srm.srm_buf_size;
+
+
+	return hdcp_cmd->out_msg.hdcp_get_srm.srm_buf;
+}
+
+static int psp_set_srm(struct psp_context *psp, uint8_t *srm, uint32_t srm_size, uint32_t *srm_version)
+{
+
+	struct ta_hdcp_shared_memory *hdcp_cmd;
+
+	if (!psp->hdcp_context.hdcp_initialized) {
+		DRM_WARN("Failed to get hdcp srm. HDCP TA is not initialized.");
+		return -EINVAL;
+	}
+
+	hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.hdcp_shared_buf;
+	memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
+
+	memcpy(hdcp_cmd->in_msg.hdcp_set_srm.srm_buf, srm, srm_size);
+	hdcp_cmd->in_msg.hdcp_set_srm.srm_buf_size = srm_size;
+	hdcp_cmd->cmd_id = TA_HDCP_COMMAND__HDCP_SET_SRM;
+
+	psp_hdcp_invoke(psp, hdcp_cmd->cmd_id);
+
+	if (hdcp_cmd->hdcp_status != TA_HDCP_STATUS__SUCCESS || hdcp_cmd->out_msg.hdcp_set_srm.valid_signature != 1 ||
+	    hdcp_cmd->out_msg.hdcp_set_srm.srm_version == PSP_SRM_VERSION_MAX)
+		return -EINVAL;
+
+	*srm_version = hdcp_cmd->out_msg.hdcp_set_srm.srm_version;
+	return 0;
+}
+
 static void process_output(struct hdcp_workqueue *hdcp_work)
 {
 	struct mod_hdcp_output output = hdcp_work->output;
@@ -88,6 +148,18 @@ static void process_output(struct hdcp_workqueue *hdcp_work)
 	schedule_delayed_work(&hdcp_work->property_validate_dwork, msecs_to_jiffies(0));
 }
 
+static void link_lock(struct hdcp_workqueue *work, bool lock)
+{
+
+	int i = 0;
+
+	for (i = 0; i < work->max_link; i++) {
+		if (lock)
+			mutex_lock(&work[i].mutex);
+		else
+			mutex_unlock(&work[i].mutex);
+	}
+}
 void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
 			 unsigned int link_index,
 			 struct amdgpu_dm_connector *aconnector,
@@ -112,6 +184,13 @@ void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
 		hdcp_w->link.adjust.hdcp2.force_type = MOD_HDCP_FORCE_TYPE_0;
 
 		if (enable_encryption) {
+			/* Explicitly set the saved SRM as sysfs call will be after we already enabled hdcp
+			 * (s3 resume case)
+			 */
+			if (hdcp_work->srm_size > 0)
+				psp_set_srm(hdcp_work->hdcp.config.psp.handle, hdcp_work->srm, hdcp_work->srm_size,
+					    &hdcp_work->srm_version);
+
 			display->adjust.disable = 0;
 			if (content_type == DRM_MODE_HDCP_CONTENT_TYPE0)
 				hdcp_w->link.adjust.hdcp2.force_type = MOD_HDCP_FORCE_TYPE_0;
@@ -301,8 +380,9 @@ void hdcp_destroy(struct hdcp_workqueue *hdcp_work)
 		cancel_delayed_work_sync(&hdcp_work[i].watchdog_timer_dwork);
 	}
 
+	kfree(hdcp_work->srm);
+	kfree(hdcp_work->srm_temp);
 	kfree(hdcp_work);
-
 }
 
 static void update_config(void *handle, struct cp_psp_stream_config *config)
@@ -338,20 +418,163 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
 	hdcp_update_display(hdcp_work, link_index, aconnector, DRM_MODE_HDCP_CONTENT_TYPE0, false);
 }
 
-struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *cp_psp, struct dc *dc)
+
+/* NOTE: From the usermodes prospective you only need to call write *ONCE*, the kernel
+ *      will automatically call once or twice depending on the size
+ *
+ * call: "cat file > /sys/class/drm/card0/device/hdcp_srm" from usermode no matter what the size is
+ *
+ * The kernel can only send PAGE_SIZE at once and since MAX_SRM_FILE(5120) > PAGE_SIZE(4096),
+ * srm_data_write can be called multiple times.
+ *
+ * sysfs interface doesn't tell us the size we will get so we are sending partial SRMs to psp and on
+ * the last call we will send the full SRM. PSP will fail on every call before the last.
+ *
+ * This means we don't know if the SRM is good until the last call. And because of this limitation we
+ * cannot throw errors early as it will stop the kernel from writing to sysfs
+ *
+ * Example 1:
+ * 	Good SRM size = 5096
+ * 	first call to write 4096 -> PSP fails
+ * 	Second call to write 1000 -> PSP Pass -> SRM is set
+ *
+ * Example 2:
+ * 	Bad SRM size = 4096
+ * 	first call to write 4096 -> PSP fails (This is the same as above, but we don't know if this
+ * 	is the last call)
+ *
+ * Solution?:
+ * 	1: Parse the SRM? -> It is signed so we don't know the EOF
+ * 	2: We can have another sysfs that passes the size before calling set. -> simpler solution
+ * 	below
+ *
+ * Easy Solution:
+ * Always call get after Set to verify if set was successful.
+ * +----------------------+
+ * |   Why it works:      |
+ * +----------------------+
+ * PSP will only update its srm if its older than the one we are trying to load.
+ * Always do set first than get.
+ * 	-if we try to "1. SET" a older version PSP will reject it and we can "2. GET" the newer
+ * 	version and save it
+ *
+ * 	-if we try to "1. SET" a newer version PSP will accept it and we can "2. GET" the
+ * 	same(newer) version back and save it
+ *
+ * 	-if we try to "1. SET" a newer version and PSP rejects it. That means the format is
+ * 	incorrect/corrupted and we should correct our SRM by getting it from PSP
+ */
+static ssize_t srm_data_write(struct file *filp, struct kobject *kobj, struct bin_attribute *bin_attr, char *buffer,
+			      loff_t pos, size_t count)
+{
+	struct hdcp_workqueue *work;
+	uint32_t srm_version = 0;
+
+	work = container_of(bin_attr, struct hdcp_workqueue, attr);
+	link_lock(work, true);
+
+	memcpy(work->srm_temp + pos, buffer, count);
+
+	if (!psp_set_srm(work->hdcp.config.psp.handle, work->srm_temp, pos + count, &srm_version)) {
+		DRM_DEBUG_DRIVER("HDCP SRM SET version 0x%X", srm_version);
+		memcpy(work->srm, work->srm_temp, pos + count);
+		work->srm_size = pos + count;
+		work->srm_version = srm_version;
+	}
+
+
+	link_lock(work, false);
+
+	return count;
+}
+
+static ssize_t srm_data_read(struct file *filp, struct kobject *kobj, struct bin_attribute *bin_attr, char *buffer,
+			     loff_t pos, size_t count)
+{
+	struct hdcp_workqueue *work;
+	uint8_t *srm = NULL;
+	uint32_t srm_version;
+	uint32_t srm_size;
+	size_t ret = count;
+
+	work = container_of(bin_attr, struct hdcp_workqueue, attr);
+
+	link_lock(work, true);
+
+	srm = psp_get_srm(work->hdcp.config.psp.handle, &srm_version, &srm_size);
+
+	if (!srm)
+		return -EINVAL;
+
+	if (pos >= srm_size)
+		ret = 0;
+
+	if (srm_size - pos < count) {
+		memcpy(buffer, srm + pos, srm_size - pos);
+		ret = srm_size - pos;
+		goto ret;
+	}
+
+	memcpy(buffer, srm + pos, count);
+
+ret:
+	link_lock(work, false);
+	return ret;
+}
+
+/* From the hdcp spec (5.Renewability) SRM needs to be stored in a non-volatile memory.
+ *
+ * For example,
+ * 	if Application "A" sets the SRM (ver 2) and we reboot/suspend and later when Application "B"
+ * 	needs to use HDCP, the version in PSP should be SRM(ver 2). So SRM should be persistent
+ * 	across boot/reboots/suspend/resume/shutdown
+ *
+ * Currently when the system goes down (suspend/shutdown) the SRM is cleared from PSP. For HDCP we need
+ * to make the SRM persistent.
+ *
+ * -PSP owns the checking of SRM but doesn't have the ability to store it in a non-volatile memory.
+ * -The kernel cannot write to the file systems.
+ * -So we need usermode to do this for us, which is why an interface for usermode is needed
+ *
+ *
+ *
+ * Usermode can read/write to/from PSP using the sysfs interface
+ * For example:
+ * 	to save SRM from PSP to storage : cat /sys/class/drm/card0/device/hdcp_srm > srmfile
+ * 	to load from storage to PSP: cat srmfile > /sys/class/drm/card0/device/hdcp_srm
+ */
+static const struct bin_attribute data_attr = {
+	.attr = {.name = "hdcp_srm", .mode = 0664},
+	.size = PSP_HDCP_SRM_FIRST_GEN_MAX_SIZE, /* Limit SRM size */
+	.write = srm_data_write,
+	.read = srm_data_read,
+};
+
+
+struct hdcp_workqueue *hdcp_create_workqueue(struct amdgpu_device *adev, struct cp_psp *cp_psp, struct dc *dc)
 {
 
 	int max_caps = dc->caps.max_links;
-	struct hdcp_workqueue *hdcp_work = kzalloc(max_caps*sizeof(*hdcp_work), GFP_KERNEL);
+	struct hdcp_workqueue *hdcp_work;
 	int i = 0;
 
+	hdcp_work = kcalloc(max_caps, sizeof(*hdcp_work), GFP_KERNEL);
 	if (hdcp_work == NULL)
+		return NULL;
+
+	hdcp_work->srm = kcalloc(PSP_HDCP_SRM_FIRST_GEN_MAX_SIZE, sizeof(*hdcp_work->srm), GFP_KERNEL);
+
+	if (hdcp_work->srm == NULL)
+		goto fail_alloc_context;
+
+	hdcp_work->srm_temp = kcalloc(PSP_HDCP_SRM_FIRST_GEN_MAX_SIZE, sizeof(*hdcp_work->srm_temp), GFP_KERNEL);
+
+	if (hdcp_work->srm_temp == NULL)
 		goto fail_alloc_context;
 
 	hdcp_work->max_link = max_caps;
 
 	for (i = 0; i < max_caps; i++) {
-
 		mutex_init(&hdcp_work[i].mutex);
 
 		INIT_WORK(&hdcp_work[i].cpirq_work, event_cpirq);
@@ -360,7 +583,7 @@ struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *c
 		INIT_DELAYED_WORK(&hdcp_work[i].watchdog_timer_dwork, event_watchdog_timer);
 		INIT_DELAYED_WORK(&hdcp_work[i].property_validate_dwork, event_property_validate);
 
-		hdcp_work[i].hdcp.config.psp.handle =  psp_context;
+		hdcp_work[i].hdcp.config.psp.handle = &adev->psp;
 		hdcp_work[i].hdcp.config.ddc.handle = dc_get_link_at_index(dc, i);
 		hdcp_work[i].hdcp.config.ddc.funcs.write_i2c = lp_write_i2c;
 		hdcp_work[i].hdcp.config.ddc.funcs.read_i2c = lp_read_i2c;
@@ -371,9 +594,17 @@ struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *c
 	cp_psp->funcs.update_stream_config = update_config;
 	cp_psp->handle = hdcp_work;
 
+	/* File created at /sys/class/drm/card0/device/hdcp_srm*/
+	hdcp_work[0].attr = data_attr;
+
+	if (sysfs_create_bin_file(&adev->dev->kobj, &hdcp_work[0].attr))
+		DRM_WARN("Failed to create device file hdcp_srm");
+
 	return hdcp_work;
 
 fail_alloc_context:
+	kfree(hdcp_work->srm);
+	kfree(hdcp_work->srm_temp);
 	kfree(hdcp_work);
 
 	return NULL;

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.h

@@ -30,6 +30,7 @@
 #include "hdcp.h"
 #include "dc.h"
 #include "dm_cp_psp.h"
+#include "amdgpu.h"
 
 struct mod_hdcp;
 struct mod_hdcp_link;
@@ -52,6 +53,12 @@ struct hdcp_workqueue {
 
 	enum mod_hdcp_encryption_status encryption_status;
 	uint8_t max_link;
+
+	uint8_t *srm;
+	uint8_t *srm_temp;
+	uint32_t srm_version;
+	uint32_t srm_size;
+	struct bin_attribute attr;
 };
 
 void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
@@ -64,6 +71,6 @@ void hdcp_reset_display(struct hdcp_workqueue *work, unsigned int link_index);
 void hdcp_handle_cpirq(struct hdcp_workqueue *work, unsigned int link_index);
 void hdcp_destroy(struct hdcp_workqueue *work);
 
-struct hdcp_workqueue *hdcp_create_workqueue(void *psp_context, struct cp_psp *cp_psp, struct dc *dc);
+struct hdcp_workqueue *hdcp_create_workqueue(struct amdgpu_device *adev, struct cp_psp *cp_psp, struct dc *dc);
 
 #endif /* AMDGPU_DM_AMDGPU_DM_HDCP_H_ */

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c

@@ -154,15 +154,18 @@ amdgpu_dm_mst_connector_late_register(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *amdgpu_dm_connector =
 		to_amdgpu_dm_connector(connector);
-	struct drm_dp_mst_port *port = amdgpu_dm_connector->port;
+	int r;
+
+	amdgpu_dm_connector->dm_dp_aux.aux.dev = connector->kdev;
+	r = drm_dp_aux_register(&amdgpu_dm_connector->dm_dp_aux.aux);
+	if (r)
+		return r;
 
 #if defined(CONFIG_DEBUG_FS)
 	connector_debugfs_init(amdgpu_dm_connector);
-	amdgpu_dm_connector->debugfs_dpcd_address = 0;
-	amdgpu_dm_connector->debugfs_dpcd_size = 0;
 #endif
 
-	return drm_dp_mst_connector_late_register(connector, port);
+	return r;
 }
 
 static void
@@ -482,11 +485,10 @@ void amdgpu_dm_initialize_dp_connector(struct amdgpu_display_manager *dm,
 				       struct amdgpu_dm_connector *aconnector)
 {
 	aconnector->dm_dp_aux.aux.name = "dmdc";
-	aconnector->dm_dp_aux.aux.dev = aconnector->base.kdev;
 	aconnector->dm_dp_aux.aux.transfer = dm_dp_aux_transfer;
 	aconnector->dm_dp_aux.ddc_service = aconnector->dc_link->ddc;
 
-	drm_dp_aux_register(&aconnector->dm_dp_aux.aux);
+	drm_dp_aux_init(&aconnector->dm_dp_aux.aux);
 	drm_dp_cec_register_connector(&aconnector->dm_dp_aux.aux,
 				      &aconnector->base);
 

drivers/gpu/drm/amd/display/dc/bios/command_table2.c

@@ -221,8 +221,8 @@ static void init_transmitter_control(struct bios_parser *bp)
 	uint8_t frev;
 	uint8_t crev;
 
-	if (BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev) == false)
+	BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev);
-		BREAK_TO_DEBUGGER();
+
 	switch (crev) {
 	case 6:
 		bp->cmd_tbl.transmitter_control = transmitter_control_v1_6;

drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c

@@ -53,25 +53,18 @@ bool dal_bios_parser_init_cmd_tbl_helper2(
 
 	case DCE_VERSION_11_2:
 	case DCE_VERSION_11_22:
+	case DCE_VERSION_12_0:
+	case DCE_VERSION_12_1:
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	case DCN_VERSION_1_0:
 	case DCN_VERSION_1_01:
-		*h = dal_cmd_tbl_helper_dce112_get_table2();
-		return true;
-#endif
-
 	case DCN_VERSION_2_0:
-		*h = dal_cmd_tbl_helper_dce112_get_table2();
-		return true;
 	case DCN_VERSION_2_1:
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
-	case DCE_VERSION_12_0:
+#endif
-	case DCE_VERSION_12_1:
-		*h = dal_cmd_tbl_helper_dce112_get_table2();
-		return true;
 
 	default:
 		/* Unsupported DCE */

drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c

@@ -703,11 +703,19 @@ static void hack_bounding_box(struct dcn_bw_internal_vars *v,
 }
 
