Merge tag 'amd-drm-next-5.7-2020-03-10' of git://people.freedesktop.o…

…rg/~agd5f/linux into drm-next

amd-drm-next-5.7-2020-03-10:

amdgpu:
- SR-IOV fixes
- Fix up fallout from drm load/unload callback removal
- Navi, renoir power management watermark fixes
- Refactor smu parameter handling
- Display FEC fixes
- Display DCC fixes
- HDCP fixes
- Add support for USB-C PD firmware updates
- Pollock detection fix
- Rework compute ring priority handling
- RAS fixes
- Misc cleanups

amdkfd:
- Consolidate more gfx config details in amdgpu
- Consolidate bo alloc flags
- Improve code comments
- SDMA MQD fixes
- Misc cleanups

gpu scheduler:
- Add suport for modifying the sched list

uapi:
- Clarify comments about GEM_CREATE flags that are not used by userspace.
  The kernel driver has always prevented userspace from using these.
  They are only used internally in the kernel driver.

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200310212748.4519-1-alexander.deucher@amd.com

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

@@ -579,6 +579,7 @@ struct amdgpu_asic_funcs {
 	/* invalidate hdp read cache */
 	void (*invalidate_hdp)(struct amdgpu_device *adev,
 			       struct amdgpu_ring *ring);
+	void (*reset_hdp_ras_error_count)(struct amdgpu_device *adev);
 	/* check if the asic needs a full reset of if soft reset will work */
 	bool (*need_full_reset)(struct amdgpu_device *adev);
 	/* initialize doorbell layout for specific asic*/

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

@@ -29,6 +29,7 @@
 #include <linux/module.h>
 #include <linux/dma-buf.h>
 #include "amdgpu_xgmi.h"
+#include <uapi/linux/kfd_ioctl.h>
 
 static const unsigned int compute_vmid_bitmap = 0xFF00;
 
@@ -224,7 +225,7 @@ void amdgpu_amdkfd_gpu_reset(struct kgd_dev *kgd)
 
 int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 				void **mem_obj, uint64_t *gpu_addr,
-				void **cpu_ptr, bool mqd_gfx9)
+				void **cpu_ptr, bool cp_mqd_gfx9)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
 	struct amdgpu_bo *bo = NULL;
@@ -240,8 +241,8 @@ int amdgpu_amdkfd_alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 	bp.type = ttm_bo_type_kernel;
 	bp.resv = NULL;
 
-	if (mqd_gfx9)
-		bp.flags |= AMDGPU_GEM_CREATE_MQD_GFX9;
+	if (cp_mqd_gfx9)
+		bp.flags |= AMDGPU_GEM_CREATE_CP_MQD_GFX9;
 
 	r = amdgpu_bo_create(adev, &bp, &bo);
 	if (r) {
@@ -501,10 +502,11 @@ int amdgpu_amdkfd_get_dmabuf_info(struct kgd_dev *kgd, int dma_buf_fd,
 					   metadata_size, &metadata_flags);
 	if (flags) {
 		*flags = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
-			ALLOC_MEM_FLAGS_VRAM : ALLOC_MEM_FLAGS_GTT;
+				KFD_IOC_ALLOC_MEM_FLAGS_VRAM
+				: KFD_IOC_ALLOC_MEM_FLAGS_GTT;
 
 		if (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)
-			*flags |= ALLOC_MEM_FLAGS_PUBLIC;
+			*flags |= KFD_IOC_ALLOC_MEM_FLAGS_PUBLIC;
 	}
 
 out_put:

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

@@ -242,6 +242,9 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
 void amdgpu_amdkfd_gpuvm_init_mem_limits(void);
 void amdgpu_amdkfd_unreserve_memory_limit(struct amdgpu_bo *bo);
 
+int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
+				struct tile_config *config);
+
 /* KGD2KFD callbacks */
 int kgd2kfd_init(void);
 void kgd2kfd_exit(void);

