Merge branch 'drm-next-4.16' of git://people.freedesktop.org/~agd5f/linux into drm-next

* 'drm-next-4.16' of git://people.freedesktop.org/~agd5f/linux: (171 commits)
  drm/amdgpu: fix test for shadow page tables
  drm/amd/display: Expose dpp1_set_cursor_attributes
  drm/amd/display: Update FMT and OPPBUF functions
  drm/amd/display: check for null before calling is_blanked
  drm/amd/display: dal 3.1.27
  drm/amd/display: Fix unused variable warnings.
  drm/amd/display: Only blank DCN when we have set_blank implementation
  drm/amd/display: Put dcn_mi_registers with other structs
  drm/amd/display: hubp refactor
  drm/amd/display: integrating optc pseudocode
  drm/amd/display: Call validate_fbc should_enable_fbc
  drm/amd/display: Clean up DCN cursor code
  drm/amd/display: fix 180 full screen pipe split
  drm/amd/display: reprogram surface config on scaling change
  drm/amd/display: Remove dwbc from pipe_ctx
  drm/amd/display: Use the maximum link setting which EDP reported.
  drm/amd/display: Add hdr_supported flag
  drm/amd/display: fix global sync param retrieval when not pipe splitting
  drm/amd/display: Update HUBP
  drm/amd/display: fix rotated surface scaling
  ...

MAINTAINERS

@@ -4811,6 +4811,15 @@ S:	Maintained
 F:	drivers/gpu/drm/tinydrm/
 F:	include/drm/tinydrm/
 
+DRM TTM SUBSYSTEM
+M:	Christian Koenig <christian.koenig@amd.com>
+M:	Roger He <Hongbo.He@amd.com>
+T:	git git://people.freedesktop.org/~agd5f/linux
+S:	Maintained
+L:	dri-devel@lists.freedesktop.org
+F:	include/drm/ttm/
+F:	drivers/gpu/drm/ttm/
+
 DSBR100 USB FM RADIO DRIVER
 M:	Alexey Klimov <klimov.linux@gmail.com>
 L:	linux-media@vger.kernel.org

drivers/gpu/drm/Kconfig

@@ -154,6 +154,10 @@ config DRM_VM
 	bool
 	depends on DRM && MMU
 
+config DRM_SCHED
+	tristate
+	depends on DRM
+
 source "drivers/gpu/drm/i2c/Kconfig"
 
 source "drivers/gpu/drm/arm/Kconfig"
@@ -183,6 +187,7 @@ config DRM_AMDGPU
 	depends on DRM && PCI && MMU
 	select FW_LOADER
         select DRM_KMS_HELPER
+	select DRM_SCHED
         select DRM_TTM
 	select POWER_SUPPLY
 	select HWMON

drivers/gpu/drm/Makefile

@@ -102,3 +102,4 @@ obj-$(CONFIG_DRM_MXSFB)	+= mxsfb/
 obj-$(CONFIG_DRM_TINYDRM) += tinydrm/
 obj-$(CONFIG_DRM_PL111) += pl111/
 obj-$(CONFIG_DRM_TVE200) += tve200/
+obj-$(CONFIG_DRM_SCHED)	+= scheduler/

drivers/gpu/drm/amd/amdgpu/Makefile

@@ -52,7 +52,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 	amdgpu_prime.o amdgpu_vm.o amdgpu_ib.o amdgpu_pll.o \
 	amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \
 	amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \
-	amdgpu_queue_mgr.o amdgpu_vf_error.o amdgpu_sched.o
+	amdgpu_queue_mgr.o amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o
 
 # add asic specific block
 amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
@@ -62,7 +62,7 @@ amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
 amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o
 
 amdgpu-y += \
-	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o mxgpu_ai.o nbio_v7_0.o
+	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o
 
 # add GMC block
 amdgpu-y += \
@@ -135,10 +135,7 @@ amdgpu-y += \
 amdgpu-y += amdgpu_cgs.o
 
 # GPU scheduler
-amdgpu-y += \
-	../scheduler/gpu_scheduler.o \
-	../scheduler/sched_fence.o \
-	amdgpu_job.o
+amdgpu-y += amdgpu_job.o
 
 # ACP componet
 ifneq ($(CONFIG_DRM_AMD_ACP),)

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

@@ -45,6 +45,7 @@
 #include <drm/drmP.h>
 #include <drm/drm_gem.h>
 #include <drm/amdgpu_drm.h>
+#include <drm/gpu_scheduler.h>
 
 #include <kgd_kfd_interface.h>
 #include "dm_pp_interface.h"
@@ -68,10 +69,9 @@
 #include "amdgpu_vcn.h"
 #include "amdgpu_mn.h"
 #include "amdgpu_dm.h"
-#include "gpu_scheduler.h"
 #include "amdgpu_virt.h"
 #include "amdgpu_gart.h"
-
+#include "amdgpu_debugfs.h"
 
 /*
  * Modules parameters.
@@ -126,6 +126,7 @@ extern int amdgpu_param_buf_per_se;
 extern int amdgpu_job_hang_limit;
 extern int amdgpu_lbpw;
 extern int amdgpu_compute_multipipe;
+extern int amdgpu_gpu_recovery;
 
 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
@@ -223,17 +224,18 @@ enum amdgpu_kiq_irq {
 	AMDGPU_CP_KIQ_IRQ_LAST
 };
 
-int amdgpu_set_clockgating_state(struct amdgpu_device *adev,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state);
-int amdgpu_set_powergating_state(struct amdgpu_device *adev,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state);
-void amdgpu_get_clockgating_state(struct amdgpu_device *adev, u32 *flags);
-int amdgpu_wait_for_idle(struct amdgpu_device *adev,
-			 enum amd_ip_block_type block_type);
-bool amdgpu_is_idle(struct amdgpu_device *adev,
-		    enum amd_ip_block_type block_type);
+int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+					   enum amd_ip_block_type block_type,
+					   enum amd_clockgating_state state);
+int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+					   enum amd_ip_block_type block_type,
+					   enum amd_powergating_state state);
+void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev,
+					    u32 *flags);
+int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev,
+				   enum amd_ip_block_type block_type);
+bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev,
+			      enum amd_ip_block_type block_type);
 
 #define AMDGPU_MAX_IP_NUM 16
 
@@ -258,15 +260,16 @@ struct amdgpu_ip_block {
 	const struct amdgpu_ip_block_version *version;
 };
 
-int amdgpu_ip_block_version_cmp(struct amdgpu_device *adev,
-				enum amd_ip_block_type type,
-				u32 major, u32 minor);
+int amdgpu_device_ip_block_version_cmp(struct amdgpu_device *adev,
+				       enum amd_ip_block_type type,
+				       u32 major, u32 minor);
 
-struct amdgpu_ip_block * amdgpu_get_ip_block(struct amdgpu_device *adev,
-					     enum amd_ip_block_type type);
+struct amdgpu_ip_block *
+amdgpu_device_ip_get_ip_block(struct amdgpu_device *adev,
+			      enum amd_ip_block_type type);
 
-int amdgpu_ip_block_add(struct amdgpu_device *adev,
-			const struct amdgpu_ip_block_version *ip_block_version);
+int amdgpu_device_ip_block_add(struct amdgpu_device *adev,
+			       const struct amdgpu_ip_block_version *ip_block_version);
 
 /* provided by hw blocks that can move/clear data.  e.g., gfx or sdma */
 struct amdgpu_buffer_funcs {
@@ -346,7 +349,8 @@ struct amdgpu_gart_funcs {
 	uint64_t (*get_vm_pte_flags)(struct amdgpu_device *adev,
 				     uint32_t flags);
 	/* get the pde for a given mc addr */
-	u64 (*get_vm_pde)(struct amdgpu_device *adev, u64 addr);
+	void (*get_vm_pde)(struct amdgpu_device *adev, int level,
+			   u64 *dst, u64 *flags);
 	uint32_t (*get_invalidate_req)(unsigned int vm_id);
 };
 
@@ -373,9 +377,6 @@ struct amdgpu_dummy_page {
 	struct page	*page;
 	dma_addr_t	addr;
 };
-int amdgpu_dummy_page_init(struct amdgpu_device *adev);
-void amdgpu_dummy_page_fini(struct amdgpu_device *adev);
-
 
 /*
  * Clocks
@@ -423,7 +424,6 @@ struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *);
 void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj);
 void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr);
 int amdgpu_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
-int amdgpu_gem_debugfs_init(struct amdgpu_device *adev);
 
 /* sub-allocation manager, it has to be protected by another lock.
  * By conception this is an helper for other part of the driver
@@ -540,6 +540,7 @@ struct amdgpu_mc {
 	u64					private_aperture_end;
 	/* protects concurrent invalidation */
 	spinlock_t		invalidate_lock;
+	bool			translate_further;
 };
 
 /*
@@ -650,12 +651,6 @@ typedef enum _AMDGPU_DOORBELL64_ASSIGNMENT
 	AMDGPU_DOORBELL64_INVALID                 = 0xFFFF
 } AMDGPU_DOORBELL64_ASSIGNMENT;
 
-
-void amdgpu_doorbell_get_kfd_info(struct amdgpu_device *adev,
-				phys_addr_t *aperture_base,
-				size_t *aperture_size,
-				size_t *start_offset);
-
 /*
  * IRQS.
  */
@@ -689,7 +684,7 @@ struct amdgpu_ib {
 	uint32_t			flags;
 };
 
-extern const struct amd_sched_backend_ops amdgpu_sched_ops;
+extern const struct drm_sched_backend_ops amdgpu_sched_ops;
 
 int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
 		     struct amdgpu_job **job, struct amdgpu_vm *vm);
@@ -699,7 +694,7 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
 void amdgpu_job_free_resources(struct amdgpu_job *job);
 void amdgpu_job_free(struct amdgpu_job *job);
 int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
-		      struct amd_sched_entity *entity, void *owner,
+		      struct drm_sched_entity *entity, void *owner,
 		      struct dma_fence **f);
 
 /*
@@ -732,7 +727,7 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 struct amdgpu_ctx_ring {
 	uint64_t		sequence;
 	struct dma_fence	**fences;
-	struct amd_sched_entity	entity;
+	struct drm_sched_entity	entity;
 };
 
 struct amdgpu_ctx {
@@ -746,8 +741,8 @@ struct amdgpu_ctx {
 	struct dma_fence	**fences;
 	struct amdgpu_ctx_ring	rings[AMDGPU_MAX_RINGS];
 	bool			preamble_presented;
-	enum amd_sched_priority init_priority;
-	enum amd_sched_priority override_priority;
+	enum drm_sched_priority init_priority;
+	enum drm_sched_priority override_priority;
 	struct mutex            lock;
 	atomic_t	guilty;
 };
@@ -767,7 +762,7 @@ int amdgpu_ctx_add_fence(struct amdgpu_ctx *ctx, struct amdgpu_ring *ring,
 struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
 				   struct amdgpu_ring *ring, uint64_t seq);
 void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx,
-				  enum amd_sched_priority priority);
+				  enum drm_sched_priority priority);
 
 int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 		     struct drm_file *filp);
@@ -1116,7 +1111,7 @@ struct amdgpu_cs_parser {
 #define AMDGPU_HAVE_CTX_SWITCH              (1 << 2) /* bit set means context switch occured */
 
 struct amdgpu_job {
-	struct amd_sched_job    base;
+	struct drm_sched_job    base;
 	struct amdgpu_device	*adev;
 	struct amdgpu_vm	*vm;
 	struct amdgpu_ring	*ring;
@@ -1170,10 +1165,10 @@ struct amdgpu_wb {
 	unsigned long		used[DIV_ROUND_UP(AMDGPU_MAX_WB, BITS_PER_LONG)];
 };
 
-int amdgpu_wb_get(struct amdgpu_device *adev, u32 *wb);
-void amdgpu_wb_free(struct amdgpu_device *adev, u32 wb);
+int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb);
+void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb);
 
-void amdgpu_get_pcie_info(struct amdgpu_device *adev);
+void amdgpu_device_get_pcie_info(struct amdgpu_device *adev);
 
 /*
  * SDMA
@@ -1238,24 +1233,6 @@ void amdgpu_benchmark(struct amdgpu_device *adev, int test_number);
  */
 void amdgpu_test_moves(struct amdgpu_device *adev);
 
-/*
- * Debugfs
- */
-struct amdgpu_debugfs {
-	const struct drm_info_list	*files;
-	unsigned		num_files;
-};
-
-int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
-			     const struct drm_info_list *files,
-			     unsigned nfiles);
-int amdgpu_debugfs_fence_init(struct amdgpu_device *adev);
-
-#if defined(CONFIG_DEBUG_FS)
-int amdgpu_debugfs_init(struct drm_minor *minor);
-#endif
-
-int amdgpu_debugfs_firmware_init(struct amdgpu_device *adev);
 
 /*
  * amdgpu smumgr functions
@@ -1410,9 +1387,6 @@ struct amdgpu_fw_vram_usage {
 	void *va;
 };
 
-int amdgpu_fw_reserve_vram_init(struct amdgpu_device *adev);
-void amdgpu_fw_reserve_vram_fini(struct amdgpu_device *adev);
-
 /*
  * CGS
  */
@@ -1428,6 +1402,80 @@ typedef void (*amdgpu_wreg_t)(struct amdgpu_device*, uint32_t, uint32_t);
 typedef uint32_t (*amdgpu_block_rreg_t)(struct amdgpu_device*, uint32_t, uint32_t);
 typedef void (*amdgpu_block_wreg_t)(struct amdgpu_device*, uint32_t, uint32_t, uint32_t);
 
+
+/*
+ * amdgpu nbio functions
+ *
+ */
+struct nbio_hdp_flush_reg {
+	u32 ref_and_mask_cp0;
+	u32 ref_and_mask_cp1;
+	u32 ref_and_mask_cp2;
+	u32 ref_and_mask_cp3;
+	u32 ref_and_mask_cp4;
+	u32 ref_and_mask_cp5;
+	u32 ref_and_mask_cp6;
+	u32 ref_and_mask_cp7;
+	u32 ref_and_mask_cp8;
+	u32 ref_and_mask_cp9;
+	u32 ref_and_mask_sdma0;
+	u32 ref_and_mask_sdma1;
+};
+
+struct amdgpu_nbio_funcs {
+	const struct nbio_hdp_flush_reg *hdp_flush_reg;
+	u32 (*get_hdp_flush_req_offset)(struct amdgpu_device *adev);
+	u32 (*get_hdp_flush_done_offset)(struct amdgpu_device *adev);
+	u32 (*get_pcie_index_offset)(struct amdgpu_device *adev);
+	u32 (*get_pcie_data_offset)(struct amdgpu_device *adev);
+	u32 (*get_rev_id)(struct amdgpu_device *adev);
+	void (*mc_access_enable)(struct amdgpu_device *adev, bool enable);
+	void (*hdp_flush)(struct amdgpu_device *adev);
+	u32 (*get_memsize)(struct amdgpu_device *adev);
+	void (*sdma_doorbell_range)(struct amdgpu_device *adev, int instance,
+				    bool use_doorbell, int doorbell_index);
+	void (*enable_doorbell_aperture)(struct amdgpu_device *adev,
+					 bool enable);
+	void (*enable_doorbell_selfring_aperture)(struct amdgpu_device *adev,
+						  bool enable);
+	void (*ih_doorbell_range)(struct amdgpu_device *adev,
+				  bool use_doorbell, int doorbell_index);
+	void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev,
+						 bool enable);
+	void (*update_medium_grain_light_sleep)(struct amdgpu_device *adev,
+						bool enable);
+	void (*get_clockgating_state)(struct amdgpu_device *adev,
+				      u32 *flags);
+	void (*ih_control)(struct amdgpu_device *adev);
+	void (*init_registers)(struct amdgpu_device *adev);
+	void (*detect_hw_virt)(struct amdgpu_device *adev);
+};
+
+
+/* Define the HW IP blocks will be used in driver , add more if necessary */
+enum amd_hw_ip_block_type {
+	GC_HWIP = 1,
+	HDP_HWIP,
+	SDMA0_HWIP,
+	SDMA1_HWIP,
+	MMHUB_HWIP,
+	ATHUB_HWIP,
+	NBIO_HWIP,
+	MP0_HWIP,
+	UVD_HWIP,
+	VCN_HWIP = UVD_HWIP,
+	VCE_HWIP,
+	DF_HWIP,
+	DCE_HWIP,
+	OSSSYS_HWIP,
+	SMUIO_HWIP,
+	PWR_HWIP,
+	NBIF_HWIP,
+	MAX_HWIP
+};
+
+#define HWIP_MAX_INSTANCE	6
+
 struct amd_powerplay {
 	struct cgs_device *cgs_device;
 	void *pp_handle;
@@ -1620,6 +1668,11 @@ struct amdgpu_device {
 	/* amdkfd interface */
 	struct kfd_dev          *kfd;
 
+	/* soc15 register offset based on ip, instance and  segment */
+	uint32_t 		*reg_offset[MAX_HWIP][HWIP_MAX_INSTANCE];
+
+	const struct amdgpu_nbio_funcs	*nbio_funcs;
+
 	/* delayed work_func for deferring clockgating during resume */
 	struct delayed_work     late_init_work;
 
@@ -1785,7 +1838,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_asic_get_config_memsize(adev) (adev)->asic_funcs->get_config_memsize((adev))
 #define amdgpu_gart_flush_gpu_tlb(adev, vmid) (adev)->gart.gart_funcs->flush_gpu_tlb((adev), (vmid))
 #define amdgpu_gart_set_pte_pde(adev, pt, idx, addr, flags) (adev)->gart.gart_funcs->set_pte_pde((adev), (pt), (idx), (addr), (flags))
-#define amdgpu_gart_get_vm_pde(adev, addr) (adev)->gart.gart_funcs->get_vm_pde((adev), (addr))
+#define amdgpu_gart_get_vm_pde(adev, level, dst, flags) (adev)->gart.gart_funcs->get_vm_pde((adev), (level), (dst), (flags))
 #define amdgpu_vm_copy_pte(adev, ib, pe, src, count) ((adev)->vm_manager.vm_pte_funcs->copy_pte((ib), (pe), (src), (count)))
 #define amdgpu_vm_write_pte(adev, ib, pe, value, count, incr) ((adev)->vm_manager.vm_pte_funcs->write_pte((ib), (pe), (value), (count), (incr)))
 #define amdgpu_vm_set_pte_pde(adev, ib, pe, addr, count, incr, flags) ((adev)->vm_manager.vm_pte_funcs->set_pte_pde((ib), (pe), (addr), (count), (incr), (flags)))
@@ -1835,23 +1888,25 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_psp_check_fw_loading_status(adev, i) (adev)->firmware.funcs->check_fw_loading_status((adev), (i))
 
 /* Common functions */
-int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job* job);
-bool amdgpu_need_backup(struct amdgpu_device *adev);
-void amdgpu_pci_config_reset(struct amdgpu_device *adev);
-bool amdgpu_need_post(struct amdgpu_device *adev);
+int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
+			      struct amdgpu_job* job, bool force);
+void amdgpu_device_pci_config_reset(struct amdgpu_device *adev);
+bool amdgpu_device_need_post(struct amdgpu_device *adev);
 void amdgpu_update_display_priority(struct amdgpu_device *adev);
 
 void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes,
 				  u64 num_vis_bytes);
 void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain);
 bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo);
-void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 base);
-void amdgpu_gart_location(struct amdgpu_device *adev, struct amdgpu_mc *mc);
+void amdgpu_device_vram_location(struct amdgpu_device *adev,
+				 struct amdgpu_mc *mc, u64 base);
+void amdgpu_device_gart_location(struct amdgpu_device *adev,
+				 struct amdgpu_mc *mc);
 int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev);
 void amdgpu_ttm_set_active_vram_size(struct amdgpu_device *adev, u64 size);
 int amdgpu_ttm_init(struct amdgpu_device *adev);
 void amdgpu_ttm_fini(struct amdgpu_device *adev);
-void amdgpu_program_register_sequence(struct amdgpu_device *adev,
+void amdgpu_device_program_register_sequence(struct amdgpu_device *adev,
 					     const u32 *registers,
 					     const u32 array_size);
 
@@ -1885,7 +1940,7 @@ void amdgpu_driver_lastclose_kms(struct drm_device *dev);
 int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv);
 void amdgpu_driver_postclose_kms(struct drm_device *dev,
 				 struct drm_file *file_priv);
-int amdgpu_suspend(struct amdgpu_device *adev);
+int amdgpu_device_ip_suspend(struct amdgpu_device *adev);
 int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon);
 int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon);
 u32 amdgpu_get_vblank_counter_kms(struct drm_device *dev, unsigned int pipe);

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

@@ -277,7 +277,7 @@ static int acp_hw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	const struct amdgpu_ip_block *ip_block =
-		amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_ACP);
+		amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_ACP);
 
 	if (!ip_block)
 		return -EINVAL;

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

@@ -93,6 +93,39 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
 				   adev->pdev, kfd2kgd);
 }
 
+/**
+ * amdgpu_doorbell_get_kfd_info - Report doorbell configuration required to
+ *                                setup amdkfd
+ *
+ * @adev: amdgpu_device pointer
+ * @aperture_base: output returning doorbell aperture base physical address
+ * @aperture_size: output returning doorbell aperture size in bytes
+ * @start_offset: output returning # of doorbell bytes reserved for amdgpu.
+ *
+ * amdgpu and amdkfd share the doorbell aperture. amdgpu sets it up,
+ * takes doorbells required for its own rings and reports the setup to amdkfd.
+ * amdgpu reserved doorbells are at the start of the doorbell aperture.
+ */
+static void amdgpu_doorbell_get_kfd_info(struct amdgpu_device *adev,
+					 phys_addr_t *aperture_base,
+					 size_t *aperture_size,
+					 size_t *start_offset)
+{
+	/*
+	 * The first num_doorbells are used by amdgpu.
+	 * amdkfd takes whatever's left in the aperture.
+	 */
+	if (adev->doorbell.size > adev->doorbell.num_doorbells * sizeof(u32)) {
+		*aperture_base = adev->doorbell.base;
+		*aperture_size = adev->doorbell.size;
+		*start_offset = adev->doorbell.num_doorbells * sizeof(u32);
+	} else {
+		*aperture_base = 0;
+		*aperture_size = 0;
+		*start_offset = 0;
+	}
+}
+
 void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 {
 	int i;

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

@@ -27,6 +27,7 @@
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
 #include "amdgpu_atombios.h"
+#include "amdgpu_atomfirmware.h"
 #include "amdgpu_i2c.h"
 
 #include "atom.h"
@@ -1699,7 +1700,7 @@ void amdgpu_atombios_scratch_regs_lock(struct amdgpu_device *adev, bool lock)
 	WREG32(adev->bios_scratch_reg_offset + 6, bios_6_scratch);
 }
 
-void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev)
+static void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev)
 {
 	uint32_t bios_2_scratch, bios_6_scratch;
 
@@ -1721,28 +1722,6 @@ void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev)
 	WREG32(adev->bios_scratch_reg_offset + 6, bios_6_scratch);
 }
 
-void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev)
-{
-	int i;
-
-	for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++)
-		adev->bios_scratch[i] = RREG32(adev->bios_scratch_reg_offset + i);
-}
-
-void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev)
-{
-	int i;
-
-	/*
-	 * VBIOS will check ASIC_INIT_COMPLETE bit to decide if
-	 * execute ASIC_Init posting via driver
-	 */
-	adev->bios_scratch[7] &= ~ATOM_S7_ASIC_INIT_COMPLETE_MASK;
-
-	for (i = 0; i < AMDGPU_BIOS_NUM_SCRATCH; i++)
-		WREG32(adev->bios_scratch_reg_offset + i, adev->bios_scratch[i]);
-}
-
 void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev,
 					      bool hung)
 {
@@ -1798,7 +1777,7 @@ void amdgpu_atombios_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le)
 #endif
 }
 
-int amdgpu_atombios_allocate_fb_scratch(struct amdgpu_device *adev)
+static int amdgpu_atombios_allocate_fb_scratch(struct amdgpu_device *adev)
 {
 	struct atom_context *ctx = adev->mode_info.atom_context;
 	int index = GetIndexIntoMasterTable(DATA, VRAM_UsageByFirmware);
@@ -1841,3 +1820,234 @@ int amdgpu_atombios_allocate_fb_scratch(struct amdgpu_device *adev)
 	ctx->scratch_size_bytes = usage_bytes;
 	return 0;
 }
+
+/* ATOM accessor methods */
+/*
+ * ATOM is an interpreted byte code stored in tables in the vbios.  The
+ * driver registers callbacks to access registers and the interpreter
+ * in the driver parses the tables and executes then to program specific
+ * actions (set display modes, asic init, etc.).  See amdgpu_atombios.c,
+ * atombios.h, and atom.c
+ */
+
+/**
+ * cail_pll_read - read PLL register
+ *
+ * @info: atom card_info pointer
+ * @reg: PLL register offset
+ *
+ * Provides a PLL register accessor for the atom interpreter (r4xx+).
+ * Returns the value of the PLL register.
+ */
+static uint32_t cail_pll_read(struct card_info *info, uint32_t reg)
+{
+	return 0;
+}
+
+/**
+ * cail_pll_write - write PLL register
+ *
+ * @info: atom card_info pointer
+ * @reg: PLL register offset
+ * @val: value to write to the pll register
+ *
+ * Provides a PLL register accessor for the atom interpreter (r4xx+).
+ */
+static void cail_pll_write(struct card_info *info, uint32_t reg, uint32_t val)
+{
+
+}
+
+/**
+ * cail_mc_read - read MC (Memory Controller) register
+ *
+ * @info: atom card_info pointer
+ * @reg: MC register offset
+ *
+ * Provides an MC register accessor for the atom interpreter (r4xx+).
+ * Returns the value of the MC register.
+ */
+static uint32_t cail_mc_read(struct card_info *info, uint32_t reg)
+{
+	return 0;
+}
+
+/**
+ * cail_mc_write - write MC (Memory Controller) register
+ *
+ * @info: atom card_info pointer
+ * @reg: MC register offset
+ * @val: value to write to the pll register
+ *
+ * Provides a MC register accessor for the atom interpreter (r4xx+).
+ */
+static void cail_mc_write(struct card_info *info, uint32_t reg, uint32_t val)
+{
+
+}
+
+/**
+ * cail_reg_write - write MMIO register
+ *
+ * @info: atom card_info pointer
+ * @reg: MMIO register offset
+ * @val: value to write to the pll register
+ *
+ * Provides a MMIO register accessor for the atom interpreter (r4xx+).
+ */
+static void cail_reg_write(struct card_info *info, uint32_t reg, uint32_t val)
+{
+	struct amdgpu_device *adev = info->dev->dev_private;
+
+	WREG32(reg, val);
+}
+
+/**
+ * cail_reg_read - read MMIO register
+ *
+ * @info: atom card_info pointer
+ * @reg: MMIO register offset
+ *
+ * Provides an MMIO register accessor for the atom interpreter (r4xx+).
+ * Returns the value of the MMIO register.
+ */
+static uint32_t cail_reg_read(struct card_info *info, uint32_t reg)
+{
+	struct amdgpu_device *adev = info->dev->dev_private;
+	uint32_t r;
+
+	r = RREG32(reg);
+	return r;
+}
+
+/**
+ * cail_ioreg_write - write IO register
+ *
+ * @info: atom card_info pointer
+ * @reg: IO register offset
+ * @val: value to write to the pll register
+ *
+ * Provides a IO register accessor for the atom interpreter (r4xx+).
+ */
+static void cail_ioreg_write(struct card_info *info, uint32_t reg, uint32_t val)
+{
+	struct amdgpu_device *adev = info->dev->dev_private;
+
+	WREG32_IO(reg, val);
+}
+
+/**
+ * cail_ioreg_read - read IO register
+ *
+ * @info: atom card_info pointer
+ * @reg: IO register offset
+ *
+ * Provides an IO register accessor for the atom interpreter (r4xx+).
+ * Returns the value of the IO register.
+ */
+static uint32_t cail_ioreg_read(struct card_info *info, uint32_t reg)
+{
+	struct amdgpu_device *adev = info->dev->dev_private;
+	uint32_t r;
+
+	r = RREG32_IO(reg);
+	return r;
+}
+
+static ssize_t amdgpu_atombios_get_vbios_version(struct device *dev,
+						 struct device_attribute *attr,
+						 char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+	struct atom_context *ctx = adev->mode_info.atom_context;
+
+	return snprintf(buf, PAGE_SIZE, "%s\n", ctx->vbios_version);
+}
+
+static DEVICE_ATTR(vbios_version, 0444, amdgpu_atombios_get_vbios_version,
+		   NULL);
+
+/**
+ * amdgpu_atombios_fini - free the driver info and callbacks for atombios
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Frees the driver info and register access callbacks for the ATOM
+ * interpreter (r4xx+).
+ * Called at driver shutdown.
+ */
+void amdgpu_atombios_fini(struct amdgpu_device *adev)
+{
+	if (adev->mode_info.atom_context) {
+		kfree(adev->mode_info.atom_context->scratch);
+		kfree(adev->mode_info.atom_context->iio);
+	}
+	kfree(adev->mode_info.atom_context);
+	adev->mode_info.atom_context = NULL;
+	kfree(adev->mode_info.atom_card_info);
+	adev->mode_info.atom_card_info = NULL;
+	device_remove_file(adev->dev, &dev_attr_vbios_version);
+}
+
+/**
+ * amdgpu_atombios_init - init the driver info and callbacks for atombios
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Initializes the driver info and register access callbacks for the
+ * ATOM interpreter (r4xx+).
+ * Returns 0 on sucess, -ENOMEM on failure.
+ * Called at driver startup.
+ */
+int amdgpu_atombios_init(struct amdgpu_device *adev)
+{
+	struct card_info *atom_card_info =
+	    kzalloc(sizeof(struct card_info), GFP_KERNEL);
+	int ret;
+
+	if (!atom_card_info)
+		return -ENOMEM;
+
+	adev->mode_info.atom_card_info = atom_card_info;
+	atom_card_info->dev = adev->ddev;
+	atom_card_info->reg_read = cail_reg_read;
+	atom_card_info->reg_write = cail_reg_write;
+	/* needed for iio ops */
+	if (adev->rio_mem) {
+		atom_card_info->ioreg_read = cail_ioreg_read;
+		atom_card_info->ioreg_write = cail_ioreg_write;
+	} else {
+		DRM_DEBUG("PCI I/O BAR is not found. Using MMIO to access ATOM BIOS\n");
+		atom_card_info->ioreg_read = cail_reg_read;
+		atom_card_info->ioreg_write = cail_reg_write;
+	}
+	atom_card_info->mc_read = cail_mc_read;
+	atom_card_info->mc_write = cail_mc_write;
+	atom_card_info->pll_read = cail_pll_read;
+	atom_card_info->pll_write = cail_pll_write;
+
+	adev->mode_info.atom_context = amdgpu_atom_parse(atom_card_info, adev->bios);
+	if (!adev->mode_info.atom_context) {
+		amdgpu_atombios_fini(adev);
+		return -ENOMEM;
+	}
+
+	mutex_init(&adev->mode_info.atom_context->mutex);
+	if (adev->is_atom_fw) {
+		amdgpu_atomfirmware_scratch_regs_init(adev);
+		amdgpu_atomfirmware_allocate_fb_scratch(adev);
+	} else {
+		amdgpu_atombios_scratch_regs_init(adev);
+		amdgpu_atombios_allocate_fb_scratch(adev);
+	}
+
+	ret = device_create_file(adev->dev, &dev_attr_vbios_version);
+	if (ret) {
+		DRM_ERROR("Failed to create device file for VBIOS version\n");
+		return ret;
+	}
+
+	return 0;
+}
+

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

@@ -195,9 +195,6 @@ int amdgpu_atombios_init_mc_reg_table(struct amdgpu_device *adev,
 bool amdgpu_atombios_has_gpu_virtualization_table(struct amdgpu_device *adev);
 
 void amdgpu_atombios_scratch_regs_lock(struct amdgpu_device *adev, bool lock);
-void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev);
-void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev);
-void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev);
 void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev,
 					      bool hung);
 bool amdgpu_atombios_scratch_need_asic_init(struct amdgpu_device *adev);
@@ -219,6 +216,7 @@ int amdgpu_atombios_get_svi2_info(struct amdgpu_device *adev,
 			      u8 voltage_type,
 			      u8 *svd_gpio_id, u8 *svc_gpio_id);
 
-int amdgpu_atombios_allocate_fb_scratch(struct amdgpu_device *adev);
+void amdgpu_atombios_fini(struct amdgpu_device *adev);
+int amdgpu_atombios_init(struct amdgpu_device *adev);
 
 #endif

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

@@ -93,7 +93,7 @@ static bool igp_read_bios_from_vram(struct amdgpu_device *adev)
 	resource_size_t size = 256 * 1024; /* ??? */
 
 	if (!(adev->flags & AMD_IS_APU))
-		if (amdgpu_need_post(adev))
+		if (amdgpu_device_need_post(adev))
 			return false;
 
 	adev->bios = NULL;

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

@@ -801,6 +801,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 				else
 					strcpy(fw_name, "amdgpu/vega10_smc.bin");
 				break;
+			case CHIP_RAVEN:
+				adev->pm.fw_version = info->version;
+				return 0;
 			default:
 				DRM_ERROR("SMC firmware not supported\n");
 				return -EINVAL;

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

@@ -343,7 +343,12 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
 				 struct amdgpu_bo *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	struct ttm_operation_ctx ctx = { true, false };
+	struct ttm_operation_ctx ctx = {
+		.interruptible = true,
+		.no_wait_gpu = false,
+		.allow_reserved_eviction = false,
+		.resv = bo->tbo.resv
+	};
 	uint32_t domain;
 	int r;
 
@@ -1150,7 +1155,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 			    union drm_amdgpu_cs *cs)
 {
 	struct amdgpu_ring *ring = p->job->ring;
-	struct amd_sched_entity *entity = &p->ctx->rings[ring->idx].entity;
+	struct drm_sched_entity *entity = &p->ctx->rings[ring->idx].entity;
 	struct amdgpu_job *job;
 	unsigned i;
 	uint64_t seq;
@@ -1173,7 +1178,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	job = p->job;
 	p->job = NULL;
 
-	r = amd_sched_job_init(&job->base, &ring->sched, entity, p->filp);
+	r = drm_sched_job_init(&job->base, &ring->sched, entity, p->filp);
 	if (r) {
 		amdgpu_job_free(job);
 		amdgpu_mn_unlock(p->mn);
@@ -1202,7 +1207,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	amdgpu_ring_priority_get(job->ring, job->base.s_priority);
 
 	trace_amdgpu_cs_ioctl(job);
-	amd_sched_entity_push_job(&job->base, entity);
+	drm_sched_entity_push_job(&job->base, entity);
 
 	ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence);
 	amdgpu_mn_unlock(p->mn);

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

@@ -28,10 +28,10 @@
 #include "amdgpu_sched.h"
 
 static int amdgpu_ctx_priority_permit(struct drm_file *filp,
-				      enum amd_sched_priority priority)
+				      enum drm_sched_priority priority)
 {
 	/* NORMAL and below are accessible by everyone */
-	if (priority <= AMD_SCHED_PRIORITY_NORMAL)
+	if (priority <= DRM_SCHED_PRIORITY_NORMAL)
 		return 0;
 
 	if (capable(CAP_SYS_NICE))
@@ -44,14 +44,14 @@ static int amdgpu_ctx_priority_permit(struct drm_file *filp,
 }
 
 static int amdgpu_ctx_init(struct amdgpu_device *adev,
-			   enum amd_sched_priority priority,
+			   enum drm_sched_priority priority,
 			   struct drm_file *filp,
 			   struct amdgpu_ctx *ctx)
 {
 	unsigned i, j;
 	int r;
 
-	if (priority < 0 || priority >= AMD_SCHED_PRIORITY_MAX)
+	if (priority < 0 || priority >= DRM_SCHED_PRIORITY_MAX)
 		return -EINVAL;
 
 	r = amdgpu_ctx_priority_permit(filp, priority);
@@ -78,19 +78,19 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 	ctx->reset_counter_query = ctx->reset_counter;
 	ctx->vram_lost_counter = atomic_read(&adev->vram_lost_counter);
 	ctx->init_priority = priority;
-	ctx->override_priority = AMD_SCHED_PRIORITY_UNSET;
+	ctx->override_priority = DRM_SCHED_PRIORITY_UNSET;
 
 	/* create context entity for each ring */
 	for (i = 0; i < adev->num_rings; i++) {
 		struct amdgpu_ring *ring = adev->rings[i];
-		struct amd_sched_rq *rq;
+		struct drm_sched_rq *rq;
 
 		rq = &ring->sched.sched_rq[priority];
 
 		if (ring == &adev->gfx.kiq.ring)
 			continue;
 
-		r = amd_sched_entity_init(&ring->sched, &ctx->rings[i].entity,
+		r = drm_sched_entity_init(&ring->sched, &ctx->rings[i].entity,
 					  rq, amdgpu_sched_jobs, &ctx->guilty);
 		if (r)
 			goto failed;
@@ -104,7 +104,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 
 failed:
 	for (j = 0; j < i; j++)
-		amd_sched_entity_fini(&adev->rings[j]->sched,
+		drm_sched_entity_fini(&adev->rings[j]->sched,
 				      &ctx->rings[j].entity);
 	kfree(ctx->fences);
 	ctx->fences = NULL;
@@ -126,7 +126,7 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
 	ctx->fences = NULL;
 
 	for (i = 0; i < adev->num_rings; i++)
-		amd_sched_entity_fini(&adev->rings[i]->sched,
+		drm_sched_entity_fini(&adev->rings[i]->sched,
 				      &ctx->rings[i].entity);
 
 	amdgpu_queue_mgr_fini(adev, &ctx->queue_mgr);
@@ -137,7 +137,7 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
 static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
 			    struct amdgpu_fpriv *fpriv,
 			    struct drm_file *filp,
-			    enum amd_sched_priority priority,
+			    enum drm_sched_priority priority,
 			    uint32_t *id)
 {
 	struct amdgpu_ctx_mgr *mgr = &fpriv->ctx_mgr;
@@ -266,7 +266,7 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 {
 	int r;
 	uint32_t id;
-	enum amd_sched_priority priority;
+	enum drm_sched_priority priority;
 
 	union drm_amdgpu_ctx *args = data;
 	struct amdgpu_device *adev = dev->dev_private;
@@ -278,8 +278,8 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 
 	/* For backwards compatibility reasons, we need to accept
 	 * ioctls with garbage in the priority field */
-	if (priority == AMD_SCHED_PRIORITY_INVALID)
-		priority = AMD_SCHED_PRIORITY_NORMAL;
+	if (priority == DRM_SCHED_PRIORITY_INVALID)
+		priority = DRM_SCHED_PRIORITY_NORMAL;
 
 	switch (args->in.op) {
 	case AMDGPU_CTX_OP_ALLOC_CTX:
@@ -385,18 +385,18 @@ struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
 }
 
 void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx,
-				  enum amd_sched_priority priority)
+				  enum drm_sched_priority priority)
 {
 	int i;
 	struct amdgpu_device *adev = ctx->adev;
-	struct amd_sched_rq *rq;
-	struct amd_sched_entity *entity;
+	struct drm_sched_rq *rq;
+	struct drm_sched_entity *entity;
 	struct amdgpu_ring *ring;
-	enum amd_sched_priority ctx_prio;
+	enum drm_sched_priority ctx_prio;
 
 	ctx->override_priority = priority;
 
-	ctx_prio = (ctx->override_priority == AMD_SCHED_PRIORITY_UNSET) ?
+	ctx_prio = (ctx->override_priority == DRM_SCHED_PRIORITY_UNSET) ?
 			ctx->init_priority : ctx->override_priority;
 
 	for (i = 0; i < adev->num_rings; i++) {
@@ -407,7 +407,7 @@ void amdgpu_ctx_priority_override(struct amdgpu_ctx *ctx,
 		if (ring->funcs->type == AMDGPU_RING_TYPE_KIQ)
 			continue;
 
-		amd_sched_entity_set_rq(entity, rq);
+		drm_sched_entity_set_rq(entity, rq);
 	}
 }
 

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

@@ -0,0 +1,792 @@
+/*
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
+ *
+ * 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.
+ *
+ */
+
+#include <linux/kthread.h>
+#include <drm/drmP.h>
+#include <linux/debugfs.h>
+#include "amdgpu.h"
+
+/*
+ * Debugfs
+ */
+int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
+			     const struct drm_info_list *files,
+			     unsigned nfiles)
+{
+	unsigned i;
+
+	for (i = 0; i < adev->debugfs_count; i++) {
+		if (adev->debugfs[i].files == files) {
+			/* Already registered */
+			return 0;
+		}
+	}
+
+	i = adev->debugfs_count + 1;
+	if (i > AMDGPU_DEBUGFS_MAX_COMPONENTS) {
+		DRM_ERROR("Reached maximum number of debugfs components.\n");
+		DRM_ERROR("Report so we increase "
+			  "AMDGPU_DEBUGFS_MAX_COMPONENTS.\n");
+		return -EINVAL;
+	}
+	adev->debugfs[adev->debugfs_count].files = files;
+	adev->debugfs[adev->debugfs_count].num_files = nfiles;
+	adev->debugfs_count = i;
+#if defined(CONFIG_DEBUG_FS)
+	drm_debugfs_create_files(files, nfiles,
+				 adev->ddev->primary->debugfs_root,
+				 adev->ddev->primary);
+#endif
+	return 0;
+}
+
+#if defined(CONFIG_DEBUG_FS)
+
+static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = file_inode(f)->i_private;
+	ssize_t result = 0;
+	int r;
+	bool pm_pg_lock, use_bank;
+	unsigned instance_bank, sh_bank, se_bank;
+
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	/* are we reading registers for which a PG lock is necessary? */
+	pm_pg_lock = (*pos >> 23) & 1;
+
+	if (*pos & (1ULL << 62)) {
+		se_bank = (*pos & GENMASK_ULL(33, 24)) >> 24;
+		sh_bank = (*pos & GENMASK_ULL(43, 34)) >> 34;
+		instance_bank = (*pos & GENMASK_ULL(53, 44)) >> 44;
+
+		if (se_bank == 0x3FF)
+			se_bank = 0xFFFFFFFF;
+		if (sh_bank == 0x3FF)
+			sh_bank = 0xFFFFFFFF;
+		if (instance_bank == 0x3FF)
+			instance_bank = 0xFFFFFFFF;
+		use_bank = 1;
+	} else {
+		use_bank = 0;
+	}
+
+	*pos &= (1UL << 22) - 1;
+
+	if (use_bank) {
+		if ((sh_bank != 0xFFFFFFFF && sh_bank >= adev->gfx.config.max_sh_per_se) ||
+		    (se_bank != 0xFFFFFFFF && se_bank >= adev->gfx.config.max_shader_engines))
+			return -EINVAL;
+		mutex_lock(&adev->grbm_idx_mutex);
+		amdgpu_gfx_select_se_sh(adev, se_bank,
+					sh_bank, instance_bank);
+	}
+
+	if (pm_pg_lock)
+		mutex_lock(&adev->pm.mutex);
+
+	while (size) {
+		uint32_t value;
+
+		if (*pos > adev->rmmio_size)
+			goto end;
+
+		value = RREG32(*pos >> 2);
+		r = put_user(value, (uint32_t *)buf);
+		if (r) {
+			result = r;
+			goto end;
+		}
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+end:
+	if (use_bank) {
+		amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+		mutex_unlock(&adev->grbm_idx_mutex);
+	}
+
+	if (pm_pg_lock)
+		mutex_unlock(&adev->pm.mutex);
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_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;
+	bool pm_pg_lock, use_bank;
+	unsigned instance_bank, sh_bank, se_bank;
+
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	/* are we reading registers for which a PG lock is necessary? */
+	pm_pg_lock = (*pos >> 23) & 1;
+
+	if (*pos & (1ULL << 62)) {
+		se_bank = (*pos & GENMASK_ULL(33, 24)) >> 24;
+		sh_bank = (*pos & GENMASK_ULL(43, 34)) >> 34;
+		instance_bank = (*pos & GENMASK_ULL(53, 44)) >> 44;
+
+		if (se_bank == 0x3FF)
+			se_bank = 0xFFFFFFFF;
+		if (sh_bank == 0x3FF)
+			sh_bank = 0xFFFFFFFF;
+		if (instance_bank == 0x3FF)
+			instance_bank = 0xFFFFFFFF;
+		use_bank = 1;
+	} else {
+		use_bank = 0;
+	}
+
+	*pos &= (1UL << 22) - 1;
+
+	if (use_bank) {
+		if ((sh_bank != 0xFFFFFFFF && sh_bank >= adev->gfx.config.max_sh_per_se) ||
+		    (se_bank != 0xFFFFFFFF && se_bank >= adev->gfx.config.max_shader_engines))
+			return -EINVAL;
+		mutex_lock(&adev->grbm_idx_mutex);
+		amdgpu_gfx_select_se_sh(adev, se_bank,
+					sh_bank, instance_bank);
+	}
+
+	if (pm_pg_lock)
+		mutex_lock(&adev->pm.mutex);
+
+	while (size) {
+		uint32_t value;
+
+		if (*pos > adev->rmmio_size)
+			return result;
+
+		r = get_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		WREG32(*pos >> 2, value);
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	if (use_bank) {
+		amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+		mutex_unlock(&adev->grbm_idx_mutex);
+	}
+
+	if (pm_pg_lock)
+		mutex_unlock(&adev->pm.mutex);
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, 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;
+
+	while (size) {
+		uint32_t value;
+
+		value = RREG32_PCIE(*pos >> 2);
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_pcie_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;
+
+	while (size) {
+		uint32_t value;
+
+		r = get_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		WREG32_PCIE(*pos >> 2, value);
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, 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;
+
+	while (size) {
+		uint32_t value;
+
+		value = RREG32_DIDT(*pos >> 2);
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_didt_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;
+
+	while (size) {
+		uint32_t value;
+
+		r = get_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		WREG32_DIDT(*pos >> 2, value);
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, 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;
+
+	while (size) {
+		uint32_t value;
+
+		value = RREG32_SMC(*pos);
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_regs_smc_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;
+
+	while (size) {
+		uint32_t value;
+
+		r = get_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		WREG32_SMC(*pos, value);
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = file_inode(f)->i_private;
+	ssize_t result = 0;
+	int r;
+	uint32_t *config, no_regs = 0;
+
+	if (size & 0x3 || *pos & 0x3)
+		return -EINVAL;
+
+	config = kmalloc_array(256, sizeof(*config), GFP_KERNEL);
+	if (!config)
+		return -ENOMEM;
+
+	/* version, increment each time something is added */
+	config[no_regs++] = 3;
+	config[no_regs++] = adev->gfx.config.max_shader_engines;
+	config[no_regs++] = adev->gfx.config.max_tile_pipes;
+	config[no_regs++] = adev->gfx.config.max_cu_per_sh;
+	config[no_regs++] = adev->gfx.config.max_sh_per_se;
+	config[no_regs++] = adev->gfx.config.max_backends_per_se;
+	config[no_regs++] = adev->gfx.config.max_texture_channel_caches;
+	config[no_regs++] = adev->gfx.config.max_gprs;
+	config[no_regs++] = adev->gfx.config.max_gs_threads;
+	config[no_regs++] = adev->gfx.config.max_hw_contexts;
+	config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_frontend;
+	config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_backend;
+	config[no_regs++] = adev->gfx.config.sc_hiz_tile_fifo_size;
+	config[no_regs++] = adev->gfx.config.sc_earlyz_tile_fifo_size;
+	config[no_regs++] = adev->gfx.config.num_tile_pipes;
+	config[no_regs++] = adev->gfx.config.backend_enable_mask;
+	config[no_regs++] = adev->gfx.config.mem_max_burst_length_bytes;
+	config[no_regs++] = adev->gfx.config.mem_row_size_in_kb;
+	config[no_regs++] = adev->gfx.config.shader_engine_tile_size;
+	config[no_regs++] = adev->gfx.config.num_gpus;
+	config[no_regs++] = adev->gfx.config.multi_gpu_tile_size;
+	config[no_regs++] = adev->gfx.config.mc_arb_ramcfg;
+	config[no_regs++] = adev->gfx.config.gb_addr_config;
+	config[no_regs++] = adev->gfx.config.num_rbs;
+
+	/* rev==1 */
+	config[no_regs++] = adev->rev_id;
+	config[no_regs++] = adev->pg_flags;
+	config[no_regs++] = adev->cg_flags;
+
+	/* rev==2 */
+	config[no_regs++] = adev->family;
+	config[no_regs++] = adev->external_rev_id;
+
+	/* rev==3 */
+	config[no_regs++] = adev->pdev->device;
+	config[no_regs++] = adev->pdev->revision;
+	config[no_regs++] = adev->pdev->subsystem_device;
+	config[no_regs++] = adev->pdev->subsystem_vendor;
+
+	while (size && (*pos < no_regs * 4)) {
+		uint32_t value;
+
+		value = config[*pos >> 2];
+		r = put_user(value, (uint32_t *)buf);
+		if (r) {
+			kfree(config);
+			return r;
+		}
+
+		result += 4;
+		buf += 4;
+		*pos += 4;
+		size -= 4;
+	}
+
+	kfree(config);
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = file_inode(f)->i_private;
+	int idx, x, outsize, r, valuesize;
+	uint32_t values[16];
+
+	if (size & 3 || *pos & 0x3)
+		return -EINVAL;
+
+	if (amdgpu_dpm == 0)
+		return -EINVAL;
+
+	/* convert offset to sensor number */
+	idx = *pos >> 2;
+
+	valuesize = sizeof(values);
+	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor)
+		r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize);
+	else
+		return -EINVAL;
+
+	if (size > valuesize)
+		return -EINVAL;
+
+	outsize = 0;
+	x = 0;
+	if (!r) {
+		while (size) {
+			r = put_user(values[x++], (int32_t *)buf);
+			buf += 4;
+			size -= 4;
+			outsize += 4;
+		}
+	}
+
+	return !r ? outsize : r;
+}
+
+static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = f->f_inode->i_private;
+	int r, x;
+	ssize_t result=0;
+	uint32_t offset, se, sh, cu, wave, simd, data[32];
+
+	if (size & 3 || *pos & 3)
+		return -EINVAL;
+
+	/* decode offset */
+	offset = (*pos & GENMASK_ULL(6, 0));
+	se = (*pos & GENMASK_ULL(14, 7)) >> 7;
+	sh = (*pos & GENMASK_ULL(22, 15)) >> 15;
+	cu = (*pos & GENMASK_ULL(30, 23)) >> 23;
+	wave = (*pos & GENMASK_ULL(36, 31)) >> 31;
+	simd = (*pos & GENMASK_ULL(44, 37)) >> 37;
+
+	/* switch to the specific se/sh/cu */
+	mutex_lock(&adev->grbm_idx_mutex);
+	amdgpu_gfx_select_se_sh(adev, se, sh, cu);
+
+	x = 0;
+	if (adev->gfx.funcs->read_wave_data)
+		adev->gfx.funcs->read_wave_data(adev, simd, wave, data, &x);
+
+	amdgpu_gfx_select_se_sh(adev, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF);
+	mutex_unlock(&adev->grbm_idx_mutex);
+
+	if (!x)
+		return -EINVAL;
+
+	while (size && (offset < x * 4)) {
+		uint32_t value;
+
+		value = data[offset >> 2];
+		r = put_user(value, (uint32_t *)buf);
+		if (r)
+			return r;
+
+		result += 4;
+		buf += 4;
+		offset += 4;
+		size -= 4;
+	}
+
+	return result;
+}
+
+static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
+					size_t size, loff_t *pos)
+{
+	struct amdgpu_device *adev = f->f_inode->i_private;
+	int r;
+	ssize_t result = 0;
+	uint32_t offset, se, sh, cu, wave, simd, thread, bank, *data;
+
+	if (size & 3 || *pos & 3)
+		return -EINVAL;
+
+	/* decode offset */
+	offset = *pos & GENMASK_ULL(11, 0);
+	se = (*pos & GENMASK_ULL(19, 12)) >> 12;
+	sh = (*pos & GENMASK_ULL(27, 20)) >> 20;
+	cu = (*pos & GENMASK_ULL(35, 28)) >> 28;
+	wave = (*pos & GENMASK_ULL(43, 36)) >> 36;
+	simd = (*pos & GENMASK_ULL(51, 44)) >> 44;
+	thread = (*pos & GENMASK_ULL(59, 52)) >> 52;
+	bank = (*pos & GENMASK_ULL(61, 60)) >> 60;
+
+	data = kmalloc_array(1024, sizeof(*data), GFP_KERNEL);
+	if (!data)
+		return -ENOMEM;
+
+	/* switch to the specific se/sh/cu */
+	mutex_lock(&adev->grbm_idx_mutex);
+	amdgpu_gfx_select_se_sh(adev, se, sh, cu);
+
+	if (bank == 0) {
+		if (adev->gfx.funcs->read_wave_vgprs)
+			adev->gfx.funcs->read_wave_vgprs(adev, simd, wave, thread, offset, size>>2, data);
+	} else {
+		if (adev->gfx.funcs->read_wave_sgprs)
+			adev->gfx.funcs->read_wave_sgprs(adev, simd, wave, offset, size>>2, data);
+	}
+
+	amdgpu_gfx_select_se_sh(adev, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF);
+	mutex_unlock(&adev->grbm_idx_mutex);
+
+	while (size) {
+		uint32_t value;
+
+		value = data[offset++];
+		r = put_user(value, (uint32_t *)buf);
+		if (r) {
+			result = r;
+			goto err;
+		}
+
+		result += 4;
+		buf += 4;
+		size -= 4;
+	}
+
+err:
+	kfree(data);
+	return result;
+}
+
+static const struct file_operations amdgpu_debugfs_regs_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_regs_read,
+	.write = amdgpu_debugfs_regs_write,
+	.llseek = default_llseek
+};
+static const struct file_operations amdgpu_debugfs_regs_didt_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_regs_didt_read,
+	.write = amdgpu_debugfs_regs_didt_write,
+	.llseek = default_llseek
+};
+static const struct file_operations amdgpu_debugfs_regs_pcie_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_regs_pcie_read,
+	.write = amdgpu_debugfs_regs_pcie_write,
+	.llseek = default_llseek
+};
+static const struct file_operations amdgpu_debugfs_regs_smc_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_regs_smc_read,
+	.write = amdgpu_debugfs_regs_smc_write,
+	.llseek = default_llseek
+};
+
+static const struct file_operations amdgpu_debugfs_gca_config_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_gca_config_read,
+	.llseek = default_llseek
+};
+
+static const struct file_operations amdgpu_debugfs_sensors_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_sensor_read,
+	.llseek = default_llseek
+};
+
+static const struct file_operations amdgpu_debugfs_wave_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_wave_read,
+	.llseek = default_llseek
+};
+static const struct file_operations amdgpu_debugfs_gpr_fops = {
+	.owner = THIS_MODULE,
+	.read = amdgpu_debugfs_gpr_read,
+	.llseek = default_llseek
+};
+
+static const struct file_operations *debugfs_regs[] = {
+	&amdgpu_debugfs_regs_fops,
+	&amdgpu_debugfs_regs_didt_fops,
+	&amdgpu_debugfs_regs_pcie_fops,
+	&amdgpu_debugfs_regs_smc_fops,
+	&amdgpu_debugfs_gca_config_fops,
+	&amdgpu_debugfs_sensors_fops,
+	&amdgpu_debugfs_wave_fops,
+	&amdgpu_debugfs_gpr_fops,
+};
+
+static const char *debugfs_regs_names[] = {
+	"amdgpu_regs",
+	"amdgpu_regs_didt",
+	"amdgpu_regs_pcie",
+	"amdgpu_regs_smc",
+	"amdgpu_gca_config",
+	"amdgpu_sensors",
+	"amdgpu_wave",
+	"amdgpu_gpr",
+};
+
+int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
+{
+	struct drm_minor *minor = adev->ddev->primary;
+	struct dentry *ent, *root = minor->debugfs_root;
+	unsigned i, j;
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_regs); i++) {
+		ent = debugfs_create_file(debugfs_regs_names[i],
+					  S_IFREG | S_IRUGO, root,
+					  adev, debugfs_regs[i]);
+		if (IS_ERR(ent)) {
+			for (j = 0; j < i; j++) {
+				debugfs_remove(adev->debugfs_regs[i]);
+				adev->debugfs_regs[i] = NULL;
+			}
+			return PTR_ERR(ent);
+		}
+
+		if (!i)
+			i_size_write(ent->d_inode, adev->rmmio_size);
+		adev->debugfs_regs[i] = ent;
+	}
+
+	return 0;
+}
+
+void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev)
+{
+	unsigned i;
+
+	for (i = 0; i < ARRAY_SIZE(debugfs_regs); i++) {
+		if (adev->debugfs_regs[i]) {
+			debugfs_remove(adev->debugfs_regs[i]);
+			adev->debugfs_regs[i] = NULL;
+		}
+	}
+}
+
+static int amdgpu_debugfs_test_ib(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct amdgpu_device *adev = dev->dev_private;
+	int r = 0, i;
+
+	/* hold on the scheduler */
+	for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
+		struct amdgpu_ring *ring = adev->rings[i];
+
+		if (!ring || !ring->sched.thread)
+			continue;
+		kthread_park(ring->sched.thread);
+	}
+
+	seq_printf(m, "run ib test:\n");
+	r = amdgpu_ib_ring_tests(adev);
+	if (r)
+		seq_printf(m, "ib ring tests failed (%d).\n", r);
+	else
+		seq_printf(m, "ib ring tests passed.\n");
+
+	/* go on the scheduler */
+	for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
+		struct amdgpu_ring *ring = adev->rings[i];
+
+		if (!ring || !ring->sched.thread)
+			continue;
+		kthread_unpark(ring->sched.thread);
+	}
+
+	return 0;
+}
+
+static int amdgpu_debugfs_get_vbios_dump(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *) m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct amdgpu_device *adev = dev->dev_private;
+
+	seq_write(m, adev->bios, adev->bios_size);
+	return 0;
+}
+
+static int amdgpu_debugfs_evict_vram(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct amdgpu_device *adev = dev->dev_private;
+
+	seq_printf(m, "(%d)\n", amdgpu_bo_evict_vram(adev));
+	return 0;
+}
+
+static const struct drm_info_list amdgpu_debugfs_list[] = {
+	{"amdgpu_vbios", amdgpu_debugfs_get_vbios_dump},
+	{"amdgpu_test_ib", &amdgpu_debugfs_test_ib},
+	{"amdgpu_evict_vram", &amdgpu_debugfs_evict_vram}
+};
+
+int amdgpu_debugfs_init(struct amdgpu_device *adev)
+{
+	return amdgpu_debugfs_add_files(adev, amdgpu_debugfs_list,
+					ARRAY_SIZE(amdgpu_debugfs_list));
+}
+
+#else
+int amdgpu_debugfs_init(struct amdgpu_device *adev)
+{
+	return 0;
+}
+int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
+{
+	return 0;
+}
+void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev) { }
+#endif

drivers/gpu/drm/amd/display/dc/basics/grph_object_id.c → drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.h