 
-unsigned int get_highest_allowed_voltage_level(uint32_t hw_internal_rev)
+unsigned int get_highest_allowed_voltage_level(uint32_t hw_internal_rev, uint32_t pci_revision_id)
 {
-	/* for dali & pollock, the highest voltage level we want is 0 */
+	/* for low power RV2 variants, the highest voltage level we want is 0 */
-	if (ASICREV_IS_POLLOCK(hw_internal_rev) || ASICREV_IS_DALI(hw_internal_rev))
+	if (ASICREV_IS_RAVEN2(hw_internal_rev))
+		switch (pci_revision_id) {
+		case PRID_DALI_DE:
+		case PRID_DALI_DF:
+		case PRID_DALI_E3:
+		case PRID_DALI_E4:
 			return 0;
+		default:
+			break;
+		}
 
 	/* we are ok with all levels */
 	return 4;
@@ -1277,7 +1285,9 @@ bool dcn_validate_bandwidth(
 	PERFORMANCE_TRACE_END();
 	BW_VAL_TRACE_FINISH();
 
-	if (bw_limit_pass && v->voltage_level <= get_highest_allowed_voltage_level(dc->ctx->asic_id.hw_internal_rev))
+	if (bw_limit_pass && v->voltage_level <= get_highest_allowed_voltage_level(
+							dc->ctx->asic_id.hw_internal_rev,
+							dc->ctx->asic_id.pci_revision_id))
 		return true;
 	else
 		return false;

drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c

@@ -63,6 +63,25 @@ int clk_mgr_helper_get_active_display_cnt(
 	return display_count;
 }
 
+int clk_mgr_helper_get_active_plane_cnt(
+		struct dc *dc,
+		struct dc_state *context)
+{
+	int i, total_plane_count;
+
+	total_plane_count = 0;
+	for (i = 0; i < context->stream_count; i++) {
+		const struct dc_stream_status stream_status = context->stream_status[i];
+
+		/*
+		 * Sum up plane_count for all streams ( active and virtual ).
+		 */
+		total_plane_count += stream_status.plane_count;
+	}
+
+	return total_plane_count;
+}
+
 void clk_mgr_exit_optimized_pwr_state(const struct dc *dc, struct clk_mgr *clk_mgr)
 {
 	struct dc_link *edp_link = get_edp_link(dc);
@@ -134,13 +153,6 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p
 
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	case FAMILY_RV:
-		if (ASICREV_IS_DALI(asic_id.hw_internal_rev) ||
-				ASICREV_IS_POLLOCK(asic_id.hw_internal_rev)) {
-			/* TEMP: this check has to come before ASICREV_IS_RENOIR */
-			/* which also incorrectly returns true for Dali/Pollock*/
-			rv2_clk_mgr_construct(ctx, clk_mgr, pp_smu);
-			break;
-		}
 		if (ASICREV_IS_RENOIR(asic_id.hw_internal_rev)) {
 			rn_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
 			break;

drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c

@@ -158,6 +158,8 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 	bool dpp_clock_lowered = false;
 	struct dmcu *dmcu = clk_mgr_base->ctx->dc->res_pool->dmcu;
 	bool force_reset = false;
+	bool p_state_change_support;
+	int total_plane_count;
 
 	if (dc->work_arounds.skip_clock_update)
 		return;
@@ -213,9 +215,11 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base,
 			pp_smu->set_hard_min_socclk_by_freq(&pp_smu->pp_smu, clk_mgr_base->clks.socclk_khz / 1000);
 	}
 
-	if (should_update_pstate_support(safe_to_lower, new_clocks->p_state_change_support, clk_mgr_base->clks.p_state_change_support)) {
+	total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context);
+	p_state_change_support = new_clocks->p_state_change_support || (total_plane_count == 0);
+	if (should_update_pstate_support(safe_to_lower, p_state_change_support, clk_mgr_base->clks.p_state_change_support)) {
 		clk_mgr_base->clks.prev_p_state_change_support = clk_mgr_base->clks.p_state_change_support;
-		clk_mgr_base->clks.p_state_change_support = new_clocks->p_state_change_support;
+		clk_mgr_base->clks.p_state_change_support = p_state_change_support;
 		if (pp_smu && pp_smu->set_pstate_handshake_support)
 			pp_smu->set_pstate_handshake_support(&pp_smu->pp_smu, clk_mgr_base->clks.p_state_change_support);
 	}

drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr.c

@@ -405,7 +405,7 @@ void rn_init_clocks(struct clk_mgr *clk_mgr)
 	clk_mgr->clks.pwr_state = DCN_PWR_STATE_UNKNOWN;
 }
 
-void build_watermark_ranges(struct clk_bw_params *bw_params, struct pp_smu_wm_range_sets *ranges)
+static void build_watermark_ranges(struct clk_bw_params *bw_params, struct pp_smu_wm_range_sets *ranges)
 {
 	int i, num_valid_sets;
 
@@ -465,16 +465,15 @@ void build_watermark_ranges(struct clk_bw_params *bw_params, struct pp_smu_wm_ra
 static void rn_notify_wm_ranges(struct clk_mgr *clk_mgr_base)
 {
 	struct dc_debug_options *debug = &clk_mgr_base->ctx->dc->debug;
-	struct pp_smu_wm_range_sets ranges = {0};
 	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
 	struct pp_smu_funcs *pp_smu = clk_mgr->pp_smu;
 
 	if (!debug->disable_pplib_wm_range) {
-		build_watermark_ranges(clk_mgr_base->bw_params, &ranges);
+		build_watermark_ranges(clk_mgr_base->bw_params, &clk_mgr_base->ranges);
 
 		/* Notify PP Lib/SMU which Watermarks to use for which clock ranges */
 		if (pp_smu && pp_smu->rn_funcs.set_wm_ranges)
-			pp_smu->rn_funcs.set_wm_ranges(&pp_smu->rn_funcs.pp_smu, &ranges);
+			pp_smu->rn_funcs.set_wm_ranges(&pp_smu->rn_funcs.pp_smu, &clk_mgr_base->ranges);
 	}
 
 }
@@ -504,7 +503,7 @@ static struct clk_mgr_funcs dcn21_funcs = {
 	.notify_wm_ranges = rn_notify_wm_ranges
 };
 
-struct clk_bw_params rn_bw_params = {
+static struct clk_bw_params rn_bw_params = {
 	.vram_type = Ddr4MemType,
 	.num_channels = 1,
 	.clk_table = {
@@ -544,7 +543,7 @@ struct clk_bw_params rn_bw_params = {
 
 };
 
-struct wm_table ddr4_wm_table = {
+static struct wm_table ddr4_wm_table = {
 	.entries = {
 		{
 			.wm_inst = WM_A,
@@ -581,7 +580,7 @@ struct wm_table ddr4_wm_table = {
 	}
 };
 
-struct wm_table lpddr4_wm_table = {
+static struct wm_table lpddr4_wm_table = {
 	.entries = {
 		{
 			.wm_inst = WM_A,

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

@@ -701,7 +701,7 @@ static bool dc_construct(struct dc *dc,
 		dc_ctx->created_bios = true;
 	}
 
-
+	dc->vendor_signature = init_params->vendor_signature;
 
 	/* Create GPIO service */
 	dc_ctx->gpio_service = dal_gpio_service_create(
@@ -761,6 +761,28 @@ static bool disable_all_writeback_pipes_for_stream(
 	return true;
 }
 
+void apply_ctx_interdependent_lock(struct dc *dc, struct dc_state *context, struct dc_stream_state *stream, bool lock)
+{
+	int i = 0;
+
+	/* Checks if interdependent update function pointer is NULL or not, takes care of DCE110 case */
+	if (dc->hwss.interdependent_update_lock)
+		dc->hwss.interdependent_update_lock(dc, context, lock);
+	else {
+		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+			struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+			struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+
+			// Copied conditions that were previously in dce110_apply_ctx_for_surface
+			if (stream == pipe_ctx->stream) {
+				if (!pipe_ctx->top_pipe &&
+					(pipe_ctx->plane_state || old_pipe_ctx->plane_state))
+					dc->hwss.pipe_control_lock(dc, pipe_ctx, lock);
+			}
+		}
+	}
+}
+
 static void disable_dangling_plane(struct dc *dc, struct dc_state *context)
 {
 	int i, j;
@@ -786,11 +808,20 @@ static void disable_dangling_plane(struct dc *dc, struct dc_state *context)
 		if (should_disable && old_stream) {
 			dc_rem_all_planes_for_stream(dc, old_stream, dangling_context);
 			disable_all_writeback_pipes_for_stream(dc, old_stream, dangling_context);
-			if (dc->hwss.apply_ctx_for_surface)
+
+			if (dc->hwss.apply_ctx_for_surface) {
+				apply_ctx_interdependent_lock(dc, dc->current_state, old_stream, true);
 				dc->hwss.apply_ctx_for_surface(dc, old_stream, 0, dangling_context);
+				apply_ctx_interdependent_lock(dc, dc->current_state, old_stream, false);
+				dc->hwss.post_unlock_program_front_end(dc, dangling_context);
 			}
-		if (dc->hwss.program_front_end_for_ctx)
+			if (dc->hwss.program_front_end_for_ctx) {
+				dc->hwss.interdependent_update_lock(dc, dc->current_state, true);
 				dc->hwss.program_front_end_for_ctx(dc, dangling_context);
+				dc->hwss.interdependent_update_lock(dc, dc->current_state, false);
+				dc->hwss.post_unlock_program_front_end(dc, dangling_context);
+			}
+		}
 	}
 
 	current_ctx = dc->current_state;
@@ -1210,15 +1241,18 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	/* re-program planes for existing stream, in case we need to
 	 * free up plane resource for later use
 	 */
-	if (dc->hwss.apply_ctx_for_surface)
+	if (dc->hwss.apply_ctx_for_surface) {
 		for (i = 0; i < context->stream_count; i++) {
 			if (context->streams[i]->mode_changed)
 				continue;
-
+			apply_ctx_interdependent_lock(dc, context, context->streams[i], true);
 			dc->hwss.apply_ctx_for_surface(
 				dc, context->streams[i],
 				context->stream_status[i].plane_count,
 				context); /* use new pipe config in new context */
+			apply_ctx_interdependent_lock(dc, context, context->streams[i], false);
+			dc->hwss.post_unlock_program_front_end(dc, context);
+		}
 	}
 
 	/* Program hardware */
@@ -1238,19 +1272,27 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	}
 
 	/* Program all planes within new context*/
-	if (dc->hwss.program_front_end_for_ctx)
+	if (dc->hwss.program_front_end_for_ctx) {
+		dc->hwss.interdependent_update_lock(dc, context, true);
 		dc->hwss.program_front_end_for_ctx(dc, context);
+		dc->hwss.interdependent_update_lock(dc, context, false);
+		dc->hwss.post_unlock_program_front_end(dc, context);
+	}
 	for (i = 0; i < context->stream_count; i++) {
 		const struct dc_link *link = context->streams[i]->link;
 
 		if (!context->streams[i]->mode_changed)
 			continue;
 
-		if (dc->hwss.apply_ctx_for_surface)
+		if (dc->hwss.apply_ctx_for_surface) {
+			apply_ctx_interdependent_lock(dc, context, context->streams[i], true);
 			dc->hwss.apply_ctx_for_surface(
 					dc, context->streams[i],
 					context->stream_status[i].plane_count,
 					context);
+			apply_ctx_interdependent_lock(dc, context, context->streams[i], false);
+			dc->hwss.post_unlock_program_front_end(dc, context);
+		}
 
 		/*
 		 * enable stereo
@@ -1318,18 +1360,12 @@ bool dc_commit_state(struct dc *dc, struct dc_state *context)
 	return (result == DC_OK);
 }
 
-bool dc_is_hw_initialized(struct dc *dc)
-{
-	struct dc_bios *dcb = dc->ctx->dc_bios;
-	return dcb->funcs->is_accelerated_mode(dcb);
-}
-
 bool dc_post_update_surfaces_to_stream(struct dc *dc)
 {
 	int i;
 	struct dc_state *context = dc->current_state;
 
-	if (!dc->optimized_required || dc->optimize_seamless_boot_streams > 0)
+	if ((!dc->clk_optimized_required && !dc->wm_optimized_required) || dc->optimize_seamless_boot_streams > 0)
 		return true;
 
 	post_surface_trace(dc);
@@ -1341,8 +1377,6 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc)
 			dc->hwss.disable_plane(dc, &context->res_ctx.pipe_ctx[i]);
 		}
 
-	dc->optimized_required = false;
-
 	dc->hwss.optimize_bandwidth(dc, context);
 	return true;
 }
@@ -1734,14 +1768,15 @@ static enum surface_update_type check_update_surfaces_for_stream(
 
 		if (stream_update->wb_update)
 			su_flags->bits.wb_update = 1;
+
+		if (stream_update->dsc_config)
+			su_flags->bits.dsc_changed = 1;
+
 		if (su_flags->raw != 0)
 			overall_type = UPDATE_TYPE_FULL;
 
 		if (stream_update->output_csc_transform || stream_update->output_color_space)
 			su_flags->bits.out_csc = 1;
-
-		if (stream_update->dsc_config)
-			overall_type = UPDATE_TYPE_FULL;
 	}
 
 	for (i = 0 ; i < surface_count; i++) {
@@ -1776,8 +1811,11 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
 
 	type = check_update_surfaces_for_stream(dc, updates, surface_count, stream_update, stream_status);
 	if (type == UPDATE_TYPE_FULL) {
-		if (stream_update)
+		if (stream_update) {
+			uint32_t dsc_changed = stream_update->stream->update_flags.bits.dsc_changed;
 			stream_update->stream->update_flags.raw = 0xFFFFFFFF;
+			stream_update->stream->update_flags.bits.dsc_changed = dsc_changed;
+		}
 		for (i = 0; i < surface_count; i++)
 			updates[i].surface->update_flags.raw = 0xFFFFFFFF;
 	}
@@ -1786,10 +1824,10 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
 		// If there's an available clock comparator, we use that.
 		if (dc->clk_mgr->funcs->are_clock_states_equal) {
 			if (!dc->clk_mgr->funcs->are_clock_states_equal(&dc->clk_mgr->clks, &dc->current_state->bw_ctx.bw.dcn.clk))
-				dc->optimized_required = true;
+				dc->clk_optimized_required = true;
 		// Else we fallback to mem compare.
 		} else if (memcmp(&dc->current_state->bw_ctx.bw.dcn.clk, &dc->clk_mgr->clks, offsetof(struct dc_clocks, prev_p_state_change_support)) != 0) {
-			dc->optimized_required = true;
+			dc->clk_optimized_required = true;
 		}
 	}
 
@@ -2093,18 +2131,14 @@ static void commit_planes_do_stream_update(struct dc *dc,
 				}
 			}
 
-			if (stream_update->dsc_config && dc->hwss.pipe_control_lock_global) {
-				dc->hwss.pipe_control_lock_global(dc, pipe_ctx, true);
-				dp_update_dsc_config(pipe_ctx);
-				dc->hwss.pipe_control_lock_global(dc, pipe_ctx, false);
-			}
 			/* Full fe update*/
 			if (update_type == UPDATE_TYPE_FAST)
 				continue;
 
-			if (stream_update->dpms_off) {
+			if (stream_update->dsc_config)
-				dc->hwss.pipe_control_lock(dc, pipe_ctx, true);
+				dp_update_dsc_config(pipe_ctx);
 
+			if (stream_update->dpms_off) {
 				if (*stream_update->dpms_off) {
 					core_link_disable_stream(pipe_ctx);
 					/* for dpms, keep acquired resources*/
@@ -2118,8 +2152,6 @@ static void commit_planes_do_stream_update(struct dc *dc,
 
 					core_link_enable_stream(dc->current_state, pipe_ctx);
 				}
-
-				dc->hwss.pipe_control_lock(dc, pipe_ctx, false);
 			}
 
 			if (stream_update->abm_level && pipe_ctx->stream_res.abm) {
@@ -2166,7 +2198,7 @@ static void commit_planes_for_stream(struct dc *dc,
 			dc->optimize_seamless_boot_streams--;
 
 			if (dc->optimize_seamless_boot_streams == 0)
-				dc->optimized_required = true;
+				dc->clk_optimized_required = true;
 		}
 	}
 
@@ -2175,6 +2207,32 @@ static void commit_planes_for_stream(struct dc *dc,
 		context_clock_trace(dc, context);
 	}
 