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

@@ -319,7 +319,6 @@ const struct kfd2kgd_calls arcturus_kfd2kgd = {
 	.address_watch_get_offset = kgd_gfx_v9_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_info =
 			kgd_gfx_v9_get_atc_vmid_pasid_mapping_info,
-	.get_tile_config = kgd_gfx_v9_get_tile_config,
 	.set_vm_context_page_table_base = kgd_set_vm_context_page_table_base,
 	.get_hive_id = amdgpu_amdkfd_get_hive_id,
 };

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

@@ -42,38 +42,6 @@ enum hqd_dequeue_request_type {
 	SAVE_WAVES
 };
 
-/* Because of REG_GET_FIELD() being used, we put this function in the
- * asic specific file.
- */
-static int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
-		struct tile_config *config)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
-
-	config->gb_addr_config = adev->gfx.config.gb_addr_config;
-#if 0
-/* TODO - confirm REG_GET_FIELD x2, should be OK as is... but
- * MC_ARB_RAMCFG register doesn't exist on Vega10 - initial amdgpu
- * changes commented out related code, doing the same here for now but
- * need to sync with Ken et al
- */
-	config->num_banks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFBANK);
-	config->num_ranks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFRANKS);
-#endif
-
-	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
-	config->num_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
-	config->macro_tile_config_ptr =
-			adev->gfx.config.macrotile_mode_array;
-	config->num_macro_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
-
-	return 0;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -805,7 +773,6 @@ const struct kfd2kgd_calls gfx_v10_kfd2kgd = {
 	.address_watch_get_offset = kgd_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_info =
 			get_atc_vmid_pasid_mapping_info,
-	.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_v7.c

@@ -84,31 +84,6 @@ union TCP_WATCH_CNTL_BITS {
 	float f32All;
 };
 
-/* Because of REG_GET_FIELD() being used, we put this function in the
- * asic specific file.
- */
-static int get_tile_config(struct kgd_dev *kgd,
-		struct tile_config *config)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
-
-	config->gb_addr_config = adev->gfx.config.gb_addr_config;
-	config->num_banks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFBANK);
-	config->num_ranks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFRANKS);
-
-	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
-	config->num_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
-	config->macro_tile_config_ptr =
-			adev->gfx.config.macrotile_mode_array;
-	config->num_macro_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
-
-	return 0;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -730,7 +705,6 @@ const struct kfd2kgd_calls gfx_v7_kfd2kgd = {
 	.address_watch_get_offset = kgd_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_info = get_atc_vmid_pasid_mapping_info,
 	.set_scratch_backing_va = set_scratch_backing_va,
-	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
 	.read_vmid_from_vmfault_reg = read_vmid_from_vmfault_reg,
 };

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c

@@ -41,31 +41,6 @@ enum hqd_dequeue_request_type {
 	RESET_WAVES
 };
 
-/* Because of REG_GET_FIELD() being used, we put this function in the
- * asic specific file.
- */
-static int get_tile_config(struct kgd_dev *kgd,
-		struct tile_config *config)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
-
-	config->gb_addr_config = adev->gfx.config.gb_addr_config;
-	config->num_banks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFBANK);
-	config->num_ranks = REG_GET_FIELD(adev->gfx.config.mc_arb_ramcfg,
-				MC_ARB_RAMCFG, NOOFRANKS);
-
-	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
-	config->num_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
-	config->macro_tile_config_ptr =
-			adev->gfx.config.macrotile_mode_array;
-	config->num_macro_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
-
-	return 0;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -676,6 +651,5 @@ const struct kfd2kgd_calls gfx_v8_kfd2kgd = {
 	.get_atc_vmid_pasid_mapping_info =
 			get_atc_vmid_pasid_mapping_info,
 	.set_scratch_backing_va = set_scratch_backing_va,
-	.get_tile_config = get_tile_config,
 	.set_vm_context_page_table_base = set_vm_context_page_table_base,
 };

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

@@ -48,28 +48,6 @@ enum hqd_dequeue_request_type {
 	RESET_WAVES
 };
 