@@ -1,5 +1,7 @@
 /*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
+ * Copyright 2008 Advanced Micro Devices, Inc.
+ * Copyright 2008 Red Hat Inc.
+ * Copyright 2009 Jerome Glisse.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,57 +21,22 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: AMD
- *
  */
 
-#include "dm_services.h"
-#include "include/grph_object_id.h"
-
-static bool dal_graphics_object_id_is_valid(struct graphics_object_id id)
-{
-	bool rc = true;
-
-	switch (id.type) {
-	case OBJECT_TYPE_UNKNOWN:
-		rc = false;
-		break;
-	case OBJECT_TYPE_GPU:
-	case OBJECT_TYPE_ENGINE:
-		/* do NOT check for id.id == 0 */
-		if (id.enum_id == ENUM_ID_UNKNOWN)
-			rc = false;
-		break;
-	default:
-		if (id.id == 0 || id.enum_id == ENUM_ID_UNKNOWN)
-			rc = false;
-		break;
-	}
-
-	return rc;
-}
-
-bool dal_graphics_object_id_is_equal(
-	struct graphics_object_id id1,
-	struct graphics_object_id id2)
-{
-	if (false == dal_graphics_object_id_is_valid(id1)) {
-		dm_output_to_console(
-		"%s: Warning: comparing invalid object 'id1'!\n", __func__);
-		return false;
-	}
-
-	if (false == dal_graphics_object_id_is_valid(id2)) {
-		dm_output_to_console(
-		"%s: Warning: comparing invalid object 'id2'!\n", __func__);
-		return false;
-	}
-
-	if (id1.id == id2.id && id1.enum_id == id2.enum_id
-		&& id1.type == id2.type)
-		return true;
-
-	return false;
-}
-
-
+/*
+ * Debugfs
+ */
+struct amdgpu_debugfs {
+	const struct drm_info_list	*files;
+	unsigned		num_files;
+};
+
+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);
+int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
+			     const struct drm_info_list *files,
+			     unsigned nfiles);
+int amdgpu_debugfs_fence_init(struct amdgpu_device *adev);
+int amdgpu_debugfs_firmware_init(struct amdgpu_device *adev);
+int amdgpu_debugfs_gem_init(struct amdgpu_device *adev);

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

@@ -28,7 +28,6 @@
 #include <linux/kthread.h>
 #include <linux/console.h>
 #include <linux/slab.h>
-#include <linux/debugfs.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_atomic_helper.h>
@@ -64,11 +63,6 @@ MODULE_FIRMWARE("amdgpu/raven_gpu_info.bin");
 
 #define AMDGPU_RESUME_MS		2000
 
-static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev);
-static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev);
-static int amdgpu_debugfs_test_ib_ring_init(struct amdgpu_device *adev);
-static int amdgpu_debugfs_vbios_dump_init(struct amdgpu_device *adev);
-
 static const char *amdgpu_asic_name[] = {
 	"TAHITI",
 	"PITCAIRN",
@@ -333,7 +327,7 @@ static void amdgpu_block_invalid_wreg(struct amdgpu_device *adev,
 	BUG();
 }
 
-static int amdgpu_vram_scratch_init(struct amdgpu_device *adev)
+static int amdgpu_device_vram_scratch_init(struct amdgpu_device *adev)
 {
 	return amdgpu_bo_create_kernel(adev, AMDGPU_GPU_PAGE_SIZE,
 				       PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
@@ -342,13 +336,13 @@ static int amdgpu_vram_scratch_init(struct amdgpu_device *adev)
 				       (void **)&adev->vram_scratch.ptr);
 }
 
-static void amdgpu_vram_scratch_fini(struct amdgpu_device *adev)
+static void amdgpu_device_vram_scratch_fini(struct amdgpu_device *adev)
 {
 	amdgpu_bo_free_kernel(&adev->vram_scratch.robj, NULL, NULL);
 }
 
 /**
- * amdgpu_program_register_sequence - program an array of registers.
+ * amdgpu_device_program_register_sequence - program an array of registers.
  *
  * @adev: amdgpu_device pointer
  * @registers: pointer to the register array
@@ -357,9 +351,9 @@ static void amdgpu_vram_scratch_fini(struct amdgpu_device *adev)
  * Programs an array or registers with and and or masks.
  * This is a helper for setting golden registers.
  */
-void amdgpu_program_register_sequence(struct amdgpu_device *adev,
-				      const u32 *registers,
-				      const u32 array_size)
+void amdgpu_device_program_register_sequence(struct amdgpu_device *adev,
+					     const u32 *registers,
+					     const u32 array_size)
 {
 	u32 tmp, reg, and_mask, or_mask;
 	int i;
@@ -383,7 +377,7 @@ void amdgpu_program_register_sequence(struct amdgpu_device *adev,
 	}
 }
 
-void amdgpu_pci_config_reset(struct amdgpu_device *adev)
+void amdgpu_device_pci_config_reset(struct amdgpu_device *adev)
 {
 	pci_write_config_dword(adev->pdev, 0x7c, AMDGPU_ASIC_RESET_DATA);
 }
@@ -392,14 +386,14 @@ void amdgpu_pci_config_reset(struct amdgpu_device *adev)
  * GPU doorbell aperture helpers function.
  */
 /**
- * amdgpu_doorbell_init - Init doorbell driver information.
+ * amdgpu_device_doorbell_init - Init doorbell driver information.
  *
  * @adev: amdgpu_device pointer
  *
  * Init doorbell driver information (CIK)
  * Returns 0 on success, error on failure.
  */
-static int amdgpu_doorbell_init(struct amdgpu_device *adev)
+static int amdgpu_device_doorbell_init(struct amdgpu_device *adev)
 {
 	/* No doorbell on SI hardware generation */
 	if (adev->asic_type < CHIP_BONAIRE) {
@@ -432,66 +426,35 @@ static int amdgpu_doorbell_init(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_doorbell_fini - Tear down doorbell driver information.
+ * amdgpu_device_doorbell_fini - Tear down doorbell driver information.
  *
  * @adev: amdgpu_device pointer
  *
  * Tear down doorbell driver information (CIK)
  */
-static void amdgpu_doorbell_fini(struct amdgpu_device *adev)
+static void amdgpu_device_doorbell_fini(struct amdgpu_device *adev)
 {
 	iounmap(adev->doorbell.ptr);
 	adev->doorbell.ptr = NULL;
 }
 
-/**
- * amdgpu_doorbell_get_kfd_info - Report doorbell configuration required to
- *                                setup amdkfd
- *
- * @adev: amdgpu_device pointer
- * @aperture_base: output returning doorbell aperture base physical address
- * @aperture_size: output returning doorbell aperture size in bytes
- * @start_offset: output returning # of doorbell bytes reserved for amdgpu.
- *
- * amdgpu and amdkfd share the doorbell aperture. amdgpu sets it up,
- * takes doorbells required for its own rings and reports the setup to amdkfd.
- * amdgpu reserved doorbells are at the start of the doorbell aperture.
- */
-void amdgpu_doorbell_get_kfd_info(struct amdgpu_device *adev,
-				phys_addr_t *aperture_base,
-				size_t *aperture_size,
-				size_t *start_offset)
-{
-	/*
-	 * The first num_doorbells are used by amdgpu.
-	 * amdkfd takes whatever's left in the aperture.
-	 */
-	if (adev->doorbell.size > adev->doorbell.num_doorbells * sizeof(u32)) {
-		*aperture_base = adev->doorbell.base;
-		*aperture_size = adev->doorbell.size;
-		*start_offset = adev->doorbell.num_doorbells * sizeof(u32);
-	} else {
-		*aperture_base = 0;
-		*aperture_size = 0;
-		*start_offset = 0;
-	}
-}
+
 
 /*
- * amdgpu_wb_*()
+ * amdgpu_device_wb_*()
  * Writeback is the method by which the GPU updates special pages in memory
  * with the status of certain GPU events (fences, ring pointers,etc.).
  */
 
 /**
- * amdgpu_wb_fini - Disable Writeback and free memory
+ * amdgpu_device_wb_fini - Disable Writeback and free memory
  *
  * @adev: amdgpu_device pointer
  *
  * Disables Writeback and frees the Writeback memory (all asics).
  * Used at driver shutdown.
  */
-static void amdgpu_wb_fini(struct amdgpu_device *adev)
+static void amdgpu_device_wb_fini(struct amdgpu_device *adev)
 {
 	if (adev->wb.wb_obj) {
 		amdgpu_bo_free_kernel(&adev->wb.wb_obj,
@@ -502,7 +465,7 @@ static void amdgpu_wb_fini(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_wb_init- Init Writeback driver info and allocate memory
+ * amdgpu_device_wb_init- Init Writeback driver info and allocate memory
  *
  * @adev: amdgpu_device pointer
  *
@@ -510,7 +473,7 @@ static void amdgpu_wb_fini(struct amdgpu_device *adev)
  * Used at driver startup.
  * Returns 0 on success or an -error on failure.
  */
-static int amdgpu_wb_init(struct amdgpu_device *adev)
+static int amdgpu_device_wb_init(struct amdgpu_device *adev)
 {
 	int r;
 
@@ -536,7 +499,7 @@ static int amdgpu_wb_init(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_wb_get - Allocate a wb entry
+ * amdgpu_device_wb_get - Allocate a wb entry
  *
  * @adev: amdgpu_device pointer
  * @wb: wb index
@@ -544,7 +507,7 @@ static int amdgpu_wb_init(struct amdgpu_device *adev)
  * Allocate a wb slot for use by the driver (all asics).
  * Returns 0 on success or -EINVAL on failure.
  */
-int amdgpu_wb_get(struct amdgpu_device *adev, u32 *wb)
+int amdgpu_device_wb_get(struct amdgpu_device *adev, u32 *wb)
 {
 	unsigned long offset = find_first_zero_bit(adev->wb.used, adev->wb.num_wb);
 
@@ -558,21 +521,21 @@ int amdgpu_wb_get(struct amdgpu_device *adev, u32 *wb)
 }
 
 /**
- * amdgpu_wb_free - Free a wb entry
+ * amdgpu_device_wb_free - Free a wb entry
  *
  * @adev: amdgpu_device pointer
  * @wb: wb index
  *
  * Free a wb slot allocated for use by the driver (all asics)
  */
-void amdgpu_wb_free(struct amdgpu_device *adev, u32 wb)
+void amdgpu_device_wb_free(struct amdgpu_device *adev, u32 wb)
 {
 	if (wb < adev->wb.num_wb)
 		__clear_bit(wb >> 3, adev->wb.used);
 }
 
 /**
- * amdgpu_vram_location - try to find VRAM location
+ * amdgpu_device_vram_location - try to find VRAM location
  * @adev: amdgpu device structure holding all necessary informations
  * @mc: memory controller structure holding memory informations
  * @base: base address at which to put VRAM
@@ -580,7 +543,8 @@ void amdgpu_wb_free(struct amdgpu_device *adev, u32 wb)
  * Function will try to place VRAM at base address provided
  * as parameter.
  */
-void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64 base)
+void amdgpu_device_vram_location(struct amdgpu_device *adev,
+				 struct amdgpu_mc *mc, u64 base)
 {
 	uint64_t limit = (uint64_t)amdgpu_vram_limit << 20;
 
@@ -594,7 +558,7 @@ void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64
 }
 
 /**
- * amdgpu_gart_location - try to find GTT location
+ * amdgpu_device_gart_location - try to find GTT location
  * @adev: amdgpu device structure holding all necessary informations
  * @mc: memory controller structure holding memory informations
  *
@@ -605,7 +569,8 @@ void amdgpu_vram_location(struct amdgpu_device *adev, struct amdgpu_mc *mc, u64
  *
  * FIXME: when reducing GTT size align new size on power of 2.
  */
-void amdgpu_gart_location(struct amdgpu_device *adev, struct amdgpu_mc *mc)
+void amdgpu_device_gart_location(struct amdgpu_device *adev,
+				 struct amdgpu_mc *mc)
 {
 	u64 size_af, size_bf;
 
@@ -632,101 +597,6 @@ void amdgpu_gart_location(struct amdgpu_device *adev, struct amdgpu_mc *mc)
 			mc->gart_size >> 20, mc->gart_start, mc->gart_end);
 }
 
-/*
- * Firmware Reservation functions
- */
-/**
- * amdgpu_fw_reserve_vram_fini - free fw reserved vram
- *
- * @adev: amdgpu_device pointer
- *
- * free fw reserved vram if it has been reserved.
- */
-void amdgpu_fw_reserve_vram_fini(struct amdgpu_device *adev)
-{
-	amdgpu_bo_free_kernel(&adev->fw_vram_usage.reserved_bo,
-		NULL, &adev->fw_vram_usage.va);
-}
-
-/**
- * amdgpu_fw_reserve_vram_init - create bo vram reservation from fw
- *
- * @adev: amdgpu_device pointer
- *
- * create bo vram reservation from fw.
- */
-int amdgpu_fw_reserve_vram_init(struct amdgpu_device *adev)
-{
-	struct ttm_operation_ctx ctx = { false, false };
-	int r = 0;
-	int i;
-	u64 vram_size = adev->mc.visible_vram_size;
-	u64 offset = adev->fw_vram_usage.start_offset;
-	u64 size = adev->fw_vram_usage.size;
-	struct amdgpu_bo *bo;
-
-	adev->fw_vram_usage.va = NULL;
-	adev->fw_vram_usage.reserved_bo = NULL;
-
-	if (adev->fw_vram_usage.size > 0 &&
-		adev->fw_vram_usage.size <= vram_size) {
-
-		r = amdgpu_bo_create(adev, adev->fw_vram_usage.size,
-			PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM,
-			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-			AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, NULL, NULL, 0,
-			&adev->fw_vram_usage.reserved_bo);
-		if (r)
-			goto error_create;
-
-		r = amdgpu_bo_reserve(adev->fw_vram_usage.reserved_bo, false);
-		if (r)
-			goto error_reserve;
-
-		/* remove the original mem node and create a new one at the
-		 * request position
-		 */
-		bo = adev->fw_vram_usage.reserved_bo;
-		offset = ALIGN(offset, PAGE_SIZE);
-		for (i = 0; i < bo->placement.num_placement; ++i) {
-			bo->placements[i].fpfn = offset >> PAGE_SHIFT;
-			bo->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
-		}
-
-		ttm_bo_mem_put(&bo->tbo, &bo->tbo.mem);
-		r = ttm_bo_mem_space(&bo->tbo, &bo->placement,
-				     &bo->tbo.mem, &ctx);
-		if (r)
-			goto error_pin;
-
-		r = amdgpu_bo_pin_restricted(adev->fw_vram_usage.reserved_bo,
-			AMDGPU_GEM_DOMAIN_VRAM,
-			adev->fw_vram_usage.start_offset,
-			(adev->fw_vram_usage.start_offset +
-			adev->fw_vram_usage.size), NULL);
-		if (r)
-			goto error_pin;
-		r = amdgpu_bo_kmap(adev->fw_vram_usage.reserved_bo,
-			&adev->fw_vram_usage.va);
-		if (r)
-			goto error_kmap;
-
-		amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
-	}
-	return r;
-
-error_kmap:
-	amdgpu_bo_unpin(adev->fw_vram_usage.reserved_bo);
-error_pin:
-	amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
-error_reserve:
-	amdgpu_bo_unref(&adev->fw_vram_usage.reserved_bo);
-error_create:
-	adev->fw_vram_usage.va = NULL;
-	adev->fw_vram_usage.reserved_bo = NULL;
-	return r;
-}
-
 /**
  * amdgpu_device_resize_fb_bar - try to resize FB BAR
  *
@@ -771,7 +641,7 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
 			      cmd & ~PCI_COMMAND_MEMORY);
 
 	/* Free the VRAM and doorbell BAR, we most likely need to move both. */
-	amdgpu_doorbell_fini(adev);
+	amdgpu_device_doorbell_fini(adev);
 	if (adev->asic_type >= CHIP_BONAIRE)
 		pci_release_resource(adev->pdev, 2);
 
@@ -788,7 +658,7 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
 	/* When the doorbell or fb BAR isn't available we have no chance of
 	 * using the device.
 	 */
-	r = amdgpu_doorbell_init(adev);
+	r = amdgpu_device_doorbell_init(adev);
 	if (r || (pci_resource_flags(adev->pdev, 0) & IORESOURCE_UNSET))
 		return -ENODEV;
 
@@ -801,7 +671,7 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
  * GPU helpers function.
  */
 /**
- * amdgpu_need_post - check if the hw need post or not
+ * amdgpu_device_need_post - check if the hw need post or not
  *
  * @adev: amdgpu_device pointer
  *
@@ -809,7 +679,7 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
  * or post is needed if  hw reset is performed.
  * Returns true if need or false if not.
  */
-bool amdgpu_need_post(struct amdgpu_device *adev)
+bool amdgpu_device_need_post(struct amdgpu_device *adev)
 {
 	uint32_t reg;
 
@@ -854,285 +724,9 @@ bool amdgpu_need_post(struct amdgpu_device *adev)
 	return true;
 }
 
-/**
- * amdgpu_dummy_page_init - init dummy page used by the driver
- *
- * @adev: amdgpu_device pointer
- *
- * Allocate the dummy page used by the driver (all asics).
- * This dummy page is used by the driver as a filler for gart entries
- * when pages are taken out of the GART
- * Returns 0 on sucess, -ENOMEM on failure.
- */
-int amdgpu_dummy_page_init(struct amdgpu_device *adev)
-{
-	if (adev->dummy_page.page)
-		return 0;
-	adev->dummy_page.page = alloc_page(GFP_DMA32 | GFP_KERNEL | __GFP_ZERO);
-	if (adev->dummy_page.page == NULL)
-		return -ENOMEM;
-	adev->dummy_page.addr = pci_map_page(adev->pdev, adev->dummy_page.page,
-					0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-	if (pci_dma_mapping_error(adev->pdev, adev->dummy_page.addr)) {
-		dev_err(&adev->pdev->dev, "Failed to DMA MAP the dummy page\n");
-		__free_page(adev->dummy_page.page);
-		adev->dummy_page.page = NULL;
-		return -ENOMEM;
-	}
-	return 0;
-}
-
-/**
- * amdgpu_dummy_page_fini - free dummy page used by the driver
- *
- * @adev: amdgpu_device pointer
- *
- * Frees the dummy page used by the driver (all asics).
- */
-void amdgpu_dummy_page_fini(struct amdgpu_device *adev)
-{
-	if (adev->dummy_page.page == NULL)
-		return;
-	pci_unmap_page(adev->pdev, adev->dummy_page.addr,
-			PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
-	__free_page(adev->dummy_page.page);
-	adev->dummy_page.page = NULL;
-}
-
-
-/* ATOM accessor methods */
-/*
- * ATOM is an interpreted byte code stored in tables in the vbios.  The
- * driver registers callbacks to access registers and the interpreter
- * in the driver parses the tables and executes then to program specific
- * actions (set display modes, asic init, etc.).  See amdgpu_atombios.c,
- * atombios.h, and atom.c
- */
-
-/**
- * cail_pll_read - read PLL register
- *
- * @info: atom card_info pointer
- * @reg: PLL register offset
- *
- * Provides a PLL register accessor for the atom interpreter (r4xx+).
- * Returns the value of the PLL register.
- */
-static uint32_t cail_pll_read(struct card_info *info, uint32_t reg)
-{
-	return 0;
-}
-
-/**
- * cail_pll_write - write PLL register
- *
- * @info: atom card_info pointer
- * @reg: PLL register offset
- * @val: value to write to the pll register
- *
- * Provides a PLL register accessor for the atom interpreter (r4xx+).
- */
-static void cail_pll_write(struct card_info *info, uint32_t reg, uint32_t val)
-{
-
-}
-
-/**
- * cail_mc_read - read MC (Memory Controller) register
- *
- * @info: atom card_info pointer
- * @reg: MC register offset
- *
- * Provides an MC register accessor for the atom interpreter (r4xx+).
- * Returns the value of the MC register.
- */
-static uint32_t cail_mc_read(struct card_info *info, uint32_t reg)
-{
-	return 0;
-}
-
-/**
- * cail_mc_write - write MC (Memory Controller) register
- *
- * @info: atom card_info pointer
- * @reg: MC register offset
- * @val: value to write to the pll register
- *
- * Provides a MC register accessor for the atom interpreter (r4xx+).
- */
-static void cail_mc_write(struct card_info *info, uint32_t reg, uint32_t val)
-{
-
-}
-
-/**
- * cail_reg_write - write MMIO register
- *
- * @info: atom card_info pointer
- * @reg: MMIO register offset
- * @val: value to write to the pll register
- *
- * Provides a MMIO register accessor for the atom interpreter (r4xx+).
- */
-static void cail_reg_write(struct card_info *info, uint32_t reg, uint32_t val)
-{
-	struct amdgpu_device *adev = info->dev->dev_private;
-
-	WREG32(reg, val);
-}
-
-/**
- * cail_reg_read - read MMIO register
- *
- * @info: atom card_info pointer
- * @reg: MMIO register offset
- *
- * Provides an MMIO register accessor for the atom interpreter (r4xx+).
- * Returns the value of the MMIO register.
- */
-static uint32_t cail_reg_read(struct card_info *info, uint32_t reg)
-{
-	struct amdgpu_device *adev = info->dev->dev_private;
-	uint32_t r;
-
-	r = RREG32(reg);
-	return r;
-}
-
-/**
- * cail_ioreg_write - write IO register
- *
- * @info: atom card_info pointer
- * @reg: IO register offset
- * @val: value to write to the pll register
- *
- * Provides a IO register accessor for the atom interpreter (r4xx+).
- */
-static void cail_ioreg_write(struct card_info *info, uint32_t reg, uint32_t val)
-{
-	struct amdgpu_device *adev = info->dev->dev_private;
-
-	WREG32_IO(reg, val);
-}
-
-/**
- * cail_ioreg_read - read IO register
- *
- * @info: atom card_info pointer
- * @reg: IO register offset
- *
- * Provides an IO register accessor for the atom interpreter (r4xx+).
- * Returns the value of the IO register.
- */
-static uint32_t cail_ioreg_read(struct card_info *info, uint32_t reg)
-{
-	struct amdgpu_device *adev = info->dev->dev_private;
-	uint32_t r;
-
-	r = RREG32_IO(reg);
-	return r;
-}
-
-static ssize_t amdgpu_atombios_get_vbios_version(struct device *dev,
-						 struct device_attribute *attr,
-						 char *buf)
-{
-	struct drm_device *ddev = dev_get_drvdata(dev);
-	struct amdgpu_device *adev = ddev->dev_private;
-	struct atom_context *ctx = adev->mode_info.atom_context;
-
-	return snprintf(buf, PAGE_SIZE, "%s\n", ctx->vbios_version);
-}
-
-static DEVICE_ATTR(vbios_version, 0444, amdgpu_atombios_get_vbios_version,
-		   NULL);
-
-/**
- * amdgpu_atombios_fini - free the driver info and callbacks for atombios
- *
- * @adev: amdgpu_device pointer
- *
- * Frees the driver info and register access callbacks for the ATOM
- * interpreter (r4xx+).
- * Called at driver shutdown.
- */
-static void amdgpu_atombios_fini(struct amdgpu_device *adev)
-{
-	if (adev->mode_info.atom_context) {
-		kfree(adev->mode_info.atom_context->scratch);
-		kfree(adev->mode_info.atom_context->iio);
-	}
-	kfree(adev->mode_info.atom_context);
-	adev->mode_info.atom_context = NULL;
-	kfree(adev->mode_info.atom_card_info);
-	adev->mode_info.atom_card_info = NULL;
-	device_remove_file(adev->dev, &dev_attr_vbios_version);
-}
-
-/**
- * amdgpu_atombios_init - init the driver info and callbacks for atombios
- *
- * @adev: amdgpu_device pointer
- *
- * Initializes the driver info and register access callbacks for the
- * ATOM interpreter (r4xx+).
- * Returns 0 on sucess, -ENOMEM on failure.
- * Called at driver startup.
- */
-static int amdgpu_atombios_init(struct amdgpu_device *adev)
-{
-	struct card_info *atom_card_info =
-	    kzalloc(sizeof(struct card_info), GFP_KERNEL);
-	int ret;
-
-	if (!atom_card_info)
-		return -ENOMEM;
-
-	adev->mode_info.atom_card_info = atom_card_info;
-	atom_card_info->dev = adev->ddev;
-	atom_card_info->reg_read = cail_reg_read;
-	atom_card_info->reg_write = cail_reg_write;
-	/* needed for iio ops */
-	if (adev->rio_mem) {
-		atom_card_info->ioreg_read = cail_ioreg_read;
-		atom_card_info->ioreg_write = cail_ioreg_write;
-	} else {
-		DRM_DEBUG("PCI I/O BAR is not found. Using MMIO to access ATOM BIOS\n");
-		atom_card_info->ioreg_read = cail_reg_read;
-		atom_card_info->ioreg_write = cail_reg_write;
-	}
-	atom_card_info->mc_read = cail_mc_read;
-	atom_card_info->mc_write = cail_mc_write;
-	atom_card_info->pll_read = cail_pll_read;
-	atom_card_info->pll_write = cail_pll_write;
-
-	adev->mode_info.atom_context = amdgpu_atom_parse(atom_card_info, adev->bios);
-	if (!adev->mode_info.atom_context) {
-		amdgpu_atombios_fini(adev);
-		return -ENOMEM;
-	}
-
-	mutex_init(&adev->mode_info.atom_context->mutex);
-	if (adev->is_atom_fw) {
-		amdgpu_atomfirmware_scratch_regs_init(adev);
-		amdgpu_atomfirmware_allocate_fb_scratch(adev);
-	} else {
-		amdgpu_atombios_scratch_regs_init(adev);
-		amdgpu_atombios_allocate_fb_scratch(adev);
-	}
-
-	ret = device_create_file(adev->dev, &dev_attr_vbios_version);
-	if (ret) {
-		DRM_ERROR("Failed to create device file for VBIOS version\n");
-		return ret;
-	}
-
-	return 0;
-}
-
 /* if we get transitioned to only one device, take VGA back */
 /**
- * amdgpu_vga_set_decode - enable/disable vga decode
+ * amdgpu_device_vga_set_decode - enable/disable vga decode
  *
  * @cookie: amdgpu_device pointer
  * @state: enable/disable vga decode
@@ -1140,7 +734,7 @@ static int amdgpu_atombios_init(struct amdgpu_device *adev)
  * Enable/disable vga decode (all asics).
  * Returns VGA resource flags.
  */
-static unsigned int amdgpu_vga_set_decode(void *cookie, bool state)
+static unsigned int amdgpu_device_vga_set_decode(void *cookie, bool state)
 {
 	struct amdgpu_device *adev = cookie;
 	amdgpu_asic_set_vga_state(adev, state);
@@ -1151,7 +745,7 @@ static unsigned int amdgpu_vga_set_decode(void *cookie, bool state)
 		return VGA_RSRC_NORMAL_IO | VGA_RSRC_NORMAL_MEM;
 }
 
-static void amdgpu_check_block_size(struct amdgpu_device *adev)
+static void amdgpu_device_check_block_size(struct amdgpu_device *adev)
 {
 	/* defines number of bits in page table versus page directory,
 	 * a page is 4KB so we have 12 bits offset, minimum 9 bits in the
@@ -1166,7 +760,7 @@ static void amdgpu_check_block_size(struct amdgpu_device *adev)
 	}
 }
 
-static void amdgpu_check_vm_size(struct amdgpu_device *adev)
+static void amdgpu_device_check_vm_size(struct amdgpu_device *adev)
 {
 	/* no need to check the default value */
 	if (amdgpu_vm_size == -1)
@@ -1180,14 +774,14 @@ static void amdgpu_check_vm_size(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_check_arguments - validate module params
+ * amdgpu_device_check_arguments - validate module params
  *
  * @adev: amdgpu_device pointer
  *
  * Validates certain module parameters and updates
  * the associated values used by the driver (all asics).
  */
-static void amdgpu_check_arguments(struct amdgpu_device *adev)
+static void amdgpu_device_check_arguments(struct amdgpu_device *adev)
 {
 	if (amdgpu_sched_jobs < 4) {
 		dev_warn(adev->dev, "sched jobs (%d) must be at least 4\n",
@@ -1220,9 +814,9 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev)
 		amdgpu_vm_fragment_size = -1;
 	}
 
-	amdgpu_check_vm_size(adev);
+	amdgpu_device_check_vm_size(adev);
 
-	amdgpu_check_block_size(adev);
+	amdgpu_device_check_block_size(adev);
 
 	if (amdgpu_vram_page_split != -1 && (amdgpu_vram_page_split < 16 ||
 	    !is_power_of_2(amdgpu_vram_page_split))) {
@@ -1230,6 +824,11 @@ static void amdgpu_check_arguments(struct amdgpu_device *adev)
 			 amdgpu_vram_page_split);
 		amdgpu_vram_page_split = 1024;
 	}
+
+	if (amdgpu_lockup_timeout == 0) {
+		dev_warn(adev->dev, "lockup_timeout msut be > 0, adjusting to 10000\n");
+		amdgpu_lockup_timeout = 10000;
+	}
 }
 
 /**
@@ -1293,9 +892,9 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = {
 	.can_switch = amdgpu_switcheroo_can_switch,
 };
 
-int amdgpu_set_clockgating_state(struct amdgpu_device *adev,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state)
+int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+					   enum amd_ip_block_type block_type,
+					   enum amd_clockgating_state state)
 {
 	int i, r = 0;
 
@@ -1315,9 +914,9 @@ int amdgpu_set_clockgating_state(struct amdgpu_device *adev,
 	return r;
 }
 
-int amdgpu_set_powergating_state(struct amdgpu_device *adev,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state)
+int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+					   enum amd_ip_block_type block_type,
+					   enum amd_powergating_state state)
 {
 	int i, r = 0;
 
@@ -1337,7 +936,8 @@ int amdgpu_set_powergating_state(struct amdgpu_device *adev,
 	return r;
 }
 
-void amdgpu_get_clockgating_state(struct amdgpu_device *adev, u32 *flags)
+void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev,
+					    u32 *flags)
 {
 	int i;
 
@@ -1349,8 +949,8 @@ void amdgpu_get_clockgating_state(struct amdgpu_device *adev, u32 *flags)
 	}
 }
 
-int amdgpu_wait_for_idle(struct amdgpu_device *adev,
-			 enum amd_ip_block_type block_type)
+int amdgpu_device_ip_wait_for_idle(struct amdgpu_device *adev,
+				   enum amd_ip_block_type block_type)
 {
 	int i, r;
 
@@ -1368,8 +968,8 @@ int amdgpu_wait_for_idle(struct amdgpu_device *adev,
 
 }
 
-bool amdgpu_is_idle(struct amdgpu_device *adev,
-		    enum amd_ip_block_type block_type)
+bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev,
+			      enum amd_ip_block_type block_type)
 {
 	int i;
 
@@ -1383,8 +983,9 @@ bool amdgpu_is_idle(struct amdgpu_device *adev,
 
 }
 
-struct amdgpu_ip_block * amdgpu_get_ip_block(struct amdgpu_device *adev,
-					     enum amd_ip_block_type type)
+struct amdgpu_ip_block *
+amdgpu_device_ip_get_ip_block(struct amdgpu_device *adev,
+			      enum amd_ip_block_type type)
 {
 	int i;
 
@@ -1396,7 +997,7 @@ struct amdgpu_ip_block * amdgpu_get_ip_block(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_ip_block_version_cmp
+ * amdgpu_device_ip_block_version_cmp
  *
  * @adev: amdgpu_device pointer
  * @type: enum amd_ip_block_type
@@ -1406,11 +1007,11 @@ struct amdgpu_ip_block * amdgpu_get_ip_block(struct amdgpu_device *adev,
  * return 0 if equal or greater
  * return 1 if smaller or the ip_block doesn't exist
  */
-int amdgpu_ip_block_version_cmp(struct amdgpu_device *adev,
-				enum amd_ip_block_type type,
-				u32 major, u32 minor)
+int amdgpu_device_ip_block_version_cmp(struct amdgpu_device *adev,
+				       enum amd_ip_block_type type,
+				       u32 major, u32 minor)
 {
-	struct amdgpu_ip_block *ip_block = amdgpu_get_ip_block(adev, type);
+	struct amdgpu_ip_block *ip_block = amdgpu_device_ip_get_ip_block(adev, type);
 
 	if (ip_block && ((ip_block->version->major > major) ||
 			((ip_block->version->major == major) &&
@@ -1421,7 +1022,7 @@ int amdgpu_ip_block_version_cmp(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_ip_block_add
+ * amdgpu_device_ip_block_add
  *
  * @adev: amdgpu_device pointer
  * @ip_block_version: pointer to the IP to add
@@ -1429,8 +1030,8 @@ int amdgpu_ip_block_version_cmp(struct amdgpu_device *adev,
  * Adds the IP block driver information to the collection of IPs
  * on the asic.
  */
-int amdgpu_ip_block_add(struct amdgpu_device *adev,
-			const struct amdgpu_ip_block_version *ip_block_version)
+int amdgpu_device_ip_block_add(struct amdgpu_device *adev,
+			       const struct amdgpu_ip_block_version *ip_block_version)
 {
 	if (!ip_block_version)
 		return -EINVAL;
@@ -1586,7 +1187,7 @@ out:
 	return err;
 }
 
-static int amdgpu_early_init(struct amdgpu_device *adev)
+static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -1695,7 +1296,7 @@ static int amdgpu_early_init(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_init(struct amdgpu_device *adev)
+static int amdgpu_device_ip_init(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -1711,7 +1312,7 @@ static int amdgpu_init(struct amdgpu_device *adev)
 		adev->ip_blocks[i].status.sw = true;
 		/* need to do gmc hw init early so we can allocate gpu mem */
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC) {
-			r = amdgpu_vram_scratch_init(adev);
+			r = amdgpu_device_vram_scratch_init(adev);
 			if (r) {
 				DRM_ERROR("amdgpu_vram_scratch_init failed %d\n", r);
 				return r;
@@ -1721,9 +1322,9 @@ static int amdgpu_init(struct amdgpu_device *adev)
 				DRM_ERROR("hw_init %d failed %d\n", i, r);
 				return r;
 			}
-			r = amdgpu_wb_init(adev);
+			r = amdgpu_device_wb_init(adev);
 			if (r) {
-				DRM_ERROR("amdgpu_wb_init failed %d\n", r);
+				DRM_ERROR("amdgpu_device_wb_init failed %d\n", r);
 				return r;
 			}
 			adev->ip_blocks[i].status.hw = true;
@@ -1762,18 +1363,18 @@ static int amdgpu_init(struct amdgpu_device *adev)
 	return 0;
 }
 
-static void amdgpu_fill_reset_magic(struct amdgpu_device *adev)
+static void amdgpu_device_fill_reset_magic(struct amdgpu_device *adev)
 {
 	memcpy(adev->reset_magic, adev->gart.ptr, AMDGPU_RESET_MAGIC_NUM);
 }
 
-static bool amdgpu_check_vram_lost(struct amdgpu_device *adev)
+static bool amdgpu_device_check_vram_lost(struct amdgpu_device *adev)
 {
 	return !!memcmp(adev->gart.ptr, adev->reset_magic,
 			AMDGPU_RESET_MAGIC_NUM);
 }
 
-static int amdgpu_late_set_cg_state(struct amdgpu_device *adev)
+static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 {
 	int i = 0, r;
 
@@ -1796,7 +1397,7 @@ static int amdgpu_late_set_cg_state(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_late_init(struct amdgpu_device *adev)
+static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
 {
 	int i = 0, r;
 
@@ -1817,12 +1418,12 @@ static int amdgpu_late_init(struct amdgpu_device *adev)
 	mod_delayed_work(system_wq, &adev->late_init_work,
 			msecs_to_jiffies(AMDGPU_RESUME_MS));
 
-	amdgpu_fill_reset_magic(adev);
+	amdgpu_device_fill_reset_magic(adev);
 
 	return 0;
 }
 
-static int amdgpu_fini(struct amdgpu_device *adev)
+static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -1856,8 +1457,8 @@ static int amdgpu_fini(struct amdgpu_device *adev)
 			continue;
 		if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_GMC) {
 			amdgpu_free_static_csa(adev);
-			amdgpu_wb_fini(adev);
-			amdgpu_vram_scratch_fini(adev);
+			amdgpu_device_wb_fini(adev);
+			amdgpu_device_vram_scratch_fini(adev);
 		}
 
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
@@ -1910,14 +1511,14 @@ static int amdgpu_fini(struct amdgpu_device *adev)
 	return 0;
 }
 
-static void amdgpu_late_init_func_handler(struct work_struct *work)
+static void amdgpu_device_ip_late_init_func_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, late_init_work.work);
-	amdgpu_late_set_cg_state(adev);
+	amdgpu_device_ip_late_set_cg_state(adev);
 }
 
-int amdgpu_suspend(struct amdgpu_device *adev)
+int amdgpu_device_ip_suspend(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -1925,10 +1526,10 @@ int amdgpu_suspend(struct amdgpu_device *adev)
 		amdgpu_virt_request_full_gpu(adev, false);
 
 	/* ungate SMC block first */
-	r = amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_SMC,
-					 AMD_CG_STATE_UNGATE);
+	r = amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_SMC,
+						   AMD_CG_STATE_UNGATE);
 	if (r) {
-		DRM_ERROR("set_clockgating_state(ungate) SMC failed %d\n",r);
+		DRM_ERROR("set_clockgating_state(ungate) SMC failed %d\n", r);
 	}
 
 	for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
@@ -1958,7 +1559,7 @@ int amdgpu_suspend(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_sriov_reinit_early(struct amdgpu_device *adev)
+static int amdgpu_device_ip_reinit_early_sriov(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -1987,7 +1588,7 @@ static int amdgpu_sriov_reinit_early(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_sriov_reinit_late(struct amdgpu_device *adev)
+static int amdgpu_device_ip_reinit_late_sriov(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -2020,7 +1621,7 @@ static int amdgpu_sriov_reinit_late(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_resume_phase1(struct amdgpu_device *adev)
+static int amdgpu_device_ip_resume_phase1(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -2043,7 +1644,7 @@ static int amdgpu_resume_phase1(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_resume_phase2(struct amdgpu_device *adev)
+static int amdgpu_device_ip_resume_phase2(struct amdgpu_device *adev)
 {
 	int i, r;
 
@@ -2065,14 +1666,14 @@ static int amdgpu_resume_phase2(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_resume(struct amdgpu_device *adev)
+static int amdgpu_device_ip_resume(struct amdgpu_device *adev)
 {
 	int r;
 
-	r = amdgpu_resume_phase1(adev);
+	r = amdgpu_device_ip_resume_phase1(adev);
 	if (r)
 		return r;
-	r = amdgpu_resume_phase2(adev);
+	r = amdgpu_device_ip_resume_phase2(adev);
 
 	return r;
 }
@@ -2211,7 +1812,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	hash_init(adev->mn_hash);
 	mutex_init(&adev->lock_reset);
 
-	amdgpu_check_arguments(adev);
+	amdgpu_device_check_arguments(adev);
 
 	spin_lock_init(&adev->mmio_idx_lock);
 	spin_lock_init(&adev->smc_idx_lock);
@@ -2229,7 +1830,8 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	INIT_LIST_HEAD(&adev->ring_lru_list);
 	spin_lock_init(&adev->ring_lru_list_lock);
 
-	INIT_DELAYED_WORK(&adev->late_init_work, amdgpu_late_init_func_handler);
+	INIT_DELAYED_WORK(&adev->late_init_work,
+			  amdgpu_device_ip_late_init_func_handler);
 
 	/* Registers mapping */
 	/* TODO: block userspace mapping of io register */
@@ -2249,7 +1851,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	DRM_INFO("register mmio size: %u\n", (unsigned)adev->rmmio_size);
 
 	/* doorbell bar mapping */
-	amdgpu_doorbell_init(adev);
+	amdgpu_device_doorbell_init(adev);
 
 	/* io port mapping */
 	for (i = 0; i < DEVICE_COUNT_RESOURCE; i++) {
@@ -2263,14 +1865,14 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 		DRM_INFO("PCI I/O BAR is not found.\n");
 
 	/* early init functions */
-	r = amdgpu_early_init(adev);
+	r = amdgpu_device_ip_early_init(adev);
 	if (r)
 		return r;
 
 	/* if we have > 1 VGA cards, then disable the amdgpu VGA resources */
 	/* this will fail for cards that aren't VGA class devices, just
 	 * ignore it */
-	vga_client_register(adev->pdev, adev, NULL, amdgpu_vga_set_decode);
+	vga_client_register(adev->pdev, adev, NULL, amdgpu_device_vga_set_decode);
 
 	if (amdgpu_runtime_pm == 1)
 		runtime = true;
@@ -2299,7 +1901,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	amdgpu_device_detect_sriov_bios(adev);
 
 	/* Post card if necessary */
-	if (amdgpu_need_post(adev)) {
+	if (amdgpu_device_need_post(adev)) {
 		if (!adev->bios) {
 			dev_err(adev->dev, "no vBIOS found\n");
 			r = -EINVAL;
@@ -2345,7 +1947,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	/* init the mode config */
 	drm_mode_config_init(adev->ddev);
 
-	r = amdgpu_init(adev);
+	r = amdgpu_device_ip_init(adev);
 	if (r) {
 		/* failed in exclusive mode due to timeout */
 		if (amdgpu_sriov_vf(adev) &&
@@ -2359,9 +1961,9 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 			r = -EAGAIN;
 			goto failed;
 		}
-		dev_err(adev->dev, "amdgpu_init failed\n");
+		dev_err(adev->dev, "amdgpu_device_ip_init failed\n");
 		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_INIT_FAIL, 0, 0);
-		amdgpu_fini(adev);
+		amdgpu_device_ip_fini(adev);
 		goto failed;
 	}
 
@@ -2397,7 +1999,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (r)
 		DRM_ERROR("registering pm debugfs failed (%d).\n", r);
 
-	r = amdgpu_gem_debugfs_init(adev);
+	r = amdgpu_debugfs_gem_init(adev);
 	if (r)
 		DRM_ERROR("registering gem debugfs failed (%d).\n", r);
 
@@ -2405,17 +2007,13 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (r)
 		DRM_ERROR("registering register debugfs failed (%d).\n", r);
 
-	r = amdgpu_debugfs_test_ib_ring_init(adev);
-	if (r)
-		DRM_ERROR("registering register test ib ring 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_vbios_dump_init(adev);
+	r = amdgpu_debugfs_init(adev);
 	if (r)
-		DRM_ERROR("Creating vbios dump debugfs failed (%d).\n", r);
+		DRM_ERROR("Creating debugfs files failed (%d).\n", r);
 
 	if ((amdgpu_testing & 1)) {
 		if (adev->accel_working)
@@ -2433,9 +2031,9 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	/* enable clockgating, etc. after ib tests, etc. since some blocks require
 	 * explicit gating rather than handling it automatically.
 	 */
-	r = amdgpu_late_init(adev);
+	r = amdgpu_device_ip_late_init(adev);
 	if (r) {
-		dev_err(adev->dev, "amdgpu_late_init failed\n");
+		dev_err(adev->dev, "amdgpu_device_ip_late_init failed\n");
 		amdgpu_vf_error_put(adev, AMDGIM_ERROR_VF_AMDGPU_LATE_INIT_FAIL, 0, r);
 		goto failed;
 	}
@@ -2466,12 +2064,11 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 	adev->shutdown = true;
 	if (adev->mode_info.mode_config_initialized)
 		drm_crtc_force_disable_all(adev->ddev);
-	/* evict vram memory */
-	amdgpu_bo_evict_vram(adev);
+
 	amdgpu_ib_pool_fini(adev);
 	amdgpu_fence_driver_fini(adev);
 	amdgpu_fbdev_fini(adev);
-	r = amdgpu_fini(adev);
+	r = amdgpu_device_ip_fini(adev);
 	if (adev->firmware.gpu_info_fw) {
 		release_firmware(adev->firmware.gpu_info_fw);
 		adev->firmware.gpu_info_fw = NULL;
@@ -2494,7 +2091,7 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 	adev->rio_mem = NULL;
 	iounmap(adev->rmmio);
 	adev->rmmio = NULL;
-	amdgpu_doorbell_fini(adev);
+	amdgpu_device_doorbell_fini(adev);
 	amdgpu_pm_sysfs_fini(adev);
 	amdgpu_debugfs_regs_cleanup(adev);
 }
@@ -2575,7 +2172,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 
 	amdgpu_fence_driver_suspend(adev);
 
-	r = amdgpu_suspend(adev);
+	r = amdgpu_device_ip_suspend(adev);
 
 	/* evict remaining vram memory
 	 * This second call to evict vram is to evict the gart page table
@@ -2583,7 +2180,6 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	 */
 	amdgpu_bo_evict_vram(adev);
 
-	amdgpu_atombios_scratch_regs_save(adev);
 	pci_save_state(dev->pdev);
 	if (suspend) {
 		/* Shut down the device */
@@ -2632,18 +2228,17 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 		if (r)
 			goto unlock;
 	}
-	amdgpu_atombios_scratch_regs_restore(adev);
 
 	/* post card */
-	if (amdgpu_need_post(adev)) {
+	if (amdgpu_device_need_post(adev)) {
 		r = amdgpu_atom_asic_init(adev->mode_info.atom_context);
 		if (r)
 			DRM_ERROR("amdgpu asic init failed\n");
 	}
 
-	r = amdgpu_resume(adev);
+	r = amdgpu_device_ip_resume(adev);
 	if (r) {
-		DRM_ERROR("amdgpu_resume failed (%d).\n", r);
+		DRM_ERROR("amdgpu_device_ip_resume failed (%d).\n", r);
 		goto unlock;
 	}
 	amdgpu_fence_driver_resume(adev);
@@ -2654,7 +2249,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 			DRM_ERROR("ib ring test failed (%d).\n", r);
 	}
 
-	r = amdgpu_late_init(adev);
+	r = amdgpu_device_ip_late_init(adev);
 	if (r)
 		goto unlock;
 
@@ -2734,7 +2329,7 @@ unlock:
 	return r;
 }
 
-static bool amdgpu_check_soft_reset(struct amdgpu_device *adev)
+static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev)
 {
 	int i;
 	bool asic_hang = false;
@@ -2756,7 +2351,7 @@ static bool amdgpu_check_soft_reset(struct amdgpu_device *adev)
 	return asic_hang;
 }
 
-static int amdgpu_pre_soft_reset(struct amdgpu_device *adev)
+static int amdgpu_device_ip_pre_soft_reset(struct amdgpu_device *adev)
 {
 	int i, r = 0;
 
@@ -2774,7 +2369,7 @@ static int amdgpu_pre_soft_reset(struct amdgpu_device *adev)
 	return 0;
 }
 
-static bool amdgpu_need_full_reset(struct amdgpu_device *adev)
+static bool amdgpu_device_ip_need_full_reset(struct amdgpu_device *adev)
 {
 	int i;
 
@@ -2795,7 +2390,7 @@ static bool amdgpu_need_full_reset(struct amdgpu_device *adev)
 	return false;
 }
 
-static int amdgpu_soft_reset(struct amdgpu_device *adev)
+static int amdgpu_device_ip_soft_reset(struct amdgpu_device *adev)
 {
 	int i, r = 0;
 
@@ -2813,7 +2408,7 @@ static int amdgpu_soft_reset(struct amdgpu_device *adev)
 	return 0;
 }
 
-static int amdgpu_post_soft_reset(struct amdgpu_device *adev)
+static int amdgpu_device_ip_post_soft_reset(struct amdgpu_device *adev)
 {
 	int i, r = 0;
 