+	for (j = 0; j < dc->res_pool->pipe_count; j++) {
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j];
+
+		if (!pipe_ctx->top_pipe &&
+			!pipe_ctx->prev_odm_pipe &&
+			pipe_ctx->stream &&
+			pipe_ctx->stream == stream) {
+			top_pipe_to_program = pipe_ctx;
+		}
+	}
+
+	if ((update_type != UPDATE_TYPE_FAST) && stream->update_flags.bits.dsc_changed)
+		if (top_pipe_to_program->stream_res.tg->funcs->lock_doublebuffer_enable)
+			top_pipe_to_program->stream_res.tg->funcs->lock_doublebuffer_enable(
+					top_pipe_to_program->stream_res.tg);
+
+	if ((update_type != UPDATE_TYPE_FAST) && dc->hwss.interdependent_update_lock)
+		dc->hwss.interdependent_update_lock(dc, context, true);
+	else
+		/* Lock the top pipe while updating plane addrs, since freesync requires
+		 *  plane addr update event triggers to be synchronized.
+		 *  top_pipe_to_program is expected to never be NULL
+		 */
+		dc->hwss.pipe_control_lock(dc, top_pipe_to_program, true);
+
+
 	// Stream updates
 	if (stream_update)
 		commit_planes_do_stream_update(dc, stream, stream_update, update_type, context);
@@ -2189,6 +2247,12 @@ static void commit_planes_for_stream(struct dc *dc,
 		if (dc->hwss.program_front_end_for_ctx)
 			dc->hwss.program_front_end_for_ctx(dc, context);
 
+		if ((update_type != UPDATE_TYPE_FAST) && dc->hwss.interdependent_update_lock)
+			dc->hwss.interdependent_update_lock(dc, context, false);
+		else
+			dc->hwss.pipe_control_lock(dc, top_pipe_to_program, false);
+
+		dc->hwss.post_unlock_program_front_end(dc, context);
 		return;
 	}
 
@@ -2224,8 +2288,6 @@ static void commit_planes_for_stream(struct dc *dc,
 			pipe_ctx->stream == stream) {
 			struct dc_stream_status *stream_status = NULL;
 
-			top_pipe_to_program = pipe_ctx;
-
 			if (!pipe_ctx->plane_state)
 				continue;
 
@@ -2270,12 +2332,6 @@ static void commit_planes_for_stream(struct dc *dc,
 
 	// Update Type FAST, Surface updates
 	if (update_type == UPDATE_TYPE_FAST) {
-		/* Lock the top pipe while updating plane addrs, since freesync requires
-		 *  plane addr update event triggers to be synchronized.
-		 *  top_pipe_to_program is expected to never be NULL
-		 */
-		dc->hwss.pipe_control_lock(dc, top_pipe_to_program, true);
-
 		if (dc->hwss.set_flip_control_gsl)
 			for (i = 0; i < surface_count; i++) {
 				struct dc_plane_state *plane_state = srf_updates[i].surface;
@@ -2317,10 +2373,31 @@ static void commit_planes_for_stream(struct dc *dc,
 					dc->hwss.update_plane_addr(dc, pipe_ctx);
 			}
 		}
-
-		dc->hwss.pipe_control_lock(dc, top_pipe_to_program, false);
 	}
 
+	if ((update_type != UPDATE_TYPE_FAST) && dc->hwss.interdependent_update_lock)
+		dc->hwss.interdependent_update_lock(dc, context, false);
+	else
+		dc->hwss.pipe_control_lock(dc, top_pipe_to_program, false);
+
+	if ((update_type != UPDATE_TYPE_FAST) && stream->update_flags.bits.dsc_changed)
+		if (top_pipe_to_program->stream_res.tg->funcs->lock_doublebuffer_enable) {
+			top_pipe_to_program->stream_res.tg->funcs->wait_for_state(
+					top_pipe_to_program->stream_res.tg,
+					CRTC_STATE_VACTIVE);
+			top_pipe_to_program->stream_res.tg->funcs->wait_for_state(
+					top_pipe_to_program->stream_res.tg,
+					CRTC_STATE_VBLANK);
+			top_pipe_to_program->stream_res.tg->funcs->wait_for_state(
+					top_pipe_to_program->stream_res.tg,
+					CRTC_STATE_VACTIVE);
+			top_pipe_to_program->stream_res.tg->funcs->lock_doublebuffer_disable(
+					top_pipe_to_program->stream_res.tg);
+		}
+
+	if (update_type != UPDATE_TYPE_FAST)
+		dc->hwss.post_unlock_program_front_end(dc, context);
+
 	// Fire manual trigger only when bottom plane is flipped
 	for (j = 0; j < dc->res_pool->pipe_count; j++) {
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j];

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

@@ -45,7 +45,7 @@
 #include "dpcd_defs.h"
 #include "dmcu.h"
 #include "hw/clk_mgr.h"
-#include "../dce/dmub_psr.h"
+#include "dce/dmub_psr.h"
 
 #define DC_LOGGER_INIT(logger)
 
@@ -599,6 +599,9 @@ static bool detect_dp(
 
 	if (sink_caps->transaction_type == DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
 		sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT;
+
+		dpcd_set_source_specific_data(link);
+
 		if (!detect_dp_sink_caps(link))
 			return false;
 
@@ -769,8 +772,16 @@ static bool dc_link_detect_helper(struct dc_link *link,
 
 	if ((link->connector_signal == SIGNAL_TYPE_LVDS ||
 			link->connector_signal == SIGNAL_TYPE_EDP) &&
-			link->local_sink)
+			link->local_sink) {
+
+		// need to re-write OUI and brightness in resume case
+		if (link->connector_signal == SIGNAL_TYPE_EDP) {
+			dpcd_set_source_specific_data(link);
+			dc_link_set_default_brightness_aux(link); //TODO: use cached
+		}
+
 		return true;
+	}
 
 	if (false == dc_link_detect_sink(link, &new_connection_type)) {
 		BREAK_TO_DEBUGGER();
@@ -818,6 +829,10 @@ static bool dc_link_detect_helper(struct dc_link *link,
 		}
 
 		case SIGNAL_TYPE_EDP: {
+			read_current_link_settings_on_detect(link);
+
+			dpcd_set_source_specific_data(link);
+
 			detect_edp_sink_caps(link);
 			read_current_link_settings_on_detect(link);
 			sink_caps.transaction_type = DDC_TRANSACTION_TYPE_I2C_OVER_AUX;
@@ -962,6 +977,9 @@ static bool dc_link_detect_helper(struct dc_link *link,
 		if ((prev_sink != NULL) && ((edid_status == EDID_THE_SAME) || (edid_status == EDID_OK)))
 			same_edid = is_same_edid(&prev_sink->dc_edid, &sink->dc_edid);
 
+		if (sink->edid_caps.panel_patch.skip_scdc_overwrite)
+			link->ctx->dc->debug.hdmi20_disable = true;
+
 		if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT &&
 			sink_caps.transaction_type == DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
 			/*
@@ -1492,6 +1510,7 @@ static enum dc_status enable_link_dp(
 	bool fec_enable;
 	int i;
 	bool apply_seamless_boot_optimization = false;
+	uint32_t bl_oled_enable_delay = 50; // in ms
 
 	// check for seamless boot
 	for (i = 0; i < state->stream_count; i++) {
@@ -1515,6 +1534,9 @@ static enum dc_status enable_link_dp(
 	if (state->clk_mgr && !apply_seamless_boot_optimization)
 		state->clk_mgr->funcs->update_clocks(state->clk_mgr, state, false);
 
+	// during mode switch we do DP_SET_POWER off then on, and OUI is lost
+	dpcd_set_source_specific_data(link);
+
 	skip_video_pattern = true;
 
 	if (link_settings.link_rate == LINK_RATE_LOW)
@@ -1538,6 +1560,17 @@ static enum dc_status enable_link_dp(
 		fec_enable = true;
 
 	dp_set_fec_enable(link, fec_enable);
+
+	// during mode set we do DP_SET_POWER off then on, aux writes are lost
+	if (link->dpcd_sink_ext_caps.bits.oled == 1 ||
+		link->dpcd_sink_ext_caps.bits.sdr_aux_backlight_control == 1 ||
+		link->dpcd_sink_ext_caps.bits.hdr_aux_backlight_control == 1) {
+		dc_link_set_default_brightness_aux(link); // TODO: use cached if known
+		if (link->dpcd_sink_ext_caps.bits.oled == 1)
+			msleep(bl_oled_enable_delay);
+		dc_link_backlight_enable_aux(link, true);
+	}
+
 	return status;
 }
 
@@ -2400,8 +2433,8 @@ bool dc_link_set_psr_allow_active(struct dc_link *link, bool allow_active, bool
 	struct dmcu *dmcu = dc->res_pool->dmcu;
 	struct dmub_psr *psr = dc->res_pool->psr;
 
-	if ((psr != NULL) && link->psr_feature_enabled)
+	if (psr != NULL && link->psr_feature_enabled)
-		psr->funcs->set_psr_enable(psr, allow_active);
+		psr->funcs->psr_enable(psr, allow_active);
 	else if ((dmcu != NULL && dmcu->funcs->is_dmcu_initialized(dmcu)) && link->psr_feature_enabled)
 		dmcu->funcs->set_psr_enable(dmcu, allow_active, wait);
 
@@ -2417,7 +2450,7 @@ bool dc_link_get_psr_state(const struct dc_link *link, uint32_t *psr_state)
 	struct dmub_psr *psr = dc->res_pool->psr;
 
 	if (psr != NULL && link->psr_feature_enabled)
-		psr->funcs->get_psr_state(psr_state);
+		psr->funcs->psr_get_state(psr, psr_state);
 	else if (dmcu != NULL && link->psr_feature_enabled)
 		dmcu->funcs->get_psr_state(dmcu, psr_state);
 
@@ -2589,7 +2622,7 @@ bool dc_link_setup_psr(struct dc_link *link,
 	psr_context->frame_delay = 0;
 
 	if (psr)
-		link->psr_feature_enabled = psr->funcs->setup_psr(psr, link, psr_context);
+		link->psr_feature_enabled = psr->funcs->psr_copy_settings(psr, link, psr_context);
 	else
 		link->psr_feature_enabled = dmcu->funcs->setup_psr(dmcu, link, psr_context);
 
@@ -2922,7 +2955,8 @@ static void update_psp_stream_config(struct pipe_ctx *pipe_ctx, bool dpms_off)
 		memset(&config, 0, sizeof(config));
 
 		config.otg_inst = (uint8_t) pipe_ctx->stream_res.tg->inst;
-		config.stream_enc_inst = (uint8_t) pipe_ctx->stream_res.stream_enc->id;
+		/*stream_enc_inst*/
+		config.stream_enc_inst = (uint8_t) pipe_ctx->stream_res.stream_enc->stream_enc_inst;
 		config.link_enc_inst = pipe_ctx->stream->link->link_enc_hw_inst;
 		config.dpms_off = dpms_off;
 		config.dm_stream_ctx = pipe_ctx->stream->dm_stream_context;
@@ -3061,6 +3095,9 @@ void core_link_enable_stream(
 		dc->hwss.unblank_stream(pipe_ctx,
 			&pipe_ctx->stream->link->cur_link_settings);
 
+		if (stream->sink_patches.delay_ignore_msa > 0)
+			msleep(stream->sink_patches.delay_ignore_msa);
+
 		if (dc_is_dp_signal(pipe_ctx->stream->signal))
 			enable_stream_features(pipe_ctx);
 #if defined(CONFIG_DRM_AMD_DC_HDCP)

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

@@ -126,22 +126,16 @@ struct aux_payloads {
 	struct vector payloads;
 };
 
-static struct i2c_payloads *dal_ddc_i2c_payloads_create(struct dc_context *ctx, uint32_t count)
+static bool dal_ddc_i2c_payloads_create(
+		struct dc_context *ctx,
+		struct i2c_payloads *payloads,
+		uint32_t count)
 {
-	struct i2c_payloads *payloads;
-
-	payloads = kzalloc(sizeof(struct i2c_payloads), GFP_KERNEL);
-
-	if (!payloads)
-		return NULL;
-
 	if (dal_vector_construct(
 		&payloads->payloads, ctx, count, sizeof(struct i2c_payload)))
-		return payloads;
+		return true;
-
-	kfree(payloads);
-	return NULL;
 
+	return false;
 }
 
 static struct i2c_payload *dal_ddc_i2c_payloads_get(struct i2c_payloads *p)
@@ -154,14 +148,12 @@ static uint32_t dal_ddc_i2c_payloads_get_count(struct i2c_payloads *p)
 	return p->payloads.count;
 }
 
-static void dal_ddc_i2c_payloads_destroy(struct i2c_payloads **p)
+static void dal_ddc_i2c_payloads_destroy(struct i2c_payloads *p)
 {
-	if (!p || !*p)
+	if (!p)
 		return;
-	dal_vector_destruct(&(*p)->payloads);
-	kfree(*p);
-	*p = NULL;
 
+	dal_vector_destruct(&p->payloads);
 }
 
 #define DDC_MIN(a, b) (((a) < (b)) ? (a) : (b))
@@ -524,9 +516,13 @@ bool dal_ddc_service_query_ddc_data(
 
 	uint32_t payloads_num = write_payloads + read_payloads;
 
+
 	if (write_size > EDID_SEGMENT_SIZE || read_size > EDID_SEGMENT_SIZE)
 		return false;
 
+	if (!payloads_num)
+		return false;
+
 	/*TODO: len of payload data for i2c and aux is uint8!!!!,
 	 *  but we want to read 256 over i2c!!!!*/
 	if (dal_ddc_service_is_in_aux_transaction_mode(ddc)) {
@@ -557,23 +553,25 @@ bool dal_ddc_service_query_ddc_data(
 			ret = dal_ddc_submit_aux_command(ddc, &payload);
 		}
 	} else {
-		struct i2c_payloads *payloads =
+		struct i2c_command command = {0};
-			dal_ddc_i2c_payloads_create(ddc->ctx, payloads_num);
+		struct i2c_payloads payloads;
 
-		struct i2c_command command = {
+		if (!dal_ddc_i2c_payloads_create(ddc->ctx, &payloads, payloads_num))
-			.payloads = dal_ddc_i2c_payloads_get(payloads),
+			return false;
-			.number_of_payloads = 0,
+
-			.engine = DDC_I2C_COMMAND_ENGINE,
+		command.payloads = dal_ddc_i2c_payloads_get(&payloads);
-			.speed = ddc->ctx->dc->caps.i2c_speed_in_khz };
+		command.number_of_payloads = 0;
+		command.engine = DDC_I2C_COMMAND_ENGINE;
+		command.speed = ddc->ctx->dc->caps.i2c_speed_in_khz;
 
 		dal_ddc_i2c_payloads_add(
-			payloads, address, write_size, write_buf, true);
+			&payloads, address, write_size, write_buf, true);
 
 		dal_ddc_i2c_payloads_add(
-			payloads, address, read_size, read_buf, false);
+			&payloads, address, read_size, read_buf, false);
 
 		command.number_of_payloads =
-			dal_ddc_i2c_payloads_get_count(payloads);
+			dal_ddc_i2c_payloads_get_count(&payloads);
 
 		ret = dm_helpers_submit_i2c(
 				ddc->ctx,
@@ -686,6 +684,10 @@ void dal_ddc_service_write_scdc_data(struct ddc_service *ddc_service,
 	uint8_t write_buffer[2] = {0};
 	/*Lower than 340 Scramble bit from SCDC caps*/
 
+	if (ddc_service->link->local_sink &&
+		ddc_service->link->local_sink->edid_caps.panel_patch.skip_scdc_overwrite)
+		return;
+
 	dal_ddc_service_query_ddc_data(ddc_service, slave_address, &offset,
 			sizeof(offset), &sink_version, sizeof(sink_version));
 	if (sink_version == 1) {
@@ -715,6 +717,10 @@ void dal_ddc_service_read_scdc_data(struct ddc_service *ddc_service)
 	uint8_t offset = HDMI_SCDC_TMDS_CONFIG;
 	uint8_t tmds_config = 0;
 
+	if (ddc_service->link->local_sink &&
+		ddc_service->link->local_sink->edid_caps.panel_patch.skip_scdc_overwrite)
+		return;
+
 	dal_ddc_service_query_ddc_data(ddc_service, slave_address, &offset,
 			sizeof(offset), &tmds_config, sizeof(tmds_config));
 	if (tmds_config & 0x1) {

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

@@ -945,6 +945,17 @@ static enum link_training_result perform_channel_equalization_sequence(
 }
 #define TRAINING_AUX_RD_INTERVAL 100 //us
 
+static void start_clock_recovery_pattern_early(struct dc_link *link,
+		struct link_training_settings *lt_settings,
+		uint32_t offset)
+{
+	DC_LOG_HW_LINK_TRAINING("%s\n GPU sends TPS1. Wait 400us.\n",
+			__func__);
+	dp_set_hw_training_pattern(link, DP_TRAINING_PATTERN_SEQUENCE_1, offset);
+	dp_set_hw_lane_settings(link, lt_settings, offset);
+	udelay(400);
+}
+
 static enum link_training_result perform_clock_recovery_sequence(
 	struct dc_link *link,
 	struct link_training_settings *lt_settings,
@@ -962,6 +973,7 @@ static enum link_training_result perform_clock_recovery_sequence(
 	retries_cr = 0;
 	retry_count = 0;
 
+	if (!link->ctx->dc->work_arounds.lt_early_cr_pattern)
 		dp_set_hw_training_pattern(link, tr_pattern, offset);
 
 	/* najeeb - The synaptics MST hub can put the LT in
@@ -1435,6 +1447,9 @@ enum link_training_result dc_link_dp_perform_link_training(
 			&lt_settings);
 
 	/* 1. set link rate, lane count and spread. */
+	if (link->ctx->dc->work_arounds.lt_early_cr_pattern)
+		start_clock_recovery_pattern_early(link, &lt_settings, DPRX);
+	else
 		dpcd_set_link_settings(link, &lt_settings);
 
 	if (link->preferred_training_settings.fec_enable != NULL)
@@ -1654,8 +1669,10 @@ enum link_training_result dc_link_dp_sync_lt_attempt(
 	dp_set_panel_mode(link, panel_mode);
 
 	/* Attempt to train with given link training settings */
-
 	/* Set link rate, lane count and spread. */
+	if (link->ctx->dc->work_arounds.lt_early_cr_pattern)
+		start_clock_recovery_pattern_early(link, &lt_settings, DPRX);
+	else
 		dpcd_set_link_settings(link, &lt_settings);
 
 	/* 2. perform link training (set link training done
@@ -1892,6 +1909,16 @@ bool dp_verify_link_cap(
 	/* disable PHY done possible by BIOS, will be done by driver itself */
 	dp_disable_link_phy(link, link->connector_signal);
 
+	dp_cs_id = get_clock_source_id(link);
+
+	/* link training starts with the maximum common settings
+	 * supported by both sink and ASIC.
+	 */
+	initial_link_settings = get_common_supported_link_settings(
+			*known_limit_link_setting,
+			max_link_cap);
+	cur_link_setting = initial_link_settings;
+
 	/* Temporary Renoir-specific workaround for SWDEV-215184;
 	 * PHY will sometimes be in bad state on hotplugging display from certain USB-C dongle,
 	 * so add extra cycle of enabling and disabling the PHY before first link training.
@@ -1902,15 +1929,6 @@ bool dp_verify_link_cap(
 		dp_disable_link_phy(link, link->connector_signal);
 	}
 
-	dp_cs_id = get_clock_source_id(link);
-
-	/* link training starts with the maximum common settings
-	 * supported by both sink and ASIC.
-	 */
-	initial_link_settings = get_common_supported_link_settings(
-			*known_limit_link_setting,
-			max_link_cap);
-	cur_link_setting = initial_link_settings;
 	do {
 		skip_video_pattern = true;
 
@@ -3165,6 +3183,23 @@ static void dp_wa_power_up_0010FA(struct dc_link *link, uint8_t *dpcd_data,
 		link->wa_flags.dp_keep_receiver_powered = false;
 }
 
+/* Read additional sink caps defined in source specific DPCD area
+ * This function currently only reads from SinkCapability address (DP_SOURCE_SINK_CAP)
+ */
+static bool dpcd_read_sink_ext_caps(struct dc_link *link)
+{
+	uint8_t dpcd_data;
+
+	if (!link)
+		return false;
+
+	if (core_link_read_dpcd(link, DP_SOURCE_SINK_CAP, &dpcd_data, 1) != DC_OK)
+		return false;
+
+	link->dpcd_sink_ext_caps.raw = dpcd_data;
+	return true;
+}
+
 static bool retrieve_link_cap(struct dc_link *link)
 {
 	/* DP_ADAPTER_CAP - DP_DPCD_REV + 1 == 16 and also DP_DSC_BITS_PER_PIXEL_INC - DP_DSC_SUPPORT + 1 == 16,
@@ -3437,6 +3472,9 @@ static bool retrieve_link_cap(struct dc_link *link)
 				sizeof(link->dpcd_caps.dsc_caps.dsc_ext_caps.raw));
 	}
 
+	if (!dpcd_read_sink_ext_caps(link))
+		link->dpcd_sink_ext_caps.raw = 0;
+
 	/* Connectivity log: detection */
 	CONN_DATA_DETECT(link, dpcd_data, sizeof(dpcd_data), "Rx Caps: ");
 
@@ -3589,6 +3627,8 @@ void detect_edp_sink_caps(struct dc_link *link)
 		}
 	}
 	link->verified_link_cap = link->reported_link_cap;
+
+	dc_link_set_default_brightness_aux(link);
 }
 
 void dc_link_dp_enable_hpd(const struct dc_link *link)
@@ -3680,7 +3720,7 @@ static void set_crtc_test_pattern(struct dc_link *link,
 			struct pipe_ctx *odm_pipe;
 			enum controller_dp_color_space controller_color_space;
 			int opp_cnt = 1;
-			int count;
+			uint16_t count = 0;
 
 			switch (test_pattern_color_space) {
 			case DP_TEST_PATTERN_COLOR_SPACE_RGB:
@@ -4146,3 +4186,148 @@ void dp_set_fec_enable(struct dc_link *link, bool enable)
 	}
 }
 
+void dpcd_set_source_specific_data(struct dc_link *link)
+{
+	const uint32_t post_oui_delay = 30; // 30ms
+
+	if (!link->dc->vendor_signature.is_valid) {
+		struct dpcd_amd_signature amd_signature;
+		amd_signature.AMD_IEEE_TxSignature_byte1 = 0x0;
+		amd_signature.AMD_IEEE_TxSignature_byte2 = 0x0;
+		amd_signature.AMD_IEEE_TxSignature_byte3 = 0x1A;
+		amd_signature.device_id_byte1 =
+				(uint8_t)(link->ctx->asic_id.chip_id);
+		amd_signature.device_id_byte2 =
+				(uint8_t)(link->ctx->asic_id.chip_id >> 8);
+		memset(&amd_signature.zero, 0, 4);
+		amd_signature.dce_version =
+				(uint8_t)(link->ctx->dce_version);
+		amd_signature.dal_version_byte1 = 0x0; // needed? where to get?
+		amd_signature.dal_version_byte2 = 0x0; // needed? where to get?
+
+		core_link_write_dpcd(link, DP_SOURCE_OUI,
+				(uint8_t *)(&amd_signature),
+				sizeof(amd_signature));
+
+	} else {
+		core_link_write_dpcd(link, DP_SOURCE_OUI,
+				link->dc->vendor_signature.data.raw,
+				sizeof(link->dc->vendor_signature.data.raw));
+	}
+
+	// Sink may need to configure internals based on vendor, so allow some
+	// time before proceeding with possibly vendor specific transactions
+	msleep(post_oui_delay);
+}
+
+bool dc_link_set_backlight_level_nits(struct dc_link *link,
+		bool isHDR,
+		uint32_t backlight_millinits,
+		uint32_t transition_time_in_ms)
+{
+	struct dpcd_source_backlight_set dpcd_backlight_set;
+	uint8_t backlight_control = isHDR ? 1 : 0;
+
+	if (!link || (link->connector_signal != SIGNAL_TYPE_EDP &&
+			link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
+		return false;
+
+	// OLEDs have no PWM, they can only use AUX
+	if (link->dpcd_sink_ext_caps.bits.oled == 1)
+		backlight_control = 1;
+
+	*(uint32_t *)&dpcd_backlight_set.backlight_level_millinits = backlight_millinits;
+	*(uint16_t *)&dpcd_backlight_set.backlight_transition_time_ms = (uint16_t)transition_time_in_ms;
+
+
+	if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_LEVEL,
+			(uint8_t *)(&dpcd_backlight_set),
+			sizeof(dpcd_backlight_set)) != DC_OK)
+		return false;
+
+	if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_CONTROL,
+			&backlight_control, 1) != DC_OK)
+		return false;
+
+	return true;
+}
+
+bool dc_link_get_backlight_level_nits(struct dc_link *link,
+		uint32_t *backlight_millinits_avg,
+		uint32_t *backlight_millinits_peak)
+{
+	union dpcd_source_backlight_get dpcd_backlight_get;
+
+	memset(&dpcd_backlight_get, 0, sizeof(union dpcd_source_backlight_get));
+
+	if (!link || (link->connector_signal != SIGNAL_TYPE_EDP &&
+			link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
+		return false;
+
+	if (!core_link_read_dpcd(link, DP_SOURCE_BACKLIGHT_CURRENT_PEAK,
+			dpcd_backlight_get.raw,
+			sizeof(union dpcd_source_backlight_get)))
+		return false;
+
+	*backlight_millinits_avg =
+		dpcd_backlight_get.bytes.backlight_millinits_avg;
+	*backlight_millinits_peak =
+		dpcd_backlight_get.bytes.backlight_millinits_peak;
+
+	/* On non-supported panels dpcd_read usually succeeds with 0 returned */
+	if (*backlight_millinits_avg == 0 ||
+			*backlight_millinits_avg > *backlight_millinits_peak)
+		return false;
+
+	return true;
+}
+
+bool dc_link_backlight_enable_aux(struct dc_link *link, bool enable)
+{
+	uint8_t backlight_enable = enable ? 1 : 0;
+
+	if (!link || (link->connector_signal != SIGNAL_TYPE_EDP &&
+		link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
+		return false;
+
+	if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_ENABLE,
+		&backlight_enable, 1) != DC_OK)
+		return false;
+
+	return true;
+}
+
+// we read default from 0x320 because we expect BIOS wrote it there
+// regular get_backlight_nit reads from panel set at 0x326
+bool dc_link_read_default_bl_aux(struct dc_link *link, uint32_t *backlight_millinits)
+{
+	if (!link || (link->connector_signal != SIGNAL_TYPE_EDP &&
+		link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
+		return false;
+
+	if (!core_link_read_dpcd(link, DP_SOURCE_BACKLIGHT_LEVEL,
+		(uint8_t *) backlight_millinits,
+		sizeof(uint32_t)))
+		return false;
+
+	return true;
+}
+
+bool dc_link_set_default_brightness_aux(struct dc_link *link)
+{
+	uint32_t default_backlight;
+
+	if (link &&
+		(link->dpcd_sink_ext_caps.bits.hdr_aux_backlight_control == 1 ||
+		link->dpcd_sink_ext_caps.bits.sdr_aux_backlight_control == 1)) {
+		if (!dc_link_read_default_bl_aux(link, &default_backlight))
+			default_backlight = 150000;
+		// if < 5 nits or > 5000, it might be wrong readback
+		if (default_backlight < 5000 || default_backlight > 5000000)
+			default_backlight = 150000; //
+
+		return dc_link_set_backlight_level_nits(link, true,
+				default_backlight, 0);
+	}
+	return false;
+}

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

@@ -153,9 +153,9 @@ bool edp_receiver_ready_T9(struct dc_link *link)
 	unsigned char edpRev = 0;
 	enum dc_status result = DC_OK;
 	result = core_link_read_dpcd(link, DP_EDP_DPCD_REV, &edpRev, sizeof(edpRev));
-	if (edpRev < DP_EDP_12)
+
-		return true;
      /* start from eDP version 1.2, SINK_STAUS indicate the sink is ready.*/
+	if (result == DC_OK && edpRev >= DP_EDP_12) {
 		do {
 			sinkstatus = 1;
 			result = core_link_read_dpcd(link, DP_SINK_STATUS, &sinkstatus, sizeof(sinkstatus));
@@ -165,6 +165,7 @@ bool edp_receiver_ready_T9(struct dc_link *link)
 				break;
 			udelay(100); //MAx T9
 		} while (++tries < 50);
+	}
 
 	if (link->local_sink->edid_caps.panel_patch.extra_delay_backlight_off > 0)
 		udelay(link->local_sink->edid_caps.panel_patch.extra_delay_backlight_off * 1000);
@@ -183,8 +184,8 @@ bool edp_receiver_ready_T7(struct dc_link *link)
 	unsigned long long time_taken_in_ns = 0;
 
 	result = core_link_read_dpcd(link, DP_EDP_DPCD_REV, &edpRev, sizeof(edpRev));
-	if (result == DC_OK && edpRev < DP_EDP_12)
+
-		return true;
+	if (result == DC_OK && edpRev >= DP_EDP_12) {
 		/* start from eDP version 1.2, SINK_STAUS indicate the sink is ready.*/
 		enter_timestamp = dm_get_timestamp(link->ctx);
 		do {
@@ -198,6 +199,7 @@ bool edp_receiver_ready_T7(struct dc_link *link)
 			finish_timestamp = dm_get_timestamp(link->ctx);
 			time_taken_in_ns = dm_get_elapse_time_in_ns(link->ctx, finish_timestamp, enter_timestamp);
 		} while (time_taken_in_ns < 50 * 1000000); //MAx T7 is 50ms
+	}
 
 	if (link->local_sink->edid_caps.panel_patch.extra_t7_ms > 0)
 		udelay(link->local_sink->edid_caps.panel_patch.extra_t7_ms * 1000);

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

@@ -46,12 +46,12 @@
 #include "dce100/dce100_resource.h"
 #include "dce110/dce110_resource.h"
 #include "dce112/dce112_resource.h"
+#include "dce120/dce120_resource.h"
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 #include "dcn10/dcn10_resource.h"
-#endif
 #include "dcn20/dcn20_resource.h"
 #include "dcn21/dcn21_resource.h"
-#include "dce120/dce120_resource.h"
+#endif
 
 #define DC_LOGGER_INIT(logger)
 
@@ -532,6 +532,51 @@ static inline void get_vp_scan_direction(
 		*flip_horz_scan_dir = !*flip_horz_scan_dir;
 }
 
+int get_num_odm_splits(struct pipe_ctx *pipe)
+{
+	int odm_split_count = 0;
+	struct pipe_ctx *next_pipe = pipe->next_odm_pipe;
+	while (next_pipe) {
+		odm_split_count++;
+		next_pipe = next_pipe->next_odm_pipe;
+	}
+	pipe = pipe->prev_odm_pipe;
+	while (pipe) {
+		odm_split_count++;
+		pipe = pipe->prev_odm_pipe;
+	}
+	return odm_split_count;
+}
+
+static void calculate_split_count_and_index(struct pipe_ctx *pipe_ctx, int *split_count, int *split_idx)
+{
+	*split_count = get_num_odm_splits(pipe_ctx);
+	*split_idx = 0;
+	if (*split_count == 0) {
+		/*Check for mpc split*/
+		struct pipe_ctx *split_pipe = pipe_ctx->top_pipe;
+
+		while (split_pipe && split_pipe->plane_state == pipe_ctx->plane_state) {
+			(*split_idx)++;
+			(*split_count)++;
+			split_pipe = split_pipe->top_pipe;
+		}
+		split_pipe = pipe_ctx->bottom_pipe;
+		while (split_pipe && split_pipe->plane_state == pipe_ctx->plane_state) {
+			(*split_count)++;
+			split_pipe = split_pipe->bottom_pipe;
+		}
+	} else {
+		/*Get odm split index*/
+		struct pipe_ctx *split_pipe = pipe_ctx->prev_odm_pipe;
+
+		while (split_pipe) {
+			(*split_idx)++;
+			split_pipe = split_pipe->prev_odm_pipe;
+		}
+	}
+}
+
 static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
@@ -541,16 +586,16 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	struct rect clip, dest;
 	int vpc_div = (data->format == PIXEL_FORMAT_420BPP8
 			|| data->format == PIXEL_FORMAT_420BPP10) ? 2 : 1;
-	bool pri_split = pipe_ctx->bottom_pipe &&
+	int split_count = 0;
-			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state;
+	int split_idx = 0;
-	bool sec_split = pipe_ctx->top_pipe &&
-			pipe_ctx->top_pipe->plane_state == pipe_ctx->plane_state;
 	bool orthogonal_rotation, flip_y_start, flip_x_start;
 
+	calculate_split_count_and_index(pipe_ctx, &split_count, &split_idx);
+
 	if (stream->view_format == VIEW_3D_FORMAT_SIDE_BY_SIDE ||
 		stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM) {
-		pri_split = false;
+		split_count = 0;
-		sec_split = false;
+		split_idx = 0;
 	}
 
 	/* The actual clip is an intersection between stream
@@ -609,23 +654,32 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	data->viewport.height = clip.height * surf_src.height / dest.height;
 
 	/* Handle split */
-	if (pri_split || sec_split) {
+	if (split_count) {
+		/* extra pixels in the division remainder need to go to pipes after
+		 * the extra pixel index minus one(epimo) defined here as:
+		 */
+		int epimo = 0;
+
 		if (orthogonal_rotation) {
-			if (flip_y_start != pri_split)
+			if (flip_y_start)
-				data->viewport.height /= 2;
+				split_idx = split_count - split_idx;
-			else {
+
-				data->viewport.y +=  data->viewport.height / 2;
+			epimo = split_count - data->viewport.height % (split_count + 1);
-				/* Ceil offset pipe */
+
-				data->viewport.height = (data->viewport.height + 1) / 2;
+			data->viewport.y += (data->viewport.height / (split_count + 1)) * split_idx;
-			}
+			if (split_idx > epimo)
+				data->viewport.y += split_idx - epimo - 1;
+			data->viewport.height = data->viewport.height / (split_count + 1) + (split_idx > epimo ? 1 : 0);
 		} else {
-			if (flip_x_start != pri_split)
+			if (flip_x_start)
-				data->viewport.width /= 2;
+				split_idx = split_count - split_idx;
-			else {
+
-				data->viewport.x +=  data->viewport.width / 2;
+			epimo = split_count - data->viewport.width % (split_count + 1);
-				/* Ceil offset pipe */
+
-				data->viewport.width = (data->viewport.width + 1) / 2;
+			data->viewport.x += (data->viewport.width / (split_count + 1)) * split_idx;
-			}
+			if (split_idx > epimo)
+				data->viewport.x += split_idx - epimo - 1;
+			data->viewport.width = data->viewport.width / (split_count + 1) + (split_idx > epimo ? 1 : 0);
 		}
 	}
 
@@ -644,58 +698,58 @@ static void calculate_recout(struct pipe_ctx *pipe_ctx)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	const struct dc_stream_state *stream = pipe_ctx->stream;
+	struct scaler_data *data = &pipe_ctx->plane_res.scl_data;
 	struct rect surf_clip = plane_state->clip_rect;
-	bool pri_split = pipe_ctx->bottom_pipe &&
+	bool pri_split_tb = pipe_ctx->bottom_pipe &&
-			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state;
+			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state &&
-	bool sec_split = pipe_ctx->top_pipe &&
+			stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM;
-			pipe_ctx->top_pipe->plane_state == pipe_ctx->plane_state;
+	bool sec_split_tb = pipe_ctx->top_pipe &&
-	bool top_bottom_split = stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM;
+			pipe_ctx->top_pipe->plane_state == pipe_ctx->plane_state &&
+			stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM;
+	int split_count = 0;
+	int split_idx = 0;
 
-	pipe_ctx->plane_res.scl_data.recout.x = stream->dst.x;
+	calculate_split_count_and_index(pipe_ctx, &split_count, &split_idx);
+
+	data->recout.x = stream->dst.x;
 	if (stream->src.x < surf_clip.x)
-		pipe_ctx->plane_res.scl_data.recout.x += (surf_clip.x
+		data->recout.x += (surf_clip.x - stream->src.x) * stream->dst.width
-			- stream->src.x) * stream->dst.width
 						/ stream->src.width;
 
-	pipe_ctx->plane_res.scl_data.recout.width = surf_clip.width *
+	data->recout.width = surf_clip.width * stream->dst.width / stream->src.width;
-			stream->dst.width / stream->src.width;
+	if (data->recout.width + data->recout.x > stream->dst.x + stream->dst.width)
-	if (pipe_ctx->plane_res.scl_data.recout.width + pipe_ctx->plane_res.scl_data.recout.x >
+		data->recout.width = stream->dst.x + stream->dst.width - data->recout.x;
-			stream->dst.x + stream->dst.width)
-		pipe_ctx->plane_res.scl_data.recout.width =
-			stream->dst.x + stream->dst.width
-						- pipe_ctx->plane_res.scl_data.recout.x;
 
-	pipe_ctx->plane_res.scl_data.recout.y = stream->dst.y;
+	data->recout.y = stream->dst.y;
 	if (stream->src.y < surf_clip.y)
-		pipe_ctx->plane_res.scl_data.recout.y += (surf_clip.y
+		data->recout.y += (surf_clip.y - stream->src.y) * stream->dst.height
-			- stream->src.y) * stream->dst.height
 						/ stream->src.height;
 
-	pipe_ctx->plane_res.scl_data.recout.height = surf_clip.height *
+	data->recout.height = surf_clip.height * stream->dst.height / stream->src.height;
-			stream->dst.height / stream->src.height;
+	if (data->recout.height + data->recout.y > stream->dst.y + stream->dst.height)
-	if (pipe_ctx->plane_res.scl_data.recout.height + pipe_ctx->plane_res.scl_data.recout.y >
+		data->recout.height = stream->dst.y + stream->dst.height - data->recout.y;
-			stream->dst.y + stream->dst.height)
-		pipe_ctx->plane_res.scl_data.recout.height =
-			stream->dst.y + stream->dst.height
-						- pipe_ctx->plane_res.scl_data.recout.y;
 
 	/* Handle h & v split, handle rotation using viewport */
-	if (sec_split && top_bottom_split) {
+	if (sec_split_tb) {
-		pipe_ctx->plane_res.scl_data.recout.y +=
+		data->recout.y += data->recout.height / 2;
-				pipe_ctx->plane_res.scl_data.recout.height / 2;
 		/* Floor primary pipe, ceil 2ndary pipe */
-		pipe_ctx->plane_res.scl_data.recout.height =
+		data->recout.height = (data->recout.height + 1) / 2;
-				(pipe_ctx->plane_res.scl_data.recout.height + 1) / 2;
+	} else if (pri_split_tb)
-	} else if (pri_split && top_bottom_split)
+		data->recout.height /= 2;
-		pipe_ctx->plane_res.scl_data.recout.height /= 2;
+	else if (split_count) {
-	else if (sec_split) {
+		/* extra pixels in the division remainder need to go to pipes after
-		pipe_ctx->plane_res.scl_data.recout.x +=
+		 * the extra pixel index minus one(epimo) defined here as:
-				pipe_ctx->plane_res.scl_data.recout.width / 2;
+		 */
-		/* Ceil offset pipe */
+		int epimo = split_count - data->recout.width % (split_count + 1);
-		pipe_ctx->plane_res.scl_data.recout.width =
+
-				(pipe_ctx->plane_res.scl_data.recout.width + 1) / 2;
+		/*no recout offset due to odm */
-	} else if (pri_split)
+		if (!pipe_ctx->next_odm_pipe && !pipe_ctx->prev_odm_pipe) {
-		pipe_ctx->plane_res.scl_data.recout.width /= 2;
+			data->recout.x += (data->recout.width / (split_count + 1)) * split_idx;
+			if (split_idx > epimo)
+				data->recout.x += split_idx - epimo - 1;
+		}
+		data->recout.width = data->recout.width / (split_count + 1) + (split_idx > epimo ? 1 : 0);
+	}
 }
 
 static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
@@ -832,6 +886,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	const struct dc_stream_state *stream = pipe_ctx->stream;
+	struct pipe_ctx *odm_pipe = pipe_ctx->prev_odm_pipe;
 	struct scaler_data *data = &pipe_ctx->plane_res.scl_data;
 	struct rect src = pipe_ctx->plane_state->src_rect;
 	int recout_skip_h, recout_skip_v, surf_size_h, surf_size_v;
@@ -869,6 +924,12 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx)
 					* stream->dst.width / stream->src.width -
 					src.x * plane_state->dst_rect.width / src.width
 					* stream->dst.width / stream->src.width);
+	/*modified recout_skip_h calculation due to odm having no recout offset caused by split*/
+	while (odm_pipe) {
+		recout_skip_h += odm_pipe->plane_res.scl_data.recout.width + odm_pipe->plane_res.scl_data.recout.x;
+		odm_pipe = odm_pipe->prev_odm_pipe;
+	}
+
 	recout_skip_v = data->recout.y - (stream->dst.y + (plane_state->dst_rect.y - stream->src.y)
 					* stream->dst.height / stream->src.height -
 					src.y * plane_state->dst_rect.height / src.height
@@ -1021,6 +1082,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 		store_h_border_left + timing->h_border_right;
 	pipe_ctx->plane_res.scl_data.v_active = timing->v_addressable +
 		timing->v_border_top + timing->v_border_bottom;
+	if (pipe_ctx->next_odm_pipe || pipe_ctx->prev_odm_pipe)
+		pipe_ctx->plane_res.scl_data.h_active /= get_num_odm_splits(pipe_ctx) + 1;
 
 	/* Taps calculations */
 	if (pipe_ctx->plane_res.xfm != NULL)
@@ -2034,7 +2097,7 @@ enum dc_status resource_map_pool_resources(
 	for (i = 0; i < context->stream_count; i++)
 		if (context->streams[i] == stream) {
 			context->stream_status[i].primary_otg_inst = pipe_ctx->stream_res.tg->inst;
-			context->stream_status[i].stream_enc_inst = pipe_ctx->stream_res.stream_enc->id;
+			context->stream_status[i].stream_enc_inst = pipe_ctx->stream_res.stream_enc->stream_enc_inst;
 			context->stream_status[i].audio_inst =
 				pipe_ctx->stream_res.audio ? pipe_ctx->stream_res.audio->inst : -1;
 

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

@@ -47,6 +47,9 @@ int dc_setup_system_context(struct dc *dc, struct dc_phy_addr_space_config *pa_c
 		 */
 		memcpy(&dc->vm_pa_config, pa_config, sizeof(struct dc_phy_addr_space_config));
 		dc->vm_pa_config.valid = true;
+
+		if (pa_config->is_hvm_enabled == 0)
+			dc->debug.nv12_iflip_vm_wa = false;
 	}
 
 	return num_vmids;
@@ -62,7 +65,7 @@ int dc_get_vmid_use_vector(struct dc *dc)
 	int i;
 	int in_use = 0;
 
-	for (i = 0; i < dc->vm_helper->num_vmid; i++)
+	for (i = 0; i < MAX_HUBP; i++)
 		in_use |= dc->vm_helper->hubp_vmid_usage[i].vmid_usage[0]
 			| dc->vm_helper->hubp_vmid_usage[i].vmid_usage[1];
 	return in_use;

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

@@ -39,7 +39,7 @@
 #include "inc/hw/dmcu.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.2.69"
+#define DC_VER "3.2.74"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
@@ -126,6 +126,7 @@ struct dc_bug_wa {
 	bool no_connect_phy_config;
 	bool dedcn20_305_wa;
 	bool skip_clock_update;
+	bool lt_early_cr_pattern;
 };
 
 struct dc_dcc_surface_param {
@@ -409,6 +410,7 @@ struct dc_debug_options {
 	bool dmub_offload_enabled;
 	bool dmcub_emulation;
 	bool dmub_command_table; /* for testing only */
+	bool psr_on_dmub;
 	struct dc_bw_validation_profile bw_val_profile;
 	bool disable_fec;
 	bool disable_48mhz_pwrdwn;
@@ -453,6 +455,7 @@ struct dc_phy_addr_space_config {
 	} gart_config;
 
 	bool valid;
+	bool is_hvm_enabled;
 	uint64_t page_table_default_page_addr;
 };
 
@@ -517,7 +520,8 @@ struct dc {
 	struct dce_hwseq *hwseq;
 
 	/* Require to optimize clocks and bandwidth for added/removed planes */
-	bool optimized_required;
+	bool clk_optimized_required;
+	bool wm_optimized_required;
 
 	/* Require to maintain clocks and bandwidth for UEFI enabled HW */
 	int optimize_seamless_boot_streams;
@@ -526,6 +530,7 @@ struct dc {
 	struct compressor *fbc_compressor;
 
 	struct dc_debug_data debug_data;
+	struct dpcd_vendor_signature vendor_signature;
 
 	const char *build_id;
 	struct vm_helper *vm_helper;
@@ -565,12 +570,14 @@ struct dc_init_data {
 	struct dc_reg_helper_state *dmub_offload;
 
 	struct dc_config flags;
-	uint32_t log_mask;
+	uint64_t log_mask;
+
 	/**
 	 * gpu_info FW provided soc bounding box struct or 0 if not
 	 * available in FW
 	 */
 	const struct gpu_info_soc_bounding_box_v1_0 *soc_bounding_box;
+	struct dpcd_vendor_signature vendor_signature;
 };
 
 struct dc_callback_init {
@@ -682,7 +689,6 @@ struct dc_3dlut {
 	struct kref refcount;
 	struct tetrahedral_params lut_3d;
 	struct fixed31_32 hdr_multiplier;
-	bool initialized; /*remove after diag fix*/
 	union dc_3dlut_state state;
 	struct dc_context *ctx;
 };
@@ -979,6 +985,20 @@ struct dpcd_caps {
 
 };
 
+union dpcd_sink_ext_caps {
+	struct {
+		/* 0 - Sink supports backlight adjust via PWM during SDR/HDR mode
+		 * 1 - Sink supports backlight adjust via AUX during SDR/HDR mode.
+		 */
+		uint8_t sdr_aux_backlight_control : 1;
+		uint8_t hdr_aux_backlight_control : 1;
+		uint8_t reserved_1 : 2;
+		uint8_t oled : 1;
+		uint8_t reserved : 3;
+	} bits;
+	uint8_t raw;
+};
+
 #include "dc_link.h"
 
 /*******************************************************************************
@@ -1075,7 +1095,6 @@ unsigned int dc_get_current_backlight_pwm(struct dc *dc);
 unsigned int dc_get_target_backlight_pwm(struct dc *dc);
 
 bool dc_is_dmcu_initialized(struct dc *dc);
-bool dc_is_hw_initialized(struct dc *dc);
 
 enum dc_status dc_set_clock(struct dc *dc, enum dc_clock_type clock_type, uint32_t clk_khz, uint32_t stepping);
 void dc_get_clock(struct dc *dc, enum dc_clock_type clock_type, struct dc_clock_config *clock_cfg);

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

@@ -432,6 +432,54 @@ struct dp_sink_hw_fw_revision {
 	uint8_t ieee_fw_rev[2];
 };
 
+struct dpcd_vendor_signature {
+	bool is_valid;
+
+	union dpcd_ieee_vendor_signature {
+		struct {
+			uint8_t ieee_oui[3];/*24-bit IEEE OUI*/
+			uint8_t ieee_device_id[6];/*usually 6-byte ASCII name*/
+			uint8_t ieee_hw_rev;
+			uint8_t ieee_fw_rev[2];
+		};
+		uint8_t raw[12];
+	} data;
+};
+
+struct dpcd_amd_signature {
+	uint8_t AMD_IEEE_TxSignature_byte1;
+	uint8_t AMD_IEEE_TxSignature_byte2;
+	uint8_t AMD_IEEE_TxSignature_byte3;
+	uint8_t device_id_byte1;
+	uint8_t device_id_byte2;
+	uint8_t zero[4];
+	uint8_t dce_version;
+	uint8_t dal_version_byte1;
+	uint8_t dal_version_byte2;
+};
+
+struct dpcd_source_backlight_set {
+	struct  {
+		uint8_t byte0;
+		uint8_t byte1;
+		uint8_t byte2;
+		uint8_t byte3;
+	} backlight_level_millinits;
+
+	struct  {
+		uint8_t byte0;
+		uint8_t byte1;
+	} backlight_transition_time_ms;
+};
+
+union dpcd_source_backlight_get {
+	struct {
+		uint32_t backlight_millinits_peak; /* 326h */
+		uint32_t backlight_millinits_avg; /* 32Ah */
+	} bytes;
+	uint8_t raw[8];
+};
+
 /*DPCD register of DP receiver capability field bits-*/
 union edp_configuration_cap {
 	struct {

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

@@ -26,6 +26,7 @@
 #ifndef DC_LINK_H_
 #define DC_LINK_H_
 
+#include "dc.h"
 #include "dc_types.h"
 #include "grph_object_defs.h"
 
@@ -128,6 +129,7 @@ struct dc_link {
 	enum edp_revision edp_revision;
 	bool psr_feature_enabled;
 	bool psr_allow_active;
+	union dpcd_sink_ext_caps dpcd_sink_ext_caps;
 
 	/* MST record stream using this link */
 	struct link_flags {
@@ -178,6 +180,21 @@ bool dc_link_set_backlight_level(const struct dc_link *dc_link,
 		uint32_t backlight_pwm_u16_16,
 		uint32_t frame_ramp);
 
+/* Set/get nits-based backlight level of an embedded panel (eDP, LVDS). */
+bool dc_link_set_backlight_level_nits(struct dc_link *link,
+		bool isHDR,
+		uint32_t backlight_millinits,
+		uint32_t transition_time_in_ms);
+
+bool dc_link_get_backlight_level_nits(struct dc_link *link,
+		uint32_t *backlight_millinits,
+		uint32_t *backlight_millinits_peak);
+
+bool dc_link_backlight_enable_aux(struct dc_link *link, bool enable);
+
+bool dc_link_read_default_bl_aux(struct dc_link *link, uint32_t *backlight_millinits);
+bool dc_link_set_default_brightness_aux(struct dc_link *link);
+
 int dc_link_get_backlight_level(const struct dc_link *dc_link);
 
 bool dc_link_set_abm_disable(const struct dc_link *dc_link);

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

@@ -118,6 +118,7 @@ union stream_update_flags {
 		uint32_t dpms_off:1;
 		uint32_t gamut_remap:1;
 		uint32_t wb_update:1;
+		uint32_t dsc_changed : 1;
 	} bits;
 
 	uint32_t raw;

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

@@ -229,7 +229,8 @@ struct dc_panel_patch {
 	unsigned int extra_t12_ms;
 	unsigned int extra_delay_backlight_off;
 	unsigned int extra_t7_ms;
-	unsigned int manage_secondary_link;
+	unsigned int skip_scdc_overwrite;
+	unsigned int delay_ignore_msa;
 };
 
 struct dc_edid_caps {

drivers/gpu/drm/amd/display/dc/dce/Makefile

@@ -29,7 +29,7 @@
 DCE = dce_audio.o dce_stream_encoder.o dce_link_encoder.o dce_hwseq.o \
 dce_mem_input.o dce_clock_source.o dce_scl_filters.o dce_transform.o \
 dce_opp.o dce_dmcu.o dce_abm.o dce_ipp.o dce_aux.o \
-dce_i2c.o dce_i2c_hw.o dce_i2c_sw.o
+dce_i2c.o dce_i2c_hw.o dce_i2c_sw.o dmub_psr.o
 
 AMD_DAL_DCE = $(addprefix $(AMDDALPATH)/dc/dce/,$(DCE))
 

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

@@ -378,6 +378,11 @@ static bool dcn10_dmcu_init(struct dmcu *dmcu)
 	struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(dmcu);
 	const struct dc_config *config = &dmcu->ctx->dc->config;
 	bool status = false;
+	struct dc_context *ctx = dmcu->ctx;
+	unsigned int i;
+	//  5 4 3 2 1 0
+	//  F E D C B A - bit 0 is A, bit 5 is F
+	unsigned int tx_interrupt_mask = 0;
 
 	PERF_TRACE();
 	/*  Definition of DC_DMCU_SCRATCH
@@ -387,6 +392,15 @@ static bool dcn10_dmcu_init(struct dmcu *dmcu)
 	 */
 	dmcu->dmcu_state = REG_READ(DC_DMCU_SCRATCH);
 
+	for (i = 0; i < ctx->dc->link_count; i++) {
+		if (ctx->dc->links[i]->link_enc->features.flags.bits.DP_IS_USB_C) {
+			if (ctx->dc->links[i]->link_enc->transmitter >= TRANSMITTER_UNIPHY_A &&
+					ctx->dc->links[i]->link_enc->transmitter <= TRANSMITTER_UNIPHY_F) {
+				tx_interrupt_mask |= 1 << ctx->dc->links[i]->link_enc->transmitter;
+			}
+		}
+	}
+
 	switch (dmcu->dmcu_state) {
 	case DMCU_UNLOADED:
 		status = false;
@@ -401,6 +415,8 @@ static bool dcn10_dmcu_init(struct dmcu *dmcu)
 		/* Set backlight ramping stepsize */
 		REG_WRITE(MASTER_COMM_DATA_REG2, abm_gain_stepsize);
 
+		REG_WRITE(MASTER_COMM_DATA_REG3, tx_interrupt_mask);
+
 		/* Set command to initialize microcontroller */
 		REG_UPDATE(MASTER_COMM_CMD_REG, MASTER_COMM_CMD_REG_BYTE0,
 			MCP_INIT_DMCU);

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

@@ -0,0 +1,25 @@
+/*
+ * Copyright 2012-16 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
+ *
+ */
+

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

@@ -27,25 +27,55 @@
 #include "dc.h"
 #include "dc_dmub_srv.h"
 #include "../../dmub/inc/dmub_srv.h"
-#include "dmub_fw_state.h"
+#include "../../dmub/inc/dmub_gpint_cmd.h"
 #include "core_types.h"
-#include "ipp.h"
 
 #define MAX_PIPES 6
 
 /**
  * Get PSR state from firmware.
  */
-static void dmub_get_psr_state(uint32_t *psr_state)
+static void dmub_psr_get_state(struct dmub_psr *dmub, uint32_t *psr_state)
 {
-	// Not yet implemented
+	struct dmub_srv *srv = dmub->ctx->dmub_srv->dmub;
-	// Trigger GPINT interrupt from firmware
+
+	// Send gpint command and wait for ack
+	dmub_srv_send_gpint_command(srv, DMUB_GPINT__GET_PSR_STATE, 0, 30);
+
+	dmub_srv_get_gpint_response(srv, psr_state);
+}
+
+/**
+ * Set PSR version.
+ */
+static bool dmub_psr_set_version(struct dmub_psr *dmub, struct dc_stream_state *stream)
+{
+	union dmub_rb_cmd cmd;
+	struct dc_context *dc = dmub->ctx;
+
+	cmd.psr_set_version.header.type = DMUB_CMD__PSR;
+	cmd.psr_set_version.header.sub_type = DMUB_CMD__PSR_SET_VERSION;
+
+	if (stream->psr_version == 0x0) // Unsupported
+		return false;
+	else if (stream->psr_version == 0x1)
+		cmd.psr_set_version.psr_set_version_data.version = PSR_VERSION_1;
+	else if (stream->psr_version == 0x2)
+		cmd.psr_set_version.psr_set_version_data.version = PSR_VERSION_2;
+
+	cmd.psr_enable.header.payload_bytes = sizeof(struct dmub_cmd_psr_set_version_data);
+
+	dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd.psr_enable.header);
+	dc_dmub_srv_cmd_execute(dc->dmub_srv);
+	dc_dmub_srv_wait_idle(dc->dmub_srv);
+
+	return true;
 }
 
 /**
  * Enable/Disable PSR.
  */
-static void dmub_set_psr_enable(struct dmub_psr *dmub, bool enable)
+static void dmub_psr_enable(struct dmub_psr *dmub, bool enable)
 {
 	union dmub_rb_cmd cmd;
 	struct dc_context *dc = dmub->ctx;
@@ -67,13 +97,13 @@ static void dmub_set_psr_enable(struct dmub_psr *dmub, bool enable)
 /**
  * Set PSR level.
  */
-static void dmub_set_psr_level(struct dmub_psr *dmub, uint16_t psr_level)
+static void dmub_psr_set_level(struct dmub_psr *dmub, uint16_t psr_level)
 {
 	union dmub_rb_cmd cmd;
 	uint32_t psr_state = 0;
 	struct dc_context *dc = dmub->ctx;
 
-	dmub_get_psr_state(&psr_state);
+	dmub_psr_get_state(dmub, &psr_state);
 
 	if (psr_state == 0)
 		return;
@@ -91,7 +121,7 @@ static void dmub_set_psr_level(struct dmub_psr *dmub, uint16_t psr_level)
 /**
  * Setup PSR by programming phy registers and sending psr hw context values to firmware.
  */
-static bool dmub_setup_psr(struct dmub_psr *dmub,
+static bool dmub_psr_copy_settings(struct dmub_psr *dmub,
 		struct dc_link *link,
 		struct psr_context *psr_context)
 {
@@ -101,8 +131,9 @@ static bool dmub_setup_psr(struct dmub_psr *dmub,
 		= &cmd.psr_copy_settings.psr_copy_settings_data;
 	struct pipe_ctx *pipe_ctx = NULL;
 	struct resource_context *res_ctx = &link->ctx->dc->current_state->res_ctx;
+	int i = 0;
 
-	for (int i = 0; i < MAX_PIPES; i++) {
+	for (i = 0; i < MAX_PIPES; i++) {
 		if (res_ctx &&
 			res_ctx->pipe_ctx[i].stream &&
 			res_ctx->pipe_ctx[i].stream->link &&
@@ -113,9 +144,11 @@ static bool dmub_setup_psr(struct dmub_psr *dmub,
 		}
 	}
 
-	if (!pipe_ctx ||
+	if (!pipe_ctx || !&pipe_ctx->plane_res || !&pipe_ctx->stream_res)
-			!&pipe_ctx->plane_res ||
+		return false;
-			!&pipe_ctx->stream_res)
+
+	// First, set the psr version
+	if (!dmub_psr_set_version(dmub, pipe_ctx->stream))
 		return false;
 
 	// Program DP DPHY fast training registers
@@ -138,10 +171,6 @@ static bool dmub_setup_psr(struct dmub_psr *dmub,
 
 	copy_settings_data->mpcc_inst				= pipe_ctx->plane_res.mpcc_inst;
 
-	if (pipe_ctx->plane_res.hubp)
-		copy_settings_data->hubp_inst			= pipe_ctx->plane_res.hubp->inst;
-	else
-		copy_settings_data->hubp_inst			= 0;
 	if (pipe_ctx->plane_res.dpp)
 		copy_settings_data->dpp_inst			= pipe_ctx->plane_res.dpp->inst;
 	else
@@ -157,18 +186,9 @@ static bool dmub_setup_psr(struct dmub_psr *dmub,
 
 	// Misc
 	copy_settings_data->psr_level				= psr_context->psr_level.u32all;
-	copy_settings_data->hyst_frames				= psr_context->timehyst_frames;
-	copy_settings_data->hyst_lines				= psr_context->hyst_lines;
-	copy_settings_data->phy_type				= psr_context->phyType;
-	copy_settings_data->aux_repeat				= psr_context->aux_repeats;
 	copy_settings_data->smu_optimizations_en		= psr_context->allow_smu_optimizations;
-	copy_settings_data->skip_wait_for_pll_lock	= psr_context->skipPsrWaitForPllLock;
 	copy_settings_data->frame_delay				= psr_context->frame_delay;
-	copy_settings_data->smu_phy_id				= psr_context->smuPhyId;
-	copy_settings_data->num_of_controllers		= psr_context->numberOfControllers;
 	copy_settings_data->frame_cap_ind			= psr_context->psrFrameCaptureIndicationReq;
-	copy_settings_data->phy_num					= psr_context->frame_delay & 0x7;
-	copy_settings_data->link_rate				= psr_context->frame_delay & 0xF;
 
 	dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd.psr_copy_settings.header);
 	dc_dmub_srv_cmd_execute(dc->dmub_srv);
@@ -178,10 +198,10 @@ static bool dmub_setup_psr(struct dmub_psr *dmub,
 }
 
 static const struct dmub_psr_funcs psr_funcs = {
-	.set_psr_enable			= dmub_set_psr_enable,
+	.psr_copy_settings		= dmub_psr_copy_settings,
-	.setup_psr				= dmub_setup_psr,
+	.psr_enable			= dmub_psr_enable,
-	.get_psr_state			= dmub_get_psr_state,
+	.psr_get_state			= dmub_psr_get_state,
-	.set_psr_level			= dmub_set_psr_level,
+	.psr_set_level			= dmub_psr_set_level,
 };
 
 /**

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

@@ -27,6 +27,7 @@
 #define _DMUB_PSR_H_
 
 #include "os_types.h"
+#include "dc_link.h"
 
 struct dmub_psr {
 	struct dc_context *ctx;
@@ -34,14 +35,14 @@ struct dmub_psr {
 };
 
 struct dmub_psr_funcs {
-	void (*set_psr_enable)(struct dmub_psr *dmub, bool enable);
+	bool (*psr_copy_settings)(struct dmub_psr *dmub, struct dc_link *link, struct psr_context *psr_context);
-	bool (*setup_psr)(struct dmub_psr *dmub, struct dc_link *link, struct psr_context *psr_context);
+	void (*psr_enable)(struct dmub_psr *dmub, bool enable);
-	void (*get_psr_state)(uint32_t *psr_state);
+	void (*psr_get_state)(struct dmub_psr *dmub, uint32_t *psr_state);
-	void (*set_psr_level)(struct dmub_psr *dmub, uint16_t psr_level);
+	void (*psr_set_level)(struct dmub_psr *dmub, uint16_t psr_level);
 };
 
 struct dmub_psr *dmub_psr_create(struct dc_context *ctx);
 void dmub_psr_destroy(struct dmub_psr **dmub);
 
 
-#endif /* _DCE_DMUB_H_ */
+#endif /* _DMUB_PSR_H_ */

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

@@ -71,6 +71,8 @@
 #define PANEL_POWER_UP_TIMEOUT 300
 #define PANEL_POWER_DOWN_TIMEOUT 500
 #define HPD_CHECK_INTERVAL 10
+#define OLED_POST_T7_DELAY 100
+#define OLED_PRE_T11_DELAY 150
 
 #define CTX \
 	hws->ctx
@@ -696,8 +698,10 @@ void dce110_enable_stream(struct pipe_ctx *pipe_ctx)
 }
 
 /*todo: cloned in stream enc, fix*/
-static bool is_panel_backlight_on(struct dce_hwseq *hws)
+bool dce110_is_panel_backlight_on(struct dc_link *link)
 {
+	struct dc_context *ctx = link->ctx;
+	struct dce_hwseq *hws = ctx->dc->hwseq;
 	uint32_t value;
 
 	REG_GET(LVTMA_PWRSEQ_CNTL, LVTMA_BLON, &value);
@@ -705,11 +709,12 @@ static bool is_panel_backlight_on(struct dce_hwseq *hws)
 	return value;
 }
 
-static bool is_panel_powered_on(struct dce_hwseq *hws)
+bool dce110_is_panel_powered_on(struct dc_link *link)
 {
+	struct dc_context *ctx = link->ctx;
+	struct dce_hwseq *hws = ctx->dc->hwseq;
 	uint32_t pwr_seq_state, dig_on, dig_on_ovrd;
 
-
 	REG_GET(LVTMA_PWRSEQ_STATE, LVTMA_PWRSEQ_TARGET_STATE_R, &pwr_seq_state);
 
 	REG_GET_2(LVTMA_PWRSEQ_CNTL, LVTMA_DIGON, &dig_on, LVTMA_DIGON_OVRD, &dig_on_ovrd);
@@ -816,7 +821,7 @@ void dce110_edp_power_control(
 		return;
 	}
 
-	if (power_up != is_panel_powered_on(hwseq)) {
+	if (power_up != hwseq->funcs.is_panel_powered_on(link)) {
 		/* Send VBIOS command to prompt eDP panel power */
 		if (power_up) {
 			unsigned long long current_ts = dm_get_timestamp(ctx);
@@ -896,7 +901,7 @@ void dce110_edp_backlight_control(
 		return;
 	}
 