-
-/* Because of REG_GET_FIELD() being used, we put this function in the
- * asic specific file.
- */
-int kgd_gfx_v9_get_tile_config(struct kgd_dev *kgd,
-		struct tile_config *config)
-{
-	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
-
-	config->gb_addr_config = adev->gfx.config.gb_addr_config;
-
-	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
-	config->num_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
-	config->macro_tile_config_ptr =
-			adev->gfx.config.macrotile_mode_array;
-	config->num_macro_tile_configs =
-			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
-
-	return 0;
-}
-
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
 {
 	return (struct amdgpu_device *)kgd;
@@ -736,7 +714,6 @@ const struct kfd2kgd_calls gfx_v9_kfd2kgd = {
 	.address_watch_get_offset = kgd_gfx_v9_address_watch_get_offset,
 	.get_atc_vmid_pasid_mapping_info =
 			kgd_gfx_v9_get_atc_vmid_pasid_mapping_info,
-	.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_gfx_v9.h

@@ -60,5 +60,3 @@ uint32_t kgd_gfx_v9_address_watch_get_offset(struct kgd_dev *kgd,
 
 bool kgd_gfx_v9_get_atc_vmid_pasid_mapping_info(struct kgd_dev *kgd,
 					uint8_t vmid, uint16_t *p_pasid);
-int kgd_gfx_v9_get_tile_config(struct kgd_dev *kgd,
-		struct tile_config *config);

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

@@ -29,6 +29,7 @@
 #include "amdgpu_vm.h"
 #include "amdgpu_amdkfd.h"
 #include "amdgpu_dma_buf.h"
+#include <uapi/linux/kfd_ioctl.h>
 
 /* BO flag to indicate a KFD userptr BO */
 #define AMDGPU_AMDKFD_USERPTR_BO (1ULL << 63)
@@ -400,18 +401,18 @@ static int vm_update_pds(struct amdgpu_vm *vm, struct amdgpu_sync *sync)
 static uint64_t get_pte_flags(struct amdgpu_device *adev, struct kgd_mem *mem)
 {
 	struct amdgpu_device *bo_adev = amdgpu_ttm_adev(mem->bo->tbo.bdev);
-	bool coherent = mem->alloc_flags & ALLOC_MEM_FLAGS_COHERENT;
+	bool coherent = mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_COHERENT;
 	uint32_t mapping_flags;
 
 	mapping_flags = AMDGPU_VM_PAGE_READABLE;
-	if (mem->alloc_flags & ALLOC_MEM_FLAGS_WRITABLE)
+	if (mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE)
 		mapping_flags |= AMDGPU_VM_PAGE_WRITEABLE;
-	if (mem->alloc_flags & ALLOC_MEM_FLAGS_EXECUTABLE)
+	if (mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE)
 		mapping_flags |= AMDGPU_VM_PAGE_EXECUTABLE;
 