@@ -2830,18 +2425,10 @@ static int amdgpu_post_soft_reset(struct amdgpu_device *adev)
 	return 0;
 }
 
-bool amdgpu_need_backup(struct amdgpu_device *adev)
-{
-	if (adev->flags & AMD_IS_APU)
-		return false;
-
-	return amdgpu_lockup_timeout > 0 ? true : false;
-}
-
-static int amdgpu_recover_vram_from_shadow(struct amdgpu_device *adev,
-					   struct amdgpu_ring *ring,
-					   struct amdgpu_bo *bo,
-					   struct dma_fence **fence)
+static int amdgpu_device_recover_vram_from_shadow(struct amdgpu_device *adev,
+						  struct amdgpu_ring *ring,
+						  struct amdgpu_bo *bo,
+						  struct dma_fence **fence)
 {
 	uint32_t domain;
 	int r;
@@ -2874,7 +2461,7 @@ err:
 }
 
 /*
- * amdgpu_reset - reset ASIC/GPU for bare-metal or passthrough
+ * amdgpu_device_reset - reset ASIC/GPU for bare-metal or passthrough
  *
  * @adev: amdgpu device pointer
  * @reset_flags: output param tells caller the reset result
@@ -2882,18 +2469,19 @@ err:
  * attempt to do soft-reset or full-reset and reinitialize Asic
  * return 0 means successed otherwise failed
 */
-static int amdgpu_reset(struct amdgpu_device *adev, uint64_t* reset_flags)
+static int amdgpu_device_reset(struct amdgpu_device *adev,
+			       uint64_t* reset_flags)
 {
 	bool need_full_reset, vram_lost = 0;
 	int r;
 
-	need_full_reset = amdgpu_need_full_reset(adev);
+	need_full_reset = amdgpu_device_ip_need_full_reset(adev);
 
 	if (!need_full_reset) {
-		amdgpu_pre_soft_reset(adev);
-		r = amdgpu_soft_reset(adev);
-		amdgpu_post_soft_reset(adev);
-		if (r || amdgpu_check_soft_reset(adev)) {
+		amdgpu_device_ip_pre_soft_reset(adev);
+		r = amdgpu_device_ip_soft_reset(adev);
+		amdgpu_device_ip_post_soft_reset(adev);
+		if (r || amdgpu_device_ip_check_soft_reset(adev)) {
 			DRM_INFO("soft reset failed, will fallback to full reset!\n");
 			need_full_reset = true;
 		}
@@ -2901,22 +2489,20 @@ static int amdgpu_reset(struct amdgpu_device *adev, uint64_t* reset_flags)
 	}
 
 	if (need_full_reset) {
-		r = amdgpu_suspend(adev);
+		r = amdgpu_device_ip_suspend(adev);
 
 retry:
-		amdgpu_atombios_scratch_regs_save(adev);
 		r = amdgpu_asic_reset(adev);
-		amdgpu_atombios_scratch_regs_restore(adev);
 		/* post card */
 		amdgpu_atom_asic_init(adev->mode_info.atom_context);
 
 		if (!r) {
 			dev_info(adev->dev, "GPU reset succeeded, trying to resume\n");
-			r = amdgpu_resume_phase1(adev);
+			r = amdgpu_device_ip_resume_phase1(adev);
 			if (r)
 				goto out;
 
-			vram_lost = amdgpu_check_vram_lost(adev);
+			vram_lost = amdgpu_device_check_vram_lost(adev);
 			if (vram_lost) {
 				DRM_ERROR("VRAM is lost!\n");
 				atomic_inc(&adev->vram_lost_counter);
@@ -2927,12 +2513,12 @@ retry:
 			if (r)
 				goto out;
 
-			r = amdgpu_resume_phase2(adev);
+			r = amdgpu_device_ip_resume_phase2(adev);
 			if (r)
 				goto out;
 
 			if (vram_lost)
-				amdgpu_fill_reset_magic(adev);
+				amdgpu_device_fill_reset_magic(adev);
 		}
 	}
 
@@ -2942,7 +2528,7 @@ out:
 		r = amdgpu_ib_ring_tests(adev);
 		if (r) {
 			dev_err(adev->dev, "ib ring test failed (%d).\n", r);
-			r = amdgpu_suspend(adev);
+			r = amdgpu_device_ip_suspend(adev);
 			need_full_reset = true;
 			goto retry;
 		}
@@ -2960,7 +2546,7 @@ out:
 }
 
 /*
- * amdgpu_reset_sriov - reset ASIC for SR-IOV vf
+ * amdgpu_device_reset_sriov - reset ASIC for SR-IOV vf
  *
  * @adev: amdgpu device pointer
  * @reset_flags: output param tells caller the reset result
@@ -2968,7 +2554,9 @@ out:
  * do VF FLR and reinitialize Asic
  * return 0 means successed otherwise failed
 */
-static int amdgpu_reset_sriov(struct amdgpu_device *adev, uint64_t *reset_flags, bool from_hypervisor)
+static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,
+				     uint64_t *reset_flags,
+				     bool from_hypervisor)
 {
 	int r;
 
@@ -2980,7 +2568,7 @@ static int amdgpu_reset_sriov(struct amdgpu_device *adev, uint64_t *reset_flags,
 		return r;
 
 	/* Resume IP prior to SMC */
-	r = amdgpu_sriov_reinit_early(adev);
+	r = amdgpu_device_ip_reinit_early_sriov(adev);
 	if (r)
 		goto error;
 
@@ -2988,7 +2576,7 @@ static int amdgpu_reset_sriov(struct amdgpu_device *adev, uint64_t *reset_flags,
 	amdgpu_gtt_mgr_recover(&adev->mman.bdev.man[TTM_PL_TT]);
 
 	/* now we are okay to resume SMC/CP/SDMA */
-	r = amdgpu_sriov_reinit_late(adev);
+	r = amdgpu_device_ip_reinit_late_sriov(adev);
 	if (r)
 		goto error;
 
@@ -3015,25 +2603,33 @@ error:
 }
 
 /**
- * amdgpu_gpu_recover - reset the asic and recover scheduler
+ * amdgpu_device_gpu_recover - reset the asic and recover scheduler
  *
  * @adev: amdgpu device pointer
  * @job: which job trigger hang
+ * @force forces reset regardless of amdgpu_gpu_recovery
  *
  * Attempt to reset the GPU if it has hung (all asics).
  * Returns 0 for success or an error on failure.
  */
-int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job)
+int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
+			      struct amdgpu_job *job, bool force)
 {
 	struct drm_atomic_state *state = NULL;
 	uint64_t reset_flags = 0;
 	int i, r, resched;
 
-	if (!amdgpu_check_soft_reset(adev)) {
+	if (!amdgpu_device_ip_check_soft_reset(adev)) {
 		DRM_INFO("No hardware hang detected. Did some blocks stall?\n");
 		return 0;
 	}
 
+	if (!force && (amdgpu_gpu_recovery == 0 ||
+			(amdgpu_gpu_recovery == -1  && !amdgpu_sriov_vf(adev)))) {
+		DRM_INFO("GPU recovery disabled.\n");
+		return 0;
+	}
+
 	dev_info(adev->dev, "GPU reset begin!\n");
 
 	mutex_lock(&adev->lock_reset);
@@ -3058,16 +2654,16 @@ int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job)
 			continue;
 
 		kthread_park(ring->sched.thread);
-		amd_sched_hw_job_reset(&ring->sched, &job->base);
+		drm_sched_hw_job_reset(&ring->sched, &job->base);
 
 		/* after all hw jobs are reset, hw fence is meaningless, so force_completion */
 		amdgpu_fence_driver_force_completion(ring);
 	}
 
 	if (amdgpu_sriov_vf(adev))
-		r = amdgpu_reset_sriov(adev, &reset_flags, job ? false : true);
+		r = amdgpu_device_reset_sriov(adev, &reset_flags, job ? false : true);
 	else
-		r = amdgpu_reset(adev, &reset_flags);
+		r = amdgpu_device_reset(adev, &reset_flags);
 
 	if (!r) {
 		if (((reset_flags & AMDGPU_RESET_INFO_FULLRESET) && !(adev->flags & AMD_IS_APU)) ||
@@ -3080,7 +2676,7 @@ int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job)
 			mutex_lock(&adev->shadow_list_lock);
 			list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) {
 				next = NULL;
-				amdgpu_recover_vram_from_shadow(adev, ring, bo, &next);
+				amdgpu_device_recover_vram_from_shadow(adev, ring, bo, &next);
 				if (fence) {
 					r = dma_fence_wait(fence, false);
 					if (r) {
@@ -3111,7 +2707,7 @@ int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job)
 			if (job && job->ring->idx != i)
 				continue;
 
-			amd_sched_job_recovery(&ring->sched);
+			drm_sched_job_recovery(&ring->sched);
 			kthread_unpark(ring->sched.thread);
 		}
 	} else {
@@ -3153,7 +2749,7 @@ int amdgpu_gpu_recover(struct amdgpu_device *adev, struct amdgpu_job *job)
 	return r;
 }
 
-void amdgpu_get_pcie_info(struct amdgpu_device *adev)
+void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
 {
 	u32 mask;
 	int ret;
@@ -3245,773 +2841,3 @@ void amdgpu_get_pcie_info(struct amdgpu_device *adev)
 	}
 }
 
-/*
- * Debugfs
- */
-int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
-			     const struct drm_info_list *files,
-			     unsigned nfiles)
-{
-	unsigned i;
-
-	for (i = 0; i < adev->debugfs_count; i++) {
-		if (adev->debugfs[i].files == files) {
-			/* Already registered */
-			return 0;
-		}
-	}
-
-	i = adev->debugfs_count + 1;
-	if (i > AMDGPU_DEBUGFS_MAX_COMPONENTS) {
-		DRM_ERROR("Reached maximum number of debugfs components.\n");
-		DRM_ERROR("Report so we increase "
-			  "AMDGPU_DEBUGFS_MAX_COMPONENTS.\n");
-		return -EINVAL;
-	}
-	adev->debugfs[adev->debugfs_count].files = files;
-	adev->debugfs[adev->debugfs_count].num_files = nfiles;
-	adev->debugfs_count = i;
-#if defined(CONFIG_DEBUG_FS)
-	drm_debugfs_create_files(files, nfiles,
-				 adev->ddev->primary->debugfs_root,
-				 adev->ddev->primary);
-#endif
-	return 0;
-}
-
-#if defined(CONFIG_DEBUG_FS)
-
-static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf,
-					size_t size, loff_t *pos)
-{
-	struct amdgpu_device *adev = file_inode(f)->i_private;
-	ssize_t result = 0;
-	int r;
-	bool pm_pg_lock, use_bank;
-	unsigned instance_bank, sh_bank, se_bank;
-
-	if (size & 0x3 || *pos & 0x3)
-		return -EINVAL;
-
-	/* are we reading registers for which a PG lock is necessary? */
-	pm_pg_lock = (*pos >> 23) & 1;
-
-	if (*pos & (1ULL << 62)) {
-		se_bank = (*pos & GENMASK_ULL(33, 24)) >> 24;
-		sh_bank = (*pos & GENMASK_ULL(43, 34)) >> 34;
-		instance_bank = (*pos & GENMASK_ULL(53, 44)) >> 44;
-
-		if (se_bank == 0x3FF)
-			se_bank = 0xFFFFFFFF;
-		if (sh_bank == 0x3FF)
-			sh_bank = 0xFFFFFFFF;
-		if (instance_bank == 0x3FF)
-			instance_bank = 0xFFFFFFFF;
-		use_bank = 1;
-	} else {
-		use_bank = 0;
-	}
-
-	*pos &= (1UL << 22) - 1;
-
-	if (use_bank) {
-		if ((sh_bank != 0xFFFFFFFF && sh_bank >= adev->gfx.config.max_sh_per_se) ||
-		    (se_bank != 0xFFFFFFFF && se_bank >= adev->gfx.config.max_shader_engines))
-			return -EINVAL;
-		mutex_lock(&adev->grbm_idx_mutex);
-		amdgpu_gfx_select_se_sh(adev, se_bank,
-					sh_bank, instance_bank);
-	}
-
-	if (pm_pg_lock)
-		mutex_lock(&adev->pm.mutex);
-
-	while (size) {
-		uint32_t value;
-
-		if (*pos > adev->rmmio_size)
-			goto end;
-
-		value = RREG32(*pos >> 2);
-		r = put_user(value, (uint32_t *)buf);
-		if (r) {
-			result = r;
-			goto end;
-		}
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-end:
-	if (use_bank) {
-		amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
-		mutex_unlock(&adev->grbm_idx_mutex);
-	}
-
-	if (pm_pg_lock)
-		mutex_unlock(&adev->pm.mutex);
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_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;
-	bool pm_pg_lock, use_bank;
-	unsigned instance_bank, sh_bank, se_bank;
-
-	if (size & 0x3 || *pos & 0x3)
-		return -EINVAL;
-
-	/* are we reading registers for which a PG lock is necessary? */
-	pm_pg_lock = (*pos >> 23) & 1;
-
-	if (*pos & (1ULL << 62)) {
-		se_bank = (*pos & GENMASK_ULL(33, 24)) >> 24;
-		sh_bank = (*pos & GENMASK_ULL(43, 34)) >> 34;
-		instance_bank = (*pos & GENMASK_ULL(53, 44)) >> 44;
-
-		if (se_bank == 0x3FF)
-			se_bank = 0xFFFFFFFF;
-		if (sh_bank == 0x3FF)
-			sh_bank = 0xFFFFFFFF;
-		if (instance_bank == 0x3FF)
-			instance_bank = 0xFFFFFFFF;
-		use_bank = 1;
-	} else {
-		use_bank = 0;
-	}
-
-	*pos &= (1UL << 22) - 1;
-
-	if (use_bank) {
-		if ((sh_bank != 0xFFFFFFFF && sh_bank >= adev->gfx.config.max_sh_per_se) ||
-		    (se_bank != 0xFFFFFFFF && se_bank >= adev->gfx.config.max_shader_engines))
-			return -EINVAL;
-		mutex_lock(&adev->grbm_idx_mutex);
-		amdgpu_gfx_select_se_sh(adev, se_bank,
-					sh_bank, instance_bank);
-	}
-
-	if (pm_pg_lock)
-		mutex_lock(&adev->pm.mutex);
-
-	while (size) {
-		uint32_t value;
-
-		if (*pos > adev->rmmio_size)
-			return result;
-
-		r = get_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		WREG32(*pos >> 2, value);
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	if (use_bank) {
-		amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
-		mutex_unlock(&adev->grbm_idx_mutex);
-	}
-
-	if (pm_pg_lock)
-		mutex_unlock(&adev->pm.mutex);
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, 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;
-
-	while (size) {
-		uint32_t value;
-
-		value = RREG32_PCIE(*pos >> 2);
-		r = put_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_pcie_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;
-
-	while (size) {
-		uint32_t value;
-
-		r = get_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		WREG32_PCIE(*pos >> 2, value);
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, 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;
-
-	while (size) {
-		uint32_t value;
-
-		value = RREG32_DIDT(*pos >> 2);
-		r = put_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_didt_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;
-
-	while (size) {
-		uint32_t value;
-
-		r = get_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		WREG32_DIDT(*pos >> 2, value);
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, 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;
-
-	while (size) {
-		uint32_t value;
-
-		value = RREG32_SMC(*pos);
-		r = put_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_regs_smc_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;
-
-	while (size) {
-		uint32_t value;
-
-		r = get_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		WREG32_SMC(*pos, value);
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
-					size_t size, loff_t *pos)
-{
-	struct amdgpu_device *adev = file_inode(f)->i_private;
-	ssize_t result = 0;
-	int r;
-	uint32_t *config, no_regs = 0;
-
-	if (size & 0x3 || *pos & 0x3)
-		return -EINVAL;
-
-	config = kmalloc_array(256, sizeof(*config), GFP_KERNEL);
-	if (!config)
-		return -ENOMEM;
-
-	/* version, increment each time something is added */
-	config[no_regs++] = 3;
-	config[no_regs++] = adev->gfx.config.max_shader_engines;
-	config[no_regs++] = adev->gfx.config.max_tile_pipes;
-	config[no_regs++] = adev->gfx.config.max_cu_per_sh;
-	config[no_regs++] = adev->gfx.config.max_sh_per_se;
-	config[no_regs++] = adev->gfx.config.max_backends_per_se;
-	config[no_regs++] = adev->gfx.config.max_texture_channel_caches;
-	config[no_regs++] = adev->gfx.config.max_gprs;
-	config[no_regs++] = adev->gfx.config.max_gs_threads;
-	config[no_regs++] = adev->gfx.config.max_hw_contexts;
-	config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_frontend;
-	config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_backend;
-	config[no_regs++] = adev->gfx.config.sc_hiz_tile_fifo_size;
-	config[no_regs++] = adev->gfx.config.sc_earlyz_tile_fifo_size;
-	config[no_regs++] = adev->gfx.config.num_tile_pipes;
-	config[no_regs++] = adev->gfx.config.backend_enable_mask;
-	config[no_regs++] = adev->gfx.config.mem_max_burst_length_bytes;
-	config[no_regs++] = adev->gfx.config.mem_row_size_in_kb;
-	config[no_regs++] = adev->gfx.config.shader_engine_tile_size;
-	config[no_regs++] = adev->gfx.config.num_gpus;
-	config[no_regs++] = adev->gfx.config.multi_gpu_tile_size;
-	config[no_regs++] = adev->gfx.config.mc_arb_ramcfg;
-	config[no_regs++] = adev->gfx.config.gb_addr_config;
-	config[no_regs++] = adev->gfx.config.num_rbs;
-
-	/* rev==1 */
-	config[no_regs++] = adev->rev_id;
-	config[no_regs++] = adev->pg_flags;
-	config[no_regs++] = adev->cg_flags;
-
-	/* rev==2 */
-	config[no_regs++] = adev->family;
-	config[no_regs++] = adev->external_rev_id;
-
-	/* rev==3 */
-	config[no_regs++] = adev->pdev->device;
-	config[no_regs++] = adev->pdev->revision;
-	config[no_regs++] = adev->pdev->subsystem_device;
-	config[no_regs++] = adev->pdev->subsystem_vendor;
-
-	while (size && (*pos < no_regs * 4)) {
-		uint32_t value;
-
-		value = config[*pos >> 2];
-		r = put_user(value, (uint32_t *)buf);
-		if (r) {
-			kfree(config);
-			return r;
-		}
-
-		result += 4;
-		buf += 4;
-		*pos += 4;
-		size -= 4;
-	}
-
-	kfree(config);
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
-					size_t size, loff_t *pos)
-{
-	struct amdgpu_device *adev = file_inode(f)->i_private;
-	int idx, x, outsize, r, valuesize;
-	uint32_t values[16];
-
-	if (size & 3 || *pos & 0x3)
-		return -EINVAL;
-
-	if (amdgpu_dpm == 0)
-		return -EINVAL;
-
-	/* convert offset to sensor number */
-	idx = *pos >> 2;
-
-	valuesize = sizeof(values);
-	if (adev->powerplay.pp_funcs && adev->powerplay.pp_funcs->read_sensor)
-		r = amdgpu_dpm_read_sensor(adev, idx, &values[0], &valuesize);
-	else
-		return -EINVAL;
-
-	if (size > valuesize)
-		return -EINVAL;
-
-	outsize = 0;
-	x = 0;
-	if (!r) {
-		while (size) {
-			r = put_user(values[x++], (int32_t *)buf);
-			buf += 4;
-			size -= 4;
-			outsize += 4;
-		}
-	}
-
-	return !r ? outsize : r;
-}
-
-static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
-					size_t size, loff_t *pos)
-{
-	struct amdgpu_device *adev = f->f_inode->i_private;
-	int r, x;
-	ssize_t result=0;
-	uint32_t offset, se, sh, cu, wave, simd, data[32];
-
-	if (size & 3 || *pos & 3)
-		return -EINVAL;
-
-	/* decode offset */
-	offset = (*pos & GENMASK_ULL(6, 0));
-	se = (*pos & GENMASK_ULL(14, 7)) >> 7;
-	sh = (*pos & GENMASK_ULL(22, 15)) >> 15;
-	cu = (*pos & GENMASK_ULL(30, 23)) >> 23;
-	wave = (*pos & GENMASK_ULL(36, 31)) >> 31;
-	simd = (*pos & GENMASK_ULL(44, 37)) >> 37;
-
-	/* switch to the specific se/sh/cu */
-	mutex_lock(&adev->grbm_idx_mutex);
-	amdgpu_gfx_select_se_sh(adev, se, sh, cu);
-
-	x = 0;
-	if (adev->gfx.funcs->read_wave_data)
-		adev->gfx.funcs->read_wave_data(adev, simd, wave, data, &x);
-
-	amdgpu_gfx_select_se_sh(adev, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF);
-	mutex_unlock(&adev->grbm_idx_mutex);
-
-	if (!x)
-		return -EINVAL;
-
-	while (size && (offset < x * 4)) {
-		uint32_t value;
-
-		value = data[offset >> 2];
-		r = put_user(value, (uint32_t *)buf);
-		if (r)
-			return r;
-
-		result += 4;
-		buf += 4;
-		offset += 4;
-		size -= 4;
-	}
-
-	return result;
-}
-
-static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
-					size_t size, loff_t *pos)
-{
-	struct amdgpu_device *adev = f->f_inode->i_private;
-	int r;
-	ssize_t result = 0;
-	uint32_t offset, se, sh, cu, wave, simd, thread, bank, *data;
-
-	if (size & 3 || *pos & 3)
-		return -EINVAL;
-
-	/* decode offset */
-	offset = *pos & GENMASK_ULL(11, 0);
-	se = (*pos & GENMASK_ULL(19, 12)) >> 12;
-	sh = (*pos & GENMASK_ULL(27, 20)) >> 20;
-	cu = (*pos & GENMASK_ULL(35, 28)) >> 28;
-	wave = (*pos & GENMASK_ULL(43, 36)) >> 36;
-	simd = (*pos & GENMASK_ULL(51, 44)) >> 44;
-	thread = (*pos & GENMASK_ULL(59, 52)) >> 52;
-	bank = (*pos & GENMASK_ULL(61, 60)) >> 60;
-
-	data = kmalloc_array(1024, sizeof(*data), GFP_KERNEL);
-	if (!data)
-		return -ENOMEM;
-
-	/* switch to the specific se/sh/cu */
-	mutex_lock(&adev->grbm_idx_mutex);
-	amdgpu_gfx_select_se_sh(adev, se, sh, cu);
-
-	if (bank == 0) {
-		if (adev->gfx.funcs->read_wave_vgprs)
-			adev->gfx.funcs->read_wave_vgprs(adev, simd, wave, thread, offset, size>>2, data);
-	} else {
-		if (adev->gfx.funcs->read_wave_sgprs)
-			adev->gfx.funcs->read_wave_sgprs(adev, simd, wave, offset, size>>2, data);
-	}
-
-	amdgpu_gfx_select_se_sh(adev, 0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF);
-	mutex_unlock(&adev->grbm_idx_mutex);
-
-	while (size) {
-		uint32_t value;
-
-		value = data[offset++];
-		r = put_user(value, (uint32_t *)buf);
-		if (r) {
-			result = r;
-			goto err;
-		}
-
-		result += 4;
-		buf += 4;
-		size -= 4;
-	}
-
-err:
-	kfree(data);
-	return result;
-}
-
-static const struct file_operations amdgpu_debugfs_regs_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_regs_read,
-	.write = amdgpu_debugfs_regs_write,
-	.llseek = default_llseek
-};
-static const struct file_operations amdgpu_debugfs_regs_didt_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_regs_didt_read,
-	.write = amdgpu_debugfs_regs_didt_write,
-	.llseek = default_llseek
-};
-static const struct file_operations amdgpu_debugfs_regs_pcie_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_regs_pcie_read,
-	.write = amdgpu_debugfs_regs_pcie_write,
-	.llseek = default_llseek
-};
-static const struct file_operations amdgpu_debugfs_regs_smc_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_regs_smc_read,
-	.write = amdgpu_debugfs_regs_smc_write,
-	.llseek = default_llseek
-};
-
-static const struct file_operations amdgpu_debugfs_gca_config_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_gca_config_read,
-	.llseek = default_llseek
-};
-
-static const struct file_operations amdgpu_debugfs_sensors_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_sensor_read,
-	.llseek = default_llseek
-};
-
-static const struct file_operations amdgpu_debugfs_wave_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_wave_read,
-	.llseek = default_llseek
-};
-static const struct file_operations amdgpu_debugfs_gpr_fops = {
-	.owner = THIS_MODULE,
-	.read = amdgpu_debugfs_gpr_read,
-	.llseek = default_llseek
-};
-
-static const struct file_operations *debugfs_regs[] = {
-	&amdgpu_debugfs_regs_fops,
-	&amdgpu_debugfs_regs_didt_fops,
-	&amdgpu_debugfs_regs_pcie_fops,
-	&amdgpu_debugfs_regs_smc_fops,
-	&amdgpu_debugfs_gca_config_fops,
-	&amdgpu_debugfs_sensors_fops,
-	&amdgpu_debugfs_wave_fops,
-	&amdgpu_debugfs_gpr_fops,
-};
-
-static const char *debugfs_regs_names[] = {
-	"amdgpu_regs",
-	"amdgpu_regs_didt",
-	"amdgpu_regs_pcie",
-	"amdgpu_regs_smc",
-	"amdgpu_gca_config",
-	"amdgpu_sensors",
-	"amdgpu_wave",
-	"amdgpu_gpr",
-};
-
-static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
-{
-	struct drm_minor *minor = adev->ddev->primary;
-	struct dentry *ent, *root = minor->debugfs_root;
-	unsigned i, j;
-
-	for (i = 0; i < ARRAY_SIZE(debugfs_regs); i++) {
-		ent = debugfs_create_file(debugfs_regs_names[i],
-					  S_IFREG | S_IRUGO, root,
-					  adev, debugfs_regs[i]);
-		if (IS_ERR(ent)) {
-			for (j = 0; j < i; j++) {
-				debugfs_remove(adev->debugfs_regs[i]);
-				adev->debugfs_regs[i] = NULL;
-			}
-			return PTR_ERR(ent);
-		}
-
-		if (!i)
-			i_size_write(ent->d_inode, adev->rmmio_size);
-		adev->debugfs_regs[i] = ent;
-	}
-
-	return 0;
-}
-
-static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev)
-{
-	unsigned i;
-
-	for (i = 0; i < ARRAY_SIZE(debugfs_regs); i++) {
-		if (adev->debugfs_regs[i]) {
-			debugfs_remove(adev->debugfs_regs[i]);
-			adev->debugfs_regs[i] = NULL;
-		}
-	}
-}
-
-static int amdgpu_debugfs_test_ib(struct seq_file *m, void *data)
-{
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
-	struct drm_device *dev = node->minor->dev;
-	struct amdgpu_device *adev = dev->dev_private;
-	int r = 0, i;
-
-	/* hold on the scheduler */
-	for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
-		struct amdgpu_ring *ring = adev->rings[i];
-
-		if (!ring || !ring->sched.thread)
-			continue;
-		kthread_park(ring->sched.thread);
-	}
-
-	seq_printf(m, "run ib test:\n");
-	r = amdgpu_ib_ring_tests(adev);
-	if (r)
-		seq_printf(m, "ib ring tests failed (%d).\n", r);
-	else
-		seq_printf(m, "ib ring tests passed.\n");
-
-	/* go on the scheduler */
-	for (i = 0; i < AMDGPU_MAX_RINGS; i++) {
-		struct amdgpu_ring *ring = adev->rings[i];
-
-		if (!ring || !ring->sched.thread)
-			continue;
-		kthread_unpark(ring->sched.thread);
-	}
-
-	return 0;
-}
-
-static const struct drm_info_list amdgpu_debugfs_test_ib_ring_list[] = {
-	{"amdgpu_test_ib", &amdgpu_debugfs_test_ib}
-};
-
-static int amdgpu_debugfs_test_ib_ring_init(struct amdgpu_device *adev)
-{
-	return amdgpu_debugfs_add_files(adev,
-					amdgpu_debugfs_test_ib_ring_list, 1);
-}
-
-int amdgpu_debugfs_init(struct drm_minor *minor)
-{
-	return 0;
-}
-
-static int amdgpu_debugfs_get_vbios_dump(struct seq_file *m, void *data)
-{
-	struct drm_info_node *node = (struct drm_info_node *) m->private;
-	struct drm_device *dev = node->minor->dev;
-	struct amdgpu_device *adev = dev->dev_private;
-
-	seq_write(m, adev->bios, adev->bios_size);
-	return 0;
-}
-
-static const struct drm_info_list amdgpu_vbios_dump_list[] = {
-		{"amdgpu_vbios",
-		 amdgpu_debugfs_get_vbios_dump,
-		 0, NULL},
-};
-
-static int amdgpu_debugfs_vbios_dump_init(struct amdgpu_device *adev)
-{
-	return amdgpu_debugfs_add_files(adev,
-					amdgpu_vbios_dump_list, 1);
-}
-#else
-static int amdgpu_debugfs_test_ib_ring_init(struct amdgpu_device *adev)
-{
-	return 0;
-}
-static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
-{
-	return 0;
-}
-static int amdgpu_debugfs_vbios_dump_init(struct amdgpu_device *adev)
-{
-	return 0;
-}
-static void amdgpu_debugfs_regs_cleanup(struct amdgpu_device *adev) { }
-#endif

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

@@ -90,7 +90,7 @@ int amdgpu_disp_priority = 0;
 int amdgpu_hw_i2c = 0;
 int amdgpu_pcie_gen2 = -1;
 int amdgpu_msi = -1;
-int amdgpu_lockup_timeout = 0;
+int amdgpu_lockup_timeout = 10000;
 int amdgpu_dpm = -1;
 int amdgpu_fw_load_type = -1;
 int amdgpu_aspm = -1;
@@ -128,6 +128,7 @@ int amdgpu_param_buf_per_se = 0;
 int amdgpu_job_hang_limit = 0;
 int amdgpu_lbpw = -1;
 int amdgpu_compute_multipipe = -1;
+int amdgpu_gpu_recovery = -1; /* auto */
 
 MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
 module_param_named(vramlimit, amdgpu_vram_limit, int, 0600);
@@ -165,7 +166,7 @@ module_param_named(pcie_gen2, amdgpu_pcie_gen2, int, 0444);
 MODULE_PARM_DESC(msi, "MSI support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(msi, amdgpu_msi, int, 0444);
 
-MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default 0 = disable)");
+MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms > 0 (default 10000)");
 module_param_named(lockup_timeout, amdgpu_lockup_timeout, int, 0444);
 
 MODULE_PARM_DESC(dpm, "DPM support (1 = enable, 0 = disable, -1 = auto)");
@@ -280,6 +281,9 @@ module_param_named(lbpw, amdgpu_lbpw, int, 0444);
 MODULE_PARM_DESC(compute_multipipe, "Force compute queues to be spread across pipes (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(compute_multipipe, amdgpu_compute_multipipe, int, 0444);
 
+MODULE_PARM_DESC(gpu_recovery, "Enable GPU recovery mechanism, (1 = enable, 0 = disable, -1 = auto");
+module_param_named(gpu_recovery, amdgpu_gpu_recovery, int, 0444);
+
 #ifdef CONFIG_DRM_AMDGPU_SI
 
 #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE)
@@ -645,7 +649,7 @@ amdgpu_pci_shutdown(struct pci_dev *pdev)
 	 * unfortunately we can't detect certain
 	 * hypervisors so just do this all the time.
 	 */
-	amdgpu_suspend(adev);
+	amdgpu_device_ip_suspend(adev);
 }
 
 static int amdgpu_pmops_suspend(struct device *dev)
@@ -850,9 +854,6 @@ static struct drm_driver kms_driver = {
 	.disable_vblank = amdgpu_disable_vblank_kms,
 	.get_vblank_timestamp = drm_calc_vbltimestamp_from_scanoutpos,
 	.get_scanout_position = amdgpu_get_crtc_scanout_position,
-#if defined(CONFIG_DEBUG_FS)
-	.debugfs_init = amdgpu_debugfs_init,
-#endif
 	.irq_preinstall = amdgpu_irq_preinstall,
 	.irq_postinstall = amdgpu_irq_postinstall,
 	.irq_uninstall = amdgpu_irq_uninstall,
@@ -912,10 +913,6 @@ static int __init amdgpu_init(void)
 	if (r)
 		goto error_fence;
 
-	r = amd_sched_fence_slab_init();
-	if (r)
-		goto error_sched;
-
 	if (vgacon_text_force()) {
 		DRM_ERROR("VGACON disables amdgpu kernel modesetting.\n");
 		return -EINVAL;
@@ -928,9 +925,6 @@ static int __init amdgpu_init(void)
 	/* let modprobe override vga console setting */
 	return pci_register_driver(pdriver);
 
-error_sched:
-	amdgpu_fence_slab_fini();
-
 error_fence:
 	amdgpu_sync_fini();
 
@@ -944,7 +938,6 @@ static void __exit amdgpu_exit(void)
 	pci_unregister_driver(pdriver);
 	amdgpu_unregister_atpx_handler();
 	amdgpu_sync_fini();
-	amd_sched_fence_slab_fini();
 	amdgpu_fence_slab_fini();
 }
 

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

@@ -187,7 +187,7 @@ int amdgpu_fence_emit_polling(struct amdgpu_ring *ring, uint32_t *s)
 
 	seq = ++ring->fence_drv.sync_seq;
 	amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr,
-			       seq, AMDGPU_FENCE_FLAG_INT);
+			       seq, 0);
 
 	*s = seq;
 
@@ -410,7 +410,6 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
 				  unsigned num_hw_submission)
 {
-	long timeout;
 	int r;
 
 	/* Check that num_hw_submission is a power of two */
@@ -434,20 +433,9 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring,
 
 	/* No need to setup the GPU scheduler for KIQ ring */
 	if (ring->funcs->type != AMDGPU_RING_TYPE_KIQ) {
-		timeout = msecs_to_jiffies(amdgpu_lockup_timeout);
-		if (timeout == 0) {
-			/*
-			 * FIXME:
-			 * Delayed workqueue cannot use it directly,
-			 * so the scheduler will not use delayed workqueue if
-			 * MAX_SCHEDULE_TIMEOUT is set.
-			 * Currently keep it simple and silly.
-			 */
-			timeout = MAX_SCHEDULE_TIMEOUT;
-		}
-		r = amd_sched_init(&ring->sched, &amdgpu_sched_ops,
+		r = drm_sched_init(&ring->sched, &amdgpu_sched_ops,
 				   num_hw_submission, amdgpu_job_hang_limit,
-				   timeout, ring->name);
+				   msecs_to_jiffies(amdgpu_lockup_timeout), ring->name);
 		if (r) {
 			DRM_ERROR("Failed to create scheduler on ring %s.\n",
 				  ring->name);
@@ -503,7 +491,7 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev)
 		}
 		amdgpu_irq_put(adev, ring->fence_drv.irq_src,
 			       ring->fence_drv.irq_type);
-		amd_sched_fini(&ring->sched);
+		drm_sched_fini(&ring->sched);
 		del_timer_sync(&ring->fence_drv.fallback_timer);
 		for (j = 0; j <= ring->fence_drv.num_fences_mask; ++j)
 			dma_fence_put(ring->fence_drv.fences[j]);
@@ -705,7 +693,7 @@ static int amdgpu_debugfs_gpu_recover(struct seq_file *m, void *data)
 	struct amdgpu_device *adev = dev->dev_private;
 
 	seq_printf(m, "gpu recover\n");
-	amdgpu_gpu_recover(adev, NULL);
+	amdgpu_device_gpu_recover(adev, NULL, true);
 
 	return 0;
 }

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

@@ -56,6 +56,51 @@
  * Common GART table functions.
  */
 
+/**
+ * amdgpu_dummy_page_init - init dummy page used by the driver
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Allocate the dummy page used by the driver (all asics).
+ * This dummy page is used by the driver as a filler for gart entries
+ * when pages are taken out of the GART
+ * Returns 0 on sucess, -ENOMEM on failure.
+ */
+static int amdgpu_gart_dummy_page_init(struct amdgpu_device *adev)
+{
+	if (adev->dummy_page.page)
+		return 0;
+	adev->dummy_page.page = alloc_page(GFP_DMA32 | GFP_KERNEL | __GFP_ZERO);
+	if (adev->dummy_page.page == NULL)
+		return -ENOMEM;
+	adev->dummy_page.addr = pci_map_page(adev->pdev, adev->dummy_page.page,
+					0, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+	if (pci_dma_mapping_error(adev->pdev, adev->dummy_page.addr)) {
+		dev_err(&adev->pdev->dev, "Failed to DMA MAP the dummy page\n");
+		__free_page(adev->dummy_page.page);
+		adev->dummy_page.page = NULL;
+		return -ENOMEM;
+	}
+	return 0;
+}
+
+/**
+ * amdgpu_dummy_page_fini - free dummy page used by the driver
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Frees the dummy page used by the driver (all asics).
+ */
+static void amdgpu_gart_dummy_page_fini(struct amdgpu_device *adev)
+{
+	if (adev->dummy_page.page == NULL)
+		return;
+	pci_unmap_page(adev->pdev, adev->dummy_page.addr,
+			PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
+	__free_page(adev->dummy_page.page);
+	adev->dummy_page.page = NULL;
+}
+
 /**
  * amdgpu_gart_table_vram_alloc - allocate vram for gart page table
  *
@@ -308,7 +353,7 @@ int amdgpu_gart_init(struct amdgpu_device *adev)
 		DRM_ERROR("Page size is smaller than GPU page size!\n");
 		return -EINVAL;
 	}
-	r = amdgpu_dummy_page_init(adev);
+	r = amdgpu_gart_dummy_page_init(adev);
 	if (r)
 		return r;
 	/* Compute table size */
@@ -340,5 +385,5 @@ void amdgpu_gart_fini(struct amdgpu_device *adev)
 	vfree(adev->gart.pages);
 	adev->gart.pages = NULL;
 #endif
-	amdgpu_dummy_page_fini(adev);
+	amdgpu_gart_dummy_page_fini(adev);
 }

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

@@ -851,7 +851,7 @@ static const struct drm_info_list amdgpu_debugfs_gem_list[] = {
 };
 #endif
 
-int amdgpu_gem_debugfs_init(struct amdgpu_device *adev)
+int amdgpu_debugfs_gem_init(struct amdgpu_device *adev)
 {
 #if defined(CONFIG_DEBUG_FS)
 	return amdgpu_debugfs_add_files(adev, amdgpu_debugfs_gem_list, 1);

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

@@ -203,7 +203,7 @@ int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev,
 
 	spin_lock_init(&kiq->ring_lock);
 
-	r = amdgpu_wb_get(adev, &adev->virt.reg_val_offs);
+	r = amdgpu_device_wb_get(adev, &adev->virt.reg_val_offs);
 	if (r)
 		return r;
 
@@ -229,7 +229,7 @@ int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev,
 void amdgpu_gfx_kiq_free_ring(struct amdgpu_ring *ring,
 			      struct amdgpu_irq_src *irq)
 {
-	amdgpu_wb_free(ring->adev, ring->adev->virt.reg_val_offs);
+	amdgpu_device_wb_free(ring->adev, ring->adev->virt.reg_val_offs);
 	amdgpu_ring_fini(ring);
 }
 

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

@@ -92,15 +92,15 @@ int amdgpu_ih_ring_init(struct amdgpu_device *adev, unsigned ring_size,
 		}
 		return 0;
 	} else {
-		r = amdgpu_wb_get(adev, &adev->irq.ih.wptr_offs);
+		r = amdgpu_device_wb_get(adev, &adev->irq.ih.wptr_offs);
 		if (r) {
 			dev_err(adev->dev, "(%d) ih wptr_offs wb alloc failed\n", r);
 			return r;
 		}
 
-		r = amdgpu_wb_get(adev, &adev->irq.ih.rptr_offs);
+		r = amdgpu_device_wb_get(adev, &adev->irq.ih.rptr_offs);
 		if (r) {
-			amdgpu_wb_free(adev, adev->irq.ih.wptr_offs);
+			amdgpu_device_wb_free(adev, adev->irq.ih.wptr_offs);
 			dev_err(adev->dev, "(%d) ih rptr_offs wb alloc failed\n", r);
 			return r;
 		}
@@ -133,8 +133,8 @@ void amdgpu_ih_ring_fini(struct amdgpu_device *adev)
 		amdgpu_bo_free_kernel(&adev->irq.ih.ring_obj,
 				      &adev->irq.ih.gpu_addr,
 				      (void **)&adev->irq.ih.ring);
-		amdgpu_wb_free(adev, adev->irq.ih.wptr_offs);
-		amdgpu_wb_free(adev, adev->irq.ih.rptr_offs);
+		amdgpu_device_wb_free(adev, adev->irq.ih.wptr_offs);
+		amdgpu_device_wb_free(adev, adev->irq.ih.rptr_offs);
 	}
 }
 

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

@@ -88,7 +88,7 @@ static void amdgpu_irq_reset_work_func(struct work_struct *work)
 						  reset_work);
 
 	if (!amdgpu_sriov_vf(adev))
-		amdgpu_gpu_recover(adev, NULL);
+		amdgpu_device_gpu_recover(adev, NULL, false);
 }
 
 /* Disable *all* interrupts */

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

@@ -28,7 +28,7 @@
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
 
-static void amdgpu_job_timedout(struct amd_sched_job *s_job)
+static void amdgpu_job_timedout(struct drm_sched_job *s_job)
 {
 	struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base);
 
@@ -37,7 +37,7 @@ static void amdgpu_job_timedout(struct amd_sched_job *s_job)
 		  atomic_read(&job->ring->fence_drv.last_seq),
 		  job->ring->fence_drv.sync_seq);
 
-	amdgpu_gpu_recover(job->adev, job);
+	amdgpu_device_gpu_recover(job->adev, job, false);
 }
 
 int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
@@ -96,7 +96,7 @@ void amdgpu_job_free_resources(struct amdgpu_job *job)
 		amdgpu_ib_free(job->adev, &job->ibs[i], f);
 }
 
-static void amdgpu_job_free_cb(struct amd_sched_job *s_job)
+static void amdgpu_job_free_cb(struct drm_sched_job *s_job)
 {
 	struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base);
 
@@ -118,7 +118,7 @@ void amdgpu_job_free(struct amdgpu_job *job)
 }
 
 int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
-		      struct amd_sched_entity *entity, void *owner,
+		      struct drm_sched_entity *entity, void *owner,
 		      struct dma_fence **f)
 {
 	int r;
@@ -127,7 +127,7 @@ int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
 	if (!f)
 		return -EINVAL;
 
-	r = amd_sched_job_init(&job->base, &ring->sched, entity, owner);
+	r = drm_sched_job_init(&job->base, &ring->sched, entity, owner);
 	if (r)
 		return r;
 
@@ -136,13 +136,13 @@ int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
 	*f = dma_fence_get(&job->base.s_fence->finished);
 	amdgpu_job_free_resources(job);
 	amdgpu_ring_priority_get(job->ring, job->base.s_priority);
-	amd_sched_entity_push_job(&job->base, entity);
+	drm_sched_entity_push_job(&job->base, entity);
 
 	return 0;
 }
 
-static struct dma_fence *amdgpu_job_dependency(struct amd_sched_job *sched_job,
-					       struct amd_sched_entity *s_entity)
+static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job,
+					       struct drm_sched_entity *s_entity)
 {
 	struct amdgpu_job *job = to_amdgpu_job(sched_job);
 	struct amdgpu_vm *vm = job->vm;
@@ -151,7 +151,7 @@ static struct dma_fence *amdgpu_job_dependency(struct amd_sched_job *sched_job,
 	struct dma_fence *fence = amdgpu_sync_get_fence(&job->sync, &explicit);
 
 	if (fence && explicit) {
-		if (amd_sched_dependency_optimized(fence, s_entity)) {
+		if (drm_sched_dependency_optimized(fence, s_entity)) {
 			r = amdgpu_sync_fence(job->adev, &job->sched_sync, fence, false);
 			if (r)
 				DRM_ERROR("Error adding fence to sync (%d)\n", r);
@@ -173,7 +173,7 @@ static struct dma_fence *amdgpu_job_dependency(struct amd_sched_job *sched_job,
 	return fence;
 }
 
-static struct dma_fence *amdgpu_job_run(struct amd_sched_job *sched_job)
+static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 {
 	struct dma_fence *fence = NULL, *finished;
 	struct amdgpu_device *adev;
@@ -211,7 +211,7 @@ static struct dma_fence *amdgpu_job_run(struct amd_sched_job *sched_job)
 	return fence;
 }
 
-const struct amd_sched_backend_ops amdgpu_sched_ops = {
+const struct drm_sched_backend_ops amdgpu_sched_ops = {
 	.dependency = amdgpu_job_dependency,
 	.run_job = amdgpu_job_run,
 	.timedout_job = amdgpu_job_timedout,

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

@@ -37,6 +37,18 @@
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
 
+static bool amdgpu_need_backup(struct amdgpu_device *adev)
+{
+	if (adev->flags & AMD_IS_APU)
+		return false;
+
+	if (amdgpu_gpu_recovery == 0 ||
+	    (amdgpu_gpu_recovery == -1  && !amdgpu_sriov_vf(adev)))
+		return false;
+
+	return true;
+}
+
 static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev);
@@ -327,7 +339,12 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
 			       uint64_t init_value,
 			       struct amdgpu_bo **bo_ptr)
 {
-	struct ttm_operation_ctx ctx = { !kernel, false };
+	struct ttm_operation_ctx ctx = {
+		.interruptible = !kernel,
+		.no_wait_gpu = false,
+		.allow_reserved_eviction = true,
+		.resv = resv
+	};
 	struct amdgpu_bo *bo;
 	enum ttm_bo_type type;
 	unsigned long page_align;

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

@@ -1278,16 +1278,16 @@ void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable)
 			/* XXX select vce level based on ring/task */
 			adev->pm.dpm.vce_level = AMD_VCE_LEVEL_AC_ALL;
 			mutex_unlock(&adev->pm.mutex);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							AMD_CG_STATE_UNGATE);
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							AMD_PG_STATE_UNGATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_CG_STATE_UNGATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_PG_STATE_UNGATE);
 			amdgpu_pm_compute_clocks(adev);
 		} else {
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							AMD_PG_STATE_GATE);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							AMD_CG_STATE_GATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_PG_STATE_GATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_CG_STATE_GATE);
 			mutex_lock(&adev->pm.mutex);
 			adev->pm.dpm.vce_active = false;
 			mutex_unlock(&adev->pm.mutex);
@@ -1584,7 +1584,7 @@ static int amdgpu_debugfs_pm_info(struct seq_file *m, void *data)
 	struct drm_device *ddev = adev->ddev;
 	u32 flags = 0;
 
-	amdgpu_get_clockgating_state(adev, &flags);
+	amdgpu_device_ip_get_clockgating_state(adev, &flags);
 	seq_printf(m, "Clock Gating Flags Mask: 0x%x\n", flags);
 	amdgpu_parse_cg_state(m, flags);
 	seq_printf(m, "\n");

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

@@ -164,7 +164,7 @@ void amdgpu_ring_undo(struct amdgpu_ring *ring)
  * Release a request for executing at @priority
  */
 void amdgpu_ring_priority_put(struct amdgpu_ring *ring,
-			      enum amd_sched_priority priority)
+			      enum drm_sched_priority priority)
 {
 	int i;
 
@@ -175,7 +175,7 @@ void amdgpu_ring_priority_put(struct amdgpu_ring *ring,
 		return;
 
 	/* no need to restore if the job is already at the lowest priority */
-	if (priority == AMD_SCHED_PRIORITY_NORMAL)
+	if (priority == DRM_SCHED_PRIORITY_NORMAL)
 		return;
 
 	mutex_lock(&ring->priority_mutex);
@@ -184,8 +184,8 @@ void amdgpu_ring_priority_put(struct amdgpu_ring *ring,
 		goto out_unlock;
 
 	/* decay priority to the next level with a job available */
-	for (i = priority; i >= AMD_SCHED_PRIORITY_MIN; i--) {
-		if (i == AMD_SCHED_PRIORITY_NORMAL
+	for (i = priority; i >= DRM_SCHED_PRIORITY_MIN; i--) {
+		if (i == DRM_SCHED_PRIORITY_NORMAL
 				|| atomic_read(&ring->num_jobs[i])) {
 			ring->priority = i;
 			ring->funcs->set_priority(ring, i);
@@ -206,7 +206,7 @@ out_unlock:
  * Request a ring's priority to be raised to @priority (refcounted).
  */
 void amdgpu_ring_priority_get(struct amdgpu_ring *ring,
-			      enum amd_sched_priority priority)
+			      enum drm_sched_priority priority)
 {
 	if (!ring->funcs->set_priority)
 		return;
@@ -263,25 +263,25 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 			return r;
 	}
 
-	r = amdgpu_wb_get(adev, &ring->rptr_offs);
+	r = amdgpu_device_wb_get(adev, &ring->rptr_offs);
 	if (r) {
 		dev_err(adev->dev, "(%d) ring rptr_offs wb alloc failed\n", r);
 		return r;
 	}
 
-	r = amdgpu_wb_get(adev, &ring->wptr_offs);
+	r = amdgpu_device_wb_get(adev, &ring->wptr_offs);
 	if (r) {
 		dev_err(adev->dev, "(%d) ring wptr_offs wb alloc failed\n", r);
 		return r;
 	}
 
-	r = amdgpu_wb_get(adev, &ring->fence_offs);
+	r = amdgpu_device_wb_get(adev, &ring->fence_offs);
 	if (r) {
 		dev_err(adev->dev, "(%d) ring fence_offs wb alloc failed\n", r);
 		return r;
 	}
 
-	r = amdgpu_wb_get(adev, &ring->cond_exe_offs);
+	r = amdgpu_device_wb_get(adev, &ring->cond_exe_offs);
 	if (r) {
 		dev_err(adev->dev, "(%d) ring cond_exec_polling wb alloc failed\n", r);
 		return r;
@@ -317,12 +317,12 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 	}
 
 	ring->max_dw = max_dw;
-	ring->priority = AMD_SCHED_PRIORITY_NORMAL;
+	ring->priority = DRM_SCHED_PRIORITY_NORMAL;
 	mutex_init(&ring->priority_mutex);
 	INIT_LIST_HEAD(&ring->lru_list);
 	amdgpu_ring_lru_touch(adev, ring);
 
-	for (i = 0; i < AMD_SCHED_PRIORITY_MAX; ++i)
+	for (i = 0; i < DRM_SCHED_PRIORITY_MAX; ++i)
 		atomic_set(&ring->num_jobs[i], 0);
 
 	if (amdgpu_debugfs_ring_init(adev, ring)) {
@@ -348,11 +348,11 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring)
 	if (!(ring->adev) || !(ring->adev->rings[ring->idx]))
 		return;
 
-	amdgpu_wb_free(ring->adev, ring->rptr_offs);
-	amdgpu_wb_free(ring->adev, ring->wptr_offs);
+	amdgpu_device_wb_free(ring->adev, ring->rptr_offs);
+	amdgpu_device_wb_free(ring->adev, ring->wptr_offs);
 
-	amdgpu_wb_free(ring->adev, ring->cond_exe_offs);
-	amdgpu_wb_free(ring->adev, ring->fence_offs);
+	amdgpu_device_wb_free(ring->adev, ring->cond_exe_offs);
+	amdgpu_device_wb_free(ring->adev, ring->fence_offs);
 
 	amdgpu_bo_free_kernel(&ring->ring_obj,
 			      &ring->gpu_addr,

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

@@ -25,7 +25,7 @@
 #define __AMDGPU_RING_H__
 
 #include <drm/amdgpu_drm.h>
-#include "gpu_scheduler.h"
+#include <drm/gpu_scheduler.h>
 
 /* max number of rings */
 #define AMDGPU_MAX_RINGS		18
@@ -154,14 +154,14 @@ struct amdgpu_ring_funcs {
 	void (*emit_tmz)(struct amdgpu_ring *ring, bool start);
 	/* priority functions */
 	void (*set_priority) (struct amdgpu_ring *ring,
-			      enum amd_sched_priority priority);
+			      enum drm_sched_priority priority);
 };
 
 struct amdgpu_ring {
 	struct amdgpu_device		*adev;
 	const struct amdgpu_ring_funcs	*funcs;
 	struct amdgpu_fence_driver	fence_drv;
-	struct amd_gpu_scheduler	sched;
+	struct drm_gpu_scheduler	sched;
 	struct list_head		lru_list;
 
 	struct amdgpu_bo	*ring_obj;
@@ -186,6 +186,7 @@ struct amdgpu_ring {
 	uint64_t                eop_gpu_addr;
 	u32			doorbell_index;
 	bool			use_doorbell;
+	bool			use_pollmem;
 	unsigned		wptr_offs;
 	unsigned		fence_offs;
 	uint64_t		current_ctx;
@@ -196,7 +197,7 @@ struct amdgpu_ring {
 	unsigned		vm_inv_eng;
 	bool			has_compute_vm_bug;
 
-	atomic_t		num_jobs[AMD_SCHED_PRIORITY_MAX];
+	atomic_t		num_jobs[DRM_SCHED_PRIORITY_MAX];
 	struct mutex		priority_mutex;
 	/* protected by priority_mutex */
 	int			priority;
@@ -212,9 +213,9 @@ void amdgpu_ring_generic_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib);
 void amdgpu_ring_commit(struct amdgpu_ring *ring);
 void amdgpu_ring_undo(struct amdgpu_ring *ring);
 void amdgpu_ring_priority_get(struct amdgpu_ring *ring,
-			      enum amd_sched_priority priority);
+			      enum drm_sched_priority priority);
 void amdgpu_ring_priority_put(struct amdgpu_ring *ring,
-			      enum amd_sched_priority priority);
+			      enum drm_sched_priority priority);
 int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 		     unsigned ring_size, struct amdgpu_irq_src *irq_src,
 		     unsigned irq_type);

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

@@ -29,29 +29,29 @@
 
 #include "amdgpu_vm.h"
 
-enum amd_sched_priority amdgpu_to_sched_priority(int amdgpu_priority)
+enum drm_sched_priority amdgpu_to_sched_priority(int amdgpu_priority)
 {
 	switch (amdgpu_priority) {
 	case AMDGPU_CTX_PRIORITY_VERY_HIGH:
-		return AMD_SCHED_PRIORITY_HIGH_HW;
+		return DRM_SCHED_PRIORITY_HIGH_HW;
 	case AMDGPU_CTX_PRIORITY_HIGH:
-		return AMD_SCHED_PRIORITY_HIGH_SW;
+		return DRM_SCHED_PRIORITY_HIGH_SW;
 	case AMDGPU_CTX_PRIORITY_NORMAL:
-		return AMD_SCHED_PRIORITY_NORMAL;
+		return DRM_SCHED_PRIORITY_NORMAL;
 	case AMDGPU_CTX_PRIORITY_LOW:
 	case AMDGPU_CTX_PRIORITY_VERY_LOW:
-		return AMD_SCHED_PRIORITY_LOW;
+		return DRM_SCHED_PRIORITY_LOW;
 	case AMDGPU_CTX_PRIORITY_UNSET:
-		return AMD_SCHED_PRIORITY_UNSET;
+		return DRM_SCHED_PRIORITY_UNSET;
 	default:
 		WARN(1, "Invalid context priority %d\n", amdgpu_priority);
-		return AMD_SCHED_PRIORITY_INVALID;
+		return DRM_SCHED_PRIORITY_INVALID;
 	}
 }
 
 static int amdgpu_sched_process_priority_override(struct amdgpu_device *adev,
 						  int fd,
-						  enum amd_sched_priority priority)
+						  enum drm_sched_priority priority)
 {
 	struct file *filp = fcheck(fd);
 	struct drm_file *file;
@@ -86,11 +86,11 @@ int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
 {
 	union drm_amdgpu_sched *args = data;
 	struct amdgpu_device *adev = dev->dev_private;
-	enum amd_sched_priority priority;
+	enum drm_sched_priority priority;
 	int r;
 
 	priority = amdgpu_to_sched_priority(args->in.priority);
-	if (args->in.flags || priority == AMD_SCHED_PRIORITY_INVALID)
+	if (args->in.flags || priority == DRM_SCHED_PRIORITY_INVALID)
 		return -EINVAL;
 
 	switch (args->in.op) {

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

@@ -27,7 +27,7 @@
 
 #include <drm/drmP.h>
 
-enum amd_sched_priority amdgpu_to_sched_priority(int amdgpu_priority);
+enum drm_sched_priority amdgpu_to_sched_priority(int amdgpu_priority);
 int amdgpu_sched_ioctl(struct drm_device *dev, void *data,
 		       struct drm_file *filp);
 

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

@@ -64,7 +64,7 @@ void amdgpu_sync_create(struct amdgpu_sync *sync)
 static bool amdgpu_sync_same_dev(struct amdgpu_device *adev,
 				 struct dma_fence *f)
 {
-	struct amd_sched_fence *s_fence = to_amd_sched_fence(f);
+	struct drm_sched_fence *s_fence = to_drm_sched_fence(f);
 
 	if (s_fence) {
 		struct amdgpu_ring *ring;
@@ -85,7 +85,7 @@ static bool amdgpu_sync_same_dev(struct amdgpu_device *adev,
  */
 static void *amdgpu_sync_get_owner(struct dma_fence *f)
 {
-	struct amd_sched_fence *s_fence = to_amd_sched_fence(f);
+	struct drm_sched_fence *s_fence = to_drm_sched_fence(f);
 
 	if (s_fence)
 		return s_fence->owner;
@@ -120,7 +120,7 @@ static void amdgpu_sync_keep_later(struct dma_fence **keep,
  * Tries to add the fence to an existing hash entry. Returns true when an entry
  * was found, false otherwise.
  */
-static bool amdgpu_sync_add_later(struct amdgpu_sync *sync, struct dma_fence *f)
+static bool amdgpu_sync_add_later(struct amdgpu_sync *sync, struct dma_fence *f, bool explicit)
 {
 	struct amdgpu_sync_entry *e;
 
@@ -129,6 +129,10 @@ static bool amdgpu_sync_add_later(struct amdgpu_sync *sync, struct dma_fence *f)
 			continue;
 
 		amdgpu_sync_keep_later(&e->fence, f);
+
+		/* Preserve eplicit flag to not loose pipe line sync */
+		e->explicit |= explicit;
+
 		return true;
 	}
 	return false;
@@ -148,12 +152,11 @@ int amdgpu_sync_fence(struct amdgpu_device *adev, struct amdgpu_sync *sync,
 
 	if (!f)
 		return 0;
-
 	if (amdgpu_sync_same_dev(adev, f) &&
 	    amdgpu_sync_get_owner(f) == AMDGPU_FENCE_OWNER_VM)
 		amdgpu_sync_keep_later(&sync->last_vm_update, f);
 
-	if (amdgpu_sync_add_later(sync, f))
+	if (amdgpu_sync_add_later(sync, f, explicit))
 		return 0;
 
 	e = kmem_cache_alloc(amdgpu_sync_slab, GFP_KERNEL);
@@ -245,7 +248,7 @@ struct dma_fence *amdgpu_sync_peek_fence(struct amdgpu_sync *sync,
 
 	hash_for_each_safe(sync->fences, i, tmp, e, node) {
 		struct dma_fence *f = e->fence;
-		struct amd_sched_fence *s_fence = to_amd_sched_fence(f);
+		struct drm_sched_fence *s_fence = to_drm_sched_fence(f);
 
 		if (dma_fence_is_signaled(f)) {
 			hash_del(&e->node);

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

@@ -76,7 +76,7 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 {
 	struct drm_global_reference *global_ref;
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	int r;
 
 	adev->mman.mem_global_referenced = false;
@@ -108,8 +108,8 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 	mutex_init(&adev->mman.gtt_window_lock);
 
 	ring = adev->mman.buffer_funcs_ring;
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_KERNEL];
-	r = amd_sched_entity_init(&ring->sched, &adev->mman.entity,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
+	r = drm_sched_entity_init(&ring->sched, &adev->mman.entity,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r) {
 		DRM_ERROR("Failed setting up TTM BO move run queue.\n");
@@ -131,7 +131,7 @@ error_mem:
 static void amdgpu_ttm_global_fini(struct amdgpu_device *adev)
 {
 	if (adev->mman.mem_global_referenced) {
-		amd_sched_entity_fini(adev->mman.entity.sched,
+		drm_sched_entity_fini(adev->mman.entity.sched,
 				      &adev->mman.entity);
 		mutex_destroy(&adev->mman.gtt_window_lock);
 		drm_global_item_unref(&adev->mman.bo_global_ref.ref);
@@ -505,7 +505,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
-	r = ttm_bo_move_ttm(bo, ctx->interruptible, ctx->no_wait_gpu, new_mem);
+	r = ttm_bo_move_ttm(bo, ctx, new_mem);
 out_cleanup:
 	ttm_bo_mem_put(bo, &tmp_mem);
 	return r;
@@ -536,7 +536,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	if (unlikely(r)) {
 		return r;
 	}
-	r = ttm_bo_move_ttm(bo, ctx->interruptible, ctx->no_wait_gpu, &tmp_mem);
+	r = ttm_bo_move_ttm(bo, ctx, &tmp_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
@@ -597,8 +597,7 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict,
 
 	if (r) {
 memcpy:
-		r = ttm_bo_move_memcpy(bo, ctx->interruptible,
-				       ctx->no_wait_gpu, new_mem);
+		r = ttm_bo_move_memcpy(bo, ctx, new_mem);
 		if (r) {
 			return r;
 		}
@@ -1270,6 +1269,101 @@ static struct ttm_bo_driver amdgpu_bo_driver = {
 	.access_memory = &amdgpu_ttm_access_memory
 };
 
+/*
+ * Firmware Reservation functions
+ */
+/**
+ * amdgpu_ttm_fw_reserve_vram_fini - free fw reserved vram
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * free fw reserved vram if it has been reserved.
+ */
+static void amdgpu_ttm_fw_reserve_vram_fini(struct amdgpu_device *adev)
+{
+	amdgpu_bo_free_kernel(&adev->fw_vram_usage.reserved_bo,
+		NULL, &adev->fw_vram_usage.va);
+}
+
+/**
+ * amdgpu_ttm_fw_reserve_vram_init - create bo vram reservation from fw
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * create bo vram reservation from fw.
+ */
+static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
+{
+	struct ttm_operation_ctx ctx = { false, false };
+	int r = 0;
+	int i;
+	u64 vram_size = adev->mc.visible_vram_size;
+	u64 offset = adev->fw_vram_usage.start_offset;
+	u64 size = adev->fw_vram_usage.size;
+	struct amdgpu_bo *bo;
+
+	adev->fw_vram_usage.va = NULL;
+	adev->fw_vram_usage.reserved_bo = NULL;
+
+	if (adev->fw_vram_usage.size > 0 &&
+		adev->fw_vram_usage.size <= vram_size) {
+
+		r = amdgpu_bo_create(adev, adev->fw_vram_usage.size,
+			PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM,
+			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+			AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS, NULL, NULL, 0,
+			&adev->fw_vram_usage.reserved_bo);
+		if (r)
+			goto error_create;
+
+		r = amdgpu_bo_reserve(adev->fw_vram_usage.reserved_bo, false);
+		if (r)
+			goto error_reserve;
+
+		/* remove the original mem node and create a new one at the
+		 * request position
+		 */
+		bo = adev->fw_vram_usage.reserved_bo;
+		offset = ALIGN(offset, PAGE_SIZE);
+		for (i = 0; i < bo->placement.num_placement; ++i) {
+			bo->placements[i].fpfn = offset >> PAGE_SHIFT;
+			bo->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
+		}
+
+		ttm_bo_mem_put(&bo->tbo, &bo->tbo.mem);
+		r = ttm_bo_mem_space(&bo->tbo, &bo->placement,
+				     &bo->tbo.mem, &ctx);
+		if (r)
+			goto error_pin;
+
+		r = amdgpu_bo_pin_restricted(adev->fw_vram_usage.reserved_bo,
+			AMDGPU_GEM_DOMAIN_VRAM,
+			adev->fw_vram_usage.start_offset,
+			(adev->fw_vram_usage.start_offset +
+			adev->fw_vram_usage.size), NULL);
+		if (r)
+			goto error_pin;
+		r = amdgpu_bo_kmap(adev->fw_vram_usage.reserved_bo,
+			&adev->fw_vram_usage.va);
+		if (r)
+			goto error_kmap;
+
+		amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
+	}
+	return r;
+
+error_kmap:
+	amdgpu_bo_unpin(adev->fw_vram_usage.reserved_bo);
+error_pin:
+	amdgpu_bo_unreserve(adev->fw_vram_usage.reserved_bo);
+error_reserve:
+	amdgpu_bo_unref(&adev->fw_vram_usage.reserved_bo);
+error_create:
+	adev->fw_vram_usage.va = NULL;
+	adev->fw_vram_usage.reserved_bo = NULL;
+	return r;
+}
+
 int amdgpu_ttm_init(struct amdgpu_device *adev)
 {
 	uint64_t gtt_size;
@@ -1312,7 +1406,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	 *The reserved vram for firmware must be pinned to the specified
 	 *place on the VRAM, so reserve it early.
 	 */
-	r = amdgpu_fw_reserve_vram_init(adev);
+	r = amdgpu_ttm_fw_reserve_vram_init(adev);
 	if (r) {
 		return r;
 	}
@@ -1330,9 +1424,11 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		struct sysinfo si;
 
 		si_meminfo(&si);
-		gtt_size = max(AMDGPU_DEFAULT_GTT_SIZE_MB << 20,
-			(uint64_t)si.totalram * si.mem_unit * 3/4);
-	} else
+		gtt_size = min(max((AMDGPU_DEFAULT_GTT_SIZE_MB << 20),
+			       adev->mc.mc_vram_size),
+			       ((uint64_t)si.totalram * si.mem_unit * 3/4));
+	}
+	else
 		gtt_size = (uint64_t)amdgpu_gtt_size << 20;
 	r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_TT, gtt_size >> PAGE_SHIFT);
 	if (r) {
@@ -1396,7 +1492,7 @@ void amdgpu_ttm_fini(struct amdgpu_device *adev)
 
 	amdgpu_ttm_debugfs_fini(adev);
 	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
-	amdgpu_fw_reserve_vram_fini(adev);
+	amdgpu_ttm_fw_reserve_vram_fini(adev);
 
 	ttm_bo_clean_mm(&adev->mman.bdev, TTM_PL_VRAM);
 	ttm_bo_clean_mm(&adev->mman.bdev, TTM_PL_TT);

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

@@ -25,7 +25,7 @@
 #define __AMDGPU_TTM_H__
 
 #include "amdgpu.h"
-#include "gpu_scheduler.h"
+#include <drm/gpu_scheduler.h>
 
 #define AMDGPU_PL_GDS		(TTM_PL_PRIV + 0)
 #define AMDGPU_PL_GWS		(TTM_PL_PRIV + 1)
@@ -55,7 +55,7 @@ struct amdgpu_mman {
 
 	struct mutex				gtt_window_lock;
 	/* Scheduler entity for buffer moves */
-	struct amd_sched_entity			entity;
+	struct drm_sched_entity			entity;
 };
 
 struct amdgpu_copy_mem {

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

@@ -116,7 +116,7 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	unsigned long bo_size;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
@@ -230,8 +230,8 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	}
 
 	ring = &adev->uvd.ring;
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-	r = amd_sched_entity_init(&ring->sched, &adev->uvd.entity,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up UVD run queue.\n");
@@ -244,7 +244,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	}
 
 	/* from uvd v5.0 HW addressing capacity increased to 64 bits */
-	if (!amdgpu_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
+	if (!amdgpu_device_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
 		adev->uvd.address_64_bit = true;
 
 	switch (adev->asic_type) {
@@ -272,7 +272,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 	int i;
 	kfree(adev->uvd.saved_bo);
 
-	amd_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);
+	drm_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);
 
 	amdgpu_bo_free_kernel(&adev->uvd.vcpu_bo,
 			      &adev->uvd.gpu_addr,
@@ -297,6 +297,8 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 	if (adev->uvd.vcpu_bo == NULL)
 		return 0;
 
+	cancel_delayed_work_sync(&adev->uvd.idle_work);
+
 	for (i = 0; i < adev->uvd.max_handles; ++i)
 		if (atomic_read(&adev->uvd.handles[i]))
 			break;
@@ -304,8 +306,6 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 	if (i == AMDGPU_MAX_UVD_HANDLES)
 		return 0;
 
-	cancel_delayed_work_sync(&adev->uvd.idle_work);
-
 	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
 	ptr = adev->uvd.cpu_addr;
 
@@ -346,6 +346,8 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 			ptr += le32_to_cpu(hdr->ucode_size_bytes);
 		}
 		memset_io(ptr, 0, size);
+		/* to restore uvd fence seq */
+		amdgpu_fence_driver_force_completion(&adev->uvd.ring);
 	}
 
 	return 0;
@@ -1153,10 +1155,10 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 		} else {
 			amdgpu_asic_set_uvd_clocks(adev, 0, 0);
 			/* shutdown the UVD block */
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							    AMD_PG_STATE_GATE);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							    AMD_CG_STATE_GATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							       AMD_PG_STATE_GATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							       AMD_CG_STATE_GATE);
 		}
 	} else {
 		schedule_delayed_work(&adev->uvd.idle_work, UVD_IDLE_TIMEOUT);
@@ -1176,10 +1178,10 @@ void amdgpu_uvd_ring_begin_use(struct amdgpu_ring *ring)
 			amdgpu_dpm_enable_uvd(adev, true);
 		} else {
 			amdgpu_asic_set_uvd_clocks(adev, 53300, 40000);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							    AMD_CG_STATE_UNGATE);
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							    AMD_PG_STATE_UNGATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							       AMD_CG_STATE_UNGATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							       AMD_PG_STATE_UNGATE);
 		}
 	}
 }

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

@@ -51,8 +51,8 @@ struct amdgpu_uvd {
 	struct amdgpu_irq_src	irq;
 	bool			address_64_bit;
 	bool			use_ctx_buf;
-	struct amd_sched_entity entity;
-	struct amd_sched_entity entity_enc;
+	struct drm_sched_entity entity;
+	struct drm_sched_entity entity_enc;
 	uint32_t                srbm_soft_reset;
 	unsigned		num_enc_rings;
 };

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

@@ -85,7 +85,7 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 {
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
 	unsigned ucode_version, version_major, version_minor, binary_id;
@@ -174,8 +174,8 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	}
 
 	ring = &adev->vce.ring[0];
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-	r = amd_sched_entity_init(&ring->sched, &adev->vce.entity,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+	r = drm_sched_entity_init(&ring->sched, &adev->vce.entity,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCE run queue.\n");
@@ -207,7 +207,7 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
 	if (adev->vce.vcpu_bo == NULL)
 		return 0;
 
-	amd_sched_entity_fini(&adev->vce.ring[0].sched, &adev->vce.entity);
+	drm_sched_entity_fini(&adev->vce.ring[0].sched, &adev->vce.entity);
 
 	amdgpu_bo_free_kernel(&adev->vce.vcpu_bo, &adev->vce.gpu_addr,
 		(void **)&adev->vce.cpu_addr);
@@ -311,10 +311,10 @@ static void amdgpu_vce_idle_work_handler(struct work_struct *work)
 			amdgpu_dpm_enable_vce(adev, false);
 		} else {
 			amdgpu_asic_set_vce_clocks(adev, 0, 0);
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							    AMD_PG_STATE_GATE);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							    AMD_CG_STATE_GATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_PG_STATE_GATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_CG_STATE_GATE);
 		}
 	} else {
 		schedule_delayed_work(&adev->vce.idle_work, VCE_IDLE_TIMEOUT);
@@ -343,10 +343,10 @@ void amdgpu_vce_ring_begin_use(struct amdgpu_ring *ring)
 			amdgpu_dpm_enable_vce(adev, true);
 		} else {
 			amdgpu_asic_set_vce_clocks(adev, 53300, 40000);
-			amdgpu_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							    AMD_CG_STATE_UNGATE);
-			amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
-							    AMD_PG_STATE_UNGATE);
+			amdgpu_device_ip_set_clockgating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_CG_STATE_UNGATE);
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCE,
+							       AMD_PG_STATE_UNGATE);
 
 		}
 	}

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

@@ -46,7 +46,7 @@ struct amdgpu_vce {
 	struct amdgpu_ring	ring[AMDGPU_MAX_VCE_RINGS];
 	struct amdgpu_irq_src	irq;
 	unsigned		harvest_config;
-	struct amd_sched_entity	entity;
+	struct drm_sched_entity	entity;
 	uint32_t                srbm_soft_reset;
 	unsigned		num_rings;
 };

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

@@ -35,7 +35,6 @@
 #include "soc15d.h"
 #include "soc15_common.h"
 
-#include "soc15ip.h"
 #include "vcn/vcn_1_0_offset.h"
 
 /* 1 second timeout */
@@ -51,7 +50,7 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work);
 int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 {
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	unsigned long bo_size;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
@@ -104,8 +103,8 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	}
 
 	ring = &adev->vcn.ring_dec;
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-	r = amd_sched_entity_init(&ring->sched, &adev->vcn.entity_dec,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_dec,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCN dec run queue.\n");
@@ -113,8 +112,8 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	}
 
 	ring = &adev->vcn.ring_enc[0];
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-	r = amd_sched_entity_init(&ring->sched, &adev->vcn.entity_enc,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_enc,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCN enc run queue.\n");
@@ -130,9 +129,9 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 
 	kfree(adev->vcn.saved_bo);
 
-	amd_sched_entity_fini(&adev->vcn.ring_dec.sched, &adev->vcn.entity_dec);
+	drm_sched_entity_fini(&adev->vcn.ring_dec.sched, &adev->vcn.entity_dec);
 
-	amd_sched_entity_fini(&adev->vcn.ring_enc[0].sched, &adev->vcn.entity_enc);
+	drm_sched_entity_fini(&adev->vcn.ring_enc[0].sched, &adev->vcn.entity_enc);
 
 	amdgpu_bo_free_kernel(&adev->vcn.vcpu_bo,
 			      &adev->vcn.gpu_addr,

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

@@ -56,8 +56,8 @@ struct amdgpu_vcn {
 	struct amdgpu_ring	ring_dec;
 	struct amdgpu_ring	ring_enc[AMDGPU_VCN_MAX_ENC_RINGS];
 	struct amdgpu_irq_src	irq;
-	struct amd_sched_entity entity_dec;
-	struct amd_sched_entity entity_enc;
+	struct drm_sched_entity entity_dec;
+	struct drm_sched_entity entity_enc;
 	unsigned		num_enc_rings;
 };
 

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

@@ -148,12 +148,23 @@ struct amdgpu_prt_cb {
 static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev,
 				      unsigned level)
 {
-	if (level != adev->vm_manager.num_level)
-		return 9 * (adev->vm_manager.num_level - level - 1) +
+	unsigned shift = 0xff;
+
+	switch (level) {
+	case AMDGPU_VM_PDB2:
+	case AMDGPU_VM_PDB1:
+	case AMDGPU_VM_PDB0:
+		shift = 9 * (AMDGPU_VM_PDB0 - level) +
 			adev->vm_manager.block_size;
-	else
-		/* For the page tables on the leaves */
-		return 0;
+		break;
+	case AMDGPU_VM_PTB:
+		shift = 0;
+		break;
+	default:
+		dev_err(adev->dev, "the level%d isn't supported.\n", level);
+	}
+
+	return shift;
 }
 
 /**
@@ -166,12 +177,13 @@ static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev,
 static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev,
 				      unsigned level)
 {
-	unsigned shift = amdgpu_vm_level_shift(adev, 0);
+	unsigned shift = amdgpu_vm_level_shift(adev,
+					       adev->vm_manager.root_level);
 
-	if (level == 0)
+	if (level == adev->vm_manager.root_level)
 		/* For the root directory */
 		return round_up(adev->vm_manager.max_pfn, 1 << shift) >> shift;
-	else if (level != adev->vm_manager.num_level)
+	else if (level != AMDGPU_VM_PTB)
 		/* Everything in between */
 		return 512;
 	else
@@ -329,9 +341,6 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 	    to >= amdgpu_vm_num_entries(adev, level))
 		return -EINVAL;
 
-	if (to > parent->last_entry_used)
-		parent->last_entry_used = to;
-
 	++level;
 	saddr = saddr & ((1 << shift) - 1);
 	eaddr = eaddr & ((1 << shift) - 1);
@@ -346,7 +355,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 
 	if (vm->pte_support_ats) {
 		init_value = AMDGPU_PTE_DEFAULT_ATC;
-		if (level != adev->vm_manager.num_level - 1)
+		if (level != AMDGPU_VM_PTB)
 			init_value |= AMDGPU_PDE_PTE;
 
 	}
@@ -386,10 +395,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 			spin_lock(&vm->status_lock);
 			list_add(&entry->base.vm_status, &vm->relocated);
 			spin_unlock(&vm->status_lock);
-			entry->addr = 0;
 		}
 
-		if (level < adev->vm_manager.num_level) {
+		if (level < AMDGPU_VM_PTB) {
 			uint64_t sub_saddr = (pt_idx == from) ? saddr : 0;
 			uint64_t sub_eaddr = (pt_idx == to) ? eaddr :
 				((1 << shift) - 1);
@@ -435,7 +443,8 @@ int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
 	saddr /= AMDGPU_GPU_PAGE_SIZE;
 	eaddr /= AMDGPU_GPU_PAGE_SIZE;
 
-	return amdgpu_vm_alloc_levels(adev, vm, &vm->root, saddr, eaddr, 0);
+	return amdgpu_vm_alloc_levels(adev, vm, &vm->root, saddr, eaddr,
+				      adev->vm_manager.root_level);
 }
 
 /**
@@ -732,7 +741,7 @@ void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev)
 
 	has_compute_vm_bug = false;
 
-	ip_block = amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GFX);
+	ip_block = amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GFX);
 	if (ip_block) {
 		/* Compute has a VM bug for GFX version < 7.
 		   Compute has a VM bug for GFX 8 MEC firmware version < 673.*/
@@ -1060,162 +1069,52 @@ static int amdgpu_vm_wait_pd(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 }
 
 /*
- * amdgpu_vm_update_level - update a single level in the hierarchy
+ * amdgpu_vm_update_pde - update a single level in the hierarchy
  *
- * @adev: amdgpu_device pointer
+ * @param: parameters for the update
  * @vm: requested vm
  * @parent: parent directory
+ * @entry: entry to update
  *
- * Makes sure all entries in @parent are up to date.
- * Returns 0 for success, error for failure.
+ * Makes sure the requested entry in parent is up to date.
  */
-static int amdgpu_vm_update_level(struct amdgpu_device *adev,
-				  struct amdgpu_vm *vm,
-				  struct amdgpu_vm_pt *parent)
+static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params,
+				 struct amdgpu_vm *vm,
+				 struct amdgpu_vm_pt *parent,
+				 struct amdgpu_vm_pt *entry)
 {
-	struct amdgpu_bo *shadow;
-	struct amdgpu_ring *ring = NULL;
+	struct amdgpu_bo *bo = entry->base.bo, *shadow = NULL, *pbo;
 	uint64_t pd_addr, shadow_addr = 0;
-	uint64_t last_pde = ~0, last_pt = ~0, last_shadow = ~0;
-	unsigned count = 0, pt_idx, ndw = 0;
-	struct amdgpu_job *job;
-	struct amdgpu_pte_update_params params;
-	struct dma_fence *fence = NULL;
-	uint32_t incr;
-
-	int r;
+	uint64_t pde, pt, flags;
+	unsigned level;
 
-	if (!parent->entries)
-		return 0;
-
-	memset(&params, 0, sizeof(params));
-	params.adev = adev;
-	shadow = parent->base.bo->shadow;
+	/* Don't update huge pages here */
+	if (entry->huge)
+		return;
 
 	if (vm->use_cpu_for_update) {
 		pd_addr = (unsigned long)amdgpu_bo_kptr(parent->base.bo);
-		r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM);
-		if (unlikely(r))
-			return r;
-
-		params.func = amdgpu_vm_cpu_set_ptes;
 	} else {
-		ring = container_of(vm->entity.sched, struct amdgpu_ring,
-				    sched);
-
-		/* padding, etc. */
-		ndw = 64;
-
-		/* assume the worst case */
-		ndw += parent->last_entry_used * 6;
-
 		pd_addr = amdgpu_bo_gpu_offset(parent->base.bo);
-
-		if (shadow) {
+		shadow = parent->base.bo->shadow;
+		if (shadow)
 			shadow_addr = amdgpu_bo_gpu_offset(shadow);
-			ndw *= 2;
-		} else {
-			shadow_addr = 0;
-		}
-
-		r = amdgpu_job_alloc_with_ib(adev, ndw * 4, &job);
-		if (r)
-			return r;
-
-		params.ib = &job->ibs[0];
-		params.func = amdgpu_vm_do_set_ptes;
 	}
 
+	for (level = 0, pbo = parent->base.bo->parent; pbo; ++level)
+		pbo = pbo->parent;
 
-	/* walk over the address space and update the directory */
-	for (pt_idx = 0; pt_idx <= parent->last_entry_used; ++pt_idx) {
-		struct amdgpu_vm_pt *entry = &parent->entries[pt_idx];
-		struct amdgpu_bo *bo = entry->base.bo;
-		uint64_t pde, pt;
-
-		if (bo == NULL)
-			continue;
-
-		spin_lock(&vm->status_lock);
-		list_del_init(&entry->base.vm_status);
-		spin_unlock(&vm->status_lock);
-
-		pt = amdgpu_bo_gpu_offset(bo);
-		pt = amdgpu_gart_get_vm_pde(adev, pt);
-		/* Don't update huge pages here */
-		if ((parent->entries[pt_idx].addr & AMDGPU_PDE_PTE) ||
-		    parent->entries[pt_idx].addr == (pt | AMDGPU_PTE_VALID))
-			continue;
-
-		parent->entries[pt_idx].addr = pt | AMDGPU_PTE_VALID;
-
-		pde = pd_addr + pt_idx * 8;
-		incr = amdgpu_bo_size(bo);
-		if (((last_pde + 8 * count) != pde) ||
-		    ((last_pt + incr * count) != pt) ||
-		    (count == AMDGPU_VM_MAX_UPDATE_SIZE)) {
-
-			if (count) {
-				if (shadow)
-					params.func(&params,
-						    last_shadow,
-						    last_pt, count,
-						    incr,
-						    AMDGPU_PTE_VALID);
-
-				params.func(&params, last_pde,
-					    last_pt, count, incr,
-					    AMDGPU_PTE_VALID);
-			}
-
-			count = 1;
-			last_pde = pde;
-			last_shadow = shadow_addr + pt_idx * 8;
-			last_pt = pt;
-		} else {
-			++count;
-		}
-	}
-
-	if (count) {
-		if (vm->root.base.bo->shadow)
-			params.func(&params, last_shadow, last_pt,
-				    count, incr, AMDGPU_PTE_VALID);
-
-		params.func(&params, last_pde, last_pt,
-			    count, incr, AMDGPU_PTE_VALID);
-	}
-
-	if (!vm->use_cpu_for_update) {
-		if (params.ib->length_dw == 0) {
-			amdgpu_job_free(job);
-		} else {
-			amdgpu_ring_pad_ib(ring, params.ib);
-			amdgpu_sync_resv(adev, &job->sync,
-					 parent->base.bo->tbo.resv,
-					 AMDGPU_FENCE_OWNER_VM, false);
-			if (shadow)
-				amdgpu_sync_resv(adev, &job->sync,
-						 shadow->tbo.resv,
-						 AMDGPU_FENCE_OWNER_VM, false);
-
-			WARN_ON(params.ib->length_dw > ndw);
-			r = amdgpu_job_submit(job, ring, &vm->entity,
-					AMDGPU_FENCE_OWNER_VM, &fence);
-			if (r)
-				goto error_free;
-
-			amdgpu_bo_fence(parent->base.bo, fence, true);
-			dma_fence_put(vm->last_update);
-			vm->last_update = fence;
-		}
+	level += params->adev->vm_manager.root_level;
+	pt = amdgpu_bo_gpu_offset(bo);
+	flags = AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(params->adev, level, &pt, &flags);
+	if (shadow) {
+		pde = shadow_addr + (entry - parent->entries) * 8;
+		params->func(params, pde, pt, 1, 0, flags);
 	}
 
-	return 0;
-
-error_free:
-	amdgpu_job_free(job);
-	return r;
+	pde = pd_addr + (entry - parent->entries) * 8;
+	params->func(params, pde, pt, 1, 0, flags);
 }
 
 /*
@@ -1225,27 +1124,29 @@ error_free:
  *
  * Mark all PD level as invalid after an error.
  */
-static void amdgpu_vm_invalidate_level(struct amdgpu_vm *vm,
-				       struct amdgpu_vm_pt *parent)
+static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev,
+				       struct amdgpu_vm *vm,
+				       struct amdgpu_vm_pt *parent,
+				       unsigned level)
 {
-	unsigned pt_idx;
+	unsigned pt_idx, num_entries;
 
 	/*
 	 * Recurse into the subdirectories. This recursion is harmless because
 	 * we only have a maximum of 5 layers.
 	 */
-	for (pt_idx = 0; pt_idx <= parent->last_entry_used; ++pt_idx) {
+	num_entries = amdgpu_vm_num_entries(adev, level);
+	for (pt_idx = 0; pt_idx < num_entries; ++pt_idx) {
 		struct amdgpu_vm_pt *entry = &parent->entries[pt_idx];
 
 		if (!entry->base.bo)
 			continue;
 
-		entry->addr = ~0ULL;
 		spin_lock(&vm->status_lock);
 		if (list_empty(&entry->base.vm_status))
 			list_add(&entry->base.vm_status, &vm->relocated);
 		spin_unlock(&vm->status_lock);
-		amdgpu_vm_invalidate_level(vm, entry);
+		amdgpu_vm_invalidate_level(adev, vm, entry, level + 1);
 	}
 }
 
@@ -1261,38 +1162,63 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_vm *vm,
 int amdgpu_vm_update_directories(struct amdgpu_device *adev,
 				 struct amdgpu_vm *vm)
 {
+	struct amdgpu_pte_update_params params;
+	struct amdgpu_job *job;
+	unsigned ndw = 0;
 	int r = 0;
 
+	if (list_empty(&vm->relocated))
+		return 0;
+
+restart:
+	memset(&params, 0, sizeof(params));
+	params.adev = adev;
+
+	if (vm->use_cpu_for_update) {
+		r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM);
+		if (unlikely(r))
+			return r;
+
+		params.func = amdgpu_vm_cpu_set_ptes;
+	} else {
+		ndw = 512 * 8;
+		r = amdgpu_job_alloc_with_ib(adev, ndw * 4, &job);
+		if (r)
+			return r;
+
+		params.ib = &job->ibs[0];
+		params.func = amdgpu_vm_do_set_ptes;
+	}
+
 	spin_lock(&vm->status_lock);
 	while (!list_empty(&vm->relocated)) {
-		struct amdgpu_vm_bo_base *bo_base;
+		struct amdgpu_vm_bo_base *bo_base, *parent;
+		struct amdgpu_vm_pt *pt, *entry;
 		struct amdgpu_bo *bo;
 
 		bo_base = list_first_entry(&vm->relocated,
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
+		list_del_init(&bo_base->vm_status);
 		spin_unlock(&vm->status_lock);
 
 		bo = bo_base->bo->parent;
-		if (bo) {
-			struct amdgpu_vm_bo_base *parent;
-			struct amdgpu_vm_pt *pt;
-
-			parent = list_first_entry(&bo->va,
-						  struct amdgpu_vm_bo_base,
-						  bo_list);
-			pt = container_of(parent, struct amdgpu_vm_pt, base);
-
-			r = amdgpu_vm_update_level(adev, vm, pt);
-			if (r) {
-				amdgpu_vm_invalidate_level(vm, &vm->root);
-				return r;
-			}
-			spin_lock(&vm->status_lock);
-		} else {
+		if (!bo) {
 			spin_lock(&vm->status_lock);
-			list_del_init(&bo_base->vm_status);
+			continue;
 		}
+
+		parent = list_first_entry(&bo->va, struct amdgpu_vm_bo_base,
+					  bo_list);
+		pt = container_of(parent, struct amdgpu_vm_pt, base);
+		entry = container_of(bo_base, struct amdgpu_vm_pt, base);
+
+		amdgpu_vm_update_pde(&params, vm, pt, entry);
+
+		spin_lock(&vm->status_lock);
+		if (!vm->use_cpu_for_update &&
+		    (ndw - params.ib->length_dw) < 32)
+			break;
 	}
 	spin_unlock(&vm->status_lock);
 
@@ -1300,8 +1226,44 @@ int amdgpu_vm_update_directories(struct amdgpu_device *adev,
 		/* Flush HDP */
 		mb();
 		amdgpu_gart_flush_gpu_tlb(adev, 0);
+	} else if (params.ib->length_dw == 0) {
+		amdgpu_job_free(job);
+	} else {
+		struct amdgpu_bo *root = vm->root.base.bo;
+		struct amdgpu_ring *ring;
+		struct dma_fence *fence;
+
+		ring = container_of(vm->entity.sched, struct amdgpu_ring,
+				    sched);
+
+		amdgpu_ring_pad_ib(ring, params.ib);
+		amdgpu_sync_resv(adev, &job->sync, root->tbo.resv,
+				 AMDGPU_FENCE_OWNER_VM, false);
+		if (root->shadow)
+			amdgpu_sync_resv(adev, &job->sync,
+					 root->shadow->tbo.resv,
+					 AMDGPU_FENCE_OWNER_VM, false);
+
+		WARN_ON(params.ib->length_dw > ndw);
+		r = amdgpu_job_submit(job, ring, &vm->entity,
+				      AMDGPU_FENCE_OWNER_VM, &fence);
+		if (r)
+			goto error;
+
+		amdgpu_bo_fence(root, fence, true);
+		dma_fence_put(vm->last_update);
+		vm->last_update = fence;
 	}
 
+	if (!list_empty(&vm->relocated))
+		goto restart;
+
+	return 0;
+
+error:
+	amdgpu_vm_invalidate_level(adev, vm, &vm->root,
+				   adev->vm_manager.root_level);
+	amdgpu_job_free(job);
 	return r;
 }
 
@@ -1319,19 +1281,19 @@ void amdgpu_vm_get_entry(struct amdgpu_pte_update_params *p, uint64_t addr,
 			 struct amdgpu_vm_pt **entry,
 			 struct amdgpu_vm_pt **parent)
 {
-	unsigned level = 0;
+	unsigned level = p->adev->vm_manager.root_level;
 
 	*parent = NULL;
 	*entry = &p->vm->root;
 	while ((*entry)->entries) {
-		unsigned idx = addr >> amdgpu_vm_level_shift(p->adev, level++);
+		unsigned shift = amdgpu_vm_level_shift(p->adev, level++);
 
-		idx %= amdgpu_bo_size((*entry)->base.bo) / 8;
 		*parent = *entry;
-		*entry = &(*entry)->entries[idx];
+		*entry = &(*entry)->entries[addr >> shift];
+		addr &= (1ULL << shift) - 1;
 	}
 
-	if (level != p->adev->vm_manager.num_level)
+	if (level != AMDGPU_VM_PTB)
 		*entry = NULL;
 }
 
@@ -1363,17 +1325,18 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p,
 	    !(flags & AMDGPU_PTE_VALID)) {
 
 		dst = amdgpu_bo_gpu_offset(entry->base.bo);
-		dst = amdgpu_gart_get_vm_pde(p->adev, dst);
 		flags = AMDGPU_PTE_VALID;
 	} else {
 		/* Set the huge page flag to stop scanning at this PDE */
 		flags |= AMDGPU_PDE_PTE;
 	}
 
-	if (entry->addr == (dst | flags))
+	if (!entry->huge && !(flags & AMDGPU_PDE_PTE))
 		return;
+	entry->huge = !!(flags & AMDGPU_PDE_PTE);
 
-	entry->addr = (dst | flags);
+	amdgpu_gart_get_vm_pde(p->adev, AMDGPU_VM_PDB0,
+			       &dst, &flags);
 
 	if (use_cpu_update) {
 		/* In case a huge page is replaced with a system
@@ -1447,7 +1410,7 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params,
 		amdgpu_vm_handle_huge_pages(params, entry, parent,
 					    nptes, dst, flags);
 		/* We don't need to update PTEs for huge pages */
-		if (entry->addr & AMDGPU_PDE_PTE)
+		if (entry->huge)
 			continue;
 
 		pt = entry->base.bo;
@@ -1688,7 +1651,8 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 
 error_free:
 	amdgpu_job_free(job);
-	amdgpu_vm_invalidate_level(vm, &vm->root);
+	amdgpu_vm_invalidate_level(adev, vm, &vm->root,
+				   adev->vm_manager.root_level);
 	return r;
 }
 
@@ -2604,7 +2568,19 @@ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size,
 		tmp >>= amdgpu_vm_block_size - 9;
 	tmp = DIV_ROUND_UP(fls64(tmp) - 1, 9) - 1;
 	adev->vm_manager.num_level = min(max_level, (unsigned)tmp);
-
+	switch (adev->vm_manager.num_level) {
+	case 3:
+		adev->vm_manager.root_level = AMDGPU_VM_PDB2;
+		break;
+	case 2:
+		adev->vm_manager.root_level = AMDGPU_VM_PDB1;
+		break;
+	case 1:
+		adev->vm_manager.root_level = AMDGPU_VM_PDB0;
+		break;
+	default:
+		dev_err(adev->dev, "VMPT only supports 2~4+1 levels\n");
+	}
 	/* block size depends on vm size and hw setup*/
 	if (amdgpu_vm_block_size != -1)
 		adev->vm_manager.block_size =
@@ -2643,7 +2619,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		AMDGPU_VM_PTE_COUNT(adev) * 8);
 	unsigned ring_instance;
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	int r, i;
 	u64 flags;
 	uint64_t init_pde_value = 0;
@@ -2663,8 +2639,8 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	ring_instance = atomic_inc_return(&adev->vm_manager.vm_pte_next_ring);
 	ring_instance %= adev->vm_manager.vm_pte_num_rings;
 	ring = adev->vm_manager.vm_pte_rings[ring_instance];
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_KERNEL];
-	r = amd_sched_entity_init(&ring->sched, &vm->entity,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
+	r = drm_sched_entity_init(&ring->sched, &vm->entity,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r)
 		return r;
@@ -2698,7 +2674,9 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		flags |= (AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
 				AMDGPU_GEM_CREATE_SHADOW);
 
-	r = amdgpu_bo_create(adev, amdgpu_vm_bo_size(adev, 0), align, true,
+	r = amdgpu_bo_create(adev,
+			     amdgpu_vm_bo_size(adev, adev->vm_manager.root_level),
+			     align, true,
 			     AMDGPU_GEM_DOMAIN_VRAM,
 			     flags,
 			     NULL, NULL, init_pde_value, &vm->root.base.bo);
@@ -2744,7 +2722,7 @@ error_free_root:
 	vm->root.base.bo = NULL;
 
 error_free_sched_entity:
-	amd_sched_entity_fini(&ring->sched, &vm->entity);
+	drm_sched_entity_fini(&ring->sched, &vm->entity);
 
 	return r;
 }
@@ -2752,26 +2730,31 @@ error_free_sched_entity:
 /**
  * amdgpu_vm_free_levels - free PD/PT levels
  *
- * @level: PD/PT starting level to free
+ * @adev: amdgpu device structure
+ * @parent: PD/PT starting level to free
+ * @level: level of parent structure
  *
  * Free the page directory or page table level and all sub levels.
  */
-static void amdgpu_vm_free_levels(struct amdgpu_vm_pt *level)
+static void amdgpu_vm_free_levels(struct amdgpu_device *adev,
+				  struct amdgpu_vm_pt *parent,
+				  unsigned level)
 {
-	unsigned i;
+	unsigned i, num_entries = amdgpu_vm_num_entries(adev, level);
 
-	if (level->base.bo) {
-		list_del(&level->base.bo_list);
-		list_del(&level->base.vm_status);
-		amdgpu_bo_unref(&level->base.bo->shadow);
-		amdgpu_bo_unref(&level->base.bo);
+	if (parent->base.bo) {
+		list_del(&parent->base.bo_list);
+		list_del(&parent->base.vm_status);
+		amdgpu_bo_unref(&parent->base.bo->shadow);
+		amdgpu_bo_unref(&parent->base.bo);
 	}
 
-	if (level->entries)
-		for (i = 0; i <= level->last_entry_used; i++)
-			amdgpu_vm_free_levels(&level->entries[i]);
+	if (parent->entries)
+		for (i = 0; i < num_entries; i++)
+			amdgpu_vm_free_levels(adev, &parent->entries[i],
+					      level + 1);
 
-	kvfree(level->entries);
+	kvfree(parent->entries);
 }
 
 /**
@@ -2803,7 +2786,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 		spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags);
 	}
 
-	amd_sched_entity_fini(vm->entity.sched, &vm->entity);
+	drm_sched_entity_fini(vm->entity.sched, &vm->entity);
 
 	if (!RB_EMPTY_ROOT(&vm->va.rb_root)) {
 		dev_err(adev->dev, "still active bo inside vm\n");
@@ -2829,7 +2812,8 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	if (r) {
 		dev_err(adev->dev, "Leaking page tables because BO reservation failed\n");
 	} else {
-		amdgpu_vm_free_levels(&vm->root);
+		amdgpu_vm_free_levels(adev, &vm->root,
+				      adev->vm_manager.root_level);
 		amdgpu_bo_unreserve(root);
 	}
 	amdgpu_bo_unref(&root);

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

@@ -24,10 +24,11 @@
 #ifndef __AMDGPU_VM_H__
 #define __AMDGPU_VM_H__
 
-#include <linux/rbtree.h>
 #include <linux/idr.h>
+#include <linux/kfifo.h>
+#include <linux/rbtree.h>
+#include <drm/gpu_scheduler.h>
 
-#include "gpu_scheduler.h"
 #include "amdgpu_sync.h"
 #include "amdgpu_ring.h"
 
@@ -69,6 +70,12 @@ struct amdgpu_bo_list_entry;
 /* PDE is handled as PTE for VEGA10 */
 #define AMDGPU_PDE_PTE		(1ULL << 54)
 
+/* PTE is handled as PDE for VEGA10 (Translate Further) */
+#define AMDGPU_PTE_TF		(1ULL << 56)
+
+/* PDE Block Fragment Size for VEGA10 */
+#define AMDGPU_PDE_BFS(a)	((uint64_t)a << 59)
+
 /* VEGA10 only */
 #define AMDGPU_PTE_MTYPE(a)    ((uint64_t)a << 57)
 #define AMDGPU_PTE_MTYPE_MASK	AMDGPU_PTE_MTYPE(3ULL)
@@ -119,6 +126,16 @@ struct amdgpu_bo_list_entry;
 #define AMDGPU_VM_USE_CPU_FOR_GFX (1 << 0)
 #define AMDGPU_VM_USE_CPU_FOR_COMPUTE (1 << 1)
 
+/* VMPT level enumerate, and the hiberachy is:
+ * PDB2->PDB1->PDB0->PTB
+ */
+enum amdgpu_vm_level {
+	AMDGPU_VM_PDB2,
+	AMDGPU_VM_PDB1,
+	AMDGPU_VM_PDB0,
+	AMDGPU_VM_PTB
+};
+
 /* base structure for tracking BO usage in a VM */
 struct amdgpu_vm_bo_base {
 	/* constant after initialization */
@@ -137,11 +154,10 @@ struct amdgpu_vm_bo_base {
 
 struct amdgpu_vm_pt {
 	struct amdgpu_vm_bo_base	base;
-	uint64_t			addr;
+	bool				huge;
 
 	/* array of page tables, one for each directory entry */
 	struct amdgpu_vm_pt		*entries;
-	unsigned			last_entry_used;
 };
 
 #define AMDGPU_VM_FAULT(pasid, addr) (((u64)(pasid) << 48) | (addr))
@@ -175,7 +191,7 @@ struct amdgpu_vm {
 	spinlock_t		freed_lock;
 
 	/* Scheduler entity for page table updates */
-	struct amd_sched_entity	entity;
+	struct drm_sched_entity	entity;
 
 	/* client id and PASID (TODO: replace client_id with PASID) */
 	u64                     client_id;
@@ -236,6 +252,7 @@ struct amdgpu_vm_manager {
 	uint32_t				num_level;
 	uint32_t				block_size;
 	uint32_t				fragment_size;
+	enum amdgpu_vm_level			root_level;
 	/* vram base address for page table entry  */
 	u64					vram_base_offset;
 	/* vm pte handling */

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

@@ -891,12 +891,12 @@ static void ci_dpm_powergate_uvd(void *handle, bool gate)
 
 	if (gate) {
 		/* stop the UVD block */
-		amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							AMD_PG_STATE_GATE);
+		amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+						       AMD_PG_STATE_GATE);
 		ci_update_uvd_dpm(adev, gate);
 	} else {
-		amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							AMD_PG_STATE_UNGATE);
+		amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+						       AMD_PG_STATE_UNGATE);
 		ci_update_uvd_dpm(adev, gate);
 	}
 }

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

@@ -755,74 +755,74 @@ static void cik_init_golden_registers(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_BONAIRE:
-		amdgpu_program_register_sequence(adev,
-						 bonaire_mgcg_cgcg_init,
-						 ARRAY_SIZE(bonaire_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 bonaire_golden_registers,
-						 ARRAY_SIZE(bonaire_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 bonaire_golden_common_registers,
-						 ARRAY_SIZE(bonaire_golden_common_registers));
-		amdgpu_program_register_sequence(adev,
-						 bonaire_golden_spm_registers,
-						 ARRAY_SIZE(bonaire_golden_spm_registers));
+		amdgpu_device_program_register_sequence(adev,
+							bonaire_mgcg_cgcg_init,
+							ARRAY_SIZE(bonaire_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							bonaire_golden_registers,
+							ARRAY_SIZE(bonaire_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							bonaire_golden_common_registers,
+							ARRAY_SIZE(bonaire_golden_common_registers));
+		amdgpu_device_program_register_sequence(adev,
+							bonaire_golden_spm_registers,
+							ARRAY_SIZE(bonaire_golden_spm_registers));
 		break;
 	case CHIP_KABINI:
-		amdgpu_program_register_sequence(adev,
-						 kalindi_mgcg_cgcg_init,
-						 ARRAY_SIZE(kalindi_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 kalindi_golden_registers,
-						 ARRAY_SIZE(kalindi_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 kalindi_golden_common_registers,
-						 ARRAY_SIZE(kalindi_golden_common_registers));
-		amdgpu_program_register_sequence(adev,
-						 kalindi_golden_spm_registers,
-						 ARRAY_SIZE(kalindi_golden_spm_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_mgcg_cgcg_init,
+							ARRAY_SIZE(kalindi_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_golden_registers,
+							ARRAY_SIZE(kalindi_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_golden_common_registers,
+							ARRAY_SIZE(kalindi_golden_common_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_golden_spm_registers,
+							ARRAY_SIZE(kalindi_golden_spm_registers));
 		break;
 	case CHIP_MULLINS:
-		amdgpu_program_register_sequence(adev,
-						 kalindi_mgcg_cgcg_init,
-						 ARRAY_SIZE(kalindi_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 godavari_golden_registers,
-						 ARRAY_SIZE(godavari_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 kalindi_golden_common_registers,
-						 ARRAY_SIZE(kalindi_golden_common_registers));
-		amdgpu_program_register_sequence(adev,
-						 kalindi_golden_spm_registers,
-						 ARRAY_SIZE(kalindi_golden_spm_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_mgcg_cgcg_init,
+							ARRAY_SIZE(kalindi_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							godavari_golden_registers,
+							ARRAY_SIZE(godavari_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_golden_common_registers,
+							ARRAY_SIZE(kalindi_golden_common_registers));
+		amdgpu_device_program_register_sequence(adev,
+							kalindi_golden_spm_registers,
+							ARRAY_SIZE(kalindi_golden_spm_registers));
 		break;
 	case CHIP_KAVERI:
-		amdgpu_program_register_sequence(adev,
-						 spectre_mgcg_cgcg_init,
-						 ARRAY_SIZE(spectre_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 spectre_golden_registers,
-						 ARRAY_SIZE(spectre_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 spectre_golden_common_registers,
-						 ARRAY_SIZE(spectre_golden_common_registers));
-		amdgpu_program_register_sequence(adev,
-						 spectre_golden_spm_registers,
-						 ARRAY_SIZE(spectre_golden_spm_registers));
+		amdgpu_device_program_register_sequence(adev,
+							spectre_mgcg_cgcg_init,
+							ARRAY_SIZE(spectre_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							spectre_golden_registers,
+							ARRAY_SIZE(spectre_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							spectre_golden_common_registers,
+							ARRAY_SIZE(spectre_golden_common_registers));
+		amdgpu_device_program_register_sequence(adev,
+							spectre_golden_spm_registers,
+							ARRAY_SIZE(spectre_golden_spm_registers));
 		break;
 	case CHIP_HAWAII:
-		amdgpu_program_register_sequence(adev,
-						 hawaii_mgcg_cgcg_init,
-						 ARRAY_SIZE(hawaii_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 hawaii_golden_registers,
-						 ARRAY_SIZE(hawaii_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 hawaii_golden_common_registers,
-						 ARRAY_SIZE(hawaii_golden_common_registers));
-		amdgpu_program_register_sequence(adev,
-						 hawaii_golden_spm_registers,
-						 ARRAY_SIZE(hawaii_golden_spm_registers));
+		amdgpu_device_program_register_sequence(adev,
+							hawaii_mgcg_cgcg_init,
+							ARRAY_SIZE(hawaii_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							hawaii_golden_registers,
+							ARRAY_SIZE(hawaii_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							hawaii_golden_common_registers,
+							ARRAY_SIZE(hawaii_golden_common_registers));
+		amdgpu_device_program_register_sequence(adev,
+							hawaii_golden_spm_registers,
+							ARRAY_SIZE(hawaii_golden_spm_registers));
 		break;
 	default:
 		break;
@@ -1246,7 +1246,7 @@ static int cik_gpu_pci_config_reset(struct amdgpu_device *adev)
 	/* disable BM */
 	pci_clear_master(adev->pdev);
 	/* reset */
-	amdgpu_pci_config_reset(adev);
+	amdgpu_device_pci_config_reset(adev);
 
 	udelay(100);
 
@@ -1866,7 +1866,7 @@ static int cik_common_early_init(void *handle)
 
 	adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type);
 
-	amdgpu_get_pcie_info(adev);
+	amdgpu_device_get_pcie_info(adev);
 
 	return 0;
 }
@@ -1974,77 +1974,77 @@ int cik_set_ip_blocks(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_BONAIRE:
-		amdgpu_ip_block_add(adev, &cik_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v7_0_ip_block);
-		amdgpu_ip_block_add(adev, &cik_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v8_2_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v7_2_ip_block);
-		amdgpu_ip_block_add(adev, &cik_sdma_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v4_2_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v2_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v8_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v7_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v4_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v2_0_ip_block);
 		break;
 	case CHIP_HAWAII:
-		amdgpu_ip_block_add(adev, &cik_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v7_0_ip_block);
-		amdgpu_ip_block_add(adev, &cik_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v8_5_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v7_3_ip_block);
-		amdgpu_ip_block_add(adev, &cik_sdma_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v4_2_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v2_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v8_5_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v7_3_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v4_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v2_0_ip_block);
 		break;
 	case CHIP_KAVERI:
-		amdgpu_ip_block_add(adev, &cik_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v7_0_ip_block);
-		amdgpu_ip_block_add(adev, &cik_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v8_1_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v7_1_ip_block);
-		amdgpu_ip_block_add(adev, &cik_sdma_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v4_2_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v2_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v8_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v7_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v4_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v2_0_ip_block);
 		break;
 	case CHIP_KABINI:
 	case CHIP_MULLINS:
-		amdgpu_ip_block_add(adev, &cik_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v7_0_ip_block);
-		amdgpu_ip_block_add(adev, &cik_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v8_3_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v7_2_ip_block);
-		amdgpu_ip_block_add(adev, &cik_sdma_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v4_2_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v2_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v8_3_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v7_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &cik_sdma_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v4_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v2_0_ip_block);
 		break;
 	default:
 		/* FIXME: not supported yet */

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

@@ -626,7 +626,7 @@ static int cik_sdma_ring_test_ring(struct amdgpu_ring *ring)
 	u32 tmp;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
 		return r;
@@ -639,7 +639,7 @@ static int cik_sdma_ring_test_ring(struct amdgpu_ring *ring)
 	r = amdgpu_ring_alloc(ring, 5);
 	if (r) {
 		DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
-		amdgpu_wb_free(adev, index);
+		amdgpu_device_wb_free(adev, index);
 		return r;
 	}
 	amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0));
@@ -663,7 +663,7 @@ static int cik_sdma_ring_test_ring(struct amdgpu_ring *ring)
 			  ring->idx, tmp);
 		r = -EINVAL;
 	}
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 
 	return r;
 }
@@ -686,7 +686,7 @@ static int cik_sdma_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	u64 gpu_addr;
 	long r;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
@@ -735,7 +735,7 @@ err1:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err0:
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 

drivers/gpu/drm/amd/amdgpu/clearstate_gfx9.h

@@ -48,7 +48,7 @@ static const unsigned int gfx9_SECT_CONTEXT_def_1[] =
     0x00000000, // DB_STENCIL_WRITE_BASE
     0x00000000, // DB_STENCIL_WRITE_BASE_HI
     0x00000000, // DB_DFSM_CONTROL
-    0x00000000, // DB_RENDER_FILTER
+    0, // HOLE
     0x00000000, // DB_Z_INFO2
     0x00000000, // DB_STENCIL_INFO2
     0, // HOLE
@@ -259,8 +259,8 @@ static const unsigned int gfx9_SECT_CONTEXT_def_2[] =
     0x00000000, // PA_SC_RIGHT_VERT_GRID
     0x00000000, // PA_SC_LEFT_VERT_GRID
     0x00000000, // PA_SC_HORIZ_GRID
-    0x00000000, // PA_SC_FOV_WINDOW_LR
-    0x00000000, // PA_SC_FOV_WINDOW_TB
+    0, // HOLE
+    0, // HOLE
     0, // HOLE
     0, // HOLE
     0, // HOLE
@@ -701,7 +701,7 @@ static const unsigned int gfx9_SECT_CONTEXT_def_7[] =
 {
     0x00000000, // VGT_GS_MAX_PRIMS_PER_SUBGROUP
     0x00000000, // VGT_DRAW_PAYLOAD_CNTL
-    0x00000000, // VGT_INDEX_PAYLOAD_CNTL
+    0, // HOLE
     0x00000000, // VGT_INSTANCE_STEP_RATE_0
     0x00000000, // VGT_INSTANCE_STEP_RATE_1
     0, // HOLE

drivers/gpu/drm/amd/amdgpu/dce_v10_0.c

@@ -145,20 +145,20 @@ static void dce_v10_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(fiji_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_fiji_a10,
-						 ARRAY_SIZE(golden_settings_fiji_a10));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_fiji_a10,
+							ARRAY_SIZE(golden_settings_fiji_a10));
 		break;
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(tonga_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_tonga_a11,
-						 ARRAY_SIZE(golden_settings_tonga_a11));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_tonga_a11,
+							ARRAY_SIZE(golden_settings_tonga_a11));
 		break;
 	default:
 		break;

drivers/gpu/drm/amd/amdgpu/dce_v11_0.c

@@ -154,28 +154,28 @@ static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_CARRIZO:
-		amdgpu_program_register_sequence(adev,
-						 cz_mgcg_cgcg_init,
-						 ARRAY_SIZE(cz_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 cz_golden_settings_a11,
-						 ARRAY_SIZE(cz_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							cz_mgcg_cgcg_init,
+							ARRAY_SIZE(cz_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							cz_golden_settings_a11,
+							ARRAY_SIZE(cz_golden_settings_a11));
 		break;
 	case CHIP_STONEY:
-		amdgpu_program_register_sequence(adev,
-						 stoney_golden_settings_a11,
-						 ARRAY_SIZE(stoney_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_golden_settings_a11,
+							ARRAY_SIZE(stoney_golden_settings_a11));
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-		amdgpu_program_register_sequence(adev,
-						 polaris11_golden_settings_a11,
-						 ARRAY_SIZE(polaris11_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							polaris11_golden_settings_a11,
+							ARRAY_SIZE(polaris11_golden_settings_a11));
 		break;
 	case CHIP_POLARIS10:
-		amdgpu_program_register_sequence(adev,
-						 polaris10_golden_settings_a11,
-						 ARRAY_SIZE(polaris10_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							polaris10_golden_settings_a11,
+							ARRAY_SIZE(polaris10_golden_settings_a11));
 		break;
 	default:
 		break;

drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c

@@ -679,55 +679,55 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_TOPAZ:
-		amdgpu_program_register_sequence(adev,
-						 iceland_mgcg_cgcg_init,
-						 ARRAY_SIZE(iceland_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_iceland_a11,
-						 ARRAY_SIZE(golden_settings_iceland_a11));
-		amdgpu_program_register_sequence(adev,
-						 iceland_golden_common_all,
-						 ARRAY_SIZE(iceland_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							iceland_mgcg_cgcg_init,
+							ARRAY_SIZE(iceland_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_iceland_a11,
+							ARRAY_SIZE(golden_settings_iceland_a11));
+		amdgpu_device_program_register_sequence(adev,
+							iceland_golden_common_all,
+							ARRAY_SIZE(iceland_golden_common_all));
 		break;
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(fiji_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_fiji_a10,
-						 ARRAY_SIZE(golden_settings_fiji_a10));
-		amdgpu_program_register_sequence(adev,
-						 fiji_golden_common_all,
-						 ARRAY_SIZE(fiji_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_fiji_a10,
+							ARRAY_SIZE(golden_settings_fiji_a10));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_golden_common_all,
+							ARRAY_SIZE(fiji_golden_common_all));
 		break;
 
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(tonga_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_tonga_a11,
-						 ARRAY_SIZE(golden_settings_tonga_a11));
-		amdgpu_program_register_sequence(adev,
-						 tonga_golden_common_all,
-						 ARRAY_SIZE(tonga_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_tonga_a11,
+							ARRAY_SIZE(golden_settings_tonga_a11));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_golden_common_all,
+							ARRAY_SIZE(tonga_golden_common_all));
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris11_a11,
-						 ARRAY_SIZE(golden_settings_polaris11_a11));
-		amdgpu_program_register_sequence(adev,
-						 polaris11_golden_common_all,
-						 ARRAY_SIZE(polaris11_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris11_a11,
+							ARRAY_SIZE(golden_settings_polaris11_a11));
+		amdgpu_device_program_register_sequence(adev,
+							polaris11_golden_common_all,
+							ARRAY_SIZE(polaris11_golden_common_all));
 		break;
 	case CHIP_POLARIS10:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris10_a11,
-						 ARRAY_SIZE(golden_settings_polaris10_a11));
-		amdgpu_program_register_sequence(adev,
-						 polaris10_golden_common_all,
-						 ARRAY_SIZE(polaris10_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris10_a11,
+							ARRAY_SIZE(golden_settings_polaris10_a11));
+		amdgpu_device_program_register_sequence(adev,
+							polaris10_golden_common_all,
+							ARRAY_SIZE(polaris10_golden_common_all));
 		WREG32_SMC(ixCG_ACLK_CNTL, 0x0000001C);
 		if (adev->pdev->revision == 0xc7 &&
 		    ((adev->pdev->subsystem_device == 0xb37 && adev->pdev->subsystem_vendor == 0x1002) ||
@@ -738,26 +738,26 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 		}
 		break;
 	case CHIP_CARRIZO:
-		amdgpu_program_register_sequence(adev,
-						 cz_mgcg_cgcg_init,
-						 ARRAY_SIZE(cz_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 cz_golden_settings_a11,
-						 ARRAY_SIZE(cz_golden_settings_a11));
-		amdgpu_program_register_sequence(adev,
-						 cz_golden_common_all,
-						 ARRAY_SIZE(cz_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							cz_mgcg_cgcg_init,
+							ARRAY_SIZE(cz_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							cz_golden_settings_a11,
+							ARRAY_SIZE(cz_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							cz_golden_common_all,
+							ARRAY_SIZE(cz_golden_common_all));
 		break;
 	case CHIP_STONEY:
-		amdgpu_program_register_sequence(adev,
-						 stoney_mgcg_cgcg_init,
-						 ARRAY_SIZE(stoney_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 stoney_golden_settings_a11,
-						 ARRAY_SIZE(stoney_golden_settings_a11));
-		amdgpu_program_register_sequence(adev,
-						 stoney_golden_common_all,
-						 ARRAY_SIZE(stoney_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_mgcg_cgcg_init,
+							ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_golden_settings_a11,
+							ARRAY_SIZE(stoney_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_golden_common_all,
+							ARRAY_SIZE(stoney_golden_common_all));
 		break;
 	default:
 		break;
@@ -5062,8 +5062,9 @@ static int gfx_v8_0_hw_fini(void *handle)
 	gfx_v8_0_cp_enable(adev, false);
 	gfx_v8_0_rlc_stop(adev);
 
-	amdgpu_set_powergating_state(adev,
-			AMD_IP_BLOCK_TYPE_GFX, AMD_PG_STATE_UNGATE);
+	amdgpu_device_ip_set_powergating_state(adev,
+					       AMD_IP_BLOCK_TYPE_GFX,
+					       AMD_PG_STATE_UNGATE);
 
 	return 0;
 }
@@ -5480,8 +5481,9 @@ static int gfx_v8_0_late_init(void *handle)
 	if (r)
 		return r;
 
-	amdgpu_set_powergating_state(adev,
-			AMD_IP_BLOCK_TYPE_GFX, AMD_PG_STATE_GATE);
+	amdgpu_device_ip_set_powergating_state(adev,
+					       AMD_IP_BLOCK_TYPE_GFX,
+					       AMD_PG_STATE_GATE);
 
 	return 0;
 }
@@ -5492,10 +5494,10 @@ static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *ade
 	if ((adev->asic_type == CHIP_POLARIS11) ||
 	    (adev->asic_type == CHIP_POLARIS12))
 		/* Send msg to SMU via Powerplay */
-		amdgpu_set_powergating_state(adev,
-					     AMD_IP_BLOCK_TYPE_SMC,
-					     enable ?
-					     AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE);
+		amdgpu_device_ip_set_powergating_state(adev,
+						       AMD_IP_BLOCK_TYPE_SMC,
+						       enable ?
+						       AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE);
 
 	WREG32_FIELD(RLC_PG_CNTL, STATIC_PER_CU_PG_ENABLE, enable ? 1 : 0);
 }
@@ -6472,10 +6474,10 @@ static void gfx_v8_0_hqd_set_priority(struct amdgpu_device *adev,
 	mutex_unlock(&adev->srbm_mutex);
 }
 static void gfx_v8_0_ring_set_priority_compute(struct amdgpu_ring *ring,
-					       enum amd_sched_priority priority)
+					       enum drm_sched_priority priority)
 {
 	struct amdgpu_device *adev = ring->adev;
-	bool acquire = priority == AMD_SCHED_PRIORITY_HIGH_HW;
+	bool acquire = priority == DRM_SCHED_PRIORITY_HIGH_HW;
 
 	if (ring->funcs->type != AMDGPU_RING_TYPE_COMPUTE)
 		return;

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

@@ -28,7 +28,6 @@
 #include "soc15.h"
 #include "soc15d.h"
 
-#include "soc15ip.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
 #include "vega10_enum.h"
@@ -65,152 +64,84 @@ MODULE_FIRMWARE("amdgpu/raven_mec.bin");
 MODULE_FIRMWARE("amdgpu/raven_mec2.bin");
 MODULE_FIRMWARE("amdgpu/raven_rlc.bin");
 
-static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] =
-{
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID0),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID0) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID1_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID1_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID1),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID1) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID2_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID2_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID2),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID2) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID3_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID3_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID3),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID3) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID4_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID4_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID4),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID4) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID5_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID5_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID5),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID5) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID6_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID6_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID6),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID6) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID7_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID7_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID7),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID7) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID8_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID8_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID8),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID8) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID9_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID9_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID9),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID9) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID10_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID10_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID10),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID10) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID11_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID11_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID11),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID11) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID12_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID12_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID12),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID12)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID13_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID13_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID13),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID13) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID14_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID14_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID14),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID14) },
-	{ SOC15_REG_OFFSET(GC, 0, mmGDS_VMID15_BASE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_VMID15_SIZE),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID15),
-	  SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID15) }
+static const struct soc15_reg_golden golden_settings_gc_9_0[] =
+{
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPC_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPG_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0xf00fffff, 0x00000420),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_GPU_ID, 0x0000000f, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3, 0x00000003, 0x82400024),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x00000001),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_0, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_1, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_2, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_SPM_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSH_MEM_CONFIG, 0x00001000, 0x00001000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQC_CONFIG, 0x03000000, 0x020a2000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x4a2c0e68),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0xb5d3f197),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080)
 };
 
-static const u32 golden_settings_gc_9_0[] =
-{
-	SOC15_REG_OFFSET(GC, 0, mmCPC_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmCPF_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmCPG_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG2), 0xf00fffff, 0x00000420,
-	SOC15_REG_OFFSET(GC, 0, mmGB_GPU_ID), 0x0000000f, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmIA_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3), 0x00000003, 0x82400024,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_ENHANCE), 0x3fffffff, 0x00000001,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_LINE_STIPPLE_STATE), 0x0000ff0f, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_0), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_1), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_2), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_SPM_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), 0x00001000, 0x00001000,
-	SOC15_REG_OFFSET(GC, 0, mmSPI_CONFIG_CNTL_1), 0x0000000f, 0x01000107,
-	SOC15_REG_OFFSET(GC, 0, mmSQC_CONFIG), 0x03000000, 0x020a2000,
-	SOC15_REG_OFFSET(GC, 0, mmTA_CNTL_AUX), 0xfffffeef, 0x010b0000,
-	SOC15_REG_OFFSET(GC, 0, mmTCP_CHAN_STEER_HI), 0xffffffff, 0x4a2c0e68,
-	SOC15_REG_OFFSET(GC, 0, mmTCP_CHAN_STEER_LO), 0xffffffff, 0xb5d3f197,
-	SOC15_REG_OFFSET(GC, 0, mmVGT_CACHE_INVALIDATION), 0x3fff3af3, 0x19200000,
-	SOC15_REG_OFFSET(GC, 0, mmVGT_GS_MAX_WAVE_ID), 0x00000fff, 0x000003ff,
-	SOC15_REG_OFFSET(GC, 0, mmWD_UTCL1_CNTL), 0x08000000, 0x08000080
-};
-
-static const u32 golden_settings_gc_9_0_vg10[] =
+static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
 {
-	SOC15_REG_OFFSET(GC, 0, mmCB_HW_CONTROL), 0x0000f000, 0x00012107,
-	SOC15_REG_OFFSET(GC, 0, mmCB_HW_CONTROL_3), 0x30000000, 0x10000000,
-	SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG), 0xffff77ff, 0x2a114042,
-	SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG_READ), 0xffff77ff, 0x2a114042,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_ENHANCE_1), 0x00008000, 0x00048000,
-	SOC15_REG_OFFSET(GC, 0, mmRMI_UTCL1_CNTL2), 0x00030000, 0x00020000,
-	SOC15_REG_OFFSET(GC, 0, mmTD_CNTL), 0x00001800, 0x00000800
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL, 0x0000f000, 0x00012107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xffff77ff, 0x2a114042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xffff77ff, 0x2a114042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0x00008000, 0x00048000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00020000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800)
 };
 
-static const u32 golden_settings_gc_9_1[] =
-{
-	SOC15_REG_OFFSET(GC, 0, mmCB_HW_CONTROL), 0xfffdf3cf, 0x00014104,
-	SOC15_REG_OFFSET(GC, 0, mmCPC_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmCPF_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmCPG_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG2), 0xf00fffff, 0x00000420,
-	SOC15_REG_OFFSET(GC, 0, mmGB_GPU_ID), 0x0000000f, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmIA_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3), 0x00000003, 0x82400024,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_ENHANCE), 0x3fffffff, 0x00000001,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_LINE_STIPPLE_STATE), 0x0000ff0f, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_0), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_1), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UTCL1_CNTL_2), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmRLC_SPM_UTCL1_CNTL), 0x08000000, 0x08000080,
-	SOC15_REG_OFFSET(GC, 0, mmTA_CNTL_AUX), 0xfffffeef, 0x010b0000,
-	SOC15_REG_OFFSET(GC, 0, mmTCP_CHAN_STEER_HI), 0xffffffff, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmTCP_CHAN_STEER_LO), 0xffffffff, 0x00003120,
-	SOC15_REG_OFFSET(GC, 0, mmVGT_CACHE_INVALIDATION), 0x3fff3af3, 0x19200000,
-	SOC15_REG_OFFSET(GC, 0, mmVGT_GS_MAX_WAVE_ID), 0x00000fff, 0x000000ff,
-	SOC15_REG_OFFSET(GC, 0, mmWD_UTCL1_CNTL), 0x08000000, 0x08000080
+static const struct soc15_reg_golden golden_settings_gc_9_1[] =
+{
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL, 0xfffdf3cf, 0x00014104),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPC_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPG_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0xf00fffff, 0x00000420),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_GPU_ID, 0x0000000f, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3, 0x00000003, 0x82400024),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x00000001),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_0, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_1, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_2, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_SPM_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0x00003120),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000000ff),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080)
 };
 
-static const u32 golden_settings_gc_9_1_rv1[] =
+static const struct soc15_reg_golden golden_settings_gc_9_1_rv1[] =
 {
-	SOC15_REG_OFFSET(GC, 0, mmCB_HW_CONTROL_3), 0x30000000, 0x10000000,
-	SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG), 0xffff77ff, 0x24000042,
-	SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG_READ), 0xffff77ff, 0x24000042,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_ENHANCE_1), 0xffffffff, 0x04048000,
-	SOC15_REG_OFFSET(GC, 0, mmPA_SC_MODE_CNTL_1), 0x06000000, 0x06000000,
-	SOC15_REG_OFFSET(GC, 0, mmRMI_UTCL1_CNTL2), 0x00030000, 0x00020000,
-	SOC15_REG_OFFSET(GC, 0, mmTD_CNTL), 0x01bd9f33, 0x00000800
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xffff77ff, 0x24000042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xffff77ff, 0x24000042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0xffffffff, 0x04048000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_MODE_CNTL_1, 0x06000000, 0x06000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00020000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01bd9f33, 0x00000800)
 };
 
-static const u32 golden_settings_gc_9_x_common[] =
+static const struct soc15_reg_golden golden_settings_gc_9_x_common[] =
 {
-	SOC15_REG_OFFSET(GC, 0, mmGRBM_CAM_INDEX), 0xffffffff, 0x00000000,
-	SOC15_REG_OFFSET(GC, 0, mmGRBM_CAM_DATA), 0xffffffff, 0x2544c382
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_INDEX, 0xffffffff, 0x00000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2544c382)
 };
 
 #define VEGA10_GB_ADDR_CONFIG_GOLDEN 0x2a114042
@@ -230,18 +161,18 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_0,
 						 ARRAY_SIZE(golden_settings_gc_9_0));
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_0_vg10,
 						 ARRAY_SIZE(golden_settings_gc_9_0_vg10));
 		break;
 	case CHIP_RAVEN:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_1,
 						 ARRAY_SIZE(golden_settings_gc_9_1));
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_1_rv1,
 						 ARRAY_SIZE(golden_settings_gc_9_1_rv1));
 		break;
@@ -249,7 +180,7 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	}
 
-	amdgpu_program_register_sequence(adev, golden_settings_gc_9_x_common,
+	soc15_program_register_sequence(adev, golden_settings_gc_9_x_common,
 					(const u32)ARRAY_SIZE(golden_settings_gc_9_x_common));
 }
 
@@ -1137,7 +1068,7 @@ static int gfx_v9_0_ngg_init(struct amdgpu_device *adev)
 	adev->gfx.ngg.gds_reserve_size = ALIGN(5 * 4, 0x40);
 	adev->gds.mem.total_size -= adev->gfx.ngg.gds_reserve_size;
 	adev->gds.mem.gfx_partition_size -= adev->gfx.ngg.gds_reserve_size;
-	adev->gfx.ngg.gds_reserve_addr = amdgpu_gds_reg_offset[0].mem_base;
+	adev->gfx.ngg.gds_reserve_addr = SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_BASE);
 	adev->gfx.ngg.gds_reserve_addr += adev->gds.mem.gfx_partition_size;
 
 	/* Primitive Buffer */
@@ -1243,7 +1174,7 @@ static int gfx_v9_0_ngg_en(struct amdgpu_device *adev)
 	}
 
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[0].mem_size,
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE),
 			           (adev->gds.mem.total_size +
 				    adev->gfx.ngg.gds_reserve_size) >>
 				   AMDGPU_GDS_SHIFT);
@@ -1259,7 +1190,7 @@ static int gfx_v9_0_ngg_en(struct amdgpu_device *adev)
 
 
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[0].mem_size, 0);
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE), 0);
 
 	amdgpu_ring_commit(ring);
 
@@ -1598,11 +1529,18 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev)
 	for (i = 0; i < 16; i++) {
 		soc15_grbm_select(adev, 0, 0, 0, i);
 		/* CP and shaders */
-		tmp = 0;
-		tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, ALIGNMENT_MODE,
-				    SH_MEM_ALIGNMENT_MODE_UNALIGNED);
-		WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp);
-		WREG32_SOC15(GC, 0, mmSH_MEM_BASES, 0);
+		if (i == 0) {
+			tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE,
+					    SH_MEM_ALIGNMENT_MODE_UNALIGNED);
+			WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp);
+			WREG32_SOC15(GC, 0, mmSH_MEM_BASES, 0);
+		} else {
+			tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE,
+					    SH_MEM_ALIGNMENT_MODE_UNALIGNED);
+			WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp);
+			tmp = adev->mc.shared_aperture_start >> 48;
+			WREG32_SOC15(GC, 0, mmSH_MEM_BASES, tmp);
+		}
 	}
 	soc15_grbm_select(adev, 0, 0, 0, 0);
 
@@ -2474,7 +2412,7 @@ static int gfx_v9_0_kiq_kcq_enable(struct amdgpu_device *adev)
 				  PACKET3_MAP_QUEUES_PIPE(ring->pipe) |
 				  PACKET3_MAP_QUEUES_ME((ring->me == 1 ? 0 : 1)) |
 				  PACKET3_MAP_QUEUES_QUEUE_TYPE(0) | /*queue_type: normal compute queue */
-				  PACKET3_MAP_QUEUES_ALLOC_FORMAT(1) | /* alloc format: all_on_one_pipe */
+				  PACKET3_MAP_QUEUES_ALLOC_FORMAT(0) | /* alloc format: all_on_one_pipe */
 				  PACKET3_MAP_QUEUES_ENGINE_SEL(0) | /* engine_sel: compute */
 				  PACKET3_MAP_QUEUES_NUM_QUEUES(1)); /* num_queues: must be 1 */
 		amdgpu_ring_write(kiq_ring, PACKET3_MAP_QUEUES_DOORBELL_OFFSET(ring->doorbell_index));
@@ -3146,6 +3084,8 @@ static void gfx_v9_0_ring_emit_gds_switch(struct amdgpu_ring *ring,
 					  uint32_t gws_base, uint32_t gws_size,
 					  uint32_t oa_base, uint32_t oa_size)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	gds_base = gds_base >> AMDGPU_GDS_SHIFT;
 	gds_size = gds_size >> AMDGPU_GDS_SHIFT;
 
@@ -3157,22 +3097,22 @@ static void gfx_v9_0_ring_emit_gds_switch(struct amdgpu_ring *ring,
 
 	/* GDS Base */
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[vmid].mem_base,
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_BASE) + 2 * vmid,
 				   gds_base);
 
 	/* GDS Size */
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[vmid].mem_size,
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_VMID0_SIZE) + 2 * vmid,
 				   gds_size);
 
 	/* GWS */
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[vmid].gws,
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_GWS_VMID0) + vmid,
 				   gws_size << GDS_GWS_VMID0__SIZE__SHIFT | gws_base);
 
 	/* OA */
 	gfx_v9_0_write_data_to_reg(ring, 0, false,
-				   amdgpu_gds_reg_offset[vmid].oa,
+				   SOC15_REG_OFFSET(GC, 0, mmGDS_OA_VMID0) + vmid,
 				   (1 << (oa_size + oa_base)) - (1 << oa_base));
 }
 
@@ -3617,13 +3557,9 @@ static void gfx_v9_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
 
 static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask, reg_mem_engine;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg;
-
-	if (ring->adev->flags & AMD_IS_APU)
-		nbio_hf_reg = &nbio_v7_0_hdp_flush_reg;
-	else
-		nbio_hf_reg = &nbio_v6_1_hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
 
 	if (ring->funcs->type == AMDGPU_RING_TYPE_COMPUTE) {
 		switch (ring->me) {
@@ -3643,13 +3579,15 @@ static void gfx_v9_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	}
 
 	gfx_v9_0_wait_reg_mem(ring, reg_mem_engine, 0, 1,
-			      nbio_hf_reg->hdp_flush_req_offset,
-			      nbio_hf_reg->hdp_flush_done_offset,
+			      adev->nbio_funcs->get_hdp_flush_req_offset(adev),
+			      adev->nbio_funcs->get_hdp_flush_done_offset(adev),
 			      ref_and_mask, ref_and_mask, 0x20);
 }
 
 static void gfx_v9_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	gfx_v9_0_write_data_to_reg(ring, 0, true,
 				   SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1);
 }
@@ -3750,10 +3688,11 @@ static void gfx_v9_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX);
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	gfx_v9_0_write_data_to_reg(ring, usepfp, true,
 				   hub->ctx0_ptb_addr_lo32 + (2 * vm_id),
@@ -3811,6 +3750,8 @@ static void gfx_v9_0_ring_set_wptr_compute(struct amdgpu_ring *ring)
 static void gfx_v9_0_ring_emit_fence_kiq(struct amdgpu_ring *ring, u64 addr,
 					 u64 seq, unsigned int flags)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	/* we only allocate 32bit for each seq wb address */
 	BUG_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 

drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c

@@ -23,7 +23,6 @@
 #include "amdgpu.h"
 #include "gfxhub_v1_0.h"
 
-#include "soc15ip.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
 #include "gc/gc_9_0_default.h"
@@ -144,8 +143,15 @@ static void gfxhub_v1_0_init_cache_regs(struct amdgpu_device *adev)
 	WREG32_SOC15(GC, 0, mmVM_L2_CNTL2, tmp);
 
 	tmp = mmVM_L2_CNTL3_DEFAULT;
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 9);
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	if (adev->mc.translate_further) {
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12);
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 9);
+	} else {
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 9);
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	}
 	WREG32_SOC15(GC, 0, mmVM_L2_CNTL3, tmp);
 
 	tmp = mmVM_L2_CNTL4_DEFAULT;
@@ -183,31 +189,40 @@ static void gfxhub_v1_0_disable_identity_aperture(struct amdgpu_device *adev)
 
 static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev)
 {
-	int i;
+	unsigned num_level, block_size;
 	uint32_t tmp;
+	int i;
+
+	num_level = adev->vm_manager.num_level;
+	block_size = adev->vm_manager.block_size;
+	if (adev->mc.translate_further)
+		num_level -= 1;
+	else
+		block_size -= 9;
 
 	for (i = 0; i <= 14; i++) {
 		tmp = RREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL, i);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, ENABLE_CONTEXT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_DEPTH,
-				    adev->vm_manager.num_level);
+				    num_level);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT,
+				    1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				PDE0_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    PDE0_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				VALID_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    VALID_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				READ_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    READ_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				WRITE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    WRITE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				PAGE_TABLE_BLOCK_SIZE,
-				adev->vm_manager.block_size - 9);
+				    PAGE_TABLE_BLOCK_SIZE,
+				    block_size);
 		/* Send no-retry XNACK on fault to suppress VM fault storm. */
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
 				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);

drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c

@@ -222,8 +222,8 @@ static void gmc_v6_0_vram_gtt_location(struct amdgpu_device *adev,
 	u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;
 
-	amdgpu_vram_location(adev, &adev->mc, base);
-	amdgpu_gart_location(adev, mc);
+	amdgpu_device_vram_location(adev, &adev->mc, base);
+	amdgpu_device_gart_location(adev, mc);
 }
 
 static void gmc_v6_0_mc_program(struct amdgpu_device *adev)
@@ -395,10 +395,10 @@ static uint64_t gmc_v6_0_get_vm_pte_flags(struct amdgpu_device *adev,
 	return pte_flag;
 }
 
-static uint64_t gmc_v6_0_get_vm_pde(struct amdgpu_device *adev, uint64_t addr)
+static void gmc_v6_0_get_vm_pde(struct amdgpu_device *adev, int level,
+				uint64_t *addr, uint64_t *flags)
 {
-	BUG_ON(addr & 0xFFFFFF0000000FFFULL);
-	return addr;
+	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
 static void gmc_v6_0_set_fault_enable_default(struct amdgpu_device *adev,

drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c

@@ -67,12 +67,12 @@ static void gmc_v7_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_TOPAZ:
-		amdgpu_program_register_sequence(adev,
-						 iceland_mgcg_cgcg_init,
-						 ARRAY_SIZE(iceland_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_iceland_a11,
-						 ARRAY_SIZE(golden_settings_iceland_a11));
+		amdgpu_device_program_register_sequence(adev,
+							iceland_mgcg_cgcg_init,
+							ARRAY_SIZE(iceland_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_iceland_a11,
+							ARRAY_SIZE(golden_settings_iceland_a11));
 		break;
 	default:
 		break;
@@ -240,8 +240,8 @@ static void gmc_v7_0_vram_gtt_location(struct amdgpu_device *adev,
 	u64 base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;
 
-	amdgpu_vram_location(adev, &adev->mc, base);
-	amdgpu_gart_location(adev, mc);
+	amdgpu_device_vram_location(adev, &adev->mc, base);
+	amdgpu_device_gart_location(adev, mc);
 }
 
 /**
@@ -480,10 +480,10 @@ static uint64_t gmc_v7_0_get_vm_pte_flags(struct amdgpu_device *adev,
 	return pte_flag;
 }
 
-static uint64_t gmc_v7_0_get_vm_pde(struct amdgpu_device *adev, uint64_t addr)
+static void gmc_v7_0_get_vm_pde(struct amdgpu_device *adev, int level,
+				uint64_t *addr, uint64_t *flags)
 {
-	BUG_ON(addr & 0xFFFFFF0000000FFFULL);
-	return addr;
+	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
 /**

drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c

@@ -120,44 +120,44 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(fiji_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_fiji_a10,
-						 ARRAY_SIZE(golden_settings_fiji_a10));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_fiji_a10,
+							ARRAY_SIZE(golden_settings_fiji_a10));
 		break;
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(tonga_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_tonga_a11,
-						 ARRAY_SIZE(golden_settings_tonga_a11));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_tonga_a11,
+							ARRAY_SIZE(golden_settings_tonga_a11));
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris11_a11,
-						 ARRAY_SIZE(golden_settings_polaris11_a11));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris11_a11,
+							ARRAY_SIZE(golden_settings_polaris11_a11));
 		break;
 	case CHIP_POLARIS10:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris10_a11,
-						 ARRAY_SIZE(golden_settings_polaris10_a11));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris10_a11,
+							ARRAY_SIZE(golden_settings_polaris10_a11));
 		break;
 	case CHIP_CARRIZO:
-		amdgpu_program_register_sequence(adev,
-						 cz_mgcg_cgcg_init,
-						 ARRAY_SIZE(cz_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							cz_mgcg_cgcg_init,
+							ARRAY_SIZE(cz_mgcg_cgcg_init));
 		break;
 	case CHIP_STONEY:
-		amdgpu_program_register_sequence(adev,
-						 stoney_mgcg_cgcg_init,
-						 ARRAY_SIZE(stoney_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_stoney_common,
-						 ARRAY_SIZE(golden_settings_stoney_common));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_mgcg_cgcg_init,
+							ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_stoney_common,
+							ARRAY_SIZE(golden_settings_stoney_common));
 		break;
 	default:
 		break;
@@ -405,8 +405,8 @@ static void gmc_v8_0_vram_gtt_location(struct amdgpu_device *adev,
 		base = RREG32(mmMC_VM_FB_LOCATION) & 0xFFFF;
 	base <<= 24;
 
-	amdgpu_vram_location(adev, &adev->mc, base);
-	amdgpu_gart_location(adev, mc);
+	amdgpu_device_vram_location(adev, &adev->mc, base);
+	amdgpu_device_gart_location(adev, mc);
 }
 
 /**
@@ -677,10 +677,10 @@ static uint64_t gmc_v8_0_get_vm_pte_flags(struct amdgpu_device *adev,
 	return pte_flag;
 }
 
-static uint64_t gmc_v8_0_get_vm_pde(struct amdgpu_device *adev, uint64_t addr)
+static void gmc_v8_0_get_vm_pde(struct amdgpu_device *adev, int level,
+				uint64_t *addr, uint64_t *flags)
 {
-	BUG_ON(addr & 0xFFFFFF0000000FFFULL);
-	return addr;
+	BUG_ON(*addr & 0xFFFFFF0000000FFFULL);
 }
 
 /**

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

@@ -25,7 +25,6 @@
 #include "gmc_v9_0.h"
 #include "amdgpu_atomfirmware.h"
 
-#include "soc15ip.h"
 #include "hdp/hdp_4_0_offset.h"
 #include "hdp/hdp_4_0_sh_mask.h"
 #include "gc/gc_9_0_sh_mask.h"
@@ -35,11 +34,10 @@
 #include "mmhub/mmhub_1_0_offset.h"
 #include "athub/athub_1_0_offset.h"
 
+#include "soc15.h"
 #include "soc15_common.h"
 #include "umc/umc_6_0_sh_mask.h"
 
-#include "nbio_v6_1.h"
-#include "nbio_v7_0.h"
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
 
@@ -74,16 +72,16 @@ static const u32 golden_settings_vega10_hdp[] =
 	0xf6e, 0x0fffffff, 0x00000000,
 };
 
-static const u32 golden_settings_mmhub_1_0_0[] =
+static const struct soc15_reg_golden golden_settings_mmhub_1_0_0[] =
 {
-	SOC15_REG_OFFSET(MMHUB, 0, mmDAGB1_WRCLI2), 0x00000007, 0xfe5fe0fa,
-	SOC15_REG_OFFSET(MMHUB, 0, mmMMEA1_DRAM_WR_CLI2GRP_MAP0), 0x00000030, 0x55555565
+	SOC15_REG_GOLDEN_VALUE(MMHUB, 0, mmDAGB1_WRCLI2, 0x00000007, 0xfe5fe0fa),
+	SOC15_REG_GOLDEN_VALUE(MMHUB, 0, mmMMEA1_DRAM_WR_CLI2GRP_MAP0, 0x00000030, 0x55555565)
 };
 
-static const u32 golden_settings_athub_1_0_0[] =
+static const struct soc15_reg_golden golden_settings_athub_1_0_0[] =
 {
-	SOC15_REG_OFFSET(ATHUB, 0, mmRPB_ARB_CNTL), 0x0000ff00, 0x00000800,
-	SOC15_REG_OFFSET(ATHUB, 0, mmRPB_ARB_CNTL2), 0x00ff00ff, 0x00080008
+	SOC15_REG_GOLDEN_VALUE(ATHUB, 0, mmRPB_ARB_CNTL, 0x0000ff00, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(ATHUB, 0, mmRPB_ARB_CNTL2, 0x00ff00ff, 0x00080008)
 };
 
 /* Ecc related register addresses, (BASE + reg offset) */
@@ -332,10 +330,7 @@ static void gmc_v9_0_gart_flush_gpu_tlb(struct amdgpu_device *adev,
 	unsigned i, j;
 
 	/* flush hdp cache */
-	if (adev->flags & AMD_IS_APU)
-		nbio_v7_0_hdp_flush(adev);
-	else
-		nbio_v6_1_hdp_flush(adev);
+	adev->nbio_funcs->hdp_flush(adev);
 
 	spin_lock(&adev->mc.invalidate_lock);
 
@@ -474,11 +469,28 @@ static uint64_t gmc_v9_0_get_vm_pte_flags(struct amdgpu_device *adev,
 	return pte_flag;
 }
 
-static u64 gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, u64 addr)
+static void gmc_v9_0_get_vm_pde(struct amdgpu_device *adev, int level,
+				uint64_t *addr, uint64_t *flags)
 {
-	addr = adev->vm_manager.vram_base_offset + addr - adev->mc.vram_start;
-	BUG_ON(addr & 0xFFFF00000000003FULL);
-	return addr;
+	if (!(*flags & AMDGPU_PDE_PTE))
+		*addr = adev->vm_manager.vram_base_offset + *addr -
+			adev->mc.vram_start;
+	BUG_ON(*addr & 0xFFFF00000000003FULL);
+
+	if (!adev->mc.translate_further)
+		return;
+
+	if (level == AMDGPU_VM_PDB1) {
+		/* Set the block fragment size */
+		if (!(*flags & AMDGPU_PDE_PTE))
+			*flags |= AMDGPU_PDE_BFS(0x9);
+
+	} else if (level == AMDGPU_VM_PDB0) {
+		if (*flags & AMDGPU_PDE_PTE)
+			*flags &= ~AMDGPU_PDE_PTE;
+		else
+			*flags |= AMDGPU_PTE_TF;
+	}
 }
 
 static const struct amdgpu_gart_funcs gmc_v9_0_gart_funcs = {
@@ -502,6 +514,14 @@ static int gmc_v9_0_early_init(void *handle)
 	gmc_v9_0_set_gart_funcs(adev);
 	gmc_v9_0_set_irq_funcs(adev);
 
+	adev->mc.shared_aperture_start = 0x2000000000000000ULL;
+	adev->mc.shared_aperture_end =
+		adev->mc.shared_aperture_start + (4ULL << 30) - 1;
+	adev->mc.private_aperture_start =
+		adev->mc.shared_aperture_end + 1;
+	adev->mc.private_aperture_end =
+		adev->mc.private_aperture_start + (4ULL << 30) - 1;
+
 	return 0;
 }
 
@@ -633,8 +653,8 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
 	u64 base = 0;
 	if (!amdgpu_sriov_vf(adev))
 		base = mmhub_v1_0_get_fb_location(adev);
-	amdgpu_vram_location(adev, &adev->mc, base);
-	amdgpu_gart_location(adev, mc);
+	amdgpu_device_vram_location(adev, &adev->mc, base);
+	amdgpu_device_gart_location(adev, mc);
 	/* base offset of vram pages */
 	if (adev->flags & AMD_IS_APU)
 		adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev);
@@ -700,8 +720,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 
 	/* size in MB on si */
 	adev->mc.mc_vram_size =
-		((adev->flags & AMD_IS_APU) ? nbio_v7_0_get_memsize(adev) :
-		 nbio_v6_1_get_memsize(adev)) * 1024ULL * 1024ULL;
+		adev->nbio_funcs->get_memsize(adev) * 1024ULL * 1024ULL;
 	adev->mc.real_vram_size = adev->mc.mc_vram_size;
 
 	if (!(adev->flags & AMD_IS_APU)) {
@@ -769,11 +788,14 @@ static int gmc_v9_0_sw_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_RAVEN:
 		adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN;
-		if (adev->rev_id == 0x0 || adev->rev_id == 0x1)
+		if (adev->rev_id == 0x0 || adev->rev_id == 0x1) {
 			amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 48);
-		else
-			/* vm_size is 64GB for legacy 2-level page support */
-			amdgpu_vm_adjust_size(adev, 64, 9, 1, 48);
+		} else {
+			/* vm_size is 128TB + 512GB for legacy 3-level page support */
+			amdgpu_vm_adjust_size(adev, 128 * 1024 + 512, 9, 2, 48);
+			adev->mc.translate_further =
+				adev->vm_manager.num_level > 1;
+		}
 		break;
 	case CHIP_VEGA10:
 		/* XXX Don't know how to get VRAM type yet. */
@@ -883,17 +905,18 @@ static int gmc_v9_0_sw_fini(void *handle)
 
 static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev)
 {
+
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						golden_settings_mmhub_1_0_0,
 						ARRAY_SIZE(golden_settings_mmhub_1_0_0));
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						golden_settings_athub_1_0_0,
 						ARRAY_SIZE(golden_settings_athub_1_0_0));
 		break;
 	case CHIP_RAVEN:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						golden_settings_athub_1_0_0,
 						ARRAY_SIZE(golden_settings_athub_1_0_0));
 		break;
@@ -913,9 +936,9 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	bool value;
 	u32 tmp;
 
-	amdgpu_program_register_sequence(adev,
-		golden_settings_vega10_hdp,
-		ARRAY_SIZE(golden_settings_vega10_hdp));
+	amdgpu_device_program_register_sequence(adev,
+						golden_settings_vega10_hdp,
+						ARRAY_SIZE(golden_settings_vega10_hdp));
 
 	if (adev->gart.robj == NULL) {
 		dev_err(adev->dev, "No VRAM object for PCIE GART.\n");
@@ -948,10 +971,7 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev)
 	WREG32_SOC15(HDP, 0, mmHDP_HOST_PATH_CNTL, tmp);
 
 	/* After HDP is initialized, flush HDP.*/
-	if (adev->flags & AMD_IS_APU)
-		nbio_v7_0_hdp_flush(adev);
-	else
-		nbio_v6_1_hdp_flush(adev);
+	adev->nbio_funcs->hdp_flush(adev);
 
 	if (amdgpu_vm_fault_stop == AMDGPU_VM_FAULT_STOP_ALWAYS)
 		value = false;

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

@@ -1682,8 +1682,8 @@ static void kv_dpm_powergate_uvd(void *handle, bool gate)
 
 	if (gate) {
 		/* stop the UVD block */
-		ret = amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							AMD_PG_STATE_GATE);
+		ret = amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							     AMD_PG_STATE_GATE);
 		kv_update_uvd_dpm(adev, gate);
 		if (pi->caps_uvd_pg)
 			/* power off the UVD block */
@@ -1695,8 +1695,8 @@ static void kv_dpm_powergate_uvd(void *handle, bool gate)
 			/* re-init the UVD block */
 		kv_update_uvd_dpm(adev, gate);
 
-		ret = amdgpu_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
-							AMD_PG_STATE_UNGATE);
+		ret = amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_UVD,
+							     AMD_PG_STATE_UNGATE);
 	}
 }
 

drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c

@@ -23,7 +23,6 @@
 #include "amdgpu.h"
 #include "mmhub_v1_0.h"
 
-#include "soc15ip.h"
 #include "mmhub/mmhub_1_0_offset.h"
 #include "mmhub/mmhub_1_0_sh_mask.h"
 #include "mmhub/mmhub_1_0_default.h"
@@ -156,10 +155,15 @@ static void mmhub_v1_0_init_cache_regs(struct amdgpu_device *adev)
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL2, INVALIDATE_L2_CACHE, 1);
 	WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL2, tmp);
 
-	tmp = mmVM_L2_CNTL3_DEFAULT;
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 9);
-	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
-	WREG32_SOC15(MMHUB, 0, mmVM_L2_CNTL3, tmp);
+	if (adev->mc.translate_further) {
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 12);
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 9);
+	} else {
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3, BANK_SELECT, 9);
+		tmp = REG_SET_FIELD(tmp, VM_L2_CNTL3,
+				    L2_CACHE_BIGK_FRAGMENT_SIZE, 6);
+	}
 
 	tmp = mmVM_L2_CNTL4_DEFAULT;
 	tmp = REG_SET_FIELD(tmp, VM_L2_CNTL4, VMC_TAP_PDE_REQUEST_PHYSICAL, 0);
@@ -197,32 +201,40 @@ static void mmhub_v1_0_disable_identity_aperture(struct amdgpu_device *adev)
 
 static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev)
 {
-	int i;
+	unsigned num_level, block_size;
 	uint32_t tmp;
+	int i;
+
+	num_level = adev->vm_manager.num_level;
+	block_size = adev->vm_manager.block_size;
+	if (adev->mc.translate_further)
+		num_level -= 1;
+	else
+		block_size -= 9;
 
 	for (i = 0; i <= 14; i++) {
 		tmp = RREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL, i);
+		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, ENABLE_CONTEXT, 1);
+		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, PAGE_TABLE_DEPTH,
+				    num_level);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				ENABLE_CONTEXT, 1);
-		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				PAGE_TABLE_DEPTH, adev->vm_manager.num_level);
-		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    DUMMY_PAGE_PROTECTION_FAULT_ENABLE_DEFAULT,
+				    1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				PDE0_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    PDE0_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				VALID_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    VALID_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				READ_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    READ_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				WRITE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    WRITE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
+				    EXECUTE_PROTECTION_FAULT_ENABLE_DEFAULT, 1);
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
-				PAGE_TABLE_BLOCK_SIZE,
-				adev->vm_manager.block_size - 9);
+				    PAGE_TABLE_BLOCK_SIZE,
+				    block_size);
 		/* Send no-retry XNACK on fault to suppress VM fault storm. */
 		tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL,
 				    RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0);

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

@@ -22,7 +22,6 @@
  */
 
 #include "amdgpu.h"
-#include "soc15ip.h"
 #include "nbio/nbio_6_1_offset.h"
 #include "nbio/nbio_6_1_sh_mask.h"
 #include "gc/gc_9_0_offset.h"
@@ -254,7 +253,7 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work)
 	}
 
 	/* Trigger recovery due to world switch failure */
-	amdgpu_gpu_recover(adev, NULL);
+	amdgpu_device_gpu_recover(adev, NULL, false);
 }
 
 static int xgpu_ai_set_mailbox_rcv_irq(struct amdgpu_device *adev,
@@ -278,7 +277,7 @@ static int xgpu_ai_mailbox_rcv_irq(struct amdgpu_device *adev,
 	int r;
 
 	/* trigger gpu-reset by hypervisor only if TDR disbaled */
-	if (amdgpu_lockup_timeout == 0) {
+	if (!amdgpu_gpu_recovery) {
 		/* see what event we get */
 		r = xgpu_ai_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION);
 

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

@@ -279,32 +279,32 @@ void xgpu_vi_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 xgpu_fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(
-						 xgpu_fiji_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 xgpu_fiji_golden_settings_a10,
-						 ARRAY_SIZE(
-						 xgpu_fiji_golden_settings_a10));
-		amdgpu_program_register_sequence(adev,
-						 xgpu_fiji_golden_common_all,
-						 ARRAY_SIZE(
-						 xgpu_fiji_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(
+								xgpu_fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_fiji_golden_settings_a10,
+							ARRAY_SIZE(
+								xgpu_fiji_golden_settings_a10));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_fiji_golden_common_all,
+							ARRAY_SIZE(
+								xgpu_fiji_golden_common_all));
 		break;
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 xgpu_tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(
-						 xgpu_tonga_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 xgpu_tonga_golden_settings_a11,
-						 ARRAY_SIZE(
-						 xgpu_tonga_golden_settings_a11));
-		amdgpu_program_register_sequence(adev,
-						 xgpu_tonga_golden_common_all,
-						 ARRAY_SIZE(
-						 xgpu_tonga_golden_common_all));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(
+								xgpu_tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_tonga_golden_settings_a11,
+							ARRAY_SIZE(
+								xgpu_tonga_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							xgpu_tonga_golden_common_all,
+							ARRAY_SIZE(
+								xgpu_tonga_golden_common_all));
 		break;
 	default:
 		BUG_ON("Doesn't support chip type.\n");
@@ -521,7 +521,7 @@ static void xgpu_vi_mailbox_flr_work(struct work_struct *work)
 	}
 
 	/* Trigger recovery due to world switch failure */
-	amdgpu_gpu_recover(adev, NULL);
+	amdgpu_device_gpu_recover(adev, NULL, false);
 }
 
 static int xgpu_vi_set_mailbox_rcv_irq(struct amdgpu_device *adev,
@@ -545,7 +545,7 @@ static int xgpu_vi_mailbox_rcv_irq(struct amdgpu_device *adev,
 	int r;
 
 	/* trigger gpu-reset by hypervisor only if TDR disbaled */
-	if (amdgpu_lockup_timeout == 0) {
+	if (!amdgpu_gpu_recovery) {
 		/* see what event we get */
 		r = xgpu_vi_mailbox_rcv_msg(adev, IDH_FLR_NOTIFICATION);
 

drivers/gpu/drm/amd/amdgpu/nbio_v6_1.c

@@ -24,7 +24,6 @@
 #include "amdgpu_atombios.h"
 #include "nbio_v6_1.h"
 
-#include "soc15ip.h"
 #include "nbio/nbio_6_1_default.h"
 #include "nbio/nbio_6_1_offset.h"
 #include "nbio/nbio_6_1_sh_mask.h"
@@ -34,7 +33,7 @@
 #define smnPCIE_CNTL2                                                                                   0x11180070
 #define smnPCIE_CONFIG_CNTL                                                                             0x11180044
 
-u32 nbio_v6_1_get_rev_id(struct amdgpu_device *adev)
+static u32 nbio_v6_1_get_rev_id(struct amdgpu_device *adev)
 {
         u32 tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
 
@@ -44,19 +43,7 @@ u32 nbio_v6_1_get_rev_id(struct amdgpu_device *adev)
 	return tmp;
 }
 
-u32 nbio_v6_1_get_atombios_scratch_regs(struct amdgpu_device *adev,
-					uint32_t idx)
-{
-	return RREG32_SOC15_OFFSET(NBIO, 0, mmBIOS_SCRATCH_0, idx);
-}
-
-void nbio_v6_1_set_atombios_scratch_regs(struct amdgpu_device *adev,
-					 uint32_t idx, uint32_t val)
-{
-	WREG32_SOC15_OFFSET(NBIO, 0, mmBIOS_SCRATCH_0, idx, val);
-}
-
-void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable)
+static void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable)
 {
 	if (enable)
 		WREG32_SOC15(NBIO, 0, mmBIF_FB_EN,
@@ -66,26 +53,23 @@ void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable)
 		WREG32_SOC15(NBIO, 0, mmBIF_FB_EN, 0);
 }
 
-void nbio_v6_1_hdp_flush(struct amdgpu_device *adev)
+static void nbio_v6_1_hdp_flush(struct amdgpu_device *adev)
 {
 	WREG32_SOC15_NO_KIQ(NBIO, 0, mmBIF_BX_PF0_HDP_MEM_COHERENCY_FLUSH_CNTL, 0);
 }
 
-u32 nbio_v6_1_get_memsize(struct amdgpu_device *adev)
+static u32 nbio_v6_1_get_memsize(struct amdgpu_device *adev)
 {
 	return RREG32_SOC15(NBIO, 0, mmRCC_PF_0_0_RCC_CONFIG_MEMSIZE);
 }
 
-static const u32 nbio_sdma_doorbell_range_reg[] =
-{
-	SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA0_DOORBELL_RANGE),
-	SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE)
-};
-
-void nbio_v6_1_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
+static void nbio_v6_1_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
 				  bool use_doorbell, int doorbell_index)
 {
-	u32 doorbell_range = RREG32(nbio_sdma_doorbell_range_reg[instance]);
+	u32 reg = instance == 0 ? SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA0_DOORBELL_RANGE) :
+			SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE);
+
+	u32 doorbell_range = RREG32(reg);
 
 	if (use_doorbell) {
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, OFFSET, doorbell_index);
@@ -93,17 +77,18 @@ void nbio_v6_1_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
 	} else
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 0);
 
-	WREG32(nbio_sdma_doorbell_range_reg[instance], doorbell_range);
+	WREG32(reg, doorbell_range);
+
 }
 
-void nbio_v6_1_enable_doorbell_aperture(struct amdgpu_device *adev,
-					bool enable)
+static void nbio_v6_1_enable_doorbell_aperture(struct amdgpu_device *adev,
+					       bool enable)
 {
 	WREG32_FIELD15(NBIO, 0, RCC_PF_0_0_RCC_DOORBELL_APER_EN, BIF_DOORBELL_APER_EN, enable ? 1 : 0);
 }
 
-void nbio_v6_1_enable_doorbell_selfring_aperture(struct amdgpu_device *adev,
-					bool enable)
+static void nbio_v6_1_enable_doorbell_selfring_aperture(struct amdgpu_device *adev,
+							bool enable)
 {
 	u32 tmp = 0;
 
@@ -122,8 +107,8 @@ void nbio_v6_1_enable_doorbell_selfring_aperture(struct amdgpu_device *adev,
 }
 
 
-void nbio_v6_1_ih_doorbell_range(struct amdgpu_device *adev,
-				bool use_doorbell, int doorbell_index)
+static void nbio_v6_1_ih_doorbell_range(struct amdgpu_device *adev,
+					bool use_doorbell, int doorbell_index)
 {
 	u32 ih_doorbell_range = RREG32_SOC15(NBIO, 0 , mmBIF_IH_DOORBELL_RANGE);
 
@@ -136,7 +121,7 @@ void nbio_v6_1_ih_doorbell_range(struct amdgpu_device *adev,
 	WREG32_SOC15(NBIO, 0, mmBIF_IH_DOORBELL_RANGE, ih_doorbell_range);
 }
 
-void nbio_v6_1_ih_control(struct amdgpu_device *adev)
+static void nbio_v6_1_ih_control(struct amdgpu_device *adev)
 {
 	u32 interrupt_cntl;
 
@@ -152,8 +137,8 @@ void nbio_v6_1_ih_control(struct amdgpu_device *adev)
 	WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL, interrupt_cntl);
 }
 
-void nbio_v6_1_update_medium_grain_clock_gating(struct amdgpu_device *adev,
-						bool enable)
+static void nbio_v6_1_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						       bool enable)
 {
 	uint32_t def, data;
 
@@ -180,8 +165,8 @@ void nbio_v6_1_update_medium_grain_clock_gating(struct amdgpu_device *adev,
 		WREG32_PCIE(smnCPM_CONTROL, data);
 }
 
-void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev,
-					       bool enable)
+static void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev,
+						      bool enable)
 {
 	uint32_t def, data;
 
@@ -200,7 +185,8 @@ void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev,
 		WREG32_PCIE(smnPCIE_CNTL2, data);
 }
 
-void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev, u32 *flags)
+static void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev,
+					    u32 *flags)
 {
 	int data;
 
@@ -215,9 +201,27 @@ void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev, u32 *flags)
 		*flags |= AMD_CG_SUPPORT_BIF_LS;
 }
 
-const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
-	.hdp_flush_req_offset = SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_GPU_HDP_FLUSH_REQ),
-	.hdp_flush_done_offset = SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_GPU_HDP_FLUSH_DONE),
+static u32 nbio_v6_1_get_hdp_flush_req_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_GPU_HDP_FLUSH_REQ);
+}
+
+static u32 nbio_v6_1_get_hdp_flush_done_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmBIF_BX_PF0_GPU_HDP_FLUSH_DONE);
+}
+
+static u32 nbio_v6_1_get_pcie_index_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX);
+}
+
+static u32 nbio_v6_1_get_pcie_data_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA);
+}
+
+static const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
 	.ref_and_mask_cp0 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP0_MASK,
 	.ref_and_mask_cp1 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP1_MASK,
 	.ref_and_mask_cp2 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__CP2_MASK,
@@ -232,12 +236,7 @@ const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg = {
 	.ref_and_mask_sdma1 = BIF_BX_PF0_GPU_HDP_FLUSH_DONE__SDMA1_MASK
 };
 
-const struct nbio_pcie_index_data nbio_v6_1_pcie_index_data = {
-	.index_offset = SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX),
-	.data_offset = SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA),
-};
-
-void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev)
+static void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev)
 {
 	uint32_t reg;
 
@@ -254,7 +253,7 @@ void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev)
 	}
 }
 
-void nbio_v6_1_init_registers(struct amdgpu_device *adev)
+static void nbio_v6_1_init_registers(struct amdgpu_device *adev)
 {
 	uint32_t def, data;
 
@@ -265,3 +264,25 @@ void nbio_v6_1_init_registers(struct amdgpu_device *adev)
 	if (def != data)
 		WREG32_PCIE(smnPCIE_CONFIG_CNTL, data);
 }
+
+const struct amdgpu_nbio_funcs nbio_v6_1_funcs = {
+	.hdp_flush_reg = &nbio_v6_1_hdp_flush_reg,
+	.get_hdp_flush_req_offset = nbio_v6_1_get_hdp_flush_req_offset,
+	.get_hdp_flush_done_offset = nbio_v6_1_get_hdp_flush_done_offset,
+	.get_pcie_index_offset = nbio_v6_1_get_pcie_index_offset,
+	.get_pcie_data_offset = nbio_v6_1_get_pcie_data_offset,
+	.get_rev_id = nbio_v6_1_get_rev_id,
+	.mc_access_enable = nbio_v6_1_mc_access_enable,
+	.hdp_flush = nbio_v6_1_hdp_flush,
+	.get_memsize = nbio_v6_1_get_memsize,
+	.sdma_doorbell_range = nbio_v6_1_sdma_doorbell_range,
+	.enable_doorbell_aperture = nbio_v6_1_enable_doorbell_aperture,
+	.enable_doorbell_selfring_aperture = nbio_v6_1_enable_doorbell_selfring_aperture,
+	.ih_doorbell_range = nbio_v6_1_ih_doorbell_range,
+	.update_medium_grain_clock_gating = nbio_v6_1_update_medium_grain_clock_gating,
+	.update_medium_grain_light_sleep = nbio_v6_1_update_medium_grain_light_sleep,
+	.get_clockgating_state = nbio_v6_1_get_clockgating_state,
+	.ih_control = nbio_v6_1_ih_control,
+	.init_registers = nbio_v6_1_init_registers,
+	.detect_hw_virt = nbio_v6_1_detect_hw_virt,
+};

drivers/gpu/drm/amd/amdgpu/nbio_v6_1.h

@@ -26,30 +26,6 @@
 
 #include "soc15_common.h"
 
-extern const struct nbio_hdp_flush_reg nbio_v6_1_hdp_flush_reg;
-extern const struct nbio_pcie_index_data nbio_v6_1_pcie_index_data;
-int nbio_v6_1_init(struct amdgpu_device *adev);
-u32 nbio_v6_1_get_atombios_scratch_regs(struct amdgpu_device *adev,
-                                        uint32_t idx);
-void nbio_v6_1_set_atombios_scratch_regs(struct amdgpu_device *adev,
-                                         uint32_t idx, uint32_t val);
-void nbio_v6_1_mc_access_enable(struct amdgpu_device *adev, bool enable);
-void nbio_v6_1_hdp_flush(struct amdgpu_device *adev);
-u32 nbio_v6_1_get_memsize(struct amdgpu_device *adev);
-void nbio_v6_1_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
-				  bool use_doorbell, int doorbell_index);
-void nbio_v6_1_enable_doorbell_aperture(struct amdgpu_device *adev,
-					bool enable);
-void nbio_v6_1_enable_doorbell_selfring_aperture(struct amdgpu_device *adev,
-					bool enable);
-void nbio_v6_1_ih_doorbell_range(struct amdgpu_device *adev,
-				bool use_doorbell, int doorbell_index);
-void nbio_v6_1_ih_control(struct amdgpu_device *adev);
-u32 nbio_v6_1_get_rev_id(struct amdgpu_device *adev);
-void nbio_v6_1_update_medium_grain_clock_gating(struct amdgpu_device *adev, bool enable);
-void nbio_v6_1_update_medium_grain_light_sleep(struct amdgpu_device *adev, bool enable);
-void nbio_v6_1_get_clockgating_state(struct amdgpu_device *adev, u32 *flags);
-void nbio_v6_1_detect_hw_virt(struct amdgpu_device *adev);
-void nbio_v6_1_init_registers(struct amdgpu_device *adev);
+extern const struct amdgpu_nbio_funcs nbio_v6_1_funcs;
 
 #endif

drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c

@@ -24,7 +24,6 @@
 #include "amdgpu_atombios.h"
 #include "nbio_v7_0.h"
 
-#include "soc15ip.h"
 #include "nbio/nbio_7_0_default.h"
 #include "nbio/nbio_7_0_offset.h"
 #include "nbio/nbio_7_0_sh_mask.h"
@@ -32,7 +31,10 @@
 
 #define smnNBIF_MGCG_CTRL_LCLK	0x1013a05c
 
-u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev)
+#define smnCPM_CONTROL                                                                                  0x11180460
+#define smnPCIE_CNTL2                                                                                   0x11180070
+
+static u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev)
 {
         u32 tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
 
@@ -42,19 +44,7 @@ u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev)
 	return tmp;
 }
 
-u32 nbio_v7_0_get_atombios_scratch_regs(struct amdgpu_device *adev,
-					uint32_t idx)
-{
-	return RREG32_SOC15_OFFSET(NBIO, 0, mmBIOS_SCRATCH_0, idx);
-}
-
-void nbio_v7_0_set_atombios_scratch_regs(struct amdgpu_device *adev,
-					 uint32_t idx, uint32_t val)
-{
-	WREG32_SOC15_OFFSET(NBIO, 0, mmBIOS_SCRATCH_0, idx, val);
-}
-
-void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable)
+static void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable)
 {
 	if (enable)
 		WREG32_SOC15(NBIO, 0, mmBIF_FB_EN,
@@ -63,26 +53,23 @@ void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable)
 		WREG32_SOC15(NBIO, 0, mmBIF_FB_EN, 0);
 }
 
-void nbio_v7_0_hdp_flush(struct amdgpu_device *adev)
+static void nbio_v7_0_hdp_flush(struct amdgpu_device *adev)
 {
 	WREG32_SOC15_NO_KIQ(NBIO, 0, mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0);
 }
 
-u32 nbio_v7_0_get_memsize(struct amdgpu_device *adev)
+static u32 nbio_v7_0_get_memsize(struct amdgpu_device *adev)
 {
 	return RREG32_SOC15(NBIO, 0, mmRCC_CONFIG_MEMSIZE);
 }
 
-static const u32 nbio_sdma_doorbell_range_reg[] =
+static void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
+					  bool use_doorbell, int doorbell_index)
 {
-	SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA0_DOORBELL_RANGE),
-	SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE)
-};
+	u32 reg = instance == 0 ? SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA0_DOORBELL_RANGE) :
+			SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE);
 
-void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
-				  bool use_doorbell, int doorbell_index)
-{
-	u32 doorbell_range = RREG32(nbio_sdma_doorbell_range_reg[instance]);
+	u32 doorbell_range = RREG32(reg);
 
 	if (use_doorbell) {
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, OFFSET, doorbell_index);
@@ -90,17 +77,23 @@ void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
 	} else
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 0);
 
-	WREG32(nbio_sdma_doorbell_range_reg[instance], doorbell_range);
+	WREG32(reg, doorbell_range);
 }
 
-void nbio_v7_0_enable_doorbell_aperture(struct amdgpu_device *adev,
-					bool enable)
+static void nbio_v7_0_enable_doorbell_aperture(struct amdgpu_device *adev,
+					       bool enable)
 {
 	WREG32_FIELD15(NBIO, 0, RCC_DOORBELL_APER_EN, BIF_DOORBELL_APER_EN, enable ? 1 : 0);
 }
 
-void nbio_v7_0_ih_doorbell_range(struct amdgpu_device *adev,
-				bool use_doorbell, int doorbell_index)
+static void nbio_v7_0_enable_doorbell_selfring_aperture(struct amdgpu_device *adev,
+							bool enable)
+{
+
+}
+
+static void nbio_v7_0_ih_doorbell_range(struct amdgpu_device *adev,
+					bool use_doorbell, int doorbell_index)
 {
 	u32 ih_doorbell_range = RREG32_SOC15(NBIO, 0 , mmBIF_IH_DOORBELL_RANGE);
 
@@ -130,8 +123,8 @@ static void nbio_7_0_write_syshub_ind_mmr(struct amdgpu_device *adev, uint32_t o
 	WREG32_SOC15(NBIO, 0, mmSYSHUB_DATA, data);
 }
 
-void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *adev,
-						bool enable)
+static void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						       bool enable)
 {
 	uint32_t def, data;
 
@@ -169,7 +162,43 @@ void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *adev,
 		nbio_7_0_write_syshub_ind_mmr(adev, ixSYSHUB_MMREG_IND_SYSHUB_MGCG_CTRL_SHUBCLK, data);
 }
 
-void nbio_v7_0_ih_control(struct amdgpu_device *adev)
+static void nbio_v7_0_update_medium_grain_light_sleep(struct amdgpu_device *adev,
+						      bool enable)
+{
+	uint32_t def, data;
+
+	def = data = RREG32_PCIE(smnPCIE_CNTL2);
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_BIF_LS)) {
+		data |= (PCIE_CNTL2__SLV_MEM_LS_EN_MASK |
+			 PCIE_CNTL2__MST_MEM_LS_EN_MASK |
+			 PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK);
+	} else {
+		data &= ~(PCIE_CNTL2__SLV_MEM_LS_EN_MASK |
+			  PCIE_CNTL2__MST_MEM_LS_EN_MASK |
+			  PCIE_CNTL2__REPLAY_MEM_LS_EN_MASK);
+	}
+
+	if (def != data)
+		WREG32_PCIE(smnPCIE_CNTL2, data);
+}
+
+static void nbio_v7_0_get_clockgating_state(struct amdgpu_device *adev,
+					    u32 *flags)
+{
+	int data;
+
+	/* AMD_CG_SUPPORT_BIF_MGCG */
+	data = RREG32_PCIE(smnCPM_CONTROL);
+	if (data & CPM_CONTROL__LCLK_DYN_GATE_ENABLE_MASK)
+		*flags |= AMD_CG_SUPPORT_BIF_MGCG;
+
+	/* AMD_CG_SUPPORT_BIF_LS */
+	data = RREG32_PCIE(smnPCIE_CNTL2);
+	if (data & PCIE_CNTL2__SLV_MEM_LS_EN_MASK)
+		*flags |= AMD_CG_SUPPORT_BIF_LS;
+}
+
+static void nbio_v7_0_ih_control(struct amdgpu_device *adev)
 {
 	u32 interrupt_cntl;
 
@@ -185,9 +214,27 @@ void nbio_v7_0_ih_control(struct amdgpu_device *adev)
 	WREG32_SOC15(NBIO, 0, mmINTERRUPT_CNTL, interrupt_cntl);
 }
 
+static u32 nbio_v7_0_get_hdp_flush_req_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmGPU_HDP_FLUSH_REQ);
+}
+
+static u32 nbio_v7_0_get_hdp_flush_done_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmGPU_HDP_FLUSH_DONE);
+}
+
+static u32 nbio_v7_0_get_pcie_index_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX2);
+}
+
+static u32 nbio_v7_0_get_pcie_data_offset(struct amdgpu_device *adev)
+{
+	return SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2);
+}
+
 const struct nbio_hdp_flush_reg nbio_v7_0_hdp_flush_reg = {
-	.hdp_flush_req_offset = SOC15_REG_OFFSET(NBIO, 0, mmGPU_HDP_FLUSH_REQ),
-	.hdp_flush_done_offset = SOC15_REG_OFFSET(NBIO, 0, mmGPU_HDP_FLUSH_DONE),
 	.ref_and_mask_cp0 = GPU_HDP_FLUSH_DONE__CP0_MASK,
 	.ref_and_mask_cp1 = GPU_HDP_FLUSH_DONE__CP1_MASK,
 	.ref_and_mask_cp2 = GPU_HDP_FLUSH_DONE__CP2_MASK,
@@ -202,7 +249,35 @@ const struct nbio_hdp_flush_reg nbio_v7_0_hdp_flush_reg = {
 	.ref_and_mask_sdma1 = GPU_HDP_FLUSH_DONE__SDMA1_MASK,
 };
 
-const struct nbio_pcie_index_data nbio_v7_0_pcie_index_data = {
-	.index_offset = SOC15_REG_OFFSET(NBIO, 0, mmPCIE_INDEX2),
-	.data_offset = SOC15_REG_OFFSET(NBIO, 0, mmPCIE_DATA2)
+static void nbio_v7_0_detect_hw_virt(struct amdgpu_device *adev)
+{
+	if (is_virtual_machine())	/* passthrough mode exclus sriov mod */
+		adev->virt.caps |= AMDGPU_PASSTHROUGH_MODE;
+}
+
+static void nbio_v7_0_init_registers(struct amdgpu_device *adev)
+{
+
+}
+
+const struct amdgpu_nbio_funcs nbio_v7_0_funcs = {
+	.hdp_flush_reg = &nbio_v7_0_hdp_flush_reg,
+	.get_hdp_flush_req_offset = nbio_v7_0_get_hdp_flush_req_offset,
+	.get_hdp_flush_done_offset = nbio_v7_0_get_hdp_flush_done_offset,
+	.get_pcie_index_offset = nbio_v7_0_get_pcie_index_offset,
+	.get_pcie_data_offset = nbio_v7_0_get_pcie_data_offset,
+	.get_rev_id = nbio_v7_0_get_rev_id,
+	.mc_access_enable = nbio_v7_0_mc_access_enable,
+	.hdp_flush = nbio_v7_0_hdp_flush,
+	.get_memsize = nbio_v7_0_get_memsize,
+	.sdma_doorbell_range = nbio_v7_0_sdma_doorbell_range,
+	.enable_doorbell_aperture = nbio_v7_0_enable_doorbell_aperture,
+	.enable_doorbell_selfring_aperture = nbio_v7_0_enable_doorbell_selfring_aperture,
+	.ih_doorbell_range = nbio_v7_0_ih_doorbell_range,
+	.update_medium_grain_clock_gating = nbio_v7_0_update_medium_grain_clock_gating,
+	.update_medium_grain_light_sleep = nbio_v7_0_update_medium_grain_light_sleep,
+	.get_clockgating_state = nbio_v7_0_get_clockgating_state,
+	.ih_control = nbio_v7_0_ih_control,
+	.init_registers = nbio_v7_0_init_registers,
+	.detect_hw_virt = nbio_v7_0_detect_hw_virt,
 };

drivers/gpu/drm/amd/amdgpu/nbio_v7_0.h

@@ -26,24 +26,6 @@
 
 #include "soc15_common.h"
 
-extern const struct nbio_hdp_flush_reg nbio_v7_0_hdp_flush_reg;
-extern const struct nbio_pcie_index_data nbio_v7_0_pcie_index_data;
-int nbio_v7_0_init(struct amdgpu_device *adev);
-u32 nbio_v7_0_get_atombios_scratch_regs(struct amdgpu_device *adev,
-                                        uint32_t idx);
-void nbio_v7_0_set_atombios_scratch_regs(struct amdgpu_device *adev,
-                                         uint32_t idx, uint32_t val);
-void nbio_v7_0_mc_access_enable(struct amdgpu_device *adev, bool enable);
-void nbio_v7_0_hdp_flush(struct amdgpu_device *adev);
-u32 nbio_v7_0_get_memsize(struct amdgpu_device *adev);
-void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instance,
-				  bool use_doorbell, int doorbell_index);
-void nbio_v7_0_enable_doorbell_aperture(struct amdgpu_device *adev,
-					bool enable);
-void nbio_v7_0_ih_doorbell_range(struct amdgpu_device *adev,
-				bool use_doorbell, int doorbell_index);
-void nbio_v7_0_ih_control(struct amdgpu_device *adev);
-u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev);
-void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *adev,
-						bool enable);
+extern const struct amdgpu_nbio_funcs nbio_v7_0_funcs;
+
 #endif

drivers/gpu/drm/amd/amdgpu/psp_v10_0.c

@@ -30,7 +30,6 @@
 #include "soc15_common.h"
 #include "psp_v10_0.h"
 
-#include "soc15ip.h"
 #include "mp/mp_10_0_offset.h"
 #include "gc/gc_9_1_offset.h"
 #include "sdma0/sdma0_4_1_offset.h"
@@ -298,9 +297,10 @@ int psp_v10_0_cmd_submit(struct psp_context *psp,
 }
 
 static int
-psp_v10_0_sram_map(unsigned int *sram_offset, unsigned int *sram_addr_reg_offset,
-		  unsigned int *sram_data_reg_offset,
-		  enum AMDGPU_UCODE_ID ucode_id)
+psp_v10_0_sram_map(struct amdgpu_device *adev,
+		unsigned int *sram_offset, unsigned int *sram_addr_reg_offset,
+		unsigned int *sram_data_reg_offset,
+		enum AMDGPU_UCODE_ID ucode_id)
 {
 	int ret = 0;
 
@@ -395,7 +395,7 @@ bool psp_v10_0_compare_sram_data(struct psp_context *psp,
 	uint32_t *ucode_mem = NULL;
 	struct amdgpu_device *adev = psp->adev;
 
-	err = psp_v10_0_sram_map(&fw_sram_reg_val, &fw_sram_addr_reg_offset,
+	err = psp_v10_0_sram_map(adev, &fw_sram_reg_val, &fw_sram_addr_reg_offset,
 				&fw_sram_data_reg_offset, ucode_type);
 	if (err)
 		return false;

drivers/gpu/drm/amd/amdgpu/psp_v3_1.c

@@ -31,7 +31,6 @@
 #include "soc15_common.h"
 #include "psp_v3_1.h"
 
-#include "soc15ip.h"
 #include "mp/mp_9_0_offset.h"
 #include "mp/mp_9_0_sh_mask.h"
 #include "gc/gc_9_0_offset.h"
@@ -410,9 +409,10 @@ int psp_v3_1_cmd_submit(struct psp_context *psp,
 }
 
 static int
-psp_v3_1_sram_map(unsigned int *sram_offset, unsigned int *sram_addr_reg_offset,
-		  unsigned int *sram_data_reg_offset,
-		  enum AMDGPU_UCODE_ID ucode_id)
+psp_v3_1_sram_map(struct amdgpu_device *adev,
+		unsigned int *sram_offset, unsigned int *sram_addr_reg_offset,
+		unsigned int *sram_data_reg_offset,
+		enum AMDGPU_UCODE_ID ucode_id)
 {
 	int ret = 0;
 
@@ -507,7 +507,7 @@ bool psp_v3_1_compare_sram_data(struct psp_context *psp,
 	uint32_t *ucode_mem = NULL;
 	struct amdgpu_device *adev = psp->adev;
 
-	err = psp_v3_1_sram_map(&fw_sram_reg_val, &fw_sram_addr_reg_offset,
+	err = psp_v3_1_sram_map(adev, &fw_sram_reg_val, &fw_sram_addr_reg_offset,
 				&fw_sram_data_reg_offset, ucode_type);
 	if (err)
 		return false;

drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c

@@ -93,12 +93,12 @@ static void sdma_v2_4_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_TOPAZ:
-		amdgpu_program_register_sequence(adev,
-						 iceland_mgcg_cgcg_init,
-						 ARRAY_SIZE(iceland_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_iceland_a11,
-						 ARRAY_SIZE(golden_settings_iceland_a11));
+		amdgpu_device_program_register_sequence(adev,
+							iceland_mgcg_cgcg_init,
+							ARRAY_SIZE(iceland_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_iceland_a11,
+							ARRAY_SIZE(golden_settings_iceland_a11));
 		break;
 	default:
 		break;
@@ -600,7 +600,7 @@ static int sdma_v2_4_ring_test_ring(struct amdgpu_ring *ring)
 	u32 tmp;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
 		return r;
@@ -613,7 +613,7 @@ static int sdma_v2_4_ring_test_ring(struct amdgpu_ring *ring)
 	r = amdgpu_ring_alloc(ring, 5);
 	if (r) {
 		DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
-		amdgpu_wb_free(adev, index);
+		amdgpu_device_wb_free(adev, index);
 		return r;
 	}
 
@@ -639,7 +639,7 @@ static int sdma_v2_4_ring_test_ring(struct amdgpu_ring *ring)
 			  ring->idx, tmp);
 		r = -EINVAL;
 	}
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 
 	return r;
 }
@@ -662,7 +662,7 @@ static int sdma_v2_4_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	u64 gpu_addr;
 	long r;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
@@ -715,7 +715,7 @@ err1:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err0:
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 

drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c

@@ -192,47 +192,47 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(fiji_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_fiji_a10,
-						 ARRAY_SIZE(golden_settings_fiji_a10));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_fiji_a10,
+							ARRAY_SIZE(golden_settings_fiji_a10));
 		break;
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(tonga_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_tonga_a11,
-						 ARRAY_SIZE(golden_settings_tonga_a11));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_tonga_a11,
+							ARRAY_SIZE(golden_settings_tonga_a11));
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris11_a11,
-						 ARRAY_SIZE(golden_settings_polaris11_a11));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris11_a11,
+							ARRAY_SIZE(golden_settings_polaris11_a11));
 		break;
 	case CHIP_POLARIS10:
-		amdgpu_program_register_sequence(adev,
-						 golden_settings_polaris10_a11,
-						 ARRAY_SIZE(golden_settings_polaris10_a11));
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_polaris10_a11,
+							ARRAY_SIZE(golden_settings_polaris10_a11));
 		break;
 	case CHIP_CARRIZO:
-		amdgpu_program_register_sequence(adev,
-						 cz_mgcg_cgcg_init,
-						 ARRAY_SIZE(cz_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 cz_golden_settings_a11,
-						 ARRAY_SIZE(cz_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							cz_mgcg_cgcg_init,
+							ARRAY_SIZE(cz_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							cz_golden_settings_a11,
+							ARRAY_SIZE(cz_golden_settings_a11));
 		break;
 	case CHIP_STONEY:
-		amdgpu_program_register_sequence(adev,
-						 stoney_mgcg_cgcg_init,
-						 ARRAY_SIZE(stoney_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 stoney_golden_settings_a11,
-						 ARRAY_SIZE(stoney_golden_settings_a11));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_mgcg_cgcg_init,
+							ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_golden_settings_a11,
+							ARRAY_SIZE(stoney_golden_settings_a11));
 		break;
 	default:
 		break;
@@ -355,7 +355,7 @@ static uint64_t sdma_v3_0_ring_get_wptr(struct amdgpu_ring *ring)
 	struct amdgpu_device *adev = ring->adev;
 	u32 wptr;
 
-	if (ring->use_doorbell) {
+	if (ring->use_doorbell || ring->use_pollmem) {
 		/* XXX check if swapping is necessary on BE */
 		wptr = ring->adev->wb.wb[ring->wptr_offs] >> 2;
 	} else {
@@ -380,10 +380,13 @@ static void sdma_v3_0_ring_set_wptr(struct amdgpu_ring *ring)
 
 	if (ring->use_doorbell) {
 		u32 *wb = (u32 *)&adev->wb.wb[ring->wptr_offs];
-
 		/* XXX check if swapping is necessary on BE */
 		WRITE_ONCE(*wb, (lower_32_bits(ring->wptr) << 2));
 		WDOORBELL32(ring->doorbell_index, lower_32_bits(ring->wptr) << 2);
+	} else if (ring->use_pollmem) {
+		u32 *wb = (u32 *)&adev->wb.wb[ring->wptr_offs];
+
+		WRITE_ONCE(*wb, (lower_32_bits(ring->wptr) << 2));
 	} else {
 		int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
 
@@ -718,10 +721,14 @@ static int sdma_v3_0_gfx_resume(struct amdgpu_device *adev)
 		WREG32(mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI + sdma_offsets[i],
 		       upper_32_bits(wptr_gpu_addr));
 		wptr_poll_cntl = RREG32(mmSDMA0_GFX_RB_WPTR_POLL_CNTL + sdma_offsets[i]);
-		if (amdgpu_sriov_vf(adev))
-			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, SDMA0_GFX_RB_WPTR_POLL_CNTL, F32_POLL_ENABLE, 1);
+		if (ring->use_pollmem)
+			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl,
+						       SDMA0_GFX_RB_WPTR_POLL_CNTL,
+						       ENABLE, 1);
 		else
-			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, SDMA0_GFX_RB_WPTR_POLL_CNTL, F32_POLL_ENABLE, 0);
+			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl,
+						       SDMA0_GFX_RB_WPTR_POLL_CNTL,
+						       ENABLE, 0);
 		WREG32(mmSDMA0_GFX_RB_WPTR_POLL_CNTL + sdma_offsets[i], wptr_poll_cntl);
 
 		/* enable DMA RB */
@@ -860,7 +867,7 @@ static int sdma_v3_0_ring_test_ring(struct amdgpu_ring *ring)
 	u32 tmp;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
 		return r;
@@ -873,7 +880,7 @@ static int sdma_v3_0_ring_test_ring(struct amdgpu_ring *ring)
 	r = amdgpu_ring_alloc(ring, 5);
 	if (r) {
 		DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
-		amdgpu_wb_free(adev, index);
+		amdgpu_device_wb_free(adev, index);
 		return r;
 	}
 
@@ -899,7 +906,7 @@ static int sdma_v3_0_ring_test_ring(struct amdgpu_ring *ring)
 			  ring->idx, tmp);
 		r = -EINVAL;
 	}
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 
 	return r;
 }
@@ -922,7 +929,7 @@ static int sdma_v3_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	u64 gpu_addr;
 	long r;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
@@ -974,7 +981,7 @@ err1:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err0:
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 
@@ -1203,9 +1210,13 @@ static int sdma_v3_0_sw_init(void *handle)
 	for (i = 0; i < adev->sdma.num_instances; i++) {
 		ring = &adev->sdma.instance[i].ring;
 		ring->ring_obj = NULL;
-		ring->use_doorbell = true;
-		ring->doorbell_index = (i == 0) ?
-			AMDGPU_DOORBELL_sDMA_ENGINE0 : AMDGPU_DOORBELL_sDMA_ENGINE1;
+		if (!amdgpu_sriov_vf(adev)) {
+			ring->use_doorbell = true;
+			ring->doorbell_index = (i == 0) ?
+				AMDGPU_DOORBELL_sDMA_ENGINE0 : AMDGPU_DOORBELL_sDMA_ENGINE1;
+		} else {
+			ring->use_pollmem = true;
+		}
 
 		sprintf(ring->name, "sdma%d", i);
 		r = amdgpu_ring_init(adev, ring, 1024,

drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c

@@ -27,7 +27,6 @@
 #include "amdgpu_ucode.h"
 #include "amdgpu_trace.h"
 
-#include "soc15ip.h"
 #include "sdma0/sdma0_4_0_offset.h"
 #include "sdma0/sdma0_4_0_sh_mask.h"
 #include "sdma1/sdma1_4_0_offset.h"
@@ -53,95 +52,83 @@ static void sdma_v4_0_set_buffer_funcs(struct amdgpu_device *adev);
 static void sdma_v4_0_set_vm_pte_funcs(struct amdgpu_device *adev);
 static void sdma_v4_0_set_irq_funcs(struct amdgpu_device *adev);
 
-static const u32 golden_settings_sdma_4[] = {
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CHICKEN_BITS), 0xfe931f07, 0x02831d07,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CLK_CTRL), 0xff000ff0, 0x3f000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), 0xfffffff7, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_PAGE_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), 0x003ff006, 0x0003c000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC0_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC1_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_UTCL1_PAGE), 0x000003ff, 0x000003c0,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_CHICKEN_BITS), 0xfe931f07, 0x02831f07,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_CLK_CTRL), 0xffffffff, 0x3f000100,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_PAGE_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_PAGE_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_POWER_CNTL), 0x003ff000, 0x0003c000,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_RLC0_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_RLC0_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_RLC1_IB_CNTL), 0x800f0100, 0x00000100,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_RLC1_RB_WPTR_POLL_CNTL), 0x0000fff0, 0x00403000,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_UTCL1_PAGE), 0x000003ff, 0x000003c0
+static const struct soc15_reg_golden golden_settings_sdma_4[] = {
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CLK_CTRL, 0xff000ff0, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_PAGE_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_POWER_CNTL, 0x003ff006, 0x0003c000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CHICKEN_BITS, 0xfe931f07, 0x02831f07),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GFX_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GFX_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_PAGE_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_PAGE_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_POWER_CNTL, 0x003ff000, 0x0003c000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC0_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC0_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC1_IB_CNTL, 0x800f0100, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC1_RB_WPTR_POLL_CNTL, 0x0000fff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_UTCL1_PAGE, 0x000003ff, 0x000003c0)
 };
 
-static const u32 golden_settings_sdma_vg10[] = {
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG), 0x0018773f, 0x00104002,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ), 0x0018773f, 0x00104002,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG), 0x0018773f, 0x00104002,
-	SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ), 0x0018773f, 0x00104002
+static const struct soc15_reg_golden golden_settings_sdma_vg10[] = {
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0018773f, 0x00104002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00104002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG, 0x0018773f, 0x00104002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00104002)
 };
 
-static const u32 golden_settings_sdma_4_1[] =
+static const struct soc15_reg_golden golden_settings_sdma_4_1[] =
 {
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CHICKEN_BITS), 0xfe931f07, 0x02831d07,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_CLK_CTRL), 0xffffffff, 0x3f000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_IB_CNTL), 0x800f0111, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), 0xfffffff7, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_POWER_CNTL), 0xfc3fffff, 0x40000051,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC0_IB_CNTL), 0x800f0111, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL), 0xfffffff7, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC1_IB_CNTL), 0x800f0111, 0x00000100,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL), 0xfffffff7, 0x00403000,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_UTCL1_PAGE), 0x000003ff, 0x000003c0
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_IB_CNTL, 0x800f0111, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_POWER_CNTL, 0xfc3fffff, 0x40000051),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_IB_CNTL, 0x800f0111, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_IB_CNTL, 0x800f0111, 0x00000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0)
 };
 
-static const u32 golden_settings_sdma_rv1[] =
+static const struct soc15_reg_golden golden_settings_sdma_rv1[] =
 {
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG), 0x0018773f, 0x00000002,
-	SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ), 0x0018773f, 0x00000002
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0018773f, 0x00000002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x0018773f, 0x00000002)
 };
 
-static u32 sdma_v4_0_get_reg_offset(u32 instance, u32 internal_offset)
+static u32 sdma_v4_0_get_reg_offset(struct amdgpu_device *adev,
+		u32 instance, u32 offset)
 {
-	u32 base = 0;
-
-	switch (instance) {
-	case 0:
-		base = SDMA0_BASE.instance[0].segment[0];
-		break;
-	case 1:
-		base = SDMA1_BASE.instance[0].segment[0];
-		break;
-	default:
-		BUG();
-		break;
-	}
-
-	return base + internal_offset;
+	return ( 0 == instance ? (adev->reg_offset[SDMA0_HWIP][0][0] + offset) :
+			(adev->reg_offset[SDMA1_HWIP][0][0] + offset));
 }
 
 static void sdma_v4_0_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_4,
 						 ARRAY_SIZE(golden_settings_sdma_4));
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_vg10,
 						 ARRAY_SIZE(golden_settings_sdma_vg10));
 		break;
 	case CHIP_RAVEN:
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_4_1,
 						 ARRAY_SIZE(golden_settings_sdma_4_1));
-		amdgpu_program_register_sequence(adev,
+		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_rv1,
 						 ARRAY_SIZE(golden_settings_sdma_rv1));
 		break;
@@ -265,8 +252,8 @@ static uint64_t sdma_v4_0_ring_get_wptr(struct amdgpu_ring *ring)
 		int me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
 
 		wptr = &local_wptr;
-		lowbit = RREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR)) >> 2;
-		highbit = RREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2;
+		lowbit = RREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR)) >> 2;
+		highbit = RREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2;
 
 		DRM_DEBUG("wptr [%i]high== 0x%08x low==0x%08x\n",
 				me, highbit, lowbit);
@@ -315,8 +302,8 @@ static void sdma_v4_0_ring_set_wptr(struct amdgpu_ring *ring)
 				lower_32_bits(ring->wptr << 2),
 				me,
 				upper_32_bits(ring->wptr << 2));
-		WREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2));
-		WREG32(sdma_v4_0_get_reg_offset(me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2));
+		WREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2));
+		WREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2));
 	}
 }
 
@@ -370,13 +357,9 @@ static void sdma_v4_0_ring_emit_ib(struct amdgpu_ring *ring,
  */
 static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
 	u32 ref_and_mask = 0;
-	const struct nbio_hdp_flush_reg *nbio_hf_reg;
-
-	if (ring->adev->flags & AMD_IS_APU)
-		nbio_hf_reg = &nbio_v7_0_hdp_flush_reg;
-	else
-		nbio_hf_reg = &nbio_v6_1_hdp_flush_reg;
+	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
 
 	if (ring == &ring->adev->sdma.instance[0].ring)
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0;
@@ -386,8 +369,8 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
 			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) |
 			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
-	amdgpu_ring_write(ring, nbio_hf_reg->hdp_flush_done_offset << 2);
-	amdgpu_ring_write(ring, nbio_hf_reg->hdp_flush_req_offset << 2);
+	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_done_offset(adev)) << 2);
+	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_req_offset(adev)) << 2);
 	amdgpu_ring_write(ring, ref_and_mask); /* reference */
 	amdgpu_ring_write(ring, ref_and_mask); /* mask */
 	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
@@ -396,6 +379,8 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 
 static void sdma_v4_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) |
 			  SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf));
 	amdgpu_ring_write(ring, SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE));
@@ -460,12 +445,12 @@ static void sdma_v4_0_gfx_stop(struct amdgpu_device *adev)
 		amdgpu_ttm_set_active_vram_size(adev, adev->mc.visible_vram_size);
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
-		rb_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_CNTL));
+		rb_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL));
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
-		ib_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_IB_CNTL));
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
+		ib_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL));
 		ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_IB_CNTL), ib_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl);
 	}
 
 	sdma0->ready = false;
@@ -522,18 +507,18 @@ static void sdma_v4_0_ctx_switch_enable(struct amdgpu_device *adev, bool enable)
 	}
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
-		f32_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL));
+		f32_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_CNTL));
 		f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_CNTL,
 				AUTO_CTXSW_ENABLE, enable ? 1 : 0);
 		if (enable && amdgpu_sdma_phase_quantum) {
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE0_QUANTUM),
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_PHASE0_QUANTUM),
 			       phase_quantum);
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE1_QUANTUM),
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_PHASE1_QUANTUM),
 			       phase_quantum);
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_PHASE2_QUANTUM),
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_PHASE2_QUANTUM),
 			       phase_quantum);
 		}
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL), f32_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_CNTL), f32_cntl);
 	}
 
 }
@@ -557,9 +542,9 @@ static void sdma_v4_0_enable(struct amdgpu_device *adev, bool enable)
 	}
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
-		f32_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_F32_CNTL));
+		f32_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL));
 		f32_cntl = REG_SET_FIELD(f32_cntl, SDMA0_F32_CNTL, HALT, enable ? 0 : 1);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_F32_CNTL), f32_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), f32_cntl);
 	}
 }
 
@@ -587,48 +572,48 @@ static int sdma_v4_0_gfx_resume(struct amdgpu_device *adev)
 		ring = &adev->sdma.instance[i].ring;
 		wb_offset = (ring->rptr_offs * 4);
 
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_SEM_WAIT_FAIL_TIMER_CNTL), 0);
 
 		/* Set ring buffer size in dwords */
 		rb_bufsz = order_base_2(ring->ring_size / 4);
-		rb_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_CNTL));
+		rb_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL));
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SIZE, rb_bufsz);
 #ifdef __BIG_ENDIAN
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_SWAP_ENABLE, 1);
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL,
 					RPTR_WRITEBACK_SWAP_ENABLE, 1);
 #endif
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
 
 		/* Initialize the ring buffer's read and write pointers */
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_RPTR), 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_RPTR_HI), 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR), 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_HI), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_HI), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), 0);
 
 		/* set the wb address whether it's enabled or not */
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_RPTR_ADDR_HI),
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_HI),
 		       upper_32_bits(adev->wb.gpu_addr + wb_offset) & 0xFFFFFFFF);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_RPTR_ADDR_LO),
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_RPTR_ADDR_LO),
 		       lower_32_bits(adev->wb.gpu_addr + wb_offset) & 0xFFFFFFFC);
 
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RPTR_WRITEBACK_ENABLE, 1);
 
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE), ring->gpu_addr >> 8);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_BASE_HI), ring->gpu_addr >> 40);
 
 		ring->wptr = 0;
 
 		/* before programing wptr to a less value, need set minor_ptr_update first */
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 1);
 
 		if (!amdgpu_sriov_vf(adev)) { /* only bare-metal use register write for wptr */
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr) << 2);
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2);
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr) << 2);
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr) << 2);
 		}
 
-		doorbell = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL));
-		doorbell_offset = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL_OFFSET));
+		doorbell = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL));
+		doorbell_offset = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET));
 
 		if (ring->use_doorbell) {
 			doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 1);
@@ -637,55 +622,53 @@ static int sdma_v4_0_gfx_resume(struct amdgpu_device *adev)
 		} else {
 			doorbell = REG_SET_FIELD(doorbell, SDMA0_GFX_DOORBELL, ENABLE, 0);
 		}
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL), doorbell);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset);
-		if (adev->flags & AMD_IS_APU)
-			nbio_v7_0_sdma_doorbell_range(adev, i, ring->use_doorbell, ring->doorbell_index);
-		else
-			nbio_v6_1_sdma_doorbell_range(adev, i, ring->use_doorbell, ring->doorbell_index);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL), doorbell);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_DOORBELL_OFFSET), doorbell_offset);
+		adev->nbio_funcs->sdma_doorbell_range(adev, i, ring->use_doorbell,
+						      ring->doorbell_index);
 
 		if (amdgpu_sriov_vf(adev))
 			sdma_v4_0_ring_set_wptr(ring);
 
 		/* set minor_ptr_update to 0 after wptr programed */
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_MINOR_PTR_UPDATE), 0);
 
 		/* set utc l1 enable flag always to 1 */
-		temp = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL));
+		temp = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_CNTL));
 		temp = REG_SET_FIELD(temp, SDMA0_CNTL, UTC_L1_ENABLE, 1);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_CNTL), temp);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_CNTL), temp);
 
 		if (!amdgpu_sriov_vf(adev)) {
 			/* unhalt engine */
-			temp = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_F32_CNTL));
+			temp = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL));
 			temp = REG_SET_FIELD(temp, SDMA0_F32_CNTL, HALT, 0);
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_F32_CNTL), temp);
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_F32_CNTL), temp);
 		}
 
 		/* setup the wptr shadow polling */
 		wptr_gpu_addr = adev->wb.gpu_addr + (ring->wptr_offs * 4);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO),
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_LO),
 		       lower_32_bits(wptr_gpu_addr));
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI),
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_ADDR_HI),
 		       upper_32_bits(wptr_gpu_addr));
-		wptr_poll_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL));
+		wptr_poll_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL));
 		if (amdgpu_sriov_vf(adev))
 			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, SDMA0_GFX_RB_WPTR_POLL_CNTL, F32_POLL_ENABLE, 1);
 		else
 			wptr_poll_cntl = REG_SET_FIELD(wptr_poll_cntl, SDMA0_GFX_RB_WPTR_POLL_CNTL, F32_POLL_ENABLE, 0);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), wptr_poll_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_WPTR_POLL_CNTL), wptr_poll_cntl);
 
 		/* enable DMA RB */
 		rb_cntl = REG_SET_FIELD(rb_cntl, SDMA0_GFX_RB_CNTL, RB_ENABLE, 1);
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_RB_CNTL), rb_cntl);
 
-		ib_cntl = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_IB_CNTL));
+		ib_cntl = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL));
 		ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_ENABLE, 1);
 #ifdef __BIG_ENDIAN
 		ib_cntl = REG_SET_FIELD(ib_cntl, SDMA0_GFX_IB_CNTL, IB_SWAP_ENABLE, 1);
 #endif
 		/* enable DMA IBs */
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_GFX_IB_CNTL), ib_cntl);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_GFX_IB_CNTL), ib_cntl);
 
 		ring->ready = true;
 
@@ -816,12 +799,12 @@ static int sdma_v4_0_load_microcode(struct amdgpu_device *adev)
 			(adev->sdma.instance[i].fw->data +
 				le32_to_cpu(hdr->header.ucode_array_offset_bytes));
 
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_UCODE_ADDR), 0);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_UCODE_ADDR), 0);
 
 		for (j = 0; j < fw_size; j++)
-			WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_UCODE_DATA), le32_to_cpup(fw_data++));
+			WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_UCODE_DATA), le32_to_cpup(fw_data++));
 
-		WREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_UCODE_ADDR), adev->sdma.instance[i].fw_version);
+		WREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_UCODE_ADDR), adev->sdma.instance[i].fw_version);
 	}
 
 	return 0;
@@ -886,7 +869,7 @@ static int sdma_v4_0_ring_test_ring(struct amdgpu_ring *ring)
 	u32 tmp;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
 		return r;
@@ -899,7 +882,7 @@ static int sdma_v4_0_ring_test_ring(struct amdgpu_ring *ring)
 	r = amdgpu_ring_alloc(ring, 5);
 	if (r) {
 		DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
-		amdgpu_wb_free(adev, index);
+		amdgpu_device_wb_free(adev, index);
 		return r;
 	}
 
@@ -925,7 +908,7 @@ static int sdma_v4_0_ring_test_ring(struct amdgpu_ring *ring)
 			  ring->idx, tmp);
 		r = -EINVAL;
 	}
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 
 	return r;
 }
@@ -948,7 +931,7 @@ static int sdma_v4_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	u32 tmp = 0;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
@@ -1000,7 +983,7 @@ err1:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err0:
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 
@@ -1156,10 +1139,11 @@ static void sdma_v4_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_SRBM_WRITE) |
 			  SDMA_PKT_SRBM_WRITE_HEADER_BYTE_EN(0xf));
@@ -1317,7 +1301,7 @@ static bool sdma_v4_0_is_idle(void *handle)
 	u32 i;
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
-		u32 tmp = RREG32(sdma_v4_0_get_reg_offset(i, mmSDMA0_STATUS_REG));
+		u32 tmp = RREG32(sdma_v4_0_get_reg_offset(adev, i, mmSDMA0_STATUS_REG));
 
 		if (!(tmp & SDMA0_STATUS_REG__IDLE_MASK))
 			return false;
@@ -1333,8 +1317,8 @@ static int sdma_v4_0_wait_for_idle(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	for (i = 0; i < adev->usec_timeout; i++) {
-		sdma0 = RREG32(sdma_v4_0_get_reg_offset(0, mmSDMA0_STATUS_REG));
-		sdma1 = RREG32(sdma_v4_0_get_reg_offset(1, mmSDMA0_STATUS_REG));
+		sdma0 = RREG32(sdma_v4_0_get_reg_offset(adev, 0, mmSDMA0_STATUS_REG));
+		sdma1 = RREG32(sdma_v4_0_get_reg_offset(adev, 1, mmSDMA0_STATUS_REG));
 
 		if (sdma0 & sdma1 & SDMA0_STATUS_REG__IDLE_MASK)
 			return 0;
@@ -1358,8 +1342,8 @@ static int sdma_v4_0_set_trap_irq_state(struct amdgpu_device *adev,
 	u32 sdma_cntl;
 
 	u32 reg_offset = (type == AMDGPU_SDMA_IRQ_TRAP0) ?
-		sdma_v4_0_get_reg_offset(0, mmSDMA0_CNTL) :
-		sdma_v4_0_get_reg_offset(1, mmSDMA0_CNTL);
+		sdma_v4_0_get_reg_offset(adev, 0, mmSDMA0_CNTL) :
+		sdma_v4_0_get_reg_offset(adev, 1, mmSDMA0_CNTL);
 
 	sdma_cntl = RREG32(reg_offset);
 	sdma_cntl = REG_SET_FIELD(sdma_cntl, SDMA0_CNTL, TRAP_ENABLE,

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

@@ -1390,65 +1390,65 @@ static void si_init_golden_registers(struct amdgpu_device *adev)
 {
 	switch (adev->asic_type) {
 	case CHIP_TAHITI:
-		amdgpu_program_register_sequence(adev,
-						 tahiti_golden_registers,
-						 ARRAY_SIZE(tahiti_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 tahiti_golden_rlc_registers,
-						 ARRAY_SIZE(tahiti_golden_rlc_registers));
-		amdgpu_program_register_sequence(adev,
-						 tahiti_mgcg_cgcg_init,
-						 ARRAY_SIZE(tahiti_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 tahiti_golden_registers2,
-						 ARRAY_SIZE(tahiti_golden_registers2));
+		amdgpu_device_program_register_sequence(adev,
+							tahiti_golden_registers,
+							ARRAY_SIZE(tahiti_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							tahiti_golden_rlc_registers,
+							ARRAY_SIZE(tahiti_golden_rlc_registers));
+		amdgpu_device_program_register_sequence(adev,
+							tahiti_mgcg_cgcg_init,
+							ARRAY_SIZE(tahiti_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							tahiti_golden_registers2,
+							ARRAY_SIZE(tahiti_golden_registers2));
 		break;
 	case CHIP_PITCAIRN:
-		amdgpu_program_register_sequence(adev,
-						 pitcairn_golden_registers,
-						 ARRAY_SIZE(pitcairn_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 pitcairn_golden_rlc_registers,
-						 ARRAY_SIZE(pitcairn_golden_rlc_registers));
-		amdgpu_program_register_sequence(adev,
-						 pitcairn_mgcg_cgcg_init,
-						 ARRAY_SIZE(pitcairn_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							pitcairn_golden_registers,
+							ARRAY_SIZE(pitcairn_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							pitcairn_golden_rlc_registers,
+							ARRAY_SIZE(pitcairn_golden_rlc_registers));
+		amdgpu_device_program_register_sequence(adev,
+							pitcairn_mgcg_cgcg_init,
+							ARRAY_SIZE(pitcairn_mgcg_cgcg_init));
 		break;
 	case CHIP_VERDE:
-		amdgpu_program_register_sequence(adev,
-						 verde_golden_registers,
-						 ARRAY_SIZE(verde_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 verde_golden_rlc_registers,
-						 ARRAY_SIZE(verde_golden_rlc_registers));
-		amdgpu_program_register_sequence(adev,
-						 verde_mgcg_cgcg_init,
-						 ARRAY_SIZE(verde_mgcg_cgcg_init));
-		amdgpu_program_register_sequence(adev,
-						 verde_pg_init,
-						 ARRAY_SIZE(verde_pg_init));
+		amdgpu_device_program_register_sequence(adev,
+							verde_golden_registers,
+							ARRAY_SIZE(verde_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							verde_golden_rlc_registers,
+							ARRAY_SIZE(verde_golden_rlc_registers));
+		amdgpu_device_program_register_sequence(adev,
+							verde_mgcg_cgcg_init,
+							ARRAY_SIZE(verde_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							verde_pg_init,
+							ARRAY_SIZE(verde_pg_init));
 		break;
 	case CHIP_OLAND:
-		amdgpu_program_register_sequence(adev,
-						 oland_golden_registers,
-						 ARRAY_SIZE(oland_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 oland_golden_rlc_registers,
-						 ARRAY_SIZE(oland_golden_rlc_registers));
-		amdgpu_program_register_sequence(adev,
-						 oland_mgcg_cgcg_init,
-						 ARRAY_SIZE(oland_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							oland_golden_registers,
+							ARRAY_SIZE(oland_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							oland_golden_rlc_registers,
+							ARRAY_SIZE(oland_golden_rlc_registers));
+		amdgpu_device_program_register_sequence(adev,
+							oland_mgcg_cgcg_init,
+							ARRAY_SIZE(oland_mgcg_cgcg_init));
 		break;
 	case CHIP_HAINAN:
-		amdgpu_program_register_sequence(adev,
-						 hainan_golden_registers,
-						 ARRAY_SIZE(hainan_golden_registers));
-		amdgpu_program_register_sequence(adev,
-						 hainan_golden_registers2,
-						 ARRAY_SIZE(hainan_golden_registers2));
-		amdgpu_program_register_sequence(adev,
-						 hainan_mgcg_cgcg_init,
-						 ARRAY_SIZE(hainan_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							hainan_golden_registers,
+							ARRAY_SIZE(hainan_golden_registers));
+		amdgpu_device_program_register_sequence(adev,
+							hainan_golden_registers2,
+							ARRAY_SIZE(hainan_golden_registers2));
+		amdgpu_device_program_register_sequence(adev,
+							hainan_mgcg_cgcg_init,
+							ARRAY_SIZE(hainan_mgcg_cgcg_init));
 		break;
 
 
@@ -1959,42 +1959,42 @@ int si_set_ip_blocks(struct amdgpu_device *adev)
 	case CHIP_VERDE:
 	case CHIP_TAHITI:
 	case CHIP_PITCAIRN:
-		amdgpu_ip_block_add(adev, &si_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 		else
-			amdgpu_ip_block_add(adev, &dce_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_dma_ip_block);
-		/* amdgpu_ip_block_add(adev, &uvd_v3_1_ip_block); */
-		/* amdgpu_ip_block_add(adev, &vce_v1_0_ip_block); */
+			amdgpu_device_ip_block_add(adev, &dce_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_dma_ip_block);
+		/* amdgpu_device_ip_block_add(adev, &uvd_v3_1_ip_block); */
+		/* amdgpu_device_ip_block_add(adev, &vce_v1_0_ip_block); */
 		break;
 	case CHIP_OLAND:
-		amdgpu_ip_block_add(adev, &si_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 		else
-			amdgpu_ip_block_add(adev, &dce_v6_4_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_dma_ip_block);
-		/* amdgpu_ip_block_add(adev, &uvd_v3_1_ip_block); */
-		/* amdgpu_ip_block_add(adev, &vce_v1_0_ip_block); */
+			amdgpu_device_ip_block_add(adev, &dce_v6_4_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_dma_ip_block);
+		/* amdgpu_device_ip_block_add(adev, &uvd_v3_1_ip_block); */
+		/* amdgpu_device_ip_block_add(adev, &vce_v1_0_ip_block); */
 		break;
 	case CHIP_HAINAN:
-		amdgpu_ip_block_add(adev, &si_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &si_dma_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &si_dma_ip_block);
 		break;
 	default:
 		BUG();

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

@@ -221,7 +221,7 @@ static int si_dma_ring_test_ring(struct amdgpu_ring *ring)
 	u32 tmp;
 	u64 gpu_addr;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to allocate wb slot\n", r);
 		return r;
@@ -234,7 +234,7 @@ static int si_dma_ring_test_ring(struct amdgpu_ring *ring)
 	r = amdgpu_ring_alloc(ring, 4);
 	if (r) {
 		DRM_ERROR("amdgpu: dma failed to lock ring %d (%d).\n", ring->idx, r);
-		amdgpu_wb_free(adev, index);
+		amdgpu_device_wb_free(adev, index);
 		return r;
 	}
 
@@ -258,7 +258,7 @@ static int si_dma_ring_test_ring(struct amdgpu_ring *ring)
 			  ring->idx, tmp);
 		r = -EINVAL;
 	}
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 
 	return r;
 }
@@ -281,7 +281,7 @@ static int si_dma_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	u64 gpu_addr;
 	long r;
 
-	r = amdgpu_wb_get(adev, &index);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
 		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
@@ -328,7 +328,7 @@ err1:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err0:
-	amdgpu_wb_free(adev, index);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 

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

@@ -34,7 +34,6 @@
 #include "atom.h"
 #include "amd_pcie.h"
 
-#include "soc15ip.h"
 #include "uvd/uvd_7_0_offset.h"
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
@@ -101,15 +100,8 @@ static u32 soc15_pcie_rreg(struct amdgpu_device *adev, u32 reg)
 {
 	unsigned long flags, address, data;
 	u32 r;
-	const struct nbio_pcie_index_data *nbio_pcie_id;
-
-	if (adev->flags & AMD_IS_APU)
-		nbio_pcie_id = &nbio_v7_0_pcie_index_data;
-	else
-		nbio_pcie_id = &nbio_v6_1_pcie_index_data;
-
-	address = nbio_pcie_id->index_offset;
-	data = nbio_pcie_id->data_offset;
+	address = adev->nbio_funcs->get_pcie_index_offset(adev);
+	data = adev->nbio_funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -122,15 +114,9 @@ static u32 soc15_pcie_rreg(struct amdgpu_device *adev, u32 reg)
 static void soc15_pcie_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 {
 	unsigned long flags, address, data;
-	const struct nbio_pcie_index_data *nbio_pcie_id;
 
-	if (adev->flags & AMD_IS_APU)
-		nbio_pcie_id = &nbio_v7_0_pcie_index_data;
-	else
-		nbio_pcie_id = &nbio_v6_1_pcie_index_data;
-
-	address = nbio_pcie_id->index_offset;
-	data = nbio_pcie_id->data_offset;
+	address = adev->nbio_funcs->get_pcie_index_offset(adev);
+	data = adev->nbio_funcs->get_pcie_data_offset(adev);
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	WREG32(address, reg);
@@ -242,41 +228,9 @@ static void soc15_se_cac_wreg(struct amdgpu_device *adev, u32 reg, u32 v)
 
 static u32 soc15_get_config_memsize(struct amdgpu_device *adev)
 {
-	if (adev->flags & AMD_IS_APU)
-		return nbio_v7_0_get_memsize(adev);
-	else
-		return nbio_v6_1_get_memsize(adev);
+	return adev->nbio_funcs->get_memsize(adev);
 }
 
-static const u32 vega10_golden_init[] =
-{
-};
-
-static const u32 raven_golden_init[] =
-{
-};
-
-static void soc15_init_golden_registers(struct amdgpu_device *adev)
-{
-	/* Some of the registers might be dependent on GRBM_GFX_INDEX */
-	mutex_lock(&adev->grbm_idx_mutex);
-
-	switch (adev->asic_type) {
-	case CHIP_VEGA10:
-		amdgpu_program_register_sequence(adev,
-						 vega10_golden_init,
-						 ARRAY_SIZE(vega10_golden_init));
-		break;
-	case CHIP_RAVEN:
-		amdgpu_program_register_sequence(adev,
-						 raven_golden_init,
-						 ARRAY_SIZE(raven_golden_init));
-		break;
-	default:
-		break;
-	}
-	mutex_unlock(&adev->grbm_idx_mutex);
-}
 static u32 soc15_get_xclk(struct amdgpu_device *adev)
 {
 	return adev->clock.spll.reference_freq;
@@ -332,25 +286,34 @@ static bool soc15_read_bios_from_rom(struct amdgpu_device *adev,
 	return true;
 }
 
-static struct amdgpu_allowed_register_entry soc15_allowed_read_registers[] = {
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS2)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE0)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE1)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE2)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGRBM_STATUS_SE3)},
-	{ SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_STATUS_REG)},
-	{ SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_STATUS_REG)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_STAT)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT1)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT2)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_STALLED_STAT3)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_CPF_BUSY_STAT)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_CPF_STALLED_STAT1)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_CPF_STATUS)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_CPC_STALLED_STAT1)},
-	{ SOC15_REG_OFFSET(GC, 0, mmCP_CPC_STATUS)},
-	{ SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG)},
+struct soc15_allowed_register_entry {
+	uint32_t hwip;
+	uint32_t inst;
+	uint32_t seg;
+	uint32_t reg_offset;
+	bool grbm_indexed;
+};
+
+
+static struct soc15_allowed_register_entry soc15_allowed_read_registers[] = {
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS2)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS_SE0)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS_SE1)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS_SE2)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGRBM_STATUS_SE3)},
+	{ SOC15_REG_ENTRY(SDMA0, 0, mmSDMA0_STATUS_REG)},
+	{ SOC15_REG_ENTRY(SDMA1, 0, mmSDMA1_STATUS_REG)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_STAT)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_STALLED_STAT1)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_STALLED_STAT2)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_STALLED_STAT3)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_BUSY_STAT)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STALLED_STAT1)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPF_STATUS)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STALLED_STAT1)},
+	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STATUS)},
+	{ SOC15_REG_ENTRY(GC, 0, mmGB_ADDR_CONFIG)},
 };
 
 static uint32_t soc15_read_indexed_register(struct amdgpu_device *adev, u32 se_num,
@@ -377,12 +340,9 @@ static uint32_t soc15_get_register_value(struct amdgpu_device *adev,
 	if (indexed) {
 		return soc15_read_indexed_register(adev, se_num, sh_num, reg_offset);
 	} else {
-		switch (reg_offset) {
-		case SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG):
+		if (reg_offset == SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG))
 			return adev->gfx.config.gb_addr_config;
-		default:
-			return RREG32(reg_offset);
-		}
+		return RREG32(reg_offset);
 	}
 }
 
@@ -390,10 +350,13 @@ static int soc15_read_register(struct amdgpu_device *adev, u32 se_num,
 			    u32 sh_num, u32 reg_offset, u32 *value)
 {
 	uint32_t i;
+	struct soc15_allowed_register_entry  *en;
 
 	*value = 0;
 	for (i = 0; i < ARRAY_SIZE(soc15_allowed_read_registers); i++) {
-		if (reg_offset != soc15_allowed_read_registers[i].reg_offset)
+		en = &soc15_allowed_read_registers[i];
+		if (reg_offset != (adev->reg_offset[en->hwip][en->inst][en->seg]
+					+ en->reg_offset))
 			continue;
 
 		*value = soc15_get_register_value(adev,
@@ -404,6 +367,43 @@ static int soc15_read_register(struct amdgpu_device *adev, u32 se_num,
 	return -EINVAL;
 }
 
+
+/**
+ * soc15_program_register_sequence - program an array of registers.
+ *
+ * @adev: amdgpu_device pointer
+ * @regs: pointer to the register array
+ * @array_size: size of the register array
+ *
+ * Programs an array or registers with and and or masks.
+ * This is a helper for setting golden registers.
+ */
+
+void soc15_program_register_sequence(struct amdgpu_device *adev,
+					     const struct soc15_reg_golden *regs,
+					     const u32 array_size)
+{
+	const struct soc15_reg_golden *entry;
+	u32 tmp, reg;
+	int i;
+
+	for (i = 0; i < array_size; ++i) {
+		entry = &regs[i];
+		reg =  adev->reg_offset[entry->hwip][entry->instance][entry->segment] + entry->reg;
+
+		if (entry->and_mask == 0xffffffff) {
+			tmp = entry->or_mask;
+		} else {
+			tmp = RREG32(reg);
+			tmp &= ~(entry->and_mask);
+			tmp |= entry->or_mask;
+		}
+		WREG32(reg, tmp);
+	}
+
+}
+
+
 static int soc15_asic_reset(struct amdgpu_device *adev)
 {
 	u32 i;
@@ -428,9 +428,8 @@ static int soc15_asic_reset(struct amdgpu_device *adev)
 
 	/* wait for asic to come out of reset */
 	for (i = 0; i < adev->usec_timeout; i++) {
-		u32 memsize = (adev->flags & AMD_IS_APU) ?
-			nbio_v7_0_get_memsize(adev) :
-			nbio_v6_1_get_memsize(adev);
+		u32 memsize = adev->nbio_funcs->get_memsize(adev);
+
 		if (memsize != 0xffffffff)
 			break;
 		udelay(1);
@@ -495,14 +494,10 @@ static void soc15_program_aspm(struct amdgpu_device *adev)
 }
 
 static void soc15_enable_doorbell_aperture(struct amdgpu_device *adev,
-					bool enable)
+					   bool enable)
 {
-	if (adev->flags & AMD_IS_APU) {
-		nbio_v7_0_enable_doorbell_aperture(adev, enable);
-	} else {
-		nbio_v6_1_enable_doorbell_aperture(adev, enable);
-		nbio_v6_1_enable_doorbell_selfring_aperture(adev, enable);
-	}
+	adev->nbio_funcs->enable_doorbell_aperture(adev, enable);
+	adev->nbio_funcs->enable_doorbell_selfring_aperture(adev, enable);
 }
 
 static const struct amdgpu_ip_block_version vega10_common_ip_block =
@@ -516,50 +511,65 @@ static const struct amdgpu_ip_block_version vega10_common_ip_block =
 
 int soc15_set_ip_blocks(struct amdgpu_device *adev)
 {
-	nbio_v6_1_detect_hw_virt(adev);
+	/* Set IP register base before any HW register access */
+	switch (adev->asic_type) {
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		vega10_reg_base_init(adev);
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (adev->flags & AMD_IS_APU)
+		adev->nbio_funcs = &nbio_v7_0_funcs;
+	else
+		adev->nbio_funcs = &nbio_v6_1_funcs;
+
+	adev->nbio_funcs->detect_hw_virt(adev);
 
 	if (amdgpu_sriov_vf(adev))
 		adev->virt.ops = &xgpu_ai_virt_ops;
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
-		amdgpu_ip_block_add(adev, &vega10_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v9_0_ip_block);
-		amdgpu_ip_block_add(adev, &vega10_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
 		if (amdgpu_fw_load_type == 2 || amdgpu_fw_load_type == -1)
-			amdgpu_ip_block_add(adev, &psp_v3_1_ip_block);
+			amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block);
 		if (!amdgpu_sriov_vf(adev))
-			amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+			amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #else
 #	warning "Enable CONFIG_DRM_AMD_DC for display support on SOC15."
 #endif
-		amdgpu_ip_block_add(adev, &gfx_v9_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v4_0_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v7_0_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v4_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v4_0_ip_block);
 		break;
 	case CHIP_RAVEN:
-		amdgpu_ip_block_add(adev, &vega10_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v9_0_ip_block);
-		amdgpu_ip_block_add(adev, &vega10_ih_ip_block);
-		amdgpu_ip_block_add(adev, &psp_v10_0_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &psp_v10_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #else
 #	warning "Enable CONFIG_DRM_AMD_DC for display support on SOC15."
 #endif
-		amdgpu_ip_block_add(adev, &gfx_v9_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v4_0_ip_block);
-		amdgpu_ip_block_add(adev, &vcn_v1_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v9_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v4_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vcn_v1_0_ip_block);
 		break;
 	default:
 		return -EINVAL;
@@ -570,10 +580,7 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 
 static uint32_t soc15_get_rev_id(struct amdgpu_device *adev)
 {
-	if (adev->flags & AMD_IS_APU)
-		return nbio_v7_0_get_rev_id(adev);
-	else
-		return nbio_v6_1_get_rev_id(adev);
+	return adev->nbio_funcs->get_rev_id(adev);
 }
 
 static const struct amdgpu_asic_funcs soc15_asic_funcs =
@@ -609,8 +616,8 @@ static int soc15_common_early_init(void *handle)
 
 	adev->asic_funcs = &soc15_asic_funcs;
 
-	if (amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP) &&
-		(amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_PSP)))
+	if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_PSP) &&
+	    (amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_PSP)))
 		psp_enabled = true;
 
 	adev->rev_id = soc15_get_rev_id(adev);
@@ -675,7 +682,7 @@ static int soc15_common_early_init(void *handle)
 
 	adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type);
 
-	amdgpu_get_pcie_info(adev);
+	amdgpu_device_get_pcie_info(adev);
 
 	return 0;
 }
@@ -709,15 +716,12 @@ static int soc15_common_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	/* move the golden regs per IP block */
-	soc15_init_golden_registers(adev);
 	/* enable pcie gen2/3 link */
 	soc15_pcie_gen3_enable(adev);
 	/* enable aspm */
 	soc15_program_aspm(adev);
 	/* setup nbio registers */
-	if (!(adev->flags & AMD_IS_APU))
-		nbio_v6_1_init_registers(adev);
+	adev->nbio_funcs->init_registers(adev);
 	/* enable the doorbell aperture */
 	soc15_enable_doorbell_aperture(adev, true);
 
@@ -878,9 +882,9 @@ static int soc15_common_set_clockgating_state(void *handle,
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
-		nbio_v6_1_update_medium_grain_clock_gating(adev,
+		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		nbio_v6_1_update_medium_grain_light_sleep(adev,
+		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_hdp_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
@@ -894,9 +898,9 @@ static int soc15_common_set_clockgating_state(void *handle,
 				state == AMD_CG_STATE_GATE ? true : false);
 		break;
 	case CHIP_RAVEN:
-		nbio_v7_0_update_medium_grain_clock_gating(adev,
+		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		nbio_v6_1_update_medium_grain_light_sleep(adev,
+		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_hdp_light_sleep(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
@@ -921,7 +925,7 @@ static void soc15_common_get_clockgating_state(void *handle, u32 *flags)
 	if (amdgpu_sriov_vf(adev))
 		*flags = 0;
 
-	nbio_v6_1_get_clockgating_state(adev, flags);
+	adev->nbio_funcs->get_clockgating_state(adev, flags);
 
 	/* AMD_CG_SUPPORT_HDP_LS */
 	data = RREG32(SOC15_REG_OFFSET(HDP, 0, mmHDP_MEM_POWER_LS));

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

@@ -29,8 +29,28 @@
 
 extern const struct amd_ip_funcs soc15_common_ip_funcs;
 
+struct soc15_reg_golden {
+	u32	hwip;
+	u32	instance;
+	u32	segment;
+	u32	reg;
+	u32	and_mask;
+	u32	or_mask;
+};
+
+#define SOC15_REG_ENTRY(ip, inst, reg)	ip##_HWIP, inst, reg##_BASE_IDX, reg
+
+#define SOC15_REG_GOLDEN_VALUE(ip, inst, reg, and_mask, or_mask) \
+	{ ip##_HWIP, inst, reg##_BASE_IDX, reg, and_mask, or_mask }
+
 void soc15_grbm_select(struct amdgpu_device *adev,
 		    u32 me, u32 pipe, u32 queue, u32 vmid);
 int soc15_set_ip_blocks(struct amdgpu_device *adev);
 
+void soc15_program_register_sequence(struct amdgpu_device *adev,
+					     const struct soc15_reg_golden *registers,
+					     const u32 array_size);
+
+int vega10_reg_base_init(struct amdgpu_device *adev);
+
 #endif

drivers/gpu/drm/amd/amdgpu/soc15_common.h

@@ -24,72 +24,28 @@
 #ifndef __SOC15_COMMON_H__
 #define __SOC15_COMMON_H__
 
-struct nbio_hdp_flush_reg {
-	u32 hdp_flush_req_offset;
-	u32 hdp_flush_done_offset;
-	u32 ref_and_mask_cp0;
-	u32 ref_and_mask_cp1;
-	u32 ref_and_mask_cp2;
-	u32 ref_and_mask_cp3;
-	u32 ref_and_mask_cp4;
-	u32 ref_and_mask_cp5;
-	u32 ref_and_mask_cp6;
-	u32 ref_and_mask_cp7;
-	u32 ref_and_mask_cp8;
-	u32 ref_and_mask_cp9;
-	u32 ref_and_mask_sdma0;
-	u32 ref_and_mask_sdma1;
-};
-
-struct nbio_pcie_index_data {
-	u32 index_offset;
-	u32 data_offset;
-};
-
 /* Register Access Macros */
-#define SOC15_REG_OFFSET(ip, inst, reg)       (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-                                                (1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-                                                    (2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-                                                        (3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-                                                            (ip##_BASE__INST##inst##_SEG4 + reg)))))
+#define SOC15_REG_OFFSET(ip, inst, reg)	(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg)
 
 #define WREG32_FIELD15(ip, idx, reg, field, val)	\
-	WREG32(SOC15_REG_OFFSET(ip, idx, mm##reg), (RREG32(SOC15_REG_OFFSET(ip, idx, mm##reg)) & ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field))
+	WREG32(adev->reg_offset[ip##_HWIP][idx][mm##reg##_BASE_IDX] + mm##reg,	\
+	(RREG32(adev->reg_offset[ip##_HWIP][idx][mm##reg##_BASE_IDX] + mm##reg)	\
+	& ~REG_FIELD_MASK(reg, field)) | (val) << REG_FIELD_SHIFT(reg, field))
 
 #define RREG32_SOC15(ip, inst, reg) \
-	RREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-		(1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-		(2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-		(3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-		(ip##_BASE__INST##inst##_SEG4 + reg))))))
+	RREG32(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg)
 
 #define RREG32_SOC15_OFFSET(ip, inst, reg, offset) \
-	RREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-		(1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-		(2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-		(3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-		(ip##_BASE__INST##inst##_SEG4 + reg))))) + offset)
+	RREG32((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) + offset)
 
 #define WREG32_SOC15(ip, inst, reg, value) \
-	WREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-		(1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-		(2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-		(3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-		(ip##_BASE__INST##inst##_SEG4 + reg))))), value)
+	WREG32((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg), value)
 
 #define WREG32_SOC15_NO_KIQ(ip, inst, reg, value) \
-	WREG32_NO_KIQ( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-		(1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-		(2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-		(3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-		(ip##_BASE__INST##inst##_SEG4 + reg))))), value)
+	WREG32_NO_KIQ((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg), value)
 
 #define WREG32_SOC15_OFFSET(ip, inst, reg, offset, value) \
-	WREG32( (0 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG0 + reg : \
-		(1 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG1 + reg : \
-		(2 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG2 + reg : \
-		(3 == reg##_BASE_IDX ? ip##_BASE__INST##inst##_SEG3 + reg : \
-		(ip##_BASE__INST##inst##_SEG4 + reg))))) + offset, value)
+	WREG32((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) + offset, value)
 
 #endif
 

drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c

@@ -412,10 +412,10 @@ static int uvd_v6_0_sw_init(void *handle)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
-		struct amd_sched_rq *rq;
+		struct drm_sched_rq *rq;
 		ring = &adev->uvd.ring_enc[0];
-		rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-		r = amd_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
+		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+		r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
 					  rq, amdgpu_sched_jobs, NULL);
 		if (r) {
 			DRM_ERROR("Failed setting up UVD ENC run queue.\n");
@@ -456,7 +456,7 @@ static int uvd_v6_0_sw_fini(void *handle)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
-		amd_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
+		drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
 
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
 			amdgpu_ring_fini(&adev->uvd.ring_enc[i]);

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

@@ -29,7 +29,6 @@
 #include "soc15_common.h"
 #include "mmsch_v1_0.h"
 
-#include "soc15ip.h"
 #include "uvd/uvd_7_0_offset.h"
 #include "uvd/uvd_7_0_sh_mask.h"
 #include "vce/vce_4_0_offset.h"
@@ -385,7 +384,7 @@ static int uvd_v7_0_early_init(void *handle)
 static int uvd_v7_0_sw_init(void *handle)
 {
 	struct amdgpu_ring *ring;
-	struct amd_sched_rq *rq;
+	struct drm_sched_rq *rq;
 	int i, r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
@@ -416,8 +415,8 @@ static int uvd_v7_0_sw_init(void *handle)
 	}
 
 	ring = &adev->uvd.ring_enc[0];
-	rq = &ring->sched.sched_rq[AMD_SCHED_PRIORITY_NORMAL];
-	r = amd_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
+	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
 				  rq, amdgpu_sched_jobs, NULL);
 	if (r) {
 		DRM_ERROR("Failed setting up UVD ENC run queue.\n");
@@ -472,7 +471,7 @@ static int uvd_v7_0_sw_fini(void *handle)
 	if (r)
 		return r;
 
-	amd_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
+	drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
 		amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
@@ -1086,6 +1085,8 @@ static void uvd_v7_0_stop(struct amdgpu_device *adev)
 static void uvd_v7_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
 				     unsigned flags)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 
 	amdgpu_ring_write(ring,
@@ -1123,6 +1124,7 @@ static void uvd_v7_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq
 static void uvd_v7_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 			u64 seq, unsigned flags)
 {
+
 	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 
 	amdgpu_ring_write(ring, HEVC_ENC_CMD_FENCE);
@@ -1141,6 +1143,8 @@ static void uvd_v7_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
  */
 static void uvd_v7_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(NBIF, 0,
 		mmHDP_MEM_COHERENCY_FLUSH_CNTL), 0));
 	amdgpu_ring_write(ring, 0);
@@ -1155,6 +1159,8 @@ static void uvd_v7_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
  */
 static void uvd_v7_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 0));
 	amdgpu_ring_write(ring, 1);
 }
@@ -1214,6 +1220,8 @@ static void uvd_v7_0_ring_emit_ib(struct amdgpu_ring *ring,
 				  struct amdgpu_ib *ib,
 				  unsigned vm_id, bool ctx_switch)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_VMID), 0));
 	amdgpu_ring_write(ring, vm_id);
@@ -1250,6 +1258,8 @@ static void uvd_v7_0_enc_ring_emit_ib(struct amdgpu_ring *ring,
 static void uvd_v7_0_vm_reg_write(struct amdgpu_ring *ring,
 				uint32_t data0, uint32_t data1)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, data0);
@@ -1264,6 +1274,8 @@ static void uvd_v7_0_vm_reg_write(struct amdgpu_ring *ring,
 static void uvd_v7_0_vm_reg_wait(struct amdgpu_ring *ring,
 				uint32_t data0, uint32_t data1, uint32_t mask)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, data0);
@@ -1283,11 +1295,12 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
-	uint32_t data0, data1, mask;
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
+	uint32_t data0, data1, mask;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2;
 	data1 = upper_32_bits(pd_addr);
@@ -1314,6 +1327,16 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 	uvd_v7_0_vm_reg_wait(ring, data0, data1, mask);
 }
 
+static void uvd_v7_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+	struct amdgpu_device *adev = ring->adev;
+
+	for (i = 0; i < count; i++)
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+
+}
+
 static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring)
 {
 	amdgpu_ring_write(ring, HEVC_ENC_CMD_END);
@@ -1324,10 +1347,11 @@ static void uvd_v7_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	amdgpu_ring_write(ring, HEVC_ENC_CMD_REG_WRITE);
 	amdgpu_ring_write(ring,	(hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2);
@@ -1681,7 +1705,7 @@ const struct amd_ip_funcs uvd_v7_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0),
+	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = uvd_v7_0_ring_get_rptr,
@@ -1700,7 +1724,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.emit_hdp_invalidate = uvd_v7_0_ring_emit_hdp_invalidate,
 	.test_ring = uvd_v7_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v7_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,

drivers/gpu/drm/amd/amdgpu/vce_v4_0.c

@@ -32,7 +32,6 @@
 #include "soc15_common.h"
 #include "mmsch_v1_0.h"
 
-#include "soc15ip.h"
 #include "vce/vce_4_0_offset.h"
 #include "vce/vce_4_0_default.h"
 #include "vce/vce_4_0_sh_mask.h"
@@ -424,7 +423,7 @@ static int vce_v4_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	size  = (VCE_V4_0_STACK_SIZE + VCE_V4_0_DATA_SIZE) * 2;
+	size  = VCE_V4_0_STACK_SIZE + VCE_V4_0_DATA_SIZE;
 	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
 		size += VCE_V4_0_FW_SIZE;
 
@@ -970,10 +969,11 @@ static void vce_v4_0_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	amdgpu_ring_write(ring, VCE_CMD_REG_WRITE);
 	amdgpu_ring_write(ring,	(hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2);

drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c

@@ -28,7 +28,6 @@
 #include "soc15d.h"
 #include "soc15_common.h"
 
-#include "soc15ip.h"
 #include "vcn/vcn_1_0_offset.h"
 #include "vcn/vcn_1_0_sh_mask.h"
 #include "hdp/hdp_4_0_offset.h"
@@ -744,6 +743,8 @@ static void vcn_v1_0_dec_ring_set_wptr(struct amdgpu_ring *ring)
  */
 static void vcn_v1_0_dec_ring_insert_start(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, 0);
@@ -761,6 +762,8 @@ static void vcn_v1_0_dec_ring_insert_start(struct amdgpu_ring *ring)
  */
 static void vcn_v1_0_dec_ring_insert_end(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, VCN_DEC_CMD_PACKET_END << 1);
@@ -777,6 +780,8 @@ static void vcn_v1_0_dec_ring_insert_end(struct amdgpu_ring *ring)
 static void vcn_v1_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
 				     unsigned flags)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 
 	amdgpu_ring_write(ring,
@@ -812,6 +817,8 @@ static void vcn_v1_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64
  */
 static void vcn_v1_0_dec_ring_emit_hdp_invalidate(struct amdgpu_ring *ring)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 0));
 	amdgpu_ring_write(ring, 1);
 }
@@ -828,6 +835,8 @@ static void vcn_v1_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
 				  struct amdgpu_ib *ib,
 				  unsigned vm_id, bool ctx_switch)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_VMID), 0));
 	amdgpu_ring_write(ring, vm_id);
@@ -846,6 +855,8 @@ static void vcn_v1_0_dec_ring_emit_ib(struct amdgpu_ring *ring,
 static void vcn_v1_0_dec_vm_reg_write(struct amdgpu_ring *ring,
 				uint32_t data0, uint32_t data1)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, data0);
@@ -860,6 +871,8 @@ static void vcn_v1_0_dec_vm_reg_write(struct amdgpu_ring *ring,
 static void vcn_v1_0_dec_vm_reg_wait(struct amdgpu_ring *ring,
 				uint32_t data0, uint32_t data1, uint32_t mask)
 {
+	struct amdgpu_device *adev = ring->adev;
+
 	amdgpu_ring_write(ring,
 		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, data0);
@@ -879,11 +892,12 @@ static void vcn_v1_0_dec_ring_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
-	uint32_t data0, data1, mask;
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
+	uint32_t data0, data1, mask;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	data0 = (hub->ctx0_ptb_addr_hi32 + vm_id * 2) << 2;
 	data1 = upper_32_bits(pd_addr);
@@ -1011,10 +1025,11 @@ static void vcn_v1_0_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
 	uint32_t req = ring->adev->gart.gart_funcs->get_invalidate_req(vm_id);
+	uint64_t flags = AMDGPU_PTE_VALID;
 	unsigned eng = ring->vm_inv_eng;
 
-	pd_addr = amdgpu_gart_get_vm_pde(ring->adev, pd_addr);
-	pd_addr |= AMDGPU_PTE_VALID;
+	amdgpu_gart_get_vm_pde(ring->adev, -1, &pd_addr, &flags);
+	pd_addr |= flags;
 
 	amdgpu_ring_write(ring, VCN_ENC_CMD_REG_WRITE);
 	amdgpu_ring_write(ring,
@@ -1077,6 +1092,17 @@ static int vcn_v1_0_process_interrupt(struct amdgpu_device *adev,
 	return 0;
 }
 
+static void vcn_v1_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+	struct amdgpu_device *adev = ring->adev;
+
+	for (i = 0; i < count; i++)
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+
+}
+
+
 static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.name = "vcn_v1_0",
 	.early_init = vcn_v1_0_early_init,
@@ -1100,7 +1126,7 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_VCN_DEC,
 	.align_mask = 0xf,
-	.nop = PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0),
+	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = vcn_v1_0_dec_ring_get_rptr,
@@ -1118,7 +1144,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.emit_hdp_invalidate = vcn_v1_0_dec_ring_emit_hdp_invalidate,
 	.test_ring = amdgpu_vcn_dec_ring_test_ring,
 	.test_ib = amdgpu_vcn_dec_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = vcn_v1_0_ring_insert_nop,
 	.insert_start = vcn_v1_0_dec_ring_insert_start,
 	.insert_end = vcn_v1_0_dec_ring_insert_end,
 	.pad_ib = amdgpu_ring_generic_pad_ib,

drivers/gpu/drm/amd/amdgpu/vega10_ih.c

@@ -25,8 +25,6 @@
 #include "amdgpu_ih.h"
 #include "soc15.h"
 
-
-#include "soc15ip.h"
 #include "oss/osssys_4_0_offset.h"
 #include "oss/osssys_4_0_sh_mask.h"
 
@@ -97,10 +95,7 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev)
 	/* disable irqs */
 	vega10_ih_disable_interrupts(adev);
 
-	if (adev->flags & AMD_IS_APU)
-		nbio_v7_0_ih_control(adev);
-	else
-		nbio_v6_1_ih_control(adev);
+	adev->nbio_funcs->ih_control(adev);
 
 	ih_rb_cntl = RREG32_SOC15(OSSSYS, 0, mmIH_RB_CNTL);
 	/* Ring Buffer base. [39:8] of 40-bit address of the beginning of the ring buffer*/
@@ -151,10 +146,8 @@ static int vega10_ih_irq_init(struct amdgpu_device *adev)
 						 ENABLE, 0);
 	}
 	WREG32_SOC15(OSSSYS, 0, mmIH_DOORBELL_RPTR, ih_doorbell_rtpr);
-	if (adev->flags & AMD_IS_APU)
-		nbio_v7_0_ih_doorbell_range(adev, adev->irq.ih.use_doorbell, adev->irq.ih.doorbell_index);
-	else
-		nbio_v6_1_ih_doorbell_range(adev, adev->irq.ih.use_doorbell, adev->irq.ih.doorbell_index);
+	adev->nbio_funcs->ih_doorbell_range(adev, adev->irq.ih.use_doorbell,
+					    adev->irq.ih.doorbell_index);
 
 	tmp = RREG32_SOC15(OSSSYS, 0, mmIH_STORM_CLIENT_LIST_CNTL);
 	tmp = REG_SET_FIELD(tmp, IH_STORM_CLIENT_LIST_CNTL,

drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c

@@ -0,0 +1,56 @@
+/*
+ * Copyright 2017 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.
+ *
+ */
+#include "amdgpu.h"
+#include "soc15.h"
+
+#include "soc15_common.h"
+#include "soc15ip.h"
+
+int vega10_reg_base_init(struct amdgpu_device *adev)
+{
+	/* HW has more IP blocks,  only initialized the blocke beend by our driver  */
+	uint32_t i;
+	for (i = 0 ; i < MAX_INSTANCE ; ++i) {
+		adev->reg_offset[GC_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i]));
+		adev->reg_offset[HDP_HWIP][i] = (uint32_t *)(&(HDP_BASE.instance[i]));
+		adev->reg_offset[MMHUB_HWIP][i] = (uint32_t *)(&(MMHUB_BASE.instance[i]));
+		adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i]));
+		adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i]));
+		adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i]));
+		adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i]));
+		adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i]));
+		adev->reg_offset[VCN_HWIP][i] = (uint32_t *)(&(VCN_BASE.instance[i]));
+		adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i]));
+		adev->reg_offset[DCE_HWIP][i] = (uint32_t *)(&(DCE_BASE.instance[i]));
+		adev->reg_offset[OSSSYS_HWIP][i] = (uint32_t *)(&(OSSSYS_BASE.instance[i]));
+		adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(SDMA0_BASE.instance[i]));
+		adev->reg_offset[SDMA1_HWIP][i] = (uint32_t *)(&(SDMA1_BASE.instance[i]));
+		adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i]));
+		adev->reg_offset[PWR_HWIP][i] = (uint32_t *)(&(PWR_BASE.instance[i]));
+		adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIF_BASE.instance[i]));
+
+	}
+	return 0;
+}
+
+

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

@@ -282,29 +282,29 @@ static void vi_init_golden_registers(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_TOPAZ:
-		amdgpu_program_register_sequence(adev,
-						 iceland_mgcg_cgcg_init,
-						 ARRAY_SIZE(iceland_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							iceland_mgcg_cgcg_init,
+							ARRAY_SIZE(iceland_mgcg_cgcg_init));
 		break;
 	case CHIP_FIJI:
-		amdgpu_program_register_sequence(adev,
-						 fiji_mgcg_cgcg_init,
-						 ARRAY_SIZE(fiji_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							fiji_mgcg_cgcg_init,
+							ARRAY_SIZE(fiji_mgcg_cgcg_init));
 		break;
 	case CHIP_TONGA:
-		amdgpu_program_register_sequence(adev,
-						 tonga_mgcg_cgcg_init,
-						 ARRAY_SIZE(tonga_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							tonga_mgcg_cgcg_init,
+							ARRAY_SIZE(tonga_mgcg_cgcg_init));
 		break;
 	case CHIP_CARRIZO:
-		amdgpu_program_register_sequence(adev,
-						 cz_mgcg_cgcg_init,
-						 ARRAY_SIZE(cz_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							cz_mgcg_cgcg_init,
+							ARRAY_SIZE(cz_mgcg_cgcg_init));
 		break;
 	case CHIP_STONEY:
-		amdgpu_program_register_sequence(adev,
-						 stoney_mgcg_cgcg_init,
-						 ARRAY_SIZE(stoney_mgcg_cgcg_init));
+		amdgpu_device_program_register_sequence(adev,
+							stoney_mgcg_cgcg_init,
+							ARRAY_SIZE(stoney_mgcg_cgcg_init));
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
@@ -449,14 +449,18 @@ static bool vi_read_bios_from_rom(struct amdgpu_device *adev,
 
 static void vi_detect_hw_virtualization(struct amdgpu_device *adev)
 {
-	uint32_t reg = RREG32(mmBIF_IOV_FUNC_IDENTIFIER);
-	/* bit0: 0 means pf and 1 means vf */
-	/* bit31: 0 means disable IOV and 1 means enable */
-	if (reg & 1)
-		adev->virt.caps |= AMDGPU_SRIOV_CAPS_IS_VF;
-
-	if (reg & 0x80000000)
-		adev->virt.caps |= AMDGPU_SRIOV_CAPS_ENABLE_IOV;
+	uint32_t reg = 0;
+
+	if (adev->asic_type == CHIP_TONGA ||
+	    adev->asic_type == CHIP_FIJI) {
+	       reg = RREG32(mmBIF_IOV_FUNC_IDENTIFIER);
+	       /* bit0: 0 means pf and 1 means vf */
+	       if (REG_GET_FIELD(reg, BIF_IOV_FUNC_IDENTIFIER, FUNC_IDENTIFIER))
+		       adev->virt.caps |= AMDGPU_SRIOV_CAPS_IS_VF;
+	       /* bit31: 0 means disable IOV and 1 means enable */
+	       if (REG_GET_FIELD(reg, BIF_IOV_FUNC_IDENTIFIER, IOV_ENABLE))
+		       adev->virt.caps |= AMDGPU_SRIOV_CAPS_ENABLE_IOV;
+	}
 
 	if (reg == 0) {
 		if (is_virtual_machine()) /* passthrough mode exclus sr-iov mode */
@@ -667,7 +671,7 @@ static int vi_gpu_pci_config_reset(struct amdgpu_device *adev)
 	/* disable BM */
 	pci_clear_master(adev->pdev);
 	/* reset */
-	amdgpu_pci_config_reset(adev);
+	amdgpu_device_pci_config_reset(adev);
 
 	udelay(100);
 
@@ -891,8 +895,8 @@ static int vi_common_early_init(void *handle)
 
 	adev->asic_funcs = &vi_asic_funcs;
 
-	if (amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_SMC) &&
-		(amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_SMC)))
+	if (amdgpu_device_ip_get_ip_block(adev, AMD_IP_BLOCK_TYPE_SMC) &&
+	    (amdgpu_ip_block_mask & (1 << AMD_IP_BLOCK_TYPE_SMC)))
 		smc_enabled = true;
 
 	adev->rev_id = vi_get_rev_id(adev);
@@ -1074,7 +1078,7 @@ static int vi_common_early_init(void *handle)
 	/* vi use smc load by default */
 	adev->firmware.load_type = amdgpu_ucode_get_load_type(adev, amdgpu_fw_load_type);
 
-	amdgpu_get_pcie_info(adev);
+	amdgpu_device_get_pcie_info(adev);
 
 	return 0;
 }
@@ -1487,115 +1491,115 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 	switch (adev->asic_type) {
 	case CHIP_TOPAZ:
 		/* topaz has no DCE, UVD, VCE */
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v7_4_ip_block);
-		amdgpu_ip_block_add(adev, &iceland_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v7_4_ip_block);
+		amdgpu_device_ip_block_add(adev, &iceland_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v2_4_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v2_4_ip_block);
 		break;
 	case CHIP_FIJI:
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v8_5_ip_block);
-		amdgpu_ip_block_add(adev, &tonga_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v8_5_ip_block);
+		amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v10_1_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v3_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v10_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v3_0_ip_block);
 		if (!amdgpu_sriov_vf(adev)) {
-			amdgpu_ip_block_add(adev, &uvd_v6_0_ip_block);
-			amdgpu_ip_block_add(adev, &vce_v3_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &uvd_v6_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &vce_v3_0_ip_block);
 		}
 		break;
 	case CHIP_TONGA:
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &tonga_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v10_0_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v3_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v10_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v3_0_ip_block);
 		if (!amdgpu_sriov_vf(adev)) {
-			amdgpu_ip_block_add(adev, &uvd_v5_0_ip_block);
-			amdgpu_ip_block_add(adev, &vce_v3_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &uvd_v5_0_ip_block);
+			amdgpu_device_ip_block_add(adev, &vce_v3_0_ip_block);
 		}
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS12:
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v8_1_ip_block);
-		amdgpu_ip_block_add(adev, &tonga_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v8_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v11_2_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v3_1_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v6_3_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v3_4_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v11_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v3_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v6_3_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v3_4_ip_block);
 		break;
 	case CHIP_CARRIZO:
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &cz_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cz_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v11_0_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v3_0_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v6_0_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v3_1_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v11_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v3_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v6_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v3_1_ip_block);
 #if defined(CONFIG_DRM_AMD_ACP)
-		amdgpu_ip_block_add(adev, &acp_ip_block);
+		amdgpu_device_ip_block_add(adev, &acp_ip_block);
 #endif
 		break;
 	case CHIP_STONEY:
-		amdgpu_ip_block_add(adev, &vi_common_ip_block);
-		amdgpu_ip_block_add(adev, &gmc_v8_0_ip_block);
-		amdgpu_ip_block_add(adev, &cz_ih_ip_block);
-		amdgpu_ip_block_add(adev, &amdgpu_pp_ip_block);
+		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
+		amdgpu_device_ip_block_add(adev, &gmc_v8_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &cz_ih_ip_block);
+		amdgpu_device_ip_block_add(adev, &amdgpu_pp_ip_block);
 		if (adev->enable_virtual_display)
-			amdgpu_ip_block_add(adev, &dce_virtual_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
 		else if (amdgpu_device_has_dc_support(adev))
-			amdgpu_ip_block_add(adev, &dm_ip_block);
+			amdgpu_device_ip_block_add(adev, &dm_ip_block);
 #endif
 		else
-			amdgpu_ip_block_add(adev, &dce_v11_0_ip_block);
-		amdgpu_ip_block_add(adev, &gfx_v8_1_ip_block);
-		amdgpu_ip_block_add(adev, &sdma_v3_0_ip_block);
-		amdgpu_ip_block_add(adev, &uvd_v6_2_ip_block);
-		amdgpu_ip_block_add(adev, &vce_v3_4_ip_block);
+			amdgpu_device_ip_block_add(adev, &dce_v11_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &gfx_v8_1_ip_block);
+		amdgpu_device_ip_block_add(adev, &sdma_v3_0_ip_block);
+		amdgpu_device_ip_block_add(adev, &uvd_v6_2_ip_block);
+		amdgpu_device_ip_block_add(adev, &vce_v3_4_ip_block);
 #if defined(CONFIG_DRM_AMD_ACP)
-		amdgpu_ip_block_add(adev, &acp_ip_block);
+		amdgpu_device_ip_block_add(adev, &acp_ip_block);
 #endif
 		break;
 	default:

drivers/gpu/drm/amd/display/TODO

@@ -105,3 +105,6 @@ useless with filtering output. dynamic debug printing might be an option.
 20. Use kernel i2c device to program HDMI retimer. Some boards have an HDMI
 retimer that we need to program to pass PHY compliance. Currently that's
 bypassing the i2c device and goes directly to HW. This should be changed.
+
+21. Remove vector.c from dc/basics. It's used in DDC code which can probably
+be simplified enough to no longer need a vector implementation.

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

@@ -431,9 +431,9 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 	adev->dm.dc = dc_create(&init_data);
 
 	if (adev->dm.dc) {
-		DRM_INFO("Display Core initialized!\n");
+		DRM_INFO("Display Core initialized with v%s!\n", DC_VER);
 	} else {
-		DRM_INFO("Display Core failed to initialize!\n");
+		DRM_INFO("Display Core failed to initialize with v%s!\n", DC_VER);
 		goto error;
 	}
 
@@ -2351,7 +2351,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 		       const struct dm_connector_state *dm_state)
 {
 	struct drm_display_mode *preferred_mode = NULL;
-	const struct drm_connector *drm_connector;
+	struct drm_connector *drm_connector;
 	struct dc_stream_state *stream = NULL;
 	struct drm_display_mode mode = *drm_mode;
 	bool native_mode_found = false;
@@ -2370,11 +2370,13 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 
 	if (!aconnector->dc_sink) {
 		/*
-		 * Exclude MST from creating fake_sink
-		 * TODO: need to enable MST into fake_sink feature
+		 * Create dc_sink when necessary to MST
+		 * Don't apply fake_sink to MST
 		 */
-		if (aconnector->mst_port)
-			goto stream_create_fail;
+		if (aconnector->mst_port) {
+			dm_dp_mst_dc_sink_create(drm_connector);
+			goto mst_dc_sink_create_done;
+		}
 
 		if (create_fake_sink(aconnector))
 			goto stream_create_fail;
@@ -2425,6 +2427,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 stream_create_fail:
 dm_state_null:
 drm_connector_null:
+mst_dc_sink_create_done:
 	return stream;
 }
 
@@ -2725,8 +2728,7 @@ static void create_eml_sink(struct amdgpu_dm_connector *aconnector)
 	};
 	struct edid *edid;
 
-	if (!aconnector->base.edid_blob_ptr ||
-		!aconnector->base.edid_blob_ptr->data) {
+	if (!aconnector->base.edid_blob_ptr) {
 		DRM_ERROR("No EDID firmware found on connector: %s ,forcing to OFF!\n",
 				aconnector->base.name);
 
@@ -4514,18 +4516,15 @@ static int dm_update_crtcs_state(struct dc *dc,
 						__func__, acrtc->base.base.id);
 				break;
 			}
-		}
 
-		if (enable && dc_is_stream_unchanged(new_stream, dm_old_crtc_state->stream) &&
-				dc_is_stream_scaling_unchanged(new_stream, dm_old_crtc_state->stream)) {
-
-			new_crtc_state->mode_changed = false;
-
-			DRM_DEBUG_DRIVER("Mode change not required, setting mode_changed to %d",
-				         new_crtc_state->mode_changed);
+			if (dc_is_stream_unchanged(new_stream, dm_old_crtc_state->stream) &&
+			    dc_is_stream_scaling_unchanged(new_stream, dm_old_crtc_state->stream)) {
+				new_crtc_state->mode_changed = false;
+				DRM_DEBUG_DRIVER("Mode change not required, setting mode_changed to %d",
+						 new_crtc_state->mode_changed);
+			}
 		}
 
-
 		if (!drm_atomic_crtc_needs_modeset(new_crtc_state))
 			goto next_crtc;
 

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

@@ -189,6 +189,8 @@ struct amdgpu_dm_connector {
 	struct mutex hpd_lock;
 
 	bool fake_enable;
+
+	bool mst_connected;
 };
 
 #define to_amdgpu_dm_connector(x) container_of(x, struct amdgpu_dm_connector, base)

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

@@ -180,6 +180,42 @@ static int dm_connector_update_modes(struct drm_connector *connector,
 	return drm_add_edid_modes(connector, edid);
 }
 
+void dm_dp_mst_dc_sink_create(struct drm_connector *connector)
+{
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+	struct edid *edid;
+	struct dc_sink *dc_sink;
+	struct dc_sink_init_data init_params = {
+			.link = aconnector->dc_link,
+			.sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST };
+
+	edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port);
+
+	if (!edid) {
+		drm_mode_connector_update_edid_property(
+			&aconnector->base,
+			NULL);
+		return;
+	}
+
+	aconnector->edid = edid;
+
+	dc_sink = dc_link_add_remote_sink(
+		aconnector->dc_link,
+		(uint8_t *)aconnector->edid,
+		(aconnector->edid->extensions + 1) * EDID_LENGTH,
+		&init_params);
+
+	dc_sink->priv = aconnector;
+	aconnector->dc_sink = dc_sink;
+
+	amdgpu_dm_add_sink_to_freesync_module(
+			connector, aconnector->edid);
+
+	drm_mode_connector_update_edid_property(
+					&aconnector->base, aconnector->edid);
+}
+
 static int dm_dp_mst_get_modes(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
@@ -306,6 +342,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 			drm_mode_connector_set_path_property(connector, pathprop);
 
 			drm_connector_list_iter_end(&conn_iter);
+			aconnector->mst_connected = true;
 			return &aconnector->base;
 		}
 	}
@@ -358,6 +395,8 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 	 */
 	amdgpu_dm_connector_funcs_reset(connector);
 
+	aconnector->mst_connected = true;
+
 	DRM_INFO("DM_MST: added connector: %p [id: %d] [master: %p]\n",
 			aconnector, connector->base.id, aconnector->mst_port);
 
@@ -389,6 +428,8 @@ static void dm_dp_destroy_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 	drm_mode_connector_update_edid_property(
 			&aconnector->base,
 			NULL);
+
+	aconnector->mst_connected = false;
 }
 
 static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
@@ -399,10 +440,18 @@ static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
 	drm_kms_helper_hotplug_event(dev);
 }
 
+static void dm_dp_mst_link_status_reset(struct drm_connector *connector)
+{
+	mutex_lock(&connector->dev->mode_config.mutex);
+	drm_mode_connector_set_link_status_property(connector, DRM_MODE_LINK_STATUS_BAD);
+	mutex_unlock(&connector->dev->mode_config.mutex);
+}
+
 static void dm_dp_mst_register_connector(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct amdgpu_device *adev = dev->dev_private;
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
 
 	if (adev->mode_info.rfbdev)
 		drm_fb_helper_add_one_connector(&adev->mode_info.rfbdev->helper, connector);
@@ -411,6 +460,8 @@ static void dm_dp_mst_register_connector(struct drm_connector *connector)
 
 	drm_connector_register(connector);
 
+	if (aconnector->mst_connected)
+		dm_dp_mst_link_status_reset(connector);
 }
 
 static const struct drm_dp_mst_topology_cbs dm_mst_cbs = {

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

@@ -31,5 +31,6 @@ struct amdgpu_dm_connector;
 
 void amdgpu_dm_initialize_dp_connector(struct amdgpu_display_manager *dm,
 				       struct amdgpu_dm_connector *aconnector);
+void dm_dp_mst_dc_sink_create(struct drm_connector *connector);
 
 #endif

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

@@ -24,7 +24,7 @@
 # It provides the general basic services required by other DAL
 # subcomponents.
 
-BASICS = conversion.o fixpt31_32.o fixpt32_32.o grph_object_id.o \
+BASICS = conversion.o fixpt31_32.o fixpt32_32.o \
 	logger.o log_helpers.o vector.o
 
 AMD_DAL_BASICS = $(addprefix $(AMDDALPATH)/dc/basics/,$(BASICS))

drivers/gpu/drm/amd/display/dc/basics/conversion.c

@@ -46,7 +46,7 @@ uint16_t fixed_point_to_int_frac(
 			arg));
 
 	if (d <= (uint16_t)(1 << integer_bits) - (1 / (uint16_t)divisor))
-		numerator = (uint16_t)dal_fixed31_32_floor(
+		numerator = (uint16_t)dal_fixed31_32_round(
 			dal_fixed31_32_mul_int(
 				arg,
 				divisor));

drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c

@@ -554,6 +554,22 @@ static inline uint32_t ux_dy(
 	return result | fractional_part;
 }
 
+static inline uint32_t clamp_ux_dy(
+	int64_t value,
+	uint32_t integer_bits,
+	uint32_t fractional_bits,
+	uint32_t min_clamp)
+{
+	uint32_t truncated_val = ux_dy(value, integer_bits, fractional_bits);
+
+	if (value >= (1LL << (integer_bits + FIXED31_32_BITS_PER_FRACTIONAL_PART)))
+		return (1 << (integer_bits + fractional_bits)) - 1;
+	else if (truncated_val > min_clamp)
+		return truncated_val;
+	else
+		return min_clamp;
+}
+
 uint32_t dal_fixed31_32_u2d19(
 	struct fixed31_32 arg)
 {
@@ -565,3 +581,15 @@ uint32_t dal_fixed31_32_u0d19(
 {
 	return ux_dy(arg.value, 0, 19);
 }
+
+uint32_t dal_fixed31_32_clamp_u0d14(
+	struct fixed31_32 arg)
+{
+	return clamp_ux_dy(arg.value, 0, 14, 1);
+}
+
+uint32_t dal_fixed31_32_clamp_u0d10(
+	struct fixed31_32 arg)
+{
+	return clamp_ux_dy(arg.value, 0, 10, 1);
+}

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

@@ -190,6 +190,7 @@ static struct graphics_object_id bios_parser_get_connector_id(
 	struct bios_parser *bp = BP_FROM_DCB(dcb);
 	struct graphics_object_id object_id = dal_graphics_object_id_init(
 		0, ENUM_ID_UNKNOWN, OBJECT_TYPE_UNKNOWN);
+	uint16_t id;
 
 	uint32_t connector_table_offset = bp->object_info_tbl_offset
 		+ le16_to_cpu(bp->object_info_tbl.v1_1->usConnectorObjectTableOffset);
@@ -197,12 +198,19 @@ static struct graphics_object_id bios_parser_get_connector_id(
 	ATOM_OBJECT_TABLE *tbl =
 		GET_IMAGE(ATOM_OBJECT_TABLE, connector_table_offset);
 
-	if (tbl && tbl->ucNumberOfObjects > i) {
-		const uint16_t id = le16_to_cpu(tbl->asObjects[i].usObjectID);
+	if (!tbl) {
+		dm_error("Can't get connector table from atom bios.\n");
+		return object_id;
+	}
 
-		object_id = object_id_from_bios_object_id(id);
+	if (tbl->ucNumberOfObjects <= i) {
+		dm_error("Can't find connector id %d in connector table of size %d.\n",
+			 i, tbl->ucNumberOfObjects);
+		return object_id;
 	}
 
+	id = le16_to_cpu(tbl->asObjects[i].usObjectID);
+	object_id = object_id_from_bios_object_id(id);
 	return object_id;
 }
 
@@ -2254,6 +2262,52 @@ static enum bp_result get_gpio_i2c_info(struct bios_parser *bp,
 	return BP_RESULT_OK;
 }
 
+static bool dal_graphics_object_id_is_valid(struct graphics_object_id id)
+{
+	bool rc = true;
+
+	switch (id.type) {
+	case OBJECT_TYPE_UNKNOWN:
+		rc = false;
+		break;
+	case OBJECT_TYPE_GPU:
+	case OBJECT_TYPE_ENGINE:
+		/* do NOT check for id.id == 0 */
+		if (id.enum_id == ENUM_ID_UNKNOWN)
+			rc = false;
+		break;
+	default:
+		if (id.id == 0 || id.enum_id == ENUM_ID_UNKNOWN)
+			rc = false;
+		break;
+	}
+
+	return rc;
+}
+
+static bool dal_graphics_object_id_is_equal(
+	struct graphics_object_id id1,
+	struct graphics_object_id id2)
+{
+	if (false == dal_graphics_object_id_is_valid(id1)) {
+		dm_output_to_console(
+		"%s: Warning: comparing invalid object 'id1'!\n", __func__);
+		return false;
+	}
+
+	if (false == dal_graphics_object_id_is_valid(id2)) {
+		dm_output_to_console(
+		"%s: Warning: comparing invalid object 'id2'!\n", __func__);
+		return false;
+	}
+
+	if (id1.id == id2.id && id1.enum_id == id2.enum_id
+		&& id1.type == id2.type)
+		return true;
+
+	return false;
+}
+
 static ATOM_OBJECT *get_bios_object(struct bios_parser *bp,
 	struct graphics_object_id id)
 {

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

@@ -387,6 +387,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 		bp->cmd_tbl.transmitter_control = transmitter_control_v1_6;
 		break;
 	default:
+		dm_error("Don't have transmitter_control for v%d\n", crev);
 		bp->cmd_tbl.transmitter_control = NULL;
 		break;
 	}
@@ -910,6 +911,8 @@ static void init_set_pixel_clock(struct bios_parser *bp)
 		bp->cmd_tbl.set_pixel_clock = set_pixel_clock_v7;
 		break;
 	default:
+		dm_error("Don't have set_pixel_clock for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(SetPixelClock));
 		bp->cmd_tbl.set_pixel_clock = NULL;
 		break;
 	}
@@ -1227,6 +1230,8 @@ static void init_enable_spread_spectrum_on_ppll(struct bios_parser *bp)
 				enable_spread_spectrum_on_ppll_v3;
 		break;
 	default:
+		dm_error("Don't have enable_spread_spectrum_on_ppll for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(EnableSpreadSpectrumOnPPLL));
 		bp->cmd_tbl.enable_spread_spectrum_on_ppll = NULL;
 		break;
 	}
@@ -1422,6 +1427,8 @@ static void init_adjust_display_pll(struct bios_parser *bp)
 		bp->cmd_tbl.adjust_display_pll = adjust_display_pll_v3;
 		break;
 	default:
+		dm_error("Don't have adjust_display_pll for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(AdjustDisplayPll));
 		bp->cmd_tbl.adjust_display_pll = NULL;
 		break;
 	}
@@ -1695,6 +1702,8 @@ static void init_set_crtc_timing(struct bios_parser *bp)
 					set_crtc_using_dtd_timing_v3;
 			break;
 		default:
+			dm_error("Don't have set_crtc_timing for dtd v%d\n",
+				 dtd_version);
 			bp->cmd_tbl.set_crtc_timing = NULL;
 			break;
 		}
@@ -1704,6 +1713,8 @@ static void init_set_crtc_timing(struct bios_parser *bp)
 			bp->cmd_tbl.set_crtc_timing = set_crtc_timing_v1;
 			break;
 		default:
+			dm_error("Don't have set_crtc_timing for v%d\n",
+				 BIOS_CMD_TABLE_PARA_REVISION(SetCRTC_Timing));
 			bp->cmd_tbl.set_crtc_timing = NULL;
 			break;
 		}
@@ -1890,6 +1901,8 @@ static void init_select_crtc_source(struct bios_parser *bp)
 		bp->cmd_tbl.select_crtc_source = select_crtc_source_v3;
 		break;
 	default:
+		dm_error("Don't select_crtc_source enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(SelectCRTC_Source));
 		bp->cmd_tbl.select_crtc_source = NULL;
 		break;
 	}
@@ -1997,6 +2010,8 @@ static void init_enable_crtc(struct bios_parser *bp)
 		bp->cmd_tbl.enable_crtc = enable_crtc_v1;
 		break;
 	default:
+		dm_error("Don't have enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(EnableCRTC));
 		bp->cmd_tbl.enable_crtc = NULL;
 		break;
 	}
@@ -2103,6 +2118,8 @@ static void init_program_clock(struct bios_parser *bp)
 		bp->cmd_tbl.program_clock = program_clock_v6;
 		break;
 	default:
+		dm_error("Don't have program_clock for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(SetPixelClock));
 		bp->cmd_tbl.program_clock = NULL;
 		break;
 	}
@@ -2324,6 +2341,8 @@ static void init_enable_disp_power_gating(
 				enable_disp_power_gating_v2_1;
 		break;
 	default:
+		dm_error("Don't enable_disp_power_gating enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(EnableDispPowerGating));
 		bp->cmd_tbl.enable_disp_power_gating = NULL;
 		break;
 	}
@@ -2371,6 +2390,8 @@ static void init_set_dce_clock(struct bios_parser *bp)
 		bp->cmd_tbl.set_dce_clock = set_dce_clock_v2_1;
 		break;
 	default:
+		dm_error("Don't have set_dce_clock for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(SetDCEClock));
 		bp->cmd_tbl.set_dce_clock = NULL;
 		break;
 	}

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

@@ -118,6 +118,7 @@ static void init_dig_encoder_control(struct bios_parser *bp)
 		bp->cmd_tbl.dig_encoder_control = encoder_control_digx_v1_5;
 		break;
 	default:
+		dm_error("Don't have dig_encoder_control for v%d\n", version);
 		bp->cmd_tbl.dig_encoder_control = NULL;
 		break;
 	}
@@ -205,6 +206,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 		bp->cmd_tbl.transmitter_control = transmitter_control_v1_6;
 		break;
 	default:
+		dm_error("Don't have transmitter_control for v%d\n", crev);
 		bp->cmd_tbl.transmitter_control = NULL;
 		break;
 	}
@@ -268,6 +270,8 @@ static void init_set_pixel_clock(struct bios_parser *bp)
 		bp->cmd_tbl.set_pixel_clock = set_pixel_clock_v7;
 		break;
 	default:
+		dm_error("Don't have set_pixel_clock for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(setpixelclock));
 		bp->cmd_tbl.set_pixel_clock = NULL;
 		break;
 	}
@@ -379,6 +383,7 @@ static void init_set_crtc_timing(struct bios_parser *bp)
 			set_crtc_using_dtd_timing_v3;
 		break;
 	default:
+		dm_error("Don't have set_crtc_timing for v%d\n", dtd_version);
 		bp->cmd_tbl.set_crtc_timing = NULL;
 		break;
 	}
@@ -498,6 +503,8 @@ static void init_select_crtc_source(struct bios_parser *bp)
 		bp->cmd_tbl.select_crtc_source = select_crtc_source_v3;
 		break;
 	default:
+		dm_error("Don't select_crtc_source enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(selectcrtc_source));
 		bp->cmd_tbl.select_crtc_source = NULL;
 		break;
 	}
@@ -565,6 +572,8 @@ static void init_enable_crtc(struct bios_parser *bp)
 		bp->cmd_tbl.enable_crtc = enable_crtc_v1;
 		break;
 	default:
+		dm_error("Don't have enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(enablecrtc));
 		bp->cmd_tbl.enable_crtc = NULL;
 		break;
 	}
@@ -661,6 +670,8 @@ static void init_enable_disp_power_gating(
 				enable_disp_power_gating_v2_1;
 		break;
 	default:
+		dm_error("Don't enable_disp_power_gating enable_crtc for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(enabledisppowergating));
 		bp->cmd_tbl.enable_disp_power_gating = NULL;
 		break;
 	}
@@ -710,6 +721,8 @@ static void init_set_dce_clock(struct bios_parser *bp)
 		bp->cmd_tbl.set_dce_clock = set_dce_clock_v2_1;
 		break;
 	default:
+		dm_error("Don't have set_dce_clock for v%d\n",
+			 BIOS_CMD_TABLE_PARA_REVISION(setdceclock));
 		bp->cmd_tbl.set_dce_clock = NULL;
 		break;
 	}

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

@@ -29,6 +29,15 @@
 #include "core_types.h"
 #include "dal_asic_id.h"
 
+/*
+ * NOTE:
+ *   This file is gcc-parseable HW gospel, coming straight from HW engineers.
+ *
+ * It doesn't adhere to Linux kernel style and sometimes will do things in odd
+ * ways. Unless there is something clearly wrong with it the code should
+ * remain as-is as it provides us with a guarantee from HW that it is correct.
+ */
+
 /*******************************************************************************
  * Private Functions
  ******************************************************************************/

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

@@ -27,6 +27,15 @@
 #include "dcn_calc_auto.h"
 #include "dcn_calc_math.h"
 
+/*
+ * NOTE:
+ *   This file is gcc-parseable HW gospel, coming straight from HW engineers.
+ *
+ * It doesn't adhere to Linux kernel style and sometimes will do things in odd
+ * ways. Unless there is something clearly wrong with it the code should
+ * remain as-is as it provides us with a guarantee from HW that it is correct.
+ */
+
 /*REVISION#250*/
 void scaler_settings_calculation(struct dcn_bw_internal_vars *v)
 {
@@ -773,11 +782,11 @@ void mode_support_and_system_configuration(struct dcn_bw_internal_vars *v)
 					v->dst_y_after_scaler = 0.0;
 				}
 				v->time_calc = 24.0 / v->projected_dcfclk_deep_sleep;
-				v->v_update_offset[k] =dcn_bw_ceil2(v->htotal[k] / 4.0, 1.0);
+				v->v_update_offset[k][j] = dcn_bw_ceil2(v->htotal[k] / 4.0, 1.0);
 				v->total_repeater_delay = v->max_inter_dcn_tile_repeaters * (2.0 / (v->required_dispclk[i][j] / (j + 1)) + 3.0 / v->required_dispclk[i][j]);
-				v->v_update_width[k] = (14.0 / v->projected_dcfclk_deep_sleep + 12.0 / (v->required_dispclk[i][j] / (j + 1)) + v->total_repeater_delay) * v->pixel_clock[k];
-				v->v_ready_offset[k] =dcn_bw_max2(150.0 / (v->required_dispclk[i][j] / (j + 1)), v->total_repeater_delay + 20.0 / v->projected_dcfclk_deep_sleep + 10.0 / (v->required_dispclk[i][j] / (j + 1))) * v->pixel_clock[k];
-				v->time_setup = (v->v_update_offset[k] + v->v_update_width[k] + v->v_ready_offset[k]) / v->pixel_clock[k];
+				v->v_update_width[k][j] = (14.0 / v->projected_dcfclk_deep_sleep + 12.0 / (v->required_dispclk[i][j] / (j + 1)) + v->total_repeater_delay) * v->pixel_clock[k];
+				v->v_ready_offset[k][j] = dcn_bw_max2(150.0 / (v->required_dispclk[i][j] / (j + 1)), v->total_repeater_delay + 20.0 / v->projected_dcfclk_deep_sleep + 10.0 / (v->required_dispclk[i][j] / (j + 1))) * v->pixel_clock[k];
+				v->time_setup = (v->v_update_offset[k][j] + v->v_update_width[k][j] + v->v_ready_offset[k][j]) / v->pixel_clock[k];
 				v->extra_latency = v->urgent_round_trip_and_out_of_order_latency_per_state[i] + (v->total_number_of_active_dpp[i][j] * v->pixel_chunk_size_in_kbyte + v->total_number_of_dcc_active_dpp[i][j] * v->meta_chunk_size) * 1024.0 / v->return_bw_per_state[i];
 				if (v->pte_enable == dcn_bw_yes) {
 					v->extra_latency = v->extra_latency + v->total_number_of_active_dpp[i][j] * v->pte_chunk_size * 1024.0 / v->return_bw_per_state[i];

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

@@ -25,37 +25,44 @@
 
 #include "dcn_calc_math.h"
 
+#define isNaN(number) ((number) != (number))
+
+/*
+ * NOTE:
+ *   This file is gcc-parseable HW gospel, coming straight from HW engineers.
+ *
+ * It doesn't adhere to Linux kernel style and sometimes will do things in odd
+ * ways. Unless there is something clearly wrong with it the code should
+ * remain as-is as it provides us with a guarantee from HW that it is correct.
+ */
+
 float dcn_bw_mod(const float arg1, const float arg2)
 {
-	if (arg1 != arg1)
+	if (isNaN(arg1))
 		return arg2;
-	if (arg2 != arg2)
+	if (isNaN(arg2))
 		return arg1;
 	return arg1 - arg1 * ((int) (arg1 / arg2));
 }
 
 float dcn_bw_min2(const float arg1, const float arg2)
 {
-	if (arg1 != arg1)
+	if (isNaN(arg1))
 		return arg2;
-	if (arg2 != arg2)
+	if (isNaN(arg2))
 		return arg1;
 	return arg1 < arg2 ? arg1 : arg2;
 }
 
 unsigned int dcn_bw_max(const unsigned int arg1, const unsigned int arg2)
 {
-	if (arg1 != arg1)
-		return arg2;
-	if (arg2 != arg2)
-		return arg1;
 	return arg1 > arg2 ? arg1 : arg2;
 }
 float dcn_bw_max2(const float arg1, const float arg2)
 {
-	if (arg1 != arg1)
+	if (isNaN(arg1))
 		return arg2;
-	if (arg2 != arg2)
+	if (isNaN(arg2))
 		return arg1;
 	return arg1 > arg2 ? arg1 : arg2;
 }

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

@@ -33,6 +33,15 @@
 #include "dcn10/dcn10_resource.h"
 #include "dcn_calc_math.h"
 
+/*
+ * NOTE:
+ *   This file is gcc-parseable HW gospel, coming straight from HW engineers.
+ *
+ * It doesn't adhere to Linux kernel style and sometimes will do things in odd
+ * ways. Unless there is something clearly wrong with it the code should
+ * remain as-is as it provides us with a guarantee from HW that it is correct.
+ */
+
 /* Defaults from spreadsheet rev#247 */
 const struct dcn_soc_bounding_box dcn10_soc_defaults = {
 		/* latencies */
@@ -878,6 +887,17 @@ bool dcn_validate_bandwidth(
 						+ pipe->bottom_pipe->plane_res.scl_data.recout.width;
 			}
 
+			if (pipe->plane_state->rotation % 2 == 0) {
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+					|| v->scaler_rec_out_width[input_idx] == v->viewport_width[input_idx]);
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+					|| v->scaler_recout_height[input_idx] == v->viewport_height[input_idx]);
+			} else {
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+					|| v->scaler_recout_height[input_idx] == v->viewport_width[input_idx]);
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+					|| v->scaler_rec_out_width[input_idx] == v->viewport_height[input_idx]);
+			}
 			v->dcc_enable[input_idx] = pipe->plane_state->dcc.enable ? dcn_bw_yes : dcn_bw_no;
 			v->source_pixel_format[input_idx] = tl_pixel_format_to_bw_defs(
 					pipe->plane_state->format);
@@ -888,6 +908,15 @@ bool dcn_validate_bandwidth(
 			v->override_vta_ps[input_idx] = pipe->plane_res.scl_data.taps.v_taps;
 			v->override_hta_pschroma[input_idx] = pipe->plane_res.scl_data.taps.h_taps_c;
 			v->override_vta_pschroma[input_idx] = pipe->plane_res.scl_data.taps.v_taps_c;
+			/*
+			 * Spreadsheet doesn't handle taps_c is one properly,
+			 * need to force Chroma to always be scaled to pass
+			 * bandwidth validation.
+			 */
+			if (v->override_hta_pschroma[input_idx] == 1)
+				v->override_hta_pschroma[input_idx] = 2;
+			if (v->override_vta_pschroma[input_idx] == 1)
+				v->override_vta_pschroma[input_idx] = 2;
 			v->source_scan[input_idx] = (pipe->plane_state->rotation % 2) ? dcn_bw_vert : dcn_bw_hor;
 		}
 		if (v->is_line_buffer_bpp_fixed == dcn_bw_yes)
@@ -985,9 +1014,9 @@ bool dcn_validate_bandwidth(
 			if (pipe->top_pipe && pipe->top_pipe->plane_state == pipe->plane_state)
 				continue;
 
-			pipe->pipe_dlg_param.vupdate_width = v->v_update_width[input_idx];
-			pipe->pipe_dlg_param.vupdate_offset = v->v_update_offset[input_idx];
-			pipe->pipe_dlg_param.vready_offset = v->v_ready_offset[input_idx];
+			pipe->pipe_dlg_param.vupdate_width = v->v_update_width[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
+			pipe->pipe_dlg_param.vupdate_offset = v->v_update_offset[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
+			pipe->pipe_dlg_param.vready_offset = v->v_ready_offset[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
 			pipe->pipe_dlg_param.vstartup_start = v->v_startup[input_idx];
 
 			pipe->pipe_dlg_param.htotal = pipe->stream->timing.h_total;
@@ -1026,9 +1055,9 @@ bool dcn_validate_bandwidth(
 					 TIMING_3D_FORMAT_SIDE_BY_SIDE))) {
 					if (hsplit_pipe && hsplit_pipe->plane_state == pipe->plane_state) {
 						/* update previously split pipe */
-						hsplit_pipe->pipe_dlg_param.vupdate_width = v->v_update_width[input_idx];
-						hsplit_pipe->pipe_dlg_param.vupdate_offset = v->v_update_offset[input_idx];
-						hsplit_pipe->pipe_dlg_param.vready_offset = v->v_ready_offset[input_idx];
+						hsplit_pipe->pipe_dlg_param.vupdate_width = v->v_update_width[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
+						hsplit_pipe->pipe_dlg_param.vupdate_offset = v->v_update_offset[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
+						hsplit_pipe->pipe_dlg_param.vready_offset = v->v_ready_offset[input_idx][v->dpp_per_plane[input_idx] == 2 ? 1 : 0];
 						hsplit_pipe->pipe_dlg_param.vstartup_start = v->v_startup[input_idx];
 
 						hsplit_pipe->pipe_dlg_param.htotal = pipe->stream->timing.h_total;
@@ -1556,35 +1585,6 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 			dc->dcn_ip->can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one,
 			dc->dcn_ip->bug_forcing_luma_and_chroma_request_to_same_size_fixed,
 			dc->dcn_ip->dcfclk_cstate_latency);
-	dc->dml.soc.vmin.socclk_mhz = dc->dcn_soc->socclk;
-	dc->dml.soc.vmid.socclk_mhz = dc->dcn_soc->socclk;
-	dc->dml.soc.vnom.socclk_mhz = dc->dcn_soc->socclk;
-	dc->dml.soc.vmax.socclk_mhz = dc->dcn_soc->socclk;
-
-	dc->dml.soc.vmin.dcfclk_mhz = dc->dcn_soc->dcfclkv_min0p65;
-	dc->dml.soc.vmid.dcfclk_mhz = dc->dcn_soc->dcfclkv_mid0p72;
-	dc->dml.soc.vnom.dcfclk_mhz = dc->dcn_soc->dcfclkv_nom0p8;
-	dc->dml.soc.vmax.dcfclk_mhz = dc->dcn_soc->dcfclkv_max0p9;
-
-	dc->dml.soc.vmin.dispclk_mhz = dc->dcn_soc->max_dispclk_vmin0p65;
-	dc->dml.soc.vmid.dispclk_mhz = dc->dcn_soc->max_dispclk_vmid0p72;
-	dc->dml.soc.vnom.dispclk_mhz = dc->dcn_soc->max_dispclk_vnom0p8;
-	dc->dml.soc.vmax.dispclk_mhz = dc->dcn_soc->max_dispclk_vmax0p9;
-
-	dc->dml.soc.vmin.dppclk_mhz = dc->dcn_soc->max_dppclk_vmin0p65;
-	dc->dml.soc.vmid.dppclk_mhz = dc->dcn_soc->max_dppclk_vmid0p72;
-	dc->dml.soc.vnom.dppclk_mhz = dc->dcn_soc->max_dppclk_vnom0p8;
-	dc->dml.soc.vmax.dppclk_mhz = dc->dcn_soc->max_dppclk_vmax0p9;
-
-	dc->dml.soc.vmin.phyclk_mhz = dc->dcn_soc->phyclkv_min0p65;
-	dc->dml.soc.vmid.phyclk_mhz = dc->dcn_soc->phyclkv_mid0p72;
-	dc->dml.soc.vnom.phyclk_mhz = dc->dcn_soc->phyclkv_nom0p8;
-	dc->dml.soc.vmax.phyclk_mhz = dc->dcn_soc->phyclkv_max0p9;
-
-	dc->dml.soc.vmin.dram_bw_per_chan_gbps = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65;
-	dc->dml.soc.vmid.dram_bw_per_chan_gbps = dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72;
-	dc->dml.soc.vnom.dram_bw_per_chan_gbps = dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8;
-	dc->dml.soc.vmax.dram_bw_per_chan_gbps = dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9;
 
 	dc->dml.soc.sr_exit_time_us = dc->dcn_soc->sr_exit_time;
 	dc->dml.soc.sr_enter_plus_exit_time_us = dc->dcn_soc->sr_enter_plus_exit_time;

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

@@ -283,19 +283,17 @@ static bool construct(struct dc *dc,
 		const struct dc_init_data *init_params)
 {
 	struct dal_logger *logger;
-	struct dc_context *dc_ctx = kzalloc(sizeof(*dc_ctx), GFP_KERNEL);
-	struct bw_calcs_dceip *dc_dceip = kzalloc(sizeof(*dc_dceip),
-						  GFP_KERNEL);
-	struct bw_calcs_vbios *dc_vbios = kzalloc(sizeof(*dc_vbios),
-						  GFP_KERNEL);
+	struct dc_context *dc_ctx;
+	struct bw_calcs_dceip *dc_dceip;
+	struct bw_calcs_vbios *dc_vbios;
 #ifdef CONFIG_DRM_AMD_DC_DCN1_0
-	struct dcn_soc_bounding_box *dcn_soc = kzalloc(sizeof(*dcn_soc),
-						       GFP_KERNEL);
-	struct dcn_ip_params *dcn_ip = kzalloc(sizeof(*dcn_ip), GFP_KERNEL);
+	struct dcn_soc_bounding_box *dcn_soc;
+	struct dcn_ip_params *dcn_ip;
 #endif
 
 	enum dce_version dc_version = DCE_VERSION_UNKNOWN;
 
+	dc_dceip = kzalloc(sizeof(*dc_dceip), GFP_KERNEL);
 	if (!dc_dceip) {
 		dm_error("%s: failed to create dceip\n", __func__);
 		goto fail;
@@ -303,6 +301,7 @@ static bool construct(struct dc *dc,
 
 	dc->bw_dceip = dc_dceip;
 
+	dc_vbios = kzalloc(sizeof(*dc_vbios), GFP_KERNEL);
 	if (!dc_vbios) {
 		dm_error("%s: failed to create vbios\n", __func__);
 		goto fail;
@@ -310,6 +309,7 @@ static bool construct(struct dc *dc,
 
 	dc->bw_vbios = dc_vbios;
 #ifdef CONFIG_DRM_AMD_DC_DCN1_0
+	dcn_soc = kzalloc(sizeof(*dcn_soc), GFP_KERNEL);
 	if (!dcn_soc) {
 		dm_error("%s: failed to create dcn_soc\n", __func__);
 		goto fail;
@@ -317,6 +317,7 @@ static bool construct(struct dc *dc,
 
 	dc->dcn_soc = dcn_soc;
 
+	dcn_ip = kzalloc(sizeof(*dcn_ip), GFP_KERNEL);
 	if (!dcn_ip) {
 		dm_error("%s: failed to create dcn_ip\n", __func__);
 		goto fail;
@@ -325,11 +326,18 @@ static bool construct(struct dc *dc,
 	dc->dcn_ip = dcn_ip;
 #endif
 
+	dc_ctx = kzalloc(sizeof(*dc_ctx), GFP_KERNEL);
 	if (!dc_ctx) {
 		dm_error("%s: failed to create ctx\n", __func__);
 		goto fail;
 	}
 
+	dc_ctx->cgs_device = init_params->cgs_device;
+	dc_ctx->driver_context = init_params->driver;
+	dc_ctx->dc = dc;
+	dc_ctx->asic_id = init_params->asic_id;
+	dc->ctx = dc_ctx;
+
 	dc->current_state = dc_create_state();
 
 	if (!dc->current_state) {
@@ -337,11 +345,6 @@ static bool construct(struct dc *dc,
 		goto fail;
 	}
 
-	dc_ctx->cgs_device = init_params->cgs_device;
-	dc_ctx->driver_context = init_params->driver;
-	dc_ctx->dc = dc;
-	dc_ctx->asic_id = init_params->asic_id;
-
 	/* Create logger */
 	logger = dal_logger_create(dc_ctx, init_params->log_mask);
 
@@ -351,11 +354,10 @@ static bool construct(struct dc *dc,
 		goto fail;
 	}
 	dc_ctx->logger = logger;
-	dc->ctx = dc_ctx;
-	dc->ctx->dce_environment = init_params->dce_environment;
+	dc_ctx->dce_environment = init_params->dce_environment;
 
 	dc_version = resource_parse_asic_id(init_params->asic_id);
-	dc->ctx->dce_version = dc_version;
+	dc_ctx->dce_version = dc_version;
 
 #if defined(CONFIG_DRM_AMD_DC_FBC)
 	dc->ctx->fbc_gpu_addr = init_params->fbc_gpu_addr;
@@ -578,7 +580,7 @@ static void program_timing_sync(
 		for (j = 0; j < group_size; j++) {
 			struct pipe_ctx *temp;
 
-			if (!pipe_set[j]->stream_res.tg->funcs->is_blanked(pipe_set[j]->stream_res.tg)) {
+			if (pipe_set[j]->stream_res.tg->funcs->is_blanked && !pipe_set[j]->stream_res.tg->funcs->is_blanked(pipe_set[j]->stream_res.tg)) {
 				if (j == 0)
 					break;
 
@@ -591,7 +593,7 @@ static void program_timing_sync(
 
 		/* remove any other unblanked pipes as they have already been synced */
 		for (j = j + 1; j < group_size; j++) {
-			if (!pipe_set[j]->stream_res.tg->funcs->is_blanked(pipe_set[j]->stream_res.tg)) {
+			if (pipe_set[j]->stream_res.tg->funcs->is_blanked && !pipe_set[j]->stream_res.tg->funcs->is_blanked(pipe_set[j]->stream_res.tg)) {
 				group_size--;
 				pipe_set[j] = pipe_set[group_size];
 				j--;
@@ -786,6 +788,8 @@ 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;
+
 	/* 3rd param should be true, temp w/a for RV*/
 #if defined(CONFIG_DRM_AMD_DC_DCN1_0)
 	dc->hwss.set_bandwidth(dc, context, dc->ctx->dce_version < DCN_VERSION_1_0);
@@ -981,6 +985,11 @@ static enum surface_update_type get_plane_info_update_type(const struct dc_surfa
 	if (u->plane_info->per_pixel_alpha != u->surface->per_pixel_alpha)
 		update_flags->bits.per_pixel_alpha_change = 1;
 
+	if (u->plane_info->dcc.enable != u->surface->dcc.enable
+			|| u->plane_info->dcc.grph.independent_64b_blks != u->surface->dcc.grph.independent_64b_blks
+			|| u->plane_info->dcc.grph.meta_pitch != u->surface->dcc.grph.meta_pitch)
+		update_flags->bits.dcc_change = 1;
+
 	if (pixel_format_to_bpp(u->plane_info->format) !=
 			pixel_format_to_bpp(u->surface->format))
 		/* different bytes per element will require full bandwidth
@@ -1178,12 +1187,6 @@ static void commit_planes_for_stream(struct dc *dc,
 	if (update_type == UPDATE_TYPE_FULL) {
 		dc->hwss.set_bandwidth(dc, context, false);
 		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];
-
-			dc->hwss.wait_for_mpcc_disconnect(dc, dc->res_pool, pipe_ctx);
-		}
 	}
 
 	if (surface_count == 0) {

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

@@ -28,6 +28,8 @@
 #include "timing_generator.h"
 #include "hw_sequencer.h"
 
+#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0]))
+
 /* used as index in array of black_color_format */
 enum black_color_format {
 	BLACK_COLOR_FORMAT_RGB_FULLRANGE = 0,
@@ -38,6 +40,15 @@ enum black_color_format {
 	BLACK_COLOR_FORMAT_DEBUG,
 };
 
+enum dc_color_space_type {
+	COLOR_SPACE_RGB_TYPE,
+	COLOR_SPACE_RGB_LIMITED_TYPE,
+	COLOR_SPACE_YCBCR601_TYPE,
+	COLOR_SPACE_YCBCR709_TYPE,
+	COLOR_SPACE_YCBCR601_LIMITED_TYPE,
+	COLOR_SPACE_YCBCR709_LIMITED_TYPE
+};
+
 static const struct tg_color black_color_format[] = {
 	/* BlackColorFormat_RGB_FullRange */
 	{0, 0, 0},
@@ -53,6 +64,140 @@ static const struct tg_color black_color_format[] = {
 	{0xff, 0xff, 0},
 };
 
+struct out_csc_color_matrix_type {
+	enum dc_color_space_type color_space_type;
+	uint16_t regval[12];
+};
+
+static const struct out_csc_color_matrix_type output_csc_matrix[] = {
+	{ COLOR_SPACE_RGB_TYPE,
+		{ 0x2000, 0, 0, 0, 0, 0x2000, 0, 0, 0, 0, 0x2000, 0} },
+	{ COLOR_SPACE_RGB_LIMITED_TYPE,
+		{ 0x1B67, 0, 0, 0x201, 0, 0x1B67, 0, 0x201, 0, 0, 0x1B67, 0x201} },
+	{ COLOR_SPACE_YCBCR601_TYPE,
+		{ 0xE04, 0xF444, 0xFDB9, 0x1004, 0x831, 0x1016, 0x320, 0x201, 0xFB45,
+				0xF6B7, 0xE04, 0x1004} },
+	{ COLOR_SPACE_YCBCR709_TYPE,
+		{ 0xE04, 0xF345, 0xFEB7, 0x1004, 0x5D3, 0x1399, 0x1FA,
+				0x201, 0xFCCA, 0xF533, 0xE04, 0x1004} },
+
+	/* TODO: correct values below */
+	{ COLOR_SPACE_YCBCR601_LIMITED_TYPE,
+		{ 0xE00, 0xF447, 0xFDB9, 0x1000, 0x991,
+				0x12C9, 0x3A6, 0x200, 0xFB47, 0xF6B9, 0xE00, 0x1000} },
+	{ COLOR_SPACE_YCBCR709_LIMITED_TYPE,
+		{ 0xE00, 0xF349, 0xFEB7, 0x1000, 0x6CE, 0x16E3,
+				0x24F, 0x200, 0xFCCB, 0xF535, 0xE00, 0x1000} },
+};
+
+static bool is_rgb_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_SRGB			||
+		color_space == COLOR_SPACE_XR_RGB		||
+		color_space == COLOR_SPACE_MSREF_SCRGB		||
+		color_space == COLOR_SPACE_2020_RGB_FULLRANGE	||
+		color_space == COLOR_SPACE_ADOBERGB		||
+		color_space == COLOR_SPACE_DCIP3	||
+		color_space == COLOR_SPACE_DOLBYVISION)
+		ret = true;
+	return ret;
+}
+
+static bool is_rgb_limited_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_SRGB_LIMITED		||
+		color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE)
+		ret = true;
+	return ret;
+}
+
+static bool is_ycbcr601_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_YCBCR601	||
+		color_space == COLOR_SPACE_XV_YCC_601)
+		ret = true;
+	return ret;
+}
+
+static bool is_ycbcr601_limited_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_YCBCR601_LIMITED)
+		ret = true;
+	return ret;
+}
+
+static bool is_ycbcr709_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_YCBCR709	||
+		color_space == COLOR_SPACE_XV_YCC_709)
+		ret = true;
+	return ret;
+}
+
+static bool is_ycbcr709_limited_type(
+		enum dc_color_space color_space)
+{
+	bool ret = false;
+
+	if (color_space == COLOR_SPACE_YCBCR709_LIMITED)
+		ret = true;
+	return ret;
+}
+enum dc_color_space_type get_color_space_type(enum dc_color_space color_space)
+{
+	enum dc_color_space_type type = COLOR_SPACE_RGB_TYPE;
+
+	if (is_rgb_type(color_space))
+		type = COLOR_SPACE_RGB_TYPE;
+	else if (is_rgb_limited_type(color_space))
+		type = COLOR_SPACE_RGB_LIMITED_TYPE;
+	else if (is_ycbcr601_type(color_space))
+		type = COLOR_SPACE_YCBCR601_TYPE;
+	else if (is_ycbcr709_type(color_space))
+		type = COLOR_SPACE_YCBCR709_TYPE;
+	else if (is_ycbcr601_limited_type(color_space))
+		type = COLOR_SPACE_YCBCR601_LIMITED_TYPE;
+	else if (is_ycbcr709_limited_type(color_space))
+		type = COLOR_SPACE_YCBCR709_LIMITED_TYPE;
+
+	return type;
+}
+
+const uint16_t *find_color_matrix(enum dc_color_space color_space,
+							uint32_t *array_size)
+{
+	int i;
+	enum dc_color_space_type type;
+	const uint16_t *val = NULL;
+	int arr_size = NUM_ELEMENTS(output_csc_matrix);
+
+	type = get_color_space_type(color_space);
+	for (i = 0; i < arr_size; i++)
+		if (output_csc_matrix[i].color_space_type == type) {
+			val = output_csc_matrix[i].regval;
+			*array_size = 12;
+			break;
+		}
+
+	return val;
+}
+
+
 void color_space_to_black_color(
 	const struct dc *dc,
 	enum dc_color_space colorspace,

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

@@ -938,8 +938,9 @@ static bool construct(
 	link->link_id = bios->funcs->get_connector_id(bios, init_params->connector_index);
 
 	if (link->link_id.type != OBJECT_TYPE_CONNECTOR) {
-		dm_error("%s: Invalid Connector ObjectID from Adapter Service for connector index:%d!\n",
-				__func__, init_params->connector_index);
+		dm_error("%s: Invalid Connector ObjectID from Adapter Service for connector index:%d! type %d expected %d\n",
+			 __func__, init_params->connector_index,
+			 link->link_id.type, OBJECT_TYPE_CONNECTOR);
 		goto create_fail;
 	}
 
@@ -1271,6 +1272,24 @@ static enum dc_status enable_link_dp(
 	return status;
 }
 
+static enum dc_status enable_link_edp(
+		struct dc_state *state,
+		struct pipe_ctx *pipe_ctx)
+{
+	enum dc_status status;
+	struct dc_stream_state *stream = pipe_ctx->stream;
+	struct dc_link *link = stream->sink->link;
+
+	link->dc->hwss.edp_power_control(link, true);
+	link->dc->hwss.edp_wait_for_hpd_ready(link, true);
+
+	status = enable_link_dp(state, pipe_ctx);
+
+	link->dc->hwss.edp_backlight_control(link, true);
+
+	return status;
+}
+
 static enum dc_status enable_link_dp_mst(
 		struct dc_state *state,
 		struct pipe_ctx *pipe_ctx)
@@ -1746,9 +1765,11 @@ static enum dc_status enable_link(
 	enum dc_status status = DC_ERROR_UNEXPECTED;
 	switch (pipe_ctx->stream->signal) {
 	case SIGNAL_TYPE_DISPLAY_PORT:
-	case SIGNAL_TYPE_EDP:
 		status = enable_link_dp(state, pipe_ctx);
 		break;
+	case SIGNAL_TYPE_EDP:
+		status = enable_link_edp(state, pipe_ctx);
+		break;
 	case SIGNAL_TYPE_DISPLAY_PORT_MST:
 		status = enable_link_dp_mst(state, pipe_ctx);
 		msleep(200);
@@ -1801,7 +1822,7 @@ static void disable_link(struct dc_link *link, enum signal_type signal)
 		link->link_enc->funcs->disable_output(link->link_enc, signal);
 }
 
-bool dp_active_dongle_validate_timing(
+static bool dp_active_dongle_validate_timing(
 		const struct dc_crtc_timing *timing,
 		const struct dc_dongle_caps *dongle_caps)
 {
@@ -1833,6 +1854,8 @@ bool dp_active_dongle_validate_timing(
 	/* Check Color Depth and Pixel Clock */
 	if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
 		required_pix_clk /= 2;
+	else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+		required_pix_clk = required_pix_clk * 2 / 3;
 
 	switch (timing->display_color_depth) {
 	case COLOR_DEPTH_666:
@@ -1907,12 +1930,18 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 {
 	struct dc  *core_dc = link->ctx->dc;
 	struct abm *abm = core_dc->res_pool->abm;
+	struct dmcu *dmcu = core_dc->res_pool->dmcu;
 	unsigned int controller_id = 0;
+	bool use_smooth_brightness = true;
 	int i;
 
-	if ((abm == NULL) || (abm->funcs->set_backlight_level == NULL))
+	if ((dmcu == NULL) ||
+		(abm == NULL) ||
+		(abm->funcs->set_backlight_level == NULL))
 		return false;
 
+	use_smooth_brightness = dmcu->funcs->is_dmcu_initialized(dmcu);
+
 	dm_logger_write(link->ctx->logger, LOG_BACKLIGHT,
 			"New Backlight level: %d (0x%X)\n", level, level);
 
@@ -1935,7 +1964,8 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 				abm,
 				level,
 				frame_ramp,
-				controller_id);
+				controller_id,
+				use_smooth_brightness);
 	}
 
 	return true;
@@ -1952,144 +1982,6 @@ bool dc_link_set_psr_enable(const struct dc_link *link, bool enable, bool wait)
 	return true;
 }
 
-bool dc_link_get_psr_state(const struct dc_link *link, uint32_t *psr_state)
-{
-	struct dc  *core_dc = link->ctx->dc;
-	struct dmcu *dmcu = core_dc->res_pool->dmcu;
-
-	if (dmcu != NULL && link->psr_enabled)
-		dmcu->funcs->get_psr_state(dmcu, psr_state);
-
-	return true;
-}
-
-bool dc_link_setup_psr(struct dc_link *link,
-		const struct dc_stream_state *stream, struct psr_config *psr_config,
-		struct psr_context *psr_context)
-{
-	struct dc  *core_dc = link->ctx->dc;
-	struct dmcu *dmcu = core_dc->res_pool->dmcu;
-	int i;
-
-	psr_context->controllerId = CONTROLLER_ID_UNDEFINED;
-
-	if (link != NULL &&
-		dmcu != NULL) {
-		/* updateSinkPsrDpcdConfig*/
-		union dpcd_psr_configuration psr_configuration;
-
-		memset(&psr_configuration, 0, sizeof(psr_configuration));
-
-		psr_configuration.bits.ENABLE                    = 1;
-		psr_configuration.bits.CRC_VERIFICATION          = 1;
-		psr_configuration.bits.FRAME_CAPTURE_INDICATION  =
-				psr_config->psr_frame_capture_indication_req;
-
-		/* Check for PSR v2*/
-		if (psr_config->psr_version == 0x2) {
-			/* For PSR v2 selective update.
-			 * Indicates whether sink should start capturing
-			 * immediately following active scan line,
-			 * or starting with the 2nd active scan line.
-			 */
-			psr_configuration.bits.LINE_CAPTURE_INDICATION = 0;
-			/*For PSR v2, determines whether Sink should generate
-			 * IRQ_HPD when CRC mismatch is detected.
-			 */
-			psr_configuration.bits.IRQ_HPD_WITH_CRC_ERROR    = 1;
-		}
-
-		dm_helpers_dp_write_dpcd(
-			link->ctx,
-			link,
-			368,
-			&psr_configuration.raw,
-			sizeof(psr_configuration.raw));
-
-		psr_context->channel = link->ddc->ddc_pin->hw_info.ddc_channel;
-		psr_context->transmitterId = link->link_enc->transmitter;
-		psr_context->engineId = link->link_enc->preferred_engine;
-
-		for (i = 0; i < MAX_PIPES; i++) {
-			if (core_dc->current_state->res_ctx.pipe_ctx[i].stream
-					== stream) {
-				/* dmcu -1 for all controller id values,
-				 * therefore +1 here
-				 */
-				psr_context->controllerId =
-					core_dc->current_state->res_ctx.
-					pipe_ctx[i].stream_res.tg->inst + 1;
-				break;
-			}
-		}
-
-		/* Hardcoded for now.  Can be Pcie or Uniphy (or Unknown)*/
-		psr_context->phyType = PHY_TYPE_UNIPHY;
-		/*PhyId is associated with the transmitter id*/
-		psr_context->smuPhyId = link->link_enc->transmitter;
-
-		psr_context->crtcTimingVerticalTotal = stream->timing.v_total;
-		psr_context->vsyncRateHz = div64_u64(div64_u64((stream->
-						timing.pix_clk_khz * 1000),
-						stream->timing.v_total),
-						stream->timing.h_total);
-
-		psr_context->psrSupportedDisplayConfig = true;
-		psr_context->psrExitLinkTrainingRequired =
-			psr_config->psr_exit_link_training_required;
-		psr_context->sdpTransmitLineNumDeadline =
-			psr_config->psr_sdp_transmit_line_num_deadline;
-		psr_context->psrFrameCaptureIndicationReq =
-			psr_config->psr_frame_capture_indication_req;
-
-		psr_context->skipPsrWaitForPllLock = 0; /* only = 1 in KV */
-
-		psr_context->numberOfControllers =
-				link->dc->res_pool->res_cap->num_timing_generator;
-
-		psr_context->rfb_update_auto_en = true;
-
-		/* 2 frames before enter PSR. */
-		psr_context->timehyst_frames = 2;
-		/* half a frame
-		 * (units in 100 lines, i.e. a value of 1 represents 100 lines)
-		 */
-		psr_context->hyst_lines = stream->timing.v_total / 2 / 100;
-		psr_context->aux_repeats = 10;
-
-		psr_context->psr_level.u32all = 0;
-
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-		/*skip power down the single pipe since it blocks the cstate*/
-		if (ASIC_REV_IS_RAVEN(link->ctx->asic_id.hw_internal_rev))
-			psr_context->psr_level.bits.SKIP_CRTC_DISABLE = true;
-#endif
-
-		/* SMU will perform additional powerdown sequence.
-		 * For unsupported ASICs, set psr_level flag to skip PSR
-		 *  static screen notification to SMU.
-		 *  (Always set for DAL2, did not check ASIC)
-		 */
-		psr_context->psr_level.bits.SKIP_SMU_NOTIFICATION = 1;
-
-		/* Complete PSR entry before aborting to prevent intermittent
-		 * freezes on certain eDPs
-		 */
-		psr_context->psr_level.bits.DISABLE_PSR_ENTRY_ABORT = 1;
-
-		/* Controls additional delay after remote frame capture before
-		 * continuing power down, default = 0
-		 */
-		psr_context->frame_delay = 0;
-
-		link->psr_enabled = true;
-		dmcu->funcs->setup_psr(dmcu, link, psr_context);
-		return true;
-	} else
-		return false;
-
-}
-
 const struct dc_link_status *dc_link_get_status(const struct dc_link *link)
 {
 	return &link->link_status;
@@ -2418,6 +2310,9 @@ void core_link_disable_stream(struct pipe_ctx *pipe_ctx, int option)
 	if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
 		deallocate_mst_payload(pipe_ctx);
 
+	if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP)
+		core_dc->hwss.edp_backlight_control(pipe_ctx->stream->sink->link, false);
+
 	core_dc->hwss.disable_stream(pipe_ctx, option);
 
 	disable_link(pipe_ctx->stream->sink->link, pipe_ctx->stream->signal);

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

@@ -1470,6 +1470,12 @@ void decide_link_settings(struct dc_stream_state *stream,
 		return;
 	}
 
+	/* EDP use the link cap setting */
+	if (stream->sink->sink_signal == SIGNAL_TYPE_EDP) {
+		*link_setting = link->verified_link_cap;
+		return;
+	}
+
 	/* search for the minimum link setting that:
 	 * 1. is supported according to the link training result
 	 * 2. could support the b/w requested by the timing

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

@@ -88,15 +88,7 @@ void dp_enable_link_phy(
 	}
 
 	if (dc_is_dp_sst_signal(signal)) {
-		if (signal == SIGNAL_TYPE_EDP) {
-			link->dc->hwss.edp_power_control(link, true);
-			link_enc->funcs->enable_dp_output(
-						link_enc,
-						link_settings,
-						clock_source);
-			link->dc->hwss.edp_backlight_control(link, true);
-		} else
-			link_enc->funcs->enable_dp_output(
+		link_enc->funcs->enable_dp_output(
 						link_enc,
 						link_settings,
 						clock_source);
@@ -138,7 +130,6 @@ void dp_disable_link_phy(struct dc_link *link, enum signal_type signal)
 		dp_receiver_power_ctrl(link, false);
 
 	if (signal == SIGNAL_TYPE_EDP) {
-		link->dc->hwss.edp_backlight_control(link, false);
 		edp_receiver_ready_T9(link);
 		link->link_enc->funcs->disable_output(link->link_enc, signal);
 		link->dc->hwss.edp_power_control(link, false);

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

@@ -498,26 +498,15 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	data->viewport_c.height = (data->viewport.height + vpc_div - 1) / vpc_div;
 
 	/* Handle hsplit */
-	if (pri_split || sec_split) {
-		/* HMirror XOR Secondary_pipe XOR Rotation_180 */
-		bool right_view = (sec_split != plane_state->horizontal_mirror) !=
-					(plane_state->rotation == ROTATION_ANGLE_180);
-
-		if (plane_state->rotation == ROTATION_ANGLE_90
-				|| plane_state->rotation == ROTATION_ANGLE_270)
-			/* Secondary_pipe XOR Rotation_270 */
-			right_view = (plane_state->rotation == ROTATION_ANGLE_270) != sec_split;
-
-		if (right_view) {
-			data->viewport.x +=  data->viewport.width / 2;
-			data->viewport_c.x +=  data->viewport_c.width / 2;
-			/* Ceil offset pipe */
-			data->viewport.width = (data->viewport.width + 1) / 2;
-			data->viewport_c.width = (data->viewport_c.width + 1) / 2;
-		} else {
-			data->viewport.width /= 2;
-			data->viewport_c.width /= 2;
-		}
+	if (sec_split) {
+		data->viewport.x +=  data->viewport.width / 2;
+		data->viewport_c.x +=  data->viewport_c.width / 2;
+		/* Ceil offset pipe */
+		data->viewport.width = (data->viewport.width + 1) / 2;
+		data->viewport_c.width = (data->viewport_c.width + 1) / 2;
+	} else if (pri_split) {
+		data->viewport.width /= 2;
+		data->viewport_c.width /= 2;
 	}
 
 	if (plane_state->rotation == ROTATION_ANGLE_90 ||
@@ -534,6 +523,11 @@ static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip
 	struct rect surf_src = plane_state->src_rect;
 	struct rect surf_clip = plane_state->clip_rect;
 	int recout_full_x, recout_full_y;
+	bool pri_split = pipe_ctx->bottom_pipe &&
+			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state;
+	bool sec_split = pipe_ctx->top_pipe &&
+			pipe_ctx->top_pipe->plane_state == pipe_ctx->plane_state;
+	bool top_bottom_split = stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM;
 
 	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 ||
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270)
@@ -568,33 +562,43 @@ static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip
 						- pipe_ctx->plane_res.scl_data.recout.y;
 
 	/* Handle h & vsplit */
-	if (pipe_ctx->top_pipe && pipe_ctx->top_pipe->plane_state ==
-		pipe_ctx->plane_state) {
-		if (stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM) {
-			pipe_ctx->plane_res.scl_data.recout.y += pipe_ctx->plane_res.scl_data.recout.height / 2;
-			/* Floor primary pipe, ceil 2ndary pipe */
-			pipe_ctx->plane_res.scl_data.recout.height = (pipe_ctx->plane_res.scl_data.recout.height + 1) / 2;
+	if (sec_split && top_bottom_split) {
+		pipe_ctx->plane_res.scl_data.recout.y +=
+				pipe_ctx->plane_res.scl_data.recout.height / 2;
+		/* Floor primary pipe, ceil 2ndary pipe */
+		pipe_ctx->plane_res.scl_data.recout.height =
+				(pipe_ctx->plane_res.scl_data.recout.height + 1) / 2;
+	} else if (pri_split && top_bottom_split)
+		pipe_ctx->plane_res.scl_data.recout.height /= 2;
+	else if (pri_split || sec_split) {
+		/* HMirror XOR Secondary_pipe XOR Rotation_180 */
+		bool right_view = (sec_split != plane_state->horizontal_mirror) !=
+					(plane_state->rotation == ROTATION_ANGLE_180);
+
+		if (plane_state->rotation == ROTATION_ANGLE_90
+				|| plane_state->rotation == ROTATION_ANGLE_270)
+			/* Secondary_pipe XOR Rotation_270 */
+			right_view = (plane_state->rotation == ROTATION_ANGLE_270) != sec_split;
+
+		if (right_view) {
+			pipe_ctx->plane_res.scl_data.recout.x +=
+					pipe_ctx->plane_res.scl_data.recout.width / 2;
+			/* Ceil offset pipe */
+			pipe_ctx->plane_res.scl_data.recout.width =
+					(pipe_ctx->plane_res.scl_data.recout.width + 1) / 2;
 		} else {
-			pipe_ctx->plane_res.scl_data.recout.x += pipe_ctx->plane_res.scl_data.recout.width / 2;
-			pipe_ctx->plane_res.scl_data.recout.width = (pipe_ctx->plane_res.scl_data.recout.width + 1) / 2;
-		}
-	} else if (pipe_ctx->bottom_pipe &&
-			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state) {
-		if (stream->view_format == VIEW_3D_FORMAT_TOP_AND_BOTTOM)
-			pipe_ctx->plane_res.scl_data.recout.height /= 2;
-		else
 			pipe_ctx->plane_res.scl_data.recout.width /= 2;
+		}
 	}
-
 	/* Unclipped recout offset = stream dst offset + ((surf dst offset - stream surf_src offset)
 	 * 				* 1/ stream scaling ratio) - (surf surf_src offset * 1/ full scl
 	 * 				ratio)
 	 */
-	recout_full_x = stream->dst.x + (plane_state->dst_rect.x -  stream->src.x)
+	recout_full_x = stream->dst.x + (plane_state->dst_rect.x - stream->src.x)
 					* stream->dst.width / stream->src.width -
 			surf_src.x * plane_state->dst_rect.width / surf_src.width
 					* stream->dst.width / stream->src.width;
-	recout_full_y = stream->dst.y + (plane_state->dst_rect.y -  stream->src.y)
+	recout_full_y = stream->dst.y + (plane_state->dst_rect.y - stream->src.y)
 					* stream->dst.height / stream->src.height -
 			surf_src.y * plane_state->dst_rect.height / surf_src.height
 					* stream->dst.height / stream->src.height;
@@ -650,7 +654,20 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	struct rect src = pipe_ctx->plane_state->src_rect;
 	int vpc_div = (data->format == PIXEL_FORMAT_420BPP8
 			|| data->format == PIXEL_FORMAT_420BPP10) ? 2 : 1;
+	bool flip_vert_scan_dir = false, flip_horz_scan_dir = false;
 
+	/*
+	 * Need to calculate the scan direction for viewport to make adjustments
+	 */
+	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_180) {
+		flip_vert_scan_dir = true;
+		flip_horz_scan_dir = true;
+	} else if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90)
+		flip_vert_scan_dir = true;
+	else if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270)
+		flip_horz_scan_dir = true;
+	if (pipe_ctx->plane_state->horizontal_mirror)
+		flip_horz_scan_dir = !flip_horz_scan_dir;
 
 	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 ||
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270) {
@@ -715,7 +732,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	}
 
 	/* Adjust for non-0 viewport offset */
-	if (data->viewport.x) {
+	if (data->viewport.x && !flip_horz_scan_dir) {
 		int int_part;
 
 		data->inits.h = dal_fixed31_32_add(data->inits.h, dal_fixed31_32_mul_int(
@@ -736,7 +753,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 		data->inits.h = dal_fixed31_32_add_int(data->inits.h, int_part);
 	}
 
-	if (data->viewport_c.x) {
+	if (data->viewport_c.x && !flip_horz_scan_dir) {
 		int int_part;
 
 		data->inits.h_c = dal_fixed31_32_add(data->inits.h_c, dal_fixed31_32_mul_int(
@@ -757,7 +774,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 		data->inits.h_c = dal_fixed31_32_add_int(data->inits.h_c, int_part);
 	}
 
-	if (data->viewport.y) {
+	if (data->viewport.y && !flip_vert_scan_dir) {
 		int int_part;
 
 		data->inits.v = dal_fixed31_32_add(data->inits.v, dal_fixed31_32_mul_int(
@@ -778,7 +795,7 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 		data->inits.v = dal_fixed31_32_add_int(data->inits.v, int_part);
 	}
 
-	if (data->viewport_c.y) {
+	if (data->viewport_c.y && !flip_vert_scan_dir) {
 		int int_part;
 
 		data->inits.v_c = dal_fixed31_32_add(data->inits.v_c, dal_fixed31_32_mul_int(