-	if (enable && is_panel_backlight_on(hws)) {
+	if (enable && hws->funcs.is_panel_backlight_on(link)) {
 		DC_LOG_HW_RESUME_S3(
 				"%s: panel already powered up. Do nothing.\n",
 				__func__);
@@ -936,9 +941,21 @@ void dce110_edp_backlight_control(
 	if (cntl.action == TRANSMITTER_CONTROL_BACKLIGHT_ON)
 		edp_receiver_ready_T7(link);
 	link_transmitter_control(ctx->dc_bios, &cntl);
+
+	if (enable && link->dpcd_sink_ext_caps.bits.oled)
+		msleep(OLED_POST_T7_DELAY);
+
+	if (link->dpcd_sink_ext_caps.bits.oled ||
+		link->dpcd_sink_ext_caps.bits.hdr_aux_backlight_control == 1 ||
+		link->dpcd_sink_ext_caps.bits.sdr_aux_backlight_control == 1)
+		dc_link_backlight_enable_aux(link, enable);
+
 	/*edp 1.2*/
 	if (cntl.action == TRANSMITTER_CONTROL_BACKLIGHT_OFF)
 		edp_receiver_ready_T9(link);
+
+	if (!enable && link->dpcd_sink_ext_caps.bits.oled)
+		msleep(OLED_PRE_T11_DELAY);
 }
 
 void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
@@ -2574,17 +2591,6 @@ static void dce110_apply_ctx_for_surface(
 	if (dc->fbc_compressor)
 		dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
-
-		if (stream == pipe_ctx->stream) {
-			if (!pipe_ctx->top_pipe &&
-				(pipe_ctx->plane_state || old_pipe_ctx->plane_state))
-				dc->hwss.pipe_control_lock(dc, pipe_ctx, true);
-		}
-	}
-
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
 
@@ -2607,20 +2613,16 @@ static void dce110_apply_ctx_for_surface(
 
 	}
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
-
-		if ((stream == pipe_ctx->stream) &&
-			(!pipe_ctx->top_pipe) &&
-			(pipe_ctx->plane_state || old_pipe_ctx->plane_state))
-			dc->hwss.pipe_control_lock(dc, pipe_ctx, false);
-	}
-
 	if (dc->fbc_compressor)
 		enable_fbc(dc, context);
 }
 
+static void dce110_post_unlock_program_front_end(
+		struct dc *dc,
+		struct dc_state *context)
+{
+}
+
 static void dce110_power_down_fe(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct dce_hwseq *hws = dc->hwseq;
@@ -2722,6 +2724,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.init_hw = init_hw,
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dce110_apply_ctx_for_surface,
+	.post_unlock_program_front_end = dce110_post_unlock_program_front_end,
 	.update_plane_addr = update_plane_addr,
 	.update_pending_status = dce110_update_pending_status,
 	.enable_accelerated_mode = dce110_enable_accelerated_mode,
@@ -2736,6 +2739,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.disable_audio_stream = dce110_disable_audio_stream,
 	.disable_plane = dce110_power_down_fe,
 	.pipe_control_lock = dce_pipe_control_lock,
+	.interdependent_update_lock = NULL,
 	.prepare_bandwidth = dce110_prepare_bandwidth,
 	.optimize_bandwidth = dce110_optimize_bandwidth,
 	.set_drr = set_drr,
@@ -2763,6 +2767,8 @@ static const struct hwseq_private_funcs dce110_private_funcs = {
 	.disable_stream_gating = NULL,
 	.enable_stream_gating = NULL,
 	.edp_backlight_control = dce110_edp_backlight_control,
+	.is_panel_backlight_on = dce110_is_panel_backlight_on,
+	.is_panel_powered_on = dce110_is_panel_powered_on,
 };
 
 void dce110_hw_sequencer_construct(struct dc *dc)

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

@@ -85,5 +85,9 @@ void dce110_edp_wait_for_hpd_ready(
 		struct dc_link *link,
 		bool power_up);
 
+bool dce110_is_panel_backlight_on(struct dc_link *link);
+
+bool dce110_is_panel_powered_on(struct dc_link *link);
+
 #endif /* __DC_HWSS_DCE110_H__ */
 

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c

@@ -316,6 +316,7 @@ bool cm_helper_translate_curve_to_hw_format(
 	struct pwl_result_data *rgb_resulted;
 	struct pwl_result_data *rgb;
 	struct pwl_result_data *rgb_plus_1;
+	struct pwl_result_data *rgb_minus_1;
 
 	int32_t region_start, region_end;
 	int32_t i;
@@ -465,9 +466,20 @@ bool cm_helper_translate_curve_to_hw_format(
 
 	rgb = rgb_resulted;
 	rgb_plus_1 = rgb_resulted + 1;
+	rgb_minus_1 = rgb;
 
 	i = 1;
 	while (i != hw_points + 1) {
+
+		if (i >= hw_points - 1) {
+			if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
+				rgb_plus_1->red = dc_fixpt_add(rgb->red, rgb_minus_1->delta_red);
+			if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
+				rgb_plus_1->green = dc_fixpt_add(rgb->green, rgb_minus_1->delta_green);
+			if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
+				rgb_plus_1->blue = dc_fixpt_add(rgb->blue, rgb_minus_1->delta_blue);
+		}
+
 		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
 		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
 		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
@@ -482,6 +494,7 @@ bool cm_helper_translate_curve_to_hw_format(
 		}
 
 		++rgb_plus_1;
+		rgb_minus_1 = rgb;
 		++rgb;
 		++i;
 	}

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c

@@ -128,8 +128,8 @@ bool hubbub1_verify_allow_pstate_change_high(
 	 * pstate takes around ~100us on linux. Unknown currently as to
 	 * why it takes that long on linux
 	 */
-	static unsigned int pstate_wait_timeout_us = 200;
+	const unsigned int pstate_wait_timeout_us = 200;
-	static unsigned int pstate_wait_expected_timeout_us = 40;
+	const unsigned int pstate_wait_expected_timeout_us = 40;
 	static unsigned int max_sampled_pstate_wait_us; /* data collection */
 	static bool forced_pstate_allow; /* help with revert wa */
 
@@ -147,8 +147,9 @@ bool hubbub1_verify_allow_pstate_change_high(
 		forced_pstate_allow = false;
 	}
 
-	/* RV2:
+	/* The following table only applies to DCN1 and DCN2,
-	 * dchubbubdebugind, at: 0xB
+	 * for newer DCNs, need to consult with HW IP folks to read RTL
+	 * HUBBUB:DCHUBBUB_TEST_ARB_DEBUG10 DCHUBBUBDEBUGIND:0xB
 	 * description
 	 * 0:     Pipe0 Plane0 Allow Pstate Change
 	 * 1:     Pipe0 Plane1 Allow Pstate Change
@@ -181,64 +182,6 @@ bool hubbub1_verify_allow_pstate_change_high(
 	 * 28:    WB0 Allow Pstate Change
 	 * 29:    WB1 Allow Pstate Change
 	 * 30:    Arbiter's allow_pstate_change
-	 * 31:    SOC pstate change request"
-	 */
-	/*DCN2.x:
-	HUBBUB:DCHUBBUB_TEST_ARB_DEBUG10 DCHUBBUBDEBUGIND:0xB
-	0: Pipe0 Plane0 Allow P-state Change
-	1: Pipe0 Plane1 Allow P-state Change
-	2: Pipe0 Cursor0 Allow P-state Change
-	3: Pipe0 Cursor1 Allow P-state Change
-	4: Pipe1 Plane0 Allow P-state Change
-	5: Pipe1 Plane1 Allow P-state Change
-	6: Pipe1 Cursor0 Allow P-state Change
-	7: Pipe1 Cursor1 Allow P-state Change
-	8: Pipe2 Plane0 Allow P-state Change
-	9: Pipe2 Plane1 Allow P-state Change
-	10: Pipe2 Cursor0 Allow P-state Change
-	11: Pipe2 Cursor1 Allow P-state Change
-	12: Pipe3 Plane0 Allow P-state Change
-	13: Pipe3 Plane1 Allow P-state Change
-	14: Pipe3 Cursor0 Allow P-state Change
-	15: Pipe3 Cursor1 Allow P-state Change
-	16: Pipe4 Plane0 Allow P-state Change
-	17: Pipe4 Plane1 Allow P-state Change
-	18: Pipe4 Cursor0 Allow P-state Change
-	19: Pipe4 Cursor1 Allow P-state Change
-	20: Pipe5 Plane0 Allow P-state Change
-	21: Pipe5 Plane1 Allow P-state Change
-	22: Pipe5 Cursor0 Allow P-state Change
-	23: Pipe5 Cursor1 Allow P-state Change
-	24: Pipe6 Plane0 Allow P-state Change
-	25: Pipe6 Plane1 Allow P-state Change
-	26: Pipe6 Cursor0 Allow P-state Change
-	27: Pipe6 Cursor1 Allow P-state Change
-	28: WB0 Allow P-state Change
-	29: WB1 Allow P-state Change
-	30: Arbiter`s Allow P-state Change
-	31: SOC P-state Change request
-	*/
-	/* RV1:
-	 * dchubbubdebugind, at: 0x7
-	 * description "3-0:   Pipe0 cursor0 QOS
-	 * 7-4:   Pipe1 cursor0 QOS
-	 * 11-8:  Pipe2 cursor0 QOS
-	 * 15-12: Pipe3 cursor0 QOS
-	 * 16:    Pipe0 Plane0 Allow Pstate Change
-	 * 17:    Pipe1 Plane0 Allow Pstate Change
-	 * 18:    Pipe2 Plane0 Allow Pstate Change
-	 * 19:    Pipe3 Plane0 Allow Pstate Change
-	 * 20:    Pipe0 Plane1 Allow Pstate Change
-	 * 21:    Pipe1 Plane1 Allow Pstate Change
-	 * 22:    Pipe2 Plane1 Allow Pstate Change
-	 * 23:    Pipe3 Plane1 Allow Pstate Change
-	 * 24:    Pipe0 cursor0 Allow Pstate Change
-	 * 25:    Pipe1 cursor0 Allow Pstate Change
-	 * 26:    Pipe2 cursor0 Allow Pstate Change
-	 * 27:    Pipe3 cursor0 Allow Pstate Change
-	 * 28:    WB0 Allow Pstate Change
-	 * 29:    WB1 Allow Pstate Change
-	 * 30:    Arbiter's allow_pstate_change
 	 * 31:    SOC pstate change request
 	 */
 
@@ -300,7 +243,7 @@ void hubbub1_wm_change_req_wa(struct hubbub *hubbub)
 			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 1);
 }
 
-void hubbub1_program_urgent_watermarks(
+bool hubbub1_program_urgent_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -308,6 +251,7 @@ void hubbub1_program_urgent_watermarks(
 {
 	struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
 	uint32_t prog_wm_value;
+	bool wm_pending = false;
 
 	/* Repeat for water mark set A, B, C and D. */
 	/* clock state A */
@@ -321,7 +265,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->a.urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->a.urgent_ns < hubbub1->watermarks.a.urgent_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->a.pte_meta_urgent_ns > hubbub1->watermarks.a.pte_meta_urgent_ns) {
 		hubbub1->watermarks.a.pte_meta_urgent_ns = watermarks->a.pte_meta_urgent_ns;
@@ -331,7 +276,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->a.pte_meta_urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->a.pte_meta_urgent_ns < hubbub1->watermarks.a.pte_meta_urgent_ns)
+		wm_pending = true;
 
 	/* clock state B */
 	if (safe_to_lower || watermarks->b.urgent_ns > hubbub1->watermarks.b.urgent_ns) {
@@ -344,7 +290,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->b.urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->b.urgent_ns < hubbub1->watermarks.b.urgent_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->b.pte_meta_urgent_ns > hubbub1->watermarks.b.pte_meta_urgent_ns) {
 		hubbub1->watermarks.b.pte_meta_urgent_ns = watermarks->b.pte_meta_urgent_ns;
@@ -354,7 +301,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->b.pte_meta_urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->b.pte_meta_urgent_ns < hubbub1->watermarks.b.pte_meta_urgent_ns)
+		wm_pending = true;
 
 	/* clock state C */
 	if (safe_to_lower || watermarks->c.urgent_ns > hubbub1->watermarks.c.urgent_ns) {
@@ -367,7 +315,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->c.urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->c.urgent_ns < hubbub1->watermarks.c.urgent_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->c.pte_meta_urgent_ns > hubbub1->watermarks.c.pte_meta_urgent_ns) {
 		hubbub1->watermarks.c.pte_meta_urgent_ns = watermarks->c.pte_meta_urgent_ns;
@@ -377,7 +326,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->c.pte_meta_urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->c.pte_meta_urgent_ns < hubbub1->watermarks.c.pte_meta_urgent_ns)
+		wm_pending = true;
 
 	/* clock state D */
 	if (safe_to_lower || watermarks->d.urgent_ns > hubbub1->watermarks.d.urgent_ns) {
@@ -390,7 +340,8 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->d.urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->d.urgent_ns < hubbub1->watermarks.d.urgent_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->d.pte_meta_urgent_ns > hubbub1->watermarks.d.pte_meta_urgent_ns) {
 		hubbub1->watermarks.d.pte_meta_urgent_ns = watermarks->d.pte_meta_urgent_ns;
@@ -400,10 +351,13 @@ void hubbub1_program_urgent_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->d.pte_meta_urgent_ns, prog_wm_value);
-	}
+	} else if (watermarks->d.pte_meta_urgent_ns < hubbub1->watermarks.d.pte_meta_urgent_ns)
+		wm_pending = true;
+
+	return wm_pending;
 }
 
-void hubbub1_program_stutter_watermarks(
+bool hubbub1_program_stutter_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -411,6 +365,7 @@ void hubbub1_program_stutter_watermarks(
 {
 	struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
 	uint32_t prog_wm_value;
+	bool wm_pending = false;
 
 	/* clock state A */
 	if (safe_to_lower || watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns
@@ -425,7 +380,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns
+			< hubbub1->watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->a.cstate_pstate.cstate_exit_ns
 			> hubbub1->watermarks.a.cstate_pstate.cstate_exit_ns) {
@@ -439,7 +396,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->a.cstate_pstate.cstate_exit_ns
+			< hubbub1->watermarks.a.cstate_pstate.cstate_exit_ns)
+		wm_pending = true;
 
 	/* clock state B */
 	if (safe_to_lower || watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns
@@ -454,7 +413,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns
+			< hubbub1->watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->b.cstate_pstate.cstate_exit_ns
 			> hubbub1->watermarks.b.cstate_pstate.cstate_exit_ns) {
@@ -468,7 +429,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->b.cstate_pstate.cstate_exit_ns
+			< hubbub1->watermarks.b.cstate_pstate.cstate_exit_ns)
+		wm_pending = true;
 
 	/* clock state C */
 	if (safe_to_lower || watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns
@@ -483,7 +446,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns
+			< hubbub1->watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->c.cstate_pstate.cstate_exit_ns
 			> hubbub1->watermarks.c.cstate_pstate.cstate_exit_ns) {
@@ -497,7 +462,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->c.cstate_pstate.cstate_exit_ns
+			< hubbub1->watermarks.c.cstate_pstate.cstate_exit_ns)
+		wm_pending = true;
 
 	/* clock state D */
 	if (safe_to_lower || watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns
@@ -512,7 +479,9 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns
+			< hubbub1->watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns)
+		wm_pending = true;
 
 	if (safe_to_lower || watermarks->d.cstate_pstate.cstate_exit_ns
 			> hubbub1->watermarks.d.cstate_pstate.cstate_exit_ns) {
@@ -526,11 +495,14 @@ void hubbub1_program_stutter_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
 			watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value);
-	}
+	} else if (watermarks->d.cstate_pstate.cstate_exit_ns
+			< hubbub1->watermarks.d.cstate_pstate.cstate_exit_ns)
+		wm_pending = true;
 
+	return wm_pending;
 }
 
-void hubbub1_program_pstate_watermarks(
+bool hubbub1_program_pstate_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -538,6 +510,7 @@ void hubbub1_program_pstate_watermarks(
 {
 	struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
 	uint32_t prog_wm_value;
+	bool wm_pending = false;
 
 	/* clock state A */
 	if (safe_to_lower || watermarks->a.cstate_pstate.pstate_change_ns
@@ -552,7 +525,9 @@ void hubbub1_program_pstate_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value);
-	}
+	} else if (watermarks->a.cstate_pstate.pstate_change_ns
+			< hubbub1->watermarks.a.cstate_pstate.pstate_change_ns)
+		wm_pending = true;
 
 	/* clock state B */
 	if (safe_to_lower || watermarks->b.cstate_pstate.pstate_change_ns
@@ -567,7 +542,9 @@ void hubbub1_program_pstate_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value);
-	}
+	} else if (watermarks->b.cstate_pstate.pstate_change_ns
+			< hubbub1->watermarks.b.cstate_pstate.pstate_change_ns)
+		wm_pending = true;
 
 	/* clock state C */
 	if (safe_to_lower || watermarks->c.cstate_pstate.pstate_change_ns
@@ -582,7 +559,9 @@ void hubbub1_program_pstate_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value);
-	}
+	} else if (watermarks->c.cstate_pstate.pstate_change_ns
+			< hubbub1->watermarks.c.cstate_pstate.pstate_change_ns)
+		wm_pending = true;
 
 	/* clock state D */
 	if (safe_to_lower || watermarks->d.cstate_pstate.pstate_change_ns
@@ -597,23 +576,33 @@ void hubbub1_program_pstate_watermarks(
 		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n\n",
 			watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value);
-	}
+	} else if (watermarks->d.cstate_pstate.pstate_change_ns
+			< hubbub1->watermarks.d.cstate_pstate.pstate_change_ns)
+		wm_pending = true;
+
+	return wm_pending;
 }
 
-void hubbub1_program_watermarks(
+bool hubbub1_program_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
 		bool safe_to_lower)
 {
 	struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
+	bool wm_pending = false;
 	/*
 	 * Need to clamp to max of the register values (i.e. no wrap)
 	 * for dcn1, all wm registers are 21-bit wide
 	 */
-	hubbub1_program_urgent_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower);
+	if (hubbub1_program_urgent_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower))
-	hubbub1_program_stutter_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower);
+		wm_pending = true;
-	hubbub1_program_pstate_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower);
+
+	if (hubbub1_program_stutter_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower))
+		wm_pending = true;
+
+	if (hubbub1_program_pstate_watermarks(hubbub, watermarks, refclk_mhz, safe_to_lower))
+		wm_pending = true;
 
 	REG_UPDATE(DCHUBBUB_ARB_SAT_LEVEL,
 			DCHUBBUB_ARB_SAT_LEVEL, 60 * refclk_mhz);
@@ -627,6 +616,7 @@ void hubbub1_program_watermarks(
 			DCHUBBUB_ARB_WATERMARK_CHANGE_DONE_INTERRUPT_DISABLE, 1,
 			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 1);
 #endif
+	return wm_pending;
 }
 
 void hubbub1_update_dchub(

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h

@@ -308,7 +308,7 @@ bool hubbub1_verify_allow_pstate_change_high(
 
 void hubbub1_wm_change_req_wa(struct hubbub *hubbub);
 
-void hubbub1_program_watermarks(
+bool hubbub1_program_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
@@ -331,17 +331,17 @@ void hubbub1_construct(struct hubbub *hubbub,
 	const struct dcn_hubbub_shift *hubbub_shift,
 	const struct dcn_hubbub_mask *hubbub_mask);
 
-void hubbub1_program_urgent_watermarks(
+bool hubbub1_program_urgent_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
 		bool safe_to_lower);
-void hubbub1_program_stutter_watermarks(
+bool hubbub1_program_stutter_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,
 		bool safe_to_lower);
-void hubbub1_program_pstate_watermarks(
+bool hubbub1_program_pstate_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
 		unsigned int refclk_mhz,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c

@@ -48,8 +48,8 @@
 #include "dc_link_dp.h"
 #include "dccg.h"
 #include "clk_mgr.h"
-
+#include "link_hwss.h"
-
+#include "dpcd_defs.h"
 #include "dsc.h"
 
 #define DC_LOGGER_INIT(logger)
@@ -82,7 +82,7 @@ void print_microsec(struct dc_context *dc_ctx,
 			us_x10 % frac);
 }
 
-static void dcn10_lock_all_pipes(struct dc *dc,
+void dcn10_lock_all_pipes(struct dc *dc,
 	struct dc_state *context,
 	bool lock)
 {
@@ -93,6 +93,7 @@ static void dcn10_lock_all_pipes(struct dc *dc,
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		pipe_ctx = &context->res_ctx.pipe_ctx[i];
 		tg = pipe_ctx->stream_res.tg;
+
 		/*
 		 * Only lock the top pipe's tg to prevent redundant
 		 * (un)locking. Also skip if pipe is disabled.
@@ -103,9 +104,9 @@ static void dcn10_lock_all_pipes(struct dc *dc,
 			continue;
 
 		if (lock)
-			tg->funcs->lock(tg);
+			dc->hwss.pipe_control_lock(dc, pipe_ctx, true);
 		else
-			tg->funcs->unlock(tg);
+			dc->hwss.pipe_control_lock(dc, pipe_ctx, false);
 	}
 }
 
@@ -900,6 +901,10 @@ static void dcn10_reset_back_end_for_pipe(
 	 * parent pipe.
 	 */
 	if (pipe_ctx->top_pipe == NULL) {
+
+		if (pipe_ctx->stream_res.abm)
+			pipe_ctx->stream_res.abm->funcs->set_abm_immediate_disable(pipe_ctx->stream_res.abm);
+
 		pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg);
 
 		pipe_ctx->stream_res.tg->funcs->enable_optc_clock(pipe_ctx->stream_res.tg, false);
@@ -1043,7 +1048,7 @@ void dcn10_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	if (opp != NULL)
 		opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true;
 
-	dc->optimized_required = true;
+	dc->clk_optimized_required = true;
 
 	if (hubp->funcs->hubp_disconnect)
 		hubp->funcs->hubp_disconnect(hubp);
@@ -1094,7 +1099,7 @@ void dcn10_plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
 				false);
 
 	hubp->power_gated = true;
-	dc->optimized_required = false; /* We're powering off, no need to optimize */
+	dc->clk_optimized_required = false; /* We're powering off, no need to optimize */
 
 	hws->funcs.plane_atomic_power_down(dc,
 			pipe_ctx->plane_res.dpp,
@@ -1263,6 +1268,7 @@ void dcn10_init_hw(struct dc *dc)
 		}
 
 		//Enable ability to power gate / don't force power on permanently
+		if (hws->funcs.enable_power_gating_plane)
 			hws->funcs.enable_power_gating_plane(hws, true);
 
 		return;
@@ -1317,6 +1323,31 @@ void dcn10_init_hw(struct dc *dc)
 		if (hws->funcs.dsc_pg_control != NULL)
 			hws->funcs.dsc_pg_control(hws, res_pool->dscs[i]->inst, false);
 
+	/* we want to turn off all dp displays before doing detection */
+	if (dc->config.power_down_display_on_boot) {
+		uint8_t dpcd_power_state = '\0';
+		enum dc_status status = DC_ERROR_UNEXPECTED;
+
+		for (i = 0; i < dc->link_count; i++) {
+			if (dc->links[i]->connector_signal != SIGNAL_TYPE_DISPLAY_PORT)
+				continue;
+
+			/*
+			 * core_link_read_dpcd() will invoke dm_helpers_dp_read_dpcd(),
+			 * which needs to read dpcd info with the help of aconnector.
+			 * If aconnector (dc->links[i]->prev) is NULL, then dpcd status
+			 * cannot be read.
+			 */
+			if (dc->links[i]->priv) {
+				/* if any of the displays are lit up turn them off */
+				status = core_link_read_dpcd(dc->links[i], DP_SET_POWER,
+								&dpcd_power_state, sizeof(dpcd_power_state));
+				if (status == DC_OK && dpcd_power_state == DP_POWER_STATE_D0)
+					dp_receiver_power_ctrl(dc->links[i], false);
+			}
+		}
+	}
+
 	/* If taking control over from VBIOS, we may want to optimize our first
 	 * mode set, so we need to skip powering down pipes until we know which
 	 * pipes we want to use.
@@ -1355,7 +1386,7 @@ void dcn10_init_hw(struct dc *dc)
 
 		REG_UPDATE(DCFCLK_CNTL, DCFCLK_GATE_DIS, 0);
 	}
-
+	if (hws->funcs.enable_power_gating_plane)
 		hws->funcs.enable_power_gating_plane(dc->hwseq, true);
 
 	if (dc->clk_mgr->funcs->notify_wm_ranges)
@@ -1576,7 +1607,7 @@ void dcn10_pipe_control_lock(
 	/* use TG master update lock to lock everything on the TG
 	 * therefore only top pipe need to lock
 	 */
-	if (pipe->top_pipe)
+	if (!pipe || pipe->top_pipe)
 		return;
 
 	if (dc->debug.sanity_checks)
@@ -2090,6 +2121,10 @@ void dcn10_get_hdr_visual_confirm_color(
 		if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) {
 			/* HDR10, ARGB2101010 - set boarder color to red */
 			color->color_r_cr = color_value;
+		} else if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) {
+			/* FreeSync 2 ARGB2101010 - set boarder color to pink */
+			color->color_r_cr = color_value;
+			color->color_b_cb = color_value;
 		}
 		break;
 	case PIXEL_FORMAT_FP16:
@@ -2512,12 +2547,17 @@ void dcn10_apply_ctx_for_surface(
 	int i;
 	struct timing_generator *tg;
 	uint32_t underflow_check_delay_us;
-	bool removed_pipe[4] = { false };
 	bool interdependent_update = false;
 	struct pipe_ctx *top_pipe_to_program =
 			dcn10_find_top_pipe_for_stream(dc, context, stream);
 	DC_LOGGER_INIT(dc->ctx->logger);
 
+	// Clear pipe_ctx flag
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+		pipe_ctx->update_flags.raw = 0;
+	}
+
 	if (!top_pipe_to_program)
 		return;
 
@@ -2531,11 +2571,6 @@ void dcn10_apply_ctx_for_surface(
 	if (underflow_check_delay_us != 0xFFFFFFFF && hws->funcs.did_underflow_occur)
 		ASSERT(hws->funcs.did_underflow_occur(dc, top_pipe_to_program));
 
-	if (interdependent_update)
-		dcn10_lock_all_pipes(dc, context, true);
-	else
-		dcn10_pipe_control_lock(dc, top_pipe_to_program, true);
-
 	if (underflow_check_delay_us != 0xFFFFFFFF)
 		udelay(underflow_check_delay_us);
 
@@ -2552,18 +2587,6 @@ void dcn10_apply_ctx_for_surface(
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
 		struct pipe_ctx *old_pipe_ctx =
 				&dc->current_state->res_ctx.pipe_ctx[i];
-		/*
-		 * Powergate reused pipes that are not powergated
-		 * fairly hacky right now, using opp_id as indicator
-		 * TODO: After move dc_post to dc_update, this will
-		 * be removed.
-		 */
-		if (pipe_ctx->plane_state && !old_pipe_ctx->plane_state) {
-			if (old_pipe_ctx->stream_res.tg == tg &&
-			    old_pipe_ctx->plane_res.hubp &&
-			    old_pipe_ctx->plane_res.hubp->opp_id != OPP_ID_INVALID)
-				dc->hwss.disable_plane(dc, old_pipe_ctx);
-		}
 
 		if ((!pipe_ctx->plane_state ||
 		     pipe_ctx->stream_res.tg != old_pipe_ctx->stream_res.tg) &&
@@ -2571,7 +2594,7 @@ void dcn10_apply_ctx_for_surface(
 		    old_pipe_ctx->stream_res.tg == tg) {
 
 			hws->funcs.plane_atomic_disconnect(dc, old_pipe_ctx);
-			removed_pipe[i] = true;
+			pipe_ctx->update_flags.bits.disable = 1;
 
 			DC_LOG_DC("Reset mpcc for pipe %d\n",
 					old_pipe_ctx->pipe_idx);
@@ -2597,21 +2620,35 @@ void dcn10_apply_ctx_for_surface(
 				&pipe_ctx->dlg_regs,
 				&pipe_ctx->ttu_regs);
 		}
+}
 
-	if (interdependent_update)
+void dcn10_post_unlock_program_front_end(
-		dcn10_lock_all_pipes(dc, context, false);
+		struct dc *dc,
-	else
+		struct dc_state *context)
-		dcn10_pipe_control_lock(dc, top_pipe_to_program, false);
+{
+	int i;
 
-	if (num_planes == 0)
+	DC_LOGGER_INIT(dc->ctx->logger);
-		false_optc_underflow_wa(dc, stream, tg);
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+		if (!pipe_ctx->top_pipe &&
+			!pipe_ctx->prev_odm_pipe &&
+			pipe_ctx->stream) {
+			struct timing_generator *tg = pipe_ctx->stream_res.tg;
+
+			if (context->stream_status[i].plane_count == 0)
+				false_optc_underflow_wa(dc, pipe_ctx->stream, tg);
+		}
+	}
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++)
-		if (removed_pipe[i])
+		if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable)
 			dc->hwss.disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++)
-		if (removed_pipe[i]) {
+		if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable) {
 			dc->hwss.optimize_bandwidth(dc, context);
 			break;
 		}
@@ -2656,7 +2693,7 @@ void dcn10_prepare_bandwidth(
 				false);
 	}
 
-	hubbub->funcs->program_watermarks(hubbub,
+	dc->wm_optimized_required = hubbub->funcs->program_watermarks(hubbub,
 			&context->bw_ctx.bw.dcn.watermarks,
 			dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000,
 			true);
@@ -2680,19 +2717,30 @@ void dcn10_optimize_bandwidth(
 		hws->funcs.verify_allow_pstate_change_high(dc);
 
 	if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) {
-		if (context->stream_count == 0)
+		if (context->stream_count == 0) {
 			context->bw_ctx.bw.dcn.clk.phyclk_khz = 0;
 
 			dc->clk_mgr->funcs->update_clocks(
 					dc->clk_mgr,
 					context,
 					true);
+		} else if (dc->clk_optimized_required || IS_DIAG_DC(dc->ctx->dce_environment)) {
+			dc->clk_mgr->funcs->update_clocks(
+								dc->clk_mgr,
+								context,
+								true);
+		}
 	}
 
+	if (dc->wm_optimized_required || IS_DIAG_DC(dc->ctx->dce_environment)) {
 		hubbub->funcs->program_watermarks(hubbub,
 				&context->bw_ctx.bw.dcn.watermarks,
 				dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000,
 				true);
+	}
+
+	dc->clk_optimized_required = false;
+	dc->wm_optimized_required = false;
 	dcn10_stereo_hw_frame_pack_wa(dc, context);
 
 	if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE)
@@ -2884,6 +2932,7 @@ void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx)
 	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct timing_generator *tg = pipe_ctx->stream_res.tg;
 	bool flip_pending;
+	struct dc *dc = plane_state->ctx->dc;
 
 	if (plane_state == NULL)
 		return;
@@ -2901,6 +2950,19 @@ void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx)
 		plane_state->status.is_right_eye =
 				!tg->funcs->is_stereo_left_eye(pipe_ctx->stream_res.tg);
 	}
+
+	if (dc->hwseq->wa_state.disallow_self_refresh_during_multi_plane_transition_applied) {
+		struct dce_hwseq *hwseq = dc->hwseq;
+		struct timing_generator *tg = dc->res_pool->timing_generators[0];
+		unsigned int cur_frame = tg->funcs->get_frame_count(tg);
+
+		if (cur_frame != hwseq->wa_state.disallow_self_refresh_during_multi_plane_transition_applied_on_frame) {
+			struct hubbub *hubbub = dc->res_pool->hubbub;
+
+			hubbub->funcs->allow_self_refresh_control(hubbub, !dc->debug.disable_stutter);
+			hwseq->wa_state.disallow_self_refresh_during_multi_plane_transition_applied = false;
+		}
+	}
 }
 
 void dcn10_update_dchub(struct dce_hwseq *hws, struct dchub_init_data *dh_data)

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h

@@ -70,11 +70,18 @@ void dcn10_reset_hw_ctx_wrap(
 		struct dc *dc,
 		struct dc_state *context);
 void dcn10_disable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx);
+void dcn10_lock_all_pipes(
+		struct dc *dc,
+		struct dc_state *context,
+		bool lock);
 void dcn10_apply_ctx_for_surface(
 		struct dc *dc,
 		const struct dc_stream_state *stream,
 		int num_planes,
 		struct dc_state *context);
+void dcn10_post_unlock_program_front_end(
+		struct dc *dc,
+		struct dc_state *context);
 void dcn10_hubp_pg_control(
 		struct dce_hwseq *hws,
 		unsigned int hubp_inst,