 	switch (adev->asic_type) {
 	case CHIP_ARCTURUS:
-		if (mem->alloc_flags & ALLOC_MEM_FLAGS_VRAM) {
+		if (mem->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) {
 			if (bo_adev == adev)
 				mapping_flags |= coherent ?
 					AMDGPU_VM_MTYPE_CC : AMDGPU_VM_MTYPE_RW;
@@ -1160,24 +1161,24 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	/*
 	 * Check on which domain to allocate BO
 	 */
-	if (flags & ALLOC_MEM_FLAGS_VRAM) {
+	if (flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) {
 		domain = alloc_domain = AMDGPU_GEM_DOMAIN_VRAM;
 		alloc_flags = AMDGPU_GEM_CREATE_VRAM_WIPE_ON_RELEASE;
-		alloc_flags |= (flags & ALLOC_MEM_FLAGS_PUBLIC) ?
+		alloc_flags |= (flags & KFD_IOC_ALLOC_MEM_FLAGS_PUBLIC) ?
 			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED :
 			AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
-	} else if (flags & ALLOC_MEM_FLAGS_GTT) {
+	} else if (flags & KFD_IOC_ALLOC_MEM_FLAGS_GTT) {
 		domain = alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
 		alloc_flags = 0;
-	} else if (flags & ALLOC_MEM_FLAGS_USERPTR) {
+	} else if (flags & KFD_IOC_ALLOC_MEM_FLAGS_USERPTR) {
 		domain = AMDGPU_GEM_DOMAIN_GTT;
 		alloc_domain = AMDGPU_GEM_DOMAIN_CPU;
 		alloc_flags = 0;
 		if (!offset || !*offset)
 			return -EINVAL;
 		user_addr = untagged_addr(*offset);
-	} else if (flags & (ALLOC_MEM_FLAGS_DOORBELL |
-			ALLOC_MEM_FLAGS_MMIO_REMAP)) {
+	} else if (flags & (KFD_IOC_ALLOC_MEM_FLAGS_DOORBELL |
+			KFD_IOC_ALLOC_MEM_FLAGS_MMIO_REMAP)) {
 		domain = AMDGPU_GEM_DOMAIN_GTT;
 		alloc_domain = AMDGPU_GEM_DOMAIN_CPU;
 		bo_type = ttm_bo_type_sg;
@@ -1198,7 +1199,7 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	}
 	INIT_LIST_HEAD(&(*mem)->bo_va_list);
 	mutex_init(&(*mem)->lock);
-	(*mem)->aql_queue = !!(flags & ALLOC_MEM_FLAGS_AQL_QUEUE_MEM);
+	(*mem)->aql_queue = !!(flags & KFD_IOC_ALLOC_MEM_FLAGS_AQL_QUEUE_MEM);
 
 	/* Workaround for AQL queue wraparound bug. Map the same
 	 * memory twice. That means we only actually allocate half
@@ -1680,10 +1681,12 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
 
 	INIT_LIST_HEAD(&(*mem)->bo_va_list);
 	mutex_init(&(*mem)->lock);
+
 	(*mem)->alloc_flags =
 		((bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
-		 ALLOC_MEM_FLAGS_VRAM : ALLOC_MEM_FLAGS_GTT) |
-		ALLOC_MEM_FLAGS_WRITABLE | ALLOC_MEM_FLAGS_EXECUTABLE;
+		KFD_IOC_ALLOC_MEM_FLAGS_VRAM : KFD_IOC_ALLOC_MEM_FLAGS_GTT)
+		| KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE
+		| KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
 
 	(*mem)->bo = amdgpu_bo_ref(bo);
 	(*mem)->va = va;
@@ -2242,3 +2245,25 @@ int amdgpu_amdkfd_remove_gws_from_process(void *info, void *mem)
 	kfree(mem);
 	return 0;
 }
+
+/* Returns GPU-specific tiling mode information */
+int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
+				struct tile_config *config)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	config->gb_addr_config = adev->gfx.config.gb_addr_config;
+	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
+	config->num_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
+	config->macro_tile_config_ptr =
+			adev->gfx.config.macrotile_mode_array;
+	config->num_macro_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
+
+	/* Those values are not set from GFX9 onwards */
+	config->num_banks = adev->gfx.config.num_banks;
+	config->num_ranks = adev->gfx.config.num_ranks;
+
+	return 0;
+}

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

@@ -1208,7 +1208,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
 	struct drm_sched_entity *entity = p->entity;
 	enum drm_sched_priority priority;
-	struct amdgpu_ring *ring;
 	struct amdgpu_bo_list_entry *e;
 	struct amdgpu_job *job;
 	uint64_t seq;
@@ -1261,9 +1260,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	priority = job->base.s_priority;
 	drm_sched_entity_push_job(&job->base, entity);
 
-	ring = to_amdgpu_ring(entity->rq->sched);
-	amdgpu_ring_priority_get(ring, priority);
-
 	amdgpu_vm_move_to_lru_tail(p->adev, &fpriv->vm);
 
 	ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence);